Geotechnical Practice for

Waste Disposal
Geotechnical Practice for
Waste Disposal

Edited by
David E. Daniel
Professor of Civil Engineering
The University of Texas at Austin
USA

SPRINGER-SCIENCE+BUSINESS MEDIA, B.V

First edition 1993
Reprinted 1995, 1996

© 1993 Springer Science+Business Media Dordrecht

Originally published by Chapman & Hall in 1993
Softcover reprint ofthe hardcover Ist edition 1993

ISBN 978-1-4613-6340-8 ISBN 978-1-4615-3070-1 (eBook)

DOI 10.1007/978-1-4615-3070-1

Apart from any fair dealing for the purposes of research or private study, or
criticism or review, as permitted under the UK Copyright Designs and Patents
Act, 1988, this publication may not be reprodueed, stored, or transmitted, in
any form or by any means, without the prior permission in writing of the
publishers, or in the ease of reprographie reproduetion only in accordanee with
the terms of the lieenees issued by the Copyright Lieensing Ageney in the UK,
or in aecordanee with the terms of lieenees issued by the appropriate
Reproduetion Rights Organization outside the UK. Enquiries eonceming
reproduetion outside the terms stated here should be sent to the publishers at the
London address printed on this page.
The publisher makes no representation, express or implied, with regard to the
accuracy of the information eontained in this book and cannot aecept any legal
responsibility or liability for any errors or omissions that may be made.

A Catalogue record for this book is available from the British Library

Library of Congress Cataloging-in-Publieation Data available

Geoteehnieal praetice for waste disposalledited by David E. Daniel.
- 1st ed.
p. em.
Ineludes index and bibliographical references
1. Waste disposal in the ground. 2. Enviromnental geoteehnology.
I. Daniel, David E. (David Edwin), 1949-
TD795.7.G455 1993
628.4'456 - de20 92-40842
CIP

§l Printed on acid-free paper, manufactured in aeeordanee with ANSIINISO

Z39.48-1992 (Permanenee of Paper)
 Contents

Contributors vii
Preface ix

PART ONE GENERAL PRINCIPLES 1

l. In trod uction 3
David E. Daniel
2. Geochemistry 15
Jim V. Rouse and Roman Z. Pyrih
3. Contaminant transport 33
Charles D. Shackelford
4. Hydrogeology 66
Keros Cartwright and Bruce R. Hensel

PART TWO NEW DISPOSAL FACILITIES 95

5. Landfills and impoundments 97
David E. Daniel
6. Leachate and gas generation 113
Robert K. Ham and Morton Barlaz
7. Clay liners 137
David E. Daniel
8. Geomembrane liners 164
Robert M. Koerner
9. Collection and removal systems 187
Robert M. Koerner
10. Water balance for landfills 214
R. Lee Peyton and Paul R. Schroeder
11. Stability of landfills 244
Issa S. Oweis
12. Mine waste disposal 269
Dirk Van Zyl
vi Contents
PART THREE REMEDIATION TECHNOLOGIES 287
13. Strategies for remediation 289
Larry A. Holm
14. Geophysical techniques for subsurface site
characterization 311
Richard C. Benson
15. Soil exploration at contaminated sites 358
Charles O. Riggs
16. Vapor analysis/extraction 379
Lyle R. Silka and David L. Jordan
17. Vertical cutoff walls 430
Jeffrey C. Evans
18. Cover systems 455
David E. Daniel and Robert M. Koerner
19. Recovery well systems 497
Bob Kent and Perry Mann
20. Bioremediation of soils 520
Raymond C. Loehr
21. In situ bioremediation of groundwater 551
Gaylen R. Brubaker
22. Soil washing 585
Paul B. Trost

PART FOUR MONITORING 605

23. Monitoring wells 607
Bob Kent and Mark P. Hemingway
24. Vadose zone monitoring 651
Lome G. Everett

Index 677
 Contributors

Morton Barlaz, Department of Civil Engineering, North Carolina State

University, Raleigh, North Carolina.
Richard C. Benson, Technos Inc., Miami, Florida.
Gaylen R. Brubaker, Remediation Technologies Inc., Chapel Hill,
North Carolina.
Keros Cartwright, Illinois State Geological Survey, Champaign,
Illinois.
David E. Daniel, Department of Civil Engineering, The University of
Texas at Austin, Texas.
Jeffrey c. Evans, Department of Civil Engineering, Bucknell
University, Lewisburg, Pennsylvania.
Lome G. Everett, Geraghty and Miller, Inc., and University of
California at Santa Barbara, Santa Barbara, California.
Robert K. Ham, Department of Civil and Environmental Engineering,
The University of Wisconsin, Madison, Wisconsin.
Mark P. Hemingway, Southwestern Laboratories, Austin, Texas.
Bruce R. Hensel, Illinois State Geological Survey, Champaign, Illinois.
Larry A. Holm, CH2M Hill, Oak Ridge, Tennessee.
David L. Jordan, Hydrosystems Inc., Sterling, Virginia.
Bob Kent, Geomatrix Consultants, Inc., Santa Anna Heights,
California.
Robert M. Koerner, Professor of Civil Engineering, Director,
Geosynthetic Research Institute, Drexel University, Philadelphia,
Pennsylvania.
Raymond C. Loehr, Department of Environmental and Water
Resources Engineering, The University of Texas at Austin, Texas.
viii Contributors

Perry Mann, International Technology Corp., Austin, Texas.

Issa S. Oweis, Converse Consultants East, Parsippany, New Jersey.
R. Lee Peyton, Department of Civil Engineering, University of
Missouri, Columbia, Missouri.
Roman Z. Pyrih, GEOCHEM Div., Terra Vac, Lakewood,
Colorado.
Charles O. Riggs, Sverdrup Environmental, Inc., St Louis, Missouri.
Jim V. Rouse, GEOCHEM Div., Terra Vac, Lakewood, Colorado.
Charles D. Shackelford, Department of Civil Engineering, Colorado
State University, Fort Collins, Colorado.
Paul R. Schroeder, US Army Engineer, Waterways Experiment Station,
Vicksburg, Mississippi.
Lyle R. Silka, Hydrosystems Inc., Sterling, Virginia.
Paul B. Trost, Waste-Tech Services Inc., Golden, Colorado.
Dirk Van Zyl, Golder Associates, Inc., Lakewood, Colorado.
 Preface

Earth scientists and geotechnical engineers are increasingly challenged

to solve environmental problems related to waste disposal facilities and
cleanup of contaminated sites. The effort has given rise to a new
discipline of specialists in the field of environmental geotechnology. To
be effective, environmental geotechnologists must not only be armed
with the traditional knowledge of fields such as geology and civil
engineering, but also be knowledgeable of principles of hydrogeology,
chemistry, and biological processes. In addition, the environmental
geotechnologist must be completely up to date on the often complex
cadre of local and national regulations, must comprehend the often
complex legal issues and sometimes mind-boggling financial impli-
cations of a project, and must be able to communicate effectively with a
host of other technical specialists, regulatory officials, attorneys, local
land owners, journalists, and others. The field of environmental geo-
technology will no doubt continue to offer unique challenges.
The purpose of this book is to summarize the current state of practice
in the field of environmental geotechnology. Part One covers broadly
applicable principles such as hydrogeology, geochemistry, and con-
taminant transport in soil and rock. Part Two describes in detail the
underlying principles for design and construction of new waste
disposal facilities. Part Three covers techniques for site remediation.
Finally, Part Four addresses the methodologies for monitoring.
The topics of 'waste disposal' and 'site remediation' are extra-
ordinarily broad. This book is written for geologists, hydrogeologists,
geotechnical engineers, environmental engineers, soil scientists, and
others with similar backgrounds. The book does not attempt to go
beyond the immediate capabilities of the environmental geotech-
nologist. For example, the book does not cover incineration of waste
because the environmental geotechnologist would not logically design
an incinerator; a combustion specialist would.
In preparing this book, the editor has assembled a group of ex-
ceptionally knowledgeable contributors to whom the editor is most
grateful. The editor thanks the numerous scientists and engineers
who have shaped his thinking over the years and expresses particular
x Preface

appreciation to his students for innumerable thoughtful questions and

many hours of sacrifice and hard work. Finally, many thanks to Cindy
Symington for her help in preparing this book.
David E. Daniel
 PART ONE

General Principles
 CHAPTER 1

Introduction
David E. Daniel

1.1 EARLY DAYS OF ENVIRONMENTAL

GEOTECHNOLOGY

In the 1940s to 1960s, the industrialized countries of the world under-

went an enormous expansion in capacity for manufacturing goods,
processing petroleum, and making new chemicals. Geotechnical
engineers and earth scientists played an important role in that ex-
pansion by identifying mineral and petroleum resources, investigating
subsurface stratigraphy and soil conditions, designing foundations for
buildings and machinery, and developing earthwork specifications.
The early roots of the field of geotechnical engineering can be traced
to the 1800s. Legendary contributions were made by Coulomb,
Rankine, Terzaghi, and many others over the past two centuries.
Although the International Society for Soil Mechanics and Foundations
Engineering held its first international conference in 1936, the modern
practice of soil mechanics and foundation engineering blossomed
throughout the world after World War II and underwent particularly
intensive growth and maturity in the 1950s and 1960s. Most of the
modern journals in geotechnical engineering, such as Geotechnique,
the Canadian Geotechnical Journal, and the American Society of
Engineer's (ASCE's) Journal of the Soil Mechanics and Foundation Division
(now Journal of Geotechnical Engineering) published their first issues in
the 1950s.
Despite the advances made in the field that we now call geotechnical
engineering in the 1950s and 1960s, there was practically no emphasis
on environmental matters in geotechnical engineering practice during
this period. The geotechnical engineer's role in environmental
decisions, however, was relatively minor until the late 1970s. The

Geotechnical Practice for Waste Disposal.

Edited by David E. Daniel.
Published in 1993 by Chapman & Hall, London. ISBN 0 412 35170 6
4 Introduction

author recalls working as a junior engineer at the San Francisco firm of

Woodward-Clyde Consultants from the period 1974-1977. In the
beginning of this experience, geotechnical engineers generally shunned
projects that involved waste materials or contamination. The author
recalls that the firm had one engineer who was particularly
experienced in evaluation of 'garbage dumps'. This individual's special
capability in this area was not viewed as particularly valuable by his
fellow engineers: there were far more glamorous projects than
evaluation of garbage dumps and preventing environmental damage
from garbage dumps. By the end of the 1970s, though, as will be seen
in the next subsection, the situation began to change markedly and
irrevocably.
Perhaps a typical example of the pre-1970s situation is the following
project from the author's project files. In the late 1950s, a petroleum
refinery in Texas underwent a major expansion, which required new
structures as well as a 600 ha 'evaporation pond' for waste water.
Waste water containing benzene (a carcinogen) was to be pumped into
the pond and allowed to evaporate. At the time, evaporation was
considered a viable 'disposal' method; today in the US, evaporation is
not considered a suitable technique for treatment or disposal of
hazardous waste.
A well known and very capable geotechnical engineering company
was hired to develop geotechnical recommendations for the plant
expansion. Approximately 20 exploratory borings were drilled to a
depth of approximately 15 m. Various laboratory tests (mostly strength
and compressibility tests to support foundation design recom-
mendations) were performed, and several hydraulic conductivity tests
were conducted on subsoils from the area where the evaporation pond
was proposed. The hydraulic conductivity tests indicated hydraulic
conductivities less than 1 X 1O~8 cm/s.
The evaporation pond was discussed in the final report that the
geotechnical engineer prepared, but only one paragraph was devoted
to the pond and to potential groundwater contamination issues. The
subsoils were described as 'impermeable' in the report. Implicitly,
groundwater contamination was not considered to be a relevant issue.
Construction began a short time after the geotechnical report was
issued, and the unlined evaporation pond was put into service in about
1960.
In 1980, a regulatory agency directed the company to install a
groundwater monitoring system. An enormous pool of contaminated
liquid was found beneath the pond (Fig. 1.1). Fortunately, the upper-
most aquifer, which was located nearly 100 m beneath the ground
surface, had not yet been impacted, thanks to a low-hydraulic-
conductivity stratum located about 75 m below the surface (Fig. 1.1).
Nevertheless, the pond was taken out of service and remediation was
 Early days of environmental geotechnology 5
Benzene-Rich Wastewater

100 m

Low-Hydraulic-Conductivity Layer

Fig. 1.1 Mound of contaminated liquid that formed over a period of 20+ years
from unanticipated seepage of waste water into soil beneath evaporation pond.

initiated by installing approximately 25 recovery wells. Remediation

will probably take decades.
Where did the original geotechnical specialists err? Subsequent
studies showed that the actual hydraulic conductivity of the subsoils
located 0-15 m below the surface was not the value of < 1 X 10- 8 cm/s
reported by the geotechnical engineer but instead was approximately 1
X 10- 4 cm/s. The soils beneath the evaporation pond were far from
impermeable. The problem was that the original testing program was
restricted to laboratory hydraulic conductivity tests on small samples of
soil. The soils in the field contained numerous cracks (probably caused
by cyclic wetting and drying) that were not included in the small
laboratory test specimens. Today's specialist in environmental geo-
technology knows that small-scale laboratory hydraulic conductivity
tests do not always give representative in situ values, but the engineer
of the 1950s had little clue that laboratory tests could produce mis-
leading results.
The problem just described is not untypical. In the 1950s and 1960s,
there was little understanding of the long-term consequences of waste
disposal to the land. Clayey soils were thought to be 'impermeable'.
The mere existence of such materials was thought to be enough to
ensure protection of the environment. There was practically no under-
6 Introduction

standing by geotechnical engineers of the relationship between disposal

of chemicals in the ground and long-term ground water impacts. Many
of the realities that we now know to exist were simply not recognized
as problems before the 1970s. Even today, the uneducated engineer or
scientist can easily repeat the common mistakes of the past if he or she
does not possess the appropriate knowledge and understanding of
recent experience and discoveries.

1.2 ORIGINS OF ENVIRONMENTAL GEOTECHNICS

The field of practice now called 'environmental geotechnics' evolved

over a period of about two decades starting in the 1970s. In the US, the
birth of environmental geotechnology can be traced largely to the
nuclear power industry. In the 1970s, an enormous amount of geo-
technical engineering work was focused on design of nuclear power
plants. One part of the process of constructing a nuclear power plant in
the US was development of an 'environmental impact statement'. No
nuclear power facility could be granted a construction permit until an
environmental impact statement was completed. Because geotechnical
engineers were the lead technical specialists in most siting inves-
tigations, it was natural that geotechnical engineers manage and
coordinate preparation of environmental impact statements. Many geo-
technical engineering firms developed the capability for preparing such
statements and hired environmental specialists, e.g., biologists and
botanists, to conduct the technical work. For many geotechnical
engineers, this was their first exposure to concern for the environment
in their technical work and their first opportunity to work with diverse
environmental specialists. These early experiences were critical in the
development of an understanding among geotechnical engineers that
geotechnical engineering principles, along with other scientific and
engineering specialties, could be brought to bear in solving important
technical problems involving the environment.
Concerns about the safety of nuclear power plants have always been
voiced, but in the mid to late 1970s, one of the most frequently ex-
pressed concerns was over the ultimate disposal of high-level radio-
active waste. Where would such wastes go? The wastes remain
dangerous for hundreds of thousands of years; how can they be safely
contained? To address these questions, numerous investigations were
performed, primarily in the US and Europe, to study the proper
disposal of high-level radioactive wastes. Geotechnical engineers played
an important role in these early studies through investigation and
characterization of suitable host rocks for waste repositories, analysis of
long-term performance of the earth materials under realistic tempera-
tures and pressures, evaluation of probable ground water impacts, and
 Origins of environmental geotechnics 7

assessment of potential risks. For many large geotechnical engineering

companies, these types of investigations represented the first signifi-
cant attempt to perform a comprehensive technical analysis of a waste
containment facility.
In the US, the next significant event in the development of environ-
mental geotechnology was the widely publicized contamination at a
site in New York called Love Canal. At this site, dangerous chemical
wastes had been buried in an old canal and covered with clayey soils.
The chemicals slowly seeped out of the canal over a period of many
years until, in the mid 1970s, suspected health impacts were identified.
The area was eventually evacuated, and unprecedented publicity was
given to the dangers posed to the residents of the area from the wastes
and to the enormous health impacts from hazardous waste in general.
For the first time, national attention was focused on the adverse
impacts of improper disposal and management of chemical wastes.
Also, in the late 1970s and early 1980s, significant discussion began to
occur concerning appropriate design standards for waste containment
facilities. Geotechnical engineers at this time began to be brought into
the process of designing waste containment facilities.
The 1980s brought an explosive expansion of environmental geo-
technology, particularly in the US, where two significant processes
were at work. First, stringent regulations drove waste containment
systems to unprecedented levels of sophistication and complexity.
Double liner systems, leak detection layers, geomembrane liners, and a
host of new concepts were not only discussed but were written into
regulations. Regulatory agencies, rather than engineers, drove the
evolution of increasingly sophisticated containment designs. Environ-
mental geotechnologists were called upon to develop and implement
the required designs and containment systems. The regulatory agencies
and environmental geotechnologists literally pushed containment
technology to unprecedented levels of sophistication.
The second process stemmed from promulgation by the US Congress
in 1980 of the Comprehensive Environmental Response, Compensa-
tion, and Liability Act (CERCLA), which became commonly referred to
as 'Superfund'. Superfund charted a course in the US for cleanup of
contaminated sites, such as the Love Canal site, and provided monies
for cleanup if the responsible parties could not be identified or were
incapable of paying for the cleanup. The requirement that contaminated
sites be cleaned up was the single most important boost to environ-
mental geotechnology in its relatively short history. Tens of thousands
of engineers are actively engaged in investigation, design, and actual
cleanup of contaminated sites in the US, and similar priorities are being
established in many other countries, as well.
Because of the complexity of environmental contamination, environ-
mental geotechnologists are as diverse as the problems they tackle.
8 Introduction
Table 1.1 Rates of generation of chemical wastes in selected countries (from
van Veen and Mensink, 1985)

country population waste production per capita

(X10 6 ) (x10 6 tonnes) waste production
(kg/person)

Switzerland 6.4 0.1 16

Denmark 5.1 0.1 20
Canada 23.9 1 42
Sweden 8.3 0.5 60
Netherlands 14.1 1 71
Italy 57 5 88
United Kingdom 56 5 89
France 53.7 5 93
West Germany 61.6 6 97
Belgium 9.9 1 101
United States 227.7 40 176

Environmental geotechnologists include civil engineers (especially those

schooled in soil and rock mechanics, geotechnical engineering, sanitary
engineering, and environmental engineering), soil scientists, geologists,
hydrogeologists, geohydrologists, engineering geologists, geological
engineers, mining engineers, and agricultural engineers. This book is
written from the point of view of the civil engineer who must not only
examine and define the scope of environmental problems but, most
importantly, solve those problems in practical, reliable, cost-effective
ways. The challenge is immense.
Those who assume that the field of environmental geotechnology is
restricted to one or two countries could not be more incorrect. All of
the industrialized countries of the world generate significant quantities
of waste, as the data in Table 1.1 indicate. Those countries with rela-
tively recently-developed industrialized bases frequently have less
mature and sophisticated environmental controls. Many of the past
mistakes made in the industrialized countries of the world can be
avoided if engineers in industrially-emerging nations will take the time
to learn from those mistakes and apply the technologies (especially the
most cost effective ones) that have been found to work in the highly
developed countries. The field of environmental geotechnology is truly
a field of world-wide importance.

1.3 REGULATORY REQUIREMENTS

No book on waste disposal would be complete without some dis-

cussion of regulatory requirements. However, the focus of this book is
 Regulatory requirements 9
technology, not regulation. Thus, the following brief discussion will
serve to summarize the regulatory framework in one country (the US),
where the regulatory climate is the most complex. This discussion will
be helpful in following some of the material in later chapters. As stated
earlier, waste containment technology has been driven by regulation,
not engineering. One simply cannot function as an environmental
geotechnologist without some appreciation for how regulations drive
technology.
In the US, two key pieces of legislation, and amendments to the two
legislative benchmarks, drive the environmentalist's technical approach
to problems. Disposal of newly-generated solid waste is regulated
under the Resource Conservation and Recovery Act (RCRA) and
the Hazardous and Solid Waste Amendments (HSWA) to RCRA.
Hazardous waste is covered in Subtitle C of RCRA, and non-hazardous
waste is regulated in Subtitle D.
Cleanup of old waste disposal sites is governed by legislation men-
tioned earlier, viz., Comprehensive Environmental Response, Com-
pensation, and Liability Act (CERCLA), which is also known as
Superfund. The original CERCLA legislation has expired and has been
superseded by the Superfund Amendments and Reauthorization Act
(SARA).
The spirit and details of RCRA and SARA are quite different. In
RCRA, stringent, explicit requirements are set forth for nearly all
aspects of hazardous and nonhazardous waste management, including
treatment, storage, and disposal of waste. With cleanup (Superfund),
the problem is different because each site is unique and must be
evaluated individually. Thus, SARA is far more flexible than RCRA.
However, RCRA requirements may have to be considered in imple-
mentation of cleanup under Superfund under the principle of ARARs,
which stands for Applicable or Relevant and Appropriate Regulations.
If RCRA applies to the waste (e.g., if the waste was disposed of after
RCRA went into effect), RCRA requirements are applicable and must
generally be followed. Even if RCRA does not apply to the waste (e.g.,
the waste was disposed of before RCRA took effect), RCRA require-
ments may still be relevant and appropriate, and if so, typically must at
least be considered.
Hazardous wastes are widely perceived to be incredibly dangerous
waste materials that require the highest degree of care through
regulation. However, a fundamental problem exists in defining what
constitutes a hazardous waste. If one assumes that a hazardous waste
is any waste that is potentially dangerous to humans, practically all
wastes would be considered hazardous because practically everything
is potentially harmful to humans. Ordinary table salt (NaCl), for
example, can be lethal if ingested in excessive doses. A health-based
definition of hazardous waste would be logical but awkward to apply
10 Introduction

due to the many complexities involved. Ultimately, any definition that

is selected for hazardous waste must be easy to understand and must
rely upon reproducible, widely-accepted test methods.
Under RCRA, a waste must first be a solid waste to be considered a
hazardous waste. Hazardous wastes are, by statutory specification, a
subset of solid wastes. A solid waste, by definition, is any material,
other than those which are specifically excluded, that is:
1. disposed of or abandoned in lieu of disposal;
2. burned, incinerated, or recycled; or
3. considered inherently waste-like.
The definition of solid waste was evidently intentionally made very
broad and inclusive.
A solid waste, unless specifically excluded, is considered a hazardous
waste if anyone of the following four criteria are met.
1. The waste is specifically listed as a hazardous waste (RCRA provides
an extensive list of waste constituents that are by definition
hazardous - these wastes are called listed wastes):
2. The waste is a mixture containing hazardous waste (i.e., any waste
mixed with a hazardous waste is also a hazardous waste).
3. The waste is derived from the treatment, storage, or disposal of
hazardous waste, i.e., once a hazardous waste, always a hazardous
waste (but see exemptions discussed later).
4. The waste exhibits anyone of four characteristics of hazardous
waste:
(a) ignitability (the waste presents a fire hazard during routine
management and either has a flash point <60°C or is not
a liquid and is capable of causing fire, e.g., by spontaneous
chemical changes, that creates a hazard);
(b) corrosivity (pH ~ 2 or pH ~ 12.5, or corrodes steel at a rate
greater than 6 mm/yr at 58°C);
(c) reactivity (unstable waste that undergoes violent reaction); or
(d) toxicity (based on the Toxicity Characteristic Leaching Pro-
cedure, or TCLP, test).
The TCLP test is a revision of the US Environmental Protection
Agency's (EPA's) earlier test procedure known as the Extraction Pro-
cedure (EP, or EP toxicity test). In the original EP toxicity test, the
pulverized solid waste was mixed with a leachant at a liquid: solids
ratio of 20: 1. The liquid and solid were mixed in a sealed container,
and then the solid and liquid phases were separated by centrifugation
and filtration. The liquid was analyzed for 14 compounds (8 metals,
4 insecticides, and 2 herbicides). Concentrations were compared to
published values; if the published values were exceeded, the waste
was defined as hazardous. If not, the waste was not considered a
 Regulatory requirements 11

Table 1.2 Toxicity characteristic constituents and regulatory levels

TCLP constituentslregulatory levels old EP constituentslregulatory levels

---~-----

Benzene O.Smg/L Arsenic S.Omg/L

Carbon tetrachloride O.Smg/L Barium 100.Omg/L
Chlordane 0.03mg/L Cadmium 1.0mg/L
Chlorobenzene 100.Omg/L Chromium S.Omg/L
Chloroform 6.0mg/L Lead S.Omg/L
m-Cresol 200.0mg/L* Mercury 0.2mg/L
o-Cresol 200.0mg/L Selenium 1.0 mg/L
p-Cresol 200.0mg/L Silver S.OmglL
1,4-Dichlorobenzene 7.Smg/L Endrin 0.02mg/L
1,2-Dichloroethane O.Smg/L Lindane O.4mg/L
1,1-Dichloroethylene 0.7mg/L Methoxychlor 1O.0mgIL
2,4-Dinitrotoluene O.13mg/L Toxaphene O.Smg/L
Heptachlor (and its 2,4 D 1O.Omg/L
hydroxide) O.OOSmg/L 2,4, ST-Silvex 1.0mg/L
Hexachloro-1,3-
butadiene O.Smg/L
Hexachlorobenzene O.13mg/L
Hexachloroethane 3.0mg/Lt
Methyl ethyl ketone 200.0mg/L
Nitrobenzene 2.0mg/L
Pentachlorophenol 100.Omg/L
Pyridine S.Omg/Lt
Tetrachloroethylene 0.7mg/L
Trichloroethylene O.Smg/L
2,4,S-Trichlorophenol 400.0mg/L
2,4,6-Trichlorophenol 2.0mg/L
Vinyl Chloride 0.2mg/L

*If 0-, m-, and p-Cresol concentrations cannot be differentiated, the total cresol
concentration is used. The regulatory level for total cresol is 200.0mg/L.
t Quantitation limit is greater than the calculated regulatory level. The quantitation limit,
therefore, becomes the regulatory level.

characteristic hazardous waste and would have to meet one of the

other criteria to be considered a hazardous waste.
In 1984, EPA was directed to reconsider the EP toxicity test. In 1986,
EPA proposed to revise and expand the EP text. After much delibera-
tion, collaboration, and revision, EPA issued the final procedure for
the revised test (i.e., the TCLP) in 1989. The TCLP procedure varies
slightly depending on the nature of possible contamination and the
form of the waste. Basically, though, the procedure still involves
mixing the solid waste with a leachant at a liquid: solid ratio of 20: 1.
The compounds analyzed in the TCLP test, and trigger values for
definition as a hazardous waste, are listed in Table 1.2.
A number of waste types are specifically excluded from categoriza-
tion as a hazardous waste. The two most notable exclusions are house-
hold refuse and small-quantity hazardous waste.
12 Introduction

In the US, it is illegal to dispose of untreated hazardous waste. All

hazardous waste must first be treated with best demonstrated available
technology (BOAT) prior to disposal. For this reason, the amount of
hazardous waste disposed of in landfills has declined while the amount
destroyed, e.g., via incineration, has increased markedly.
In the US, the volume of household waste (often termed municipal
solid waste, or MSW) that must be landfilled has remained relatively
constant or increased slightly. Increased recycling efforts have served
more to slow the rate of growth of MSW that must be landfilled rather
than to reduce the amount going to landfills. Landfills are, and will
continue to be for the foreseeable future, the primary means for
disposal of MSW. The need for landfills for all categories of waste will
remain because society can never fully rid itself of all solid residual
material.

1.4 WASTE CHARACTERISTICS

Hazardous waste cover a broad spectrum of materials, particularly

when one considers newly generated, treated wastes and old, untreated
wastes buried in the ground some years in the past. Figure 1.2 shows a
distribution of industrial waste constituents in the mid 1980s in the US.
Many wastes are mixtures of materials. Any attempt to characterize
such materials would be pointless; virtually all waste forms can be
found. The wastes range from strongly acidic to neutral to strongly

Other

Rubber and Plastics Products

Fabricated Metal Products

Chemicals and
Allied Products
Petroleum and Coal
Products

Fig. 1.2 Distribution of industrial waste in the US in the mid 1980s.

 Waste characteristics 13
alkaline. Some wastes are rich in metals, some are rich in organics, and
some are mixtures that contain both metals and organics.
One special type of waste deserves mention at this point to introduce
terminology. Organic liquids may be divided into two groups: water
soluble liquids (e.g., ethyl alcohol) and non-water soluble liquids (e.g.,
trichloroethylene). Actually, virtually all organic liquids are soluble to
some extent in water, but with many organic liquids the solubilities are
measured in parts per million or parts per billion of organic constituent
that can be dissolved in water. Non-water-soluble organic liquids are
called non-aqueous-phase liquids (NAPLs). There are two types of
non-aqueous-phase organic liquids; dense non-aqueous phase liquids
(DNAPLs) are heavier than water and light non-aqueous phase liquids
(LNAPLs) are lighter than water. One would expect to find a LNAPL
(e.g., gasoline) floating on the water table and a DNAPL (such as the
chlorinated solvent perchloroethylene) perhaps below the water table
at the interface between an aquifer and a lower-hydraulic-conductivity
formation. Cleanup of LNAPLs and especially DNAPLs poses unique
challenges, as discussed later in this book.
Municipal solid waste (MSW) is somewhat more consistent than
industrial waste. The distribution of MSW produced in the US is
shown in Fig. 1.3 (by weight) and Fig. 1.4 (by volume). As recycling
becomes more prevalent, the distribution can be expected to change.
An important feature of MSW is that the waste decomposes and

Other

Food Wastes

Paper
Plastics

Glass

Metals

Yard Wastes

Fig. 1.3 Distribution of municipal solid waste generated in the US by weight

(USEPA, 1990).
14 iJ1troducfirm

Food Wastes

Plastics

Metals

Fig. 1.4 Distribution of municipal solid waste 10 US landfills by volume

(USEPA, 1990).

produces gas, including methane, that can be very dangerous if

improperly controlled.
The liquid that is derived from waste is called leachate. Leachate can
be produced directly from buried liquid wastes or consolidation of
fluid-bearing wastes, by decomposition or chemical reactions, or by the
leaching action of water moving through the waste. The control of
leachate and gas is the single most important design requirement for
new waste disposal facilities.

REFERENCES

USEPA (1990) Characterization of municipal solid waste in the United States: 1990
update, Office of solid waste and emergency response, Washington, DC,
EP A/530-SW -90-042.
van Veen, F. and J.A. Mensink (1985) Brief survey of legislation and arrange-
ments for the disposal of chemical waste in a number of industrialized
countries, Haz. Waste & Haz. Mat., 2(3), 333-53.
 CHAPTER 2

Geochemistry
Jim V. Rouse and Roman Z. Pyrih

2.1 INTRODUCTION

Not so many years ago when the subject of geochemistry was men-
tioned, topics such as crustal abundance of the various elements,
radiochemical age dating, or computer modeling of stable chemical
states immediately came to mind. Most engineers in the geotechnical
professions equated geochemistry with analytical services. However,
there is a rapidly emerging side to the discipline of geochemistry which
relates directly to geotechnical practices in waste disposal.
This relationship between the two disciplines is based on the prin-
ciple that an interdisciplinary approach to problem solving has the best
chance of success. This is especially true when dealing with waste
disposal. It is essential for the geotechnical professional to be able to
communicate with the geochemist, to ask the right questions, and to
understand the significance of geochemical data. At the same time,
it is the obligation of the geochemist to provide data in usable form,
to voice alerts and anticipate problems, and, most importantly, to
integrate the principles of geochemistry into the engineering and
design of waste disposal facilities, soil cleanup strategies, and ground-
water restoration programs. When integrated with sound engineering
design, geochemistry can provide an additional level of environmental
safeguards and control.
The purpose of this chapter is to introduce the geotechnical pro-
fessional to two very useful concepts of geochemistry: the concept of
attenuation and the idea of phased ground-water monitoring. When
integrated into the design of a waste-disposal facility, attenuation in-
troduces one additional environmental safeguard against the migration
of potential ground-water contaminants. Likewise, taking a phased

Geotechnical Practice for Waste Disposal.

Edited by David E. Daniel.
Published in 1993 by Chapman & Hall, London. ISBN 0412351706
16 Geochemistry
approach to the monitoring of chemical constituents during and after
the operational life of the facility is not only cost-effective, but is
more environmentally protective than the approach of collecting and
analyzing for 'everything'.
The discussion of the basic geochemical principles involved in these
concepts is not intensive; it is provided so that the geotechnical
engineer can relate the concepts to everyday problems. Geochemical
literature cited throughout the chapter will provide additional insight
into various aspects of environmental geochemistry, for those who
wish to pursue the subject further. The examples of geochemical pro-
cesses which are referenced in the chapter focus on geochemical and
biogeochemical interactions between ground water and the natural
geological material through which it flows. As such, the examples
illustrate the effect of geochemical and biogeochemical interactions on
chemical constituents present in seepage solutions and ground water.

2.2 GEOCHEMICAL ATTENUATION

Natural geological material has the ability to interact geochemically

with chemical constituents of ground water. The results of this inter-
action can lead to the partial or total immobilization of potential
ground-water contaminants. This process of immobilizing and retarding
the chemical constituents from moving with ground-water flow is
called attenuation.
The ability of natural geological material such as soil to immobilize
potential ground-water contaminants and attenuate their movement is
well known. Under the right conditions, soils and sediments can act as
geochemical 'traps' and prevent the movement of various chemical
constituents. Early in the development of geotechnical practices in
waste disposal, Griffin et al. (1976, 1977) described and quantified the
ability of clay-bearing material to remove heavy metals and other trace
elements from seepage solutions resulting from landfill leachate. The
authors of this early effort made the innovative suggestion that overall
pollution from landfills could be reduced if liners of natural earth
material were designed for higher hydraulic conductivities. Higher
hydraulic conductivities would allow for contaminant attenuation and
avoid the development of 'bathtub' conditions. Unfortunately, recent
regulatory reliance on specific design criteria runs counter to this often
reasonable and technically-sound approach.
As described by Rouse and Pyrih (1985), a number of naturally
occurring geochemical processes are effective in removing ground-
water contaminants. These mechanisms can include cation and anion
exchange with clays, adsorption of cations and anions on hydrous
oxides of iron and manganese, sorption on organic maUer, direct
 Geochemical attenuation 17
precipitation of ions from solution, and co-precipitation by adsorption.
In addition, volatilization and biodegradation can be effective mech-
anisms for attenuating the movement of organic contaminants (Dragun,
1988).
Most geological material in contact with ground water contains some
percentage of clay minerals. Such geological material usually consists
of complex mixtures of clay minerals, iron and manganese hydrous
oxides, and organic matter. Under the proper conditions, the clay
minerals, the hydrous oxides and the organic matter impart to natural
earth material the ability to scavenge and to concentrate cations and
anions from seepage solutions or migrating ground water (Dragun,
1988; Hornick, 1976).
A mechanism with the potential to affect significant attenuation is
cation and anion exchange between clay minerals and ions in solution.
Ion-exchange or ion-replacement reactions can occur to some extent in
all clay minerals (Grim, 1968). Analogous reactions resulting in ion-
exchange from solution can occur not only with clay minerals but also
at the surface of iron and manganese hydrous-oxides (Jenne, 1968) and
organic matter (Schmidt-Collerur, 1978) that can be found associated
with natural geological materials.
Other mechanisms can be activated as seepage solution or ground
water contacts geological materials. Precipitation and co precipitation
can remove chemical constituents from solution as insoluble pre-
cipitates. Often, precipitation reactions are initiated by changes in
chemical parameters such as pH and Eh. As a general rule, neutral-
ization of pH optimizes conditions for geochemical removal of many
constituents from solutions. Geochemical attenuation mechanisms are
most active in a pH range between 5 and 8.
The extent to which geological materials will function as a geo-
chemical trap and attenuate the movement of potential ground-water
contaminants will depend upon:
1. the chemical composition of the seepage solution or ground water;
2. the geochemical and mineralogical properties of the geological
material; and
3. the pH and Eh conditions that are established during contact of the
water with the geological material.

2.2.1 Examples of geochemical attenuation

The environmental fate and geochemical behavior of chromium and
selenium have been under intensive investigation (Rouse and Pyrih
1989; Rouse, 1988; Dragun, 1988; Hornick, 1976). Griffin et al. (1976)
reported that clay-bearing material could remove trace elements
including chromium and selenium from seepage solutions resulting
18 Geochemistry

from landfill leachate. Above values of pH 6, trivalent chromium was

immobilized as a result of precipitation. Below pH 4, the trivalent
species was adsorbed by both kaolinite and montmorillonite clays.
Between pH 4 and 6, the combination of adsorption and precipitation
mechanisms rendered trivalent chromium immobile. The removal of
selenium (as the selenite ion) from leachate solution by kaolinite and
montmorillonite clays reached a maximum in the pH range of 2 to 3,
but quickly decreased as the pH increased. The adsorption of selenium
became negligible as the pH approached 10. Fritz and Hall (1988)
established that selenium adsorption occurs not only on clays but on
hydrous oxides and organic matter. Maximum adsorption of selenite
occurred between pH 4 and 6; above pH 8.5 a significant decrease in
selenite adsorption was observed.
Experiences with the subsurface migration of arsenic have shown
that this potential ground-water contaminant can be and is attenuated
by geochemical mechanisms (Pyrih and Rouse, 1989). Under slightly
acidic pH conditions, arsenic will readily anion-exchange with clay
minerals such as illite and smectite. In addition, arpenic is sorbed by
hydrous oxides of iron and manganese and precipitated as an insoluble
arsenate by metals such as iron, copper, or zinc. Iron/arsenic and
iron/chromium reactions are the basis for newly-evolving process
technology designed to remove traces of arsenic and chromium from
natural water systems (Peck, 1990).
The environmental fate of free cyanide and of metal-cyanide com-
plexes has also been under intensive investigation. Recent laboratory
studies and field investigations of cyanide attenuation by natural geo-
logical material indicate that free cyanide and metal-cyanide complexes
move only short distances through soil before being immobilized. This
attenuation of cyanide is due to geochemical and biogeochemical pro-
cesses that naturally occur in the subsurface. Several geochemical
mechanisms can account for the removal of cyanide from solutions.
These mechanisms include sorption on mineral surfaces or organic
detritus; precipitation of insoluble metal-cyanide complexes typically
found with iron; chemical conversion to thiocyanate; and chemical
oxidation to cyanate. Furthermore, cyanide can be altered by biological
mechanisms and metabolized by plants. Under aerobic conditions
occurring in near-surface soils and ground water, cyanide will be
decomposed to ammonia, nitrogen or even nitrate, and to carbon
dioxide. Under anaerobic conditions, cyanide will be decomposed to
ammonium ion, nitrogen, thiocyanate, and carbon dioxide. (University
of California at Berkeley, 1988; Smith, 1988; Rouse and Pyrih, 1988;
Chatwin and Trepanowski, 1987; Simovic et al., 1984; Schmidt et al.,
1981; Knowles, 1976).
Volatilization of free cyanide is an important physicochemical mech-
anism that can limit the mobility of free cyanide and weak metal-
 Geochemical attenuation 19

cyanide complexes in the near-surface environment. Free cyanide

volatilization becomes quite significant when the pH in soil or ground
water is reduced below 9.4. This mechanism is especially important
since most soils are capable of neutralizing alkaline seepage and
lowering the solution pH. Photodecomposition of metal-cyanide com-
plexes accelerates the volatilization process by breaking down metal-
cyanide complexes in the presence of sunlight. The extent of sunlight
penetration will determine the degree of this photodecomposition.
(Smith, 1988; Huiatt et al., 1983; Schmidt et al., 1981).
Moderately strong and strong metal-cyanide complexes such as
ferro-and ferri-cyanide complexes are exceptionally stable and have a
geochemistry that is totally different from the free forms of cyanide,
such as molecular hydrogen cyanide and simple ionic cyanide. Such
iron-cyanide complexes are better able to interact geochemically and
biogeochemically with soil and aquifer materials. These interactions
lead to immobilization of the cyanide complexes and eventual decom-
position by microbial activity (Smith and Struhsacker, 1988; Scott, 1984;
Huiatt et al., 1983).

2.2.2 Quantifying attenuation capacities

The effectiveness and capacity of natural geological material to attenuate
the movement of chemical constituents that may be present in seepage
solutions or ground water can be quantified in the laboratory. To
quantify attenuation capacities in a geological material, a phased geo-
chemical program of testing and analyses should be designed that
includes the following steps:
1. analysis of the earth material for mineralogical and geochemical
properties;
2. conduct of sequential batch-contact measurements to quantity
attenuation capacities;
3. conduct of long-term column experiments to confirm attenuation
capacity; and
4. field demonstration.

Geochemical properties
As an initial effort to quantify the geochemical attenuation capacity of
geological material, representative samples should be analyzed by x-
ray diffraction to determine bulk mineralogy and the mineralogy of the
clay-size fractions. In addition, samples should be analyzed for a list of
geochemical parameters which experience has shown can affect the
degree of attenuation taking place in natural earth material. This list of
geochemical parameters includes at least the following:
20 Geochemistry

• clay content
• cation-exchange capacity and exchangeable cations
• iron and manganese hydrous-oxide content
• organic carbon content
• soil pH
• acid and base neutralizing potential
A good source of information on procedures for measuring the
parameters listed above is Klute (1986).

Laboratory evaluation
After data are developed on the geochemical character of site soils and
the degree of variability of the geological material, the next task is to
evaluate the material's attenuation capacity. The geochemical reactions
which are likely to occur when seepage solutions percolate through
soils or ground water flows through geological material can be simulated
in the laboratory. Two procedures have been shown to be useful in
demonstrating these reactions and evaluating the effectiveness of the
attenuation mechanisms, namely:
• sequential batch-contact tests
• column percolation tests
Each procedure has certain advantages and disadvantages. Sequential
batch-contact tests are useful in evaluating attenuation properties on a
small sample of geological material and obtaining estimates of attenua-
tion capacity. The testing can be completed relatively quickly and is
ideally suited for determining attenuation characteristics on a number
of samples at one time. However, the results from column testing
generally describe attenuation processes more truly. Unfortunately,
column tests may require several months to generate useful data for
materials with low hydraulic conductivity.

2.2.3 Laboratory methodology and procedures

Sequential batch-contact testing

Sequential batch-contact testing is often selected as the laboratory tech-
nique to obtain preliminary indications of attenuation capacities. The
sequential batch procedure simulates seepage solution percolating
through columns of soil material wherein successive solution comes in
contact continuously with fresh material. This testing procedure can
simulate years of potential field seepage or ground-water flow in a
few days of laboratory testing. The methodology and procedures for
sequential batch-contact testwork are described in detail by Houle and
Long (1980).
 Geochemical attenuation 21
Sequential batch-contact testing (SBT) of a soil's attenuation property
proceeds by placing weighed amounts of a bulk sample in poly-
ethylene containers and adding measured volumes of representative
seepage solution or ground water containing the potential contaminant.
The proportions of solution to bulk sample are controlled in order to
achieve the desired liquid to solid ratios. The resulting slurries are
gently agitated over a period of 24 hours. At the conclusion of each 24-
hour period, the slurry is decanted and filtered. The moist solids and
the appropriate volume of leachate solution from the previous test are
advanced in sequence according to the matrix illustrated in Fig. 2.1.
Liquids are advanced left to right; solids are advanced from top
to bottom. Care is taken to transfer all solids including any slimes
adhering to the filter paper. The remainder of the leachate solution
from each test is saved for chemical analysis. Inorganic leachate
samples for laboratory analyses are usually filtered through a 0.45
micron filter, split into appropriate sample portions, and preserved as
required. In principle, the same procedures can be used for organics
with the following precautions:
1. materials must be non-absorbent (e.g., glass);
2. there must be no opportunity for volatilization (i.e., no air in the
mixing chamber); and
3. samples must be handled and transferred in a way that ensures no
losses of volatile organics.

Solids Solids Solids

Liquids to
Solids Ratio
1:1 Liquids

Solids
2:1 Liquids Advance
Top to
Bottom

3:1 Liquids
•

o Liquids Advance Left to Right 0

Fig. 2.1 Schematic of a typical sequential batch-contact test.
22 Geochemistry

The liquid to solid ratios that are selected for the sequential batch tests
control the volume of solution that a given weight of soil or geological
material would contact. These ratios can be related to pore volume
and, based on these ratios, an estimate of equivalent pore volume
throughput can be calculated. At the liquid to solid ratios specified in
the schematic of Fig. 2.1, leachate sample A3 would be equivalent to 0
to approximately 2 pore volumes of seepage throughput. At the other
extreme of the schematic, leachate sample C1 would be equivalent to
approximately 20 through 40 pore volumes of seepage throughput.
The attenuation capacity of the soil or geological material for a
chemical constituent present in the solution or ground-water sample
can be directly determined from the SBT data. The attenuation capacity
is calculated by summing the amount of the constituent removed from
solution during each contact with solids material, and by dividing this
amount by the weight of the soil or geological sample in contact with
the solution. This calculation is repeated for each contact of batch A, B,
and C, and the results are summed over the entire test sequence. In
contacts where a constituent was obviously desorbed from the earth
material sample, the attenuation capacity is adjusted downward by
subtracting the amount which was dissolved. The results of the cal-
culations are reported as milligrams (mg) of constituent attenuated per
kilogram (kg) of earth material.
A simpler form of the sequential batch-contact procedure can also
be used to determine parameters useful for modeling contaminant
transport in the ground. In a single-batch extraction test, leachate is
mixed with soil for 24 hours. The concentration of contaminant con-
stituents in the leachate is measured before and after 24 hours of gentle
agitation. If the constituent does not tend to partition from the aqueous
phase to the solid phase, the concentration in solution (aqueous phase)
will be the same before and after agitation. Such a constituent is said to
be 'conservative', 'non-reactive', or 'non-sorbed'.
With most waste constituents, the solution concentration will be
lower after 24 hours of gentle mixing because some of the constituent
will be sorbed by the soil. Such constituents are said to be 'sorbed',
'reactive', or 'attenuated' species. The mass of constituent sorbed, M,
is easily calculated from the difference in solution concentration before
and after agitation, and from the volume of leachate. A graph is plotted
of the mass of chemical constituent sorbed per mass of soil (5) versus
the concentration of the chemical constituent (C) in the solution. An
example of such a plot is shown in Fig. 2.2. This type of graph is called
an 'adsorption isotherm'. The relationship between 5 and C is some-
times linear, particularly for very low concentrations of metals and
organics in aqueous solutions. If the isotherm is linear, the slope, Rd , is
 Geochemical attenuation 23

CHEMICAL A

o
W
III
II:
o
!/)

zo
1-....1
w!/)
:)u..
~o
!/)!/)
z!/)
0<
u:::E
u..
o
!/)
!/)
<
:::E

SOLUTION CONCENTRATION, C

Fig. 2.2 Linear adsorption isotherms.

5
Rd = - (2.1)
C
where Rd is denoted' distribution ratio' or 'partition coefficient'. In Fig.
2.2, the Rd of chemical A is greater than Rd of chemical B, indicating
that chemical A is more strongly sorbed by soil than chemical B.
If the soil and solution are in equilibrium after 24 hours of agitation,
then Rd is equivalent to Kd, the 'distribution coefficient':
5
Kd=Rd=C (2.2)

The distribution coefficient, Kd , is often an input parameter to con-

taminant transport models.
With many chemical/soil combinations, and particularly for batch
tests performed on leachates containing high concentrations of solutes,
the adsorption isotherms are non-linear. A finite sorption capacity
results in the typical nonlinear adsorption isotherm shown in Fig. 2.3.
A common model used to describe such isotherms is the Freundlich
equation
24 Geochem is try

til

C
W
aJ
IE:
otil
I-..J
ffiB
::JtIl
I-u.
j::0
til til
ZtIl
0<
u:!:
u.
o
til
til
<
:!:

SOLUTION CONCENTRATION, C

Fig. 2.3 Non-linear adsorption isotherm.

(2.3)
where 5 is the mass of chemical constituent sorbed per mass of soil, C
is the concentration of the chemical constituent in the solution and Kd
and n are coefficients that depend upon conditions of the system.
If one uses a contaminant transport model that assumes a linear
isotherm, but the isotherm is in fact non-linear, the appropriate con-
centration range must be selected to estimate the value of Kd to imput
to the model.
The reader is reminded of the many factors, such as pH, that can
affect attenuation. These factors must be carefully controlled in batch
testwork, and must accurately reflect field conditions.

Column percolation testing

Column tests are at the other end of available laboratory methodology.
Column tests generally provide the most realistic information on the
attenuation properties of natural geological materials. If properly
designed, the column test procedure can more accurately describe
attenuation processes by simulating conditions that will occur in the
field. Unlike the sequential batch-contact test, the column test is
 Geochemical attenuation 25
PRESSURE
CHECK VALVE

CHECK V~L"~-~

INTERNAL FEED
SOlUTION RESE RVOI R

REMOlDED CLAY MATERIAL

Fig. 2.4 Operation of a typical column test.

usually a long-term procedure often taking several months to com-

plete, depending upon the permeability of the clay-bearing material
that is being evaluated.
The objective of the column testing is to force representative solutions
that contain chemical constituents of concern through columns packed
with samples of earth material. During operation of the columns,
effluent samples are periodically sampled for chemical analysis. By
determining the chemical composition of the effluent that has percolated
through the column material and comparing the results to the starting
'feed' solution, the attenuation properties of the soil or geological
material can be quantified. Such tests frequently are performed in
flexible-wall permeameters or in relatively simple rigid-wall columns
fabricated from large diameter sections of plastic tubing or pipe.
The operation of a typical down-flow column test performed on clay-
bearing material is illustrated in the generalized sketch of Fig. 2.4.
It is not uncommon for column tests to show decreased hydraulic
conductivities over time. This decrease in hydraulic conductivity is
usually due to the formation of secondary minerals in interstitial voids
of the clay-bearing material or to transformations in clay-mineral
phases.
As with sequential batch-contact testing, attenuation capacity can be
26 Geochemistry

calculated directly from analyses of column effluent samples. Attenua-

tion capacities are calculated by summing the amount of the con-
stituent that was removed from solution during each pore volume of
column throughput. If effluent analyses indicate an increase in con-
stituent concentration (suggesting desorption), the attenuation capacity
is adjusted downward by subtracting the amount that was desorbed.
The results of the calculations are reported as milligrams (mg) of con-
stituent attenuated per kilograms (kg) of natural geological material
packed in the column and contacted by the column 'feed' solution.

Biogeochemical testing
Biogeochemical degradation has been shown to be extremely effective
in attenuating the movement of many organic ground-water con-
taminants. Several laboratory testing programs have been developed to
evaluate the effectiveness of these biogeochemical processes (Brubaker
and Crockett, 1986). The purpose of such laboratory scale testwork is
four-fold, namely:
1. to determine the availability of indigenous organisms and their
capacity to survive in the subsurface environment;
2. to ascertain if these organisms have the ability to biodegrade soil
and ground-water contaminants;
3. to evaluate critical factors limiting the growth of such organisms;
and
4. to determine if enhancing environmental conditions for bio-
degradation might mobilize other contaminants or reaction
products.
Biological degradation is discussed further in later chapters in this
book.

2.2.4 Field methodology

Soil water samplers, commonly referred to as pressure/vacuum
lysimeters, are field sampling devices that have been used successfully
for monitoring water quality in the unsaturated zone. In the past, such
lysimeters have been used in the mining industry to monitor the
unsaturated zone for solution losses from heap leach facilities and for
leachate seeping from tailing ponds (Pyrih, 1991). In more recent
applications, lysimeters have been installed at various depths in ore
heaps in order to evaluate the efficiency of the leaching process and to
determine the completeness of heap neutralization. Soil water samplers
provide a means for field verification of geochemical attenuation.
The pressure/vacuum lysimeter consists of a cylindrical chamber
made of PVC and a low permeability porous ceramic cup. Pore water
 Phased ground-water monitoring 27
neoprene tube

~
casing

4-inch auger
boring wall

bentonite backfill

pressure-vacuum
tube (black)
~~~+- water discharge
tube (white)
pressure-vacuum ----oJ.;..:
• •,.,•.....,.-:-:":,
Iysimeter chamber bentonite
....................... >.:: backfill

1'l4----- - 2-inch split-spoon

setting hole

porous rt-1H~---- silica flour

ceramic - - ---1l4-. slurry
cup

NOT TO SCALE

Fig. 2.5 Schematic of a typical pressure/vacuum lysimeter.

in the form of soil moisture (for example, seepage) is drawn by vacuum

into the chamber through the porous ceramic and then retrieved
through access tubes to the surface. The construction of a typical
pressure/vacuum lysimeter is illustrated in Fig. 2.5.
The ability of a soil in the unsaturated zone to attenuate the move-
ment of potential ground-water contaminants can be confirmed and
quantified in field percolation tests using pressure/vacuum lysimeters.
The lysimeters are used to sample the soil moisture as the seepage
containing the contaminants of concern migrates downward. Field
installation and operation of a 'nest' of soil moisture lysimeters is
illustrated in Fig. 2.6. Further information on lysimeters is given in
Chapter 24, Vadose Zone Monitoring.

2.3 PHASED GROUND-WATER MONITORING

The fact that geochemical attenuation is predictable and quantifiable

can be used to advantage in the design of a phased ground-water
28 Geochemistry

water
ground surface drum

bentonite
Iysimeter seal
sampling
tubes
silica flour
.~:
2-inch split spoon
slurry setting hole
porous
ceramic NOT TO SCALE
cup

Fig_ 2.6 Field installation of a nest of pressure/vacuum lysimeters.

monitoring program. Such a phased monitoring program is more cost

effective and environmentally protective than the more common ap-
proach of collecting and analyzing samples for esoteric, nonconservative
parameters such as heavy metals or complex organics. The phased
monitoring system makes use of the fact that most environmentally
sensitive contaminants are retarded during migration through geologic
material, relative to the more-common major-ion components.
Experience has shown that plumes of migrating ground-water con-
tamination can be divided into three zones, as illustrated in Fig. 2.7
(Cherry, Sheperd and Morin, 1982; Rouse and Pyrih, 1990). The 'core'
zone, located closest to the source, is a zone in which the geochemical
attenuation capacity of the subsurface material has been depleted. The
ground water in the core zone is similar to that of the source. The
'active' zone is the site of intense geochemical interaction between
subsurface materials and migrating ground-water contaminants. The
active zone is the zone in which the previously-described reactions
primarily occur, and marks the site of active precipitate formation. The
downgradient 'neutralized' zone contains only conservative com-
ponents which generally do not enter into geochemical attenuation
reactions. These can include most of the major-ion components of
the migrating seepage plume, but may include oxidized species of
chromium or selenium.
 Phased ground-water monitoring 29
ground level

. . .~~"""""""""""
""""""""""""
""""""""""""
.. , ..
- ~"",. . "','
""""""",~~""""
",., "" ...
Fig. 2.7 Conceptual geochemical model of zones in a contaminant plume.

This conceptual model can be used to advantage in the design,

operation, and interpretation of a phased monitoring system. A well
located downgradient of a migrating seepage plume will yield ground
water essentially identical in chemical composition to ambient ground
water. The onset of the arrival of contamination will be characterized
by a shift in major-ion chemistry and the detection of other conservative
contaminants. Initial monitoring parameters should include only major
ions (Ca, Mg, Na, K, HC0 3 , C03 , 504, Cl) and those parameters
which geochemical testing has shown to be present in the seepage
solution that are not attenuated by geochemical mechanisms. Graphical
techniques such as the trilinear plot (Hill, 1940) or the Stiff diagram
(Stiff, 1951) are useful in detecting such shifts in major-ion composition.
There is no advantage to sampling for non-conservative ions during
the initial monitoring phase, since such ions will not migrate as rapidly
as the more-conservative major ions.
After major-ion sampling has demonstrated the arrival of the
neutralized zone at a well, monitoring should be intensified to include
those parameters which geochemical testing has shown to be least
attenuated. The order of arrival of non-conservative contaminants will
be inversely proportional to the degree of parameter attenuation, that
is, those parameters showing relatively little attenuation will arrive
before those demonstrating significant attenuation.
The phased monitoring approach has environmental advantages
in addition to the obvious economic advantage of reduced analytical
costs. By focusing attention on a specific suite of parameters and using
30 Geochemistry
graphical techniques which consider the overall percentage composi-
tion in a water sample, it is possible to detect subtle changes which
might be overlooked when viewed in terms of an extensive suite of
parameters. The detection of such a shift can, in turn, lead to the
initiation of more intensive sampling. The authors have seen several
investigations lead to a false sense of security by focusing on non-
conservative parameters as the indication of the presence or absence
of contamination. Major-ion analysis documented a more extensive
plume in which geochemical attenuation had been effective in retarding
the spread of non-conservative contaminants. Examples of such errors
in interpretations include the attenuation and biotransformation of
cyanide to nitrate and the reduction of nitrate to ammonia followed by
geochemical attenuation of the ammonia.

2.4 SUMMARY

The movement of potential ground-water contaminants in the sub-

surface is controlled by geochemical processes which take place when
chemical constituents of ground water interact with soils, sediments
and bedrock. Under the right conditions, the mobility of the con-
stituents is retarded by these interactions with the result that the
contaminants become fixed in geochemical 'traps' below the surface.
This behavior can be described by a relatively simple geochemical!
geohydrological conceptual model.
To be effective, ground-water monitoring programs and cleanup
strategies need to consider these geochemical interactions. Cost effective
and environmentally sound on-site and in-place treatment technologies
must be based on the geochemistry of the contaminants of concern.
Geochemical attenuation is a predictable, dependable, and quantifi-
able mechanism which can be used to advantage in the design of a
waste-handling facility and in the operations of a monitoring system.
Geotechnical engineers can more effectively design waste disposal
systems when they understand and utilize geochemical principles as
an integral part of the overall system design.

REFERENCES

Brubaker, G.R. and Crockett, E.L. (1986) In situ aquifer remediation using
enhanced bioreciamation, in Proceedings of the Second Annual Hazardous
Materials Management Conference West, Long Beach, CA.
Chatwin, T.D. and Trepanowski, J.J. (1987) Utilization of soils to mitigate
cyanide releases, in Proc., 3rd Western Reg. Conf. on Precious Metals, Coal,
and Environ., Rapid City, SO, September 23-26, 151-70.
Cherry, J.A., Shepherd, T.A. and Morin, K.A. (1982) Chemical composition and
 References 31

geochemical behavior of contaminated ground water of uranium tailings impound-

ments, Preprint 82-114, SME-AIME, February 1982.
Dragun, J. (1988) The Soil Chemistry of Hazardous Materials, Hazardous Materials
Control Research Institute, Silver Spring, MD.
Freeze, RA. and Cherry, J.A. (1979) Groundwater, Prentice Hall, Englewood
Cliffs, NJ.
Fritz, S.J. and Hall, S.D. (1988) Efficacy of various sorbic media in attenuation
of selenium, in J. Environ. Qual., 17(3), 480-4.
Griffin, RA., Frost, R.R, Au, A.K., Robinson, G.D. and Shimp, N.F. (1977)
Attenuation of Pollutants in Municipal Landfill Leachate by Clay Minerals: Part 2
- Heavy Metal Adsorption, Environmental Geology Notes, Illinois State
Geological Survey (79) April.
Griffin, RA., Keros c., Shrimp, N.F., Steel, J.D., Ruch, R.R., White, W.A.,
Hughes, G.M. and Gilkeson, R.H. (1976) Attenuation of Pollutants in
Municipal Landfill Leachate by Clay Minerals: Part 1 - Column Leaching and
Field Verification, Environmental Geology Notes, Illinois State Geological
Survey (78), November.
Grim, RE. (1968) Clay Mineralogy, McGraw-Hill Book Co., New York.
Hill, R.A. (1940) Geochemical patterns in Coachella Valley, California, in Trans.
Amer. Geophys. Union, 21.
Hornick, S.B. (1976) The interaction of soils with waste constituents, in Soil
Chemistry, Volume A: Basic Elements, (eds G.H. Bolt and M.G.M.
Bruggenwert) Elsevier Scientific Publishing Company, New York, 4-19.
Houle, J.J. and Long, D.E. (1980) Interpreting results from serial batch
extraction tests of wastes and soils, Proc., 6th Ann. Res. Symp., EPA-600/9-
80-010, 60-81.
Huiatt, J.L., Kerrigan, J.E., Ferron, A.O. and Potter, G.L. (1983) Cyanide from
mineral processing, Proc., Nat. Sc. Foundation Workshop, Salt Lake City, UT,
February 2-3, 1982.
Jenne, E.A. (1968) Controls on Mn, Fe, Co, Ni, Cu, and Zn concentrations in
soils and water: the significant role of hydrous Mn and Fe oxides, Adv. in
Chem. (73),337-87.
Klute, A. (1986) Methods of Soil Analysis, Amer. Soc. of Agronomy, Madison,
WI, Parts 1 and 2.
Knowles, c.J. (1976) Microorganisms and cyanide, Bact. Rev., Sept., 652-781.
Peck, G.A. (1990) Surface treatment processes for heavy metals removal, in
Proc., Environ. Haz. Conj. and Exposition, Seattle, WA, May 15-17.
Pyrih, RZ. (1991) Use of lysimeters to monitor the effects of tailings
dewatering on pore-water chemistry, Proc., Randol Gold Forum CAIRNS '91,
April.
Pyrih, R.Z. and Rouse, J.V. (1989) Attenuation processes: a viable regulatory
alternative, in Proc., Environ. Haz. Conj. and Exposition, Seattle, WA.
Rouse, J.V. (1988) Copper, chromium and arsenic in the environment: natural
concentrations and geochemical attenuation, in Proc., Amer. Wood
Preservers' Assoc., 84, 110-3.
Rouse, J.V. and Pyrih, R.Z. (1985) Natural geochemical attenuation of
contaminants contained in acidic seepage, in Proc., lnt. Conj. on New
Frontiers for Haz. Waste Management, EPA/600/9-85/025, 192-9.
Rouse, J.V. and Pyrih, RZ. (1988) Natural geochemical attenuation of trace
elements in migrating precious-metal process solutions, in Proc., Perth lnt.
Gold Conj., Oct. 77-81.
Rouse, J.V. and Pyrih, R.Z. (1989) In-place cleanup of chromium contamination
of soil and ground water, Proc., Environ. Haz. Conj., Houston, TX.
Rouse, J.V. and Pyrih, RZ. (1990) In-place cleanup of heavy metal
32 Geochemistry

contamination of soil and ground water at wood preservation sites, in

Proc., Amer. Wood Preserver's Assoc., 86.
Schmidt, J.W., Simovic, L. and Shannon, E. (1981) Natural degradation of cyanide
in gold milling effluents, Industry Seminar, Waste Water Technology Centre,
Technology Development Board, Water Pollution Control Directorate,
Environment Canada.
Schmidt-Collerur, J.J. (1978) Investigations of the relationship between organic matter
and uranium deposits, Bendix Field Engineering Corporation, final report,
subcontract no. 76-030-E, Denver Research Institute.
Scott, J.5. (1984) An overview of cyanide treatment methods for gold mill
effluents, in Proc., Conf. on Cyanide and the Environ., Tucson, AZ, Dec.
11-14,307-30.
Simovic, L., Snodgrass, W.J., Murphy, K.L. and Schmidt, J.W. (1984)
Development of a model to describe the natural degradation of cyanide in
gold mill effluents, in Proc., Conf. on Cyanide and the Environ., Tucson, AZ,
Dec. 11-14, pp. 413-32.
Smith A. (1988) Cyanide degradation and detoxification in a heap leach, in
Introduction to Evaluation, Design, and Operation of Precious Metal Heap
Leaching Projects, Society of Mining Engineers, Littleton, CO, pp. 293-305.
Smith, A. and Struhsacker, D.W. (1988) Cyanide geochemistry and
detoxification regulations, in Introduction to Evaluation, Design, and
Operation of Precious Metal Heap Leaching Projects, Society of Mining
Engineers, Littleton, CO, pp. 275-92.
Stiff, H.A., Jr. (1951) The interpretation of chemical water analysis by means of
patterns, J. Petroleum Tech., 3(10), 15-71.
University of California at Berkeley (1988) Mining Waste Study, Final Report,
prepared by Mining Waste Study Team of the University of California at
Berkeley, commissioned by the California State Legislature and funded by
State Water Resources Control Board, the Department of Health Services,
and the California Department of Conservation.
 CHAPTER 3

Contaminant transport
Charles D. Shackelford

3.1 INTRODUCTION

The purpose of this chapter is to present the basic concepts for de-
scribing contaminant transport in and through porous materials. The
presentation is based on the premise that the reader has little or no
formal background in contaminant transport theory. As a result, the
material covered is limited to one-dimensional transport of miscible
contaminants (i.e., solutes) in saturated porous media. Several refer-
ences are recommended for additional study.

3.2 TRANSPORT PROCESSES

3.2.1 Advection
Advection is the process by which solutes are transported along with
the flowing fluid or solvent, typically water, in response to a gradient
in total hydraulic head. Due to advection, non-reactive solutes (i.e.,
solutes which are not subject to chemical or biological reactions) are
transported at an average rate equal to the seepage velocity of the
fluid, or
v
vs =- (3.1)
n
where Vs is the seepage or average linear velocity of the water (solvent),
n is the total porosity of the porous material, and v is the flux of water
(i.e., quantity of flow per unit area per unit time). Equation (3.1)
presupposes that all of the voids in a porous material are equally

Geotechnical Practice for Waste Disposal.

Edited by David E. Daniel.
Published in 1993 by Chapman & Hall, London. ISBN 0412351706
34 Contaminant transport

effective in conducting flow. The flux is given by Darcy's law which,

for one-dimensional flow, can be written as

v = 9= - K ah = Ki (3.2)
A ax
where Q is the volumetric flow rate of the water, A is the total cross-
sectional area (solids plus voids) perpendicular to the direction of flow,
K is the hydraulic conductivity, h is the total hydraulic head, x is the
direction of flow, and i is the dimensionless hydraulic gradient. The
seepage velocity, v" reflects the fact that the fluid actually can flow
only through the void space of the porous material whereas the flux, v,
represents the volumetric flow of fluid through the total cross-sectional
area of the material.
The time required for a nonreactive solute to migrate through a
saturated soil of thickness L, known as the solute transit time, due to
advection can be estimated using the seepage velocity as follows.
L nL
t =-=- (3.3)
v, Ki
where t is the transit time. When Eq. (3.3) is used to estimate the
transit time of the solute, it is implied that all solute constituents are
carried along with the fluid at the velocity VS. Under these conditions,
the advective mass flux of a particular chemical species can be
calculated as
fA = vc = Kic = nv,c (3.4)
where fA is the advective mass flux (mass flowing through a unit
cross-sectional area in a unit of time) and c is the concentration of the
solute in the liquid phase of the porous material based on the volume
of solution in the porous material (i.e., mass of solute per unit volume
of mixture).
With some porous materials, e.g., fractured soil or rock, most of the
advective flow occurs through only part of the total void space of the
material. For these materials, an 'effective porosity', ne, is defined as
the volume of fluid conducting pores divided by the total volume
(pores plus solids) of the material. For materials with ne < n, ne should
be substituted for n in Eqs. (3.1), (3.3), and (3.4). For more discussion
of effective porosity, the reader is advised to consult Freeze and Cherry
(1979), Horton et al. (1985), and Peyton et al. (1986).

3.2.2 Diffusion
Diffusion may be thought of as a transport process in which a chemical
or chemical species migrates in response to a gradient in its con centra-
 Transport processes 35

tion, although the actual driving force for diffusive transport is the
gradient in chemical potential of the solute (Robinson and Stokes,
1959). A hydraulic gradient is not required for transport of contaminants
by diffusion.
The fundamental equation for diffusion is Fick's first law which, for
one-dimensional transport, can be written as

Jo =
ac
-Do~ (3.5)
ax
where Jo is the diffusive mass flux, x is the direction of transport, and
Do is the 'free-solution' diffusion coefficient. Equation (3.5) represents
the one-dimensional form of Fick's first law describing diffusion in
aqueous or free solution (i.e., no porous material).
For diffusion in saturated porous material, a modified form of Fick's
first law is used

Jo = -l:Don~
ac (3.6)
ax
or

Jo = -D*n~
ac (3.7)
ax
where 't is a dimensionless tortuosity factor and D* is the 'effective'
diffusion coefficient. The porosity term is required in Eq. (3.7) since the
diffusive flux, J0, is defined with respect to the total cross-sectional
area of the porous medium. The tortuosity factor accounts for the
increased distance of transport and the more tortuous pathways ex-
perienced by solutes diffusing through porous media. Tortuosity is
expressed as

't = (tef (3.8)

where L is the macroscopic, straight-line distance between two points

defining the flow path, and Le is the actual, microscopic or effective
distance of transport between the same two points. Since Le > L, 't <
1.0 and D* < Do. Therefore, mass transport due to diffusion in porous
materials is slower than mass transport due to diffusion in free or
aqueous solution. Typical values of 't are reported in the range 0.01 to
0.67 (Perkins and Johnston, 1963; Freeze and Cherry, 1979; Daniel and
Shackelford, 1988; Shackelford, 1989; Shackelford and Daniel, 1991).
Factors affecting the determination of 't are described by Shackelford
and Daniel (1991). In some instances, the porosity term in Eq. (3.7)
is included in the definition of the effective diffusion coefficient
(Shackelford, 1988a; Shackelford and Daniel, 1991; Shackelford, 1991).
36 Contaminant transport

3.2.3 Coupled flow processes

In general, water can flow through porous media not only in response
to a gradient in total hydraulic head but also in response to gradients in
chemical composition (chemico-osmosis), electricity (electro-osmosis),
and temperature (thermal osmosis). Also, in the absence of advective
flow, solutes can migrate through porous materials not only in response
to a concentration gradient but also in response to an electrical gradient
(electro-phoresis) and a thermal gradient (thermal diffusion). Solutes
also may be filtered (ultrafiltration) by some porous materials. These
additional flow processes typically are referred to as coupled flow
processes in order to distinguish them from the direct flow processes
represented by advection and diffusion. The description of coupled
flow processes is based on the theory of the thermodynamics of
irreversible processes. A description of this theory is beyond the scope
of this chapter, but an excellent review of the relation of the theory to
coupled and direct flow processes can be found in Bear (1972). Mitchell
(1976) provides examples of the application of coupled flow processes
to typical geotechnical engineering problems.
In the formulation of the contaminant transport equations, coupled
flow processes often are neglected or are considered to be insignificant
for at least two reasons. First, coupled flow processes are significant
only at the relatively low flow rates which typically are associated with
flow through fine-grained soils (e.g., clays and silts). Since most of the
applications of contaminant transport theory traditionally have been
associated with problems pertaining to the transport of contaminants
through highly permeable, granular materials (e.g., aquifers), the
effects of coupled flow processes typically have been assumed to be
negligible. Nevertheless, coupled flow processes may be significant in
materials with high activity and/or low void ratio (Olsen, 1969;
Greenberg et al., 1973). Second, the determination of the phenomeno-
logical coefficients associated with the equations describing coupled
flow processes is difficult, at best. As a result, such determinations
usually can be performed only under highly controlled conditions in
the laboratory and extrapolation of the results to field situations is
uncertain.

3.2.4 Mechanical dispersion

In traditional contaminant transport theory, a mechanical dispersive
flux, JM, is added to the total mass flux of the solute to account for the
spreading of the solute due to variations in the seepage velocity, v s,
which occur during transport in and through porous materials. On a
microscopic level, these variations are thought to be related to the three
different effects illustrated in Fig. 3.1 (Fried, 1975; Bear, 1979; Freeze
 Transport processes 37
pore wall

direction of velocity
flow distribution
(a) -----+ across
pore
channel

pore wall

direction of
(b) A, flow
Vz> v,

equation of continuity: A, v, = Azvz

(c)

Fig. 3.1 Microscopic dispersion in soil: (a) effect of velocity distribution across
single pore; (b) effect of variation in pore sizes: and (c) effect of tortuous nature
of flow paths [after Freeze and Cherry (1979); and Fried (1975)].

and Cherry, 1979). First, the velocity of flow across any given pore
channel within the material will be greater in the middle of the pore
channel than it is near the walls of the pore channel (Fig. 3.1a). This is
the same effect which is known to occur in fluid flow through pipes
and in rivers, streams, and channels. Second, the equation of con-
tinuity predicts that the flow velocity across a smaller pore opening will
be greater than that across a larger pore opening as illustrated in Fig.
3.1 b. Finally, velocity variations will result due to the tortuous nature
of the flow paths existing in nearly all porous materials (Fig. 3.1c).
On a larger scale, mechanical dispersion is thought to be caused by
the different flow rates resulting from heterogeneities that typically are
encountered whenever transport occurs over relatively large areas or
regions. This type of dispersion is illustrated in Fig. 3.2 where flow
through the bulk sand medium is shown to be interrupted by the
existence of the low-hydraulic-conductivity clay lenses dispersed
38 Contaminant transport
Ground Surface

..
Direction of
Flow

II

..
..
II

..
..
..

Fig. 3.2 Mechanical dispersion on large or regional scale.

throughout the sand layer. Due to the existence of the clay lenses, the
transport proceeds around rather than through the clay resulting in a
situation which is analogous to that shown in Fig. 3.1c for microscopic
dispersion. Therefore, the dispersion illustrated in Fig. 3.2 may be
thought to be due to a 'macroscopic tortuosity effect'.
The mechanical dispersive flux typically is assumed to be a Fickian
process which, for one dimension, can be expressed as
ac
JM = -D",n ax (3.9)

where JM is the mechanical dispersive flux and Om is the mechanical

dispersion coefficient. Since mechanical dispersion results from
variations in the magnitude of the seepage velocity, the mechanical
dispersion coefficient often is assumed to be a function of the seepage
velocity, or
(3.10)
where aL is the longitudinal dispersivity of the porous medium in the
direction of transport, and ~ is an empirically determined constant
between 1 and 2 (Freeze and Cherry, 1979). In most applications, the
exponent, ~, is assumed to be unity, i.e., 0", is assumed to be a linear
function of Vs in Eg. (3.10). However, ~ may be greater than unity in
many situations (Anderson, 1979, 1984; Bear and Verruijt, 1987). Also,
the dispersivity, aL, is probably scale dependent with larger values for
aL being associated with greater transport distances (Pickens and
Grisak, 1981). For example, values of aL reported from the results of
field studies may be as much as four to six orders-of-magnitude greater
than the corresponding laboratory measured values which commonly
 Transient transport 39
are found to range between 0.1 and 10mm (Freeze and Cherry, 1979).
Regardless of the appropriate values for at and ~, the relationship for
Om in Eq. (3.10) indicates that the effect of mechanical dispersion
increases as the seepage velocity increases and vice versa. As a result,
for typical problems involving low-hydraulic-conductivity barrier
materials, mechanical dispersion does not appear to be significant due
to the relatively low flow rates (Rowe, 1987).

3.2.5 Combined transport

In the absence of coupled flow processes, the total mass flux, I, of the
solute or contaminant species is the sum of the advective, diffusive,
and dispersive fluxes, or
1= IA + 10 + 1M (3.11)
Based on Eqs. (3.4), (3.7), and (3.9), the total mass flux for one-
dimensional transport in saturated porous material is

I = nv c - O*n ac - 0 n ac
5 ax 11/ ax
(3.12)

or
ac
I = nvsc - O"n ax (3.13)

where Oh is the hydrodynamic dispersion coefficient given by

0" = 0* + 011/ = 0* + atVs (3.14)
The exponent ~ in Eq. (3.10) has been assumed to be unity in the
formulation of Eq. (3.14). The hydrodynamic dispersion coefficient
accounts for dispersion of the solute due to both diffusion and mech-
anical dispersion.

3.3 TRANSIENT TRANSPORT

3.3.1 Conservation of mass

The transient transport of a chemical species through saturated
material is based on the conservation of mass for a representative
elementary volume (REV) of soil, or (Freeze and Cherry, 1979)
net rate of mass flux mass flux increase (or
mass increase into out of ± decrease) in mass
within the REV the REV the REV due to chemical
reactions occur-
ring within the
REV
40 Contaminant transport

A REV is the minimum volume that will allow for the application of the
continuum approach to flow or transport through porous media (Bear,
1972, 1979; Bear and Verruijt, 1987). In mathematical terminology, the
conservation of mass is reflected by the continuity equation, or
am = -v· J ±
- R ± Am (3.15)
at
where m is the total (adsorbed plus liquid phase) mass of solute per
unit volume of soil, A is a general rate constant used to describe such
reactions as radioactive and/or biological decay and R is a general term
representing all other chemical and biological reactions. The positive
signs (+) is Eq. (3.15) are used for concentration source terms (e.g.,
mineral dissolution) whereas the ( - ) signs are used for terms
representing concentration sinks (e.g., precipitation).

3.3.2 Advection-dispersion equation

A number of simplifying assumptions often are required in practice to
reduce Eq. (3.15) to a more usable form. For example, the porous
medium often is assumed to be homogeneous, isotropic, and non-
deformable, and transport is assumed to be governed by steady flow
of an incompressible fluid. Also, only trace concentrations of solutes
typically are considered so that changes in fluid density due to changes
in solute concentrations can be neglected. Coupled flow processes
usually are neglected and only equilibrium exchange reactions (e.g.,
reversible sorption reactions) are included routinely in the modelling of
the transport of reactive solutes through porous media. When these
assumptions are acceptable, and transport is assumed to occur only in
one direction (say, the x-direction), Eq. (3.13) can be used for the flux
term, I, and Eq. (3.15) can be reduced to (Freeze and Cherry, 1979)
ac o"a 2c vsac
(3.16)
at RdaX2 - Rdax
where Rd is the dimensionless retardation factor. In the study of con-
taminant transport, Eq. (3.16) commonly is referred to as the one-
dimensional advection-dispersion equation. When the seepage velocity
is sufficiently low such that mechanical dispersion is negligible, the
advection-dispersion equation (3.16) effectively reduces to an advection-
diffusion equation since the hydrodynamic dispersion coefficient from
Eq. (3.14) is dominated by the effective diffusion coefficient (i.e., Oh ""
0*).
The retardation factor represents the relative rate of fluid flow to the
transport rate of a reactive solute (Freeze and Cherry, 1979), or

(3.17)
 Transient transport 41
where VR is the transport rate, or velocity, for the center of mass of
the reactive solute. For nonreactive (nonadsorbing) solutes, such as
chloride (Cl-), the retardation factor is unity and the solute is trans-
ported at the rate of the seepage velocity in accordance with Eq. (3.17).
As a result, Eq. (3.16) may be written as

ae = 0 a2e _ v ae (3.18)
at "ax2 ax 5

However, for reactive (adsorbing solutes), the retardation factor is

greater than unity, i.e., Rd > l. Therefore, Eq. (3.17) indicates that
adsorbing solutes are transported at a reduced rate, VR, relative to
nonadsorbing solutes. In such cases, it often is convenient to rewrite
the advection-dispersion equation (3.16) for adsorbing solutes by
utilizing Eq. (3.17) as

(3.19)

where DR = Oh/Rd'
The value of Rd typically is determined in the laboratory from the
results of either column tests (discussed later) or batch equilibrium
tests (discussed in Chapter 2) using the relationship

Rd =1+ PbKI' (3.20)

n
where Pb is the bulk (dry) density of the soil and Kp is the 'partition
coefficient'. The partition coefficient relates the change In the adsorbed
concentration relative to a change in the equilibrium concentration of
the chemical species as
dS
KI' = de (3.21)

where 5 is the adsorbed concentration, or mass of solute adsorbed per

mass of soil. Equation (3.21) represents the slope of a plot of 5 versus e,
also known as an adsorption isotherm. The term 'partition coefficient'
applies when the adsorption isotherm is nonlinear and, therefore,
when Kp is dependent on the equilibrium concentration, i.e., Kp = f(e).
Therefore, the partition coefficient of a nonlinear adsorption isotherm
is equal to the tangential slope of the adsorption isotherm evaluated at
a specific value for the equilibrium concentration, e. When the adsorp-
tion isotherm is linear, the partition coefficient is constant and is called
the 'distribution coefficient, K/.
In general, there are two types of nonlinear adsorption isotherms -
concave and convex - as well as the linear adsorption isotherms which
are of interest in contaminant transport (Melnyk, 1985; Shackelford and
Daniel, 1991). In chemical engineering, concave and convex adsorption
42 Contaminant transport

Equilbrium concentration, C

Fig. 3.3 General types of adsorption isotherms [after Melnyk (1985); and
Shackelford and Daniel (1990)].

isotherms often are referred to as favorable and unfavorable isotherms,

respectively (Weber and Smith, 1987). Concave isotherms are the more
common type of nonlinear isotherms for transport in soil. The differ-
ence in the three types of isotherms is illustrated schematically in Fig.
(3.3). In a concave isotherm, the slope of the isotherm at a lower
equilibrium concentration is greater than it is at a higher equilibrium
concentration, whereas the opposite is true for a convex isotherm
(compare points for C1 and C2 in Fig. 3.3). Since the rate of transport for
a reactive solute (VR) is inversely proportional to the retardation factor
and, therefore, the partition coefficient (compare Eqs. (3.17), (3.20),
and (3.21)), a reactive solute will be transported faster at a higher
concentration for a concave isotherm, whereas the opposite is true for
a convex isotherm (see Figs. 3.4a and 3.4b). As a result, the concen-
tration profile described by a convex isotherm (Fig. 3.4b) tends to
spread out more during transport and forms what is known as a
'concentration wave'. However, since it is physically impossible for a
higher concentration of a given solute to reach a specified distance
before a lower concentration of the same solute reaches the same
distance, the concentration profile shown in Fig. 3.4a for concave
adsorption behavior tends to form what is known as a 'concentration
step'. As a result, the limiting partition coefficient for reactive solutes
described by concave isotherms is based on the use of secant lines
 Transient transport 43

C2C1 ~w,.,
C1 ............ ~
I t1 ~
o x o X

(a) (b)

C C

(c) (d)

Fig. 3.4 Effects of different types of adsorption isotherms: (a) concentration-

distance curve for concave adsorption behavior; (b) concentration-distance
curve for convex adsorption behavior; (c) secant formulation for concave
isotherm; and (d) distribution coefficient for linear adsorption isotherm.

instead of tangent lines, as shown in Fig. 3.4c (Melnyk, 1985). The use
of secant lines forces all equilibrium concentrations of a solute up to the
value at the intersection of the secant line and the adsorption isotherm
(i.e., C2 in Fig. 3.4c) to have the same value for the partition coefficient
and, therefore, the same transport rate. In essence, the slope of the
secant line is analogous to the slope associated with a constant, dis-
tribution coefficient for a linear isotherm. The above concepts are
described in more detail by Melnyk (1985).
In most applications, nonlinear adsorption behavior is described by
either a Freundlich or a Langmuir isotherm equation, as
5 = Kfc N (3.22)
or

5= KLMc (3.23)
1 + KLc
where 5 is the adsorbed concentration, Kf and N are the Freundlich
isotherm equation parameters (Kd and n, respectively, in Eg. (2.3)), and
KL and M are the Langmuir isotherm equation parameters.
44 Contaminant transport
For the cases of a nonlinear, convex adsorption isotherm described
by a Freundlich or a Langmuir isotherm, the retardation factor becomes,
respectively

(3.24)

or

Rd = 1 + PI> KLM (3.25)

n (1 + KLC)2
When the adsorption behavior is described by nonlinear, concave
isotherms, the limiting partition coefficient can be evaluated using a
secant formulation, or

(3.26)

where Co is the equilibrium concentration which defines the particular

slope of the secant line used for the evaluation of K" (e.g., C2 in Fig.
3.4c). For secant lines passing through the origin of the adsorption
isotherm, the limiting retardation factors based on the Freundlich and
Langmuir isotherm equations are, respectively

Rd -- 1 + Pb KfCo N~l (3.27)

n
and

Rd =1+ Pb KLM (3.28)

n 1 + KLCo
Retardation factors based on Eqs. (3.27) or (3.28) can be used to
estimate the extent of migration of the center of mass of a contaminant
plume in the case of nonlinear, concave adsorption behavior. How-
ever, it is unlikely that accurate estimates of the distribution of con-
taminants in the soil can be made using retardation factors based on
Eqs. (3.27) or (3.28) due to the approximation of the nonlinear adsorp-
tion behavior by linear secant formulations. Davidson et al. (1976)
illustrate the application of Eq. (3.27) to describe the migration of high
concentrations of pesticides from hazardous waste landfill sites.
For the case of a linear adsorption isotherm (Fig. 3.4d), the retarda-
tion factor is written simply as

(3.29)

As a result, all concentrations have the same transport rate, i.e., the
slope of the isotherm is constant at all equilibrium concentrations.
 Transient transport 45
3.3.3 Transport of reactive (adsorbing) organic compounds
The adsorption of hydrophobic organic contaminants at relatively low
concentrations also can be described adequately by Eq. (3.29) provided
the distribution coefficient is defined as
(3.30)
where foe is the fraction (by weight) of organic carbon in the soil and Koe
is the organic carbon partition coefficient (Karichoff et aI., 1979). The
organic carbon partition coefficient has been correlated empirically
with a number of parameters, particularly the octanol-water partition
coefficient, Kow. These empirical correlations are covered in more detail
by Griffin and Roy (1985).

3.3.4 Transport of radioactive chemical species

In the disposal of nuclear wastes, the migration of radioactive chemical
species, or radioisotopes, is a primary environmental concern. In
addition, radioisotopes often are used as tracers in the study of solute
transport through porous materials. In such circumstances, the decay
term, Am, in Eq. (3.15) should be evaluated. If the half-life of the
radioactive isotope is short relative to the time frame of transport,
the decay term should be included in the general formulation of the
advection-dispersion equation (3.16) to give
oc Dh o2c vsoc A
ot = Rd OX2 - Rdox - C (3.31)

where A is the decay constant of a linear or first-order decay reaction

given by
dm
- = -Am (3.32)
dt
In the formulation of Eqs. (3.31) and (3.32), the decay constant, A, of
the solute in both the liquid and the solid phases of the soil has been
assumed to be the same. Based on Eq. (3.32), the decay constant can be
evaluated as follows:

J O.Smdm
- = -A
Itl dt
°
2

m m (3.33)

or

(3.34)
46 Contaminant transport
where t~ is the half-life of the radioisotope, i.e., the time required for
half of the radioisotope to decay. Equation (3.34) indicates that as tl
increases, the decay constant A decreases. Therefore, for first-ordef
decay, the decay term in Eq. (3.15) can be neglected without incurring
significant error in the analysis when the radioactive solutes have long
half-lives.

3.3.5 Fick's second law

At low values for the seepage velocity, diffusion is more significant
relative to advection. In the limit, i.e., as Vs ~ 0, Eqs. (3.16) and (3.18)
reduce to the following two expressions for Fick's 2nd law for diffusion
of adsorbing and nonadsorbing solutes, respectively:
ac 0*a 2c a2c a2c
-=--=0*-=0- (3.35)
at Rd aX2 A a~ s ax2

and

(3.36)

where OA in Eq. (3.35) is referred to as an 'apparent diffusion

coefficient' (Li and Gregory, 1974) and Os is known as the 'effective
diffusion coefficient of the reactive solute' (Gillham et al., 1984; Quigley
et al., 1987). The result of using OA or Os in Eq. (3.35)is that only one
unknown must be solved instead of the two unknowns (since OA = Ds
= O*IRd). Shackelford and Daniel (1991) and Shackelford (1991) discuss
the use of OA or Os in application and summarize several other
expressions for Fick's 2nd law of diffusion which exist due to
differences in the definitions of 0* and differences in the expression of
the solute concentration.

3.4 EFFECTS OF MATERIAL PROPERTIES

The effects of several material properties (Rd, 0"/1 0*, K, and n) on the
transport of solutes in saturated porous media can be illustrated with
the aid of solute 'breakthrough curves'. Breakthrough curves represent
the temporal variation in the concentration of a solute at the effluent
end of a column of porous material. Breakthrough curves can be
measured using laboratory columns by:
1. establishing steady-state fluid flow conditions;
2. continuously introducing at the influent end of the column a liquid
containing a solute at concentration co; and
3. monitoring the solute concentration at the effluent end.
 Effects of material properties 47
A schematic representation of a solute breakthrough curve for a
continuous source concentration for a column of length L is presented
in Fig. 3.5a. The effluent concentration, c, is expressed for convenience
as a relative concentration, clco.
With respect to Fig. 3.5a, the breakthrough curve can be divided into
three distinct zones. Zone 1 represents the time required (to) for the
solute to reach the effluent end of the column. Zone 3 represents the
steady-state transport condition with respect to the solute which occurs
after time tfwhere the effluent concentration is the same as the influent
concentration, Co. It is important to note that it takes some time before
steady-state solute transport occurs whereas steady-state fluid (solvent)
flow is a pre-condition for the column test. Zone 2 is a transient,
transitional zone wherein the effluent concentration is gradually rising
from zero to Co. Due to the spreading effect of the solute front in zone
2, there are an infinite number of possible transit times depending on
the particular choice of the relative concentration, clco. However, the
'usual' practice has been to define the transit time with respect to a
relative concentration of 0.5, which is time tL in Fig. 3.5a (Bowders et
al., 1985). The concentration-distance curves for this case are illustrated

Q
tf
zone 1 zone 2 zone 3
5- 1 .0
~
"E
g 0.5 -+ . . . . . . . . . . . . . . . . . . . . .
Q)
+ ....................................... /
8
Q)
>
.~ 0 +----......;-=------i----+----+
~ 0 tL tf
transit time

(a)

c:-
.2 1.0 +-_~-"""""""--= . . . . . . . . . . . . . . . . . . . . + . . . . . . . . . . . . . . . . . . . . . . . . . ..
~
"E
8~ 0 .5
Q)
>
';
_....
Q) 0 +-___ ~~ ___ ~_L- __ ~ _ _ _+

o X = L
distance of transport, x

(b)

Fig. 3.5 Solute breakthrough curve for constant source concentration: (a)
concentration-time profile; and (b) concentration-distance profiles for different
times.
48 Contaminant transport

in Fig. 3.5b for three different elapsed times after introduction of the
solute at the influent end of the column.

3.4.1 Effect of retardation

The effect of retardation on the transport of a solute is illustrated in
Fig. 3.6a. The abscissas for the breakthrough curves in Fig. 3.6 have
been defined in terms of pore volumes of flow instead of time. The
number of pore volumes of flow (T) is equal to the cumulative volume
of flow through the sample divided by the volume of void space in the
material, or

T = vAt =!!.!..- = ~ = Vs (3.37)

LAn Ln (Lit) VR

where A is the cross-sectional area of the material and L is the length of

the column. Under steady-state fluid flow conditions, Eq. (3.37)
indicates that the number of pore volumes of flow is directly
proportional to time. For purely advective transport of nonreactive
solutes, the velocity of the solute, VR, is the same as the seepage
velocity, v s , and (provided ne = n) complete breakthrough of the solute
will occur at one pore volume of flow (curve 1 in Fig. 3.6a). This ideal
condition often is referred to as 'piston' or 'plug' flow since there is no
spreading of the solute front.
For solutes subject to reversible sorption reactions, i.e., reactive
solutes, VR < Vs and the solute will be retarded. This retardation effect
is manifested as an offset in the solute breakthrough curve of the
reactive solute relative to the nonreactive solute (A in Fig. 3.6a). With
respect to the advective transport of an adsorbing solute, VR (=Llt) in
Eq. (3.35) represents the 'velocity' of the retarded solute. The transit
time based solely on the advective transport of adsorbing solutes can
be estimated from the retarded solute velocity as
L LRd nLRd
t=-=-=-- (3.38)
VR Vs Ki
Since Rd > 1.0 for retarded solutes, the transit time based on Eq. (3.38)
for adsorbing solutes will be greater than the transit time for
nonadsorbing solutes (Eq. 3.3).
Again, it is important to distinguish clearly between steady-state

Fig. 3.6 (opposite) Effect of material properties on solute breakthrough curves

for constant source concentration: (a) effect of retardation-piston flow; (b) effect
of effective porosity-piston flow; (c) effect of mechanical dispersion; (d) effect of
diffusion. [Note: curves 1 and 3 - nonreactive solutes; curves 2 and 4 - reactive
solutes].
 A
1.0
I- ·1
1 .... 2 ....

.£ 0.5
Co

0
0 1.0 2.0
pore volumes of flow
(a)

Rd(1 - B) Rd B
·1-
1.0 I- B
I- -I-
A
:1
3 .... 14-1 14-4 14-2
.£ 0.5
Co

0
0 0.75 1.0 1.50 2.0
pore volumes of flow
(b)

A
1.0
·1

.£ 0.5
Co

0
0 1.0 2.0
pore volumes of flow
(c)

1.0

.£ 0.5
Co

0
0 1.0 2.0
pore volumes of flow
(d)
50 Contaminant transport

fluid flow and steady-state solute transport. The retardation factor, Rd ,

as defined by Eq. (3.17) is a measure of how much slower a retarded
solute moves relative to the steady-state movement of the fluid, or
solvent. Once complete breakthrough of the solute has occurred, the
solute is no longer retarded (as the term is used here) since 'steady-
state' solute transport, by definition, means that the influent and
effluent mass fluxes of the solute are equal. As a result, steady-state
transport is achieved only after the liquid-phase concentration of the
solute is in equilibrium with the solid-phase concentration of the solute
throughout the entire length of the column.
A comparison of Eqs. (3.17) and (3.37) reveals that the retardation
factor, R d , is equal to the number of pore volumes of flow, T, it takes
for the solute to break through the effluent end of the column. This
relationship can be utilized to determine the value for the retardation
factor for an adsorbing solute. For example, the breakthrough curve in
Fig. 3.6a for the retarded solute is given by curve 2, so the retardation
factor is 2.0, i.e., Rd = T = 2.0.

3.4.2 Effective porosity

In fine-grained soils, some of the fluid in the pore space may be
immobile due to dead-end pores and/or attraction of fluid molecules to
the surface of the soil particles (Bear, 1972, 1979). As a result, the
portion of the pore space available for solute transport may be
significantly less than the total pore space. In such cases, an effective
porosity, net should be used instead of the total porosity, n, in Eq. (3.1)
to determine the seepage velocity. Effective porosities on the order of
2-100% of the total porosity have been measured in compacted soils
(Horton et al., 1985; Liao and Daniel, 1989).
The effect of effective porosity on purely advective (piston) transport
of solutes is illustrated by the offset, B, in Fig. 3.6b. Since ne < n, the
seepage velocity is greater than that predicted using the total porosity,
and actual breakthrough of the solutes occurs earlier than expected.
For the case shown in Fig. 3.6b, B represents 0.25 pore volumes of flow
and, therefore, only 75% of the total pore space is effective in
conducting the fluid flow (i.e., ne = 0.75n). As a result, breakthrough
of nonreactive (nonadsorbing) solutes will occur after only 0.75 pore
volumes of flow (curve 3) rather than after one pore volume of flow
(curve 1). On the other hand, complete breakthrough (i.e., assuming
piston transport) of reactive solutes occurs after Rd(l - B) pore
volumes of flow, or RdB pore volumes sooner than expected. For
example, Rd for the reactive solute in Fig. 3.6 is 2 and, therefore,
breakthrough of the reactive solute is expected at 2 pore volumes of
flow (curve 2). However, due to the effective porosity effect,
breakthrough of the reactive solute actually occurs after only 1.5 pore
volumes of flow (curve 4).
 Effects of material properties 51
3.4.3 Mechanical dispersion
In general, complete breakthrough of a solute will not occur instan-
taneously. Instead, there is a gradual rise in the solute concentration
from C = 0 to C = Co due to the spreading effect caused, in part, by
mechanical dispersion. This spreading effect gives the solute break-
through curve its characteristic S-shape shown in Fig. 3.5a. The effects
of mechanical dispersion on the breakthrough curves illustrated in
Fig. 3.6a are shown in Fig. 3.6c. At relatively high seepage velocities,
mechanical dispersion dominates the mixing process and the solute
breakthrough curves, including the effects of mechanical dispersion,
intersect those for piston flow at a relative concentration, clco, of 0.5, as
shown in Fig. 3.6c.

3.4.4 Effect of diffusion

The breakthrough curves illustrated in Fig. 3.6c represent spreading of
the solute front primarily due to mechanical dispersion, i.e., spreading
due to diffusion is negligible. This is the case commonly referred to in
groundwater hydrology textbooks because the primary concern is with
contaminant migration in aquifers, i.e., coarse-grained, water-bearing
strata subject to relatively high seepage velocities. Curves 3 and 4 in
Fig. 3.6d represent breakthrough curves for nonreactive and reactive
solutes, respectively, when the seepage velocity is sufficiently low such
that the effect of diffusion is not masked by the effect of advection,
including mechanical dispersion. For comparison, the breakthrough
curves in Fig. 3.6c also are provided in Fig. 3.6d. The spreading effect
is still noticeable in Fig. 3.6d, but curves 1 and 3 and curves 2 and 4 do
not intersect each other at c/co = 0.5. Instead, curves 3 and 4 are
displaced to the left of curves 1 and 2, respectively. This displacement
due to diffusion can result in transit times (at c/c o = 0.5) which, in
some cases, can be much less than those predicted by considering only
advection. This effect previously has been recognized analytically by
De Weist (1965) and experimentally by Biggar and Nielsen (1960).
The decrease in the transit time, illustrated in Fig. 3.6d, due to
diffusion is a function of the magnitude of the seepage velocity. This
dependence is illustrated in Fig. 3.7 where the transit time for purely
advective, purely diffusive, and advective-dispersive transport is
illustrated as a function of the logarithm of the seepage velocity.
Although no data are provided in Fig. 3.7, the general trends are
apparent from previous analyses (Shackelford, 1988b, 1989). The
horizontal distance between the purely advective and the advective-
dispersive curve in Fig. 3.7 represents the offset distance in Fig. 3.6d
for a given seepage velocity. The independence of pure diffusion on
hydraulic conductivity is indicated by the vertical line. From the trends
presented in Fig. 3.7, it is apparent that the effect of diffusion on
52 Contaminant transport

~
'u
o
Qi
>~
Q)~
0l0l
~.Q
Q)~
,
pure
advection
(piston flow)

Q)

F
(f)

dispersion

o transit time
(a)

pure
advection
(piston flow)

advection- / ~ pure
dispersion diffusion

o transit time
(b)

Fig. 3.7 Schematic relationship between transit time and seepage velocity for
a transport distance, L, and a constant source concentration: (a) nonreactive
solutes; and (b) reactive solutes [after Shackelford (1988b, 1989, 1991)].

transit time increases as the seepage velocity decreases. The seepage

velocity at which diffusion begins to dominate solute transport is
indicated by the inflection point in the advective-dispersive curve.
Available evidence indicates that diffusion can be a significant
transport process when the seepage velocity is in the range 0.064-
0.09m/yr (Rowe et al., 1988), and a dominant transport process when
Vs is ~ O.OOSm/yr (Gillham et a/., 1984; Shackelford, 1988b, 1989).
The effect of diffusion on solute breakthrough curves is similar to the
effect of effective porosity (compare Figs. 3.6b and 3.6d). Whereas the
effect of diffusion is noticeable only at relatively low seepage velocities,
the effect of effective porosity is expected to be independent of the
seepage velocity. However, there is some indication that the value of ne
can be a function of the hydraulic gradient (and, therefore, v s ) imposed
in the study, with lower values of ne being observed at lower hydraulic
gradients (Liao and Daniel, 1989). Therefore, attempts to discern the
effective porosity of a soil, e.g., by performing column tests at elevated
seepage velocities, may not prove successful.
 Applications for one-dimensional problems 53
3.5 APPLICATIONS FOR ONE-DIMENSIONAL PROBLEMS

3.5.1 Waste disposal scenarios

Analytical solutions may be used, for example, to estimate both the
flux and the concentration of contaminants at the effluent end of a
column of porous material or at the bottom of a containment barrier.
The appropriate analytical solution to be used depends on the
susceptibility of the contaminant to geochemical reactions (i.e., non-
reactive versus reactive solutes) and the overall flow and boundary
conditions. In general, there are three possible scenarios or cases to
consider for containment systems in which mechanical dispersion is
negligible (Shackelford, 1989):
1. diffusion without advection (i.e., pure diffusion);
2. diffusion with positive advection; and,
3. diffusion with negative advection.
Each of these three cases is illustrated in Fig. 3.8.
For the case shown in Fig. 3.8a, the hydraulic gradient across a clay
barrier is zero, so there is no advective flow. However, since the
concentration of the contaminants in the leachate is greater than that in
the natural soil, a concentration gradient is established across the
barrier and contaminants will diffuse through the barrier. Whereas the
diffusive flux of contaminants is described by Eq. (3.7), the transient
diffusive transport of contaminants can be modelled using analytical
solutions to Eq. (3.35), i.e., analytical solutions to Eq. (3.16) with Vs
equal to zero.
For the case depicted in Fig. 3.8b, a hydraulic gradient has been
established across the barrier such that advective transport of con-
taminants occurs in the same direction as the diffusive transport. The
advective transport is termed 'positive' since it results in an increase in
the contaminant concentration in the natural soil (Shackelford, 1989).
This case commonly is described in texts on contaminant transport.
(Bear, 1972, 1979; Freeze and Cherry, 1979; Bear and Verruijt, 1987). For
this case, the total mass flux of the contaminant is given by Eq. (3.12)
where Oh equals 0* (i.e., Om = 0). The time-dependent concentration
of contaminants occurring at the barrier bottom, assuming a homo-
geneous soil, can be determined with analytical solutions to the
advective-dispersive equation, e.g., Eq. (3.16).
In some instances, the direction of advective transport may be
opposite to that of diffusive transport, as shown in Fig. 3.8c. This
situation may occur in practice, for example, when the barrier (either
natural or man-made) is located over a confined aquifer under artesian
pressure, or when vertical barriers, such as slurry walls, are used to
isolate a contaminated area (see Fig. 3.9). In this case, the advective
54 Contaminant transport
OlfeClion of Oire Cl lon of
Advective Flow lezometer Oll!u slon

1
None

I Natural Soli (c=O)1

(s)

Leachate (c :> 0)

1 I Natural Soli

(b)
(c=O)1 1
,
Leachate (c :> 0)

1 I Natura l Soli

(c )
(c-oll
1
Fig. 3.8 Waste disposal scenarios: (a) pure diffusion; (b) diffusion with
positive advection; and (c) diffusion with negative advection [after Shackelford
(1989)] .

flow is termed 'negative' since it works to prevent the escape of

pollutants frem the contaminated area (Shackelford, 1989). Gray and
Weber (1984) analyzed this case using Eq. (3.16) simply by substituting
a '-v/ for 'vs'. However, the boundary and flow conditions for this
case can be complex. For example, if the concentration, C2, in Fig. 3.9 is
initially zero, the initial flow of water into the containment area may
dilute the concentration C1 with time and, therefore, reduce the
concentration gradient for outward (positive) diffusive transport. The
contaminant migration front, if any, will depend on the offsetting
effects of the negative advective and positive diffusive fluxes. The
situation for reactive solutes is more complex. Regardless, significant
diffusive transport of contaminants may still result if a relatively thin
barrier is built to contain relatively high concentrations of pollutants
over extended periods (Gray and Weber, 1984).
 Applications for one-dimensional problems 55
slurry wall

ground surface

W.T.
W.T.

Idiffusive transpj@

c,

Fig. 3.9 Slurry wall scenario illustrating case for diffusion against advection
[after Gray and Weber (1984); and Shackelford (1989)].

3.5.2 Steady-state flux determination

The steady-state (long-term) flux of a contaminant through a barrier of
thickness L due to purely diffusive transport (i.e., Fig. 3.8a) is given by
the following form of Eq. (3.7).

Jo = _O*n fic = _o*nficl (3.39)

fix L ~x=L
Here the differentials in Eq. (3.7) have been replaced by difference
operators to indicate steady-state conditions. The implicit assumption
in the use of Eq. (3.39) is that the contaminant exiting the bottom of the
barrier is instantaneously and continuously flushed so that the
concentration gradient across the barrier is maintained at the steady-
state value, i.e., so that fie/fix 1= f(t).
For example, consider the clay barrier scenario shown in Fig. 3.lOa.
The clay barrier, with a thickness L (=1 m) and porosity n (=0.4), is
used to contain a leachate with a constant concentration, Co (= 10 mg/L),
of a chemical species (e.g., Cl-). The barrier is underlain by a leachate
collection system for collection and removal of leachate which migrates
through the barrier. For the case of pure diffusion, the flow rate of the
leachate through the barrier is considered to be negligible so that
diffusion is the dominant mechanism of contaminant transport through
the barrier (see Fig. 3.7). Under these conditions, the contaminant
concentration profiles in the barrier as a function of time will resemble
those shown in Fig. 3.10b (Quigley et ai., 1987; Daniel and Shackelford,
56 Contaminant transport

Concentration Concentratio n
Leachate
0 c
0
0 c0
0 0

.~ -!:.
:.
In In

'0 '0
L (J) (J)

.5 .5
.r::. .r::.
.,
Q. .,
Q.
0 0
x=L x=L
CL

Flush I > I > I > I > I I > I

5 4 3 2 I 5

(a) (b) (e)

Fig. 3.10 Steady-state diffusive flux scenario: (a) barrier configuration; (b)
concentration profiles as a function of time; and (c) establishment of steady-
state concentration gradient across barrier.

1988). For example, in Fig. 3.10b, the contaminant species reaches the
bottom of the barrier after an elapsed time t3 , and the concentration of
the contaminant species in the leachate collection system continues to
rise as a function of time. If none of the contaminant is removed from
the leachate collection system, the concentration at the bottom of the
barrier will eventually reach the level, Co, at which time migration via
diffusion will stop since the concentration gradient across the barrier
will be zero (i.e., ~c = Co - Co = 0). However, if the leachate collection
system is continuously flushed so as to maintain a constant concentra-
tion, CL (=2mg/L), in the leachate collection system, as shown in Fig.
3.lOc, then a steady-state concentration gradient eventually will be
established across the barrier (e.g., after time ts in Fig. 3.10c), and the
steady-state diffusive flux can be determined using Eq. (3.39).

JD-- -
0* nCL--- Co
-
L

;J
( 2mg m-
--10 g)
= - (1 X10-9~2) (0.4) LIm L (1000 ~3) (31.536 X106
= 101 ( ~g )
m yr
(3.40)

Here a value of 1.0 x 10- 9 m 2 /s has been used for the effective dif-
fusion coefficient, 0*, for chloride, a nonreactive chemical species
 Applications for one-dimensional problems 57

(Shackelford, 1988a). An additional example of the use of Eq. (3.39) to

estimate the flux of organic contaminants at the bottom of a natural
clayey barrier is given by Johnson et al. (1989).
The situation for the case of diffusion with 'positive' advection (Fig.
3.8b) is not as straightforward as that for the case of pure diffusive
transport. When the concentration of contaminant species in the leachate
is assumed to be constant (e.g., co), breakthrough of the contaminant
species through the bottom of the barrier will resemble the solute
breakthrough curve shown in Fig. 3.5a, and the concentration profiles
within the barrier as a function of time will resemble those shown in
Fig. 3.5b. In accordance with Fig. 3.5a, steady-state transport con-
ditions are established only after an elapsed time tf. At this time, the
concentration of the specified chemical species at the barrier bottom
reaches the value Co corresponding to the concentration in the leachate
(i.e., clco = I), and the concentration gradient across the barrier is zero
(i.e., Ac = Co - Co = 0). Therefore, there is no diffusive flux after time
tf , and the steady-state flux of the contaminant is due solely to the
advective component of flux given by Eq. (3.4). Even if an attempt is
made to maintain a reduced concentration of the contaminant species
in the leachate collection system by flushing, steady-state diffusive flux
conditions still will not be established because time-dependent con-
centration profiles (i.e., c = f(x, t)) within the barrier still exist due to
the advective component of transport. Due to these time-dependent
concentration profiles, the differential concentration gradient in Eq.
(3.7) cannot be replaced by the difference concentration gradient in Eq.
(3.39) (i.e., oclox"* AclAx) and, therefore, steady-state diffusive flux
conditions cannot exist (i.e., since aclax = f(t)).

3.5.3 Analytical solutions

Mathematical solutions to the transport equations for the complex
geometries, heterogeneous transport parameters, and realistic boundary
conditions characteristic of many practical problems typically require
the application of numerical techniques (e.g., finite differences or
finite elements). However, the application of numerical methods to the
solution of practical contaminant transport problems is beyond the
scope of this chapter. An excellent review of the application of
numerical techniques to the solution of groundwater problems can be
found in Huyakorn and Pinder (1983).
Although restricted in use, analytical solutions can be beneficial
when utilized for preliminary design, rudimentary evaluation of
remedial action, and/or verification of more sophisticated numerical
solutions. Analytical solutions to the various forms of the one-
dimensional advection-dispersion equation (i.e., (3.16), (3.18), and
(3.31)) can be found in Lapidus and Amundson (1952), Bastian and
58 Contaminant transport

Lapidus (1956), Ogata and Banks (1961), Brenner (1962), Lindstrom et

al. (1967), Gershon and Nir (1969), Ogata (1970), Bear (1972, 1979),
Selim and Mansell (1976), Van Genuchten (1981), Van Genuchten and
Alves (1982), and Javendel et al. (1984). Semi-analytical solutions are
provided by Javendel et al. (1984) and Rowe and Booker (1985). Three
analytical solutions to Eq. (3.31) are presented in Table 3.1 (Van
Genuchten, 1981; Van Genuchten and Alves, 1982). The second
analytical solution in case A is required for the situation in which the
decay constant, A, is zero since division by zero in the first solution is
mathematically impossible.

Table 3.1 Analytical solutions to the one-dimensional advection-dispersion

equation with simultaneous reversible sorption and linear, first-order decay
reactions (after Van Genuchten, 1981; and Van Genuchten and Alves, 1982)

Case A: Initial and Boundary Conditions

Initial condition: c =0 (x;;:: 0; t = 0)
1st boundary condition: (- D" ac + Vsc) I = VsCo I
ax x~o' x~o (x = 0; t > 0)
2nd boundary condition: ac = 0
ax (x = 00; t > 0)

Solution 1: (Rd ;;:: 1; A > 0)*

c(x,-t)= ( -Vs-) exp [(V s - U)x] erc
- f [RdX - Ut]
Co Vs + U 2D" 2YD"Rdt
+ ( -V- s ) exp [(V s + U)X] erc
f [Kt Ut]
X+
Vs - U 2D" 2YD"Rdt
ifs ) (RdX + Ut)
+ (2Dh RdA exp Dh -
(VsX )
At erfc 2YD Rdt
h

Solution 2: (Rd ;;:: 1; A = 0)

Case B: Initial and Boundary Conditions

Initial condition c= 0 (x;;:: 0; t = 0)
1st boundary condition: c = Co (x = 0; t > 0)
2nd boundary condition: ~~ = 0 (x = 00; t> 0)
Solution: (Rd ;;:: 1; A ;;:: 0)*
t) = -1 exp [(V s - U)X] er f c [Kt x - Ut] + -1 exp [(V s + U)x] er fc [RdX + Ut]
c(x,-
-
Co 2 2D" 2YD"Rd t 2 2D" 2VDhRdt
I

* U = (v; + 4Dh R,/A)2 [Note: when Ie = 0, U = v,l

 Applications for one-dimensional problems 59
Both cases in Table 3.1 correspond to a constant source concentration
for semi-infinite porous media. However, the first boundary condition
for each of the two cases is different. In case A, the first boundary
condition represents the conservation of mass required at the liquid-
soil interface. The first boundary condition for case B represents an
approximation of the first boundary condition for case A. In reality, the
first boundary condition for case B is impossible in that it represents a
discontinuity in the concentration at the liquid-soil interface (Kreft
and Zuber, 1978; Parker and Van Genuchten, 1984). However,
Gershon and Nir (1969) compared breakthrough curves based on one-
dimensional analytical solutions derived from the initial and boundary
conditions represented by cases A and B and found that the approxi-
mate boundary condition in case B resulted in differences of up
to only 5% in the region of cleo = 0.5. Therefore, the analytical sol-
ution represented by case B in Table 3.1 should be sufficient for most
practical purposes (i.e., in lieu of the analytical solution presented
for case A).
The 'erfc' appearing in the analytical solutions in Table 3.1 stands for
the 'complementary error function' which, for any argument z, is given
by
erfc (z) = 1 - erf (z) (3.41)
and erf (z) is the 'error function' of the argument, z. Values of erf(z)
and erfc(z) are tabulated in several texts (Carslaw and Jaeger, 1959;
Abramowitz and Stegun, 1972; Crank, 1975; Freeze and Cherry, 1979).
As an alternative, values of erf(z) may be determined with the fol-
lowing series solution (Carslaw and Jaeger, 1959; Abramowitz and
Stegun, 1972), or
2 (_1)I1I Z2111+1
2:
00

erf (z) = - (3.42)

Vn 111=0 m! (2m + 1)
A small computer program (~100 lines) can be written to perform the
series summation. Experience indicates that fifty terms (m = 50) are
required in Eq. (3.42) to achieve an accuracy to six decimal places for
values of erf(z) (Shackelford, 1990). For situations in which the quantity
VsX » Dh and/or t is small, the series summation represented by
Eq. (3.42) may not be appropriate for convergence of the analytical
solutions in Table 3.1 (Brenner, 1962; Cadena, 1989). In such cases,
other approximations can be used for evaluation of erf(z) (Abramowitz
and Stegun, 1972).
The analytical solutions in Table 3.1 may be used to derive many of
the analytical solutions given in other references. For example, in the
absence of decay (i.e., A = 0), the analytical solution given by case B
reduces to the well-known Ogata and Banks (1961) solution (also see
Lapidus and Amundson, 1952; Ogata, 1970; Bear, 1972, 1979; Freeze
60 Contaminant transport

and Cherry, 1979; Van Genuchten, 1981; and Van Genuchten and
Alves, 1982), or

c(x, t) = ! [erfC (Rd X- Vst) + exp (VsX) erfc (Rd X+ Vst)] (3.43)
Co 2 2YR d Dh t Dh 2YRd Dh t

Dimensionless parameters
The arguments for analytical solutions often can be expressed for con-
venience in terms of dimensionless parameters. For example, Eq. (3.43)
can be expressed as

c(X,t)
--;;;- =
1{ erfc (l-TR)
2:
(l+TR)}
2YT )P + exp (Pd erfc 2YT RIP
1 L L
(3.44)

where

(3.45)

and
_
P L- VsX -_ VsLI (3.46)
Dh Dh x=L

Equation (3.45) represents the number of pore volumes, T, divided by

the retardation factor, Rd , whereas PL is the Peclet number defined
with respect to the barrier thickness, L (Bear, 1972). Equation (3.44) can
be presented graphically as a plot of clco versus T R for various values
of PL (Ogata and Banks, 1961; Ogata, 1970; Goldman et al., 1986;
Shackelford, 1990).

3.5.4 Transit time determinations

Several performance criteria can be used to provide a definition for
transit time for the design of waste containment barriers (Goldman
et al., 1986). Two specific transit-time analyses are considered here.
First, the time required for the concentration of a specific leachate
component to reach a specific value at the barrier bottom is illustrated.
Second, the time required to reach a specified flux of a particular
contaminant at the barrier bottom is considered.

Specified leachate concentration

Equation (3.43) can be used directly to determine the concentration, c,
of a specified leachate constituent at the bottom of a clay barrier of
thickness, L, for any given time t. For given values of c, co, v s , Dh , and
L (=x), all variables in Eq. (3.43) are defined except for the time, t. An
 Case histories 61

iterative procedure must be used with Eq. (3.43) to solve for t.

Shackelford (1990) illustrates the iterative procedure using Eq. (3.44) to
estimate the barrier thickness required to provide the necessary transit
time for the concentration of a specified leachate constituent. The
iterative procedure involves the following steps:
1. assume a liner thickness, L;
2. calculate P L from Eq. (3.46);
3. determine T R from a graph of T R versus clco for the desired value of
clco and the calculated value of PL; and
4. use Eq. (3.45) to determine the transit time, t.
This procedure is repeated until t is greater than or equal to the design
life of the containment barrier.

Specified leachate flux

A similar approach to the one presented above for a specified leachate
concentration can be developed for a specified leachate flux by using
the analytical solution for the concentration, c(x, t), in the determina-
tion of the advective, diffusive, and dispersive fluxes using Eqs. (3.4),
(3.7), and (3.9). For example, if Eq. (3.44) is used to represent the
solute concentration, then the total flux, J, of the solute at a distance L
and any time t can be represented by a dimensionless flux number, FNf
as follows (Shackelford, 1990):
JL 1
FN = - 0 = -2{P LQl + Q2} (3.47)
nco h
where

1 - TR ) Q2
2 exp[ -( 1 - TR
2vt;;JIiZ
fJ (3.48)
erfc ( 2VTRIP ;
=
Ql =
~n~R
L

A graphical solution to Eq. (3.47) is presented by Shackelford (1990) for

the case where Oh = 0*. The graphical solution of Eq. (3.47) may be
used to estimate the barrier thickness required to provide the necessary
transit time for a specified contaminant flux by substituting FN for clco
in the procedure outlined for the specified leachate concentration.

3.6 CASE HISTORIES

Several case histories illustrating the migration of contaminants

beneath landfills have appeared in the literature within the past 15
years or so. For example, Goodall and Quigley (1977) and Quigley et al.
62 Contaminant transport

(1984) evaluated measured concentration profiles below the Confedera-

tion Road sanitary landfill site near Sarnia, Ontario, using Eq. (3.43).
They concluded that diffusion is an important transport mechanism of
solutes in landfill leachate under low-flow situations.
Johnson et al. (1989) evaluated the migration of chloride (Cl-) and
organic compounds beneath a hazardous waste disposal site located
15 km southwest of Sarnia, Ontario. The results indicated that the
influence of advection was negligible and that diffusion was the
dominant mechanism of transport in the natural clay. Johnson et al.
(1989) contend that the study has important implications with regard to
the design of relatively thin (1 m thick) clay liners, or even double
liners, with low hydraulic conductivities. They concluded that dif-
fusion of contaminants through clay liners can result in relatively
short breakthrough times, and the resulting fluxes may be significant
especially for the common priority pollutants.
An excellent and detailed case study describing the effect of
longitudinal dispersion on the migration of contaminants from beneath
an abandoned landfill in Ontario, Canada, is presented in a series of
seven papers in a special issue of the Journal of Hydrology (Cherry,
1983). Landfill-derived contamination was delineated by monitoring
elevated distributions of chloride, sulfate, and electrical conductance in
an underlying sand aquifer. The extent of the contaminant plume
resulting from migration of the landfill leachate was detected to be of
the order of 600 m in width, 700 m in length, and 20 m in depth
beneath the landfill. Irregular concentration profiles beneath and near
the landfill which gradually became smoother downgradient from
the landfill, where maximum concentrations were much lower, were
attributed to the strong influence of longitudinal dispersion in the sand
aquifer.

REFERENCES

Abramowitz, M. and Stegun, LA. (1972) Handbook of Mathematical Functions.

Dover Publishers, Inc., New York, N.Y.
Anderson, M.P. (1979) Using models to simulate the movement of con-
taminants through groundwater flow systems, Crit. Rev. of the Environ.
Controls, eRe, 9(2), 97-156.
Anderson, M.P. (1984) Movement of contaminants in groundwater: ground-
water transport - advection and dispersion, in Groundwater Contamination,
Studies in Geophysics, National Academy Press, Washington, D.C.,
pp.37-45.
Bastian, W.c. and Lapidus, L. (1956) Longitudinal diffusion in ion exchange
and chromatographic columns. Finite column, J. Phys. Chem., 60, 816-7.
Bear, J. (1972) Dynamics of Fluids in Porous Media. American Elsevier Publishing,
Inc., New York, N.Y.
 References 63

Bear, J. (1979) Hydraulics of Groundwater. McGraw-Hill Book Co., New York,

N.Y.
Bear, J. and Verruijt, A. (1987) Theory and Applications of Transport In Porous
Media, D. Reidel Pub!. Co., Dordrecht, Holland.
Biggar, J.W. and Nielson, D.R. (1960) Diffusion effects in miscible displacement
occurring in saturated and unsaturated porous materials, 1- Geophys. Res.
65(9), 2887-95.
Bowders, J.J., Daniel, D.E., Broderick, G.P. and Liljestrand, H.M. (1985)
Methods for testing the compatibility of clay liners with landfill leachate,
Haz. and Solid Waste Testing, in Proc. 4th Symp., STP 886, Amer. Soc. for
Testing and Materials, Philadelphia, PA.
Brenner, H. (1962) The diffusion of longitudinal mixing in beds of finite length:
numerical values, Chem. Eng. Sc., 17:229-43.
Carslaw, H.S. and Jaeger, J.C (1959) Conduction of Heat in Solids, 2nd edn.
Oxford University Press, Oxford, England.
Cadena, F. (1989) Numerical approach to solution of pollutant transport
models using personal computers, Computers in Education, Division of
ASEE, 9(2), 34-6.
Cherry, J.A. (guest ed) (1983) Migration of contaminants in groundwater at a
landfill: a case study, 1- of Hydrology, 63, 1-192.
Crank, J. (1975) The Mathematics of Diffusion, 2nd edn. Oxford University Press,
London, England.
Daniel, D.E. and Shackelford, CD. (1988) Disposal barriers that release
contaminants only by molecular diffusion, Nuclear and Chemical Waste
Management, 8, 299-305.
Davidson, J.M., Ou, L.-T. and Rao, P.S.C (1976) Behavior of high pesticide
concentrations in soil water systems, Residual management by land
disposal, Proc., Haz. Waste Res. Symp., USEPA, Cincinatti, Ohio, EPA-6001
9-76-015, 206-12.
De Weist, R.J.M. (1965) Geohydrology, John Wiley and Sons, Inc., New York,
N.Y.
Freeze, R.A. and Cherry, J.A. (1979) Groundwater, Prentice-Hall, Inc.,
Englewood Cliffs, N.J.
Fried, J.J. (1975) Groundwater Pollution, Elsevier Scientific Publishers,
Amsterdam.
Gershon, N.D. and Nir, A. (1969) Effects of boundary conditions of models on
tracer distribution in flow through porous mediums, Water Resources Res.,
5(4), 830-9.
Gillham, R.W., Robin, M.J.L., Dytynyshyn, D.J. and Johnston, H.M. (1984)
Diffusion of nonreactive and reactive solutes through fine-grained barrier
materials, Canadian Geotech. J., 21, 541-50.
Goldman, L.J., Truesdale, R.S., Kingsbury, G.L., Northeim, CM. and Damle,
A.S. (1986) Design, construction, and evaluation of clay liners for waste
management facilities, EPA Draft Technical Resource Document, Hazardous
Waste Engineering Laboratory, Washington, DC, EPA/530-SW-86-007
[NTIS PB86-184496].
Goodall, D.C and Quigley, R.M. (1977) Pollutant migration from two sanitary
landfill sites near Sarnia, Ontario, Canadian Geotech. J., 14, 223-36.
Gray, D.H. and Weber, W.J., Jr. (1984) Diffusional transport of hazardous
waste leachate across clay barriers, Proc., 7th Ann. Madison Waste Conj.,
Sept 11-12, Univ. of Wisconsin, Madison, 373-89.
Greenberg, J.A., Mitchell, J.K. and Witherspoon, P.A. (1973) Coupled salt and
water flows in a groundwater basin, 1- Geophys. Res., 78(27), 6341-53.
64 Contaminant transport
Griffin, RA. and Roy, W.R (1985) Interaction of organic solvents with saturated
soil-water systems, Environmental Institute for Waste Management Studies,
University of Alabama, Tuscaloosa, Alabama.
Horton, R, Thompson, M.L. and McBride, J.F. (1985) Estimating transit times
of noninteracting pollutants through compacted soil materials, Proc., 11th
Ann. Solid Waste Res. Symp., Apr 29-May I, Cincinnati, Ohio, EPA/600/9-
85/013, 275-82.
Huyakorn, P.S. and Pinder, G.F. (1983) Computational Methods in Subsurface
Flow, Academic Press, Inc., Orlando, Florida.
Javendel, I., Doughty, e. and Tsang, e.-F. (1984) Groundwater Transport:
Handbook of Mathematical Mode/s. Water Resources Monograph Series 10,
American Geophysical Union, Washington, De.
Johnson, RL., Cherry, J.A. and Pankow, J.F. (1989) Diffusive contaminant
transport in natural clay: a field example and implications for clay-lined
waste disposal sites, Environ. Sc. and Tech., 23, 340-9.
Karichoff, S.W., Brown, D.S. and Scott, T.A. (1979) Sorption of hydrophobic
pollutants on natural sediments, Water Research, 13, 241-8.
Kreft, A. and Zuber, A. (1978) On the physical meaning of the dispersion
equation and its solutions for different initial and boundary conditions,
Chem. Eng. Sc., 33, 1471-80.
Lapidus, L. and Amundson, N.R. (1952) Mathematics of adsorption in beds.
VI. The effect of longitudinal diffusion in ion exchange and
chromatographic columns, J. Phys. Chem., 56, 984-8.
Li, Y-H. and Gregory, S. (1974) Diffusion of ions in sea water and in deep-sea
sediments, Geochimica et Cosmochimica Acta, 38, 703-14.
Liao, W.P. and Daniel, D.E. (1989) Time of travel of contaminants through soil
liners, Proc., 12th Ann. Madison Waste Conf., Sept 20-21, Univ. of
Wisconsin, Madison, 367-76.
Lindstrom, F.T., Haque, R., Freed V.H. and Boersma, L. (1967) Theory on the
movement of some herbicides in soils, linear diffusion and convection of
chemicals in soils, Environ. Sc. and Tech., 1(7), 561-5.
Melnyk, T.W. (1985) Effects of sorption behavior on contaminant migration, Atomic
Energy of Canada Limited (AECL-8390), Whiteshell Nuclear Research
Establishment, Pinawa, Manitoba ROE 1LO.
Mitchell, J.K. (1976) Fundamentals of Soil Behavior, John Wiley and Sons, Inc.,
New York, N.Y.
Ogata, A. (1970) Theory of dispersion in granular medium, US Geol. Surv. Prof.
Paper 411-1.
Ogata, A. and Banks, R.B. (1961) A solution of the differential equation of
longitudinal dispersion in porous media, US Geol. Surv. Prof. Paper 411-A.
Olsen, H.W. (1969) Simultaneous fluxes of liquid and charge in saturated
kaolinite, Proc., Soil Sc. Soc. of Amer., 33(3), 338-44.
Parker, J.e. and Van Genuchten, M. Th. (1984) Flux-averaged and volume-
averaged concentrations in continuum approaches to solute transport,
Water Resources Res., 20(7), 866-72.
Perkins, T.K. and Johnston, O.e. (1963) A review of diffusion and dispersion
in porous media. J. of Soc. of Petrol. Engin., 3, 70-83.
Peyton, G.R., Gibb, J.P., LeFaivre, M.H., Ritchey, J.D., Burch, S.L. and
Barcelona, M.J. (1986) Effective porosity of geologic materials, Land
disposal, remedial action, incineration, and treatment of hazardous waste,
Proc., 12th Ann. Res. Symp., Cincinnati, Ohio, EPA/600/9-86/022, 21-8.
Pickens, J.F. and Grisak, G.E. (1981) Scale-dependent dispersion in a stratified
granular aquifer, Water Resources Res., 17(4), 1151-211.
 References 65

Quigley, R.M., Crooks, V.E. and Fernandez, F. (1984) Engineered clay liners
(an overview report). Preprints of the Seminar on the Design and
Construction of Municipal and Industrial Waste Disposal Facilities, June
6-7, Toronto, Ontario, Canada, 115-34.
Quigley, R.M., Yanful, E.K. and Fernandez, F. (1987) Ion transfer by diffusion
through clayey barriers, Geotech. Practice for Waste Disposal, ASCE special
publication No. 13, (ed. R.D. Woods), 137-58.
Robinson, R.A. and Stokes, R.H. (1959) Electrolyte Solutions, 2nd edn.
Butterworths Scientific Publications, London, England.
Rowe, R.K. (1987) Pollutant transport through barriers, Geotech. Practice for
Waste Disposal, ASCE special publication No. 13, (ed. R.D. Woods), 159-
81.
Rowe, R.K. and Booker, J .R. (1985) 1-D pollutant migration in soils of finite
depth, J. Geotech. Eng., ASCE, 111(4), 479-99.
Rowe, R.K., Caers, CJ. and Barone, F. (1988) Laboratory determination of
diffusion and distribution coefficients of contaminants using undisturbed
clayey soil, Canadian Geotech. J., 25, 108-18.
Selim, H.M. and Mansell, R.S. (1976) Analytical solution of the equation for
transport of reactive solutes through soils, Water Resources Res., 12(3),
528-32.
Shackelford, CD. (1988a) Diffusion of inorganic chemical wastes in compacted
clay, Ph.D. Dissertation, University of Texas, Austin, TX.
Shackelford, CD. (1988b) Diffusion as a transport process in fine-grained
barrier materials, Geotech. News, 6(2), 24-7.
Shackelford, CD. (1989) Diffusion of contaminants through waste containment
barriers, Transportation Research Record 1219, Transportation Research
Board, National Research Council, Washington, DC, 169-82.
Shackelford, CD. (1990) Transit-time design of earthen barriers, Eng. Geol.,
Elsevier Pub!., Amsterdam, 29, 79-94.
Shackelford, CD. (1991) Laboratory diffusion testing for waste disposal-A
review, J. Contaminant Hydrology, Elsevier Pub!., Amsterdam, 7(3), 177-
217.
Shackelford, CD. and Daniel, D.E. (1991) Diffusion in saturated soil:
I. Background, ASCE J. Geotech. Eng., 117(3),467-84.
Van Genuchten, M. Th. (1981) Analytical solutions for chemical transport with
simultaneous adsorption, zero-order production, and first-order decay,
J. Hydrology, 49:213-33.
Van Genuchten, M. Th. and Alves, W.J. (1982) Analytical solutions of the
one-dimensional convective-dispersive solute transport equation, US
Department of Agriculture, Technical Bulletin No. 1661.
Weber, W.J., Jr. and Smith, E.H. (1987) Simulation and design models for
adsorption processes, Environ. Sc. and Tech., 21(11), 1040-50.
 CHAPTER 4

Hydrogeology
Keros Cartwright and Bruce R. Hensel

4.1 INTRODUCTION

The basic theory and concepts of hydrogeology will be discussed in

this chapter. Sections 4.2-4 will deal primarily with the definitions and
equations of ground-water flow in the saturated and unsaturated zone.
Section 4.5 will examine the geology of hydrogeology, and is especially
important because the movement of fluid through the subsurface and
the equations that describe this movement are dictated by the geology
of the subsurface materials. Geologic characteristics such as perme-
ability, porosity, anisotropy, and homogeneity, are functions not only
of the type of rock or soil material, but also of the depositional en-
vironment, diagenesis, and tectonic processes such as folding and
faulting that may have occurred after deposition. This information
is needed to build a geologic framework in which hydrogeologic prin-
ciples may be applied before ground-water flow and contaminant
transport can be evaluated.

4.2 BASIC PRINCIPLES

4.2.1 Definition of geologic materials

Two basic types of geologic materials will be discussed in this chapter.
Soil is used in the engineering sense to describe nonlithified and/or
unconsolidated sand, silt, and/or clay deposits that overlay the bedrock
surface. Rock is used to describe all igneous, metamorphic, and
lithified sedimentary rocks. When these materials have like permeability
and porosity characteristics, they make-up a hydro stratigraphic unit
that will either allow or be a barrier to ground-water flow.

Geotechnical Practice for Waste Disposal.

Edited by David E. Daniel.
Published in 1993 by Chapman & Hall, London. ISBN 0 412 35170 6
 Basic principles 67
4.2.2 Definition of ground water
Ground water is defined as water that occurs in the voids of saturated
rock and soil material (Freeze and Cherry, 1979; Todd, 1980). Water
found in unsaturated materials is defined as vadose water (Todd,
1980). There is no distinction between the two types of water other
than occurrence, and water can move from the unsaturated to saturated
zone and vice versa. Therefore, for the purpose of this chapter, the
term ground water will be used to identify all water that occurs in
voids beneath the ground surface.

4.2.3 Porosity
The voids in which ground water are found can have many shapes and
forms (Fig. 4.1). The primary porosity of a material is the ratio of the
volume of intergranular pore space relative to the total volume of the
material. This value may be quite high, 45-50% in clays, or quite low,
0-5% in crystalline rocks. This type of porosity is often referred to as
intergranular porosity. The secondary porosity of a material refers to
the porosity formed subsequent to deposition. Types of secondary
porosity are fractures, weathered rock or soil zones, solution openings,

(a) (e) (e)

,.---,.-----r---,

(b) (d)

Fig. 4.1 Types of porosity commonly found in geologie materials. A-D are
examples of primary porosity, the primary porosity of case D has been
diminished by digenetic processes, cases E and F are examples of secondary
porosity, from Meinzer (1923). Diagram showing several types of rock
interstices and the relation of rock texture to porosity: A, Well-sorted
sedimentary deposit having high porosity; B, poorly sorted sedimentary
deposit having low porosity; C, well-sorted sedimentary deposit consisting of
pebbles that are themselves porous, so that the deposit as a whole has a very
high porosity; D, well-sorted sedimentary deposit whose porosity has been
diminished by the deposition of mineral matter in the interstices; E, rock
rendered porous by solution; F, rock rendered porous by fracturing.
68 Hydrogeology
and macropores caused by animal burrowing, root action, and
dehydration of soils. This type of porosity is sometimes referred to as
fracture porosity. The total porosity of a soil or rock is the sum of the
primary and secondary porosity. Table 4.1 lists porosity characteristics
for flow through common geologic materials.
An important concept in the evaluation of contaminant transport is
that of effective porosity. Whereas total porosity is the ratio of void
volume to sample volume in a rock or soil material, effective porosity is
the percentage of void volume that is interconnected, allowing a fluid
to pass through the material. The effective porosity of a soil or rock
material must, by definition, be equal to or less than the total porosity
of the material. In general, the effective porosity of coarse-grained
materials such as sand approaches the total porosity. The effective
porosity of fine-grained materials such as clay may be 10-20% of total
porosity (Horton, Thompson, and McBride, 1988).
A term related to porosity that is important in the analysis of
unsaturated flow is moisture content. The moisture content, defined
here as a volumetric term, is the ratio of water filled pores to total
sample volume. Another related term, degree of saturation, or

Table 4.1 Porosity characteristics of ground-water flow through various

geologic material

type of porosity geologic characteristic

sedimentary igneous and

metamorphic
lithified unlithified

intergranular gravel
sand
silt
loess
peat
outwash
intergranular weather limestone/ clay-marine/ tuff
and fractures dolomite lacustrine weathered basaltic
oolitic limestone till rocks
chalk weathered granitic
sandstone rocks
shale
coal
fracture limestone basaltic rocks
dolomite granitic rocks
marble quartzite
gneiss
 Basic principles 69
moisture content, is the ratio of water filled pores to total porosity. For
a saturated material, the moisture content will be equal to the total
porosity and the degree of saturation will be equal to 1.0.

4.2.4 Hydraulic head

Freeze and Cherry (1979) noted that ground-water flow, in its simplest
form, is a mechanical process. When water moves through a porous
material, mechanical energy is transformed to thermal energy as the
water molecules encounter the frictional resistance of the matrix.
Therefore, the direction of flow must be away from areas where the
mechanical energy per unit mass of a fluid is high, and towards areas
where mechanical energy is low. This energy concept is the basis for
the definition of hydraulic head, in that the fluid potential, or head, is
the mechanical energy per unit mass of fluid.
Hydraulic head has three components: the elevation head, the
pressure head, and the velocity head (Freeze and Cherry, 1979). In
most situations, ground-water velocity is low and the velocity head is
negligible compared to pressure and elevation head. The elevation
head is a measure of the potential energy of the water. It describes the
potential for downward movement of water in response to gravity.
Pressure head describes the fluid pressure and is measured as the
height a column of water at the measurement point will rise or fall to in
a manometer (Fig. 4.2).

4.2.5 Saturated and unsaturated materials

When a rock or soil material is saturated, the pressure head will be
zero or a positive value. All of the void spaces in a saturated material
are usually filled with water. Ground water in saturated materials will
flow into an opening, such as a well, under conditions of normal
atmospheric pressure.
In an unsaturated or tension-saturated material the pressure head
will have a negative value (Fig. 4.2). Tension, or suction, is the
absolute value of pressure head in unsaturated material. Void spaces
in an unsaturated material will often be filled with both water and
gasses. Ground water in such materials will not flow into an opening
unless a pressure lower than the pressure head is applied to the
opening. Restating the last sentence in soil physics terms, soil water
will not flow into an opening unless a suction greater than the tension
is applied to the opening.
In many soils, particularly in fine grained soils, a capillary zone may
exist where most or all of the pore spaces are saturated with water
even though the pressure head is negative. In both the unsaturated
70 Hydrogeology
round surface

~l.c..

ab!
______ _ unsalural d
salur l d

r------
. ...... ... ...........

dalum

Fig. 4.2 Schematic drawing illustrating head in the saturated and unsaturated
zone. Relative head value is indicated by length of line, positive values are
indicated by upward pointing arrow, negative values by downward pointing
arrows.

and the capillary zones, ground water does not flow to an opening at
atmospheric pressure because the capillary forces hold the water
molecules to the soil particles.

4.2.6 Conservation of mass

Ground water is subject to the principle of conservation of mass. That
is, water is not created or consumed in the subsurface (except as a
byproduct of certain chemical reactions). Therefore for any volume of
soil or rock material
(4.1)

Where Vin is the volume of water that enters the system, Vout is the
volume of water that exits the system and S is the change in the
volume of water stored in the system. This is a simplified version of
the continuity equation. The concept of continuity is the basis of the
ground-water flow equation that describes the value of head at any
point within a three-dimensional flow field.
 Basic principles 71

4.2.7 Hydraulic conductivity and intrinsic permeability

The hydraulic conductivity of a rock or soil material describes the ease
with which a particular fluid, usually water, may flow through that
material. Hydraulic conductivity is neither a rock or fluid property, it is
dependent on the density and viscosity of the fluid as well as the
intrinsic permeability of the rock or soil material. The intrinsic perme-
ability of a material is dependent upon its grain size, packing, shape,
and pore distribution; thus it is a property related to the soil or rock
material. Freeze and Cherry (1979, p. 26-9) present a comprehensive
discussion on this topic. The terms hydraulic conductivity and perme-
ability are often interchanged, with intrinsic permeability used to
describe the material property that affects fluid movement. In this
chapter, hydraulic conductivity will be used to describe the movement
of water in the subsurface and permeability will hereafter be used in
place of intrinsic permeability.

4.2.8 Infiltration and recharge

Almost all ground water originates as surface water (Todd, 1980).
Infiltration describes the process of water entering the soil at the
ground surface. The source of this water may be rainfall, melting
snow, seepage from surface water bodies, or irrigation water. Recharge
describes the process of water entering the saturated zone. Figure 4.3
illustrates the difference between recharge and infiltration. Discharge
occurs when ground water exits the flow system as baseflow, evapo-
transpiration, springs, or through wells. Most discharge is baseflow to
surface water bodies (Todd, 1980) such as oceans, lakes, and rivers.
Before water can be made available for recharge, it must infiltrate
into the soil. Freeze and Cherry (1979) define infiltration as 'entry into
the soil of water made available at the ground surface, together with
associated flow away from the ground surface within the unsaturated
zone'. Not all water made available at the ground surface infiltrates
and not all water which infiltrates into the soil will recharge ground
water. Water which infiltrates may take several courses: it may be
drawn back to the surface by evaporation or through plant roots as
transpiration; it may redistribute through the unsaturated soil column
replenishing soil-moisture deficits without recharging ground water;
for certain geologic and topographic conditions it may flow to a nearby
surface water body as subsurface storm flow; or it may seep to the
saturated zone as recharge water. Infiltration will occur for almost any
event where water is applied to an unsaturated ground surface; but
only water that is not returned to the atmosphere as evapotranspira-
tion, not discharged to surface water bodies through the unsaturated
zone, and is not used to replenish soil-moisture deficits is available for
recharge.
72 Hydrogeology

precipitation

j j
-.... evapotranspiration
&
.........
& - ----- runoff
~

infiltration

Fig. 4.3 Infiltration and recharge. Infiltration is water that percolates into the
ground. Recharge is water that reaches and flows downward from the water
table. Only that portion of infiltrating water that is not used to replenish
soil-moisture deficits and is not discharged back to the atmosphere through
evapotranspiration is available for recharge.

Ground-water recharge is defined as 'entry into the saturated zone of

water made available at the water table, together with the associated
flow (of ground water) away from the water table within the saturated
zone' (Freeze and Cherry, 1979). If the water table is at the ground
surface and flow is downward, then water made available at the
ground surface can recharge the ground-water flow system. However,
if flow is upwards from the water table, discharge, rather than recharge
will occur. Recharge has also been defined as the amount of water that
reaches an aquifer. Strictly speaking, this definition is incorrect. When
this definition of recharge is used, it should be termed as recharge to
an aquifer.
Numerous field studies have shown that, in mid-latitude regions,
ground water recharge is most likely in the spring, at which time snow
melts, precipitation is high and evapotranspiration is low; so that
the water table is high and antecedent soil moisture is high (Walton,
1965; Rehm, Groenewold and Peterson, 1982; Steenhuis et al., 1985).
Steenhuis et al. (1985) measured recharge on Long Island and estimated
that most (75-90%) precipitation from mid-October to mid-May
becomes recharge, while very little precipitation occurring during late
spring, summer and early fall infiltrates to the water table to replenish
 Basic principles 73
9 .
8
/o~ /
POTENTIAL
EVAPOTRANSPIRATION
7

! ~-j.-"-.
/0
2.--
3 /
\.-\.~._._. ",c",,,,,,,,
I
1i /
00 -=7
0

I
'-0
o~
0"-0_0
I I

J FMAMJ J ASOND
MONTH

Fig. 4.4 Precipitation vs. evapotranspiration in the midwestern US from

Bowman and Collins (1987).

ground water. In general, during the summer months, evapotranspira-

tion and soil moisture requirements use all infiltrating water (Fig. 4.4)
except during periods of excessive rainfall (Walton, 1965).

4.2.9 Hydrostratigraphic units

Geologic materials may be classified into three hydrogeologic types:
aquifers, aquitards, and aquifuges. Aquifers are hydrogeologic units
that are capable of sustaining a water supply. Aquitards will not
sustain a water supply. The distinction between aquifers and aquitards
is subject to scale and interpretation. A hydrogeologic unit that
supplies adequate water to a domestic well, and is considered an
aquifer by the well owner, may be considered a portion of an aquitard
to the owner of a municipal water supply system. An aquifuge is a
unit with no interconnected porosity (effective porosity is essentially
0.0). Aquitards and aquifuges are also commonly referred to as con-
fining units.
Seabfl (1988) formally defines a hydro stratigraphic unit as a volume
of geologIC material distinguished by its porosity and permeability. An
example of a formal hydrostratigraphic classification is one proposed in
Illinois by Cartwright (1983). In this classification there are three aqui-
systems based on the nature of the geologic material: the Crystalline
Rock Aquisystem consisting of the PreCambrian basement rocks; the
Indurated Rock Aquisystem consisting of all Paleozoic sedimentary
rocks; and the Non-Indurated Rock Aquisystem consisting of uncon-
solidated Mesozoic and Cenozoic materials. The aquisystems are then
further subdivided into aquigroups, with the top of an aquigroup being
a major aquitard. Aquigr0ups then are subdivided into aquifers and
minor aquitards.
74 Hydrogeology

4.3 FLOW IN UNSATURATED MATERIALS

4.3.1 Definitions
The infiltration rate in a soil is the volume of water flowing into a soil
profile per unit of ground surface area and unit time. The infiltration
rate is therefore the flux of water entering the ground. Horton (1940)
defined infiltration capacity as the maximum infiltration rate that
occurs when the rate of water made available at the surface exceeds the
ability of the soil to adsorb water. Hillel (1971) defined the term
infiltrability for the infiltration rate when water at atmospheric
pressure is made freely available at the ground surface. When water is
ponded at the surface, the infiltration rate can be expected to exceed
the infiltrability because the ponded water will have a head greater
than atmospheric pressure. When water is supplied to the surface
slowly, so that ponding does not occur, the infiltration rate will be less
than or equal to the infiltrability. In the first case, the infiltration
capacity of the soil will dictate the infiltration rate. In the second case,
water will infiltrate as fast as it is applied so the supply rate dictates the
infiltration rate (Hillel, 1980).
Field capacity is the moisture content in a soil after the initial rapid
stage of downward gravity drainage (redistribution) through the soil
column. This redistribution of water is a continual process (Hillel, 1980)
that is relatively rapid in wet soils and slow in dry soils. As the soil
moisture content of a wet soil approaches field capacity, the rate of
downward redistribution exponentially decreases (Richards et ai.,
1956). Foth and Turk (1972) suggest that soils commonly reach field
capacity when about half the pore space is occupied by water. A soil
with a moisture content below field capacity has a soil moisture deficit.
When water infiltrates into a soil with a moisture deficit it is drawn into
the pores, and little is available for downward redistribution until the
moisture content of the soil is replenished to a state greater than the
soil's field capacity.

4.3.2 Equations of unsaturated flow

Lehr (1988) presented a discussion on the microscopic processes that
affect the movement of water in the unsaturated zone. Movement in
the unsaturated zone is in response to suction forces and gravity. The
suction forces consist of adhesion, where water molecules are attracted
to rock or soil particles and cohesion, where water molecules are
attracted to other water molecules. Collectively, these forces are also
referred to as capillary forces. The capillary forces are a result of
hydrogen bonding between the water molecules and other water/soil!
rock molecules. Capillary forces tend to be stronger in fine-grained
 Flow in unsaturated materials 75

materials than in coarse-grained materials because the fine-grained

materials have a greater surface area allowing more molecules for
adhesion.
Capillary forces dominate in dry soils; however, as the soil wets,
gravity becomes the more dominant force. When pressure head is
negative (unsaturated conditions), water movement will only occur in
the liquid phase through continuous water films around particles. As
the moisture content increases, small pores are filled with water before
larger pores, the number and volume of potential flow paths increases,
and the overall hydraulic conductivity of the unsaturated material
increases.
The mathematical representation of unsaturated flow on a macro-
scopic scale is derived from Darcy's law and the continuity equation.
This equation, known as the Richards Equation, is presented below
using the notation of Freeze and Cherry (1979).

a (K(IjI) ah)
ax ax + aya (K(IjI) ah)
ay + aza (K(IjI) ah)
az =
af)
at (4.2)

is the unsaturated hydraulic conductivity that varies with respect

K(IjI)
to pressure head ("'), h is total hydraulic head, and e is the moisture
content of the soil at a given pressure head. As can be seen in Eq. (4.2),
flow in the unsaturated zone is a function of the moisture content and
the pressure head.

4.3.3 Moisture content, tension, and hydraulic conductivity

The hydraulic conductivity in an unsaturated material is lower than the
hydraulic conductivity of that material when it is saturated. Hydraulic
conductivity in an unsaturated material is dependent upon the
moisture content and tension in the material. For any given soil and
condition of wetting or drying, there is a unique relationship between
moisture content and tension. A soil moisture characteristic curve is a
graphical representation of this relationship. Figure 4.5 shows a
hypothetical soil characteristic curve. When tension is zero or greater,
the moisture content is equal to porosity (n) and the material's
hydraulic conductivity is equal to its saturated (maximum) hydraulic
conductivity. As the soil begins to dry, tension increases; but the
moisture content and unsaturated hydraulic conductivity decrease little
because capillary forces continue to hold water in the pores. At some
tension, the pores begin to drain and moisture content and hydraulic
conductivity decrease rapidly.
If the dry soil is re-wetted, moisture content and hydraulic
conductivity increase; however, for any given point on the wetting
curve both moisture content and hydraulic conductivity will be lower
than if the soil were being dried. This effect, called hysteresis, is
 76 Hydrogeology

........
"0
....,Q)
til
-h
....,~
til
rn
~

I=: 2-
0 0
-....
I'/l +
I=:
Q)
...,.;

"0
....,Q)
oj
3
....,
water held in pores by capillary forces

~~O~------------------------------~--------
o n (porosity of sample)
moisture content
volumetric
Fig. 4.5 Soil characteristic curve showing effect of hysteresis.

attributed to several processes. One of the most important processes is

termed by Hillel (1982) as the 'ink bottle' effect. This term describes the
phenomena where the tension at which a saturated pore will drain is
greater than the tension at which it will fill with water. A pore in a
drying soil will drain when the tension reaches some specific value;
however when the soil is rewetting that same pore will fill when
tension reaches a value lower than the value at which the pore
drained. Thus, in a given soil at a specific value of tension, there will
be more pores filled with water during drying than during wetting, so
the moisture content of a drying soil is greater than a wetting soil at
that specific value of tension. Other processes effecting hysteresis are
the contact angle of the meniscus, entrapped air, and soil shrinking/
swelling (Hillel, 1982).
One example of the potential importance of the hysteresis effect
could be the case of characteristic curves for a compacted soil liner. If
this curve is obtained by laboratory analysis for drying, the lab-
estimated unsaturated hydraulic conductivity may be higher at any
given tension than the field values because the field liner would likely
be wetting.
 Flow in saturated materials 77
4.4 FLOW IN SATURATED MATERIALS

4.4.1 Gradient
Ground water in the saturated zone moves from areas of high energy
to areas of low energy. The change in energy over a given distance is
defined as the gradient. There are several types of gradients which can
drive ground-water flow in the saturated zone. A temperature gradient
would cause water to flow from areas of high temperature to areas of
low temperature, an electrical gradient can cause movement of water
from areas of high voltage to areas of low voltage, a density gradient
could cause ground-water flow from zones of high salinity to zones of
low salinity, and a chemical gradient may cause water to flow from
areas of low salinity to areas of high salinity. However, the hydraulic
gradient, where water flows from high head to low head, is generally
the dominant driving force of ground-water flow. The hydraulic
gradient is the maximum change in head over distance from a given
point.
The hydraulic gradient is a vector, therefore it has magnitude and
direction. The magnitude is the change in head over distance, and the
direction is that of the maximum change in head over distance.
Hydraulic gradient cannot be determined with two points, or with any
number of points that lie on a straight line. Hydraulic gradient must be
determined from a surface, either the water table or the potentio-
metric surface, so that the steepest value may be obtained. On a
potentiometric surface, hydraulic gradient is perpendicular to lines of
equal potential. In isotropic materials, flow lines will be parallel to the
hydraulic gradient. However, in non-isotropic materials, flow lines
will not necessarily be parallel to hydraulic gradient - see sections 4.4.3
(Homogeneity and isotropy) and 4.5.3 (Aquifers with flow in
secondary pore openings).

4.4.2 Equations of saturated flow

The mathematical equation that describes ground-water flow is Darcy's
Law. This empirical equation, often given in this simplified form for
one-dimensional flow in isotropic media,
Q = -KiA (4.3)
defines the volumetric flux of fluid (Q) through a given cross-sectional
area (A) for a material with hydraulic conductivity (K) under a
hydraulic gradient (i). A more general form of Darcy's law is
v = -Ki (4.4)
78 Hydrogeology

where v is defined by Freeze and Cherry (1979) as the specific dis-

charge. They make this distinction because v is a flux, not a velocity.
Darcy's law describes flow through a continuum at a macroscopic
scale, and v describes the time it would take for a particle of water to
traverse a distance assuming a porosity value of 1.0. In reality, ground
water must flow more rapidly than the specific discharge would
suggest because a significant portion of the continuum is occupied by a
solid matrix. Just as a river's water velocity must increase when the
channel is constricted, ground-water velocity must increase when the
cross-sectional area available for flow is decreased. An approximation
of macroscopic ground-water velocity is given by the equation
_ -Ki
v=-- (4.5)

v
where ne is effective porosity and is the average linear velocity or
seepage velocity. Because ne is always less than 1.0, v
is always
greater than the specific discharge (v). The average linear velocity
(v) approximates the velocity of ground water at a macroscopic
scale. However, because water molecules must take a tortuous path
around the solid matrix particles, v is not a true measure of the
actual movement of ground-water molecules on a microscopic scale.
However, it is sufficient for evaluation of ground-water flow and con-
taminant transport problems that are also evaluated on a macroscopic
scale.
The equation of ground-water flow for steady state conditions is
derived by combining a three-dimensional form of the Darcy equation
with the continuity equation

~(Kx~)
ax ax + ~(K
ay Y ~)
ay +
az ah)
~(KZaz = 0 (4.6)

where KXI K y, and Kz

are the hydraulic conductivity values in the x,
y, and z directions. This equation describes the head distribution
throughout a three-dimensional flow field for steady state conditions in
which the cartesian coordinates (x, y, z) are parallel to the principal
hydraulic conductivity axes. A similar equation for a transient flow
system would be

(4.7)

where Ss is the specific storage, h is hydraulic head, and t is time.

This equation describes the head distribution throughout a three-
dimensional flow field at any given point in time and space. The left
sides of Eqs. (4.6) and (4.7) are identical; however the right sides differ.
 Flow in saturated materials 79
Equation (4.6) defines steady state flow where head is constant over
time with no change in storage, while Eq. (4.7) defines transient flow
where head changes with respect to time.

4.4.3 Homogeneity and isotropy

Equations (4.6) and (4.7) define head distribution in an anisotropic,
heterogeneous material. If the material was isotropic then K would be
constant in any direction so that Kx = Ky = Kz. If the material was also
homogeneous then K would be constant at any place in the material.
For an isotropic, homogeneous material, with steady state flow, Eq.
(4.6) reduces to Laplace's equation.
a2h a2h a2h
- 2+ - + - 2= 0 (4.8)
ax ay2 az
Homogeneous and isotropic conditions are often assumed for many
analytical solutions to ground-water problems. However, for most
geologic materials, particularly for materials deposited in fluvial
environments, heterogeneous and anisotropic conditions are more

(a) (b)

(c) (d)
Fig. 4.6 Diagram illustrating concepts of heterogeneity and anisotrop~.
(a) Homogeneous, isotropic in a uniform sand, (b) homogeneous, amsotroplc
in a stratified sand, (c) heterogeneous, isotropic, uniform sands, facies change,
(d) heterogeneous, anisotropic, stratified sands, facies change.
80 Hydrogeology

common. Figure 4.6 illustrates the concepts of homogeneity and

isotropy. Both homogeneity and isotropy are subject to scale. For
example, a sandstone layer may be isotropic when viewed by itself; but
as one layer in a sequence of sandstone layers with slightly different
hydraulic conductivities it would be a portion of a vertically anisotropic
flow system. Similarly, the hydraulic conductivity of a till may vary
within a couple of meters; but on a large scale the range of values may
have a normal distribution so that the unit may be considered to be
homogeneous.
An example of an anisotropic, heterogeneous flow system might be
an alluvial valley fill aquifer. The uneven distribution of sand, silt, and
clay deposits commonly found in such deposits would mean changes
in hydraulic conductivity from point to point within the flow system
that could vary by orders of magnitude. Horizontal stratification of
sands and silts in such deposits results in horizontal hydraulic
conductivities that are commonly 2-10 times larger than vertical
conductivities (Revelle, 1941), with factors of 100 or more being
possible.
An example of a homogeneous, isotropic system could be a
sandstone deposited in an aeolian environment. The sorting of grains
by wind action would tend to remove irregularities that could cause
anisotropy and heterogeneity.

4.5 GROUND-WATER FLOW REGIMES

4.5.1 Flow systems

The route ground water takes to a discharge point is known as a flow
path. A flow system is a set of flow paths with common recharge and
discharge areas (Cartwright and Hunt, 1981). Toth (1963) developed
the concept of flow systems to distinguish flow in local, intermediate,
and regional systems. Water in a local flow system will flow to a
nearby discharge area, such as a pond or stream. Water in a regional
flow system will travel a greater distance than the local flow system,
and often will discharge to oceans, large lakes, or rivers (Fig. 4.7).
Flow systems are dependent on both the hydrogeologic charac-
teristics of the soil/rock material and landscape position. Areas of steep
relief tend to have dominant local flow systems, and areas of flat relief
tend to have dominant intermediate and regional flow systems (Freeze
and Witherspoon, 1967). Geological heterogeneities affect the
interrelationship between regional and local flow systems, the pattern
of recharge and discharge areas, and the quantity of ground water that
flows through the systems (Freeze and Cherry, 1979).
 Ground-water flow regimes 81

Fig. 4.7 Local and regional flow systems and relationship to recharge and
discharge areas (from Cartwright and Sherman, 1969).

4.5.2 Recharge and discharge areas

In a typical watershed, the recharge area will be larger than the dis-
charge area (Freeze and Witherspoon, 1967), with discharge occurring
over only 5-30% of the watershed (Freeze and Cherry, 1979). In
general, ground water is recharged in highlands and discharges in
lowland areas; however geology, topography, vegetation, and land
usage can complicate the determination of recharge and discharge
areas.
Rates of recharge are affected by the infiltrability of the soil, the
hydraulic conductivity of the unsaturated materials, depth to the water
table, ambient soil-moisture content, landscape slope and position,
land use, and availability of recharge water. In general, if recharge
water is available, recharge rates will be more rapid in coarse-grained
materials than in fine-grained materials because coarse-grained materials
have higher hydraulic conductivity and lower field capacity than fine-
grained materials.

4.5.3 Types of aquifers

There are many ways to classify the materials through which ground
water flows. One basic division that has already been discussed is that
of an aquifer versus an aquitard. Aquifers can further be divided by
hydraulic characteristics (confined vs. unconfined), porosity character-
istics (primary vs. secondary), and geologic characteristics. Figure 4.8
and Table 4.2 lists ranges of hydraulic conductivity for many of these
materials.
82 Hydrogeology

HYDRAULIC CONDUCTIVITY

gal/ft 2 /day
10 S 10' 10 3 10' 10 10- 1 10- 2 10-4
I I I I I I I I

ft/day
10' 103 10' 10 1 10- 1 10- 3 10-' 10-5
I I I I 1 L I I I
m/day
10' 10 3 10' 10 10- 1 10- 3 10-4 10- 5
I I I I I I I I
em/sec
10 10-' 10- 2 10- 3 10-4 10- 5 10- 7 10- 8
I I I ! I I I I

RELATIVE HYDRAULIC CONDUCTIVITY

Very high High Moderate Low Very low

REPRESENTATIVE SOILS

Clean gravel • Clean sand, and

sand and gravel • Fine sand • Silt, clay and mixtures
of sand, silt, and clay
• Massive clay

REPRESENTATIVE ROCKS
Cavernous limestone or
dolomite; vesicular lava
• Clean sandstone;
fractured limestone,
• Laminated sandstone,
shale, mudstone; • Massive igneous
and metamorphic
dolomite or igneous slightly fractured rocks
and metamorphic limestone or dolomite
rocks

Fig. 4.8 Range of hydraulic conductivity values in selected soil and rock
materials (modified from US Department of Interior, 1981).

Confined vs unconfined aquifers

A confined aquifer is overlain by an aquitard and contains water under
pressure so that the water level in a penetrating well rises above the
top of the aquifer. The elevation to which water in a confined aquifer
will rise in a well is called the potentiometric surface (Fig. 4.9). Freeze
and Cherry (1979) note that a map of the potentiometric surface is
really a map of head using only two or three dimensions. If there is a
vertical component of flow in the system then any interpretations
based on a potentiometric surface may be in error.
Unconfined aquifers generally occur at relatively shallow depths.
The top of the saturated zone in an unconfined aquifer is within the
rock or soil strata of the aquifer. This surface is the water table (Fig.
4.9). Unlike a confined aquifer, an unconfined aquifer is not necessarily
bounded at the top by an aquitard. Using a hydro stratigraphic
definition, the upper boundary of an unconfined aquifer would be top
of the hydro stratigraphic unit that makes up the aquifer. Using an
older definition, the upper boundary of an unconfined aquifer is the
water table. Under the latter definition, the upper boundary of an
unconfined aquifer is not fixed in space or time because the water table
rises and falls in response to seasonal and man-induced stresses. In
general, the water table will not coincide with the potentiometric
surface in areas where both types of aquifer are present (Freeze and
Cherry, 1979).
Table 4.2 Typical hydraulic conductivity, porosity, and effective porosity
values

geologic material hydraulic porosity sources

conductivity (m/s) total effective
--,---- --- ------ --------

Igneous/Meta. 0-5 5,11

fractured 10- 10 _10- 13 0-10 5
weathered 10- 4 _10- 8 40-50 11
Basalt
dense 10- 9 _10- 13 1-8 2,4
permeable 10- 2 _10- 7 5-50 5-10 2,5,6
lava 10- 1 _10- 17 12
Tuff 10- 3 _10- 8 10-40 2,9, 11, 12
Granite 10- 9 _10- 13 0-2 4
weathered 10- 4 _10- 5 0-45 0-8 2,6,9
Quartzite 1 4
Gneiss 10- 9 _10- 13 0-2 4
Sedimentary rocks
Limestone/dolomite 10- 6 _10- 9 0-20 1,5
fractured 10- 4 _10- 9 0-20 0-18 1,6
Limestone
oolitic 10- 6 _10- 7 1-25 1
karst 10+ 2 _10- 6 5-50 1,5
marble 10- 5 _10- 8 0-2 1
chalk 10- 5 _10- 9 8-45 1,4,6
weathered 10- 3 _10- 5 4
Sandstone/siltstone 10- 4 _10- 10 10-50 10-48 3,4,6,9, 11
Shale 10- 9 _10- 13 0-10 3,5,9
Coal 10- 6 _10- 11 11
Unconsolidated sediments
Gravel 10- 1 _10- 4 25-60 7-40 2,5,6,9, 11
Sand
clean 10- 2 _10- 6 25-55 24-40 2,5,6,9, 11
silty 10- 3 _10- 7
Silt 10- 6 _10- 10 35-60 2,5,9, 11
Loess 10- 5 _10- 11 2,9,10,11
Lacustrine silt/clay 10- 9 _10- 13 40-70 5, 10
fractured 10- 8 _10- 11 10
Marine clay 10- 9 _10- 12 40-70 5
Clay/silt compacted 10- 9 _10- 11 30-50 <0.1 7,8
Peat 10- 6 _10- 9 60-92 9,11
Glacial deposits
Outwash 10- 3 _10- 7 35 6,10
Tillibasal 10- 7 _10- 11 5
weathered 10- 6 _10- 9 10
fractured 10- 5 _10- 9 10
Till/ su praglacial 10- 5 _10- 9 10
weathered 10- 5 _10- 9 10
fractured 10- 5 _10- 9 10

Sources: 1. Brahana et al. (1988); 2. Davis (1969) - K values assume water at 20°C; 3. Davis
(1988); 4. de Marsily (1986); 5. Freeze and Cherry (1979); 6. Gelhar et al. (1985); 7. Horton,
Thompson and McBride (1988); 8. Krapac et al. (1991); 9. Morris and Johnson (1967); 10.
Stephenson, Fleming and Mickelson (1988); 11. Walton (1985); 12. Wood and Fernandez
(1988). Note: 1, 2, 3, 6 and 12 are compilations of data from other sources.
84 Hydrogeology

water level
~~A _~~~~B
(water table) (potentiometric SUrface)..,

~------~/--------------~--~
---
water table
.
'"

. sand

'. . .

. Dsa~dand· gra~el!:>:

Fig.4.9 Confined and unconfined aquifers, from Cartwright and Hunt (1981).

Aquifers with intergranular flow

Aquifers in which water flow is through intergranular (primary) pore
openings may consist of lithified sedimentary deposits such as
sandstone and conglomerate, or unlithified sand and gravel deposits.
The porosity, permeability, anisotropy and heterogeneity of these
deposits are functions of the depositional environment and the aquifer
material.
Darcy's law is generally assumed to be valid in aquifers with inter-
granular flow. Ground water moves through the pore openings in
response to head differentials. The pore openings in many porous
media aquifers are sufficiently large and interconnected so that effective
porosity approaches total porosity.
Flow is tortuous on a microscopic scale as the water weaves its way
between the grains of the aquifer matrix. However, flow is fairly
uniform on a macroscopic scale, and the direction of flow will be fairly
uniform from recharge areas to discharge areas unless there is a
disturbance in the flow system. Disturbances may be caused by
addition or extraction of water at sources or sinks such as wells and
small lakes and streams; or by changes in the geology/permeability of
the aquifer such as a facies change, fault, or intrusive body.
Contaminant transport in porous media aquifers will be affected
primarily by the hydraulic conductivity and effective porosity of the
aquifer, and by dispersion (Fig. 4.10). As a plume moves down-
gradient, it may move more rapidly in zones of higher hydraulic con-
ductivity and may take on a fingering configuration if the zones of
 Ground-water flow regimes 85

Fig. 4.10 Contaminant plume in an aquifer with intergranular flow, from

USEPA (1980).

higher hydraulic conductivity are extensive. On a microscopic scale,

molecules of contaminant and water are taking a tortuous flow path
around the sediment grains, causing mixing with non-contaminated
water so the area of the plume increases as the concentration decreases.
Examples of aquifers with intergranular flow are given below. Each
represents a different type of depositional environment.

Example 1
Large alluvial aquifers can be found in the valleys of many major river
systems. These aquifers consist of sands and gravels deposited in
either a fluvial or glaciofluvial environment. Examples of this type of
aquifer would be the sand and gravel deposits of the Ohio, Missouri,
and Mississippi River Valleys. These aquifers often can be subdivided
into a lower substratum where hydraulic conductivity is usually high
and the flow system is often heterogeneous and isotropic (Foreman
and Sharp, 1981) and an upper top-stratum where hydraulic conduc-
tivity is variable and tends to be greatest in the direction of the paleo-
current of a given channel deposit (Sharp, 1988). The direction of
ground-water flow in alluvial valley aquifers is generally towards the
river although some flow in the downstream direction is also common.
Recharge may come from underlying and adjacent bedrock formations,
precipitation, and seepage from tributary streams, and discharge is
commonly to the river.

Example 2
Sands and gravels within glacial deposits occur in different forms
related to different glacial environments. Common types of glacial
deposits that may contain sand and gravel are till, ice-contact, and
outwash.
86 Hydrogeology

Till is deposited directly from glacial ice and is usually poorly sorted
(Davis and DeWiest, 1966). The textural and hydrogeologic properties
of tills vary greatly, and are related to the source rock of the till and the
distance that the material has been transported by ice. In the mid-
continental US, till is largely derived from shale and carbonate rocks,
furthermore this material has been transported great distances. These
tills are typically fine-grained and have low hydraulic conductivity, and
are considered to be confining beds. In areas where tills are derived
from nearby precambrian bedrock or younger sandstones, such as in
the Canadian Shield area or New England, the tills are often coarse-
grained with high hydraulic conductivity and may be valuable aquifers
(Stephenson, Fleming and Mickelson, 1988).
Ice-contact deposits are sediments that were glaciofluvially deposited
in contact with ice (Stephenson, Fleming and Mickelson, 1988). They
may range in texture from silt and clay to sand and gravel and may be
well or poorly sorted. Ice contact sands and gravels may be discon-
tinuous and of limited extent (lOOs to lOs of meters or less), in which
case they may only yield sufficient water for domestic purposes.
However, when they are laterally extensive, these formations may be
productive aquifers.
Outwash describes glaciofluvial sediments deposited by streams
flowing away from the ice (Stephenson, Fleming and Mickelson, 1988).
These types of deposits are generally coarse-textured and well sorted.
Some alluvial aquifers described above may be partially or wholly
made up of outwash deposits. Outwash from older glaciations may be
buried by younger glacial deposits, as is the case with some highly
productive buried bedrock valley aquifers such as the Mahomet Aquifer
of central Illinois (Stephenson, 1967).
Outwash deposits tend to be more extensive and more homo-
geneous than ice contact deposits. Tills are usually finer-grained and
with lower hydraulic conductivity than glaciofluvial deposits. Some
sandy tills and outwash deposits may be fairly isotropic; ice contact
deposits will tend to be vertically and horizontally anisotropic.
Sand and gravel deposits within glacial materials are often bounded
by clayey till deposits with hydraulic conductivities that may be four or
more orders of magnitude lower than the aquifer (Stephenson, Fleming
and Mickelson, 1988), although fractures in the till may decrease that
ratio considerably. Recharge to these aquifers may be in the form of
precipitation to surficial aquifers or seepage through overlying till units
to buried aquifers. Discharge may be to a lake or stream if the aquifer is
hydraulically connected, or to the surrounding till unit if the aquifer is
of limited extent. Buried bedrock valley aquifers and other deposits in
contact with the underlying bedrock may be hydraulically connected if
both units are aquifers.
 Ground-water flow regimes 87
Example 3
Clastic sedimentary rock aquifers commonly are comprised of sand-
stone or conglomerate. These aquifers are important on a world-wide
basis because 25% of the sedimentary rocks in the world are estimated
to be sandstone (Freeze and Cherry, 1979). Sandstone can be deposited
in aeolian, fluvial, shoreline, or shelf and slope environments. The
aeolian deposits are relatively homogeneous, compared to fluvial
deposits, and as isotropic as any deposits found in nature (Freeze and
Cherry, 1979). The fluvial, shoreline, and shelf deposits are commonly
horizontally stratified and vertically isotropic; thus horizontal hydraulic
conductivity may exceed vertical hydraulic conductivity by as much as
three- to la-fold (Davis, 1988). However, vertical fracturing can result
in increased vertical hydraulic conductivity compared to horizontal.
Sandstone formations commonly have permeability values that are
lower than the sands from which they formed because cementing
materials, primarily silicate and carbonate minerals, precipitate be-
tween the grains and because compaction reduces the porosity of the
formation during lithification (Freeze and Cherry, 1979; Davis, 1988).
However, if the porosity of clastic rocks is low «0.15) then perme-
ability may be controlled more by secondary fractures than by porosity,
which may result in hydraulic conductivity as high as or higher than
the source sands (Davis, 1988).
Flow systems in large sandstone aquifers will typically fit the model
described by Toth (1963). Recharge to regional aquifers will primarily
occur at outcrop areas, although some recharge will also occur as
leakage through overlying geologic units where the aquifer is buried.
Regional discharge will usually be to large lakes and rivers, although
local discharge to smaller surface water bodies is also likely.

Aquifers with flow in secondary pore openings

Some geologic formations that lack significant primary porosity can still
be valuable aquifers because ground water is able to flow through
fractures and solution openings in the rock. In fact, many such
formations constitute major regional aquifers. Flow through secondary
pore openings is most common in sedimentary carbonate and in
igneous/metamorphic rocks.
For highly fractured rocks, it may be valid to assume that the
continuum approach of Darcy's law can be used to describe ground-
water flow on a large scale. For rock materials with widely spaced
fractures or very large cave-like openings, the continuum approach
generally is not valid.
An example of an aquifer where the continuum approach is valid on
a large scale is the weathered portion of the Silurian dolomite aquifer in
88 Hydrogeology

northeastern Illinois and eastern Wisconsin. An example of an aquifer

where the continuum approach is not valid is the karst carbonate
aquifer of central Kentucky.
Most fracture flow aquifers fall between these extremes; so the
common approach is to assume a large scale continuum. In other
words, local ground-water flow in a fractured aquifer on the scale of a
landfill may not be Darcian; but flow for the entire aquifer, which may
underlie several counties, may be treated as Darcian.
Geologic controls are the rule for aquifers with secondary porosity. A
tectonically folded aquifer may have greater permeability at the crest of
the anticline and in the synclinal trough than on the flanks (Davis and
DeWiest, 1966) because added stress at those points may have caused a
greater degree of fracturing.

A4E A5E A6E A7E A8E A 9 E

DEKALS----l----------f KANE -----l---------- 1----------,--=-CO~cO:cK·-·-------" T

T
42
N . i ~
i
__.-L I
I
.,
T i .,
T

N i N
i

T T
'0 .0
N N

T K2 T
39 39
N N

• K1

T T
38 38
N N

WILL

T F-8 T
37 37
N N

A. E 6 E A 7 E A 8 E A 9 E

I\t m,les ot-r_ _ _----.---rl--__

~
---.-I12
;8 kIlometers

Fig.4.11 Rose diagram for fractured carbonate rocks, from Curry et al. (1988).
 Ground-water flow regimes 89
Fractured aquifers may be highly anisotropic, particularly in the
horizontal orientation. That is because geologic formations commonly
have a vertical fracture pattern where there is a primary and secondary
fracture orientation (Fig. 4.11). The effect of this fracture orientation on
ground-water flow is that hydraulic conductivity is greatest in the
direction of primary fracturing, lower in the direction of secondary
fracturing, and least in orientations where there are few fractures.
Thus, if the hydraulic gradient is tangential to the primary directions of
fracturing, flow will not be orthogonal to lines of equal potential and
parallel to the hydraulic gradient; but rather in a direction that is highly
controlled by anisotropic hydraulic conductivities dictated by
fracturing. It may be almost 90 degrees to the gradient in extremely
anisotropic conditions.
Contaminant transport in fractured aquifers can be extremely
complex because of the extreme anisotropy and heterogeneity of the
aquifer (Fig. 4.12), and because transport rates can be rapid. The
effective fracture porosity of fractured aquifers may be significantly less
than 1%. As can be seen from Eq. (4.5), formations with very low
effective porosities can have very rapid ground-water velocities. Thus,
contaminant transport in a fractured aquifer may be very rapid due to
rapid fluid movement through the fractures, particularly through large
fractures.
An extreme example of contaminant transport through large
openings is in karst aquifers. In this hydrogeologic environment, the
best method to monitor a waste disposal/storage site is to locate and
monitor all of the discharge points of the karst system underlying the
potential source of contaminants (Quinlan and Ewers, 1985).

fractured rock

Fig. 4.12 Contaminant plume in an aquifer with flow through fractures, from
USEPA (1980).
90 Hydrogeology

4.5.4 Flow in aquitards

Aquitards are hydrostratigraphic units that transmit ground water less
readily than aquifers. The definition of an aquitard is use dependent.
However, for characterizing ground-water flow systems, an aquitard
may be considered as any continuous hydrogeologic unit that has a
hydraulic conductivity 2 or more orders of magnitude less than
overlaying and/or underlying aquifers. Units comprised of shale or clay
are commonly aquitards, although massive, non-fractured carbonates
and crystalline rocks can also be aquitards. However, it is difficult to
classify a particular rock type as an aquitard without extensive field
testing because unexpected geologic conditions may allow fairly rapid
transmission of water.
An example of a rock that may not necessarily be an aquitard would
be shale. Davis (1988) notes that fractures may cause shales at shallow
depths to be permeable enough to transmit water to a well or away
from waste disposal trenches; however shales at deep depths may be
as close to being absolutely impermeable as any geologic material
within a few kilometers of the Earth's surface.
The permeability and hydraulic conductivity of aquitard materials
tends to be low because fracture and pore openings tend to be too
small or too poorly connected to allow significant fluid movement. In
essence, the effective porosity of these materials approaches zero. The
distinction between effective porosity and total porosity is important
because the total porosity of some aquitard materials such as clay and
some volcanic rocks may be an order of magnitude or more greater
than the effective porosity. As noted earlier, the permeability of all of
these units can be increased considerably by open fractures.
When aquitards are underlain by aquifers, it is common to have a
situation where flow is vertical through the aquitard; so that flow is
downward in recharge areas and upward in discharge areas. This
condition may be enhanced if vertical fractures are present.
Clays and shales are commonly touted as confining materials to
contaminant transport. Stephenson, Fleming and Mickelson (1988)
compiled ground-water velocities for clay-rich glacial tills and found
intergranular ground water velocities of 0.005-11 mm/yr. Other studies
have shown that such velocities may be 300 mm/yr (Hughes et al.,
1971). At these low ground-water velocities diffusion becomes an
important contaminant transport mechanism (Freeze and Cherry,
1979). However, the data ranges compiled by Stephenson, Fleming and
Mickelson also showed that sand lenses or fractures in the till can
result in overall velocities as great as 40 m/yr.
 References 91
4.6 PRACTICAL IMPLICATIONS

When characterizing ground-water flow at a waste disposal site, care

must be used when defining the confining layer. A hydrostratigraphic
unit cannot be classified as a confining layer based only on its rock or
soil type. Nor can this distinction be made based on limited, small-
scale testing. For example, the hydraulic conductivity of a glacial till
may have been determined in situ as 1 x 10- 11 mis, and isotope
analysis on water extracted from cores from the boring for that well
may show that the unit contains ancient water. But, fractures within
meters of the well may be rapidly transmitting modern water to an
underlying aquifer. Characterizing a confining layer for waste disposal
requires detailed knowledge of the hydrogeology of the entire unit,
which means one must understand the geologic processes that have
affected the unit during and after deposition.
The movement of ground water is dictated by the physical char-
acteristics of the hydro stratigraphic units within the flow system.
Therefore, it is necessary to establish a good geologic framework before
characterizing any ground-water flow problem. If this framework is
established, so that the character, depositional environment, and post-
depositional history of the geologic materials is understood, it is then
possible to apply hydrogeologic principles to characterize ground-
water movement.

REFERENCES

Bowman, J.A. and Collins, M.A. (1987) Impacts of irrigation and drought on
groundwater resources, Illinois State Water Survey Report of Investigations
109.
Brahana, J.V., Thrailkill, J., Freeman, T. and Ward, W.e. (1988) Carbonate
rocks in Hydrogeology: The Geology of North America (eds W. Back, J.S.
Rosenshein and P.R. Seaber) V. 0-2, Geological Society of America, pp.
333-52.
Cartwright, K. (1983) Classification of hydrostratigraphic units in Illinois,
Illinois State Geological Survey draft memorandum, open file.
Cartwright, K. and Hunt, e.S. (1981) Hydrogeologic aspects of coal mining in
Illinois: an overview, Illinois State Geological Survey EGN 90.
Cartwright, K. and Sherman, F.B. (1969) Evaluating sanitary landfill sites in
Illinois, Illinois State Geological Survey EGN 27.
Curry, B.B., Graese, A.M., Hasek, M.J., Vaiden, R.e., Bauer, R.A.,
Schumacher, D.A., Norton, K.A. and Dixon, W.G. Jr. (1988) Geological-
geotechnical studies for siting the superconducting super collider in
Illinois: results of the 1986 drilling program, Illinois State Geological Survey
EGN 122.
Davis, S.N. (1969) Porosity and permeability of natural materials, in Flow
through porous media (ed. R.J.M. De Wiest) Academic Press, pp. 53-89.
Davis, S.N. (1988) Sandstones and shales, in Hydrogeology: The Geology of North
92 Hydrogeology

America, (eds W. Back, J.S. Rosenshein and P.R. Seaber) V. 0-2, Geological
Society of America, pp. 323-32.
Davis, S.N. and DeWiest, R.J.M. (1966) Hydrogeology, John Wiley and Sons,
Inc. de Marsily, G. (1986) Quantitative hydrogeology, Academic Press.
Foreman, T.L. and Sharp, J.M. Jr. (1981) Hydraulic properties of a major
alluvial aquifer; an isotropic, inhomogeneous system, J. Hydrology, 53,
247-58.
Foth, H.D. and Turk, L.M. (1972) Fundamentals of Soil Science, 5th edn, John
Wiley and Sons, Inc.
Freeze, R.A and Cherry, J.A (1979) Groundwater, Prentice-Hall, Inc.
Freeze, R.A. and Witherspoon, P.A (1967) Theoretical analysis of regional
groundwater flow, 2: effect of water-table configuration and subsurface
permeability variation. Water Resources Res., 3, 623-34.
Gelhar, L.W., Mantoglou, A, Welty, C. and Rehfeldt K.R. (1985) A review of
field-scale physical solute transport processes in saturated and unsaturated porous
media, Electric Power Research Institute EA-4190.
Hillel, D. (1971) Soil and water, physical principles and processes, Academic Press.
Hillel, D. (1980) Applications of soil physics, Academic Press.
Hillel, D. (1982) Introduction to soil physics, Academic Press.
Horton, R., Thompson, M.L. and McBride, J.F. (1988) Determination of effective
porosity of soil materials, USEPA No. EPA/600/2-88/045, NTIS PB88-242391.
Horton, R.E. (1940) An approach toward a physical interpretation of infiltration
capacity, Proc., Soil Sc. Soc. of Amer., 5, 399-417.
Hughes, G.M., Landon, R.A and Farvolden, R.N. (1971) Hydrogeology of solid
waste disposal sites in northeastern Illinois, USEPA Solid Waste Management
Series SW-12d.
Krapac, I.G., Cartwright, K., Hensel, B.R., Herzog, B.L., Larson, T.H., Panno,
S.V., Risatti, J.B., SU, W.J. and Rehfeldt, K.R. (1991) Construction,
Monitoring, and Performance of Two Soil Liners, Illinois State Geological
Survey Environmental Geology 141.
Lehr, J.H. (1988) The misunderstood world of unsaturated flow, G. Water
Monit. Rev., 8(2), pp. 4-6.
Meinzer, O.E. (1923) The occurrence of groundwater in the United States, USGS
Water Supply Paper 489.
Morris, D.A and Johnson, AI. (1967) Summary of hydrologic and physical pro-
perties of rock and soil materials, as analyzed by the Hydrologic Laboratory of
the U.S. Geological Survey 1948-1960, USGS Water Supply Paper 1839-D.
Quinlan, J.F. and Ewers, R.O. (1985) Ground water flow in limestone terrains:
strategy rationale and procedure for reliable, efficient monitoring of
ground water in karst areas, in Proc., 5th Nat!. Symp. and Exp. on Aquifer
Rest. and G. Water Monit., National Water Well Association, pp. 197-234.
Rehm, B.W., Groenewold, G.H. and Peterson, W.M. (1982) Mechanisms,
distribution, and frequency of ground water recharge in an upland area of western
North Dakota, North Dakota Geological Survey Report of Investigations 75.
Revelle, R. (1941) Criteria for recognition of sea waters in ground-waters, in
Trans. Amer. Geophys. Union, 22, 593-7.
Richards, L.A, Gardner, W.R. and Ogata, G. (1956) Physical Processes
determining water loss from soil, in Proc., Soil Sc. Soc. of Amer., 20, pp.
310-4.
Seaber, P.R. (1988) Hydrostratigraphic units, in Hydrogeology: The Geology of
North America, (eds W. Back, J.S. Rosenshein and P.R. Seaber) V. 0-2,
Geological Society of America, pp. 9-14.
 References 93
Sharp, J.M. Jr. (1988) Alluvial aquifers along major rivers, in Hydrogeology: The
Geology of North America (eds W. Back, J.S. Rosenshein and P.R. Seaber) V.
0-2, Geological Society of America, pp. 273-82.
Steenhuis, T.S., Jackson, e.D., Kung, S.K.J. and Brutsaert, W. (1985)
Measurement of recharge on eastern Long Island, New York, USA, J.
Hydrology, 79, 145.,...69.
Stephenson, D.A. (1967) Hydrogeology of glacial deposits of the Mahomet Bedrock
Valley in east-central Illinois, lllinois State Geological Survey Circular 409.
Stephenson, D.A., Fleming, A.H. and Mickelson, D.M. (1988) Glacial deposits,
in Hydrogeology: The Geology of North America (eds W. Back, J.S. Rosenshein
and P.R. Seaber) V. 0-2, Geol. Soc. of Amer., pp. 301-14.
Todd, D.K. (1980) Groundwater Hydrology, 2nd edn, John Wiley and Sons, Inc.
Toth, J.A. (1963) A theoretical analysis of ground-water flow in small drainage
basins, J. Geophys. Res., 68, pp. 4795-812.
US Department of Interior (1981) Groundwater manual, John Wiley and Sons.
USEPA (1980) Procedures manual for groundwater monitoring at solid waste disposal
facilities, Solid Waste Management Series report SW-611, Cincinnati, OH.
Walton, W.e. (1965) Ground-water recharge and runoff in Illinois, Illinois State
Water Survey Report of Investigations 48.
Walton, W.e. (1985) Practical aspects of ground water modeling, 2nd edn, National
Water Well Association, Worthington, OH.
Wood, W.W. and Fernandez, L.A. (1988) Volcanic rocks, in Hydrogeology: The
Geology of North America (eds W. Back, J.S. Rosenshein and P.R. Seaber) V.
0-2, Geological Society of America, pp. 353-65.
 PART TWO

New Disposal Facilities

 CHAPTER 5

Landfills and impoundments

David E. Daniel

5.1 INTRODUCTION

Landfills are the final repositories for unwanted or unusable wastes.

Until the middle of this century, nearly all wastes were discarded in
open, unengineered dumps. Waste was often burned to conserve
space. Topographical anomalies that lended themselves naturally to
dumping were typically selected for dump sites. The most common
waste dumps were natural depressions (creeks, low-lying areas, and
flood plains) that were otherwise of little use and mined-out areas,
e.g., sand or gravel quarries. The practice of open dumping changed
little, until a few decades ago.
The sanitary landfill began to become commonplace shortly after
World War II. A sanitary landfill consists of a refuse disposal area in
which the waste is disposed of in cells that range in thickness up to
about 5 m. Within each cell, waste is covered with a 150- to 300-mm-
thick layer of soil (called daily cover) at the end of each working day.
Today, several foams are available that in some cases have been
substituted for daily soil cover. The configuration of cells is sketched in
Fig. 5.1 for a waste containment unit that contains a liner. In fact,
however, early sanitary landfills had no man-made liner; liners did not
become common in sanitary landfills until the 1970s, and even today,
not all sanitary landfills have liners.
The sanitary landfill represented a dramatic improvement over the
open dump. Controlled placement of waste in sanitary landfills
(particularly daily covering) greatly reduced the number of rodents and
insects, dramatically reduced public health risks, and generally
contributed to major aesthetic improvements in waste disposal.
Engineered liners for waste disposal facilities did not become routine
until the 1970s. The author recalls first seeing a diagram of a double

Geotechnical Practice for Waste Disposal.

Edited by David E. Daniel.
Published in 1993 by Chapman & Hall, London. ISBN 0 412 35170 6
98 Landfills and impoundments

Fig. 5.1 Daily cells in a landfill.

liner for a landfill in about 1978. Sophisticated waste containment units

with multiple liners and fluid collection systems did not become
common in the US until they were mandated by the US Environmental
Protection Agency (EPA) for hazardous waste landfills in the early
1980s.
As mentioned in Chapter I, regulations have driven landfill practices
in the US and most other countries. Requirements for landfills vary
with the type of waste and other factors such as the hydrogeology of a
site, climate, and type of waste to be buried. In this chapter, general
principles of landfill design will be discussed. In the remaining
chapters that comprise Part II of the book, specific details on the design
of various landfill components will be discussed.

5.2 OBJECTIVES OF WASTE DISPOSAL FACILITIES

The objective of a waste disposal facility are frequently not spelled out
in clear terms that can be agreed upon by all parties involved in the
development of a facility. Further, people frequently have different
understandings about the objective of a waste disposal facility. For an
owner, the objective may be to dispose of waste that poses a public
health risk at a cost that can be borne by the local citizens. For the
engineer, the objective is frequently to design a facility that optimizes
the tradeoff between adequate waste containment and cost. For an
individual living near a landfill, the objective is for the facility to be oper-
ated with zero impact upon the individual or the individual's property.
A critical question that is often asked is: Will the landfill leak? The
 Siting of landfills 99

answer to this question depends upon the definition that one selects
for the term 'leak'. If one assumes that a leak is an unanticipated
discharge, then the landfill need not leak. If one assumes that a leak is
any release, then it is clear that all landfills either leak or have the
potential to leak. No liner material known to mankind is forever im-
permeable to all chemicals. Some chemicals will eventually migrate by
advection or diffusion through all liner materials unless an inward
gradient producing a chemical flux toward the containment unit is
maintained. The objective of the landfill design should not be to stop
the release of all chemicals from the facility for an infinite period of
time. Such an objective is unrealistic. The real issue with respect to
containment of waste in the disposal facility revolves around how
much waste material in leachate (the liquid that seeps from the landfill)
will be released from the containment facility over a given period of
time and what the environmental impact of the release will be. For well
designed facilities, the quantities of chemicals released from the waste
disposal facility are minute and environmental impacts (short and long
term) are negligible.
The objective of a waste disposal facility is to contain the waste
in a manner that is protective of human health and the environment.
Because no endeavor of mankind can be undertaken without some
risk, there is always a risk that a landfill will fail to perform up to
expectations. The best that can humanly be expected of designers is to
ensure that the risk posed by a disposal facility is extremely low.
Monitoring systems are installed around landfills to determine if the
facility is performing in an unexpected manner.
Regulations dictate the minimum technology that is required to keep
risks associated with waste containment facilities small. Usually, the
owner's and engineer's objectives are to comply with regulatory
requirements. It is assumed by most engineers that so long as the
minimum regulatory requirements are met, the environmental risk will
be low. However, the author encourages the engineer to think about
and evaluate environmental risks independently of regulatory require-
ments and to satisfy himself or herself that the risks are suitably low
with the design that is ultimately selected.

5.3 SITING OF LANDFILLS

An entire book could be written on site selection for waste containment

units. Some of the factors that must be considered include proximity to
waste generators, availability of transportation systems for moving
waste to the site, climate, geology, hydrogeology, surface hydrology,
proximity of airports, demographics, land use, impacts on the local
community, and others. It is widely understood that landfills are
100 Landfills and impoundments

perceived to an undesirable addition to a community. An attitude of

NIMBY (not in my back yard) is often used to describe the attitude of
residents of local communities. Citizens of an area will typically accept
that landfills are needed (although many suffer from the misconception
that waste minimization and recycling can eliminate the need for
landfills), but not in their community. The problem, of course, is that
the landfill must be located in someone's community. The process of
getting local citizens to accept the siting of a waste disposal facility in
their community is frequently a more significant problem than dealing
with technical issues. Most experienced landfill siting specialists have
found that effective communication at a very early stage is usually the
most effective way to deal with NIMBY attitudes. Many citizens
completely misunderstand what a modern waste disposal facility is
like; visions of smoldering wastes in an open dump are common.
Educating people who have never had the opportunity, or taken the
time, to learn how a modern waste disposal disposal facility is
designed and operated is usually a critical prerequisite in gaining
community support for a new waste disposal facility.
Another common approach for dealing with the NIMBY problem is
to build new waste containment units on top of (vertical expansion) or
next to (lateral expansion) an existing landfill. There may still be strong
local opposition to expansion of a landfill (e.g., because of concerns
over truck traffic), but there are fewer NIMBY obstacles to overcome
in expanding a facility compared to siting a new facility. Another
approach is to build a new landfill in an area that already has one or
more landfills.
The specific geotechnical problems related to site selection are site
characterization and analysis. The primary issues are definition of
subsurface stratigraphy, identification of groundwater conditions (in-
cluding identification of aquifers and aquitards, measurement of piez-
ometric levels in aquifers, and characterization of background water
quality and geochemical conditions), determination of the properties
of major subsurface units (with particular emphasis on hydraulic
conductivity, geochemical attenuation characteristics, and strength),
evaluation of the availability of construction materials (e.g., clay for
liners and granular material for drains), and analysis of the probable
performance of the site, including assessment of probable groundwater
impacts and analysis of the stability of the facility.
An ideal site is one that is located reasonably close to the source of
the waste, has convenient transportation access, is not situated in a
low-lying area or floodplain, is underlain by suitably strong materials,
and has favorable hydrogeological characteristics. Some sites are so
complex that their characterization is at best very difficult and, in some
cases, impossible~ If the site is so complex that it cannot be adequately
 Containment technology 101
characterized, the site is not a favorable one to host a waste disposal
facility.

5.4 CONTAINMENT TECHNOLOGY

One of the first questions that the landfill designer must address is
whether the waste containment units will be above or below ground.
As indicated in Fig. 5.2, the waste can be disposed of either above the
existing surface, below the surface, or both above and below. Above-
ground disposal is particularly attractive for sites with shallow water
tables. Above-ground landfills have the advantage that leachate can be
drained by gravity, the facility is conspicuous and therefore is not
easily forgotten and ignored, and construction of liner and drainage-
(A) Above·Ground landfill

Une.

(6) Above- and Below-Ground landfill

(e) Below-Ground Landfill

Fig. 5.2 Landfills: (a) above-ground; (b) above- and below-ground; (c) below-
ground.
102 Landfills and impoundments

system components occurs on more or less level ground, which

simplifies construction. Vertical expansions of existing facilities essen-
tially constitute above-ground landfills. Disposal units that are partly
below ground have the advantage of allowing more waste to be
disposed of in a given area, offer more efficient use of construction
materials (excavated material can be used for daily cover or other
purposes), and potentially could allow for productive use of the land
(e.g., for a park) if the final surface is relatively flat.
Several systems can be employed to control release of waste con-
stituents. The primary objectives of control systems are:
1. to minimize infiltration of water into the waste through the cover or
sidewalls of the facility;
2. to collect and remove gas and leachate; and
3. to minimize release of leachate to the subsurface.
In the past, natural soils were often used as liners (Fig. S.3a), but this
practice is becoming less common because it is very difficult to prove

(A) Waste Buried in a Natural Soil Liner

(B) Waste Buried Above a Natural Soil Liner

(A) Waste Buried in Unit with Engineered Liner

Fig. 5.3 Lining of waste disposal units with (a) natural soils, (b) vertical cutoff
walls, and (c) engineered liners.
 Containment technology 103
that the natural liner is free of high-hydraulic-conductivity zones or
secondary features such as cracks and joints. Vertical cutoff walls are
occasionally used in conjunction with a natural soil liner (Fig. 5.3b), but
this practice is unusual for new disposal facilities. (However, sur-
rounding a contaminated area with a cutoff wall is a common practice
for site remediation, as discussed in Chapter 17). A more typical
situation is shown in Fig. 5.3c in which an engineered liner is em-
ployed to control movement of contaminants out of the disposal unit.
In the US, Subtitles C and 0 of the Resource Conservation and
Recovery Act give the USEP A the authority to establish minimum
containment requirements for hazardous and nonhazardous solid
waste, respectively. Figure 5.4 illustrates the minimum containment
requirements that are usually imposed. For hazardous waste landfills
(Fig. 5.4a), a double liner system is required with a leachate collection
layer located above the primary liner and a leak detection layer located
between the two liners. The Subtitle 0 requirements (Fig. 5.4b) were
promulgated a few weeks before final revisions were made to this
chapter and do not go into effect until 1993. Nevertheless, the minimum
lin~~ consists of a single composite liner with a leachate collection
system. Recommended European designs for waste containment

(a) primary geomembrane

liner

leak detection and

removal system
secondary
geomembrane
liner
composite
secondary
liner

(b)
leachate collection
removal system
geomembrane

composite
liner

Fig. 5.4 Minimum liner requirements of the US Environmental Protection

Agency: (a) for hazardous waste landfills (b) for non-hazardous waste landfills.
104 Landfills and impoundments

facilities call for a single composite liner with a leachate collection

system (Fig. 5.4b).
For a hazardous waste facility, the EPA requires that the compacted
soil liner be at least 0.9 m thick and have a hydraulic conductivity ~ 1
X 10- 7 cm/s. Drainage layers are typically required to have a hydraulic
conductivity ~1 em Is, and the leak detection system must be capable
of detecting a leak within 24 hours. Geosynthetic drainage materials
(typically geonets) are commonly used for the leak detection layer
(especially on side slopes). Geomembranes must be at least O.76mm
thick. For non-hazardous waste facilities, the requirements are similar,
but the minimum thickness of the compacted soil liner is O.6m.
The final cover system for a waste disposal facility may consist of
several layers (Fig. 5.5). A surface layer controls erosion and, for sites
where vegetation grows on the surface, supports vegetative growth.
The surface layer usually consists of topsoil that is seeded with appro-
priate vegetation. Occasionally, at arid sites, gravel or cobbles are
placed at the surface. A protective layer isolates the surface environ-
ment from the underlying waste and components of the cover system.
The protective layer also stores water that infiltrates through the
surface layer; later, the stored water can be returned to the atmosphere
by evaporation or evapotranspiration from plants. At sites with sub-
stantial precipitation, a drainage layer is often placed in the cover. The
drainage layer serves three purposes:
1. to drain water away quickly so that the water is not available for
percolation through the cover;

Surface Layer

Fig. 5.5 Components of a cover system.

 Containment technology 105
2. to minimize the head of water on the liner, which minimizes
percolation; and
3. to reduce water pressures that could trigger instability of the cover.
Nearly all engineered cover systems (except perhaps those at arid sites)
have a barrier layer. The barrier layer is generally either a single
geomembrane, a layer of low-hydraulic-conductivity compacted soil, or
a composite geomembrane-soil liner. A new material, termed a geo-
synthetic clay liner, is experiencing increased usage in cover systems.
Many waste disposal regulations require that the barrier layer in a
cover system be no more permeable than the liner system in the
disposal facility. The reason for this requirement is to ensure that the
containment facility does not fill with liquid should the leachate col-
lection cease to be operated or malfunction. A gas collection layer is
recommended for wastes that produce significant quantities of poten-
tially harmful or dangerous gas. More details on cover systems and
their components are given in Chapters 10 and 18.
In the US, an extra component is usually added to the minimum
liner requirements for hazardous waste landfills. At most facilities, the
primary liner consists of a geomembrane/clay composite liner. For
example, one possible configuration of liner and cover system com-
ponents is shown in Fig. 5.6. The purpose of adding the clay com-
ponent to the primary liner is to minimize the flux of liquid into the
leak detection layer. The compacted clay component of the primary
liner system is shown in Fig. 5.6 to cover the floor and sidewalls of the
disposal unit. In reality, the clay is usually not placed on the sideslopes
(just on the base of the unit) because compaction of clay on a side
slope directly above geosynthetic materials can be very difficult and
risks damaging the geosynthetic components. A thin, manufactured

Cover SOIl
Drainage Layer
Geomembrane
Compacted Clay
Gas ColIectIOfl Layer
Waste

Primary leachate COUectlOfl Layer

:--- Cc~lcted Clay

leak DetectIOn Layer

Fig. 5.6 Double composite liner system and multiple-component cover system
in a waste containment unit.
106 Landfills and impoundments

clay blanket (called geosynthetic clay liner) is recommended by many

as the clay component of the primary liner. A design recommended by
Daniel and Koerner (1991) is sketched in Fig. 5.7. The geosynthetic clay
liner (discussed more in Chapter 7) has the advantages that the liner
can be placed with lightweight equipment, which minimizes risk of
damage to underlying components, and the dry clay blanket does not
tend to yield water due to consolidation (which a compacted clay liner
does when it is loaded). Consolidation water entering the leak detection
layer is usually misinterpreted as possible liner leakage.
Because composite liners are critical components of many waste
disposal units, it is appropriate to discuss in more detail the rationale
behind the composite liner. Defect-free geomembranes are practically
impermeable: water passes through an intact geomembrane via dif-
fusion at an extremely slow rate. The problem with geomembranes is
that defects, such as pinholes or tiny flaws in seams, can occur. The
size and number of holes is controlled by the construction methodology
and degree of construction quality assurance. Darilek, Laine, and Parra
(1989) used an electrical resistivity method to survey 28 completed
liners for leaks. They located 542 leaks (0-79 leaks/site) with an
average areal density of 26Ieaks/ha. Leaks were found at all sites,
except for two small geomembrane-lined tanks. Of the leaks detected,
18% were in parent material and 82% were in seams and at details
such as sumps and pipe protrusions, which also include seams.
Giroud and Bonaparte (1989a) independently evaluated leaks in
geomembrane liners. They concluded that 1 defect per 10 m of seam
can be expected for geomembranes installed without independent
quality assurance and that an average of 1 defect per 300 m of field

Individual Components:

Major System
Segments:

11~1 y Leachate CoIleclJOn System

[!! ~ Pnmary Composite Uner
~leak DetectIOn System

9:;..;....;.......-"'-i Secondary Composite Liner

Subgrade

Fig.5.7 Lining system recommended by Daniel and Koerner (1991).

 Containment technology 107

seam can be expected with reasonably good installation practice and

independent quality assurance. For typical panel widths, seam defects
are likely to result in 3-5leaks/ha with good quality assurance.
Bonaparte and Gross (1990) examined leakage rates measured in the
leak detection layer of double liner systems at actual landfills and
impoundments. In 19 geomembrane-lined landfills, leakage occurred
through the geomembrane liner in all 19 cells. Of the 6 surface im-
poundments, leakage occurred in only 2. However, with the impound-
ments, liners were leak tested by filling the cells with water prior to
operation - any leaks found during testing were repaired.
A composite liner overcomes the major problem (occasional defects)
with a single geomembrane liner. The way that a composite works is
sketched in Fig. 5.8 and is contrasted with individual geomembranes
and soil liners. If there is a hole in a geomembrane liner, liquid will
easily move through the hole, assuming the subgrade soil does not
impede seepage. With a soil liner alone, seepage takes place over the
entire area of the liner. With a composite liner, liquid moves easily
through any hole in the geomembrane but will then encounter low-
permeability soil. The low-permeability soil impedes further migration
of the liquid. Thus, leakage through a hole in a geomembrane is
minimized by placing a low-permeability soil beneath the membrane.
Similarly, leakage through a soil liner is reduced by placing it in contact
with a geomembrane which, despite occasional holes or defects in
seams, greatly reduces the area of flow through the soil liner and
thereby significantly decreases the rate of flow through the soil liner.
To achieve good composite action, the geomembrane must be placed
against the soil with good hydraulic contact (often called intimate
contact). One would normally not separate the geomembrane and soil
liner with a highly permeable material, such as a bed of sand or a
geotextile, because intimate hydraulic contact would be jeopardized
(Fig. 5.9). If stones of a size and shape that could puncture the geo-
membrane exist in the soil liner, the stones must be removed or a

Geomembrane Liner Soil Liner Comgosite Liner

" " "

-
<if: <if: s
-t- I)
Rapid Seepage Seepage
Seepage through through
through Entire Liner Small Area
Defect

Fig. 5.8 Seepage patterns through geomembrane, soil, and composite liners.
108 Landfills and impoundments

Preferred Not Recommended

Defect / Geomembrane

"""g""" 'Intimate
Contact

High-Permeability
Soil or Geotextile

Low-Permeability Soil Low-Permeability Soil

Fig. 5.9 Proper design of composite liner for intimate hydraulic contact
between geomembrane and compacted soil.

separate layer of select soil liner material must be placed in contact

with the geomembrane. To achieve intimate contact, the surface of a
compacted soil liner on which the geomembrane is placed should be
smooth-rolled with a steel-drum roller. Also, the geomembrane should
be placed and backfilled in a way that minimizes wrinkles.
Seepage rates through geomembrane liners, soil liners, and com-
posite liners may be calculated using equations published by Giroud
and Bonaparte (1989a) and Giroud et al. (1989). The following example
calculations are presented to compare calculated flow rates through
different lining systems. Assumptions made for the calculations are
that the head of liquid, h, is 300 mm; the hydraulic conductivity of the
soil liner, K, is 10- 8 cmls (best case), 10- 7 em Is (average case), or
1O- 6 cm/s (worst case); the geomembrane contains holes with an area
of 0.1 cm2; and the number of holes per hectare is 2 (best case), 20
(average case), or 60 (worst case). Computed flow rates are sum-
marized in Table 5.1.
Calculated flow rates through the composite liner are typically at
least 100 times less than through the geomembrane or soil liner alone.
Further, even if the soil liner is relatively permeable (l0-6 cm/s) and
there are 20 holeslha in the geomembrane, the calculated flow rate
through the composite liner is far less than the flow rate through
relatively good-quality soil liners or geomembrane liners on their own.
A compelling attribute of composite liners is that neither the clay soil
nor the geomembrane component needs to be constructed perfectly to
realize excellent performance of the composite liner. This fact is critical
to designers who cannot always be assured that the liner will be built
to the desired high standards.
The performance of composite liners to date has generally been very
good. Bonaparte and Gross (1990) report leakage rates measured in
 Containment technology 109
Table 5.1 Example calculations of flow rates

type of liner flow rate (Uhalday)

best case average case worst case

Geomembrane alone 2500 25000 75000

holes/ha 2 20 60
Compacted soil alone 115 1150 11500
K (cm/s) 10- 8 10- 7 10- 6
Composite 0.8 47 770
holes/ha 2 20 60
K (cm/s) 10- 8 10- 7 10- 6
contact poor poor poor

leak detection layers for double-liner systems in landfills and impound-

ments. Analysis of the data is complicated by the fact that most, if not
all, of the liquid collected initially in a leak detection system beneath a
composite primary liner is the result of consolidation of the clay-liner
component of the primary composite liner. For example, if a 0.6 m
thick layer of saturated soil compresses 3% in thickness over a period
of two years, the average flow rate due to consolidation would be
270Llha/day, which is likely to be far greater than the long-term leak
rate.
Bonaparte and Gross (1990) report that geosynthetic clay liners
(GCLs) were used as the lower component of the primary composite
liner in seven liner systems. For these systems, there was no consolida-
tion water produced and interpretation of the leak rate through the
composite liner was unambiguous. No flow was detected in the leak
detection system of any of the seven composite liners. For this reason,
many designers in the US are beginning to employ a geomembrane/
geosynthetic clay liner as a composite liner for a double liner systems.
One difficulty with composite liners is interfacial shear between the
geomembrane and soil liner. The interfacial shear resistance at this
surface is often low and can be lowered even more if positive water
pressure develops at the interface. Engineers should measure the
geomembrane-soil friction angle with adequately large shear boxes
and employ appropriate geotechnical methods for slope analysis (see
Chapter 11). Anchorage may be needed for the geomembrane, or
operational constraints during filling of a waste cell may be necessary
to minimize development of shear stresses along the interface.
Another potential problem is that the temperature of an exposed
geomembrane reach up to 65°C during summer days. This high tem-
perature can dry out and crack the underlying clay liner. When the
geomembrane cools, moisture can condense beneath the membrane
and produce a thin layer of water at the clay/geomembrane interface.
110 Landfills and impoundments
Condensate can migrate down-slope and pool at the toe of a slope.
The recommended solution is to cover the geomembrane as soon as
possible.

5.5 DISPOSAL UNIT

Nearly all landfills (except perhaps those that contain huge quantities
of mining wastes) consist of several units or cells rather than one
massive landfill. The reasons for dividing the disposal area into waste
containment units or cells are many. For the owner, the process of
obtaining a waste disposal permit can be lengthy and expensive.
Development of a site for waste disposal is generally not worth the
owner's effort unless permission can be granted to dispose of waste at
a site for many years. In the US, waste containment facilities with a
double composite lining system are constructed at a cost on the order
of $150,000 to $300,000 per acre ($370,000 to $740,000 per ha). To limit
the initial financial investment, it is wise to develop a portion of the
site initially rather than the whole site. Also, construction logistics
favor tractable disposal units rather than one massive landfill. In
addition, some of the components of a landfill cannot be exposed to
extreme weather conditions or ultraviolet light for extended periods;
the simplest way to deal with these problems is to cover the materials
as soon as possible, which is a practice that favors many small cells
rather than one massive landfill.
Landfills that are at least partly below ground usually contain an
earthen access ramp to allow trucks and other traffic to drive down into
the disposal cell. The ramp may be placed on top of the liner system,
or the liner system (all or part) may be placed on top of the ramp. An
access road leads to the earthen ramp. One of the major forces that can
act on a lining system is the force from a heavy truck that brakes while
moving down the earthen ramp. It is beneficial to place a curve in the
access road, just before the earthen ramp that leads down into a
disposal cell, to ensure that trucks slow down before entering the
disposal cell. The curve should ideally be designed to allow the truck
driver to get a good view of the ramp prior to driving onto the ramp.

5.6 OPERATIONS

A number of operational details have to be carefully planned from a

geotechnical standpoint. The first layer of waste placed on top of a
leachate collection layer is usually select waste. Select waste is not so
much waste that is selected to contain controlled material as it is waste
that is screened to make sure that it does not contain large objects that
could puncture a underlying liner or contain fine materials (like indus-
 References 111
trial sludges) that could plug the leachate collection system. In a few
landfills, the first layer of waste is a genuine select waste: shredded
tires. The tires provide very large hydraulic conductivity and protect
the leachate collection system and underlying liner from puncture.
The waste must be placed in a way that does not create shearing
stresses that could trigger instability. The slippage at the Kettleman
Hills facility (Seed et al., 1990) occurred in part as a result of the way in
which wastes were placed. The facility was located on a gentle slope in
a valley (such units are often called valley landfills), and the facility
was filled in a way that created a large mound of waste and a
significant slope. Sometimes the stability of a facility may hinge on the
way in which the disposal cell is filled. If so, the filling procedures
must be spelled out clearly by the designers.
Daily cover serves a critically important function of keeping insects
and rodents out of the waste. One problem that daily cover can create
is hydraulic isolation of one cell from another. If the daily cover
consists of relatively impermeable soil, water cannot migrate uniformly
through the waste. Instead, water will be channeled in the landfill.
Some cells may be saturated with water and others may be virtually
dry. Wide variation in moisture conditions leads to problems with
differential settlement and leachate collection. If leachate will be rein-
troduced to the disposal unit (leachate recirculation), it is particularly
important that daily cover has a high hydraulic conductivity. A
common manifestation of low-hydraulic-conductivity daily cover is the
appearance of leachate seeps on landfill covers: leachate flows laterally
along the surface of daily cover rather than infiltrating downward,
until the leachate 'daylights' on the sloping cover of a landfill.
In cold climates, it is sometimes necessary to spread waste over the
surface of the leachate collection layer to provide adequate depth of
cover over an underlying compacted clay liner to prevent the clay liner
from freezing. Designers of the leachate collection layer must bear in
mind the need to drain not only leachate but precipitation that falls in
the cell prior to placement of waste. For large cells, it is sometimes cost
effective to divide the leachate collection layer into separate areas so
that rainfall that falls in an area where waste has not yet been placed
can be collected and released without the need for treatment.
An important operations consideration is the cleanout of leachate
collection pipes. If pipes will be cleaned, the design must accommodate
access points for the cleanout equipment.

REFERENCES

Bonaparte, R. and Gross, B.A. (1990) Field behavior of double liner systems, in
Waste Containment Systems: Construction, Regulation, and Performance, (ed. R.
Bonaparte), American Society of Civil Engineers, New York, pp. 52-83.
112 Landfills and impoundments

Daniel, D.E. and Koerner, R.M. (1991) Landfill liners, Civil Eng., 61(12), 46-9.
Darilek, G.T., Laine, D.L. and Parra J.O. (1989) The electrical leak location
method for geomembrane liners: development and applications,
Geosynthetics '89, Industrial Fabrics Association International, Minneapolis,
Minn., 2, 456-66.
Giroud, J.P. and Bonaparte, R. (1989a) Leakage through liners constructed with
geomembranes - Part I: geomembrane liners, Geotextiles and Geomembranes,
8,27-67.
Giroud, J.P. and Bonaparte, R. (1989b) Leakage through liners constructed
with geomembranes - Part II: composite liners, Geotextiles and
Geomembranes, 8, 71-111.
Giroud, J.P., Khatami, A. and Badu-Tweneboah, K. (1989) Evaluation of the
rate of leakage through composite liners, Geotextiles and Geomembranes, 8,
337-40.
Seed, R.B., Mitchell, J.K. and Seed H.B. (1990) Kettleman hills waste landfill
slope failure II: stability analyses, J. Geotech. Eng., 116(4), 669-90.
 CHAPTER 6

Leachate and gas generation

Morton A. Barlaz and Robert K. Ham

6.1 INTRODUCTION

The generation of leachate and gas from landfills is a well documented

phenomenon which impacts landfill design and operation. Leachate
and gas production are both influenced by the contents of the landfill,
the local climate, and the manner in which the facility is operated. Gas
composition and volume will be discussed in this chapter as will
leachate composition. The quantity of leachate produced from a landfill
is a function of site surface hydrology, which in turn is a function
of climate, the presence of groundwater and the numerous factors
affecting infiltration of surface water into the landfill. Leachate quantity
is discussed in a later chapter.
The central theme for this chapter is leachate and gas generation
from landfills containing municipal solid waste and other non-
hazardous waste. This theme is consistent with present US regulations
governing the generation, treatment and disposal of wastes classified
as hazardous. Under present regulations, the burial of untreated
hazardous wastes in landfills is almost completely prohibited, particu-
larly the burial of materials in a physical state from which they could
exert a significant influence on gas and leachate generation. Nonethe-
less, it must be recognized that there are numerous landfills in which
hazardous wastes were buried together with municipal solid wastes
in the years before promulgation of the aforementioned regulations.
Hazardous wastes also enter municipal solid waste landfills from
unregulated small quantity sources. In addition, co-disposal of
municipal solid wastes with specific hazardous wastes is permitted in
other countries and is considered desirable by some. While the dis-
cussion to follow will concentrate on gas and leachate as produced

Geotechnical Practice for Waste Disposal.

Edited by David E. Daniel.
Published in 1993 by Chapman & Hall, London. ISBN 0 412 35170 6
114 Leachate and gas generation

Table 6.1 Composition of municipal refuse by

component

component* percent by wet weightt

Paper 40.0
Yard 17.6
Metal 8.S
Plastics 8.0
Food 7.4
Glass 7.0
Other 11.6t

* As reported by EPA (1990) for municipal solid waste in

1988.
tThe moisture content of refuse with this composition,
based on typical moisture contents for each component of
the refuse, is 19.8% (wet weight) (Tchobanoglous et aI.,
1977).
tlncludes rubber, leather, textiles, wood, and miscellaneous
wastes.

from municipal solid waste landfills, comments on the impacts of

hazardous waste will be inserted as appropriate.

6.1.1 Refuse composition

Traditionally, municipal refuse has been classified according to sortable
categories such as glass, paper, metals, etc. (Table 6.1). While such
data are needed for recycling studies and overall solid waste manage-
ment planning, data on the chemical composition of refuse are more
directly applicable to a discussion of refuse decomposition and its
impact on gas and leachate characteristics. Recently, data on the
chemical composition of refuse have been published and these data
have been used to calculate a methane potential for each chemical
constituent (Barlaz et a!., 1989a). Chemical constituent data are
presented in Table 6.2. The data in Table 6.2 show that the principal
biodegradable constituents of refuse are cellulose and hemicellulose.
These two constituents account for 91 % of the methane potential of
refuse (Barlaz et aI., 1989a). The remainder of the methane potential of
refuse consists of protein (8.3%), and soluble sugars (0.5%). Lignin,
the other major organic component of refuse, does not decompose to
any significant extent under anaerobic conditions (Young and Frazer,
1989). Knowing that refuse contains a significant biodegradable fraction,
one can begin to understand the manner in which refuse decomposes
and how this affects leachate and gas production.
 Landfill microbiology 115

Table 6.2 Composition and methane potential of

municipal refuse by chemical constituent

chemical % dry methane

constituent weight* potentialt

Cellulose 51.2 73.4

Hemicellulose 11.9 17.1
Protein+ 4.2 8.3
Lignin 15.2 0
Starch 0.5 0.7
Pectin§ <3.0
Soluble sugars 0.35 0.5

Total volatile
solidsll 78.6 ~

* Adopted from Barlaz (1988).

t Data expressed as a percentage of the total methane
potential based on the cellulose, hemicellulose, protein,
sugar and starch data. The methane contribution of pectin
was nut calculated because of the uncertainty associated
with the pectin concentration in refuse. Methane potential
was calculated from the stoichiometry given by Parkin
and Owen (1986) on the basis of 100'},0 conversion of a
constituent to carbon dioxide and methane. It should be
recognized that 100% of the degradable constituents will
not be decomposed as some fraction is surrounded by
lignin and is not accessible for anaerobic degradation.
t Determined by multiplication of the total Kjeldahl
nitrogen by 6.25. The actual protein content of refuse is
probably lower than the value given here because some of
the Kjeldahl nitrogen is actually nitrogen contained in
humic materials and structural proteins which are not
easily degradable (Burns and Martin, 1986).
§ Actual value is probably less than 3% but could not be
quantified.
II An independent measure of volatile solids based on
weight loss on ignition at 550°C. Not a total of previous
entries in the table.
~I The volatile solids concentration is presented to illustrate
that the chemical constituent analyses account for 110%
of the volatile solids. No methane potential is calculated
for volatile solids because this measurement includes both
degradable and non-degradable carbon.

6.2 LANDFILL MICROBIOLOGY

The decomposition of refuse to methane in sanitary landfills is a

microbially mediated process which requires the coordinated activity of
several trophic groups of bacteria. As discussed above, the principal
substrates which decompose to methane in landfills are cellulose
116 Leachate and gas generation
and hemicellulose. Similar substrates fuel the production of methane
in other ecosystems including the rumen, marshes, rice paddies and
sludge digesters (Wolfe, 1979). In the first part of this section the
general pathway for anaerobic decomposition, as it has been docu-
mented to occur in other anaerobic ecosystems, is reviewed. Implica-
tions of this pathway for sanitary landfills are also presented. Following
this general pathway review, refuse decomposition is characterized in
four phases defined by different gas and leachate characteristics.

6.2.1 The microbiology of anaerobic decomposition

Three trophic groups of anaerobic bacteria are required for the produc-
tion of methane from biological polymers (cellulose, hemicellulose, and
protein) as illustrated in Fig. 6.1 (Wolfe, 1979; Zehnder et al., 1982). The
first group of microorganisms, referred to as the Hydrolytic and
Fermentative microorganisms in Fig. 6.1, is responsible for the
hydrolysis of biological polymers. The initial products of polymer
hydrolysis are soluble sugars, amino acids, long chain carboxylic acids
and glycerol. Hydrolytic and Fermentative microorganisms then fer-
ment these initial products to short-chain carboxylic acids, alcohols,
carbon dioxide and hydrogen. Acetate, a direct precursor of methane is
also formed. The second group of bacteria active in the conversion
of biological polymers to methane is the obligate proton-reducing
acetogens. They oxidize the fermentation products of the first group of
microorganisms to acetate, carbon dioxide and hydrogen. The con-
version of fermentation intermediates like butyrate, propionate and
ethanol is only thermodynamically favorable at very low hydrogen
concentrations. Thus, these substrates are only utilized when the
obligate proton-reducing acetogenic bacteria can function in syntrophic
association with a hydrogen scavenger such as a methane producing
or sulfate reducing organism. The importance of acetogenic bacteria
which convert hydrogen plus carbon dioxide to acetate has not been
established in the refuse ecosystem. This microbiological activity
probably competes weakly with the hydrogenophilic methanogens for
hydrogen.
The third group of bacteria necessary for the production of methane
are the methanogens. The methanogens can utilize only a limited
number of substrates including formate, methanol, methylarnines,
hydrogen plus carbon dioxide, and acetate (Wolin and Miller, 1985). In
sludge digesters, it is estimated that 70% of the methane produced
originates from acetate (Zeikus, 1980; Mah et al., 1978). This value has
not been investigated in the landfill ecosystem. The production of
methane from acetate yields only 31 kJ/mole CH4 produced. This is
barely enough energy for the generation of adenosine triphosphate
 Landfill microbiology 117

biological
polymers
hydrolytic and
fermentative
microorganisms

4% 20%
alcohols,
carboxylic
acids
(except acetate)
obligate
proton
reducers

Fig. 6.1 Substrate flow in anaerobic ecosystems (Zehnder et aI., 1982).

Complete multistep methanogenesis as it occurs in lacustrine and sulfate
depleted marine sediments, bogs, marshes, trees and digesting sludge
(reprinted by permission from author).

(ATP) which requires 30.6kJ/mole, so the growth of methanogens on

acetate is relatively slow. The conversion of hydrogen and carbon
dioxide to methane yields 135.6kJ/mole CH4 produced. Thus, the latter
reaction is energetically more favorable. The methanogens are most
active in the pH range 6.8 to 7.4 (Zehnder, 1978).
The importance of the methanogens in anaerobic digestion has been
summarized by Zeikus (1980). As a group, the methanogens:
1. control the pH of their ecosystem by the consumption of acetate;
2. regulate the flow of electrons by the consumption of hydrogen,
creating thermodynamically favorable conditions for the catabolism
of alcohols and acids; and
3. excrete organic growth factors including vitamins and amino acids
which are used by other heterotrophic bacteria in the ecosystem.
118 Leachate and gas generation

Should the activity of the fermentative organisms exceed that of the

acetogens and methanogens, there will be an imbalance in the eco-
system. Carboxylic acids and hydrogen will accumulate and the pH of
the system will fall, thus inhibiting methanogenesis.

6.2.2 Microbiology of refuse decomposition

When refuse is placed in a landfill, biological decomposition resulting
in methane formation as described in the previous section does not
occur immediately. A period ranging from months to years is necessary
for the proper growth conditions and the required microbiological
system to become established. Refuse decomposition phases ranging in
number from three to six or more have been identified by different
investigators depending on the data base and purposes of each study.
A four-phase characterization of refuse decomposition describing
chemical and microbiological characteristics of decomposition is sum-
marized here (Barlaz et al., 1989b). Refuse decomposition is described
in an aerobic phase, an anaerobic acid phase, an accelerated methane
production phase, and a decelerated methane production phase as
follows and as summarized in Fig. 6.2. This description is based on
data from laboratory scale lysimeters. The relationship between these
data and field scale landfills will be discussed.

Phase 1 - aerobic phase

Oxygen is unavoidably present in the void space when refuse is land-
filled. This oxygen, plus oxygen dissolved in the refuse-associated
moisture triggers aerobic decomposition. In the aerobic phase, both
oxygen and nitrate are consumed, with soluble sugars serving as the
carbon source for microbial activity. All of the trophic groups required
for refuse methanogenesis are present in fresh refuse (cellulolytics,
acetogens, and methanogens), although there is little change in their
populations. Leachate strength will be relatively low during the aerobic
phase and gas composition will be nearly 100% CO 2 ,
The aerobic phase in newly placed refuse in a full-scale landfill is
likely to last for only a few days. During the aerobic phase, refuse is
typically below field capacity so any leachate produced from freshly
buried refuse will likely have flowed through a channel in the refuse.
In addition, since refuse is likely to be buried on top of older material
and leachate is generally collected at the bottom of a landfill, leachate
characteristic of the aerobic phase may not be observed. Instead, liquid
percolating through fresh and then older refuse will reflect the charac-
teristics of the older refuse, which may be in any of the other three
phases of decomposition described below. Leachate generated from
freshly buried refuse has been analyzed from test landfills and
 Landfill microbiology 119

aerobic anaerobic accelerated decelerated

100· acid methane methane ~
100C c: ...
~ 0>-
:;::;6:>
g-"
c: 750 "O~
o
:0=
e"O
.~ c. ....
C/)
500 Ql ...
E 40 c:.l!l
<ll:=
8 250
.J::
a5
:g E
0>

0> I Q)
c: 0.
'c 100 C/)
::l
'Cij 8 '§
E .75 75
~ C/)E
"O~
§ .5 7 50 .- u
~~
!0.25 25 C/)
6 <ll

.s
0>

z
a..
::2i

o 7 2027344148 69 90 111
day

Fig. 6.2 Summary of observed trends in refuse decomposition with leachate

recycle. Gas volume data were corrected to dry gas at standard temperature
and pressure. The acids are expressed as acetic acid equivalents. Solids
remaining is the ratio of the cellulose plus hemicellulose removed from a
container divided by the weight of cellulose plus hemicellulose added to the
container initially. Methanogen MPN data are the log of the average of the
acetate- and H2 /C0 2- utilizing populations. (Reprinted from Barlaz et al., 1989b)

laboratory cells. This leachate has sometimes been observed to be

very strong, with chemical oxygen demand (COD) values of tens
of thousands to one hundred thousand mg/l. Such numbers reflect
primarily squeezings from compaction of the refuse.

Phase 2 - anaerobic acid phase

Eventually, oxygen entrained with the refuse is depleted by aerobic
microbial activity. As a given refuse mass becomes covered by other
refuse, replenishment of oxygen becomes insignificant and the system
becomes anaerobic. In the anaerobic phase, carboxylic acids accumulate,
the pH decreases, there is some cellulose and hemicellulose decom-
position, and some methane may be detected in landfill gas. The first
120 Leachate and gas generation

part of the acid phase is characterized by the rapid accumulation of

carboxylic acids and a decrease in the pH of the refuse ecosystem.
This pH decrease can be attributed to the accumulation of acidic end-
products of sugar fermentation. In addition to carboxylic acids, it is
likely that a wide variety of organic intermediates are produced. High
CO 2 concentrations, first reported by Farquhar and Rovers (1973),
result from fermentative activity, as does the accumulation of hydrogen
gas. Both CO 2 and H2 are products of sugar fermentation. No methane
is detected early in the acid phase of refuse decomposition.
Carboxylic acid concentrations increase and the pH decreases
throughout the acid phase. The data presented in Fig. 6.2 are based on
a system in which the leachate was collected from the bottom of
laboratory-scale landfills, externally neutralized and added back to the
top of the test system, i.e. the leachate was recycled. Leachate recycle
and neutralization is not typically practiced in field-scale landfills. The
acid phase explains the long lag time between refuse burial and the
onset of methane production in sanitary landfills in that low pH
inhibits refuse methanogenesis. In a field-scale landfill, acids leaching
from refuse in the acid phase of decomposition would be rapidly
consumed by refuse in decomposition phases three or four described
below.
With progression through the acid phase, the carbon dioxide con-
centration decreases as the methane concentration increases and newly
produced gas purges carbon dioxide from the landfill. Cellulose and
hemicellulose hydrolysis is not consistent in the acid phase and there is
relatively little solids hydrolysis.
Measurable methane production denotes the beginning of the third
phase of decomposition, the accelerated methane production phase.
Leachate was externally neutralized in the research reported in Fig. 6.2.
The resulting increased pH stimulated microbial activity and allowed
decomposition to proceed to the third phase. In a full-scale landfill
without leachate neutralization, it may take months to years for the
microbiological system to progress to the third phase of decomposition.
Acid concentrations may be reduced and the pH increased by slow
microbial acid consumption, or by rinsing the acids to lower levels of
the landfill or to the leachate collection system by water infiltration.
It should be stressed that the low pH which results from an acid
accumulation, and not high acid concentrations alone, limits pro-
gression of refuse decomposition to the accelerated methane produc-
tion phase.

Phase 3 - accelerated methane production phase

In the accelerated methane production phase there is a rapid increase
in the rate of methane production to some maximum value. Methane
 Landfill microbiology 121

concentrations of 50-70% are typical of this phase with the balance

of the gas being carbon dioxide. There is a decreased accumulation of
carboxylic acids in the third phase of refuse decomposition. Carboxylic
acids are consumed faster than they are produced as the methane
production rate increases. Subsequently, the pH of the refuse eco-
system increases. There is little solids hydrolysis during this phase of
decomposition (Fig. 6.2). Increases in the populations of cellulolytic,
acetogenic and methanogenic bacteria are observed.
The data presented in Fig. 6.2 show an increasing methane produc-
tion rate and decreasing carboxylic acid concentration in the accelerated
methane production phase. However, it must be remembered that
these data were collected from laboratory-scale lysimeters. Under field
conditions, the measured methane production rate may represent an
average of refuse in numerous states of decomposition. The sharp
increase and decrease illustrated in Fig. 6.2 will be considerably
dampened. Similarly, the decrease in carboxylic acid concentrations is a
relative effect. Thus, one measurement from a landfill will probably not
be sufficient to identify the state of refuse decomposition. Neverthe-
less, decreasing carboxylic acid concentrations typically correlate with
increased rates of methane production. Estimation of the methane
potential of a landfill by the use of solids analysis, in concert with
measurement of the methane production rate, will be discussed in a
later section.

Phase 4 - decelerated methane production phase

The final phase of refuse decomposition may be described as the
decelerated methane production phase. It is characterized by a decrease
in carboxylic acid concentrations to concentrations below 100 mg/I.
With the depletion of carboxylic acids there is a further increase in the
pH of the ecosystem. Though the acids are depleted, there is still some
COD exerted by the leachate. While relatively low, it is likely to consist
of less degradable compounds, including humic materials.
The rate of methane production decreases even though the methane
and carbon dioxide concentrations remain constant at about 60% and
40%, respectively. The rate of cellulose and hemicellulose decom-
position in the decelerated methane production phase is higher than
that exhibited in any other phase of refuse decomposition. Solids
hydrolysis controls the rate of methane production in that there is no
longer an accumulation of carboxylic acids to serve as soluble substrate.
In earlier phases of refuse decomposition, the hydrolysis of cellulose
and hemicellulose leads to accumulations of carboxylic acids. In the
fourth phase of refuse decomposition, where the rate of polymer
hydrolysis exceeds that exhibited earlier, no accumulations are ob-
served. The difference may be explained by the makeup of the refuse
122 Leachate and gas generation

ecosystem; specifically by increases in methanogenic and ace to genic

activity. It is interesting to observe that the rate of methane production
and the concentration of carboxylic acids decreased in parallel. Though
there has been speculation that high concentrations of carboxylic acids
inhibit methane production, acetate and butyrate concentrations of
9753 mg/l and 6956 mg/l, respectively, did not inhibit methane produc-
tion in the containers described by Fig. 6.2 (Barlaz et al., 1989c). In the
second and third phases of refuse decomposition it is the utilization of
carboxylic acids which limits the onset and rate of methane production.
After consumption of the initial accumulation of carboxylic acids, it is
polymer hydrolysis which limits the rate of methane production.
The identification and description of these four phases of refuse
decomposition were developed based on data from a laboratory experi-
ment. The characteristics of refuse decomposition described here
represent what would happen under ideal circumstances for a given
volume of newly placed refuse. The data presented in Fig. 6.2 have
time as the abscissa. The times shown on the abscissa should not be
applied to other conditions. These times were influenced by the fre-
quency of leachate recycle and neutralization and the incubation
temperature. Neither leachate recycle nor shredded refuse are typically
used in landfills. However, no external additions of bacteria were made
to the refuse used by Barlaz et al. (1989b). It is believed that the major
differences in refuse decomposition under different conditions will be
the length of time required for the different phases of refuse decom-
position to evolve, the methane production rate, and possibly the
methane yield.

6.2.3 Implications of microbial activity on gas and leachate

characteristics
Several observations concerning refuse decomposition and the landfill
ecosystem can be made based on an understanding of the groups of
organisms involved. The groups of organisms involved in methane
production obtain energy by fermentation only. Under aerobic con-
ditions, the organic polymers will be oxidized to carbon dioxide. When
present, nitrates or nitrites will serve as electron acceptors and be
reduced to nitrogen as the biodegradable components of the refuse are
converted to carbon dioxide. Like oxygen, there can be no methane
production in the presence of nitrate or nitrite. If sulfate is present
in the ecosystem, it is reduced by the sulfate reducing bacteria to
hydrogen sulfide in the presence of hydrogen. Sulfate is not reduced
by bacteria in the presence of oxygen. Hydrogen is also used by the
methanogenic bacteria to reduce carbon dioxide to methane. Competi-
tion between sulfate reducing and methanogenic bacteria for hydrogen
 Landfill microbiology 123

has been reported for other ecosystems but appears less important in
the refuse ecosystem (Barlaz et al., 1989b).
The methanogens are most active in the pH range 6.8 to 7.4
(Zehnder, 1978). Should the activity of the fermentative organisms
exceed that of the acetogens and methanogens, there will be an
imbalance in the ecosystem. Carboxylic acids and hydrogen will
accumulate and the pH of the system will fall, thus inhibiting
methanogenesis. The potential for a rapid accumulation of acids is
particularly acute in the refuse ecosystem. Food and garden wastes
make up about 26% of the solid waste stream (Table 6.1). These
materials contain soluble sugars. In addition, there will be some
spontaneous lysis of plant cells under anaerobic conditions. This lysis
results in the release of additional soluble sugars (Greenhill, 1964a,b,c).
These sugars will be converted to carbon dioxide in the presence of
oxygen or nitrate, or carboxylic acids in the absence of oxygen and
nitrate.
Based on the measured concentrations of soluble sugar (3.46 mg/dry
gm) and nitrate (0.015 mg/dry gm) in refuse, and an assumed porosity
of 25%, Barlaz et al. (1989a) calculated that the amounts of oxygen and
nitrate present in freshly buried refuse were sufficient to support the
oxidation to carbon dioxide of only 10% of fresh refuse soluble sugars.
This means that 90% of the sugars are fermented to carboxylic acids
which accumulate, reducing the pH of the refuse ecosystem and in-
hibiting methanogenesis.
When the fermentative bacteria are active, large quantities of soluble
organics are produced which may percolate from the landfill in
leachate. As the methanogens become active, these soluble organics
are metabolized. Thus, the chemical oxygen demand (COD) or total
organic carbon (TOC) of landfill leachate is typically highest before the
onset of steady state methanogenesis (Ham and Bookter, 1982; Barlaz et
al., 1989b). Essentially all organic chemicals exert a COD. Prior to the
decelerated methane production phase carboxylic acids account for
60% to 90% of the COD of leachate. Thus, high organic acid concen-
trations and high COD may be considered equivalent. The high
organic acid concentrations plus dissolved carbon dioxide causes acidic
pH levels in leachate. This in turn causes more dissolution of inorganic
constituents. A further cause of high inorganic concentrations is com-
plex formation between organic and inorganic constituents, holding
higher concentrations of some metals in solution than would otherwise
be expected.
124 Leachate and gas generation

6.3 THE RELATIONSHIP BETWEEN LABORATORY AND

FIELD-SCALE DATA ON LEACHATE
AND GAS CHARACTERISTICS

6.3.1 Gas production and characteristics

There is general agreement that landfill gas will contain between 50%
and 70% methane with carbon dioxide and trace constituents making
up the balance. Similar results have been reported from both field and
laboratory-scale tests. Trace constituents which have been identified in
landfill gas include light petroleum compounds (benzene, toluene,
xylene), chlorinated compounds, hydrogen, and hydrogen sulfide.
Vinyl chloride, a known carcinogen, has also been detected in landfill
gas. While the volume of trace constituents is seemingly insignificant,
they may present a problem for downstream gas processing equip-
ment. In some landfills, concentrations of hydrogen as high as 20%
have been observed, indicative of phase two decomposition some-
where in the landfill. This is generally a transient condition, and once
methane production is established little hydrogen should be observed.
Comparison of methane production rate data between field-scale
landfills and laboratory experiments is difficult because there is essen-
tially no data in the open literature on methane production rates in
field-scale facilities. Data from field-scale landfills is complicated by
questions regarding the volume or mass of refuse responsible for the
production of a measured volume of gas. There is a wider body of
methane production data collected under laboratory conditions. How-
ever, the laboratory data are not perfectly comparable in that experi-
mental conditions (moisture, particle size, temperature, etc.) are not
uniform between studies. In addition, most laboratory experiments
were conducted to explore techniques for enhancing methane produc-
tion. The enhanced methane production rates would not be expected at
field-scale landfills until certain of these techniques are employed in
the field.
Methane yields of 42-120 liters CH4/dry kg refuse have been
reported in laboratory tests conducted with leachate recycle and
neutralization (Buivid et al., 1981; Barlaz et al., 1987; Kinman et al.,
1987; Barlaz 1988). These studies show significant variation in methane
production rate and methane yield. Certain of the differences can be
explained by differences in experimental design. For example, the data
reported by Barlaz et al. (1987) and Barlaz (1988) differ in reactor
volume (100 vs. 2 liters), temperature (25°C vs. 41°C) and the rate of
leachate recycle. Buivid et al. (1981) used refuse with an abnormally
high paper content.
Methane yields were measured in field-scale tests cells as part of the
Controlled Landfill Project in Mountain View California (Pacey, 1989).
 The relationship between laboratory and field-scale data 125

Yields of 38.6-92.2 liters CH4/dry kg refuse were measured after 1597

days of monitoring. However, mass balance data suggest that signifi-
cant volumes of methane were not measured in certain test cells. The
value of 92.2liters CH4/dry kg refuse is for a cell with buffer addition
only and appears to be accurate based upon good correlation between
the measured volume and mass balance data. The measured yield in
the control cell was 76.6 liters CH4/dry kg refuse and it is likely that not
all methane produced was actually measured. Biochemical methane
potential tests indicated a methane potential of 92.2 liters CH4 @ STP/
dry kg refuse remained in the control cell. Thus, the total yield for the
control cell may be as high as 168.8liters CH4/dry kg refuse (92.2 +
76.6) if decomposition proceeds to completion.
A number often used as an estimate of methane production rate in
field scale landfills is 0.1ft3 CH 4/wet lb-yr. Assuming refuse buried at
20% moisture, this converts to 7.8 liters CH4/dry kg-yr, a number com-
parable to some of the lower values reported in the literature.

6.3.2 Leachate composition characteristics

Concentrations of leachate constituents can vary by orders of mag-
nitude between different field-scale landfills. This has been illustrated
by Halvadakis et al. (1983). For example, the COD at twelve sites varied
from 0 to 89520 mg/I. The composition of the soluble phase of the
refuse ecosystem is influenced by the state of refuse decomposition as
explained in a previous section. However, leachate produced from a
field-scale landfill is not likely to be representative of refuse in one state
of decomposition, but rather is more representative of an average
leachate derived from several different refuse cells which may be in
different states of decomposition. The wide range of concentrations
makes the design of on-site leachate treatment facilities difficult. It
is thus important to provide storage for flow and leachate strength
equalization.
Perhaps the most comprehensive compilation of leachate quality
data applicable to full-scale landfills was provided by Ehrig (1988). He
obtained leachate composition data from 15 West German landfills
ranging in age from 0 to 12 years. These data are summarized in Tables
6.3 and 6.4. Wide ranges in concentration data are apparent. Con-
stituents for which the solubility is relatively unaffected by pH, such as
chlorides, show no clear shift in going from acidic leachate, character-
istic of the anaerobic acid phase, to the methane producing phases.
Constituents which are more soluble under acidic conditions, such as
iron, show a clear decrease in concentration in progressing from
the acid phase to the methane producing phases. Ehrig (1988) also
measured the COD in numerous landfill leachates. He reported wide
variation in COD concentrations (700-30000 mg/l) for landfills less
126 Leachate and gas generation
Table 6.3 Leachate analysis (parameters with
differences between acetic and methanogenic phase)
after Ehrig (1988) (Reprinted by permission of
5pringer-Verlag)

average range

acetic phase
pH(-) 6.1 4.5 - 7.5
BODs (mg/I) 13000 4000 -40000
COD (mg/I) 22000 6000 -60000
BODs/COD (-) 0.58
504 (mg/l) 500 70 - 1750
Ca (mg/I) 1200 10 - 2500
Mg (mg/I) 470 50 - 1150
Fe (mg/I) 780 20 - 2100
Mn (mg/l) 25 0.3 - 65
Zn (mg/l) 5 0.1 - 120
methanogenic phase
pH(-) 8 7.5 - 9
BODs (mg/I) 180 20 550
COD (mg/I) 3000 500 - 4500
BODs/COD (-) 0.06
504 (mg/I) 80 10 - 420
Ca (mg/I) 60 20 600
Mg (mg/l) 180 40 - 350
Fe (mg/I) 15 3 - 280
Mn (mg/l) 0.7 0.03- 45
Zn (mg/l) 0.6 0.03- 4

than eight years old. As the biological and mass transport systems in
the landfills 'matured', the COD stabilized at about 8000 mgll. This
occurred at about eight years.

6.3.3 Impact of hazardous wastes on leachate

and gas characteristics
There are a wide variety of inorganic and organic wastes which are
classified as hazardous. Thus it is only possible to make general state-
ments on potential effects of hazardous wastes on leachate and gas
characteristics. During the aerobic phase, there is an increase in tem-
perature due to the waste heat of microbial metabolism. This, com-
bined with vigorous CO 2 production during the aerobic and initial part
of the anaerobic acid phase will enhance gas stripping of volatile
chemicals.
Gas stripping does not necessarily mean release of an organic com-
pound to the environment. As volatile organic compounds pass through
soil separating the refuse from the atmosphere, there is an additional
 The relationship between laboratory and field-scale data 127

Table 6.4 Leachate analysis (no difference between

phases could be observed) after Ehrig (1988)
(Reprinted by permission of Springer-Verlag)

average range
----,-._-

Cl (mgll) 2100 100 - 5000

Na (mg/l) 1350 50 - 4000
K (mg/l) 1100 10 - 2500
alkalinity (mg CaC0 3 /1) 6700 300 -11500
NH4 (mg Nil) 750 30 - 3000
orgN (mg nil) 600 10 - 4250
total N (mgN/I) 1250 50 - 5000
N0 3 (mg Nil) 3 0.1- 50
N0 2 (mg Nil) 0.5 0 - 25
total P (mg P/I) 6 0.1- 30
AOX (Ilg Cl/I)* 2000 320 - 3500
As (Ilg/l) 160 5 - 1600
Cd (Ilg/l) 6 0.5- 140
Co (Ilg/l) 55 4 - 950
Ni (Ilg/l) 200 20 - 2050
Pb (Ilg/l) 90 8 - 1020
Cr (Ilg/l) 300 30 - 1600
Cu (Ilg/l) 80 4 - 1400
Hg (Ilg/l) 10 0.2- 50

* adsorbable organic halogen

opportunity for degradation by soil microorganisms. Whether this is a

significant sink for volatile organics buried in landfills has not been
studied. In addition to gas stripping, there are other potential fates of
organics in landfills. Refuse in an active state of methane production
represents a healthy ecosystem which may be capable of decomposing
substrates other than carbohydrates and proteins. Thus some hazardous
organics may be anaerobically degraded in sanitary landfills although
there is very little data in this area. The existing data do show that
phenol is readily degraded by refuse microorganisms. Watson-Craik
and Senior (1989) reported on the conversion of phenol to methane in
laboratory-scale lysimeters. Leachate recycle was necessary for complete
degradation of the 188 mg/l phenol solution. Halogenated solvents
such as perchloroethylene and trichloroethylene have been shown
to undergo a biologically mediated reductive dehalogenation reaction
under methanogenic conditions (Fathepure et al., 1987). One end-
product of reductive de halogenation is vinyl chloride. The presence of
vinyl chloride in landfill gas suggests that chlorinated solvents were
previously buried in the landfill. Organics may also preferentially sorb
to the non-degraded organic fraction of the decomposed refuse, thus
reducing or eliminating transport.
128 Leachate and gas generation

The pH of refuse decreases during the anaerobic acid phase and this
is likely to increase metals solubility and therefore metals mobility.
However, this does not necessarily mean that metal concentrations
will increase in leachate. Once refuse is in the accelerated methane
production phase, indicative of active anaerobic metabolism, sulfate
is reduced to hydrogen sulfide. Most metal sulfides are extremely
insoluble and their formation would reduce metal mobility. As refuse
decomposition proceeds to the decelerated methane production
phase, humic materials are produced. These materials can be expected
to behave as natural chelating agents, enhancing metal mobility.
Carboxylic acids also act as chelating agents (Francis and Dodge, 1986).
Pohland and Gould (1986), in an evaluation of the co-disposal of metal
sludges with municipal solid waste, reported an increase in metal
concentrations in leachate as the refuse became well decomposed.

6.4 FACTORS LIMITING THE ONSET

OF METHANE PRODUCTION

Research on the decomposition of solid waste in sanitary landfills

was first reported by Merz and Stone (1962). Since then, numerous
researchers have tried to enhance refuse methanogenesis by manipula-
tion of the landfill ecosystem (Barlaz et al., 1990). Many factors
influence the onset and rate of methane production including moisture
content, pH, nutrient concentrations, and temperature among others
(Farquhar and Rovers, 1973). It is important to note that all of the
microorganisms necessary for the conversion of refuse to methane
have been detected in fresh refuse (Barlaz et al., 1989b). As a result of a
review of data on both laboratory- and field-scale tests, it appears
that the two variables most important to refuse methanogenesis are
moisture content and pH.
The moisture content of fresh refuse ranges from 15% to 45% and is
typically about 20% on a wet weight basis. While there is no definitive
answer as to either the minimum or optimal moisture content required
for refuse decomposition, 20% is clearly low. Wujcik and Jewell (1980)
studied the effect of moisture content on the batch fermentation of
wheat straw and dairy manure - compounds which are analogous in
chemical composition to refuse. Methane yields decreased at moisture
contents below 70% and the yield at 30% moisture was 22% of the
yield at 70% moisture. In many laboratory-scale studies on the effects
of moisture on refuse decomposition, methane yields have been too
low for quantification of yield as a function of moisture content.
The broadest data sets where moisture content can be evaluated are
those of Emberton (1986) and Jenkins and Pettus (1985). Emberton
 Factors limiting the onset of methane production 129

evaluated methane production rate data for landfills across the US and
categorized the landfills based on annual precipitation. Jenkins tested
the effect of moisture content in refuse sampled from landfills. In both
studies, the methane production rate exhibited an upward trend with
increasing moisture contents; confounding factors such as density,
refuse age and refuse composition notwithstanding.
A second key factor influencing the rate and onset of methane pro-
duction is pH. As discussed in the section on anaerobic microbiology,
microorganisms responsible for the conversion of refuse to methane
are quite sensitive to pH. Their pH optimum is between 6.8 and 7.4
and methane production rates decrease sharply at pH values below
about 6.5. Theoretically, refuse pH is an excellent indicator of potential
methanogenic activity. Unfortunately, as discussed previously, leachate
composition data at a field-scale landfill may well not be indicative of
the decomposition state of the refuse. Leachate with a pH below 5 has
been observed in landfills actively producing methane. Such leachate
has most likely been in contact with some refuse in the anaerobic acid
phase as well as refuse actively producing methane. Hence, leachate
pH is not always a useful indicator of methanogenic activity.
Leachate recycle and neutralization has been shown to enhance the
onset and rate of methane production (Pohland, 1975; Buivid et al.,
1981; Barlaz et al., 1987). Given that moisture and pH are reported to
be the two most significant factors limiting methane production, the
stimulatory effect of leachate recycle and neutralization is logical.
Recycling neutralized leachate back through a landfill increases refuse
moisture content, substrate availability, and provides a degree of
mixing in what may otherwise be an immobilized batch reactor.
Neutralization of the leachate provides a means of externally raising
the pH of the refuse ecosystem. There is limited field experience with
leachate recycle systems and more is needed to fully document its
value in a field-scale situation.
Moisture content is the factor which most often limits methane pro-
duction. This is to be expected in dry climates where there is little
opportunity for infiltration, and in wet climates where biological activity
may be limited because landfills are typically designed to minimize
water infiltration. Other conclusions have been reached in laboratory-
scale experiments where moisture content was not limiting (Barlaz
et al., 1990). Prior to the onset of methane production, the rate at which
the methanogenic bacteria convert acetate to methane limits the onset
of methane production. In laboratory-scale work, seeding fresh refuse
with anaerobically decomposed refuse reduced the time to the onset of
methane production significantly (Barlaz et ai., 1987). With the onset of
methane production and depletion of the soluble substrates, the rate at
which polymers (cellulose, hemicellulose and protein) are hydrolyzed
limits the rate of methane production.
130 Leachate and gas generation

6.4.1 Regulatory factors influencing methane production

Another factor limiting the onset and rate of methane production is
regulatory policy and philosophy. Historically, municipal solid waste
was dumped in holes and occasionally covered with whatever material
was available. This often led to groundwater contamination and the
release of contaminated stormwater to surface water. Subsequently,
states began to regulate the burial of municipal solid waste, bringing
about the modern sanitary landfill, complete with liners and leachate
collection as described in Chapter 1.
Sanitary landfill design standards were developed to minimize the
amount of moisture which came in contact with refuse, thus minimizing
leachate production. However, at the time that the philosophy of a dry
landfill was adopted, methane recovery for energy was in its infancy.
The design of landfills to enhance methane production, by allowing the
moisture content of fresh refuse to increase by surface water infiltra-
tion, was not given serious consideration. More recently, some solid
waste regulations recognize that leachate recycle and neutralization
for enhancement of methane production may be advantageous. In
addition to energy recovery, enhanced methane production and
recovery as an energy source offers other advantages.
1. With the onset of methane production, there is a reduction in
leachate COD, thus reducing leachate treatment costs and the
potential for groundwater contamination.
2. Refuse settles as it decomposes and the resulting settlement
necessitates significant maintenance of the landfill cover to repair
cracks. More complete refuse decomposition prior to placement
of the final cover would minimize settlement and long term main-
tenance costs.
3. Enhanced methane production would make energy recovery projects
more economical, thus more would be implemented. This would
reduce emissions of methane to the atmosphere. Methane is a gas
which has a greenhouse effect at least twenty times more damaging
than carbon dioxide on a volume basis.
4. Enhanced refuse decomposition reduces the impact of future leachate
and gas emissions, most likely reducing long-term monitoring and
care requirements.
5. Enhanced decomposition means that most of the gaseous and
leachate products of decomposition are released during the period
when the gas and leachate control systems are most likely to func-
tion as designed, and when responsible parties are present to
monitor and repair the site as necessary.
Landfills designed and operated in a mode of leachate recycle and
neutralization may be recognized as more desirable in the future. It will
 Mass balance analysis of refuse decomposition 131

then be necessary to design landfills not only with liners and leachate
collection systems similar to those used today, but also with systems
for distributing leachate over refuse without creating odor or side slope
seepage problems.

6.5 MASS BALANCE ANALYSIS OF REFUSE DECOMPOSITION

A knowledge of the chemical composition of refuse remaining in a

landfill at any time makes it possible to estimate the volume of methane
which may yet be produced, remaining reserves of degradable matter,
and the potential for loss of refuse materials and settlement as the
refuse decomposes. The methane potential of refuse, combined with
an estimate of the current rate of methane production as measured in a
pump test, is useful for evaluation of the economics of landfill gas
recovery projects.
The mass of methane which would be produced if all of a given
constituent were converted to carbon dioxide, methane and ammonia
may be calculated from Eq. (6.1) (Parkin and Owen, 1986).

CHObN + ( n -4 3C)
a- - -b+ - HO~
na c 242

(-n2- 8
a- + -b +-3C) CO2 + (n- + a- - -b - -3C) CH4 + cNH3
48 2848
(6.1)

Cellulose (C6H100S) and hemicellulose (C SH S04) comprise 91 % of the

methane potential of fresh refuse (Table 6.2). These two constituents,
along with carboxylic acids will comprise at least 90% of the methane
potential of refuse in virtually any state of decomposition. Based on Eq.
(6.1), 415 and 424 liters of methane at S.T.P. would be expected for
every kilogram of cellulose and hemicellulose degraded, respectively.
The methane potential of carboxylic acids can be calculated from
the reactions governing the conversion of valerate to propionate and
acetate, butyrate and propionate to acetate and hydrogen; hydrogen
and carbon dioxide to methane; and acetate to methane and carbon
dioxide (McInerney and Bryant, 1981). Calculated methane potentials
are 720.8, 643.7, 537.0 and 373.0 liters CH4/kg of valerate, butyrate,
propionate and acetate, respectively.
Barlaz et al. (1989a) performed mass balances on shredded refuse
incubated in laboratory-scale lysimeters with leachate recycle. Carbon
recoveries of 64-111 % were obtained. After correcting for explainable
error, this range improved to 87-111 %. Mass balances were useful
for documenting the decomposition of specific chemical constituents
and demonstrating the importance of cellulose and hemicellulose to
methane production.
132 Leachate and gas generation

Mass balances may be used to estimate the methane potential

remaining in a landfill by sampling the refuse, performing the appro-
priate chemical analyses and calculating the methane potential. Ideally,
the chemical composition and methane potential of the refuse at burial
would also be known, in which case comparison of the initial methane
potential of the refuse with that at the time of sampling will provide
information on the fraction of the refuse which has been degraded. It
must be recognized that representative sampling of a full-scale sanitary
landfill is not realistic; however, it is possible to obtain multiple
samples at apparently representative locations within the landfill and
get an idea of the range and extent of loss of decomposable materials.
Samples should be as large as can reasonably be handled and be
reduced by proven techniques.

6.5.1 The final state of refuse composition

As described here, the calculated methane potential assumes complete
loss of degradable organics. Some fraction of these constituents is
surrounded by lignin and not readily available for anaerobic decom-
position. Thus, the actual methane potential of a refuse sample will be
less than that calculated. Bookter and Ham (1982) measured cellulose
concentrations as low as 8.2% for nine year old shredded refuse which
was well decomposed as evidenced by its humus like appearance. As
part of this same study various landfills around the country were
sampled and cellulose concentrations as low as 6.6% were reported
for non-shredded refuse. Refuse sampled from a landfill in Wisconsin
had cellulose concentrations of 10.0%, 6.1 % and 3.1 % for refuse
buried in 1948, 1954 and 1957, respectively. Kinman et al. (1989) found
between 1.6% and 62.0% paper in 20 year-old refuse excavated from
the Mallard North landfill in Chicago, IL. The average paper concen-
tration was 32.0% and no cellulose concentration data were reported.
It is not possible at present to make a general statement as to the
maximum amount of cellulose and hemicellulose decomposition which
can be expected in a landfill. Only a few researchers have measured
this parameter and there is little data in the open literature. Complete
cellulose and hemicellulose disappearance is not expected though the
data of Bookter and Ham (1982) suggest that values below 10% are
possible. However, excavated refuse may be diluted with cover soil,
thus diluting measured cellulose concentrations. Closely controlled
laboratory experiments and additional data on the composition of
refuse several years after burial is needed to estimate the fraction
of the cellulose and hemicellulose in effect not available for anaerobic
decomposition.
 Case study 133

6.6 CASE STUDY

The 36ha Omega Hills Landfill, owned and operated by Waste

Management of Wisconsin, Inc. (WMWI), is located 25 kms from
downtown Milwaukee. Since commencing operation in 1972, approxi-
mately 9 million cubic meters of municipal, industrial, and commercial
wastes have been deposited, mostly from the Milwaukee metropolitan
area. Average refuse depth is 37m feet with a maximum depth of 64m.
A monitoring program in 1984 at the landfill documented gas
migration and vegetative stress in the final clay cover. In addition, odor
from escaping gas was an aesthetic concern. In response to these site
conditions, a gas extraction system was designed and construction
began in the spring of 1985. The gas extraction system was designed to
provide a zero pressure envelope throughout most of the landfill by
withdrawing gas at a rate equal to its production.
Sixty-five gas extraction points are located throughout the Omega
Hills Landfill, consisting of 45 wells, 11 trenches, and 9 leachate col-
lection system connections. An average well spacing of 92 m was used
to provide zone of influence overlap. The gas collectors are connected
by a 5.8-km-Iong gas collection header system designed to handle up to
200000 m 3 of gas per day. The header system consists of high density
polyethylene pipe (HOPE) ranging from 100-600 mm in diameter. A
flexible coupling connects each well assembly to the header and allows
for differential settlement between the two components.
The header system is designed to handle both gas and liquid flow by
use of minimum slopes to facilitate drainage of condensate to six low
points. Barometric drip legs at the low points allow gravity drainage of
condensate from the header without air intrusion. Condensate is dis-
charged into the existing leachate control piping system which flows to
the on-site leachate pretreatment facility and from there to the sanitary
sewer. To minimize head loss, the header lines generally were sized for
maximum gas velocities of 9-11 m/sec.
Landfill gas is withdrawn from the wells and is transported through
the header to the 150 m 2 gas compressor building. In the compressor
building, the gas first flows through water wash scrubbers for par-
ticulate removal and then to compressors. The wash water is treated
along with the leachate. From the compressor building, the gas is
transported to the 290 m 2 turbine building where it fuels the gas fired
turbine engines which drive the electrical generators. A gas flaring
station is used only as a backup should the electrical generation system
be down for any length of time.
The compressors were manufactured by Hall Systems, Inc., of Tulsa,
Oklahoma. The first stage of each compressor consists of a centrifugal
blower capable of producing up to a 1.5 m water column head. The
centrifugal blower provides the head and flow rates required to
134 Leachate and gas generation

withdraw landfill gas from the collection system. The gas then enters
two additional stages of compression, with each stage followed by
cooling of the gas. After compression, gas at a delivery pressure of
28 kPa is fed to underground piping and transported to the turbine
generator sets. To prevent liquid condensation between the com-
pressors and the turbines, a portion of the gas bypasses the final
cooling stage and heats the rest of the gas. Each compressor is driven
by a 450 kW electric motor. The compressor package, which includes
the centrifugal blower, a piston type, two-stage compressor, motor, gas
scrubbers, and gas chiller unit, is skid mounted.
The two Centaur turbine generator sets were obtained from Solar
Turbine, Inc. Each set is skid mounted and weighs approximately
18 tonnes. Each turbine generator set can produce approximately
3.3 Mw of electricity. At this loading the turbine requires approximately
45 standard cubic meters per minute of landfill gas at 50% methane.
Electricity is generated at 4160 volts and is stepped up to 26.4kV and
purchased by the Wisconsin Electric Power Company.
Special attention was paid to the acoustical treatment of the facilities
because much of the equipment produces significant noise levels which
cannot be permitted to adversely impact adjacent communities.
The system went on-line December 17, 1985, and has since been
operating continuously with no obvious trends in gas generation rates
or composition. Gas flow meets the quantity and quality requirements
of the two turbines, which are basically operated at capacity. The plant
has exceeded the estimated on-line time with only minor maintenance
and other downtime periods having been experienced in the first four
and a half years of operation. The owner estimates a payback period of
seven to eight years.
In the fall of 1988, a third Centaur turbine-generator set was placed
on-line, increasing the plant capacity by 50%. There have been periods
when insufficient gas was available to operate all three turbines
simultaneously at capacity, largely because of liquid accumulations in
gas wells which limits their productivity.

REFERENCES

Barlaz, M.A. (1988) Microbial and chemical dynamics during refuse decom-
position in a simulated sanitary landfill, Ph.D. thesis, Dept. of Civil and
Environmental Engineering, Univ. of Wisconsin, Madison, WI.
Barlaz, M.A., Ham, R.K. and Milke, M.W. (1987) Waste Management and Res.,
5,27.
Barlaz, M.A., Ham, R.K. and Schaefer, D.M. (1989a) J. Environ. Eng., ASCE,
115(6), 1088-102.
Barlaz, M.A., Schaefer, D.M. and Ham, R.K. (1989b) App!. Env. Microbia!.,
 References 135
55(1), 55-65.
Barlaz, M.A., Schaefer, D.M. and Ham, R.K. (1989c) Appl. Biochem. Biotechnol.,
20/21, 181-205.
Barlaz, M.A, Ham, R.K. and Schaefer, D.M. (1990) CRC Critical Reviews in
Environ. Control, 19(6), 557-84.
Bookter, T.J. and Ham, R.K. (1982) J. Environ. Eng., ASCE, 108(EE6), 1089.
Buivid, M.G., Wise, D.L., Blanchet, M.J. et al. (1981) Resource Recovery and
Conservation, 6, 3.
Burns, R.G. and Martin, J.P. (1986) Biodegradation of organic residues in soil,
in Microflora and Faunal Interactions in Natural and Agro-Ecosystems, (eds M.J.
Mitchell and J.P. Nakas) Martinus NighofflDr. W. Junk Publ., Bordrecht,
The Netherlands.
Ehrig, H.J. (1988) Water and element balances of landfills, in Lecture Notes in
Earth Sciences. (ed. P. Baccini) Springer-Verlag, Berlin.
Emberton, J.R. (1986) The biological and chemical characterization of landfills,
Proc., Energy from Landfill Gas, Solihull, West Midlands, UK, Oct. 30-I.
EPA (1990) Characterization of Municipal Solid Waste in the United States; 1990
Update, EPA/530-SW-90-042, US EPA, Washington, DC, PB 90-215112.
Fathepure, B.Z., Nengu, J.P. and Boyd, S.A. (1987) Appl. Env. Microbiol.,
53(11), 2671-4.
Farquhar, G.J. and Rovers, F.A. (1973) Water, Air, and Soil Pollution, 2, 483.
Francis, AJ. and Dodge, c.J. (1986) Arch. Environ. Contam. Toxicol., 15, 611-16.
Greenhill, W.L. (1964a) J. British Grasslands Soc., 19, 30-7.
Greenhill, W.L. (1964b) J. British Grasslands Soc., 19, 231-6.
Greenhill, W.L. (1964c) J. British Grasslands Soc., 19, 336-9.
Halvadakis, c.P., Robertson, AP. and Leckie, J.O. (1983) Landfill Methano-
genesis: Literature Review and Critique, Technical Report No. 271, Depart-
ment of Civil Engineering, Stanford University.
Ham, R.K. and Bookter, T.J. (1982) J. Environ. Eng., ASCE, 108(EE6), 1147.
Jenkins, R.L. and Pettus, J.A. (1985) The use of in vitro anaerobic landfill
samples for estimating gas generation rates, in Biotechnological Advances in
Processing Municipal Wastes for Fuels and Chemicals (ed. A.A. Antonopoulos)
Argonne Nat!. Lab. Report ANLlCNSV-TM-167, p. 419.
Kinman, R.N., Nutini, P.L., Walsh, J.L. et al. (1987) Waste Management and
Res., 5, 13.
Kinman, R.N., Rickabaugh, J. and Berg, G. (1989) Analysis of 20 year-old
refuse from the Mallard North landfill in Chicago, Illinois, Proc. Purdue Ind.
Waste Conf., May 9-11, West Lafayette, IN.
Mah, R.A., Smith, M.R. and Baresi, L. (1978) Appl. Environ. Microbiol., 35(6),
1174.
McInerney, M.J. and Bryant, M.P. (1981) Basic principles of bioconversions in
anaerobic digestion and methanogenesis, in Biomass Conversion Processes for
Energy and Fuels, (eds S.S. Sofar and O. Zaborsky) Plenum Publishing
Corp., New York.
Merz, R.C. and Stone, R. (1962) Public Works, 93(9), Sept., 103.
Pacey, J. (1989) Enhancement of degradation: large-scale experiments, in
Sanitary Landfilling: Process Technology and Environmental Impact, (eds T.
Christensen, R. Cossu and R. Stegmann) Academic Press, London, pp.
103-19.
Parkin, G.F. and Owen, W.F. (1986) J. Environ Eng., ASCE, 112(5), 867.
Pohland, F.G. (1975) Sanitary Landfill Stabilization with Leachate Recycle and
Residual Treatment, Georgia Institute of Technology, EPA Grant No. R-
801397.
136 Leachate and gas generation

Pohland, F.G. and Gould, J.P. (1986) Wat. Sci. Tech., 18(12), 177.
Tchobanoglous, G., Theisen, H. and Eliassen, R. (1977) Solid Wastes, McGraw-
Hill, Inc., N.Y., 334.
Watson-Craik, I.A. and Senior, E. (1989) Water Res., 13(10), 1293-303.
Wolfe, R.S. (1979) Methanogenesis, in Microbial Biochemistry, Int. Rev. of
Biochemistry, 21, (ed. J.R. Quayle) University Park Press, Baltimore, MD.
Wolin, M.J. and Miller, T.L. (1985) in Biology of Industrial Microorganisms, (eds
A.L. Demsin and N.A. Solomon) Benjamin/Cumings Publ. Co. Inc. Menlo
Park, CA., pp. 189-221.
Wujcik, W.J. and Jewell, W.J. (1980) Biotech. Bioeng. Symp. No. 10, 43-65.
Young, L.Y. and Frazer, A.C. (1987) Geomicrobiology J., 5, 26l.
Zehnder, A.J.B. (1978) Ecology of methane formation, in Water Pollution
Microbiology, 2, (ed. Ralph Mitchell) John Wiley & Sons, N.Y. 1978, p. 349.
Zehnder, A.J.B., Ingvorsen, K. and Marti, T. (1982) Microbiology of methane
bacteria, in Anaerobic Digestion, (ed. D.E. Hughes) Elsevier Biomedical
Press B.V., Amsterdam, p. 45.
Zeikus, J.G. (1980) Microbial populations in digesters, in Anaerobic Digestion,
(ed. D.A. Stafford) Appl. Sci. Publishers, London, p. 61.
 CHAPTER 7

Clay liners
David E. Daniel

7.1 INTRODUCTION

Low-hydraulic-conductivity soil liners are given various names, in-

cluding soil liner and clay liner. In this chapter, the term clay liner is
used even though other minerals in the liner material (e.g., sand) may
be present in larger quantities than clay. The term clay is emphasized
because clay is largely responsible for the low hydraulic conductivity of
earthen liners. Attention in this chapter is focused on three types of
clay liners:
1. naturally-occurring clay liners;
2. compacted clay liners; and
3. geosynthetic clay liners.
A clay liner serves as a hydraulic barrier to flow of fluids. Clay liners
are used to minimize infiltration of water into buried waste (cover
systems) or to control release of leachate from the waste (liner systems).
To meet these objectives, clay liners must have low hydraulic con-
ductivity over long periods of time. Further, one must be able to verify
that the hydraulic conductivity will be suitably low, which is often
the most difficult problem to be resolved. In addition, clay liners are
expected to attenuate the movement of leachate, to prolong release of
chemicals in leachate, and to serve other site-specific functions.

7.2 NATURAL CLAY LINERS

Natural clay liners are naturally-occurring formations of low-hydraulic-

conductivity, clay-rich soil. Waste can be buried above or within a

Geotechnical Practice for Waste Disposal.

Edited by David E. Daniel.
Published in 1993 by Chapman & Hall, London. ISBN 0412 35170 6
138 Clay liners
natural liner. Natural liners normally contain significant amounts of
clay minerals and have hydraulic conductivities less than or equal to
1 x 10- 6 to 1 X 10- 7 cm/s. Natural liners more typically serve as a
back-up to engineered liners, but occasionally (for old landfills or,
where regulations allow, for new landfills), a natural liner may rep-
resent the only liner at a waste disposal facility.
The continuity and hydraulic conductivity of natural liner materials
are critical issues. To function effectively, the natural liner must be
continuous and be free from major hydraulic imperfections such as
fractures, joints, and holes.
An evaluation of a liner's continuity begins with a geologic evaluation
and includes a careful study of local and regional hydrogeology (see
Chapter 4). Exploratory borings are an essential investigatory tool;
surface and bore hold geophysics often provide a wealth of valuable
information. Analysis of radioisotopic concentrations of certain con-
stituents in ground water can lead to the determination of the age of
ground water; knowledge of the age of ground water can help to
establish that a soil liner hydrogeologically isolates one aquifer from
another. Another useful way to investigate the ability of a natural liner
to isolate one stratum from another is to pump from a well in an
underlying aquifer and to observe changes in water levels in wells
installed in an overlying aquifer, or vice versa.
Hydraulic conductivity of a natural liner should be studied with
a combination of laboratory and in situ hydraulic conductivity tests.
Laboratory test are usually performed with flexible-wall cells used to
permeate samples obtained by pushing a thin-walled tube into soil that
underlies a borehole (Daniel et al., 1984). The results of such tests
should be viewed with suspicion because if the liner contains hydraulic
defects (e.g., cracks, fissures, slickensides, or root holes), the defects
will probably be missed and the measured hydraulic conductivity will
be too low (Olson and Daniel, 1981; Daniel, Trautwein, and McMurtry,
1985; Keller, van der Kamp, and Cherry, 1986; Bradbury and Muldoon,
1990).
In situ hydraulic conductivity tests are an essential part of a credible
effort to characterize the hydraulic conductivity of natural soil liners.
The normal type of test involves a single cased borehole and either a
constant or a falling head. Several equations may be used to compute
hydraulic conductivity; see Olson and Daniel (1981) and Chapuis (1989)
for details. The number of tests that are needed to characterize the
hydraulic conductivity of a natural deposit of soil or rock varies from
site to site and depends upon hydrogeologic complexity, the required
accuracy to which the hydraulic conductivity needs to be determined,
available funding, and other factors. Typically, of the order of 10-20 in
situ hydraulic conductivity tests are appropriate. Great care must be
taken to seal the boreholes so that the hydraulic integrity of the liner is
 Compacted clay liners 139

not compromised. Sealing techniques are discussed in Part 4 of this

book.
It is extremely difficult and expensive to prove that a naturally-
occurring stratum of soil or rock uniformly possesses low hydraulic
conductivity. For this reason, use of a natural soil liner as the sole
means for protecting ground water from contamination is not normally
recommended. An exception might be the case of an extraordinarily
uniform, massive, and well-characterized stratum of material, but such
strata are rare. Another exception might be site remediation cases in
which ground water flow and contamination patterns are monitored
and contingencies are made in case strata are more permeable than
expected.
The interested reader is encouraged to consult Keller, van der Kamp,
and Cherry (1986) and Bradbury and Muldoon (1990) for examples of
well-conceived evaluations of natural soil liners.

7.3 COMPACTED CLAY LINERS

7.3.1 Introduction
Compacted clay liners are constructed primarily from natural soil
materials, although the liner may contain processed materials such as
bentonite or even synthetic materials such as polymers. Clay liners are
constructed in layers called lifts. On side slopes, the lifts can be
horizontal or parallel to the slope, although parallel lifts are not recom-
mended for side slopes steeper than 2.5 to 3 on 1 (horizontal to
vertical). As suggested by Fig. 7.1, lifts parallel to the side slopes are
preferred because the effect of a zone of poor material, or imperfect
bonding of lifts, is less with parallel lifts. If horizontal lifts are used, it
may help to place a slight inward inclination on the lift interfaces (Fig.
7.1) to minimize the tendency for leachate to flow along lift interfaces.

7.3.2 Compaction requirements

The objective of compaction is to remold chunks (clods) of soil into a
homogeneous mass that is free of large, continuous interclod voids. If
this objective is accomplished with suitable soil materials, low hydraulic
conductivity (::::; 1 X 10- 7 cm/s) will result.
Experience has shown that the water content of the soil, method
of compaction, and compactive effort have a major influence on the
hydraulic conductivity of compacted soil liners. Laboratory studies have
demonstrated that low hydraulic conductivity is easiest to achieve
when the soil is compacted wet of optimum water content with a
high level of kneading-type compactive energy (Mitchell, Hooper, and
140 Clay liners

(a)
improper material

(b)
improper material

Fig. 7.1 Side slopes constructed with (a) parallel and (b) horizontal lifts.

Campanella, 1965). Figure 7.2 illustrates the influence of molding water

content and compactive energy upon hydraulic conductivity. The soil
must be sufficiently wet so that, upon compaction, clods of clayey soil
will mold together, eliminating large inter-clod pores. Kneading the
soil during compaction with a high level of compactive energy helps to
remold clods and to eliminate large pore spaces.
Studies have also demonstrated lifts of soil must be bonded together
to eliminate, to the extent possible, highly permeable zones at lift
interfaces. The idea is illustrated in Fig. 7.3; if permeable inter-lift
zones are eliminated, hydraulic connection between 'defects' in each
lift is destroyed and a low overall hydraulic conductivity is achieved.

7.3.4 Materials
The minimum requirements recommended to achieve a hydraulic con-
ductivity ::::: 1 x 10- 7 cm/s for most soil liners materials are as follows:
percentage fines: ~20-30%
plasticity index: ~7-10%
percentage gravel: :::::30%
maximum particle size: 25-50mm
 Compacted clay liners 141

1 X 10- 5

6'
w
(J)

E
~
1 X 10- 6

1 X 10- 7

1 X 10- 8

116
112
108
104
100
.increasing compactive effort
96
92~~~~--~~~~L-~~~=-~~
12 14 16 18 20 22 24
molding water content (%)

Fig. 7.2 Effect of molding water content and compactive energy on hydraulic
conductivity, from Mitchell et al. (1965).

Percentage fines is defined as the percent by dry weight passing the

US No. 200 sieve, which has openings of 75 ~m. Plasticity index may
be determined by ASTM 04318. Percentage gravel is defined as the
percent by dry weight retained on a No.4 sieve (4.76mm openings).
Local experience may dictate more stringent requirements, and, for
some soils, more restrictive criteria may be appropriate. However, if
the criteria tabulated above are not met, it is unlikely that a natural soil
liner material will be suitable without additives such as bentonite.
Recent work conducted to evaluate gravel content bears mentioning.
Shelley (1991) mixed kaolinite and mine spoil with varying percentages
of gravel (maximum particle size: 20 mm), moistened the soil to a
few percent wet of optimum, compacted the soil/gravel mixtures
using standard Proctor compaction procedures (ASTM 0698), and then
permeated the compacted specimens. Results are summarized in Fig.
7.4. Shelley found that the soil could contain up to 50-60% gravel
without a detrimental impact upon hydraulic conductivity. Shelley
reported that at gravel percentages ~50-60%, clay particles plugged
the voids between the gravel particles. Shakoor and Cook (1990) report
142 Clay liners

Good bonding of lifts causes Poor bonding of lifts causes

hydraulic defects in adjacent hydraulic defects in adjacent
lifts to be hydraulically lifts to be hydraulically
unconnected connected to each other

Fig. 7.3 Effect of (a) good and (b) poor bonding of lifts on the performance of
a compacted clay liner.
10-5~------r------.-------'-------'-------'
I
Note: Hydraulic Conductivity of
Gravel Alone = 170 cm/s

------l--- - - l - - + - - --f±------I

I
-+--+---+--/-I---------t-----1

o 20 40 60 80 100

Percent Gravel (by Weight)

Fig. 7.4 Effect of percent of gravel in clay on hydraulic conductivity, after

Shelley (1991).

similar results. However, in both investigations, the soil and gravel

were carefully and uniformly mixed in the laboratory; in the field,
the mixing will be less perfect. The main issue is not necessarily how
much gravel is present (assuming the gravel content is ~50-60%) but,
rather, with the likelihood that pockets of gravel (segregation of gravel)
can occur during construction . The potential for gravel segregation to
occur depends on the soil material and construction procedures . The
author recommends that the gravel content not exceed about 30%, but
notes that this value should be increased or decreased as appropriate
for a given material and construction process.
If suitable materials are unavailable locally, local soils can be blended
with commercial clays, e.g., bentonite, to achieve low hydraulic con-
 Compacted clay liners 143
(j)
E W- 4 .---....-~~.,...~---r---.-.----....-~...,...........
~

::: 10- 5
:~
(3
::J
"0
r::
o
()

.52
::J
~
~ 10-8~~~~~~~---L-~~~~
I 0 2 4 6 8 10 12

Percent Bentonite

Fig. 7.5 Effect of percent of bentonite on hydraulic conductivity.

ductivity. A relatively small amount of bentonite can lower hydraulic

conductivity as much as several orders of magnitude (Fig. 7.5).
One should be cautious about using highly plastic soils (soils with
plasticity indices >30-40%) because these materials form hard clods
when the soil is dry and are very sticky when the soil is wet. Highly
plastic soils, for these reasons, are difficult to work with in the field.

7.3.5 Construction

Processing
Some liner materials need to be processed to break down clods of soil
(Benson and Daniel, 1990; Shackelford and Javed, 1991), to sieve out
stones and rocks, to moisten the soil, or to incorporate additives. Clods
of soil can be broken down with tilling equipment. Stones can be
sieved out of the soil with large vibratory sieves or mechanized 'rock
pickers' passed over a loose lift of soil. Road reclaimers can process soil
in a loose lift and crush stones or large clods.
If the soil must be wetted or dried more than 2-3 percentage points
in water content, the soil should be processed by spreading it in a loose
lift about 300mm thick. Water can be added and mixed into the soil
with a tiller, or the soil can be disced or tilled to allow it to dry
uniformly. It is essential that time be allowed for the soil to wet or dry
uniformly. At least 1-3 days is usually needed for adequate hydration
or dehydration. Frozen soil should never be used to construct a soil
liner.
Additives such as bentonite can be introduced in two ways. One
technique is to mix soil and additive in a pugmill. Water can also be
added in a pugmill either concurrently with bentonite or in a separate
processing step. Alternatively, the soil can be spread in a loose lift
144 Clay liners
200-300 mm thick, the additive spread over the surface, and rototillers
used to mix the materials. Several passes of the tiller over a given spot
are usually needed. Water can be added in the tiller during mixing or
later, after mixing is complete. The pugmill is more reliable in providing
thorough, controlled mixing, but, done carefully, the other method can
provide adequate mixing. For more information on bentonite and soil!
bentonite testing, the reader is referred to Alther (1983), Alther (1987),
Chapuis (1990).

Surface preparation
It is crucial that each lift of a soil liner be effectively bonded to the
overlying and underlying lifts (Fig. 7.3). The surface of a previously-
compacted lift must be rough rather than smooth. If the surface has
been smoothed, e.g., with a smooth steel-drummed roller, the surface
should be excavated to a depth of 20-30 mm with a disc or other
suitable device.

Soil placement
Soil is placed in a loose lift that is no thicker than about 230 mm (9 in.).
If grade stakes are used to gauge thickness, the stakes must be
removed and the hole left by the stakes sealed. Other techniques, e.g.,
use of lasers, are preferable for control of elevations. After the soil is
placed, a small amount of water may be added to offset evaporative
losses, and the soil may be tilled one last time prior to compaction.

Compaction
Heavy, footed compactors with large feet that fully penetrate a loose
lift of soil (Fig. 7.6) are ideal. Recommended specifications include:
minimum weight: 18000kg (40 000 lbs);
minimum foot length: 180-200mm; and
minimum number of passes: 5.
More passes may sometimes be needed. A 'pass' is defined as one pass
of the compactor, not just an axle, over a given area, and the recom-
mended minimum of five passes is for a vehicle with front and rear
drums. In the US, the Caterpillar 815B and 825C are examples of
equipment in widespread use that have led to satisfactory results in
most cases.
Statically operated compactors are preferred over vibratory com-
pactors for soil liners. The weight of the compactor must be compatible
with the soil; relatively dry soils with firm clods require a very heavy
 Compacted clay liners 145
(a)

~u(~------f'l0 ("'d of soil

(b)

loose lift
of soil

Fig. 7.6 (a) Fully and (b) partIy penetrating feet on a footed roller.

compactor whereas relatively wet soils with soft clods require a roller
that is not so heavy that it becomes bogged down in the soil.

Protection
After compaction of a lift, the soil must be protected from desiccation
and freezing. Desiccation can cause cracking of the clay (Boynton and
Daniel, 1985; Kleppe and Olson, 1985). Freeze-thaw changes the struc-
ture and fabric of compacted clay in a way that increases hydraulic
conductivity (Fig. 7.7) (Chamberlain and Cow, 1979; Zimmie and La
Plante, 1990; Kim and Daniel, 1992). Desiccation can be minimized in
several ways: the lift can be temporarily covered with a sheet of plastic
(but one must be careful that the plastic does not heat excessively
and itself dry the clay), the surface can be smooth-rolled to form
a relatively impermeable layer at the surface, or the soil can be
periodically moistened. The compacted lift can be protected from
damage by frost by avoiding construction in freezing weather or by
temporarily covering the lift with an insulating layer of material. The
protective measures discussed in this section apply to each lift as well
as to the completed liner or cover barrier.

Quality control tests

A critical component in construction quality assurance are quality
control (QC) tests. For soil liners, the tests fall into two categories:
146 Clay liners

if)
E
s ---0-- Green River

-
~
---0-- Durango
:~ ----6-- Slick Rock
t5
:J Rifle
"0
c
o
o
.~
:J
C1l
-0>,
I 10-9~~~~~~~~~~~~~~
o 5 10 15 20
Number of Freeze-Thaw Cycles

Fig. 7.7 Effect of freeze-thaw on hydraulic conductivity of compacted clay,

after Chamberlain et a/. (1990).

1. tests to verify that the materials of construction are adequate; and

2. tests and observation to verify that the compaction process is
adequate.
Recommended tests and minimum testing frequencies are summarized
by Daniel (1990).

7.3.6 Water content and dry unit weight

A critical step in design of a compacted soil liner is determination of
the range of acceptable water content and dry unit weight of the soil. If
the soil is too dry at the time of compaction, suitably low hydraulic
conductivity may be unachievable. If the soil is too wet, a variety
of problems may ensue, e.g., problems with construction equipment
operating on soft, weak soils and potential slope instability caused by
low strength of the soil.
Once an acceptable water content range has been selected, the soil
must be compacted with adequate compactive energy to compress
large voids and to remold clods of soil into a homogeneous, relatively
impermeable mass. The dry unit weight of the soil can be a useful
indicator of the effectiveness of compaction.
One problem confronting the designer is that both the water content
of the soil and the compactive energy delivered to soil during construc-
tion of a soil liner vary. Further, precise duplication of field compac-
tion in the laboratory is impossible. Accordingly, the recommended
approach (Daniel and Benson, 1990) for establishing water content and
dry unit weight requirements during the design stage is to utilize a
range in water content that more than spans the range anticipated
in the field and to compact the soil with three compactive energies
 Compacted clay liners 147

o
1: Maximum
O'l
.0.;

-
L:;. Allowable
~ L:;. L:;.
·c L:;. L:;.
::::>
0
~ 0 0
0
0
0
80
0 0
(A) (6)

Molding Water Content Molding Water Content

Modified
Acceptable
1: Zone
O'l
.0.;

-
~
·c
::::>
L:;.

~
0
o
o 0
(C)
0 o (D)
o o

Molding Water Content Molding Water Content

Fig. 7.8 Procedure for establishing acceptable zone for water content and dry
unit weight, after Daniel and Benson (1990).

that represent estimates of the lowest compactive energy anticipated,

average energy, and highest compactive energy. The author recom-
mends standard and modified Proctor (ASTM 0698 and 01557) for
the average and high compactive energies, respectively, and 'reduced
Proctor' for the lowest energy. 'Reduced Proctor' is the same as
standard Proctor but with only 15 drops of the compactive ram per lift
rather than the usual 25 drops.
The recommended procedure is illustrated in Fig. 7.8. One compacts
5-6 samples of soil with three different compactive efforts and plots
three compaction curves (Fig. 7.8a). Next, the compacted soils are
permeated and hydraulic conductivity is measured (Fig. 7.8b). The
compaction points are replotted (Fig. 7.8c) with solid symbols used for
test specimens that had adequately low hydraulic conductivity and
open symbols used for test specimens that were too permeable. An
148 Clay liners
Acceptable Zone Based
on Shear Strength

Acceptable Zone

Acceptable Zone Based

Hydraulic Conductivity

Molding Water Content

Fig. 7.9 Overall acceptable zone based on hydraulic conductivity, shear

strength, and shrinkage upon desiccation.

'acceptable zone' is drawn (usually with some judgement applied) that

encompasses the solid points. Finally, the acceptable zone is modified
to account for any other relevant factors, e.g., shear strength
considerations or local construction practices (Fig. 7.8d). Figure 7.9
illustrates how an acceptable zone can be defined from hydraulic
conductivity, shear strength, and desiccation shrinkage criteria.

7.3.7 Test pads

The construction of a test pad prior to building a full-sized liner has
many advantages. By constructing a test pad, one can experiment with
compaction water content, construction equipment, number of passes
of the equipment, lift thickness, etc. Most importantly, though, one
can conduct extensive testing, including quality control testing and in
situ hydraulic conductivity testing, on the test pad.
It is usually recommended that the test pad have a width of at least
three construction vehicles (>10 m), and an equal or greater length.
The pad should ideally be the same thickness as the full-sized liner, but
the trial pad may be thinner than the full-sized liner. (The full-
thickness liner should perform at least as well as, and probably better
than, a thinner test section because defects in anyone lift become less
important as the number of lifts increases.) The in situ hydraulic
conductivity may be determined in many ways. The large sealed
double-ring infiltrometer is usually the best large-scale test (Daniel,
1989; Sai and Anderson, 1990), although the Boutwell test (Daniel,
1989, and references therein) is enjoying increased popularity due to its
ease of operation and relatively short testing times.
One problem with in situ tests on test pads is that the test pad is
subjected to essentially zero overburden stress. Hydraulic conductivity
 Compacted clay liners 149

---

o
effective stress

Fig. 7.10 Recommended procedure for adjusting the hydraulic conductivity

measured in the field on a test pad for the influence of compressive stress.

decreases with increasing compressive stress. The author recommends

that the hydraulic conductivity measured on a test pad with in situ
methods be corrected for the effects of overburden stress based on
results of laboratory hydraulic conductivity tests performed over a
range in compressive stress (Fig. 7.10).

7.3.8 Chemical attack by waste

Waste liquids may attack and effectively destroy earthen liners. It
is convenient to consider acids and bases, neutral inorganic liquids,
neutral organic liquids, and leachates separately. Testing protocols
have been described by Bowders et ai. (1986).

Acids and bases

Strong acids and bases can dissolve solid material in the soil, form
channels, and increase hydraulic conductivity. Some acids, e.g.,
hydrofluoric and phosphoric acid, are particularly aggressive and dis-
solve soil readily. Concentration of acid, duration of reaction, liquid-
solid ratio, type of clay, and temperature are also important variables
(Grim, 1953). Leachates with pH <3 or >11 are usually of the most
concern.
When concentrated acid is passed through clayey soil, hydraulic
conductivity often declines initially but later increases. Soils have a
high capacity to buffer acid; many pore volumes of flow are usually
needed before the full effect of the acid is observed. Examples include
Nasiatka et al. (1981), Peterson and Gee (1986), and Bowders and
Daniel (1987). Soils that are composed primarily of sand, with a small
amount of bentonite, are particularly susceptible to attack by acids
150 Clay liners

because the small mass of bentonite is readily dissolved (Nasiatka et al.

1981).

Neutral, inorganic liquids

The effects of neutral, inorganic liquids may be evaluated with the
Gouy-Chapman theory (Mitchell, 1976), which states that the thick-
ness (T) of the diffuse double layer varies with the dielectric constant
of the pore fluid (0), the electrolyte concentration (no), and the cation
valence (v) as follows:
(7.1)
For solutions containing mainly water, the dielectric constant of the
liquid is relatively constant, and thus the main parameters are no and v.
As the diffuse double layer of adsorbed water and cations expands,
hydraulic conductivity decreases because flow channels become con-
stricted. Attempts to validate quantitatively the effect of 0, no, and v
on conductivity have generally failed. Qualitatively, however, the
Gouy-Chapman theory explains the observed patterns. Aqueous
solutions with few electrolytes, e.g., distilled water, tend to expand the
double layer and to produce low hydraulic conductivity. A strong
(high no) solution containing polyvalent cations tends to produce the
largest conductivity. Further details and supporting data are reported
by Fireman (1944) and McNeal and Coleman (1966).

Neutral, organic liquids

Most organic chemicals have lower dielectric constants than water.
Low 0 tends to cause low T, as shown in Eq. (7.1) and thus high
hydraulic conductivity. In addition, low-dielectric-constant liquids
cause clay particles to flocculate and cause the soil to shrink and
to crack (Anderson, 1982; Fernandez and Quigley, 1985, 1988). This
phenomenon is called 'syneresis' and the cracks 'syneresis cracks'.
Numerous studies have shown that organic chemicals can cause large
increases in hydraulic conductivity (Anderson, 1982; Acar et aI., 1985;
Fernandez and Quigley, 1985; Foreman and Daniel, 1986; and others).
High compressive stress causes the soil to compact when an organic
solvent passes through the soil rather than to crack (Broderick and
Daniel, 1990; Fernandez and Quigley, 1991). Thus, soil liners perform
much better at high compressive stress than low stress when they are
permeated with organic liquids.
Dilute organic liquids do not tend to alter hydraulic conductivity
significantly (Bowders and Daniel, 1987, and references therein). If a
small amount of low-dielectric-constant liquid, e.g., trichlorethylene (0
 Geosynthetic clay liners 151
= 3), is mixed with water, the dielectric constant of the mixture is only
slightly less than that of water (80). Tests have indicated that the
dielectric constant must be less than 30-50 for hydraulic conductivity
to increase. Experience indicates that soil liners are not attacked by an
organic liquid if
1. the solution consists of at least 50% water, and
2. there is no separation of phases, i.e., all of the organic liquid is
dissolved in the water and none exists as a separate phase.
A recent, comprehensive review is provided by Budhu et al. (1991).

7.3.9 Reliability of compacted clay liners

Many examples can be cited of soil liners that failed to function
effectively as hydraulic barriers. Soil liners are not the problem; inade-
quate investigations (natural soil liners) or inadequate construction or
quality control (compacted soil liners) are the main causes of problems.
Gordon et al. (1989), Cartwright and Krapac (1990), Johnson et al.
(1990), and Reades et al. (1990) provide examples of well-built, full-scale
compacted soil liners that have in situ hydraulic conductivities :::::; 1 X
10- 7 m/s. Good-quality soil liners that will perform effectively can be
constructed.

7.4 GEOSYNTHETIC CLAY LINERS

A relatively new type of manufactured clay liner is receiving wide-

spread attention as a potential hydraulic barrier in liner and cover
systems at waste disposal facilities. The liner consists of a thin layer of
clay sandwiched between two geotextiles or glued to a geomembrane.
Four companies currently manufacture these types of materials, and
new products of a similar design are expected to appear on the market
in the next year or two.
Various terms have been used to describe these materials, e.g.,
clay mat, bentonite matting, bentonitic clay liner, prefabricated clay
blanket, prefabricated clay liner, etc. Recently, the term geosynthetic
clay liner (GCL) has received widespread use for these materials and is
used in the remainder of this chapter.

7.4.1 Types of geosynthetic clay liners

Four GCLs are currently manufactured: Bentofix®, Bentomat®,
Claymax®, and Gundseal. The configurations, summarized in Fig. 7.11,
fall into 2 categories:
152 Clay liners

(a)
needlepunched
non-woven geolexlile
fibers "

~ non-woven geolexlile

(b)
woven polypropylene
geoleX1ile

spun-lace polyester backing

(c)

high-density polyethylene sheet

Fig_ 7_11 Geosynthetic clay liners: (a) Bentofix @ and Bentomat @; (b)
Claymax @; (c) Gundseal.

L bentonite sandwiched between two geotextiles (Bentofix®,

Bentomat®, and Claymax®); and
2. bentonite mixed with an adhesive and glued to a geomembrane
(Gundseal).
All the GCLs contain approximately 5 kg/m2 of bentonite. The materials
are manufactured in panels with widths of approximately 4-5 m and
lengths of 25-60m.
The panels are placed on rolls at the factory, stored, shipped to the
construction site, and unrolled in their final location. All four GCL's
are said to be self-sealing at overlaps between panels (Fig. 7.12): when
 Geosynthetic clay liners 153

(a) 1S0- 22Smm

additional bentonite

(b) ;!: 1S0mm

(e)
high-densi ty polyethylene
geomembrane

Fig. 7.12 Overlapped geosythetic clay liners: (a) Bentofix ® and Bentomat ®
seam; (b) Claymax ® seam; (c) Gundseal seam.

water hydrates the clay in the GCL, the clay swells and automatically
seals the overlap. Because no mechanical seaming of joints between
panels is needed, GCLs can be installed very rapidly.
After the GCL is placed, it must be covered immediately. If the GCL
becomes wet, e.g., from a rainstorm, the clay swells in an uneven
manner and the material generally may not properly self-seam. Thus,
the GCL cannot be installed when precipitation is threatening, and the
GCL cannot be left exposed.
Bentofix® is manufactured in Germany and Canada by Naue-
Fasertechnik. Sodium-activated bentonite (German product) is placed
between two thick, nonwoven, high density polyethylene (HDPE)
geotextiles and then the geotextiles are needlepunched together. The
lower and upper geotextiles weigh 800 g/m2 and 400 g/m2, respectively.
The purpose of needlepunching the geotextiles together is twofold:
154 Clay liners

1. to hold the GCL together during handling and deployment; and

2. to provide increased in-plane shear strength after deployment.

In the field, the Bentofix® sheets are unrolled and overlapped.

Additional granular bentonite is placed between overlapped panels to
make the material self-sealing at overlaps. Further information is avail-
able from Scheu et al. (1990).
Bentomat® is manufactured by American Colloid Company in Villa
Rica, Georgia. Several geotextiles and grades of bentonite can be used
in making Bentomat®. The usual configuration consists of regular or
contaminant-resistant sodium bentonite sandwiched between two geo-
textiles (each 170 g/m2) that are needlepunched together. Conceptually,
the product is similar to Bentofix®, but Bentomat® is made with thinner
geotextiles and a natural, sodium bentonite.
At overlapped sections of Bentomat®, O.4kg/m of additional bentonite
is placed between the GCL panels to assist in self-sealing upon
hydration.
Claymax® is manufactured by the James Clem Corporation in
Fairmont, Georgia. With this GCL, bentonite is mixed with a water-
soluble adhesive and sandwiched between two geotextiles. The upper
geotextile usually consists of a 119 g/m2 woven polypropylene needle-
punched with a 17 g/m2 nylon fabric to form a 136 g/m2 primary geo-
textile for the GCL. The other geotextile is a 25 g/m2 open-weave,
spun-lace polyester backing. The adhesive mixed with the bentonite
serves to hold the GCL together during manufacture and installation.
After Claymax® has been installed, the adhesive no longer serves a
purpose; hence, the solubility of the adhesive is insignificant in terms
of long-term performance.
Claymax® panels are overlapped in the field, but no extra bentonite
is required; the bentonite is said to ooze through the openings in the
geotextiles (especially the open-weave polyester backing) to self-seal
the overlapped panels when the bentonite is hydrated.
Further information about Claymax® may be obtained from Schubert
(1987), Shan (1990), and Shan and Daniel (1991).
Gundseal is manufactured by Gundle Lining Systems in Spearfish,
South Dakota. The material consists of sodium bentonite that is mixed
with an adhesive and attached to a high density polyethylene (HOPE)
geomembrane through a calendaring process. The geomembrane is
usually a 0.5 mm smooth HOPE sheet, but textured geomembranes and
very low density polyethylene (VLDPE) geomembranes can be used.
Gundseal is unrolled in the field with the HOPE facing either up or
down. Overlapped areas are said to be self-sealing at the bentonite/
polyethylene contact; no mechanical seaming is necessary, although
the polyethylene sheets could be welded together, if desired.
If the bentonite is facing downward, Gundseal serves as a composite
 Geosynthetic clay liners 155
geomembrane/clay liner with the geomembrane on the top. If one
wished to place a conventional geomembrane on top of Gundseal to
form a composite liner, the bentonite in the Gundseal would face
upward as shown in Fig. 7.1; the composite liner would consist of
a separate geomembrane underlain by the bentonite component of
Gundseal underlain by the polyethylene geomembrane component
of Gundseal.

7.4.2 Engineering properties

Hydraulic conductivity
Various organizations have measured the hydraulic conductivity of
GCLs. The tests have been performed with flexible-wall permeameters
over a range in effective stress. The tests on Gundseal were performed
by punching holes in the geomembrane (otherwise the geomembrane
would restrict cross-plane flow). Figure 7.13 summarizes the available
hydraulic conductivity data. Information about the tests and sources of
data may be found in Daniel and Estornell (1990), Scheu et al. (1990),
and Estornell (1991).
The hydraulic conductivity to water varies between approximately
1 x 10- 10 and 1 x lO- B cm/s, depending on compressive stress. There
are some differences in hydraulic conductivity between the various
GCLs, but those differences are difficult to sort out given the very

10 -8

•
~
E •
~ • •
.~
>
'u • • •
:l
U
c
0
10 -9
• •• •
••
0
.2
• •
~ •
Bentofix
-g Bentomat •
I • Claymax
• •
• Gundseal

10 -10~__~~~~~~__~__~~~~~__~~~~~~
1 10 100 1000

Compressive Stress (kPa)

Fig. 7.13 Data on hydraulic conductivity of geosynthetic clay liners.

156 Clay liners

limited data base for some of the materials and differences in testing
procedures employed to measure hydraulic conductivity. The engineer
who compares the hydraulic conductivity of one GCL to that of
another GCL should be careful to compare hydraulic conductivities
determined at the same compressive stress and similar test conditions.
Estornell (1991) built several tanks to evaluate the hydraulic integrity
of overlapped seams. The tanks measured 1.2m in width by 2.4m in
length. Samples of three GCLs (Bentomat®, Claymax®, and Gundseal)
were placed on a drainage layer on the bottom of the tank and then
covered with 300-600 mm of gravel. The GCLs were flooded with
water to a depth of 300-600 mm, and the flow rate in the drainage
layer located beneath the GCL was monitored. Tests were performed
with control samples containing no overlapped seam and with over-
lapped samples. Estornell found that the fluxes through the control
samples and the overlapped samples were consistently low and were
about the same. The overlapped sections did in fact self-seal under in
these large-scale, controlled tests.
Shan and Daniel (1991) describe tests in which one geosynthetic clay
liner was permeated with a variety of chemicals. Shan and Daniel
found that the liner maintained low hydraulic conductivity to a broad
range of chemicals when the bentonite was fully hydrated with fresh
water prior to introduction of the chemical. However, when the dry
GCL was permeated directly with an organic chemical, the bentonite
did not hydrate, did not swell, and did not attain a low hydraulic
conductivity. The designer should be careful to utilize the results of
hydraulic conductivity tests performed under the most critical con-
ditions of hydration that can be expected in the field.
Shan and Daniel (1991) also describe tests aimed at documenting the
ability of one GCL to self heal when damaged by puncture, desicca-
tion, or freeze/thaw. Shan and Daniel found that if the dry material
is punctured, the bentonite will swell when hydrated and fill small
punctures. If the GCL is hydrated and dried, severe desiccation cracks
can form. However, Shan and Daniel found that when the GCL
was rewetted, the bentonite swelled, and the hydraulic conductivity
returned to the original, low value for the undamaged material. Several
cycles of wet/dry did not cause a permanent change in the hydraulic
conductivity of the hydrated material under controlled laboratory test
conditions. Similarly, freeze/thaw cycles did not cause an increase
in hydraulic conductivity of one GCL investigated. The very high
swelling capacity of bentonite was presumably responsible for the self
healing capability that was observed.
One topic of interest concerning geosynthetic clay liners is the
behavior of composite geomembrane/geosynthetic clay liner. The
question is the following: if a geomembrane is placed on a geosynthetic
clay liner with a geotextile separating the geomembrane from the
 Geosynthetic clay liners 157
bentonite (Fig. 7.14a), will the highly transmissive geotextile transmit
water laterally (in which case the composite action is less than ideal), or
will bentonite seal the geotextile (Fig. 7.14b) such that lateral flow is
minimal? Estornell (1991) used the large tanks mentioned earlier to
study this issue for two GeLs that had geotextiles on both the upper
and lower surface and one GeL that had no geotextile on one surface.
Estornell placed a geomembrane that contained several punctures
directly on the GeLs and found that lateral wicking of water through
the geotextile in two of the GeLs was significant. For the GeL that did
not contain a geotextile on one surface of the material, no lateral
flow at the geomembranelbentonite interface was noted (in fact, the
bentonite was only wetted to a distance of 50-75 mm from the puncture
in the geomembrane after about three months of wetting). These tests
illustrate that composite action with a geomembrane may be affected

Puncture
Geomembrane

Puncture
Geomembrane

Very Little Lateral Flow

at Geomembrane/Clay
Interface

Fig. 7.14 (a) Poor and (b) good composite action between geomembrane and
geosynthetic clay liner.
158 Clay liners
100

80
•
ro-
a..
•
6
rJl 60
•
rJl
~
Ci5
til
40 •
Ql
.c
U)
20

20 40 60 80 100 120 140

Normal Stress (kPa)

Fig. 7.15 Drained direct shear data on fully hydrated geosynthetic clay liners.

by a geotextile that separates the geomembrane from the clay. Further

study of this phenomenon is in progress, and at least one manu-
facturer is attempting to add agents to the geotextile in contact with
a geomembrane that will reduce the in-plane transmissivity of the
geotextile. Also, the reader should realize the effectiveness of the
composite action between a geomembrane and a GCL depends on
overburden stress; the higher the overburden stress, the less the lateral
flow along the interface.

Shear strength
Fully-drained direct shear tests have been performed on Bentomat®,
Claymax®, and Gundseal. The tests were performed on 60-mm-
diameter samples that were subjected to a normal stress, hydrated, and
then sheared very slowly. Separate consolidation tests were performed
to obtain the parameters needed to determine the time to failure
required to ensure fully drained conditions. Times to failure were
typically 4-8 days, although the tests were continued to residual con-
ditions over a shearing period of about four weeks.
Results of the tests are plotted in Fig. 7.15. The straight lines shown
in the figure were determined from linear regression; regression
coefficients are given in Table 7.1.

7.4.3 Current applications

To date, there have been two primary applications of GCLs for waste
containment applications:
 Geosynthetic clay liners 159

Table 7.1 Regression coefficients

geosynthetic clay effective cohesion effective angle

liner (kPa) of internal friction
(degrees)

Bentomat® 30 26
Claymax® 4 9
Gundseal 8 8

1. as the clay component of a primary composite liner for double liner

systems; and
2. as the clay component of a composite liner in final cover systems for
landfills or site remediation projects.
Geosynthetic clay liners have also been used by themselves as a barrier
layer in liners and covers, although to a lesser extent. Secondary
containment structures have also been lined with GeLs (Bruton, 1991).

7.4.4 Advantages and disadvantages of GeLs

Geosynthetic clay liners are usually viewed as an alternative to a
compacted clay liner (Grube, 1991; Grube and Daniel, 1991). Hence,
the following discussion is addressed at the relative advantages of
GeLs compared to eeLs.
Table 7.2 contrasts the differences between geosynthetic clay liners
and compacted clay liners. The main advantages of GeLs are that
GeLs can be installed much more quickly, lightweight construction
equipment can be used (which is especially important when the clay
liner is placed on top of other components of a liner system that might
be punctured or damaged by heavy equipment), GeLs are installed dry
and are therefore not as vulnerable to damage from desiccation during
construction, and dry GeLs do not produce water upon loading (wet,
compacted clay liners consolidate when loaded and the consolidation
water can be misinterpreted as leakage if the liner is placed on top of a
leak detection layer). The main disadvantages of GeLs are a general
lack of experience, the vulnerability of a thin GeL to puncture, ques-
tionable composite behavior of some GeLs with an adjacent geo-
membrane, less leachate attenuation capacity than a thick liner, and
questions about stability of hydrated bentonite.
GeLs and eeLs both enjoy advantages and disadvantages. The
engineer should weigh the pluses and minuses for each project and
make a decision about which type of clay liner is most appropriate
based on such an evaluation.
160 Clay liners

Table 7.2 Comparison of geosynthetic clay liners with conventional com-

pacted clay liners

compacted clay liner (CCL) geosynthetic clay liner (GCL)

- - - - - - - - - - --~-- -.----

thick (0.6-1.5 m) thin (:::::lOmm)

field constructed manufactured
hard to build correctly easy to build (unroll and place)
impossible to puncture possible to damage and puncture
constructed with heavy equipment light construction equipment can be used
often requires test pad at each site repeated field testing not needed
site-specific data on soils needed manufactured product; data available
large leachate-attenuation capacity small leachate-attenuation capacity
relatively long containment time shorter containment time
large thickness takes up space little space is taken
cost is highly variable more predictable cost
soil has low tensile strength higher tensile strength
can desiccate and crack can't crack until wetted
difficult to repair not difficult to repair
vulnerable to freeze/thaw damage less vulnerable to freeze/thaw damage
performance is highly dependent hydraulic properties are less sensitive to
upon quality of construction construction variabilities
slow construction much faster construction

REFERENCES

Acar, Y.B., Hamidon, A., Field, S.D. and Scott, L. (1985) The effect of organic
fluids on hydraulic conductivity of compacted kaolinite, Hydraulic Barriers
in Soil and Rock, ASTM STP 874, (eds A.1. Johnson, R.K. Frobel,
N.J. Cavalli and CB. Pettersson), Am. Soc. for Testing and Materials,
Philadelphia, pp. 171-87.
Alther, C.R. (1983) The methylene blue test for bentonite liner quality control,
Geotech. Testing J., 6(3), 128-32.
Alther, C.R. (1987) The qualifications of bentonite as a soil sealant, Eng. Geo!.,
23, 177-91.
Anderson, D.C (1982) Does landfill leachate make clay liners more perme-
able?, Civ. Eng., 52(9), 66-9.
Benson, CB. and Daniel, D.E. (1990) The influence of clods on the hydraulic
conductivity of a compacted clay, ]. Geotech. Eng., 116(8), 1231-48.
Bowders, J.J., Daniel, D.E., Broderick, C.P. and Liljestrand, H.M. (1986)
Methods for testing the compatibility of clay liners with landfill leachate,
Haz. and Ind. Solid Waste Testing, 4th Symp., ASTM STP 886, (eds J.K.
Petros, W.J. Lacey and R.A. Conway), Am. Soc. for Testing and Materials,
Philadelphia, pp. 233-50.
Bowders, J.J. and Daniel, D.E. (1987) Hydraulic conductivity of compacted
clays to dilute organic chemicals, ]. Geotech. Eng., 113(12), 1432-48.
Boynton, 55. and Daniel, D.E. (1985) Hydraulic conductivity tests on com-
pacted clay, ]. Geotech. Eng., 111(4), 465-78.
Bradbury, K.R. and Muldoon, M.A. (1990) Hydraulic conductivity
determinations in unlithified glacial and fluvial materials, in Ground Water
 References 161

and Vadose Zone Monitoring, STP 1053, (eds D.M. Nielsen and A.I.
Johnson), Am. Soc. for Testing and Materials, Philadelphia, pp. 138-5l.
Broderick, G.P. and Daniel, D.E. (1990) Stabilizing compacted clay against
chemical attack, J. Geotech. Eng., 116(10), 1549-67.
Bruton, D. (1991) Bentonite mats meet secondary containment standards,
Geotech. Fabrics Rep., 9(4), 26-7.
Budhu, M., Giese, R.F., Campbell, G. and Baumgrass, L. (1991) The
permeability of soils with organic fluids, Canadian Geotech. J., 28(1), 140-7.
Cartwright, K. and Krapac, I.G. (1990) Construction and performance of a
long-term earthen liner experiment, in Waste Containment Systems:
Construction, Regulation, and Performance, (ed R. Bonaparte), ASCE, New
York, pp. 135-55.
Chamberlain, E.J. and Gow, A.J. (1979) Effects of freezing and thawing on the
permeability and structure of soil, Eng. Geol., 13, 73-92.
Chamberlain, E.J., Iskander, I. and Hunsicker, S.E. (1990) Effect of freeze-thaw
cycles on the permeability and macrostructure of soils, Proc., Int. Symp. on
Frozen Soil Impacts on Agricultural, Range, and Forest Lands, CRREL Special
Report 90-1, U.s. Army Corps of Engineers Cold Regions Research &
Engineering Laboratory, Hanover, Ne Hampshire, 145-55.
Chapuis, R.P. (1989) Shape factors for permeability tests in boreholes and
piezometers, Ground Water, 27(5), 647-54.
Chapuis, R.P. (1990) Soil-bentonite liners: predicting permeability from
laboratory tests, Canadian Geotech. J., 27(1), 47-57.
Daniel, D.E. (1984) Predicting hydraulic conductivity of clay liners, J. Geotech.
Eng., 110(2), 285-300.
Daniel, D.E. (1989) In situ hydraulic conductivity tests for compacted clay, J.
Geotech. Eng., 115(9), 1205-26.
Daniel, D.E. (1990) Summary review of construction quality control for
compacted soil liners, in Waste Containment Systems: Construction,
Regulation, and Performance, (ed R. Bonaparte), Am. Soc. of Civil Engineers,
New York, pp. 175-89.
Daniel, D.E. and Benson, C.H. (1990) Water content-density criteria for
compacted soil liners, J. Geotech. Eng., 116(12), 1811-30.
Daniel, D.E. and Estornell, P.M. (1990) Compilation of information on alternative
barriers for liner and cover systems, USEPA, EPA 600/2-91/002, Cincinnati,
Ohio.
Daniel, D.E., Trautwein, S.J., Boynton, 5.5. and Foreman, D.E. (1984)
Permeability testing with flexible-wall permeameters, Geotech. Testing T.,
7(3), 113-22.
Daniel, D.E., Trautwein, S.J. and McMurtry, D. (1985) A case history of
leakage from a surface impoundment, in Seepage and Leakage from Dams and
Impoundments, (eds R.L. Volpe and W.E. Kelly), Am. Soc. of Civil
Engineers, New York, pp. 220-35.
Estornell, P.M. (1991) Bench-scale hydraulic conductivity tests of bentonitic
blanket materials for liner and cover systems, M.S. Thesis, Univ. of Texas,
Austin, TX.
Fernandez, F. and R.M. Quigley (1985) Hydraulic conductivity of natural clays
permeated with simple liquid hydrocarbons, Canadian Geotech. J., 22(2),
205-14.
Fernandez, F. and Quigley, R.M. (1988) Viscosity and dielectric constant
controls on the hydraulic conductivity of clayey soils permeated with
water-soluble organics, Canadian Geotech. J., 25(4), 582-9.
Fernandez, F. and Quigley, R.M. (1991) Controlling the destructive effects of
162 Clay liners

clay - organic liquid interactions by application of effective stress, Canadian

Geotech. J., 28(3), 388-98.
Fireman, M. (1944) Permeability measurements on disturbed soil sample, Soil
Sc., 58, 337-55.
Foreman, D.E. and Daniel, D.E. (1986) Permeation of compacted clay with
organic chemicals, J. Geotech. Eng., 112(7), 669-81.
Gordon, M.E., Huebner, P.M. and Miazga, T.J. (1989) Hydraulic conductivity
of three landfill clay liners, J. Geotech. Eng., 15(8), 1148-60.
Grim, R.E. (1953) Clay Mineralogy, McGraw-Hill, New York.
Grube, W.E. Jr. (1991) Soil barrier alternatives, in Remedial Action, Treatment,
and Disposal of Hazardous Waste, USEPA, Cincinnati, Ohio, EPA/600/9-911
002, pp. 436-44.
Grube, W.E., Jr. and Daniel D.E. (1991) Alternative barrier technology for
landfill liner and cover systems, presented at the 84th Ann. Meeting &
Exhib. of the Air & Waste Management Assoc., Vancouver, British
Columbia, Paper 91-5.9.
Johnson, G.M., Crumbley, W.S. and Boutwell, G.P. (1990) Field verification of
clay liner hydraulic conductivity, in Waste Containment Systems:
Construction, Regulation, and Performance, (ed K Bonaparte), ASCE, New
York, pp. 226-45.
Keller, CK., van der Kamp, G. and Cherry, J.A. (1986) Fracture permeability
and groundwater flow in clayey till near Saskatoon, Saskatchewan,
Canadian Geotech. J., 23, 229-40.
Kim, W.H. and Daniel, D.E. (1992) Effects of freezing on the hydraulic
conductivity of a compacted clay, J. Geotech. Eng., 118(7).
Kleppe, J.H. and Olson, KE. (1985) Desiccation cracking of soil barriers, in
Hydraulic Barriers in Soil and Rock, ASTM STP 874, Am. Soc. for Testing and
Materials, Philadelphia, pp. 263-75.
McNeal, B.L. and Coleman, N.T. (1966) The effect of solution of soil hydraulic
conductivity, Proc., SoilSc. Soc. of Amer., 30(3), 308-12.
Mitchell, J.K. (1976) Fundamentals of Soil Behavior, John Wiley & Sons, New
York.
Mitchell, J.K., Hooper, D.R. and Campanella, KG. (1965) Permeability of
compacted clay, J. Soil Mech. and Foundations Div., ASCE, 91(SM4), 41-65.
Nasiatka, D.M., Shepherd, T.A. and Nelson, J.D.' (1981) Clay liner
permeability in low pH environments in Proc., Symp. on Uranium Mill
Tailings Management, Colorado State Univ., Fort Collins, Colorado, pp.
627-45.
Olson, KE. and Daniel, D.E. (1981) Measurement of the hydraulic conductivity
of fine-grained soils, in Permeability and Ground Water Contaminant
Transport, ASTM STP 746, (eds T.F. Zimmie and CO. Riggs), Am. Soc. for
Testing and Materials, Philadelphia, pp. 18-64.
Peterson, S.K and Gee, G.W. (1986) Interactions between acidic solutions and
clay liners: permeability and neutralization, in Hydraulic Barriers in Soil and
Rock, ASTM STP 874, Am. Soc. for Testing and Materials, Philadelphia,
pp.229-45.
Reades, D.W., Lahti, L.K, Quigley, KM. and Bacopoulos, A. (1990) Detailed
case history of clay liner performance, in Waste Containment Systems:
Construction, Regulation, and Performance, (ed K Bonaparte), ASCE, New
York, pp. 156-74.
Sai, J.O. and Anderson, D.C (1990) Field hydraulic conductivity test for
compacted soil liners, Geotech. Testing J., 13(3), 215-25.
Scheu, C, Johannben, K. and Saathoff, F. (1990) Non-woven bentonite fabrics
 References 163
- a new fibre reinforced mineral liner system, in Geotextiles, Geomembranes
and Related Products, (ed D. Hoet), Balkema Publishing, Rotterdam, pp.
467-72.
Schubert, W.K (1987) Bentonite matting in composite lining systems, in
Geotechnical Practice for Waste Disposal '87, (ed KD. Woods), ASCE, New
York, pp. 784-96.
Shackelford, CD. and Javed, F. (1991) Large-scale laboratory permeability
testing of a compacted clay soil, Geotech. Testing J., 14(2), 171-91.
Shakoor, A. and Cook, B.D. (1990) The effect of stone content, size, and shape
on the engineering properties of a compacted silty clay, Bul. Assoc. of Eng.
Geol., XXVII(2), 245-53.
Shan, H.Y. (1990) Laboratory tests on bentonitic blanket, M.S. Thesis, Univ. of
Texas, Austin, TX.
Shan, H.Y. and Daniel, D.E. (1991) Results of laboratory tests on a geotextile/
bentonite liner material, in Geosynthetics 91, Industrial Fabrics Assoc. Int.,
St. Paul, Minnesota, 2, pp. 517-35.
Shelley, T.L. (1991) Effect of gravel on hydraulic conductivity of compacted soil
liners, M.S. Thesis, Univ. of Texas, Austin, Texas.
Zimmie, T.F. and La Plante, C (1990) The effects of freeze/thaw cycles on the
permeability of a fine-grained soil, in Proc., 22nd Mid-Atlantic Ind. Waste
ConI, Drexel Univ., Philadelphia.
 CHAPTER 8

Geomembrane liners
Robert M. Koerner

As defined by the American Society for Testing and Materials (ASTM),

geomembranes are 'very low permeability synthetic membrane liners
or barriers used with any geotechnical engineering related material
so as to control fluid migration in a man-made project, structure or
system'. They are used exclusively as liquid or vapor barriers and have
wide application in broad areas of environmental and transportation
engineering practice, as well as in geotechnical engineering. Within
the environmental applications area is the containment of solid waste
materials which is the focus of this book. Geomembranes, which are
the focus of this particular chapter, form an essential part of the overall
liner system.
The outline of the chapter will be the following sections.
8.1 Historical perspective of polymeric liners
8.2 Current use in waste disposal practice
8.3 Geomembrane types and manufacture
8.4 Test methods and properties
8.5 Design models and examples
8.6 Construction quality control and assurance
8.7 Durability and other concerns
8.8 Future trends and growth

8.1 HISTORICAL PERSPECTIVE OF POLYMERIC LINERS

The genesis of modern liners, which in the context of geosynthetics

are called geomembranes, is intrinsically tied to the development and
growth of the polymer industry. According to Staff (1984) the earliest
type of thin prefabricated polymer sheets placed on a prepared soil

Geotechnical Practice for Waste Disposal.

Edited by David E. Daniel.
Published in 1993 by Chapman & Hall, London. ISBN 0412351706
 Historical perspective of polymeric liners 165

base as currently practiced, was the use of polyvinyl chloride as

swimming pool liners in the early 1930s. Bell and Yoder (1957) describe
transportation related trials that were practiced by the Bavarian
Highway Department using low density polyethylene seepage barriers
in the late 1930s. For large fabricated sheets, however, it seems as
though the synthetic thermoset materials made the first significant
impact. Potable water reservoirs sealed with butyl rubber liners are
reported by Staff (1969) in the late 1940s. A testing and design oriented
approach toward the lining of water canals was undertaken by the US
Bureau of Reclamation in the 1950s (Hickey, 1957). The research was
performed by Luritzen (1957) who worked closely with manufacturers
and resin suppliers. A bevy of worldwide activity using a wide range
of polymeric material ensued. As seen in Fig. 8.1, polyvinyl chloride
canal liner installations were made in Canada, Russia, Taiwan, and in
Europe throughout the 1960s and 1970s. Chlorosulfonated polyethylene
was developed in the USA and Europe and created a major impact.
Polyethylene liners were developed in West Germany and spread to all
of Europe, Africa, Australia and North America. South Africa was very
much involved in polyethylene liner development which it exported
and then further developed in North America.
Today, polymeric geomembranes are indeed worldwide in both their
availability and applicability. They are truly a subset of geosynthetics

-"- g~omembranes ________ _ Influence of Ihennoplastic

).-,( stili little used and thermoset liners

L -_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

Fig. 8.1 Initial development and movement of geomembranes on a

worldwide basis, after Koerner (1990a).
166 Geomembrane liners

and.a major category in their own right. In the applications of interest

in this book, i.e., solid waste landfill liners, they are a primary focus of
attention for the liner system containing the waste. Note that the liner
system referred to is often a composite liner consisting of a geomem-
brane placed in intimate contact with a clay soil of the type described in
Chapter 7 or with a geosynthetic clay liner of the type described in
Chapter 18. In such cases that other geosynthetics may be used for
drainage, filtration, separation or reinforcement (such as geotextiles,
geonets, geogrids and geocomposites) they are auxiliary to the proper
and long-term functioning of the liner system.

8.2 CURRENT USE IN WASTE DISPOSAL PRACTICE

Perhaps the most significant research effort leading to the mandated

use of geomembranes in waste disposal facilities is that of Anderson
and Brown (Anderson et al., 1981). They presented hydraulic con-
ductivity data of different clay soils using various organic solvents
which resulted in extremely high values over permeation of the same
clays with water. While this data was somewhat flawed by virtue of
their experimental procedures (e.g., use of 100% solvents and fixed
wall permeameters), it resulted in immediate action by the USEPA
which issued the following regulation in July, of 1982.
Prevention (via geomembranes), rather than minimization (via
clay soil liners), of leachate migration similarly produces better
environmental results in the case of surface impoundments (land-
fills) used to dispose of hazardous wastes. A liner that prevents
rather than minimizes leachate migration provides added assurance
that environmental contamination will not occur.
This regulation set in motion a number of different liner systems which
consisted of single geomembranes, redundant geomembranes, com-
posite geomembrane/clay liners, redundant composite,liners, and other
variations which appeared on almost an annual basis (Table 8.1). As a
result of this progression of events, the type of liner system listed as a
double geocomposite appears is shown in Fig. 8.2. It is illustrated with
a variety of other geosynthetic materials such as geotextiles, geonets,
geocomposites and geogrids, since many solid waste facilities have
some or all of these polymeric materials contained within them. The
drainage and filtration geosynthetic materials will be described in
Chapter 9. Also illustrated in Fig. 8.2 is the cover system above the
waste with its own composite liner plus a variety of other possible
geosynthetic materials. Included in this cover system is the possible
use of geosynthetic clay liners. These aspects are covered in detail in
Chapters 7 and 18. This chapter, however, will concentrate completely
 Geomembrane types and manufacturing 167
on geomembrane liners. Other references, e.g., Koerner (1990a), are
available on the entire gambit of geosynthetic materials.

8.3 GEOMEMBRANE TYPES AND MANUFACTURING

There are three categories of polymers that can be used to make

geomembranes; thermoset elastomers, thermoplastics and bituminous
types. At this point in time, thermoset elastomers are rarely used
in waste containment due primarily to the difficulty in making field
seams. An adhesive tape is used which generally does not show good
durability in accelerated aging tests. Also, bituminous geomembranes

Table 8.1 Genesis of liner systems used in the USA, after Koerner (1990b)

approx. type of liner systems PLCR system P-GM LDCR S-GM

date system
--"~--.-~

pre-1982 single clay soil/pipe clay none none

1982 single GM soil/pipe GM none none
1983 double GM soil/pipe GM soil/pipe GM
1984 single GM, single GC soil/pipe GM soil/pipe GM/clay
1985 single GM, single GC soil/pipe GM GN GM/clay
1987 double GC soil/pip GM/clay GT/GN GM/clay
1990 double GC GT/GC or GM/GCL GT/GN GM/clay
GT/GN/GN

Abbreviations: GC = geocomposite liner or drain; GCL = geosynthetic clay liner; GM =

geomembrane barrier; GN = geonet drain; GT = geotextile filter; LDCR = leak detection,
collection and removal system; PLCR = primary leachate collection and removal system;
P-CM = primary geomembrane barrier; S-GM = secondary geomembrane barrier.

~ ~~~~~~~~~~~~ geocomposite drain

geomembrane
eos nt ellc Ca geolextiiefiller
gOOlexli., gas venl

)~~~~~m~~~~~~::ge~ot.exti,e filter
geonet

secondary g~:':::~~d~~~e
geomembrane

Fig. 8.2 Cross-section of solid waste landfill liner system and closure system
illustrating the major use of geosynthetics.
168 Geomembrane liners

Table 8.2 Typical range of formulations for thermoplastic geomembranes

type of resin plasticizer carbon black and filler additivet

geomembrane* (%) (%) (%) (%)
~---

HOPE 96-97 0 2-2.5 0.5-1

VLOPE 96-97 0 2-2.5 0.5-1
CPE or CPE-R 60-75 10-15 20-30 3-5
CSPE-R 45-50 2-5 45-50 2-4
EIA-R 50-65 10-20 20-30 3-5
PVC 45-50 35-40 10-15 3-5

* the 'R' designation refers to fabric reinforcement

t includes antioxidant, processing aids and lubricants

are rarely used in North America, but still see some use in Europe,
particularly in France. Thus essentially all of the geomembranes being
used today in North America are in the thermoplastic category. By
thermoplastic it is meant that the polymer can be melted and with the
removal of heat will revert to its original structure (or nearly so).
Currently, the most commonly used thermoplastic geomembranes are
the following.
• high density polyethylene (HOPE) - either smooth or textured
• very low density polyethylene (VLOPE) - either smooth or textured
• chlorinated polyethylene (CPE) - either nonreinforced or reinforced
with a fabric scrim
• chlorosulfonated polyethylene (CSPE) - generally reinforced with a
fabric scrim
• ethylene interpolymer alloy (EIA) - always reinforced with a fabric
• polyvinyl chloride (PVC)
Note, however, that within the above generic categories various
additives, fillers and/or plasticizers are used with the base resin which
gives the geomembrane its name (Table 8.2).
Thus it is seen that the type of geomembrane selected is actually a
compound, which the geomembrane formulator makes into the final
sheet product. The manufacturing can be done in one of three ways;
extrusion, calendering or spread coating (Fig. 8.3).
Both HOPE and VLOPE geomembranes are made by extrusion pro-
cesses. A continuous screw extruder accepts the formulated compound
from a hopper, melts it under vacuum and delivers it through rollers in
either a wide flat sheet final product form, or in the form of a circular
disc. The circular disc is carried upward in the form of a large cylinder,
brought over a nip bar, cut, unfolded and rolled into the final sheet
form. This latter technique is called the 'blown film' process. Widths of
up to 10.7 m can be made by the various types of extrusion processes.
 Geomembrane types and manufacturing 169

polymer resin
and
additives

Fig. 8.3 Various processes to manufacture geomembranes, after Haxo (1986).

They are sent directly to the construction site in rolls for final place-
ment and field seaming.
CPE, CPE-R, CSPE-R and PVC are made by the calendering pro-
cess. The formulated compound is blended in a high shear-flow mixer
into a viscous form and fed into the calender which consists of counter
rotating sets of rollers forming the individual plys of the geomembrane.
Multiple laminates can be built up using an 'inverted-L' calender and
often a fabric scrim is embedded between the individual plys. In such
cases, the geomembrane is said to be scrim reinforced and carries
an 'R' designation. Calendered geomembranes are usually made into
1.8 m wide rolls which are then factory seamed into wider panels,
consisting of 5 or 6 roll widths, in a fabrication facility. They are then
accordion folded in both their width and their length directions, placed
on wooden pallets, and sent to the construction site for final placement
and field seaming.
The spread coating process of manufacturing geomembranes is the
least widely practiced. It consists of the formulated compound being
screeded over a fabric reinforcing substrate. It is often placed on both
sides of the fabric, as with EIA-R geomembranes, or can be on one-
side only as with certain geomembrane/geotextile composite liners. The
finished product may be factory seamed into wider panels or sent
directly to the site in rolls for final placement and field seaming.
A closing comment in this section should be made as to quality
170 Geomembrane liners

control of geomembrane manufacturing in a factory environment. As

might be expected, manufacturing quality control can be, and usually
is, very good. This is to be expected, since factory temperature is
generally controlled, harsh weather conditions are eliminated, and the
work force is stable and (hopefully) dedicated to a single task.

8.4 TEST METHODS AND PROPERTIES

The individual properties of the polymeric geomembranes just described

can be grouped into categories as shown below. Regarding test methods
for geomembranes per se, an ASTM subcommittee under the Soil and
Rock Committee, designated as 018.20, Impermeable Barriers, was
formed in 1982. In 1985, however, a special subcommittee of 0-35
began considering geomembranes. Today, 0-35 has been renamed as
'Committee 035 on Geosynthetics'. Many standards are completed
within this committee and still others are in various stages of develop-
ment. Note that there are many additional standards with geomem-
brane applicability in the plastics, rubber, roofing, and water canal
divisions of ASTM, the National Sanitation Foundation (NSF) and the
Geosynthetic Research Institute (GRI). Whenever possible the ASTM
035 test method will be referenced. Other ASTM standards will have
the next priority, followed by the other groups that were mentioned.
Lastly, references from the open literature will be cited if no standard-
ized test procedures exist. This testing section is subdivided into the
following categories, as per Koerner (1990b):
• physical properties
• mechanical properties
• chemical properties
• biological properties
• thermal properties

8.4.1 Physical properties

Included in this grouping are the thickness of the geomembrane, via
ASTM 0751, the density (or specific gravity of the polymer), via ASTM
0792 or 01505, and the mass per unit area, via ASTM 01910. Also the
water vapor transmission, via ASTM E96, may be considered in this
category as well as solvent vapor transmission (Haxo, 1988).

8.4.2 Mechanical properties

Tensile behavior of geomembranes is obviously important and there
are a large number of index tests including ASTM 0412, 0638, and
0882. The test specimens are narrow width or dumbbell shaped,
 Test methods and properties 171
in these particular tests. For a wide width performance test, ASTM
04885 is recommended. In this case the test specimen is 200 mm wide
in an attempt to simulate plane strain conditions. Three dimensional,
out-of-plane, tension tests are just emerging and GRI GM-4 can be
used as an interim guide until a more formalized test method appears.
Seam tests in both shear and peel modes are important and NSF No. 54
(1991) should be consulted for details which depend upon the particu-
lar type of geomembrane being evaluated.
Other index strength tests for geomembranes include tear tests via
ASTM 02263, 01004, 0751, 01424, 02261 or 01938, which all cover
the general topic. Impact is handled via ASTM 01709, 03029 or 03998.
Puncture resistance is measured via ASTM 02582 or 03787, but also
Federal Test Method No. 10lC is often referenced in the literature.
The important test evaluation of interface friction between geomem-
branes and other geosynthetics or soil has no prescribed test procedure.
Most testing laboratories use a variation of the direct shear soil test
which is covered under ASTM 0-3080. There are many unanswered
questions, however, in using this test for geomembranes such as the
following.
• length and width of the shear box
• method of geomembrane fixity to shear box
• saturation conditions of test specimens
• shear rate of conducting the test
• interpretation of an adhesion value when the results indicate a
vertical axis intercept
Geomembrane anchorage behavior has been briefly evaluated and can
be addressed by GRI Method GM-2.
Stress cracking of semi-crystalline geomembranes (such as HOPE) is
conventionally evaluated by the bent strip test, ASTM 01693, but one
should also consider a notched constant load test such as GRI Method
GM-5. The current recommendation in this test is to require at least 100
hours transition time between the ductile and brittle behavior of the
geomembrane test specimens (Halse et ai., 1989).

8.4.3 Chemical resistance

The chemical resistance of geomembranes to water is evaluated
indirectly using swelling resistance via ASTM 0570, ozone resistance
via ASTM 01149 and ultraviolet light resistance via ASTM 03334. In
this latter regard there are a number of accelerated laboratory testing
methods covered under ASTM G23, G26 and G53.
Chemical resistance of geomembranes to leachate is generally per-
formed using EPA Method 9090 (USEPA, 1985). Note, however, that
ASTM has a task group on this activity and should have a test method
172 Geomembrane liners

finalized in the very near future. In the evaluation procedure one takes
replicate samples of the candidate geomembrane and incubates them
in the site specific leachate at a constant temperature. Two sets of
tests are necessary, one at 23°C, the other at 50 0c. The samples are
sequentially retrieved at 30, 60, 90 and 120 days. A suite of index
strength tests are then performed to note any changes that may have
occurred from the original, as-received, material. This requires a sub-
jective assessment of the data as to the degree of chemical resistance
of the candidate geomembrane. The assessment procedure is not
covered by either the EPA or the ASTM (proposed) test method. Note
should also be mentioned that the 50°C incubation is generally the
more aggressive situation and shows the greater variation from the as-
received geomembrane material.

8.4.4 Biological resistance

Geomembrane resistance to fungi can be evaluated via ASTM G21 and
resistance to bacteria via ASTM G22. Quite clearly there is little concern
for biological degradation of the resin (with its very high molecular
weight), or of the inert fillers that may be present. Questions may
be raised, however, as far as plasticizers (recall Table 8.2), but no
authoritative information is available to the author's knowledge.
Burrowing animals are another matter. If an animal with sharp and
hard teeth wants to penetrate a geomembrane there is a high likelihood
that this may occur. The possible exception being thick HOPE geo-
membranes. Why an animal would want to do this is another matter
entirely. Bio-barriers, consisting of large rocks have been designed to
thwart the animals in this regard.

8.4.5 Thermal properties

The behavior of geomembranes to extreme swings in temperature can
be evaluated for high temperatures via ASTM 0794 and for cold tem-
peratures via simulated testing conditions.
The coefficient of linear expansion/contraction is evaluated by means
of ASTM 02102 and 02259 for contraction and 01042 for expansion
and dimensional change. It should be mentioned that polyethylene
geomembranes have the highest values of thermal expansion/contrac-
tion of the currently manufactured products.

8.5 DESIGN MODELS AND EXAMPLES

The essence of design for any engineering material is to compare an

allowable (or test) property with a required (or design) property for a
 Design models and examples 173
resulting factor-of-safety. This is conventionally viewed as follows:

factor-o f-sa fety (5 allowable (test) property

F ) = ------'-----'---"-~-""-- (8.1)
required (design) property
Since we have just covered the major available test methods from
which the properties are obtained, focus now shifts to various design
methods and models. For geomembranes, a number of these models
are available. Among the more important ones are the following:
(a) thickness design due to bending
(b) strength design due to subsidence
(c) strength design due to unbalanced shear stresses
(d) anchorage design due to embedment
for which the free-body diagrams for each situation are shown on Figs.
8.4a-d respectively. The requisite design equations are as follows:
for bending thickness
p x
t reqd = --A - - (tan i5 u + tan i5d (8.2)
cos I-' Gallow

FS (8.3)

for subsidence strength

2DL2p
G -
reqd - 3t(D2 + L2)
-----=---'--~ (8.4)

Gallow
FS (8.5)

for tensile strength

Treqd = [(cau - cad + p cos (J) (tan i5 u - tan i5L )] LW (8.6)
Tallow
FS (8.7)
Treqd
for anchorage design
Gallow(t) = p tan 15 (L RO ) + 2(KoGvave )(tan (5)(d AT ) (8.8)
which is solved implicitly for LRO and dAT and a suitable multiplier is
then used for the desired factor-of-safety value. The above formula-
tions are derived (along with illustrative problems) in Koerner (1990b)
and Koerner and Hwu (1991). The following constants are defined for
these equations.
t = geomembrane thickness
P = mobilizing pressure, e.g., the full weight of waste
 x

(a)

cover soil
unit weight, y

o geomembrane

(b)

Fig. 8.4 Free-body diagrams for various geomembranes design scenarios: (a)
thickness design due to bending; (b) strength design due to subsidence; (c)
strength design due to unbalance shear; and (d) anchorage design due to
embeddment.
 T (in geomembrane)

-~

(c) where

tu = Cau + (W cos 0)) tan Ou

tl = Cal + (W COS 0)) tan Ol

cover soil

imaginary and
frictionless pulleys

(d)
176 Geomembrane liners

O"allow = allowable geomembrane stress; which is yield for

HDPE, fabric break for CPE-R, CSPE-R and EIA-R
and the stress at 50% strain for CPE, VLDPE and
PVC
O"reqd = required geomembrane stress
Tallow = allowable geomembrane strength; which is yield for
HDPE, fabric break for CPE-R, CSPE-R and EIA-R
and the strength at 50% strain for CPE, VLDPE and
PVC (Note: Tallow = O"allowt)
Treqd = required geomembrane strength
X = anchorage mobilization distance
~, ill = various slope angles
Ou and OL = friction angles (upper and lower) with respect to the
geomembrane
Cau and CaL = adhesion values (upper and lower) with respect to the
geomembrane
= various dimensional values
= length of runout
= depth of anchor trench
In performing the above calculations for factor-of-safety it must be
emphasized that the test value for allowable properties of thickness,
stress, strength, etc., must simulate the in situ behavior of the material.
In this regard, partial factors-of-safety might be considered for labora-
tory generated results which do not simulate the anticipated field
performance, e.g., for degradation and aging of the geomembrane.

8.6 CONSTRUCTION QUALITY CONTROL AND ASSURANCE

(CQC AND CQA)

Construction quality control of geomembranes during manufacturing is

an important consideration and most manufacturers and many large
owner/operators have excellent guides in this regard. They should be
consulted for the many details that go beyond this brief coverage of the
subject.
Storage at the manufacturing facility, shipment to the site and storage
at the construction site are important considerations. Roll or panel
deployment is the first field activity for CQC operations and while
much can be said, one aspect is very important, i.e., protection from
wind uplift damage. Clearly, large areas of exposed geomembrane are
very dangerous since the number of sandbags required to hold geo-
membranes down under the wind uplift forces becomes huge as wind
velocities increase (Wayne and Koerner, 1988). The geomembrane
should be covered immediately after field seaming and the acceptance
 Construction quality control and assurance 177
of these seams by the parties involved. This acceptance procedure
often goes beyond CQC and involves a separate inspection team,
called the construction quality assurance (CQA) auditor. This group
is an added level of inspection to see that the construction activities
are indeed being undertaken according to the project's plans and
specifications.
Regarding the field seaming of geomembranes, there are three major
categories; they are the extrusion welding, thermal fusion or melt bonding
and solvent/adhesive seaming methods. Each will be explained along
with their specific variations so as to give an overview of field seaming
technology, for additional detail see Frobel (1984) and Koerner (1990d).
Extrusion welding is used exclusively on geomembranes made from
polyethylene. It is a direct parallel of metallurgical welding in that a
ribbon of molten polymer is extruded over the edge of, or in between,
the two surfaces to be joined. The surfaces upon which the extrudate is
to be placed must be ground to remove oxides and waxes prior to the
extrudate being placed. The hot extrudate causes the surfaces of the
sheet material to become molten and the entire mass then cools and
fuses together. The technique is called extrusion fillet welding when
the extrudate is placed over the leading edge of the seam, and is called
extrusion flat welding when the extrudate is placed between the two
sheets to be joined. It should be noted that extrusion fillet welding is
essentially the only method for seaming polyethylenes that are used
for patching and in poorly accessible areas like sump bottoms and
around pipes.
There are three seaming methods based on thermal fusion, or
melt bonding, that can be used on both semicrystalline and flexible
thermoplastic geomembranes. In each of them, portions of the opposing
surfaces are melted without the addition of new polymer. This being
the case, temperature, pressure and seaming rate all play important
roles in that too much melting weakens the geomembrane and too little
melting results in poor seam strength. The hot air method makes use
of a device consisting of a resistance heater, a blower, and temperature
controls to blow hot air between two sheets to melt the opposing
surfaces. Usually, temperatures of the 'gun' greater than 250°C (480 OF)
are required. Immediately following the melting of the surfaces,
pressure is applied by counter-rotating knurled rollers on the top and
bottom of the seamed area. The hot wedge or hot shoe method is the
widest practiced method for joining polyethylene geomembranes. It
consists of an electrically heated resistance element in the shape of a
wedge that travels between the two sheets to be sealed. As it melts the
surfaces, a shear flow occurs across the upper and lower surfaces of
the wedge and then roller pressure is applied. Hot wedge units are
automated as far as temperature, amount of pressure applied and
travel rate is concerned. A standard hot wedge creates a single uniform
178 Geomembrane liners

width seam, while a dual hot wedge (or 'split' wedge) forms two
parallel seams with a uniform unbonded space between them. This
space can be used to evaluate seam quality and continuity by pressuriz-
ing it with air and monitoring any drop in pressure that may signify a
leak in the seam. A newest variation of thermal fusion seaming of
geomembranes is the ultrasonic seaming method. Here a vibrating
knurled horn is passed between the two geomembranes to be bonded.
The vibrational energy (about 40 000 Hz) causes melting of the surfaces
which when forced together by pressure from rollers results in the
completed seam.
Regarding the solvent and/or adhesive seam types; solvent seams
make use of a liquid solvent squirted from a squeeze bottle between
the two geomembrane sheets to be joined. After a few seconds,
pressure is applied to make complete contact. As with any of the
solvent-seaming processes to be described, a portion of the two
adjacent materials to be bonded is truly dissolved into a liquid phase.
Too much solvent will weaken the adjoining sheets, and too little
solvent will result in a weak seam. Therefore great care is required in
providing the proper amount of solvent with respect to the particular
type and thickness of geomembrane. Care must also be exercised in
allowing the proper amount of time to elapse before contacting the two
surfaces, and in applying the proper pressure and duration of rolling.
Bodied solvent seams are similar to solvent seams except that 5-15%
of the parent lining compound is dissolved in the solvent which is then
used to make the seam. The purpose of this procedure is to com-
pensate for the lost material while the seam is in a liquid state and
to fill in surface dimples or undulations on reinforced liners. The
viscous liquid is brushed onto the two opposing surfaces to be bonded.
Pressure is necessary, and the use of heat guns or radiant heaters
accelerates the curing process. Solvent adhesive seams make use of an
adherent, within a bodied solvent solution, which is left after dissipa-
tion of the solvent. The adherent thus becomes an additional element
in the system. Brushing is used to apply the three-part liquid (solvent,
parent liner compound, and adhesive) onto the two opposing surfaces.
Sufficient pressure must be used and heat guns or radiant heaters are
often used to accelerate the curing process. Contact adhesive seams
have the widest applicability to all geomembrane types but are mainly
focused on the thermoset elastomeric geomembranes such as EPDM,
butyl, nitrile, etc. The solution is applied to both mating surfaces by
brush and then rolled. After reaching the proper degree of tackiness,
the two sheets are placed on top of one another, followed by roller
pressure. The adhesive forms the bond and is an additional element in
the system. Due to the lack of current use of thermoset geomembranes,
the technique is not used at this time.
In order to give an overview as to which seaming methods are used
 Durability and other concerns 179

Table 8.3 Most commonly used field seaming methods for various
geomembranes

type of seaming types of geomembranes

method - -------- -----~

HDPE VLDPE CPE CPE-R CSPE-R EIA-R PVC

----------- ----- - - - - - - - -------

extrusion fillet P P nla nla n/a n/a n/a

extrusion flat P n/a nla n/a n/a n/a n/a
hot air S P S S S P S
hot wedge P P S S S P S
solvent n/a n/a S S S n/a P
bodied n/a n/a P P P n/a P
solvent adhesive n/a n/a S S S n/a S

Note: P = primary method used; 5 = secondary method used; nla = method is 'not
applicable' .

for the various thermoplastic geomembranes that are currently being

used (Table 8.3).
Upon fabrication of the geomembrane field seams, selective destruc-
tive test samples should be taken and from these, shear and peel tests
performed. Typical sampling frequency is one per 150-300 m. The
sample is often 1 m in length, with ~ being evaluated on the job site
by the CQC personnel, ~ being sent to the CQA firm for laboratory
testing and the last ~ being archived for future reference, if necessary.
The shear and peel tests should be assessed on the basis that the sheet
material adjacent to the seam area fails without the seam delaminating.
This is called a 'film tear bond' (Haxo, 1988). Additionally, the strength
of the seam in shear should be 70-90% of the strength of the percent
sheet. For the seam tested in peel, this value is from 50-70% for
polyethylene and above 1.7kN/m for other geomembranes (NSF 54,
1991).
Lastly, all geomembrane field seams in between the locations where
destructive samples are removed, should be evaluated by a suitable
nondestructive test method. Table 8.4 gives an overview of these test
methods, but see references in Koerner (1990c) for additional details. A
conference proceedings on the seaming of geosynthetics with twenty
papers on geomembrane seaming and inspection is also available,
Koerner (1990d).

8.7 DURABILITY AND OTHER CONCERNS

The degradation of polymer materials is known to occur via a number

of different phenomena and mechanisms. Table 8.5 gives an over-
view of the situation where it can be seen that for most cases proper
Table 8.4 Overview of nondestructive seam tests, after Koerner (1990c)

nondestructive test primary user general comments

method
installation project design third-party cost of speed of cost of type of recording operator
contractor engineer auditor equipment tests tests result method dependency
(CQC) (CQA) (CQA)
----

air lance yes $200 fast nil yes-no manual very high
pick test yes nil fast nil yes-no manual very high
electric wire yes yes $500 fast nil yes-no manual high

dual seam yes yes $200 fast mod. yes-no manual low
(positive pressure)
vacuum chamber yes yes $1,000 slow very high yes-no manual very high
(negative pressure)

ultrasonic pulse echo yes yes $5,000 moderate high yes-no automatic moderate
ultrasonic impedance yes yes $7,000 moderate high qualitative automatic unknown
ultrasonic shadow yes yes $5,000 moderate high qualitative automatic moderate

electric field yes yes yes $20,000 slow high yes-no manual and low
automatic
acoustic sensing yes yes yes $1,000 fast nil yes-no manual moderate
Table 8.5 Degrada.tion phenomena in geomembranes (from a geosynthetics engineering perspective), after Koerner, Halse and
Lord (1991)

degradation degradation initial change in material subsequent change in material preventative

------------- ----- ---
classification mechanism measure
laboratory* jield t laboratoryt jield§
-------- -------_._---------
ultraviolet • chain scission • mol. wt. • color • elongation • color • screening agent
• bond breaking • stress crack • crazing • modulus • cracking • antidegradient
resist. • strength • cover the
geomembrane
radiation • chain scission • mol. wt. • color • elongation • color • cover for f3 and
• stress crack • crazing • modulus • cracking a-rays
resist. • strength • shield for
neutrons
• reduce dosage
for y rays
chemical • reaction with • carbonyl • texture • elongation • texture • proper resin
structure index • color • modulus • cracking • proper additives
• reaction with • IR • crazing • strength • reactions
additives • mol. wt. • TCA
swelling • liquid • TCA • thickness • thickness • thickness • proper resin
absorption • color • modulus • softness • proper
• texture • strength manufacturing
extraction • additive • TCA • texture • elongation • texture • proper
expulsion • IR • color • modulus • color compounds
• thickness • thickness • proper
manufacturing
Table 8.5 Continued

Table 8.5 Degradation phenomena in geomembranes (from a geosynthetics engineering perspective), after Koerner, Halse and
Lord (1991)

degradation degradation initial change in material subsequent change in material preventative

-----------
classification mechanism measure
laboratory' jie/d t laboratoryt jield§
----- - - -

delamination • adhesion loss • thickness • thickness • thickness • layer separation • proper

• edge effects • edge effects • ply adhesion • thickness manufacturing
• protect
geomembrane
edges
oxidation • reaction with • IR • color • elongation • color • anti-oxidant
structure • carbonyl • crazing • modulus • cracking • cover with soil
index • strength • cover with liquid
biological • reactions with • mol. wt. • texture • elongation • texture • avoid sensitive
additives • carbonyl • surface film • modulus • surface layer additives
index • strength • cracking • biocide
• IR
* Initial laboratory changes are generally sensed by chemical fingerprinting methods: mol. wt. molecular weight; IR = infrared spectroscopy;
TGA = thermal geometric analysis; OIT = oxidation induction time.
t Initial field changes are generally sensed on a qualitative basis.
t Subsequent laboratory changes can be sensed by numerous physical and mechanical tests. Listed in the table are those considered to be the most
sensitive parameters.
§Subsequent field changes are a continuation of the initial changes until physical and mechanical properties begin to visually change.
 Future trends and growth 183

materials, construction, and timely covering will eliminate many of

the potential problems. This, however, is not the case for oxidation
degradation which will eventually bring the geomembrane from its
normally ductile behavior into a brittle mode. Thus the geomembrane
will eventually become sensitive to ground movements, subsidence,
excessive stresses, etc. It should be noted that elevated temperature
over a normal background level greatly increases the reaction and its
negative effects.
Thus it is necessary to provide insight into different predictive
methods for assessing lifetime. There are at least three such methods:
1. the rate process method, using ductile-to-brittle experimental
curves, to extrapolate the response into the site-specific tempera-
tures and minimum acceptable stress levels;
2. Arrhenius modeling (again using elevated temperature) obtaining a
reaction rate which is assumed to simulate time at the site-specific
temperature; and
3. Hoechst multiparameter approach which superimposes ductile-to-
brittle behavior, stress relaxation behavior and field obtained strain
behavior to form the basis of predictive technique.
These methods are discussed in Koerner et al. (1991) and also in a
recent seminar on the durability and aging of geosynthetics (Koerner,
1989). It is perhaps interesting to comment on the author's personal
observation that geomembranes are not easy to place properly, but will
generally serve their intended function for time frames far longer than
most people suspect.

8.8 FUTURE TRENDS AND GROWTH

Projections into the future status of geomembranes are difficult at best.

Furthermore, they are usually biased on the part of the person making
the prediction by reason of personal knowledge of certain specific
situations. The position from the author's point of view is that of a very
strong environmental awareness in the area of groundwater pollution.
Solid waste containment (of the type described in this book) is clearly
powering the geomembrane market in North America. There is every
indication that this 'mind-set' will continue and be spread world-
wide. Thus environmental applications will continue to be strong with
a long-term extended growth pattern.
With an ever-increasing worldwide population growth, water
supplies will become more scarce and require storage reservoir and
surface impoundment liners. Similarly, canal liners for the transporta-
tion of this water to areas of their primary use is an important applica-
tion. Recent concerns of potable water reservoir pollution from the
 184 Geomembrane liners

atmosphere has ushered in the concept of floating geomembrane

covers.
With industrialization on a worldwide basis comes the necessity of
canal liners to transport chemicals and waste waters. The application
here is among the most demanding of all geomembrane applications.
With continued applications of geomembranes in the above men-
tioned environmentally related categories will come an awareness on
the part of many transportation and geotechnical engineers for the use
of geomembranes in many applications where a liquid or vapor barrier
is necessary. Certainly rehabilitation of our infrastructure such as
remediation of concrete, masonry and soil/rock dams is an area which
can use geomembranes of a wide variety of types (Cazzuffi, 1987).
With familiarity, comes additional use and even forays into areas
where geomembranes have not previously been used.
The flow chart of Fig. 8.5 of potential use and opportunities (along

I geomembrane applications
I
>¥ >¥ >¥
r
transportation
applications
1
1 environmental
applications I
geotechnical
applications
I
I
r materials 1-
I design ~ , r
-I opportunities
I testing ~
r inspection ~
r research 1-
r technical papers, books1-

r conferences, symposia1-
r education
I university courses ~
r professional courses 1- ~
familiarization
r mfgr.'s seminars

1
strong growth for
geomembranes
(20-25%/year over
next 10 years)

Fig. 8.5 Future growth and trends in the use of geomembranes.

 References 185

with education) should bring about a strong growth for geomembranes

in the near term, e.g., 20-25% over the next 10 years, and, quite
possibly, will be continued even beyond that time frame.

REFERENCES

Anderson, D.C., Brown, K.W. and Green, J. (1981) Organic leachate effects on
the permeability of clay liners, Proc. Natl. COIlf. 011 Mgmt. of Uncontrolled
Hazardous Waste Substances, HMCRI, Silver Spring, MD, pp. 223-9.
Bell, J .R. and Yoder, E.J. (1957) Plastic moisture barrier for highway subgrade
protection, Proc. Highway Research Board, Washington, DC, 36, 713-35.
Cazzuffi, D. (1987) The use of geomembranes in Italian dams, IntI. J. Water
Power and Dam Construction, March, 44-52.
Frobel, R.K. (1984) Methods of constructing and evaluating geomembrane
seams, Proc. Inti. Conf. on Geomembranes, Denver, CO, IFAI, 359-64.
Halse, Y.H., et aI. (1989) Methods to evaluate stress crack resistance of HDPE
FML sheets and seams, Proc. 1990 EPA Research Seminar.
Haxo, H.E., Jf. (1986) Quality assurance of geomembranes used as linings for
hazardous waste containment, ]. Geotextiles and Geomembranes, Elsevier
App!. Sci. Pub!. Ltd., England, 225-47.
Haxo, H.E. (1988) Lining of waste containment and other impoundment facilities,
US EPA, 600/2-88/052.
Hickey, M.E. (1957) Evaluation of plastic films as canal lining materials, US
BuRec Lab. Report B-25.
Koerner, R.M. (ed.). (1989) Durability and Aging of Geosynthetics, Elsevier App!.
Sci. Pub!., London.
Koerner, KM. (1990a) Geomembrane overview, significance and background,
in Geomembranes: Identification and Performance Testing, (eds A.L. Rollin and
J.-M. Rigo) RILEM TC-103.
Koerner, R.M. (1990b) Designing with Geosynthetics, 2nd edn, Prentice Hall,
Englewood Cliffs, NJ.
Koerner, KM. (1990c) Preservation of the environment via geosynthetic
containment systems, Proc. 4th Int!. Conf. on Geotextiles and Geomembranes,
The Netherlands, June 1990 (reprinted in Geotech. Fabrics Rep. Sept/Oct.
1990 and NovlDec. 1990 issues).
Koerner, KM. (ed.). (1990d) The Seaming of GeosYllthetics, Elsevier App!. Sci.
Pub!., London, 1990.
Koerner, R.M., Halse, Y.H. and Lord, A.E., Jf. (1991) Long-term durability and
aging of geomembranes, in Waste Containment Systems: Construction,
Regulation and Performance (ed. R. Bonaparte), American Society of Civil
Engineers, New York, pp. 106-34.
Koerner, R.M. and Hwu, B-L. (1991) Stability and Tension Considerations
Regarding Cove Soils on Geomembrane Lined Slopes, J. Geotextiles and
Geomembranes, 10(4), 335-55.
Luritzen, C. W. (1957) Seepage control with plastic films, in Irrigation
Engineering and Maintenance, pp. 18-33.
National Sanitation Foundation Standard No. 54, (1991), Ann Arbor, Michigan.
Staff, C.E. (1969) Seepage prevention with impermeable membranes, Civ. Eng.,
37(2).
Staff, C.E. (1984) The foundation and growth of the geomembrane industry in
186 Geomembrane liners

the United States, Proc. Inti. Conf. on Geomembranes, Denver, CO, IFAI,
5-8.
USEPA (1985) Compatibility Test for Wastes and Membrane Liners, Method 9090,
Washington, DC.
Wayne, M.H. and Koerner, R.M. (1988) Effect of wind uplift on liner systems,
Geotech. Fabrics Rep., IFAI, 6(4), 26-9.
 CHAPTER 9

Collection and removal

systems
Robert M. Koerner

The collection of liquids in waste containment systems and their proper

removal represents an important element in the successful functioning
of these important facilities. Focus in this chapter is on primary and
secondary (leak detection) leachate collection systems beneath the
waste and on surface water removal in the cover system above the
waste. The gas collection system in the closure system will also be
addressed. The individual sections are as follows.
9.1 Overview and background
9.2 Hydraulic concepts, definitions and test methods
9.3 Primary leachate collection and removal system
9.4 Secondary leachate collection and removal system (i.e., leak
detection)
9.5 Surface water collection and removal system
9.6 Methane gas collection and removal system

9.1 OVERVIEW AND BACKGROUND

The general area of 'liquid management' is the focus of this chapter.

This refers to the timely and efficient removal of liquids out of, and
away from, the waste containment facility where they can be properly
treated and then disposed. The liquid itself comes primarily from
rainfall and snowmelt, but some is also contained in the solid waste as
it is brought to the landfill. Taken collectively, these liquids combine to
form leachate which has a wide variation of characteristics.
After solid waste is placed in the landfill, the leachate flows gravita-

Geotechnical Practice for Waste Disposal.

Edited by David E. Daniel.
Published in 1993 by Chapman & Hall, London. ISBN 0 412 35170 6
188 Collection and removal systems

tionally downward until it reaches the primary leachate collection

system (PLCS) as shown in Fig. 9.1. This layer consists of a drainage
material (usually gravel) protected by a filter layer (sand or geotextile) .
The leachate must gravitationally flow within this drainage material to
a sump area where it is removed by a submersible pump within a
manhole or large diameter pipe. Taken together, the drain, filter, sump
and removal manhole/pipe is called the primary leachate collection and
removal system, or PLCRS.
When dealing with solid waste liner systems having double liners,
i.e., primary and secondary, a secondary leachate collection system
(SLCS) must be included between the two liners (Fig. 9.1). This layer is
also called a leak detection layer or witness drain. It consists of either
gravel or a geonet when the primary liner is a geomembrane by itself.
When the primary liner is a composite liner (geomembrane with clay
beneath), the leak detection drainage material must be protected by a
filter/separator consisting of sand (for gravel) or a geotextile (for
a geonet). The leachate passing through the primary liner, i.e., the
'leaking liquid', flows gravitationally downward to its own sump area
where it is collected and removed. Removal is generally by small
diameter pipes since, hopefully, the quantities involved are very
low, or even nonexistent. Taken together this drain, filter, sump and
removal piping is called the secondary leachate collection and removal
system, or SLCRS.
Above the waste in the closure system, concern must be focused on
a surface water collection system (SWCS) above the cap geomembrane
as shown in Fig. 9.1. The drainage material is either gravel or a geo-
composite sheet drain. Whatever the choice, the material must be
protected by an appropriate filter since cover soil is placed above it. If
the drainage media is gravel the filter will be sand or a geotextile, while

Fig. 9.1 Cross-section of solid waste facility illustrating various drainage

layers involved in a liquid management program.
 Hydraulic concepts, definitions and test methods 189

if it is a geocomposite the filter will be a geotextile. The surface water

which reaches this drainage layer flows gravitationally to the perimeter
of the facility where it is removed by a perforated pipe or a geo-
composite edge drain. Taken together the drain, filter and removal
system is called the surface water collection and removal system, or
SWCRS.
Lastly, one must consider landfill gases which can rise within the
waste to the lower side of the cover system and will require a gas
drainage layer and venting system. This drainage layer is usually sand,
but can also be a thick geotextile. Gases in this drainage layer flow up-
gradient beneath the clay/geomembrane barrier and must be removed
by a penetration through the cap barrier system in the form of a vent,
flare, or gas collection and utilization piping system.
In the above description of the four aspects of liquid/gas collection
and removal systems, natural soils were always counterpointed with
their geosynthetic alternative. (The different geosynthetic alternatives
are shown in Fig. 9.2, i.e., geonets, geocomposite sheet drains and
geocomposite edge drains.) This is done in recognition that most
regulations allow for alternates to natural soils, if they can be shown to
be technically equivalent. Thus design of the geosynthetic alternative is
necessary, whereas the natural soil is often mandated by minimum
technology guidance (MTG). Some of the practical and economic trade-
offs in considering natural soil versus the geosynthetic alternative are
given in Table 9.1.
Regarding a cost comparison between natural soils and their much
thinner geosynthetic alternatives, it is important to note that with
vertical 'air' space in landfills at such a premium it certainly behooves
an owner/operator/designer to consider a geosynthetic replacement.
For example, at a $10/kN waste tipping fee (the 1992 price in northeast
USA), the savings of a 7 mm geonet drain over 30 cm of gravel in air
space alone is approximately $2000000 for a 20 ha site.

9.2 HYDRAULIC CONCEPTS, DEFINITIONS AND TEST

METHODS

The design of the various liquid management drainage and filtration

components just described is covered by hydraulic concepts included
in the areas of hydraulics and hydraulic systems. Many textbooks used
in geotechnical engineering cover the particulars for natural soils in
flow through porous media, but relatively few include information
on geosynthetics. In this latter regard see Fluet (1987) and Koerner
(1990a,b).
It is important to note that with some of the high flow geosynthetic
materials, like geonets, sheet drain geocomposites and edge drain
 (a)

(b)

(e)

Fig. 9.2 Types of geosynthetic drainage materials: (a) geonets; (b)

geocomposite sheet drains; and (c) geocomposite edge drains.
 Hydraulic concepts, definitions and test methods 191

Table 9.1 Selected comments on natural soils vis-a-vis geosynthetics used in

landfill drainage and filtration applications

type advantages disadvantages

Drainage media
pipe (perforated) common usage takes vertical space
traditional design creep deformation
rapid transmission needs filter
not likely clogged
soil (gravel) common usage takes vertical space
traditional design slow transmission
durable particulate clogging
biological clogging
needs filter soil
moves under load
geonet saves vertical space intrusion
rapid transmission creep deformation
not likely clogged needs geotextile filter
does not move new technology
geocomposite saves vertical space intrusion
(sheet drain or rapid transmission creep deformation
edge drain) not likely clogged needs geotextile filter
does not move very new technology
Filtration media
soil (sand) common usage takes vertical space
traditional design particulate clogging
durable biological clogging
moves under load
geotextile saves vertical space particulate clogging
easy placement biological clogging
does not move installation survivability
new technology

geocomposites (Fig. 9.2), laminar flow conditions do not exist. This

being the case one cannot work with Darcy's formula in the conven-
tional sense of considering a permeability coefficient, or hydraulic
conductivity, value. In these cases where turbulent flow exists, flow
rate at a given hydraulic gradient will have to be explicitly stated.
For the various drainage designs, flow is assumed to be within the
plane of the material as shown in Fig. 9.3(a). If laminar conditions exist
as in a soil or geotextile;
q = kiA = ki(W x t) = (kt)iW = (O)iW
(9.1)
Q' = qlW = Oi
192 Collection and removal systems

(a)

q/A

(b)

Fig. 9.3 Various hydraulic concepts illustrating (a) in-plane drainage

conditions and (b) cross-plane filtration conditions.

where q = flow rate, Q' = flow rate per unit width, k = hydraulic
conductivity (or 'permeability coefficient'), i = hydraulic gradient, W,
L, t are various geometric properties, and a = (kt) = transmissivity.
Note that the flow rate per unit width, Q', is equal to the trans-
missivity, a, if and only if, the hydraulic gradient is unity. If the flow
is turbulent as in a geonet or drainage geocomposite, however, it is
recommended to work directly with the flow rate per unit width, Q', at
a given value of hydraulic gradient (and at a given value of applied
normal stress).
For the filtration portion of the various designs~ flow is assumed to
be across the plane of the material as shown in Fig. 9.3(b). Since
laminar conditions usually exist in filter media, e.g., in sands or
geotextiles, Darcy's formula can be used directly.
 Hydraulic concepts, definitions and test methods 193

· = k -~h A = -k
q = kzA ~hA
t t

k _q_
~hA

Q" = \jI(~h) (9.2)

where Q" = flow rate per unit area, and 'I' = kit = permittivity.
If there is concern regarding laminar flow conditions and the use of
permittivity, '1', or if bulk flow is being considered, one can alterna-
tively work with the term, Q", which is flow rate per unit area, or
'flux'.
Using these hydraulic definitions one can design against a factor-of-
safety concept in the traditional manner:
for drainage problems
FS = Q' allowlQ' reqd (9.3)
or
FS = 8allow/8reqd

where FS = factor-of-safety, Q' = flow rate per unit width, and 8

transmissivity.
for filtration problems
FS = Q"allowIQ"reqd (9.4)
or
FS = \jIallowl\jlreqd
where FS = factor-of-safety, Q" = flow rate
per unit area (or flux), and
'I' = permittivity.
In the above relationships the values for allowable properties are
obtained from an appropriate test method. Clearly, ASTM is the lead-
ing standardization group in this regard. Committee 0-18 has stan-
dards for soil properties and Committee 0-35 has standards for
geosynthetic properties. The most important test methods for, use in
the type of problems to be presented in this chapter are the following:
• 02434 Permeability of soil
• 02412 Strength of plastic pipe
• 04716 Transmissivity of geosynthetics
• 04491 Permittivity of geotextiles
• 04751 Apparent opening size of geotextiles
These tests methods (and others) will be referenced further in their
appropriate design sections.
194 Collection and removal systems

9.3 PRIMARY LEACHATE COLLECTION AND REMOVAL

SYSTEM

The primary leachate collection and removal system (PLCRS) is located

above the primary geomembrane as shown in Fig. 9.1. It consists of
a drainage layer which is usually gravel, perforated pipe collector
system, sump collection area, removal system (manhole or large
diameter pipes) and a protective filter layer. Each topic will be de-
scribed in this section along with selected comments on long-term filter
performance.

9.3.1 Drainage considerations

Immediately after a landfill cell is constructed, the PLCRS serves as
a construction dewatering system. After waste begins to be placed,
attenuation of much of the precipitation occurs and the flow rate
significantly decreases. The 1985 EPA minimum technology guidance
(MTG) for hazardous waste landfills calls for the following properties
of the PLCRS.
• 30 cm thick layer
• hydraulic conductivity of 0.01 cm/sec or higher, (note that new
regulations may increase this value to 1 cm/sec, or higher)
• slopes of greater than, or equal to 2%
• must include a perforated pipe removal system
• must include filter soil above the drainage soil
• must cover both bottom and sidewalls of the landfill facility
In all primary leachate collection systems known to the author, natural
soil is used as per the above stated MTG or state mandated require-
ment, whichever takes precedence. Problems are encountered, how-
ever, in holding these sands and gravels on side slopes, particularly
when using low friction geomembranes as the primary liner. Many
problems of surface sliding have occurred. Options in this regard are
the following:
• use a sufficiently low side slope so as to assure drainage soil stability;
• use high friction or roughened geomembrane sheets to increase the
frictional stability;
• reinforce the PLCS with a geogrid or geotextile in the form of veneer
stability;
• construct the PLCS as the waste is being placed, i.e., one lift at a
time; and
• use a geocomposite sheet drain or geonet on the side slopes for the
PLCS.
 Primary leachate collection and removal system 195

Of these options, the roughened geomembranes and/or veneer stability

techniques are being used most frequently, see Koerner and Hwu
(1990) for design details. The use of a geosynthetic drainage system
(geonet or sheet drain) will be discussed later in the SLCRS (leak
detection) section.

9.3.2 Collection considerations

If one considers using the above minimum values for the drainage
stone design, the mound equation, after Moore (1983), can be used to
obtain the perforated collector pipe spacing as per Fig. 9.4.

.J....J""'"'b...h...X:J~~!oo,.;b....t. ....If<...4-~(geotextile
or fisand)
lter
"

_ ...J:)~::_ ~~~hmra:. ___-:-::-:--~ geomembrane

primary

~ S '" collector pipe spacing

~
(a)

inflow

~~IOOO\1~~~IOOO\1~~~Mm~~M~dt<i~d~~~.....-filter
(geotextile
or sand)

I'" S '" collector pipe spacing

(b)

Fig. 9.4 Illustrations of leachate 'mounds' occurring between collection pipes:

(a) uniform slope between pipes; (b) crested slope between pipes.
196 Collection and removal systems

he sVc
=- - [tan2
---a + 1 tan-
- - a v'tan 2 a + cJ (9.5)
2 c c
where he = maximum height of leachate between adjacent collector
pipes, 5 = spacing between adjacent pipes, a = slope of ground
surface between adjacent pipes, c = q/k (dimensionless), q = infiltration
intensity, and k = hydraulic conductivity (permeability) of collection
stone.
In this formulation, the value of q is most significant. Immediately
after construction of the primary liner system, the PLCR5 must act as a
construction dewatering system. Here the infiltration intensity, or flow
rate, will be at a maximum and site specific weather data must be used.
Once waste is placed, precipitation will be greatly attenuated and a
water balance method can be used; see Chapter 10. Flow rates are
proportionately reduced as additional waste is placed in the facility.
Next, the size (diameter) of the piping must be determined. This is
usually done using Manning's equation (Gupta, 1989), which, in 51
units, is as follows:

v = ! RHO.66S050
n (9.6)
Q= VA
where V = velocity of flow, A = cross-section area, RH = hydraulic
radius (= flow area/wetted perimeter) = 0/4, for pipes flowing full,
o = inside diameter of the pipe, 5 = slope of the pipe, and n =
Manning's roughness coefficient (dimensionless).
Table 9.2 gives typical values of roughness coefficient n after Fox and
McDonald (1985). With the current tendency toward use of plastic pipe
in landfills, it should be mentioned that smooth interior PVC or HOPE
pipe has a Manning coefficient varying from 0.009 to 0.010, while a

Table 9.2 Values of Manning roughness coefficient, n, for representative

pipe interior surfaces, after Fox and McDonald (1985)

type of pipe surface representative

n value

Lucite, glass, or plastic 0.010

Wood or finished concrete 0.013
Unfinished concrete, well-laid brickwork, concrete or cast iron
pipe 0.015
Riveted or spiral steel pipe 0.017
Smooth, uniform earth channel 0.022
Corrugated flumes, typical canals, river free from large stones
and heavy weeds 0.025
Canals and rivers with many stones and weeds 0.035
 Primary leachate collection and removal system 197

profiled pipe interior has a Manning coefficient ranging from 0.018 to

0.025. The Manning formula is greatly aided by use of a nomograph
which is available in many hydraulic's textbooks (Fig. 9.5).
The load carrying capability of the pipe network must obviously be
assured. Here the modified Iowa formula is generally used (Spangler,
1971; Moser, 1990).

(9.7)

where ~x = horizontal deflection or change in diameter of the pipe,

DL = deflection lag factor (varies from 1.0 to 1.5), K = bedding con-
stant (varies from 0.83 to 0.110), We = prism load per unit length of
pipe, E = modulus of elasticity of the pipe material, I = moment of
inertia of the pipe wall per unit length, r = mean radius of the pipe,
and E' = modulus of soil reaction.
In this equation the value of We is calculated as the weight of waste
above the pipe and E' is usually taken from the US Bureau of Reclama-
tion studies (Howard, 1977). The horizontal deflection is assumed to be
equal to the vertical shortening (as per ASTM 0-2412) and should be

0.2

0.5
0.3 0.5
0.2
0.3
0.1

O@~
~

0.05 0.5 OJ
:;
0.1 g 0.04
E "
..c::
gp §
:g 0.02
~
«)
0.05 ~ 0.01
0.02 0.3 0
~

5
"...
0.01 , 0.007 0.2 bl)

:g
~
<: en
·2
:a"
<:: 0.005
0.005 " 0.03 <::

0
oj
~
..c:: g 0.1
OJ

6 .g0
~ ...
"
>- 0.002
...
bi)
is 0.01 ~
·0 <l
"<:: 0.001 6 E 0.05
oj
0- >
:>
~ bl)
rJ"J 0.0005 0.005 is 0.03
<:
.§
6
0.0002 0.02 '"
E-

0.0001
0.001 0.01
0.00005
0.0005
0.00002
0.0003
0.00001

Fig. 9.5 Solution chart for the Manning formula

1
V = -RjlrS0 50, after Hwang (1981).
n
198 Collection and removal systems

limited to approximately 5% of the pipe diameter. Note that plastic

pipe is governed by a limiting deflection rather than an ultimate crush
strength as is rigid pipe.

9.3.3 Removal considerations

There are many possible configurations for the collection and removal
network on the bottom of the cell (Fig. 9.6). All of them, however,

['\.. • /

...
3:1
~2% ~2% ~l
h !::>-....<~

............
1/
" /"

~2% 2%~
-+
3:1

W~"'Y~~~
/" ............

Legend
@ = sump/manhole
0= manhole
• = clean out riser

Fig. 9.6 Plan views of several configurations for leachate removal systems,
after Jordan (1984).
 Primary leachate collection and removal system 199

gravitationally drain to a sump area where a leachate removal facility is

located. Here the variations are enormous with some options being as
follows:
• manhole extending vertically through the waste (and final cover)
for leachate removal;
• pipe risers extending up the sidewall slope located above the
primary geomembrane; and
• sloped pipe extending through the liner system to a location beyond
the landfill cell area.
Of these options, the current tendency is toward side slopes risers to
avoid the downdrag problems associated with vertical manholes and
the liner penetrations associated with continuous piping systems
(Richardson and Koerner, 1988). Except in very arid areas with dry
waste materials, most owner/operators of landfills use submersible
pumps in the sump area to pump the collected leachate to trucks,
storage tanks, reservoirs or directly to treatment facilities.

9.3.4 Filtration considerations

The granular filter above the leachate collection drainage stone is
required by MTG to be at least 15 cm (6.0 in) of soil. This soil must be
appropriately designed considering the overlying waste and under-
lying drainage soil. In this regard, conventional soil filter design is
used such as the following.
dIS (of filter) < 4 to 5 < --=-=--'------'-
--"-"-'------'- dIS (of filter) (9.8)
dRS (of soil) dIS (of soil)
As stated by Cedergren (1989), the left half of the inequality is
rationalized as follows.
The 15% size (dIS) of the filter material must be not more than four
or five times the 85% size (d 8s ) of the protected soil. The ratio of
dIS of a filter to d8s of a soil is called the piping ratio.
The right half of the inequality is rationalized as follows.
The 15% size (dIS) of a filter material should be at least four or five
times the 15% size (dIS) of the protected soil. The intent is to
guarantee sufficient permeability to prevent the buildup of large
seepage forces and hydrostatic pressures.
The above design conditions are usually performed on a graph of
the various particle size distributions of the soils involved and are
explained in all geotechnical engineering textbooks.
When considering placement of a soil filter around slotted or
200 Collection and removal systems

perforated pipe, the US Army, et al. (1971) criteria is generally used.

These criteria are as follows.
for slotted pipe

dS5 (of filter)

-=--'----"- > 1. 2 (9.9a)
slot width
for perforated pipe
dS5 (of filter)
> 1. 0
-:---"-=--":"'---'- (9.9b)
hole diameter
The option of using a geotextile filter above the drainage media or
around a slotted or perforated pipe is very compelling since the savings
in air space represents additional revenue to the owner of the facility
and in many situations is cost effective in its own right. The design of a
geotextile filter must be assured in the same manner as with a soil
filter, i.e., adequate soil retention and sufficient permeability. The
approach uses factor-of-safety concepts as follows; see Koerner (1990a)
for additional details.
for soil retention

0 95 < (2 to 3) dS5 (9.10)

for adequate permeability

FS = 'l'allow (9.11)
'l'reqd

where 0 95 = 95% opening size of the geotextile (as per ASTM 04751),
dS5 = 85% size of the protected soil (or waste), 'l'allow = permittivity of
the geotextile (as per ASTM 04491), and 'l'allow = required permittivity
as per the in flowing liquid (estimated precipitation or via a water
balance procedure; see Chapter 10).

9.3.5 Filter clogging

The long-term flow performance of the filter above the PLCRS (either
natural soil or geotextile) must be assured for the length of the landfill
filling operation and for the post-closure care period which might
include leachate recycling procedures. This is typically for 30 to 40
years. Thus a question must be raised as to potential clogging of the
filter by the precipitates in the leachate, by biological micro-organisms,
or by a synergistic combination of both. Considering the characteristics
of leachate from domestic waste landfills (not necessarily from hazard-
ous waste or ash monofils), the tendency for such clogging is certainly
possible.
 Primary leachate collection and removal system 201

In a recent study by Koerner and Koerner (1990), six domestic

landfill leachates have been evaluated for their tendency to clog both
geotextile and natural soil filters. Using ninety-six, 100 mm diameter
test columns, the effects of aerobic and anaerobic conditions, four
different geotextiles, with soil (a #40 mesh Ottawa sand) and without
soil, were evaluated. Flow tests of a falling head variety were con-
ducted and within six months the following trends were observed.
no clogging (0%-25% flow reduction)6 of 96 columns
= 7%
minor clogging (25% -50% flow reduction) 4 of 96 columns
= 4%
moderate clogging (50%-75% flow reduction) 37 of 96 columns
= 38%
major clogging (75% -95% flow reduction) 35 of 96 columns
= 36%
severe clogging (95%-100% flow reduction) 14 of 96 columns
= 15%

Clearly the situation of filter clogging must be addressed and should

be done in the design stage. A number of possibilities exist. One
option that a designer has is to provide access to the removal piping
system as per cleanout risers of the type shown in the sketches of
Fig. 9.6. With such cleanout capability one might consider a leachate
backflush, a water backflush, a nitrogen gas backflush or a vacuum
extraction. (Trials are currently ongoing for each of these procedures
where the water backflush appears to be the most promising.) In this
case, a very open geotextile could be used. Another option might be
not to use a filter at all. While this will generally require approval of
the permitting agency, it might be a viable option when using high
permeability drainage stone, e.g., hydraulic conductivity greater than
1.0 cm/sec. With adequate gradient, the fine particulates and the
accompanying micro-organisms should be carried along with the
leachate to the sump and be pumped out of the withdrawal manhole.
The last option might be to not worry about filter clogging at all. If the
drainage material beneath the filter is not saturated (which is probably
the case), the hydraulic head of leachate buildup is on the filter and is
not on the geomembrane. Thus the MTG restriction of less than 30cm
head on the liner is not violated. In such a case the leachate mound
will then build up in the waste until it becomes sufficiently large to
cause breakthrough of the clogged filter. At breakthrough the leachate
will flow rapidly into the underlying PLCRS until further clogging
occurs in the filter and the process repeats itself.
Whatever the option taken in regard to potential filter clogging it
should be done in an open and candid manner.
202 Collection and removal systems

9.4 SECONDARY LEACHATE COLLECTION AND REMOVAL

SYSTEM (I.E., LEAK DETECTION)

The secondary leachate collection and removal system (SLCRS) is

located between the primary and secondary liner systems and serves as
a leak detection layer, or witness drain. It must gravitationally flow to a
sump area located at the low point of the landfill cell where it is
periodically sampled so as to assess the adequacy of the primary liner
against leakage of leachate.

9.4.1 Drainage considerations

The 1985 EPA minimum technical guidance (MTG) for hazardous waste
landfills calls for the following:
• 30 em thick layer
• hydraulic conductivity of 0.01 em/sec, or higher
• must cover both bottom and side walls
• must have a response time from liner leak to detection of less than
24 hours
This layer, when constructed of natural soils (as it certainly can be),
presents some very real challenges. Side slope stability and pipe
installation, along with placement of the primary liner system above it,
are all difficult to construct. Furthermore, the response time criterion
might be difficult to meet; see Richardson and Koerner (1988) for such
calculations.
When considering that the amount of leachate the SLCRS must
handle is only a small fraction of PLCRS, one should seriously consider
use of a geonet (Fig. 9.2a). The savings in air space by itself is consider-
able (recall the calculation in section 9.1) plus the above difficulties are
avoided, e.g., no pipe system need be used with a geonet. Thus this
section will consider the SLCS to be a geonet, recognizing that natural
soil can be used if so desired.
The design of the geonet can be based on transmissivity, but it
is straightforward (and technically more correct) to proceed with
flow rate per unit width at the project specific hydraulic gradient and
applied normal stress, i.e.,
FS = Q'allow
Q'reqd

where Q' allow = allowable flow rate per unit width of the candidate
geonet as per ASTM D4716, and Q'reqd = required flow rate per unit
width coming from the leaking primary liner.
Regarding the ASTM test for Q' allow, a cross-section and typical data
set is given in Fig. 9.7. Note, however, that this type of data is usually
 Secondary leachate collection and removal system 203
Constant Load

Outlet Reservoir

Overflow
Q
Overflow

5
E,!.
M<E 4 ., i=l.OO
g

--
B 3 • i=O.50

" • i=O.25

---
0::
•
"
i=O.J3
c 2
• • •
:- ; :
i:I: i=O.06
•
:
c i=O.03
• .....
~ II
0
0 200 400 600 800 1000
Nonnal Stress (kPa)

Fig. 9.7 Schematic diagram of lateral flow rest device following ASTM D4716
and typical geonet flow rate data between rigid platens.

presented for the geonet placed between two rigid geomembranes.

When the primary liner is a composite liner this is certainly not the
field simulated situation. In this case, a geotextile (functioning as a
filter/separator) must be placed over the geonet and under the clay
component of the overlying composite liner. The stress on the clay will
cause an intrusion of the geotextile into the apertures of the geonet
reducing a portion of its planar flow capability. Hwu et al. (1990) have
given insight into the amount of flow reduction for a wide range of
conditions, soils and geotextiles. If the reduction in flow rate becomes
excessive (for example, from very large landfill heights), it may be
necessary to consider a double geonet (Koerner and Hwu, 1989).
Regarding the quantity Q'reqd, one must estimate the flow rate of
leachate coming through the primary liner. At the minimum one
must consider the vapor diffusion rates for the 'ideal' liner system.
This amount varies according to type and thickness of primary
204 Collection and removal systems

geomembrane, but is of the order of 101Iha-day. Beyond this one must

estimate the number and size of holes that may occur in the primary
geomembrane, and whether or not it is a single geomembrane or a
composite geomembrane/clay primary liner. Guidance in this regard
can be found in Giroud and Bonaparte (1989a,b).

9.4.2 Collection considerations

When using a geonet as a SLCRS there is no network of perforated
removal pipes that are necessary. The response time from liner leak to
sump detection is surely less than 24 hours since flow in the geonet's
apertures is essentially uninhibited.

9.4.3 Removal considerations

Flow within the SLCS is gravitational and usually follows the same
grading scheme as the overlying PLCS, recall the patterns of Fig. 9.6.
Since both systems are draining to the same down gradient location in
the landfill cell, the removal sumps are usually immediately adjacent to
one another. What differs, however, is the amount of leachate in the
SLCRS versus the PLCRS. Instead of requiring submersible pumps,
one should have need for no more than well samplers and, at that,
using small diameter riser pipes. Typically 100-150 mm smooth, solid-
wall plastic pipes coming up the side slope between the primary and
secondary liners are adequate. Sampling frequency is intermittent and
done in accordance with the permit for the facility. It is important to
recognize that if liquid is found in the SLCRS it might not be from a
leaking primary liner. Gross et al. (1990) gives insight as to other
possible liquids, for example:
• compression water if a moist soil drainage layer is used;
• consolidation water from the saturated clay component if a com-
posite primary liner is used;
• infiltration water from the secondary (bottom) liner in areas of high
water table; or
• penetration water from the termination area around the periphery
of the site, i.e., between the anchorages of the primary and secondary
liner systems.
Thus one should always chemically analyze the liquid in the SLCRS via
mass spectroscopy/gas chromatography (MS/GC) methods before
assuming that the primary liner is leaking.
In closing, it should be mentioned that response action plans (RAPs)
will soon be mandated for all landfill facilities and will require feedback
of leakage rates found in the SLCRS. At that time, design values
should become more realistic than they are currently.
 Surface water collection and removal system 205

9.5 SURFACE WATER COLLECTION AND REMOVAL SYSTEM

Above the completed landfill facility will be either a temporary or a

permanent liner system (see Chapter 18), and above this liner will
generally be a surface water collection and removal system (SWCRS).
This might not be the case in extremely arid regions of the southwest
USA where the nominal amount of moisture that does occur should be
preserved, rather than drained away from the site. In general, how-
ever, the water that does penetrate the cover soil should be efficiently
and expeditiously removed off of the liner and away from the site;
recall Fig. 9.1.

9.5.1 Drainage considerations

The 1989 EPA minimum technology guidance (MTG) for final covers on
hazardous waste landfill calls for the SWCRS to include the following.
• 30 cm thick layer
• hydraulic conductivity of 0.01 cm/sec, or higher
(or geosynthetic material with equivalent performance)
• minimum final slope of 3%
Clearly, natural soils can be, and have been, used for surface water
collection systems. Savings of air space is not a consideration since
minimum cover thickness over the geomembrane/clay liner can easily
accommodate the 30 cm thickness requirement. Indeed most designs
respond to the above stated MTG using granular soil drainage layers,
taking into consideration the conditions of stability against the cover
soil sliding off of the geomembrane. On steep side slopes of
aboveground landfills this consideration may be troublesome requiring
high-friction geomembranes or some type of geosynthetic reinforce-
ment (Koerner and Hwu, 1990).
The alternate geosynthetic material referenced in the MTG makes
reference to either geonets or geocomposite sheet drains as shown in
Fig. 9.2. Since geonets were just described, the following comments
will be focused on geocomposite sheet drains. There are many com-
mercially available products in this regard; see Table 9.3. All are
designed as were the geonets described earlier, i.e.,

FS = Q'allow
Q'reqd

The allowable flow rate is obtained from an ASTM 04716 test, recall
Fig. 9.7, and compensated for the intrusion of the geotextile filter/
separator. The required flow rate is obtained from a suitable water
balance method considering site-specific precipitation, runoff, evapo-
transpiration and infiltration. The infiltration is, of course, the liquid
Table 9.3 Summary of geocomposite sheet drainage systems

manufacturer/agent product name core structure core polymer geotextile* roll size (ft) thickness
widthllength (mils)
--- ._--------- --- -------- ----
Akzo Enkamat 7010 3-D web nylon 6 3.2/492 400
Enkamat 7020 3-D web nylon 6 3.2/330 800
Enkamat 9010 3-D web nylon 6 polyester 3.0/99 400
Enkamat 9120 3-D web nylon 6 polyester 3.0/99 800
American Wick Amerdrain 480 mat Nippled core polyethylene polypropylene 41104 375
Drain Corp.
Burcan Industries Hitek 6c cuspated polyethylene polypropylene 3.0/450 255
Hitek 20c cuspated polyethylene polypropylene 3.6/125 785
Hitek 40c cuspated polyethylene polypropylene 3.5/80 1575
Exxon Tiger Drain cuspated polyethylene polypropylene 4/100 600
Exxon Battle Drain cuspated composite polypropylene 41100 600
Greenstreak Sheet Drain raised core polystyr~;te nonwoven-PP 4/ X 10/ panels 375
II
Deck Drain woven-PP 375
Geotech Systems Geotech Drain Board EPS panel 4.0/4.0 1000
Huesker Synthetic Ha Te-Drainma tte polypropylene PES 13/328 260
JDR Enterprises J-Drain 100 extruded rib polyethylene polypropylene 250
J-Drain 200 extruded rib polyethylene polypropylene 250
Mirafi Miradrain 6000 raised core polystyrene polypropylene 4/8,25,50 380
Miradrain 6200 raised core polystyrene polypropylene 4/8,25,50 380
+ backup film
Miradrain 9000 raised core polystyrene polypropylene 4/8, 25, 50 380
Miradrain 8000 raised core polyvinyl chloride polypropylene 4/8,25, 50 380
Miradrain 5000 raised core perforated polystyrene polypropylene 4/8,25,50 380
Miradrain 4000 cuspated polystyrene polypropylene 4/8,25, 50 750
Table 9.3 Continued

manufacturer/agent product name core structure core polymer geotextile* roll size iff) thickness
width/length (mils)

Monsanto Hydraway Drain raised cyl. tubes polyethylene polypropylene 12, 18/400 1000
Nilex Nudrain A polyethylene polypropylene 1.6/49,98 1575
Nudrain C polyethylene polypropylene 3.6/98 787
NW Fabrics Permadrain cuspated polyethylene polypropylene any size
Pro Drain Systems PDS20 cuspated polyethylene polypropylene 3.7/10-500 750
PDS40 cuspated p~lyethylene polypropylene 3.3110-250 1500
Rhone-Poulenc Megadrain 3-D web nylon 6 polyester 375

• All geotextile filters are nonwoven; either needle-punched or heat-bonded.

208 Collection and removal systems

that must be carried in the drainage network. The HELP computer

code is well suited for this determination; see Chapter 10.

9.5.2 Collection considerations

As shown in Fig. 9.1, the surface water drainage system gathers
and transports the infiltrating water to a collection system around the
periphery of the cell or landfill. This collector system is usually a
perforated pipe and is designed similar to a highway underdrain
system (Cedergren, 1989). However, as with highway underdrains,
geocomposite edge drains are seeing use due in large part to their low
cost and ease of placement (Fig. 9.2c). Table 9.4 gives an overview of
the available products. In the design of the collector system it must be
remembered that the required flow rate is cumulative over the length
of the landfill cover and over the length of pipe or edge drain to the
outlet.

9.5.3 Filter considerations

The 1989 cover MTG calls for a 150 mm thick filter soil over the
drainage soil. This soil is designed as with any other soil filter as was
described in Section 9.3. If a geocomposite sheet drain is used, the
filter will be a geotextile which is also designed as was described in
Section 9.3.
It should be mentioned that the design of the surface water drain
and its filter is much more straightforward than with the PLCRS drain
since water is the permeant and not leachate. Furthermore, the normal
stresses in the landfill cover are much less than beneath the waste so
that crushing of the core and intrusion of the geotextile are easier to
meet with a wide variety of geosynthetic products.

9.6 METHANE GAS COLLECTION AND REMOVAL SYSTEM

When dealing with domestic waste landfills, the anaerobic decomposi-

tion of the waste will generate numerous gases. The two principal
types are carbon dioxide (C0 2 ) and methane (CH 4). Carbon dioxide is
heavier than air and will migrate to the bottom of the waste facility and
will be removed with the leachate. Methane, however, is lighter than
air and will rise to the top of the waste (or to the sides) where it will
meet the cover liner system and be trapped beneath it (Fig. 9.1). Thus,
it is necessary to have a gas collector system beneath the liner and
have it uniformly graded to the high point of the cover. Here it is
either vented, flared, or captured and used for energy production.
Additionally, certain chemicals within the landfill may also emit gases
Table 9.4 Properties of commercially available highway edge drains as listed in manufacturers literature

company trademark core characteristics geotextile type

material shape thickness strength* flow rate*

Monsanto Hydraway 2000 PE straight column 1.00 13700 27gpm/ft. @21Opsfi = 0.2 PET nonwoven
27 gpm/ft. @ 2090psf i = 0.2 needle-punched
Con tech Strip drain 75 PE tapered column 0.75 5800 26 gpm/ft. @ 500 psf i = 1.0 PET nonwoven
22gpm/ft. @ 1500psf i = 1.0 needle-punched
Stripdrain 100 PE tapered column 1.00 8000 15 gpm/ft. @ 1440 psf i = 0.1 PET nonwoven
3.2 gpm/ft. @ 1500 psf i = 0.01 needle-punched
4.7gpmlft. @ 1500psfi = 0.02
Pro Drain Systems PDS-20 PE tapered column n/a n/a n/a n/a
PDS-30 PE tapered column 1.20 n/a n/a PP nonwoven
needle-punched
ICE Akwadrain B PE double cuspated 1.00 8000 12 gpm/ft. @ 1000 psf i = 1.0 PP nonwoven
12 gpm/ft. @ 2000 psf i = 1.0 needle-punched
Akwadrain 125 PE double cuspated 1.25 n/a n/a PP nonwoven
needle-punched
Adv. Drainage Advan Edge PE corrugated 1.60 4320 7 gpm/ft. @ 10 psi i = 0.1 PP nonwoven
Systems (ADS) with columns melt bonded
Nilex Nudrain™ 'A' PE double cuspated 1.60 2700 5.9 I1secl250 mm width PP nonwoven
melt-bonded
Greenstreak Pave Drain™ PE tapered column 1.10 8500 8 gpm/ft. @ i = 0.05 PP nonwoven
(tapered) 38 gpm/ft. @ i = 0.5 needle-punched
Drysol Drainage Cell PP cellular 1.20 n/a n/a PP nonwoven
needle-punched

*The values of flow rate are assumed to be at a hydraulic gradient of 1·0, in which case it is numerically equal to transmissivity. The values,
however, are taken directly from manufacturers' literature where considerable variation in test method, manner of presentation of results, and
concepts involved, tend to vary.
n/a = not available
210 Collection and removal systems
or vapors that require venting. The collection layer itself is not covered
in detail by MTG and is clearly site specific and waste specific. Design
proceeds as with any drainage system in that a factor-of-safety is
formulated as follows:

E
6 -- .. .. .. .. ..
... • •
12 oil
0
5
. ..
(\)
.!!? EI Ua=24.1 kPa
~
•
~
.2:
4
•• • • • •
• Ua=13.8 kPa
• •
....
Ul
Ul
'EUl
3 a Ua=6.9 kPa
c:: • • • Ua=3.4 kPa
~
2 • • 0
...
'iij 1• • Ua=O.7 kPa

0 •• • • • • • •
0 50 100 150
stress (kPa)

(a)

250

200
~
III Ua=24.1 kPa
~

~== •
~ 150
Ua=13.8 kPa

: ::
:is
~

~:
E 100 a Ua=6.9 kPa
(\)
c.
.iij 0 Ua=3.4 kPa
50 ~---~.----.------

0
- • • • • • • • Ua=O.7 kPa

0 50 100 150

stress (kPa)

(b)

Fig. 9.8 (a) Air transmissivity and (b) in-plane coefficient of air permeability,
versus applied normal stress for one layer of 600 g/m2 needle-punched
nonwoven geotextile, after Koerner and Bove (1984).
 Methane gas collection and removal system 211

FS = Q'allow
Q'reqd

The allowable flow rate, Q' allow, is not at issue when using natural soils
since gas flow-to-water flow is in rough proportion to the differences in
viscosity, i.e., air flows about 1000 times easier than does water. Most
soil layers for gas venting are 15-30 cm in thickness. Flow rate is of

(a)

lop soil and

~~i~~~iii~~cover soilwaler
surface
drain

. •.•.•.....•.•.•.•••...•••.•.•••••.•.•.••••••• . ••••••••..••••..•'._•... ]
.•.•••. :.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.: fold geolexille ..:.:.:.:.:.:.
~;;~::~:~i!~~Oil layer
gas drain layer .:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.: 3 or 4 layers thick':':':':':': was Ie
(geolextlle) :.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:•••.•.•...•.. :.:.:.:.:.:.:.:.

(b)

-----exifusionweld
or ;c;:2
~ stainless steel clamp

r
(e)
Fig. 9.9 Schemes for methane gas removal though final cover liner system:
(a) natural soil vent system, after EPA (1989); (b) geotextile vent system; (c)
suggested bellows at geomembrane penetration.
212 Collection and removal systems

issue, however, if one is considering the use of a geotextile gas drain.

Since this layer represents a savings of air space (it is within the landfill
enclosure), it should be considered and properly investigated. Figure
9.8 gives air transmissivity and air permeability in the plane of a
600 g/m2 needle punched nonwoven geotextile. Here the data is seen to
be sensitive to the incoming gas flow rate and the applied normal
stress.
Concerning the required flow rate, Q'reqd, Emcon (1980) has deter-
mined gas recovery rates from actual field measurements at six landfills.
Using data from recovery projects or gas control programs, annual gas
recovery rates from 0.27 to 0.77 cubic meters per kilometer of waste
have been estimated.
Of particular note is the extremely long time frames for decomposition
of the waste (along with the accompanying methane gas generation) to
occur. Baron et al. (1981) estimate the following.
• 50% decomposition (i.e., half-life) = 23 yrs
• 90.5% decomposition = 80yrs
• 99.1% decomposition = 160yrs
Thus the detail of methane gas collection and removal is important to
the proper long-term functioning of the closure system.
Regarding removal of the collected gas at the high point of the
closure (or along the flow path at regular intervals), the cover barrier
system must be penetrated. Several designs are possible, but dif-
ferential settlement of the underlying waste must be considered. Figure
9.9 gives some ideas as to different removal schemes with a granular
venting layer or with a geotextile. The idea of a bellows arrangement at
the geomembrane penetration is attractive. Prefabricated geomembrane
boot attachments are recommended in this regard.

REFERENCES

Baron, J.L. et al. (1981) Landfill methane utilization technology workbook, US DOE,
Contract No. 31-109-38-5686, Washington, DC.
Cedergren, H.R. (1989) Seepage, Drainage and Flow Nets, J. Wiley and Sons, NY,
NY.
Emcon Associates (1980) State-of-the-art of methane gas enhancement in landfills,
Project 343.1.1, San Jose, CA.
Fluet, J.E. (1987) Geotextile testing and the design engineer, ASTM STP 952,
Philadelphia, PA.
Fox, R.W. and McDonald, A.T. (1985) Introduction to Fluid Mechanics, 3rd edn,
J. Wiley and Sons, NY, NY.
Giroud, J.-P. and Bonaparte, R. (1989a) Leakage through liners constructed
with geomembranes - Part I: geomembrane liners, J. Geotextiles and
Geomembranes, Elsevier App!. Sci. Pub!., London, 8(1), 27-68.
Giroud, J.-P. and Bonaparte, R. (1989b) Leakage through liners constructed
 References 213
with geomembranes - Part II: composite liners, ]. Geotextiles and
Geomembranes, Elsevier App!. Sci. Pub!., Long, 8(2), 7l-112.
Gross, B.A., Bonaparte, K and Giroud, I.-P. (1990) Evaluation of flow from
landfill leachate detection layers, Proc. 4th Inti. Conf. on Geosynthetics, A.A.
Balkema Pub!., Rotterdam, The Netherlands, 481-6.
Gupta, R.A. (1989) Hydrology and Hydraulic Systems, Prentice-Hall, Inc.,
Englewood Cliffs, NJ.
Howard, A.K. (1977) Soil reaction for buried flexible pipe, J. Geotech. Eng. Div.,
ASCE, 33-43.
Hwang, H.C. (1981) Fundamentals of Hydraulic Engineering Systems, Prentice-
Hall, Inc., Englewood Cliffs, NJ.
Hwu, B-L., Koerner, KM. and Sprague, c.J. (1990) Geotextile intrusion into
geonets, Proc. 4th IntI. Geosynthetic Conf., The Hague, Netherlands, 351-6.
Jordan, E.c. Co. (1984) Performance standard for evaluating leak detection, US EPA
Contract No. 68-01-6871, Washington, DC.
Koerner, G.K and Koerner, R.M. (1990) Biological activity and potential
remediation involving geotextile landfill leachate filters, ASTM STP 1081 (ed.
KM. Koerner), Philadelphia, PA.
Koerner, R.M. (1990a) Designing with Geosynthetics, Prentice-Hall Pub!., Co.,
Englewood Cliffs, NJ.
Koerner, KM. (1990b) Geosynthetic testing for waste containment applications,
ASTM STP 1081, Philadelphia, PA.
Koerner, R.M., Bove, J.A. and Martin, J.P. (1984) Water and air transmissivity
of geotextiles, J. Geotextiles and Geomembranes, Elsevier App!. Sci. Pub!.,
3(1),57-75.
Koerner, R.M. and Hwu, B-L. (1989) Behavior of double geonet drainage
systems, Geotech. Fabrics Rep., Sept./Oct., 39-44.
Koerner, R.M. and Hwu, B-L. (1990) Stability and tension considerations
regarding cover soils on geomembrane lined slopes, IntI. J. Geotextiles and
Geomembranes, Elsevier App!. Sci. Pub!., London.
Moore, C. (1983) Landfill and surface impoundment performance evaluation, US
EPA, SW-869, Washington, DC.
Moser, A.P. (1990) Buried Pipe Design, McGraw-Hill, NY, NY.
Richardson, G.N. and Koerner, R.M. (1988) Geosynthetic design guidance for
hazardous waste landfill cells alld surface impoundments, US EPA Contract No.
68-03-3338, Cincinnati, OH.
Spangler, M.G. (197l) Soil Engineering, 2nd edn, International Textbook Co.,
Scranton, PA.
US Army, Navy and Air Force (1971) Dewatering and groundwater conrol in deep
excavations, TM5-818-5, NAVFAC P-418, AFM 88-5, Ch. 6, Washington,
DC.
US EPA (1985) Minimum technology guidance on double liner systems for landfills
and surface impoundments: design, construction and operations, EP A/530-SW -84-
014.
US EPA (1989) Final covers on hazardous waste landfills and surface impoundments,
EPA/530-SW-89-047, OSW, Washington, DC.
 CHAPTER 10

Water balance for landfills

R. Lee Peyton and Paul R. Schroeder

10.1 INTRODUCTION

A key step in the design of a solid waste disposal facility is the

execution of a 'water balance' or 'water budget' analysis. A water
balance is an accounting of the final disposition of precipitation falling
on a site. Water balance analysis can be used to estimate the potential
leachate production and liner/drain system performance and to com-
pare the relative effectiveness of alternative cover and liner/drain
designs. Knowledge of the possible range of leachate production is
important for sizing the leachate collection system (e.g., pipes) and in
making decisions about how to manage treatment of the leachate.
Similarly, prediction of liner leakage and the depth of leachate buildup
in a drain layer is important in the selection of liner and drain materials
and in the design of collection pipe spacing and liner slope.
A water balance analysis accounts for the effects of many hydrologic
processes on water movement at a site. Once precipitation strikes a
site, it may either:
• be intercepted by vegetation and subsequently evaporated;
• be stored temporarily on the surface in the form of snow or ice if the
temperature is freezing and later evaporated or melted;
• run off the surface of the cover system; or
• infiltrate into the cover.
Water infiltrating into the cover may:
• be transpired to the atmosphere via plants that draw the water from
the ground through their roots;
• evaporate from the surface soil into the atmosphere;
• be stored in the cover soil; or

Geotechnical Practice for Waste Disposal.

Edited by David E. Daniel.
Published in 1993 by Chapman & Hall, London. ISBN 0 412 35170 6
 Water balance calculations 215
• percolate downward through the cover soil and past the evaporative
zone.
Water percolating through past the evaporative zone may:
• be collected in the cover if a lateral drainage layer is present;
• be stored in the underlying waste or soil; or
• contribute to the leachate production which is either collected in
the leachate collection system or lost as leakage through the liner
system.
This chapter presents a description of hydrologic processes affecting
leachate generation, and methods for estimating the water balance. In
addition, several case studies comparil'lg the measured water balance at
several facilities with predictions generated by the Hydrologic Evalua-
tion of Landfill Performance (HELP) model are presented. Finally, the
sensitivity of various landfill design parameters on the water balance
components is shown using the HELP model to predict the effects.

10.2 WATER BALANCE CALCULATIONS

Potential pathways of water movement are shown in Fig. 10.1. The

two principal methods of analyzing water routing are a hand method
(Thornthwaite method) and a computer method (HELP model).

SIMULATION PROCESSES
IN THE HELP MODEL
RAlNF ALL/SNOW INTERCEPTION

PLANT GROWTH

VERTICAL PERCOLATION
. .'. ...5J..- DEPTH OF
...
. . ,' , .•. , ' HEAD

Fig, 10.1 Water pathways.

216 Water balance for landfills

10.2.1 Thornthwaite method

The earliest, comprehensive method for water balance analysis was
developed by Thornthwaite and Mather (1957). Fenn et al. (1975)
developed for the US Environmental Protection Agency (EPA) a water
balance method for predicting leachate generation for solid waste
disposal sites based on the Thornthwaite method, proposing runoff
coefficients for landfills and moisture storage values for municipal
waste. The water balance method employs average monthly values of
precipitation and other climatic parameters. To determine monthly
infiltration (IN) into the cover, one subtracts monthly runoff (R) from
monthly precipitation (P).
IN = P- R (10.1)
Runoff can be calculated from precipitation, as follows
R = CP (10.2)
where C is a runoff coefficient that can be estimated from the guidance
provided by Fenn et al. (1975):
description of the
grass-covered soil slope of ground surface runoff coefficient (C)
sandy soil flat «2%) 0.05-0.10
mild (2 - 7%) 0.10-0.15
steep (>7%) 0.15-0.20
clayey soil flat «2%) 0.13-0.17
mild (2 - 7%) 0.18-0.22
steep (>7%) 0.25-0.35
Potential evapotranspiration (PED, which depends on mean tem-
perature, heat index, and hours of sunlight, can be calculated from
tables provided by Thornthwaite and Mather (1957). The cumulative
monthly infiltration minus potential evapotranspiration (IN - PET) is
calculated. A negative number for IN - PET indicates that the cover
has a tendency to dry out; a positive value indicates that there is a
tendency for the cover soil to become wetter. If IN - PET is negative,
water may evapotranspire from the cover soil, but if the soil is already
dry, no further drying will occur. The amount of drying depends not
only on IN - PET but also on the water content of the soil.
If IN - PET is positive, water may be stored in the cover soil (which
would produce an increase in water content). However, if the water
content is already very high, the soil can store no additional water
and water will percolate downward through the cover soil. The field
capacity of the soil is the maximum water content that a soil can attain
without draining water by gravity. Field capacity can be determined by
 Water balance calculations 217
allowing a saturated soil to drain by gravity; the water content when
equilibrium is reached is the field capacity. Thornthwaite and Mather
(1957) and Fenn et al. (1975) recommended values of field capacity for
various soils. One must allocate IN - PET to either soil storage or,
when field capacity would be exceeded, to downward percolation.
An excellent description of the Thornthwaite method as applied by
Fenn et al. (1975) including its use and limitations is given by Kmet
(1982). Hand calculations are easily performed with this technique,
which is particularly suitable for analysis with spreadsheet software.
The main disadvantage is that numerous simplifying assumptions
must be made.

10.2.2 Computer analysis with program HELP

The state of the art in water balance analysis for landfills is a computer
program known as HELP. The HELP model was developed by the US
Army Engineer Waterways Experiment Station for the US EPA and is
described in detail by Schroeder et al. (1984a, b). Version 3 of the model
is being released in 1993 by the US EPA Risk Reduction Engineering
Laboratory.
The HELP model was adapted from the Hydrologic Simulation
Model for Estimating Percolation at Solid Waste Disposal Sites of the
US EPA (Perrier and Gibson, 1980; Schroeder and Gibson, 1982), the
Chemical Runoff and Erosion from Agricultural Management Systems
(CREAMS) (Knisel, 1980), and Simulator for Water Resources in Rural
Basins (SWRRB) (Williams et al., 1985). The following subsections
summarize the main features and assumptions of the model. The
User's Manuals (Schroeder et al., 1984a,b; 1988a,b) should be consulted
for further detail.
The HELP model is a quasi-two-dimensional, gradually varying,
deterministic, computer-based water budget model. It is termed
quasi-two-dimensional because it contains a one-dimensional vertical
drainage model and a one-dimensional lateral drainage model coupled
at the base of lateral drainage layers or the tops of liners incorporated
in the cover system or located below the waste. The program computes
free downward vertical drainage to the top of a liner, at which point
the liner restricts drainage and a zone of saturation can develop. The
models for lateral drainage and leakage or percolation through the liner
then use the height of saturated material above the liner to compute
simultaneously the rates of lateral drainage to collection systems and
vertical leakage through the liner, respectively. The model is termed
gradually varying because the simulation progresses through time
using analyses that are assumed steady for each time period. Version 2
of the model (the latest version at the time of this writing) uses a time
period of 6 hours (Schroeder et al., 1988b). The model is deterministic
218 Water balance for landfills

and quantitative. Finally, the HELP model is a computer-based water

budget model that runs on a personal computer; the model uses a
computer to apportion the precipitation and initial water content of
soil layers into estimates of the following water budget components:
surface runoff, evapotranspiration, changes in snow storage, changes
in water content of the soil, lateral drainage collected in each drain
system, and leakage or percolation through each liner in the system
(Fig. 10.2). The HELP program can generate daily, monthly, annual,
and long-term average water budgets.
The HELP model is a tool developed specifically to aid analysis in the
evaluation and comparison of alternative landfill designs. A secondary
goal in development of the model was to provide an accurate predic-
tion of water budget components. Testing, verification studies, and
refinements are ongoing to improve the accuracy of HELP. In general,
the accuracy and precision of the model is limited by uncertainty and
variability in the properties of material existing in landfills.

Infiltration
Daily infiltration into the landfill is determined indirectly from a sur-
face water balance. Infiltration equals the sum of rainfall and snowmelt,
minus the sum of runoff and surface evapotranspiration. Runoff and
surface evaporation are in part a function of interception. Precipitation
on days having a mean temperature below DoC is treated as snowfall
and is added to the surface snow storage. Decreases in snow storage
occur by snowmelt and surface evaporation.
Snowmelt is computed using a slightly modified version of the
simple degree-day method. The analysis permits a small quantity of
snowmelt to occur at mean daily temperatures between -5°C and
DoC to account for the variation in temperature during a day. It also
accounts for the fact that landfills often have higher soil temperatures
because of heat generated from biodegradation of waste. Snowmelt
contributes to runoff, evaporation, and infiltration.
Interception is modeled in the method of Horton (1919). Interception
approaches a maximum value exponentially as the rainfall increases to
about 5 mm. The maximum interception is a function of the quantity of
aboveground biomass of leaf area index and is limited to a maximum of
1.3 mm. The interception evaporates from the surface.
The HELP model uses the US Soil Conservation Service (SCS) curve
number method for estimating surface runoff (US Department of
Agriculture, 1972). The SCS curve number method is an empirical
procedure developed for small watersheds (12-200 hectares) with mild
slopes (about 3-7%). The method correlates daily runoff with daily
rainfall for watersheds with a variety of soils, types of vegetation, land
 Water balance calculations 219
PRECIPITATION
~ EVAPOTRANSPIRATION

r VEGETATION t RUNOFF
111111111 1111 II !I! II III II 1111 III I!! II II I "'' ' 't' ' INFILTRATION
' ' IIIII!II!'' ' ' ' 1!1 II II I!! II II Ii"" "III I!lIWiiiiiiIlIl!IIIII!
<D PERCOLATION
VERTICAL
LAYER TOPSOIL
I
.

® LATERAL DRAINAGE LAYER SAND LATERAL DRAINAGE

(FROM COVER) ___

LSLOPE
® BARRIER SOIL LINER CLAY
PERCOLATION

,
I
I
I
I
I

WASTE

SAND

LA TERAL DRAINAGE
LATERAL DRAINAGE LAYER (LEACHA TE COLLECTION)

®-_-I
FLEXIBLE MEMBRANE LINER

-- -
SAND
LATERAL DRAINAGE
LATERAL DRAINAGE LAYER (LEAKAGE DETECTION)

® FLEXIBLE MEMBRANE LINER ~ LSLOPE

BARRIER SOIL LINER MAXIMUM DISTANCE
CLAY DRAINAGE

PERCOLATION (LEAKAGE)

Fig. 10.2 Water balance that can be analyzed in program HELP.

management practices, and antecedent moisture conditions (level of

prior rainfall). Version 3 of the model (scheduled for release in 1993)
will include a procedure to adjust the curve number as a function
of surface slope since surface slopes greater than 20% can produce
significantly greater runoff.
220 Water balance for landfills

Several limitations merit mentioning.

• The SCS curve number method is applied to landfills, which are
much smaller than the watersheds for which the method was devel-
oped. Verification studies have shown good agreement between the
predicted and observed cumulative annual volume of runoff.
• Cumulative volume of runoff is independent of rainfall duration and
intensity since over a long simulation period, a variety of precipita-
tion events will occur. The predicted value represents an average of
the measured runoff for the typical variety of rainfall events.
• No surface run-on from surrounding areas is permitted by the
model.
• Estimates of runoff greater than predicted by the SCS curve number
method are produced when the surface soils are saturated or limit
infiltration due to very low hydraulic conductivity.

Evapotranspiration
Evapotranspiration consists of evaporation of water from the surface,
from the soil, and from plants. Each component is calculated separately.
Evaporation of water from the surface is limited to the smaller of
the potential evapotranspiration and the sum of snow storage and
interception. The HELP model uses a modified Penman method to
compute evapotranspiration (Ritchie, 1972). The potential evapo-
transpiration is a function of vegetative cover, daily temperature, and
daily solar radiation.
The HELP model uses Ritchie's (1972) method of evaporation from
soil. The method involves a two-stage square-root-of-time routine. In
stage one, the soil evaporation equals the evaporative demand placed
on the soil. Demand is based on energy and is equal to the potential
evapotranspiration discounted for surface evaporation and shading
from ground cover. A vegetative growth model is used to compute the
total quantity of vegetation, both active and dormant, which provides
shading. In stage two, evaporation from the soil is limited by low soil
moisture and low rates of water vapor transport to the surface. Stage
two soil evaporation is a function of the square root of the length of
time that the soil has been in dry condition.
The HELP model estimates plant transpiration by assuming that the
potential plant transpiration is a linear function of the potential evapo-
transpiration and the active leaf area index.
Key limitations are listed below.
• The potential evapotranspiration is a function solely of energy
available at the surface and, therefore, is not affected by energy
produced in the landfill, by wind, or by humidity.
• A constant value is used for the albedo (i.e., the fraction of incident
 Water balance calculations 221

solar radiation that is reflected). The value is typical for brown soils
and grasses and is modified only when the surface is covered with
snow.
• The program uses a constant evaporative zone depth. This depth is
the maximum depth to which soil suction can draw water to the
surface. This depth is a function of soil properties, design, vegeta-
tion, and climatic conditions.
• The vegetative growth model produces representative leaf area
indices and biomass estimates that are sufficient to estimate
interception, surface shading, and plant transpiration.

Subsurface water routing

Subsurface water routing includes vertical unsaturated drainage,
percolation through saturated soil liners, leakage through geo-
membranes, and lateral drainage in drainage layers. In modeling these
processes, the soil moisture of each layer (with the exception of barrier
layers) is computed by sequential analysis. The soil moisture controls
the rate of subsurface water movement, but the rate of movement also
affects the moisture content. Consequently, an iterative analysis pro-
cedure is used.
The HELP model simulates unsaturated vertical drainage by assum-
ing unit hydraulic gradient. The unsaturated hydraulic conductivity is
calculated by the Campbell (1974) equation which is based on the
Brooks and Corey (1964) model. The model does not allow drainage
from one layer at a rate greater than the maximum infiltration rate of
the underlying layer. This allows the simulation of a layer having a
lower hydraulic conductivity below a layer of higher hydraulic con-
ductivity without treating the lower layer as a liner.
Vertical drainage through soil liners is termed percolation in the
HELP model. Soil liners are assumed to remain saturated; however,
percolation occurs only when there is a zone of saturation directly
above the liner. Percolation is calculated from Darcy's law.
Leakage through geomembranes is modeled as a reduction of the
cross-sectional area of flow through the subsoil below the liner. The
rate of flow through the leaking subsoil is computed as the percolation
rate through a saturated barrier soil liner. This method provides
good results for composite liners but is not very good for just a geo-
membrane. Version 3 will include an improved leakage model for
geomembranes based on the works of Brown et al. (1987) and Giroud
et al. (1989a,b,c).
The HELP model simulates lateral drainage using a steady-state
analytical approximation of the Boussinesq equation (McEnroe and
Schroeder, 1988). The analytical approximation was developed by
converting the Boussinesq equation to a nondimensional form and
222 Water balance for landfills

solving it for the extremes in nondimensional average saturated soil

depth. These two solutions were then fitted with an approximation
that covers the rest of the range of nondimensional depths. The
approximation involves less than 1% error.

Vegetative growth
The HELP model accounts for seasonal variation in active and dormant
aboveground biomass and leaf area index through a general vegetative
growth model. The maximum value of leaf area index depends on type
of vegetation, soil fertility, climate, and management factors.

Input data
Required climatic input data include daily precipitation, daily mean
temperature, daily solar evaporation, maximum leaf area index, grow-
ing season, and evaporative zone depth. The computer program
includes default daily weather data for 102 US cities and can syntheti-
cally generate weather data for 183 US cities, if desired.
The input soil data required for each layer include porosity, field
capacity, wilting point, and saturated hydraulic conductivity. Default
data are included for 15 soil types as well as solid waste. Design data
include SCS runoff curve number (defaults are provided), surface area,
and miscellaneous control parameters such as number of layers in the
profile and thicknesses of the layers.

Output data
The output from the HELP consists primarily of percolation or leakage
through each liner and depth of saturation on the surface of liners
(e.g., in drainage layers). Incremental and cumulative qualities of water
budget for the various components are computed and printed.

10.3 CASE STUDIES

Available data on landfill leachate production are very limited, especially

for periods of record that extend significantly beyond the initial water-
balance equilibration period, which may last up to several years.
Available data on the important facets of the water balance, such as
runoff, evaporation, rainfall, soil moisture, leachate ponding depths,
percolation rates and detailed soil characteristics, are even more limited.
This may be affected to some degree by the reluctance of waste dis-
posal operators to subject their facilities to public scrutiny. Also, this
type of data collection can be costly. However, the verification of
 Case studies 223
water balance models and the understanding of processes involved in
leachate production and migration are highly dependent on obtaining
this information.
This section presents published water balance data for 17 landfill
cells at six sites. Leachate drainage rates were measured from all cells.
Runoff was measured from 11 cells at two sites. Daily rainfall was
measured at one site. Soil moisture, leachate ponding depths and
barrier-soil percolation rates were not adequately or consistently
measured in any of the cells. Most sites had at least limited data on the
hydraulic conductivity of the clay liner, although the testing methods,
and therefore the meaning of the test results, varied widely. Data
describing cover soils or the extent of surface vegetation were generally
lacking.

10.3.1 University of Wisconsin lysimeters

From 1970 to 1977, eight large lysimeters filled with either shredded or
unprocessed refuse were monitored for surface runoff and leachate
production at the University of Wisconsin at Madison (Ham 1980).
Each cell was 18 m long by 9 m wide. The depth of refuse was either
1.2 m or 2.4 m. Refuse was underlain with a 100 mm layer of crushed
granite over a polyethylene barrier. Bottom slopes, of approximately
3%, directed leachate to a collection box at the center of the cell where
it was periodically pumped and the volume of leachate was measured.
Four cells were covered with 150 mm of sandy silt soil; the remaining
four cells were left uncovered. The top surfaces of all cells were sloped
at 3% toward one of the 18 m long walls where surface runoff was
collected and measured.
Climatological conditions at this location are as follows:
1. average temperature is 7°C;
2. average annual precipitation is 790 mm;
3. minimum daily temperature falls below freezing on 163 days per
year; and
4. average daily solar radiation is 330 langleys.
The vegetative cover was mixed volunteer vegetation, comparable
to meadow grass, which became established on both covered and
uncovered cells over a several year period. This vegetation grew more
quickly and more densely on the uncovered cells.
The water balance measurements are shown in Table 10.1 expressed
in terms of percent precipitation based on US National Oceanic and
Atmospheric Administration (NOAA) precipitation records for a gage
located approximately two miles from the landfill site. It is important
to highlight the variability in the measured results between similar
cells. The measured runoff varied from 71-129% of the mean for the
224 Water balance for landfills

Table 10.1 Cumulative measurements of water balance components for

landfill cells at University of Wisconsin at Madison
cell number period of runoff ET + LIst leachate
measurement* (% precip.) (% precip.) drainage
(yr) (% precipJ
- - - - - - _ . _ - _ . _ .. _ - - - - -

Covered cells
1 4 5.4 67.7 26.9
2 4 6.0 67.1 26.9
3 5 9.1 73.6 17.3
8 3 9.8 67.5 22.7
Mean 7.6 68.9 23.5
Standard deviation 2.2 3.1 4.6
Uncovered cells
4 5 3.6 79.2 17.2
5 2 3.4 90.1 6.5
6 2 3.9 92.6 3.5
7 3 2.8 73.5 23.7
Mean 3.4 83.9 12.7
Standard deviation 0.5 9.1 9.4

* Excludes first two years of landfill operation.

t ET + tl.S = evapotranspiration + change in landfill moisture storage
= 100 - runoff % - leachate drainage %

covered cells and 82-115% for the uncovered cells. The measured
drainage varied from 74-114% of the mean for the covered cells and
28-187% for the uncovered cells. The difference between the cell
designs (besides the presence of a soil cover) was depth of the waste
layer and presence of shredded refuse, neither of which should sig-
nificantly affect the water balance after a two-year equilibration period.
Overall, evapotranspiration accounted for 69-84% of the water balance
and was 2.9-6.6 times as large as the leachate drainage volume. Runoff
was the smallest component of the water balance ranging from 3-8%
of precipitation.

10.3.2 Sonoma County test cells

A solid-waste stabilization project was conducted in Sonoma County,
California from 1971 to 1974 to determine the effect of applying water,
septic tank pumpings, or recycled leachate to waste in landfills (EMCON
Associates 1975). Each of the five cells studied was 15m long by 15m
wide and 2.4m deep. The landfill contained a soil liner; no flexible
membrane liner was present. The soil liner material was native sandy
clay. Pervious lenses were encountered during excavation and were
replaced by a O.6m thickness of sandy clay. The landfill cover material
 Case studies 225
Table 10.2 Cumulative measurements of water balance components for
landfill cells at Sonoma County, California

cell period of runoff ET + L15 + leachate

measurement* ('Yo precip.) PERC t drainage
(yr) ('Yo precip.) ('Yo precip.)
------ ---------- -----

A 2 59.3 39.5 1.26

B 2 64.1 32.7 3.21
C 2 -+ -+ 314.49
D 2 -+ + 996.53
E 2 59.7 35.9 4.35
Mean§ 61.0 36.0 2.94
Standard deviation§ 2.7 3.4 1.56

* Excludes first year of landfill operation.

t ET + dS + PERC = evaporation + change in landfill moisture storage + percolation
through bottom soil liner
= 100 - runoff % - leachate drainage %
tNo measurement
§Computed only for cells A, f3 and E.

was a 0.6m thick layer of sandy clay. There was no documentation of

surface vegetation.
Cells A, Band E were constructed and operated as typical sanitary
landfills except that the moisture content of the refuse was brought to
field capacity with water prior to capping in cell B and with septic tank
pumpings in cell E. Cells C and D were constructed with an inflow
pipe network placed on top of a 300 mm thick sand/gravel distribution
medium installed between the waste layer and the soil cover. Water
was continuously added to the waste through this network in cell C,
whereas landfill leachate was continuously recycled through the waste
layer in cell D.
This test site was located 45 miles north of San Francisco. The mean
temperature is 14°C, with the daily minimum temperature falling
below freezing on 39 days per year. Mean annual precipitation is
790 mm. Ninety-five per cent of this precipitation occurs from October
to April. The mean daily solar radiation is approximately 410 langleys.
Daily precipitation was measured at the landfill site.
The water balance measurements are shown in Table 10.2. In this
case, the largest component of the water balance was runoff at 61 % of
precipitation. The evapotranspiration variable accounted for 36% of the
precipitation, approximately 12 times the leachate drainage percentage.
The leachate drainage from cell C was 314% of precipitation compared
to inflow through the pipe network of 388%. For cell D, the leachate
drainage was 997% of precipitation compared to inflow of 1018%.
226 Water balance for landfills

10.3.3 Boone County test cell

A test cell was studied from 1971 to 1980 in Boone County, Kentucky,
to evaluate volume and characteristics of the leachate, composition
of gases, internal temperature, settlement, and clay liner efficiency
(EM CON Associates, 1983; Wigh, 1984). The cell consisted of a 9-m
wide by 45-m long trench with vertical walls. The bottom of the trench
at both ends was sloped toward the center at 14%. The middle 15m
were sloped at 7% to the transverse centerline. A 0.8mm synthetic
liner, 9 m wide by 15 m long, was centered over the base of the cell. A
percolation collection pipe embedded in gravel was placed on the
synthetic liner at the bottom of the cell. A 450-mm thick compacted
clay liner was placed over the synthetic liner and collection pipe. This
liner was found to have an average in-place hydraulic conductivity of 4
X 10- 7 cmls at the conclusion of the cell study. A second pipe was
embedded in a gravel-filled section of the clay liner directly above the
lower pipe to collect lateral leachate drainage above the clay liner. A
polyethylene sheet was placed beneath this pipe to prevent leachate
from short-circuiting to the lower pipe. Residential refuse was placed
and compacted above this liner system, and a 0.6 m layer of cover soil
was deposited onto the completed waste layer.
At the end of the study, a field assessment was conducted. The
cover soil was classified as CL according to the Unified Soil Classification
System (USCS) and was found to have an average in-place hydraulic
conductivity of 5 x 10- 5 cm/s. Also, secondary openings were found in
the soil cover through which relatively rapid infiltration could occur.
The test site is located approximately 30 km south of Cincinnati,
Ohio. The mean annual temperature is 12°C, with the daily minimum
temperature falling below freezing on 111 days per year. The mean
annual precipitation is 1.1 m. The mean daily solar radiation is approxi-
mately 360 langleys. Precipitation at the test site was recorded once or
more per week throughout the study period. These values were used
to adjust daily precipitation records from the nearest NOAA weather
station, located approximately 24 kms away at Covington, Kentucky.
At this landfill, leachate drainage amounted to 28.8% of the precipita-
tion over a 7-yr measurement period. Runoff was not measured. A
negligible amount of percolation was measured from the percolation
collection pipe.

10.3.4 Wisconsin County landfills

Three landfills located in Brown, Eau Claire and Marathon Counties,
Wisconsin were monitored for periods ranging from 3 to 7 yr (Gordon
et al., 1984). The Brown County landfill began in 1976 with a planned
size of 23 ha. The landfill is located near Green Bay, Wisconsin, where
 Case studies 227

the mean annual temperature is 7°C. The daily minimum temperature

falls below freezing on 163 days per year. The mean annual precipita-
tion is 690 mm, and the mean daily solar radiation is approximately
330 langleys. Daily precipitation data were measured at the NOAA
weather station at Green Bay.
The base of the Brown County landfill consists of a 1.2-m thick
compacted clay liner and a leachate collection system. Phases 1 through
3 (approximately 7ha) were designed with a 1% base slope, a leachate
flow distance of 91 m, and a 300-mm thick sand blanket on the base
and sidewalls; the leachate flow distance was shortened to 30 m for
Phase 4 (approximately 3 ha). In-place hydraulic conductivity tests
resulted in an average hydraulic conductivity of the liner of 7 x
10- 8 cm/s. The average depth of waste was 23 m. The waste initially
consisted of municipal and commercial refuse. Later, industrial waste
consisting of flyash and water treatment plant sludge was added. Daily
cover is a 150mm thickness of silty clay soil (USCS classification CL).
During the period of leachate monitoring presented here, much of the
landfill area was overlain only with daily cover. Final cover consists of
0.6 m of compacted clay and 150 mm oftopsoil.
The Eau Claire County landfill opened in December 1978 in the
vicinity of Eau Claire, Wisconsin. The data presented here covers the
period of landfill expansion up to 6 ha.
The mean annual temperature is 6°C with 172 days per year experienc-
ing a minimum temperature below freezing. The mean annual pre-
cipitation is 740 mm, and the mean daily solar radiation is approximately
330 langleys.
The base of the landfill consists of a 1.2 m thick compacted clay liner
overlaid with a 300-mm thick sand blanket and a leachate collection
system. The liner slope is 1%, and the maximum leachate flow distance
along the base is approximately 40 m. The waste is primarily municipal
and commercial refuse with minor amounts of industrial wastes. Daily
cover at this landfill is a 150 mm layer of sand. A 750 mm interim layer
of papermill sludge covered a large percentage of the site during the
period considered here. Final capping will include a 300mm sand
blanket over the sludge and then 150 mm of topsoil. The ultimate
maximum fill thickness will be approximately 15 m.
The Marathon County landfill was opened in December 1980 in
central Wisconsin near the city of Wausau. Approximately 6ha were
under development at the end of the data collection period presented
in this section. The mean annual temperature is 6°C with 170 days per
year experiencing a minimum temperature below freezing. The mean
annual precipitation is 790 mm, and the mean daily solar radiation is
approximately 330 langleys.
The facility is designed with a 1.2-m-thick compacted clay liner and a
leachate collection system. The liner is sloped at 1% toward 200-mm-
228 Water balance for landfills

diameter perforated PVC pipes. The pipes are embedded iri a 300-mm-
deep trench oriented at approximately 45 degrees to the slope of the
liner. The clay liner thickness increases to 1.5 m in the vicinity of these
trenches. The maximum leachate flow distance is about 75 m in Phases
1 and 2 (3.6ha total). Phase 3 (2ha) was designed for a maximum flow
distance of 30 m. Following construction of the clay liner and installa-
tion of the collection pipe, a minimum 300-mm-thick silty sand drainage
blanket was placed over the base and sidewalls. The waste is 75%
municipal refuse and 25% papermill sludge. The thickness of the waste
layer ranges from 15 to 25 m. The daily cover is 150 to 300 mm of sand.
The final cover is 0.6 m of clay covered by 150 mm of silty sand and
150 mm of topsoil.
The leachate drainage results for these three landfills are shown in
Table 10.3. The leachate collected ranged from 3.1 to 7.7% of precipita-
tion. Runoff was not measured. It is possible that steady-state moisture
conditions had not been reached during the period of these measure-
ments because the landfills remained active during this period and
because of their large depth. If this was the case, the drainage percent-
ages shown in Table 10.3 are probably lower than the long-term steady
state values.

10.4 SENSITIVITY OF WATER BALANCE TO

LANDFILL DESIGN PARAMETERS

10.4.1 Introduction
This section examines the sensitivity of a landfill water balance to
numerous landfill design variables using the HELP model. This in-
formation is useful in a variety of ways. It can aid the design engineer
in selecting preliminary design alternatives for municipal or hazardous
waste landfills. It can serve as a basis for regulatory agencies to establish
and evaluate technical guidelines. It can also provide additional insight

Table 10.3 Cumulative measurements of water

balance components for Wisconsin County landfills

landfill period of measurement* leachate drainage

(yr.) (% precip.)
-------~~--- - -

Brown 5 3.1
Eau Claire 5 7.7
Marathon 3 6.9

* Excludes the first year of landfill operation, except for

Marathon which includes the first year.
 Sensitivity of water balance to landfill design parameters 229

on the importance and interaction of specific design variables on the

water balance. Finally, it can assist in evaluating the suitability of
methodologies used in the computer model. The analyses include
examination of cover systems and lateral drainage/liner systems.

10.4.2 Comparison of typical cover designs

Design parameters
Three locations were studied to determine the effect of various clima-
tological regimes on cover performance - Santa Maria, California;
Schenectady, New York; and Shreveport, Louisiana. These locations
represent a wide range in levels of_precipitation, temperature, and
solar radiation.
Two cover designs were examined as shown in Fig. 10.3. One is
typical of newer RCRA landfills where 600 mm of topsoil overlies a
300-mm-thick lateral drainage layer having a saturated hydraulic
conductivity of 3 x 1O-2 cm/sec, a slope of 0.03m/m and a maximum

VEGETATION

a) Hazardou. Wa.t. Landfill Cov.r D •• llln

VEGETATION
.\

BARRIER SOIL

b) Municipal Landfill Cov.r D •• llln

Fig.l0.3 Typical cover profiles: (a) hazardous waste landfill cover design; and
(b) municipal landfill cover design.
230 Water balance for landfills

drainage length of 60 m. The drainage layer is underlain by a 0.6-

m-thick soil liner having a saturated hydraulic conductivity of 1 X
10- 7 cm/sec. The other design is typical of older municipal sanitary
landfills where a topsoil layer overlies a 0.6-m-thick soil liner having a
saturated hydraulic conductivity of 1 x 10- 6 cm/sec.
Two types of topsoil were considered in the cover designs: sandy
loam and silty clayey loam. The topsoil type was used to select soil
porosity, field capacity, wilting point and hydraulic conductivity,
besides influencing the selection of the runoff curve number. In addi-
tion to two types of topsoil, two thicknesses of topsoil were examined:
450 and 900mm.
The vegetative cover was designated as being either a good stand of
grass or a poor stand of grass. This selection dictated the values for
leaf area index, evaporative depth and runoff curve number, and
influenced the value used for the saturated hydraulic conductivity of
the topsoil.

Effects of vegetation
A good stand of grass represents three times the quantity of vegetation
as that of a poor stand. Table 10.4 presents the water balance results
for both cover systems at all three sites as a function of level of vegeta-
tion. The results are given in units of percent of the precipitation.
Vegetation reduces surface runoff and increases evapotranspiration.

Table 10.4 Effects of climate and vegetation

two layer cover design* three layer cover design t

location location

CA LA NY CA LA NY
(% precipitation) (% precipitation)

Poor grass
Runoff 5.6 4.6 5.5 3.0 4.4 2.2
Evapotranspiration 51.8 53.0 52.1 51.6 51.9 50.3
Lateral drainage 41.2 40.6 44.0
Percolation 42.6 42.4 42.4 4.2 3.1 2.5
Good grass
Runoff 3.1 0.2 3.5 0.0 0.2 0.0
Evapotranspiration 55.0 57.2 55.3 52.6 53.0 51.0
Lateral drainage 43.2 43.7 45.5
Percolation 42.9 42.6 41.2 4.2 3.1 2.5

* 900 mm of sandy loam topsoil and 0.6 m of 10- 6 em/sec clay liner.
t 450mm of sandy loam topsoil, 300mm of 0.03 em/sec sand with 60m drain length at 3%
slope, and 0.6m of 10- 7 em/sec clay liner.
 Sensitivity of water balance to landfill design parameters 231

Evapotranspiration is greater because of the plant demand for moisture

and a greater quantity of water is available for evapotranspiration due
to greater infiltration and a greater evaporative zone. Runoff is less
because vegetation increases the minimum infiltration rate, drying
rate, interception, and surface roughness resulting in a decrease in
the runoff curve number. The influence of surface vegetation on the
volume of lateral drainage and percolation or leakage from the cover is
varied. However, the quantity of vegetation tends to have very little
effect on the percolation or leakage through the cover system. For the
cover with lateral drainage, the increase in infiltration with good grass
was greater than the increase in evapotranspiration, resulting in a
larger volume of lateral drainage and a negligible change in percola-
tion. For the cover without lateral drainage, the increase in infiltration
yielded high heads or depths of saturation above the liner that permitted
greater evapotranspiration by maintaining higher moisture contents in
the evaporative zone. Consequently, the increase in evapotranspiration
was greater than the increase in infiltration. This resulted in a trend
toward a small decrease in percolation for a higher level of vegetation.
The opposite trend may occur for vegetative layers having a lower
saturated hydraulic conductivity and a higher plant available water
capacity (field capacity minus wilting point). The results were similar at
all three sites despite quite different climates.
In summary, vegetation decreases runoff and increases evapo-
transpiration but tends to have little effect on the rest of the water
balance. The magnitude of the effects is design dependent and to a
lesser degree climate dependent. The main function of vegetation is to
control erosion.

Effects of topsoil thickness

Two topsoil thicknesses were examined: 450 and 900 mm. Table 10.5
presents the water balance results for the two layer cover system at all
three sites as a function of topsoil thickness. The results are given in
units of percent of the precipitation. The cover system with lateral
drainage was not used in this analysis because lateral drainage would
negate the effects by preventing or minimizing the intrusion of the
saturated zone above the liner into the evaporative zone.
Significant differences existed between the 450- and 900-mm topsoil
depth simulations in the absence of lateral drainage. The effects were
similar at all three sites. Runoff and evapotranspiration were greater
for the shallower depth to the liner, indicating that the head above the
barrier soil layer maintained higher moisture contents in the evapora-
tive zone. The percolation was consequently less than the cases with
greater topsoil thickness. The 900-mm depth to the liner permits larger
heads and longer sustaining heads since a greater thickness of material
232 Water balance for landfills

Table 10.5 Effects of climate and topsoil thickness

45.7 em of topsoil 91.4 em of topsoil

location location

CA LA NY CA LA NY
(% precipitation) (% precipitation)

Runoff 11.2 7.5 13.4 5.6 4.6 5.5

Evapotranspiration 51.9 56.9 54.5 51.8 53.0 52.1
Percolation 36.9 35.6 32.1 42.6 42.4 42.4

Sandy loam topsoil with a poor stand of grass underlain by O.6m of 1O-6 cmisec clay
liner.

is below the evaporative zone and is free from abstraction of water

by evapotranspiration. The larger heads provide a greater pressure
gradient to increase the leakage rate through the cover system.
In general, the effects of topsoil thickness would vary greatly as
the thickness increases from several centimeters to several meters.
Throughout the transition, the quantity of runoff should continue to
decrease until the depth to the liner becomes sufficiently great so as to
prevent the zone of saturation from ever climbing into the evaporative
zone. Similarly, the percolation through the liner should continue to
increase until there is no interaction between the saturation zone and
the evaporative zone. The evapotranspiration is expected to increase
initially as the available storage in the evaporative zone increases, that
is until the depth to the liner equals the maximum depth that evapo-
transpiration can reach. At greater depths the evapotranspiration should
continue to decrease until the depth to the liner is sufficient to prevent
any further interactions between the evaporative and saturation zones.
While percolation increases with topsoil thickness given identical
properties for all layers in the cover system, adequate thickness must
be provided in a design to insure the integrity of the cover system.
A small topsoil thickness would not provide adequate water storage
to support vegetation, maintain soil stability and control erosion.
Similarly, a shallow depth to the liner would promote desiccation or
freezing of the liner which may greatly increase its permeability and
therefore the percolation.

Effects of topsoil type

Two topsoil types were examined: sandy loam and silty clayey loam.
Table 10.6 presents the water balance results for the three layer cover
system at all three sites as a function of topsoil type. The results are
given in units of percent of the precipitation. The cover system without
 Sensitivity of water balance to landfill design parameters 233

Table 10.6 Effects of climate and topsoil types

three layer cover design*

sandy loam silty clayey loam
location location

CA LA NY CA LA NY
(% precip.) (% precip.)

Runoff 3.0 4.4 2.2 21.6 22.3 19.2

Evapotranspiration 51.6 51.9 50.3 61.2 64.4 58.6
Lateral drainage 41.2 40.6 44.0 15.0 11.3 20.3
Percolation 4.2 3.1 2.5 2.2 2.0 1.9

• 450 mm of topsoil with poor stand of grass, 300 mm of 0.03 cm/sec sand with 60 m
drain length at 3% slope, and 0.6 m of 10- 7 cm/sec clay liner.

lateral drainage was not used in this analysis because the intrusion of
the saturated zone above the liner into the evaporative zone would
decrease the magnitude of the effects.
The results show that the clayey topsoil significantly increased both
runoff and evapotranspiration, which in turn greatly decreased lateral
drainage and percolation. The results were similar at all three sites.
Runoff increased from about 3% to 20% of the precipitation, due
primarily to the larger runoff curve number selected for the clayey
soil based on its lower minimum infiltration rate. Evapotranspiration
increased approximately from 51 % to 61 % of precipitation, due to the
lower hydraulic conductivity of the clayey soil and, more importantly,
the larger plant available water capacity. The lower hydraulic con-
ductivity of the clayey soil slowed the drainage rate, maintaining
moisture contents above field capacity for longer periods of time and
allowing greater evapotranspiration. The larger plant available water
capacity of the clayey soil provided a larger moisture reservoir avail-
able for evapotranspiration after gravity drainage ceased. The lateral
drainage was reduced from about 42% to 16% of the precipitation, and
the percolation was reduced from about 3% to 2% of precipitation.

Use of lateral drainage layer

In general, the use of a lateral drainage layer would be expected to
decrease the height of the saturation zone above the liner by draining
some of the infiltrated water from the cover system. As such, percola-
tion through clay liners would decrease slightly. In addition, runoff
and evapotranspiration would also tend to decrease but the magnitude
of the change would be design dependent. Topsoil thickness, topsoil
type, vegetation and climate would have impacts.
234 Water balance for landfills

Effects of climate
The effects of climate were examined in each of the previous sections.
Climate affects the absolute magnitude, in millimeters, of the water
budget components. However, Tables 10.4 through 10.6 show that
climate has a much smaller effect on the relative magnitude of the
water budget components in terms of percent of the precipitation. The
relative proportions of the water budget components are primarily
design dependent while the magnitudes are strongly dependent on the
magnitude of the precipitation.

10.4.3 Liner/drain systems

This section examines the effects of liner/drain system design on the
performance of the drain system under conditions typical of cover
systems and leachate collection systems in open and closed landfills.
Performance was determined by the apportionment of the drainage
into the drain layer between lateral drainage and percolation through
the liner. In addition, the effect of design on the resulting depth of
saturation was also examined. For the cover system or open landfill the
drainage into the drain layer was 1.3 m/yr, distributed temporally in
accordance with the precipitation at Shreveport, LA. For the closed
landfill the drainage into the drain layer was distributed uniformly
through time at a rate of 200 mm/yr.

Clay liner/drain systems

Saturated hydraulic conductivities
The liner/drain system used in this analysis is shown in Fig. lO.4a.
The value of KD (the saturated hydraulic conductivity of the drain
layer) ranged from 0.001 to 1 cm/sec while the value of KP (the
saturated hydraulic conductivity of the clay liner) ranged from 10- 8 to
10- 5 cm/sec. The slope of the liner surface toward the drainage col-
lector was 3%, and the maximum drainage length to the collector was
23 m. The results of the drainage efficiency determinations for the
various combinations of KD and KP are shown in Fig. 10.5 where the
average annual volumes of lateral drainage and percolation expressed
as a percentage of annual inflow are plotted.
Summarizing the results shown in Fig. 10.5, the saturated hydraulic
conductivity of the liner is the primary control of leakage through a
clay liner. At hydraulic conductivities below about 10- 6 cm/sec the
leakage is nearly proportional to the value of KP; that is, an order
of magnitude decrease in the value of KP yields nearly an order of
magnitude decrease in percolation. The value of KD has only a small
effect on the leakage through liners having a KP of 10- 7 cm/sec or less.
 (a) (b) (c)

(d) (e) (f)

~:~¢:~R!:%;P;h~
. ~ waste layer - ~
Q;q'2Oh~O. H'sYq:,Oq·ciF

{(f.{!?~;~~\~ri{~~~;··ll~!r

Fig. 10.4 Alternative designs considered.

100 0

~~-::-~-o.,_,
\'-."
90 10

80
'-, '\ \ "" ", 20

j ' 70 0 ' 30
0 . \', \ \ """·0 ~
~ 60 - - - - - KD = 1em/s ~ \ . \ 40~
~ 50 - - - KD = 0.1em/s .~ \ \. \ of!.
50 -;;
II>
i'
KD = 0.01 em/s 1 \ 0 \ \ 0
40 6O~
~\
"e'0'"
- - KD = 0.001 em/s '0" \ \

30 ~ "',. 0 70
~II>
l'!!II> .~ \\
Q.
0 50 in.lyr inflow
20 "'- 80
1ii
--'
10
0 8 in.lyr 55 inflow
~ --'B 90
0 100
10-8 10- 7 10-6 10-5
KP (em/s)

Fig. 10.5 Effect of saturated hydraulic conductivity on lateral drainage and

percolation.
236 Water balance for landfills

Changing the value of KD by three orders of magnitude when using

these low permeability liners yields much less than an order of mag-
nitude change in percolation.
Figures 10.6 and 10.7 show that percolation tends to dominate at
ratios of KD/KP below 106 . This is particularly true as the depth of
saturation or inflow decreases. When heads remain constant, the ratio
of lateral drainage to percolation is a linear function of KD/KP. Using
the maximum head allowed by RCRA of 300 mm and the current
minimum KD/KP ratio implied by RCRA of 105 , a percolation of 2.3%
of inflow results; however, an unusually large steady-state inflow of
2 m/yr is required to achieve this condition. Therefore when using the
RCRA guidance design, the peak and steady-state average heads will
be considerably smaller than 300 mm at virtually all locations.

Slope and drainage length

The volumes of lateral drainage and percolation vary little with changes
in slope and drainage length under both steady and unsteady inflows.
A ninefold increase in slope reduced the percolation by a maximum of

8 In./yr 55 Inflow

* KD - 0.001 em/s

n.
C?J
o"-
C?J
55Q"<0.1 In.
55 Q" - 0.2 In.
55 Q" := 1.5 In.
55 Q" := 11 In.

KO/KP
Fig. 10.6 Effect of ratio of drainage layer saturated hydraulic conductivity to
soil liner saturated hydraulic conductivity on ratio of lateral drainage to
percolation for a steady-state (55) inflow of 8 in/yr (200 mm/yr).
 Sensitivity of water balance to landfill design parameters 237

25% for the unsteady inflow and 13% for the steady inflow. As the
drainage length reduces and the slope increases, the lateral drainage
rate increases. As a result, the head decreases and is maintained at
smaller depths for shorter durations. Consequently, the percolation
decreases since it is a function of the head on the liner. A ninefold
decrease in drainage length reduced the percolation by a maximum of
50% for the unsteady inflow and 25% for the steady inflow. A ninefold
increase in slope and decrease in length decreased the percolation by
about 60% for the unsteady inflow and about 30% for the steady
inflow.
The head in the drain layer varies greatly with changes in slope
and drainage length. For a steady inflow the average head increases
linearly with an increase in drainage length and an increase in the
inverse of the slope. A similar relationship exists between the peak
average head during the simulation and the ratio LIS (drainage length
divided by drainage slope) for unsteady inflow. The average head is
slightly influenced by the product of the slope and drainage length
when the head is similar to this product.

103

lcf!

10 1
1 em/s

0.. 100
c:J
KD = 0.01 cm/s
"- ¢
0
c:J 10- 1 * KD = 0.001 cm/s

- . - Peak Y <= 2.5 In.

10- 2
- - - - - Peak y <= gin.

--Peak Y <= 24 in.

10- 3
- - - Peak Y <= 55 in.

KO/KP
Fig. 10.7 Effect of ratio of drainage layer saturated hydraulic conductivity to
soil liner saturated hydraulic conductivity on ratio of lateral drainage to
percolation for an unsteady inflow of 50 inlyr (1.3 m/yr).
238 Water balance for landfills

Geomembrane/drain suptems
A single synthetic liner under a drain layer as shown in Fig. 10.4 is
examined in this section. It is assumed that the synthetic liner was laid
directly on a 3 m-thick layer of native subsoil. The drainage layer had a
saturated hydraulic conductivity of 10- 2 cm/sec, a slope of 3% and
a drainage length of 23 m. This case will be used to demonstrate
the influence of the synthetic liner leakage fraction and the saturated
hydraulic conductivity of the native subsoil on the liner system per-
formance. The properties of the subsoil ranged from sand to clay in the
analysis.

Liner leakage fraction

Brown et al. (1987) conducted laboratory experiments and developed
predictive equations to quantify leakage rates through various size
holes in synthetic liners over soil. They assumed that the measured
leakage rates corresponded to a uniform vertical percolation rate
equal to the saturated hydraulic conductivity through a circular cross-
sectional area of the soil liner directly beneath the hole. Using the data
relating leakage and cross-sectional area of flow, Brown et al. (1987)
developed predictive equations for the radius or area of this flow cross
section as a function of hole size, depth of leachate ponding, and
saturated hydraulic conductivity of the soil. Figure 10.8 presents their
results. The radius of saturated flow through the subsoil was signifi-
cantly greater than the radius of the hole in the synthetic liner. The
cross-sectional area of saturated flow was multiplied by the number of
holes per unit area of synthetic liner to compute the synthetic liner
leakage fraction. Liner leakage fraction is simply defined as the total
horizontal area of saturated flow through the subsoil beneath all of the
liner holes divided by the horizontal area of the liner.
Liner leakage fraction is a function of many parameters, some
quantitatively defined and others qualitatively. Liner leakage fraction
increases linearly with increases in the number of holes of the same
size and shape. Shape also has a strong effect on the leakage; tears
have larger leakage than punctures. Increasing the size of circular holes
yields only a slight increase in the leakage while increasing the length
of a tear or bad seam increases the leakage nearly linearly. Leakage also
increases nearly linearly with increases in head or depth of saturation
above the liner. The leakage fraction is also affected by the gap width
between the liner and the subsoil. Gap width is a measure of the seal
between the liner and the subsoil. The smaller the gap the better the
seal. The seal is a function of the subsoil, installation, liner placement
and subsoil preparation. Installation of the liner on coarse-grained
subsoil, clods, debris or filter fabric provide a poor seal as will wrinkles
in the liner. Coarse-grained subsoils decrease the leakage fraction while
 Sensitivity of water balance to landfill design parameters 239

-
105
oJ

Upper bound is for 0.08 - cm - dia. open ings.

Lower bound is for 1.27 - cm - dia. openings. OJ
[]I
Gl 104 c
L- -
0
10- 4 cm/s c(I)
a: I2Z?ZI KP a 3.4 X
n
L-
CD \03 E=::J KP 3.4 X 10 - 7 cm/s Cl

n c:
(I)
10 CD
III
C>
c: '"
oJ
CD
102 20 w
c:
(I) C>
n c:
-u
...
0
50
a
a 10 1 n
rn
L- 100
"'0
Gl
.D '-
E
Z
::J 100
200
'"E
L-
a
500
10'1 c:
10, 6 10"5 10, 3 10' ~ 10, 1 =>
10" 100

Synlhetic Liner Leaka ge Fraction . LF

Fig. 10.8 Synthetic liner leakage fraction as a function of density of holes, size
of holes, head on the liner, and saturated hydraulic conductivity of the liner.

greatly increasing the leakage. The greater permeability of coarse

materials allows greater flow through a smaller area of saturated flow,
reducing the spreading required to accommodate the leakage through
the liner.

System performance
The percolation rate through a leaking synthetic liner is a linear
function of the leakage fraction for a given subsoil when the average
head on the liner is constant. The percolation rate expressed as a
percentage of inflow rate is shown graphically in Fig. 10.9 as a function
of the leakage fraction. This relationship is shown for a range of values
of the average head and for a steady inflow rate of 200 mm/yr. Figure
10.9 emphasizes the significant influence of average head or inflow on
controlling the distribution of the inflow between vertical percolation
and lateral drainage. This figure shows that to maintain the vertical
percolation rate at less than 1% of the inflow rate for heads greater
than 2.5 mm, the leakage fraction for a clay subsoil (KP = 10-6 cm/sec)
must be less than 5 X 10- 4 and for a sandy subsoil (KP = 10-3 cm/sec)
must be less than 5 x 10- 7 • The overall effectiveness of a flexible
membrane liner is equivalent to a soil liner having a saturated hydraulic
conductivity equal to the product of the leakage fraction and the
saturated hydraulic conductivity of the subsoil when the conductivity
240 Water balance for landfills

3: 3:
o o

o
B

LF x KP (em/s)
Fig. 10.9 Effect of liner leakage fraction on system performance.

of the subsoil is equal to or less than the conductivity of the material

above the liner.

Double liner systems

A comparison of designs in Fig. lO.4c-f was made by Schroeder and
Peyton (1990). Their results are summarized here. They found that the
Design (f) in Fig. 10.4f is the most effective in detecting the earliest
leaks with the least amount of vertical leakage through the primary
liner and also through the bottom soil liner. Design (d) yields the same
quantity of leakage through the primary liner; however, leakage in
design (d) would probably never be detected or collected. Therefore,
the bottom liner in design (d) is not functional. Designs (d) and
(e) yield the same leakage through the bottom liner but design (e)
detects leakage through the primary liner at the lowest leakage frac-
tion. Design (c) also detects leaks at very small leakage fractions but
allows significant vertical percolation through the bottom soil liner
before detection. The leakage through the primary liner in designs (c)
and (e) is large even at low leakage fractions. Therefore, synthetic
membranes placed on highly permeable subsoils are ineffective except
for very low inflows and for very low leakage fractions. Synthetic
membranes are best used in conjunction with a low-permeability soil as
 Sensitivity of water balance to landfill design parameters 241

a composite liner. Comparison of designs (b) and (c) demonstrates this

point. Both designs are composed of one synthetic membrane and one
soil liner, but the leakage from the composite liner (design b) is much
lower than the leakage from the double liner system (design c).
It is interesting to compare the single liner performance of design (b)
to the double liner performance of design (d), assuming the soil liner
saturated hydraulic conductivity in design (b) is the same as design (d).
The vertical percolation leaving the system in design (b) is essentially
the same as that leaving the secondary liner in design (d). The
secondary liner in design (d) is nonfunctional since the percolation rate
of the second soil liner is generally equal to or greater than the leakage
rate.

10.4.4 Summary of sensitivity analysis

The interrelationship between variables influencing the hydrologic per-
formance of a landfill cover is complex. It is difficult to isolate one
parameter and exactly predict its effect on the water balance without
first placing restrictions (sometimes severe restrictions) on the values
of the remaining parameters. With this qualification in mind, the fol-
lowing general summary statements are made.
The primary importance of the topsoil depth is in controlling the
extent or existence of overlap between the evaporative depth and the
head in the lateral drainage layer. The greater this overlap, the greater
will be evapotranspiration and runoff. Surface vegetation has a sig-
nificant effect on evapotranspiration from soils with long flow-through
travel times and large plant available water capacities; otherwise,
the effect of vegetation on evapotranspiration is small. The general
influence of surface vegetation on lateral drainage and percolation is
difficult to predict outside the context of an individual cover design.
Clay soils increase runoff and evapotranspiration and decrease lateral
drainage and percolation. Landfills in colder climates and in areas of
lower solar radiation are likely to show less evapotranspiration and
greater lateral drainage and percolation.
The ratio of lateral drainage to percolation is a positive function of
the ratio of KD/KP and the average head above the liner. However, the
average head is a function of lateral drainage, percolation and LIS. The
quantity of lateral drainage, and therefore also the average head, is
in turn a function of the infiltration. Therefore, the ratio of lateral
drainage to percolation increases with increases in infiltration and the
ratio of KD/KP for a given drain and liner design. The ratio of lateral
drainage to percolation for a given ratio of KD/KP increases with
increases in infiltration and the term S/L. The percolation and average
head above the liner is a positive function of the term LIS.
Leakage through a flexible membrane liner increases with the
242 Water balance for landfills

number and size of holes, the depth of water buildup on the liner, the
hydraulic conductivity of the subsoil and the gap between the liner and
the subsoil. Flexible membrane liners reduce leakage through liner
systems by reducing the area of saturated flow through the subsoil.
The overall effectiveness of a flexible membrane liner system is
equivalent to a soil liner having a saturated hydraulic conductivity
equal to the product of the saturated hydraulic conductivity of the
subsoil and the ratio of the reduced area of flow through the subsoil to
the area of the liner. Composite liners provide the best reduction in
leakage. Drain systems which yield low head buildup on the FML
improve the performance of an FML system.

REFERENCES

Brooks, R.H. and Corey, A.T. (1964) Hydraulic properties of porous media,
Hydrology Paper No.3, Colorado State University.
Brown, KW., Thomas, J.c., Lytton, R.L., Jayawikrama, P. and Bahrt, S.c.
(1987) Quantification of leak rates through holes in landfill liners, EPA/6001
S2-87-062, USEPA, Cincinnati, OH.
Campbell, G.S. (1974) A simple method for determining unsaturated hydraulic
conductivity from moisture retention data, Soil Sc., 117(6), 311-4.
EMCON Associates (1975) Sonoma County Solid Waste Stabilization Study. EPA
530-SW-65d.1, US Environmental Protection Agency, Washington DC,
283pp.
EMCON Associates (1983) Field Assessment of Site Closure, Boone County,
Kentucky. EPA 600/S2-83-058, US Environmental Protection Agency,
Cincinatti, Ohio, 6pp.
Fenn, D.G., Hanley, K.J. and DeGeare, T.V. (1975) Use of the water balance
method for predicting leachate generation from solid waste disposal sites, EP A/5301
SW-168, USEPA, Cincinnati, OH.
Giroud, J.P. and Bonaparte, R. (1989a) Leakage through liners constructed with
geomembranes - Part I. Geomembrane liners, Geotextiles and Geomembranes,
8(1), 27-67.
Giroud, J.P. and Bonaparte, R. (1989b) Leakage through liners constructed
with geomembranes - Part II. Composite liners, Geotextiles and Geo-
membranes, 8(2),71-111.
Giroud, J.P., Khatami, A. and Badu-Tweneboah, K (1989) Evaluation of the
rate of leakage through composite liners, Geotextiles and Geomembranes, 8(4),
337-40.
Gordon, M.E., Huebner, P.M. and Kmet, P. (1984) An evaluation of the
performance of four clay-lined landfills in Wisconsin. Seventh Annual Madison
Waste Conference, University of Wisconsin, Madison Extension, Madison,
Wisconsin, 62pp.
Ham, R.K (1980) Decomposition of residential and light commercial solid waste in test
lysimeters. SW-190c, US Environmental Protection Agency, Washington
DC, 103pp.
Horton, R.E. (1919) Rainfall interception, Monthly Weather Rev., US Weather
Bureau, 47(9), 603-23.
 References 243

Kmet, P. (1982) EPA's water balance method - its use and limitations. Wisconsin
DNR, Bureau of Solid Waste Management, Madison, WI.
Knisel, W.G. (ed.) (1980) CREAMS: a field scale model for chemical runoff and
erosion from agricultural management systems, Vols. I, II, and III, USDA-SEA,
AR, Cons. Res. Report 24.
McEnroe, B.M. and Schroeder, P.R (1988) Leachate collection in landfills:
steady case, J. Envir. Eng., ASCE, 114(5), 1052-62.
Perrier, E.R and Gibson, A.C (1980) Hydrologic simulation on solid waste disposal
sites, EPA-SW-868, USEPA, Cincinnati, OH.
Ritchie, J.T. (1972) A model for predicting evaporation from a row crop with
incomplete cover, Water Resources Res., 8(5), 1204-13.
Schroeder, P.R and Gibson, A.c. (1982) Supporting documentation for the
hydrologic simulation model for estimating percolation at solid waste
disposal sites (HSSWDS), draft report, USEP A, Cincinnati, OH.
Schroeder, P.R. and Peyton, R.L. (1990) Evaluation of landfill-liner designs.
J. Environ. Engin. Div., ASCE, 116(3), 421-37.
Schroeder, P.R, Morgan, J.M., Walski, T.M. and Gibson, A.C. (1984a) The
hydrologic evaluation of landfill performance (HELP) model: Vol. I. User's Guide
for Version 1. Technical Resource Document, EPA/530-SW-84-009, USEPA,
Cincinnati, Ohio.
Schroeder, P.R, Gibson, A.C and Smolen, M.D. (1984b) The hydrologic
evaluation of landfill performance (HELP) model: Vol. II. Documentation for
Version 1. Tech. Res. Document, EPA/530-SW-84-01O, USEPA, Cincinnati,
Ohio.
Schroeder, P.R, Peyton, R.L., McEnroe, B.M. and Sjostrom, J.W. (1988a) The
hydrologic evaluation of landfill performance (HELP) model: Volume III.
User's Guide for Version 2, Internal Working Document, USAE
Waterways Experiment Station, Vicksburg, MS.
Schroeder, P.R, McEnroe, B.M., Peyton, R.L. and Sjostrom, J.W. (1988b) The
hydrologic evaluation of landfill performance (HELP) model: Volume IV.
Documentation for Version 2, Internal Working Document, USAE
Waterways Experiment Station, Vicksburg, MS.
Thornthwaite, CW. and Mather, J.R (1957) Instruction and tables for
computing potential evapotranspiration and the water balance, Publication
in Climatology, 10(3), 185-311, Drexel Institute of Technology, Centerton,
NJ.
USDA, Soil Conservation Service (1972) National Engineering Handbook, section
4, Hydrology. US Government Printing Office, Washington, DC, 631pp.
Wigh, RJ. (1984) Landfill research at the Boone County field site. EPA 600/2-84-
050, US Environmental Protection Agency, Cincinatti, Ohio, 116pp.
Williams, J.R., Nicks, A.D. and Arnold, J.G. (1985) Simulator for Water
Resources in Rural Basins, J. Hydraulic Eng., ASCE, 111(6), 970-86.
 CHAPTER 11

Stability of landfills
Issa S. Oweis

11.1 INTRODUCTION

The stability of landfills is controlled in broad terms by the following

factors:
1. the properties of the supporting soil;
2. the strength characteristics and weight of refuse;
3. inclination of the slope;
4. leachate levels and movements within the landfill;
5. type of cover; and
6. cover resistance to erosion.
In all cases, the presence of water acts as a destabilizing agent in
reducing the strength and increasing the destabilizing force.
Assessment of the stability of solid waste landfills is somewhat less
reliable than for soil embankments. The unit weight of refuse and its
strength are difficult to determine and could vary over a wide range.
Assessment of these variables is largely based on case histories and
site-specific investigations (Oweis and Khera, 1990). Because of the
extreme variability in refuse composition, the usual soil sampling and
testing of soils on relatively small samples is not applicable for refuse
such as typical municipal waste.
The purpose of this chapter is to present procedures for stability
evaluation. Potential instability could occur in the foundation soil, the
refuse, or the cover. In all cases the safety margin is expressed in terms
of the factor of safety, F, and is defined as
F = Available strength along the potential failure surface (11.1)
Mobilized strength along the potential failure surface

Geotechnical Practice for Waste Disposal.

Edited by David E. Daniel.
Published in 1993 by Chapman & Hall, London. ISBN 0 412 35170 6
 Types of instability 245

11.2 TYPES OF INSTABILITY

Figures 11.1 and 11.2 illustrate two stability situations. In Fig. 11.1 a
landfill rests on a firm base such as rock or compact sand. The landfill
has a liner and a leachate collection system and no leachate mound is
expected to develop. The refuse mass in this case could slide along the
liner (potential failure surface abc), or if the slopes are steep, the refuse
could fail along a failure surface such as de. Modern landfills are
designed with typical slopes of 3: 1 horizontal to vertical and in a dry
condition, refuse is known to stand almost vertically for cuts as deep as
6m or more. From observations of the Global landfill failure, the steep
chasm in refuse at the top of the failure zone was about 12 m high
(Oweis et ai., 1985). If the leachate collection system fails, a leachate
mound will develop, resulting in a lower factor of safety.
Figure 11.2 depicts an old landfill on soft ground. It is desired to
increase the height of the landfill from level ab to cd. Because no
leachate collection system is present, a leachate mound should be
expected unless a water balance analysis indicates otherwise. In this

a d

~~~,.;::~s:.1~Drainage Pipe Geomembrane LFirm Foundation

- - - Potential Failure Surface

Potential Shape After
-."r.rrTT
Failure.

Fig. 11.1 Stability of a landfill on firm base.

e d

- - - - Potential Failure Surface

- ' - ' - Possible Slape Geometry
After Failure

Fig. 11.2 Stability of a landfill expansion on soft base.

246 Stability of landfills

case, then, the potential failure surface could include the weak founda-
tion of the landfill.
If shear stresses are a large percentage of the strength of soft founda-
tion soils, lateral movements at the toe and beyond could impact cutoff
walls, perimeter leachate collection systems, or surface water drains.
The cover of a landfill could also fail by sliding along a plane parallel
to slope. An initially safe, dry cover may fail at a later time because of
saturation. Saturation increases the weight of the cover and decreases
the frictional resistance along the interface. The source of saturation is
usually precipitation, cracking of the cover and poor drainage, but
saturation could also result from leachate seeps from the side slopes of
the landfill.
In addition failure can occur by slippage between components of a
liner or cover system. Shear along the contact between two surfaces is
often called 'interfacial shear'. The interface between a geomembrane
and geonet, and between a geomembrane and a wet compacted clay,
are especially critical. A failure at the Kettleman Hills facility is an
example of instability associated with interfacial shear (Seed et al.,
1990).

11.3 TOTAL AND EFFECTIVE STRESSES

In soil mechanics, changes in strength and volume are considered to be

due solely to the changes in effective stress. In a saturated soil mass,
the normal stress on any plane is due to the sum of two components:
a) pressure carried by water in the pore space and b) effective stress
carried by the soil skeleton. The total stress is produced by the over-
burden and stresses produced by surface loading. The effective stress
(cr') is the total stress (cr) less the pore water pressure u
cr' = cr - u (11.2)
Because soils contain water in the void space, any volume change
will require either a decrease or increase in porewater pressure. For a
saturated soil to experience compression, water would have to be
expelled from the voids, whereas for the soil to expand the effective
stress would have to decrease and water drawn into the soil. Consider
a deposit of soft saturated clay subjected to loading from a sanitary
landfill. The clay tends to compress over time. Because of the low
hydraulic conductivity of the clay, the rate at which the water is
expelled is much slower than ,the rate at which the load is applied.
This results in excess porewater pressure generated in the clay. For
saturated clay, the excess porewater pressure could be estimated based
on Skempton (1954):
 Total and effective stresses 247

(11.3)
where ()(Yl and ()(Y3 are the additional major and minor principal
stresses, respectively, produced by the loading, ()u is the change
in excess porewater pressure, and B and A are porewater pressure
coefficients. If the soil is saturated, B is essentially equal to 1 for
compressible soils. The parameter A at failure depends on the degree
of overconsolidation and ranges from about -0.5 for highly overcon-
solidated clays to 3 for very sensitive to quick clays. The range for
normally consolidated clay is 0.7-1.3 (Leonards, 1962).
The in situ effective stresses are those responsible for the consolida-
tion of the soil prior to load application. A soil profile with a horizontal
surface is shown in Fig. 11.3. The in situ vertical effective stress (Y~)
and horizontal (Y;') effective stress are also principal stresses, (Yl and
(Y3, respectively, because the horizontal and vertical planes are free
from shear stresses. (YI, could be larger than (Y~ for highly overcon-
solidated clays, and in such cases, the major principal stress would be
(Yj,. The total vertical stress is

(11.4)
where 'Ym and 'Yare total unit weights for zones shown in Fig. 11.3. The
effective vertical stress is
(Y~ = (Y" - uo (11.5)
where Uo is the initial (natural) porewater pressure equal to 'Yuh, where
'Yw is the unit weight of water.
For a soil element along a slope with a slope angle of i with the
horizontal, the vertical and horizontal stresses are no longer principal
stresses. In this case, the normal total, effective, and shear stresses are
respectively:

Ym
... --1L-... _ _

(a) (b)

Fig. 11.3 Total stresses beneath (a) a level, i.e. horizontal ground surface and
(b) within a slope.
248 Stability of landfills

(11.6a)
Ci;, = Ci ll - Yw hw cos 2 i (l1.6b)
T = yh sin i cos i (11.6c)
The effective initial horizontal stress for a sedimentary deposit is
computed based on:
Ci/, = KoCi;, (11.7)
where K o, the earth pressure coefficient at rest, depends on the pro-
perties of the clay and the stress history. For normally consolidated
soils Ko can be estimated based on
Ko = 1 - sin <p' (11.8)
where <p' is the effective angle of shearing resistance (effective friction
angle).
The increases in total principal stresses OCi} and OCi3 are computed
using stress distribution theories. The changes in effective principal
stresses are
OCi} = OCi} - ou (11.9a)
OCi] = OCi3 - OU (11.9b)
In cohesionless soils under normal static loading, the excess porewater
pressure is dissipated as the load is applied, OU is taken as zero, and
the increase in total stress produces an immediate equivalent increase
in effective stress. In clays, the development of excess porewater
pressure ou creates a hydraulic gradient and this excess porewater
pressure dissipates over time. The rate of dissipation depends on the
hydraulic conductivity of the clay and other parameters (Lambe and
Whitman, 1969). After full dissipation of the excess pore water pressure,
the porewater pressure returns to natural porewater pressure, uo. This
process will produce a strength increase.

11.4 FAILURE CRITERION

The revised Coulomb equation (Terzaghi and Peck, 1967) is used to

describe failure conditions in soils. The criterion can be expressed as
follows:
Tf = c' + (Ci;,) tan <p' (11.10)
where Tj = shear resistance (strength) along the plane of failure, c' =
effective cohesion, Ci;, = effective stress normal to the plane of failure,
and <p' is the effective angle of shearing resistance. For initial principal
stresses Ci} and Ci], failure would occur when the effective stress circle
 Failure criterion 249
becomes tangent to the failure envelope as a result of the changes in
the principal stresses generated by loading the soil. In soft, saturated
clay, the effective cohesion component is usually small (Bishop, 1971).
In order to apply effective stress relationships to field conditions,
excess pore pressures generated by loading must be known. If effective
stress parameters are to be used to assess the stability of saturated
clays, then a monitoring program is required to verify the calculated
porewater pressures. However, to avoid the need for field pore pressure
measurements, the stability of cohesive soils may be evaluated by
using the undrained shear strength (<I> = 0 analysis). The undrained
shear strength, 51/! as determined in the triaxial test, is usually taken as
one-half the principal stress difference (01 - (3) at failure:
1
5 11 = 2: (01 - (3)j (11.11)

In the unconfined compression test, 03 is zero and 5 11 is half the uncon-

fined compressive strength, ql/"
An empirical relationship in terms of the plasticity index has been
found to be applicable to a wide range of soft, sedimented clays and
may be used for a rough evaluation of the undrained shear strength as
determined from the unconfined compression or vane tests (Terzaghi
and Peck, 1967). The relationship (Skempton, 1954) is
5 11
-----; = 0.11 + 0.0037 PI (11.13)
0"
The undrained shear strength is sensitive to the type of testing pro-
cedure used, the orientation of the failure planes, and the direction of
shear stresses. The triaxial compression test usually provides undrained
strengths higher than the triaxial extension test where the horizontal
stress is the major principal stress. The direct simple shear test provides
values somewhere in between (Ladd et al., 1977; Mayne, 1988). Mayne
(1988) made a statistical evaluation of published data on many clays
and found that the ratio 5,J0~, varies from 0.25 to 0.55 under isotro-
pically consolidated, undrained triaxial conditions (ClUC) with an
average of 0.35. The range and the average under conditions of simple
shear (055) are (0.15-0.35) and 0.23, respectively.
The choice of the type of the strength test to be used requires
judgement and experience. For reasonably homogeneous clays, the
simple direct shear test is applicable along the base of the failure plane
(e'f' in Fig. 11.2) where the shape of the failure plane is nearly
horizontal. The triaxial extension undrained strength is applicable in
the region of the failure plane beyond the toe of the (H') slope. The
triaxial compression is applicable for the portion of the potential failure
surface in the same quadrant as the slope (ee').
The undrained shear strength of overconsolidated (OC) clays depends
250 Stability of landfills

on the clay type and stress history. The ratio (sulcr~)oc is defined as

(~)
(J~ OC
= (~)
(J~, He
(OCR)1n (11.14)

where OCR is the overconsolidation ratio (Ladd et al., 1977) deter-

mined in the consolidation test and the subscripts oc and lIe indicate
overconsolidated and normally consolidated, respectively. The expo-
nent m typically varies from 0.7 to 0.85, depending on the type of test
used (Mayne, 1988).
The effective stress parameters (cr' and <p') are used for cohesionless
soils under normal static loading conditions. In most projects, the
angle of shearing resistance <p, is estimated based on the soil classifi-
cation and estimated in situ relative density based on field penetration
tests (Gibbs and Holtz, 1957), such as the Standard Penetration Test
(ASTM D1586). Under seismic loading conditions, the same problem is
faced in assessing the developed excess porewater pressure, and con-
sequently, the undrained shear strength for cohesionless soils is used
for evaluating the potential for liquefaction. Liquefaction occurs when
the porewater pressure reduces the effective stress in a sand deposit to
practically zero during ground shaking, leading to loss of strength
which transforms the liquified layer into a viscous fluid. The excess
porewater pressure(> in saturated cohesionless soils develop during
shock or earthquake loading for the same reason as fine grained soils
during a much slower loading rate. The soil tends to compress, and
if the drainage rate is not fast enough, excess pore water pressures
develop. The sulcr~ for cohesionless soils has been correlated with the
Standard Penetration Test N value, the amount of fines in the soil, and
earthquake duration, as characterized by earthquake magnitude (Seed
and De Alba, 1986; Seed, 1987).

11.5 CHARACTERIZATION OF REFUSE

Unlike soils, refuse is not a geologic material and the composition

varies depending on the type of waste generated. Through experience,
however, the mechanical behavior of refuse is expressed in terms of
an apparent frictional parameter <Pa and cohesion Ca' The apparent
cohesion is a frictional parameter characterizing the lateral reinforce-
ment present in the waste, which is analogous to the apparent cohesion
in reinforced earth. The parameters <Pa and Ca are empirical in nature
and may be defined in laboratory tests at an arbitrary strain level that is
considered by the user to be excessive. Alternatively, these parameters
are backcalculated from actual cases of failure or cases where large
deformations in refuse have occurred (Oweis et al., 1985), or conserva-
tively backcalculated from slopes that have not failed, by assuming a
 Characterization of refuse 251
factor of safety of 1.0. The apparent reinforcement effect in refuse
allows the material to withstand shear strains substantially larger
than those for soils at peak strength. This could be significant when
considering refuse fills on weak clays where the factor of safety using
peak strength is marginal. The soil along the potential failure surface
may develop its post peak or residual strength without any obvious
manifestations such as cracks on the top of the slope. For embank-
ments on clays, a factor of safety of 1.5 is usually considered adequate.
~;uch a factor may not be applicable for refuse because refuse properties
could vary over a wider range.
In order to conduct a stability analysis, data on the composition, unit
weight of refuse and leachate mounding are needed.

11.5.1 Refuse composition

The composition of municipal refuse was discussed in Chapter 1. The
unit weight of refuse could be estimated from the composition of
various components, their compactability, and the contribution of daily
cover and moisture adsorption of refuse (Owe is and Khera, 1990).
Table 11.1 shows some published values for the unit weight. It should
be noted that with more recycling, the unit weights will be larger

Table 11.1 Unit weight of municipal refuse

r~unicipal solid waste (M5W) form unit weight

kNlm 3 pcf

MSW, moderate to good compaction 4.7-6.3 30-40

MSW, good to excellent compaction 8.6-9.4 55-60
MSW, baled waste 8.6-14.1 55-90
MSW, active landfill with a leachate mound 6.6 42
MSW, old landfill 9.7 62
MSW, from test pits 8.9-16.2
MSW, after degredation and settlement 9.9-11 63-70
Incinerator residue as received 7.2-12.7 46-81
Incinerator residue (maximum dry density-standard
compaction) 13.5 86
Incinerator residue-old and recent residue (in-place
densities) 14.9-16.6 94.7-105.7
Wood chips (dry) 2.3 14.6
Wood chips (with 64% moisture) 3.8 23.9
Leaf compost as delivered 2.6 16.7
In windows (40% moisture) 5.2 33.3
Final product 3.5 22.2

~;ource: NSWMA (1985), Landva and Clark (1986), Schoenberger and Fungaroli (1971),
I'oran and Ali (1989), Oweis and Khera (1990).
252 Stability of landfills

because the light waste (e.g. paper, aluminum cans, etc.) will be taken
out of the waste stream. A baling operation substantially reduces
the refuse volume and increases the unit weight up to 11 kN/m3 ,
(Tchobanoglous et al., 1977). For submerged refuse, a saturated unit
weight of 10.3-11.1 kN/m 3 could be used. In the absence of site specific
data, the unit weights in Table 11.1 could be assumed for average
municipal waste in communities with minimal recycling.
Old landfills that have undergone substantial settlement are expected
to have larger unit weight. The long-term unit weight could be
estimated knowing the age of the fill and deformation characteristics.
For example, if an old landfill eventually settles 30% of its original
volume by the time it is completed (Oweis and Khera, 1986), and
undergoes additional long term settlement of 10%, the unit weight
of refuse would increase by 67%, which is consistent with reported
values (see Table 11.1). In communities with an active recycling pro-
gram, an approximate estimate could be made for the unit weight
based on the composition of the waste stream (Oweis and Khera,
1990). If, for example, paper products, plastics, and glass are excluded
from the waste stream and replaced by food waste, the unit weight
may increase by about 30% or more, which is significant in the stability
assessment.
Incineration is used in many locations to reduce the volume of
combustible components of municipal refuse which, is often about 70%
of the waste stream. Well-burned residue from a modern incinerator
may weigh about 16kN/m3 (100pd) or more after compaction. For
existing landfills, the unit weight could be established by field density
measurements.

11.5.2 Shear strength of waste

The modified Coloumb failure criterion could be used to characterize
the shear strength of waste. Considering an apparent friction angle <l>a
and apparent cohesion Ca , the shear strength T along a given plane is a
function of the effective stress cr;, normal to that plane. The laboratory
determination of strength is usually given at a strain level of 15-20%
because the samples do not actually collapse. Fang et al. (1977) used the
double punch test for determining the tensile resistance of compacted
refuse, and together with the unconfined compression test, values of Ca
and <l>a of 63 kPa and 19°, respectively, were derived. Drained direct
shear tests on specimens that measured 287 mm by 434 mm reported by
Landva and Clark (1990) indicated the following results.
Shredded refuse: Ca = 23 kPa (480 psf)
Old refuse: Ca = 16 kPa (334 psf)
 Shear along interfaces 253
Artificial refuse: ea = 0 <Pn = 2'r to 41°
Fresh artificial refuse: ea = 0 <Pn = 36°
These results are consistent with the en and <Pn combinations inferred
from the failure of the Global landfill and field testing of refuse
described by Oweis et al. (1985). Triaxial tests on anistopically con-
solidated samples of 2-year-old milled refuse yielded <Pn of 40° (Stoll,
1971), and 15° was recommended to forestall lateral spreading.
Essentially all tests on refuse are affected by the soil content, and
results can vary over a wide range, especially when small samples are
tested. Direct shear tests on 106-mm-diameter samples of mixed refuse
yielded inferred friction angles from 39° to 81° (Siegel et al., 1990).
Refuse with more granular soil typically exhibits higher friction.
The age of the refuse is expected to affect the field strength. As the
refuse ages it becomes denser and hence, stronger. Decomposition,
however, may produce the opposite effect. It is wise, therefore, to
use conservative strength parameters for stability calculations. Another
factor is the expected variation in the character of refuse as the propor-
tion of recyclables becomes less. With more garbage waste, the refuse
would be expected to be heavier and the strength is expected to be less
than current data indicate.
Higher strengths were reported in the literature for incinerator
residue. A friction angle of 45° was reported by Schoenberger and
Fungaroli (1971) on a residue with a density of 15.4kN/m3 using the
direct shear test. The high angularity of the particles was cited as a
probable reason for the high strength. Poran and Ali (1989) used
unconsolidated-undrained triaxial tests to determine the strength of
partially saturated residue with about 20% by weight passing No. 200
sieve (0.074 mm openings) and reported apparent friction angles of 43°
to 45° at a maximum dry density of 13.5 kN 1m3 and an optimum
moisture content of 23.5% (60-70% saturation). It is probable that such
apparently high friction angles are partially due to the development
of negative porewater pressures. If waste will eventually become
saturated in the landfill, tests for stability assessment should be based
on saturated samples. Unless supported by laboratory tests, the angle
of shearing resistance for stability assessment should not exceed an
equivalent granular soil.

11.6 SHEAR ALONG INTERFACES

In modern landfill designs for liners and covers, geomembranes, geo-

textiles, drainage nets and reinforcement may be used in conjunction
with earthen materials. The stability is expressed in terms of the safety
254 Stability of landfills
I

factor against sliding along the interface. The interface friction angle, 0,
depends on the type of the materials on both sides of the interface,
moisture conditions along the interface, resin type, surface texturing,
stiffness of the geotextile or the geomembrane, settlement of the sub-
grade and other factors related to field placement and quality control.
The value is usually measured using a direct shear test that is not
standard. The size of the shear box used varies from 101.6 mm square
(Martin et al., 1984) to 300 mm square (Williams and Houlihan, 1987).
Takasumi et al. (1991) summarize all the testing techniques and shear
boxes that have been used.
The material to be tested is placed on the lower (fixed) portion of the
box and the other material is made to slide over it. The reported value
for a geomembrane on sand ranges from 17° to 36°, whereas for geo-
textiles on sand the reported range is 22° to 40°. The 0 value for
geotextiles on fine-grained soils is in the same order as the angle of the
shearing resistance of the soil. The soil particles tend to be lodged
between the yarns forcing the failure plane above the interface (Williams
and Houlihan, 1987). The reported friction angle of the interface be-
tween geotextiles and geomembranes varied over a very wide range of
6° to 24°, which is roughly the same range as geomembranes over
drainage nets. The problem with possible low values led some designers
to tack weld the drainage net to the geomembrane (Ouplancic et al.,
1987). Seed et al. (1990) carried out interface tests to conduct an analysis
for a landfill slope failure. Based on the tests, they used a residual
friction angle of 8.5° between an HOPE liner and compacted clay, with
the clay at its as-compacted water content. When the interface was
moistened, a residual friction of 8 . 0° was used. For a wet HOPE/clay
interface, an interface shear strength of 43kPa was found to represent a
more critical failure mechanism than a residual friction of 8 . 0°.
Textured geomembranes with a rough surface are now produced and
used. The manufacturers claim an improvement in the interface friction
of 60% to more than 100%. Such improvement is significant, especially
in cover design, where side slopes can be steep.
If the interface frictional characteristics are not adequate to ensure
stability, cover soils can be reinforced with geogrids or high-strength
geotextiles.

11.7 STRENGTH OF COMPACTED COHESIVE SOIL

Compacted cohesive soil is frequently used as a barrier in a liner

or a cover structure. The undrained shear strength is sensitive to the
moisture content and the method of compaction (Fig. 11.4). Clays
compacted dry of the optimum are initially stronger than if compacted
 Methods of stability analysis 255

wet of the optimum. The loss of strength upon soaking, however, can
be substantial.
In terms of total stresses, the undrained strength of the clay liners
and covers is difficult to assess because of the difficulty of predicting
the field degree of saturation. In terms of effective stresses, the strength
is expressed in terms of the effective strength parameters c' and <p'.
The failure envelope is usually curved showing essentially no effective
cohesion at low effective stresses. For a rough estimate, <p' may be
estimated from correlations with the plasticity index (Terzaghi and
Peck, 1967).

11.8 METHODS OF STABILITY ANALYSIS

11.8.1 General
Analysis of slope stability in soils is usually based on limit equilibrium
concepts. At failure, a slip surface is assumed to occur with simul-
taneous mobilization of the shear strength along that surface. The mass
above the slip surface is assumed to move as a rigid body. In waste
disposal sites, the driving force causing the development of a slip
surface is the weight of the waste and leachate. The resisting (restoring)
forces are due to the shear strength of the waste and soil and the
weights of soil, water and waste located near the toe of the slope.
Seismic forces could affect the stability by adding to the disturbing
forces and reduction of shear strength. The techniques for stability
assessment differ in the shape of the slip surface assumed and con-
ditions of equilibrium considered.
The assumptions inherent in the traditional methods of stability
analysis are rarely satisfied. Deformations do occur along the slip
surface and the mass above it prior to failure even in very sensitive clay
(Clausen et al., 1984). It is possible to monitor subsoil movements and
pore pressures to control filling to avoid failure when the factor of
safety is marginal. The assumption of simultaneous mobilization of
shear strength along the entire failure surface is likely to be violated.
Most soils exhibit a peak strength and lesser residual strength at larger
deformations. The failure is generally progressive in nature (Terzaghi
and Peck, 1967). The highly stressed zones reach their peak strength
first while others would be at less than peak. At failure, some zones
are at peak strength mobilization while others may have weakened to
below peak.
The safety of the slope is expressed in terms of the factor of safety F
which is the ratio of the available shear strength to the mobilized
strength given in Eq. (11.1). The factor of safety for the frictional
256 Stability of landfills
Optimum
Dry Of Water Wet Of
-....;;O~Pt;,:im:'cu=m----"C'ont ent Optimum
"0
6~~--~---r~----~~
~
·5(\1
erE I
- - Static
Compaction
- --- Vibratory
\ I
Q) () 51---'=--t---~
0::, Compaction
1/)01
I/) .¥
- - - Kneading

::=:==: I~~ct""
Q).
!:c:
en ·0
......
_en
0-

.~~ I. \~
Q)IO

21----+---+1\~
oQ)
o
... I/)
I. \
t-----t----+-Ij ' -.1.. ___
:::l
.co
o.U Static
~~ I '_ Vibratory
!:
en I - - Kneading
0 12 14 16 18 20 22 24
_ 115 I
u..
:::l I a
~ 113t-----t----+-1~'%+---+---j
a.. I ~
:¥ 1111----+-- 1- 9,. -+---1
~ I ~
.~ 109 ~ ~
~ ~.
9,
o>- 107 -%
\\
o
105~12~~--1~6~~18=--=~~~o~24
Molding Water Content-Percent

Fig. 11.4 Strength of compacted silty clay (Seed et al., 1960).

component of the strength is assumed to be the same as the cohesive

component. Thus
C;" = c'IF (ll.lSa)
tan <P;" = tan <P' IF (ll.lSb)
where c;" and <p;" are the mobilized cohesion and friction angle,
respectively.
The procedure for stability analysis generally involves the following
steps.
1. Assess the refuse and soil properties.
2. Assess leachate levels.
3. Determine the type of analysis required.
For waste facilities on soft, saturated clay or moderately over-
consolidated clay, a short-term condition may be appropriate for
analysis. In this case, the undrained shear strength of the deposit is
 Methods of stability analysis 257

390 80 70 60 50 40 30 20 10 if
Values Of Slope Angle, /I

Fig. 11.5 Taylor stability chart for c, <I> material (Terzaghi and Peck, 1967).

used and total stresses are considered. For waste facilities on heavily
overconsolidated clay, or where cuts are made in such deposits,
an effective stress analysis using long-term porewater pressures
(assumed to be the same as ambient) may be more critical.
4. Select the geometry of the slip surface.
5. Compute the factors of safety for varied dimensions of the slip
surface. The minimum value is the factor of safety for the case
investigated.
The shape of the slip surface could be curved, planar or a combi-
nation of both. For homogeneous material an arc of a circle has been
found to be satisfactory. A planar type slip surface is useful for
analysis of soil covers where the slip surface is parallel to the
slope or, in other cases, where planar surfaces are dictated by site
conditions.

11.8.2 Stability charts

For simple slopes composed of homogenous materials and no seepage
surface, the Taylor charts (Terzaghi and Peck, 1967) could be used for
estimating the factor of safety. The chart in Fig. 11.5 is applicable for a
constant (c, <\» material within the slope and the base. All the critical
circles for friction angles more than about 3° are toe circles. The chart,
therefore, could be used for refuse embankments on firm bases. The
chart is based on full mobilization of the frictional component of
strength and the cohesion value is the mobilized cohesion Cm. In order
to apply the factor of safety to both c, and <\>, an iterative procedure is
used, as illustrated by the following example.
258 Stability of landfills

II r---.-,-----,-,---,--r--..----.--,

~UIOr--+--+--+--1-~---r~
II
en
Z 9r--+--+--+--4-~--~
.2u
~ 8r--+--+--+--4-~---+
~
~ 7r-_+--+-_+--1_~­
c
u;
~ 6r--+--+--+~~
l-
een
Q)

'"
~

Values Of Slope Angle,,e

Fig. 11.6 Taylor stability chart for <p = 0 analysis (Terzaghi and Peck, 1967).

Example 1
Estimate the factor of safety for a refuse embankment 40 feet high
founded on stiff clay. The side slopes are 1 vertical on 1 horizontal, so
the slope angle = 'tan- 1 1.0 = 45°. The unit weight is 7.85kN/m3
(50 pef), C = 14.4 kPa (300 psf), and <I> = 20°.
For trial 1 assume F = 2.0. Then using Eq. 11.15b
<1>", = tan- 1 (tan 20°/2) = 10.3°
From Fig. 11.5, n = 9.2 so Cm = (50)(40)/9.2 = 217.4psf, and F =
300/217.4 = 1.38.
For trial 2 assume F = (1.38 + 2)/2 = 1.69. Then
<1>", = tan- 1 (tan 20°11.69) = 12.15°
From Fig. 11.5, Ns = 10.55 so Cm = (50)(40)110.55 = 189.6, and F =
(300/189.5) = 1.58.
For trial 3 assume F = (1.58 + 1.69)/2 = 1.635. Then
<l>m = tan- 1 (tan 20°11.635) = 12.6°
with Ns = 10.82, so that c", = (50)(40)/10.82 = 184.8 psf, and F =
300/184.8 = 1.62.
The Taylor chart for <I> = 0 is shown in Fig. 11.6 and is applicable
when the slope and the base have uniform cohesion c. In order to
use the chart for preliminary estimates, it is necessary to estimate an
equivalent cohesion value along the critical circular arc. The position of
the critical circle may be estimated using Fig. 11.7. The center of the
 Methods of stability analysis 259
5

A ..,,1/
-g4 I 7~"'-
c::
... I
I
--,/
v' <:-y
~ 3 I
r1 - - - .-- <:-~
..' -
/'

l 1---- ---1-- .......... ,/';f!L

-
~ 0
.c

,/'"1
0.2
Q)
o ,------ ~-- I--~

' ......... B 1----

-0---

160° 50° 40° 30° 20° 10° 0°

Values Of fJ

Firm Base

Fig. 11.7 Position of the critical failure surface for <p = 0 material (Terzaghi
and Peck, 1967).

circle could be assumed to be along a vertical line through the mid

point of the slope. After the critical circle is drawn, the equivalent
cohesion of the refuse and frictional layers may be estimated as
follows.
1. Find the average inclination a of the slip surface through the layer
of interest. For a layer having a friction angle <1>, compute the
equivalent cohesion, Ceq' as
Ceq = (W' cos a) tan <I>/L (11.16)
where W' = effective weight of soil and refuse (i.e., buoyant weight
under water) above length L, L = approximate length of the slip
surface through the layer, and a = average inclination of the slip
surface through the layer.
2. Compute the weighted average of cohesion Cav for the entire
slip surface based on the cohesion and slip surface arc length in
each layer.
260 Stability of landfills

Trial Circular
Failure Surface

Fig. 11.8 The method of slices.

11.8.3 The method of slices

This designation is applied to a variety of methods where the potential
sliding mass is divided into slices, as illustrated in Fig. 11.8. The
equilibrium of each slice, with its base on an assumed failure surface, is
considered. The differences between various methods are the shape of
the failure surface and the number of equilibrium conditions satisfied.
In the ordinary method of slices (also known as the Swedish method or
the Fellenius method), the interslice forces are assumed zero. The
factor of safety is defined as the ratio of resisting moment around the
center of a circular arc to the driving moment.
F = Ic(of) + (W cos a - u(of)) tan <p
(11.17)
IW sin a
where of is the length of the base of the slice, W is the total weight of
the slice, a is the inclination of the base of the slice, u is the average
porewater pressure acting on the base of the slice, and I indicates
summation over the total number of slices.
In the modified Bishop method, the factor of safety is defined as the
ratio of the shear strength to the mobilized shear strength. In order to
calculate the factor of safety, the vertical equilibrium of each slice is
considered and side forces are set to zero. The resulting expressions
are as follows.
 Methods of stability analysis 261

h
.8,
----T+-L--+-----r-----==~';7 ... ~

Fig. 11.9 The translational wedge method.

F = LC(b[) cos a + (W - U(b[) cos a) tan <PIma

(11.18)
LW sin a

ma = (1 +
tan a tan
F
<p) cos a (11.19)

The factor of safety appears on both sides of the equation for F. A trial
and error procedure is used to achieve a solution. The angle a is
positive when the inclination of the base of the slice is in the same
quadrant as the ground surface.
A number of computer programs are available to obtain solutions
with various methods of slices. The programs will compute the factor
of safety for many failure surfaces and locate the surface that gives the
lowest factor of safety.

11.8.4 The translational wedge method

This method is described in the Naval Facilities Engineering Command
Design Manual DM7 (1982). The potential failure surface is approxi-
mated by planar segments, as shown in Fig. 11.9. The mass of soil or
refuse above each segment forms a potential sliding wedge. At failure,
the inter wedge force is assumed horizontal (i.e., smooth vertical
planes) and equivalent to the active force calculated by considering the
262 Stability of landfills

equilibrium of the wedge. In general, where the factor of safety is other

than 1, the inter-wedge force is calculated using the mobilized cohesion
elF and mobilized friction (tan <I>/F). This active condition prevails
where the inclination of the failure planes is in the same quadrant as
the slope (zone abcde in Fig. 11.9). For the a wedges, the failure plane
through each layer is assumed to be at an angle a = (45 + <1>/2) unless
the angle is dictated by stratigraphy (e.g. a4 in Fig. 11.9). Where for a
particular wedge a is small (e.g. a4 in Fig. 11.9), the a force may have
a negative sign indicating a net resisting force.
For planar surfaces at angles that are not in the same quadrant as the
slope, the wedges sliding along those planes are passive wedges. At a
factor of safety of 1.0, the inter-wedge forces are assumed horizontal
and equivalent to the passive pressure force using c, and <I> for the
material. In general, this force (~ force) is computed using the mobilized
friction and cohesion. The angle ~ is assumed to be equal to 45 - <1>12
through each layer.
In order to perform the analysis, the equilibrium of the central
wedge is considered. The a forces are produced by the sliding mass on
each plane (a forces P u). On the passive side, the ~ forces are com-
puted. For the equilibrium of the central wedge, the following must be
satisfied.
(11.20)
For each active wedge
Pa = (W - c",L sin a - P", cos a)Fu - (c",L cos a - P", sin a) (11.21)
For each passive wedge
P ll = (W + c",L sin ~ - P", cos ~)FJl + (c",L cos ~ + Pw sin~) (11.22)
where

F _ _ta_n_a_---'.[-'-(ta_n---,-,<I>)'--../F-".]
(11.23)
u - 1 + (tan a tan <I>/F)
F _ tan ~ + [(tan <I»/F]
(11.24)
Jl - 1 - (tan ~ tan <I»/F

The functions Fa and FJl are shown in Figures 11.10 and 11.11,
respectively.

11.8.5 Stability of the cover

The stability of the cover could be assessed based on the translational
wedge method. Figure 11.12 shows the elements of the method as
applied to the cover system. The interface strength is characterized by
 Methods of stability analysis 263
4.0

3.5 0.100

3.0 J
V 0.200

V
2.5

/
"
iy ~V'v:
ir 2.0 0.300
(tanq,)/F
0.400
1.5

~ ~ ~ ~ ~ 0.800
0.500
0.600
1.0 ./ 0.700

0.5 k ~ ::::::
~ r::;...- ~ ~ ~ ~ ~ ~,
0.900
~.OOO

0
~~~ ~~~~~"'"
o 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80
a (degrees)

Fig. 11.10 The function Fu'

1.000 ~Ol .800.700

0.0

9.0
I I
I
II
8.0

7 I
R
~ / / // /
5 .•
(tanq,)IF_
J II /
40 // I / / /
~ ~ '/ / /
~ :8 v::
3.~

"/ ~ ~
~
,,/ ,,/ .,..300
?O

1.0
o
r-lalr::=::6=:::::~
~~~~_ : : : _r:---::: - ~~:--:::..:~ .200

I
-.100

o
I I I
4 8 12 16 20 24 28 ~ 12 36 40 44 48
~(degrees)

Fig. 11.11 The function F~.

Ciand Oi. Cover reinforcement per unit width is indicated by force T.

The potential failure surface is abed. Considering the horizontal
equilibrium of the central wedge, the stability equation is written as
(11.24)
264 Stability of landfills

Fig. 11.12 Stability of the cover.

where
Pa = [Wa - (elF)La sin a - PWtl cos a]F'l
- [(elF)La cos a - Pwa sin a] (11.25)
Pe = [(We - ((c;lF)Le + T) sin is - Pwe cos is)]Fe
- [((cilF)Lc + T) cos is - Pwe sin is] (11.26)
PI' = [WI' + (elF) sin ~ - PlOP cos ~]FI3
+ [(elF)LI' cos ~ + Pwp sin~] (11.27)
where
Fa = (F tan a - tan ~)/(F + tan a tan ~) (11.28)
Fe = (F tan is - tan o;)/(F + tan is tan 0;) (11.29)
F[3 = (F tan ~ + tan ~)/(F - tan ~ tan ~) (11.30)
The angles a and ~ are chosen to maximize Pa and minimize P[3. The
angles could be computed by considering the equilibrium of active and
passive wedges and a factor of safety of 1.0. The resulting expressions
are
tan a = tan ~ + VI + tan 2~ - tan ie/sin ~ cos ~ (11.31a)
tan ~ = -tan ~ + VI + tan 2 ~ - tan iclsin ~ cos ~ (11.31b)
for ie = 0, a = 45 + ~/2, and ~ = 45 - ~/2.
 Methods of stability analysis 265

For long slopes, P" and Pf3 become relatively small and the equilibrium
condition reduces to Pc = 0.0. For a cover of uniform thickness h, and a
seepage surface parallel to the slope at depth h", above the interface,
the stability equation reduces to
F = (111 - t) [(c/yh) (lIsin i, cos is)
+ (tan 8Jtan is) (1 - y".hw/yh)] (11.32)
where
(11.33)

11.8.6 Seismic considerations

Ground motion induced by earthquakes add to the destabilizing force
and hence reduce the factor of safety. During ground shaking, the
factor of safety could be reduced to 1.0 for a very short time before a
reversal in the direction of the seismic force or the end of the signifi-
cant duration of ground motion. Newmark (1965) evaluated the seismic
effects on embankments in terms of deformation. The elements of the
Newmark method are further explained by Franklin and Griffin (1981),
where the method was applied using earthquake records. Using
simplifying assumptions and assuming downhill motion only, the
slope movement could be approximated by (Newmark, 1965) as

U = 2~: ~(1 - ~) (11.34)

where U = displacement, V = ground velocity, A = ground accelera-

tion, and N = acceleration required to produce a factor of safety of 1.0
(i.e., resistance factor).
The parameter N could be estimated by a pseudo-static analysis
where a horizontal static force is substituted for the dynamic force
(Sarma, 1975; Chang et al., 1984). The ground velocity could be
estimated from ground response analyses or available correlations
(Krinitzsky and Chang, 1988; Krinitzsky et aI., 1988).
Surveys of landfills in California after earthquakes revealed cracks
that may have been generated prior to earthquake occurrence (Siegel
et aI., 1990). The OIl landfill in Southern California, about 24 m high
and founded on rock was monitored to assess earthquake effects
(Siegel et aI., 1990). The Pasadena earthquake (Dec. 3, 1988, M = 5.0.
epicenter located 15 km from the landfill) produced a horizontal
acceleration of 0.22 g at the base and 0.1 g at the top of the landfill.
Amplification occurs at frequencies less than 2 Hz, which constituted a
small segment of the frequency content. Attenuation occurred for fre-
quencies greater than 2 Hz. The Malibu earthquake (January 18, 1989,
M = 5, epicenter located 50 km from the landfill) produced about 0.01 g
266 Stability of landfills

at the base of the landfill and top of the landfill. Frequencies lower
than 3 Hz amplified while attenuation occurred for the above 3 Hz
frequency content. Considering a predominant frequency of 2 Hz, the
average shear wave velocity of the landfill would be 195 mls (640 ftls)
(uniform layer over rigid base). Sharma et al. (1990) reported a velocity
of 200 mls (650 ftls) for a landfill in Richmond, California, from down-
hole shear wave velocity measurements.
There are no case histories of failure or damage to landfills due to
liquefaction of saturated refuse. Refuse in landfills with a leachate
mound could conceivably experience a loss of strength. The hydraulic
conductivity of typical municipal refuse is in the order of 10- 3 cmls
(Oweis et aI., 1990) and under strong seismic shaking excess pore
pressures may develop. Because the undrained shear strength for
refuse is usually not known, it is prudent to apply a larger than usual
factor of safety on the static strengths for stability assessment of refuse
embankment with leachate mounds in seismically active areas.

REFERENCES

Bishop, A. W. (1971) Shear strength parameters for undisturbed and remolded

soil specimens, in Stress Strain Behavior of Soils (ed. RH.G. Parry), G.T.
Foulis & Co., Ltd., Henley-on-Thames, Oxfordshire, pp. 3-58.
Chang, e.-J., Chen, W.F. and Yao, T.P. (1984) Seismic displacement in slopes
by limit analysis, f. Geotech. Eng., ASCE, 110(7), 860-74.
Clausen, e.-J.F., Graham, J. and Wood, D.M. (1984) Yielding of soft clay at
Mastemyr, Norway, Geotechnique, 34(4), 581-600.
Collins, RJ. (1978) Highway construction use of incinerator residue, Geotech.
Pract. for Disposal of Solid Waste Materials, ASCE, New York, 246-66.
Duplancic, N., Dayal, U. and Colella, J.e. (1987) Hazardous waste landfill cap
system stability, Geotech. Pract. for Waste Disposal 'S7, Geotechnical Special
Publication 13 (ed Richard D. Woods), ASCE, New York, 432-46.
Fang, H.Y., Slutter, RJ. and Koerner, R.M. (1977) Load bearing capacity of
compacted waste materials, in Proc., Specialty Session on Geotech. Eng. and
Environ. Control, 9th Int. Conf. on Soil Mechanics and Foundation Eng., Tokyo,
IV/2,265-78.
Franklin, A.G. and Griffin, M.E.H. (1981) Dynamic analysiS of embankment
sections, Richard B. Russell dam, in Earthquakes and Earthquake Engineering
- Eastern United States (ed. James E. Beavers), Ann Arbor Science, Ann
Arbor, Michigan, pp. 623-42.
Gibbs, H.J. and Holtz, W.G. (1957) Research in determining the density of
sands by spoon penetration testing, Proc., 4th Int. Conf. on Soil Mechanics
and Foundation Engineering, London, England, 35-9.
Krinitzsky, E.L. and Chang, F.K. (1988) Intensity related earthquake ground
motion, Bulletin of the Assoc. of Eng. Geol., XXV(4), 425-35.
Krinitzsky, E.L., Chang, F.K. and Nuttli, O.W. (1988) Magnitude-related
earthquake ground motion, Bulletin of the Assoc. of Eng. Geol., XXV(4),
399-423.
 References 267
Ladd, c.c., Foott, R., Ishihara, K., Schlosser, F. and Poulos, H.G. (1977)
Stress-deformation and strength characteristics, Proc., 9th Int. Conj. on Soil
Mechanics and Foundation Eng., 421-94.
Lambe, T.W. and Whitman, R.V. (1969) Soil Mechanics, John Wiley & Sons,
Inc., New York.
Landva, A.O. and Clark, U.J. (1986) Geotechnical testing of waste fill, Proc.
39th Canadian Geotech. Conj., Ottawa, Canada, 371-85.
Landva, A.O. and Clark, J.I. (1990) Geotechnics of waste Jill - theory and practice,
ASTM STP 1070 (eds A.O. Landva and G.D. Knowles).
Leonards, G.A. (1962) Engineering properties of soils, Foundation Eng. (ed.
G.A. Leonards), McGraw Hill, New York, p. 213.
Martin, J.P., Koerner, R.M. and Whitty, J.E. (1984) Experimental friction
evaluation of slippage between geomembranes, geotextiles and soils, Proc.,
Int. Conj. on Geomembranes, Denver, CO, Industrial Fabrics Association
International, St. Paul, Minnesota, 191-6.
Mayne, P.W. (1988) Determining OCR in clays from laboratory strength,
/. Geotech. Eng., 114(1), 76-92.
National Solid Waste Management Association (1985) Basic data: solid waste
amounts, composition and management systems, National Solid Waste
Management Assoc. Technical Bulletin No. 85-6, Dublin, Ohio.
Naval Facilities Engineering Command (1982) Design Manual 7.1, Soil
Mechanics, Department of the Navy, Alexandria, VA.
Newmark, N.M. (1965) Effects of earthquakes on dams and embankments,
Geotechnique, 15(2), 139-60.
Oweis, I. and Khera, R. (1986) Criteria for geotechnical construction on
sanitary landfills, Environ. Geotech. (ed. H.Y. Fang), Envo Publishing Co.,
205-22.
Oweis, I. and Khera, R. (1990) Geotechnology of Waste Management, Butterworth
Publishers, London, England.
Oweis, I., Mills, W., Leung, A. and Scarino, J. (1985) Stability of sanitary
landfills, Geotechnical Aspects of Waste Management, Foundation and Soil
Mechanics Group, Metropolitan Section, ASCE, New York.
Oweis, I., Smith, D.A., Ellwood, R.B. and Green, D.S. (1990) Hydraulic
characteristics of refuse, /. Geotech. Eng., ASCE, 116(4), 539-53.
Poran, c.J. and Ali, F.A. (1989) Properties of solid waste incinerator fly ash,
/. Geotech. Eng., ASCE, 115(5), 1118-33.
Sarma, S.K. (1975) Seismic stability of earth dams and embankments,
Geotechnique, 25(4), 743-6l.
Schoenberger, R.J. and Fungaroli, A.A. (1971) Incinerator residue - fill site
investigation, /. SMFD, ASCE, 97(10), 1431-43.
Seed, H.B. (1987) Design problems in soil liquefaction. /. Geotech. Engin.,
ASCE, 113(8), 827-45.
Seed, H.B. and De Alba, P. (1986) Use of SPT and CPT tests for evaluating the
liquefaction resistance of sands, in Use of In Situ Tests in Geotechnical
Engineering (ed Samuel P. Clemence), Geotechnical special publication no.
6, ASCE, New York, pp. 281-302.
Seed, R.B., Mitchell, J.K. and Seed, H.B. (1990) Kettleman Hills waste landfill
slope failure II: stability analyses, /. Geotech. Eng., 116(4), 669-90.
Seed, H.B., Mitchell, J.K., and Chan, C.K. (1960) The strength of compacted
cohesive soils, Proc. Res. Conj. on Shear Strength of Cohesive Soils, ASCE,
New York, 897-964.
Sharma, D.H., Dukes, M.T. and Olsen, D.M. (1990) Field measurements of
dynamic moduli and Poisson's ratios of refuse and underlying soils at a
268 Stability of landfills

landfill site, Geotechnics of Waste Fills - Theory and Practice, ASTM STP 1070
(eds A.O. Landva and G.D. Knowles), Am. Soc. for Testing and Materials.
Siegel, A.R., Robertson, J.R. and Anderson, D.G. (1990) Slope stability
investigation at a landfill in Southern California, Geotechnics of Waste Fills -
Theory and Practice, ASTM STP 1070 (eds A.O. Landva and G.D. Knowles),
Am. Soc. for Testing and Materials.
Skempton, A.W. (1954) The pore-pressure coefficients A and B. Geotechnique, 4,
143-7.
Stoll, O.W. (1971) Mechanical properties of milled refuse, ASCE National Water
Resources Engineering Meeting, Phoenix, Arizona, January 11-15.
Takasumi, D.L., Green, K.R. and Holtz, R.D. (1991) Soil geosynthetics
interface strength characteristics: a review of state-of-the-art testing
procedures, Geosynthetics '91, Industrial Fabrics Assoc. Int., St. Paul,
Minnesota, 1, 87-100.
Tchobanoglous, G., Theisen, H. and Eliassen, R. (1977) Solid Wastes,
Engineering Principles and Management Issues, McGraw Hill, New York, p.
209.
Terzaghi, K. and Peck, R.B. (1967) Soil Mechanics in Engineering Practice, John
Wiley and Sons, Inc., New York.
Williams, N.D. and Houlihan, M.F. (1987) Evaluation of interface friction
properties between geosynthetics and soils, Geosynthetics '87, Industrial
Fabrics Assoc. Int., St. Paul, Minnesota, 2, 616-27.
 CHAPTER 12

Mine waste disposal

Dirk Van Zyl

12.1 INTRODUCTION

Mine wastes are a broad group of waste materials resulting from the
extraction of metals and non-metals. The wastes include solid, as well
as liquid waste, and can be inert or can contain hazardous constituents.
In general, mine waste consists of high volume, low toxicity wastes
(EPA, 1985).
It was estimated in 1982 that about 500 and 30 million tonnes of
waste were produced in the US from copper and molybdenum mines,
respectively. The main reason for the high volumes of mine waste is
the very low concentration of metals contained in ores. For example,
gold mining is presently conducted using heap leach technology for
ores containing as little as 0.5 g per tonne of gold. It is, therefore, clear
that virtually every tonne of rock which is mined remains as waste
after extraction of the gold. In open-pit mining, there is also a con-
siderable volume of overburden and waste rock produced which
contains no mineralization or levels of mineralization too low to make
extraction economical. The term stripping ratio in open pit mining
refers to the tonnes of waste stripped (discarded as waste rock) as a
ratio of the tonnes of ore mined. Stripping ratios in the order of 2 to 6
are not unusual. Consider the case of an open-pit copper mine where
66 000 tonnes of copper ore is mined per day and the average grade of
the ore is 1.5% (i.e., one tonne of ore contains 1.5kg of copper) and
the stripping ratio is 3. In this case, 200 000 tonnes of waste rock is
discarded daily, while 65 000 tonnes of other waste (typically tailings)
remain after extraction of the copper. In a year with 350 days of
production, 69 000 000 tonnes of waste rock and 22 750 000 tonnes of
tailings are produced.

Geotechnical Practice for Waste Disposal.

Edited by David E. Daniel.
Published in 1993 by Chapman & Hall, London. ISBN 0 412 35170 6
270 Mine waste disposal

Mine wastes remain after the extraction of materials which are sold
on the commodity markets. The markets are not controlled by the
producers but are controlled by the free market system of supply and
demand. The extra cost of mine waste disposal can, therefore, not be
passed on to the consumer but must be absorbed within the economic
constraints of the project. This aspect, combined with an appreciation
of the high volume and low toxicity characteristics of the waste are
important in considering the overall waste disposal strategies for mine
waste.
The disposal of mine waste represents capital and operating cost to a
mining operation. For this reason, mine waste has historically been
disposed of at the lowest cost. This often resulted in considerable
environmental impacts. For example, tailings containing hazardous
constituents were directly discharged onto the land without contain-
ment, or in some cases, directly into rivers, while waste rock was
dumped as close as possible to the mine adit or shaft without con-
sideration of environmental impacts. Some of these and other historical
impacts of mine waste on the environment have been surveyed by
Smith (1987).
Some of the historical mine waste disposal practices led to catastro-
phic failures. While not all such failures have been well documented
some of the recent failures have been evaluated. Three such failures
which caused a considerable number of deaths were the Buffalo Creek,
Aberfan and Chilean tailings impoundments (Dobry and Alvarez, 1967;
Bishop et al., 1969; Wahler and Assoc, 1973). These failures focused
considerable attention on the geotechnical aspects of designing safe
mine waste disposal facilities. This attention was the impetus for
developing the science related to geotechnically sound mine waste
disposal facilities over the last three decades (NBRI, 1959; U.S. Dept. of
Interior, 1975; ASCE, 1979; Vick 1983; Van Zyl and Vick, 1988).
Attention has been focused over the last decade on the environ-
mental considerations related to mine waste disposal. Activity in this
area has recently increased many fold (Van Zyl, 1985; University of
California at Berkeley, 1988; Lootens et ai., 1991). Much has been
learned about the environmental impacts of mine waste; however, it
is an area for potential fruitful future research contributions. In the
early 1980s, cyanide use (mostly for the extraction of gold and silver)
attracted considerable attention and the disposal of such wastes was
considered a big environmental risk. Research and practical experience
since that time have shown that cyanide waste can be disposed of
safely and that cyanide can best be considered a 'transient pollutant'.
Of much greater concern is the disposal and long-term effects of acid-
generating wastes, also referred to as acid mine drainage (AMD) or
acid rock drainage (ARD), (Steffen, Robertson and Kirsten, 1989). This
area of research has attracted much attention and practical methods are
 Waste rock 271

available, while others are being developed, to safely dispose of acid-

generating wastes or to treat effluents from such facilities.
Chemical characteristics of tailings solids and solutions are deter-
mined by the characteristics of the ore, as well as the chemicals used
during extraction of the metals. It is beyond the scope of this book
to discuss chemical characteristics of mine waste. However, it is an
important topic, which in most cases, determines the final design of a
mine waste facility.
This chapter presents a brief review of mine waste disposal practices
for various types of mine waste. It is not intended to be exhaustive, as
this topic could fill a multi-volume text. The references will concentrate
on recently published summary documents instead of an extensive
list of individual papers, except where such papers present specific
information.

12.2 WASTE ROCK

Waste rock consists of unmineralized rock, as well as rock containing

mineralization which is too low to extract economically with existing
technologies. Materials overlying the orebody, often referred to as
overburden, are also usually included in the waste rock category. It
is obvious that because waste rock can contain some mineralization,
today's waste may become tomorrow's ore. This fact has been proven
many times and especially so more recently with the development of
low cost hydrometallurgical processes for extraction of gold (such as
heap leaching). Many waste rock dumps are presently being reworked
as ore.
The characteristics of waste rock depend on the characteristics of the
host rock, as well as the rock surrounding the ore body. Waste rock is
very often competent, durable, hard rock consisting of particle sizes
from 2m diameter to silt and clay size. However, waste rock can also
consist of very weak materials, such as weathered volcanics or sedi-
ments, which can produce considerable stability concerns. A recent
summary of waste rock disposal practice was edited by McCarter (1985).
Waste rock and overburden may be disposed of in valley fills, side-
hill dumps, or open piles. The type of waste rock disposal facility is a
function of the volume of waste, site topography and drainage, and
haul distance from the mine. Unless located in an arid climate, these
waste management units will have flood runoff protection ponds for
purposes of sediment control, and surface runoff diversion ditches to
minimize natural runoff intruding on the pile. Even in arid climates,
there may be a need to install flood diversions to deal with flash flood
events.
Disposal of waste rock and overburden takes place by end-dumping
272 Mine waste disposal

from trucks, or stacking using conveyors. The material is dumped over

the face of a lift. The lift may be the full height of the dump, i.e., up to
100 m or more, or 15- to 30-m lifts with intermediate benches. The
working face of a lift is at the angle of repose of the material. As
the dumped material cascades down the face, segregation of particles
occurs. The large boulders roll to the bottom and the gravels, sands,
and fines remain near the top. The higher the working face of a lift, the
more prominent the segregation. As a result, the bottom of disposal
facilities is frequently free-draining rock, while the upper elevations
and working surfaces are less permeable.
The upper surfaces of the pile are usually flat. They are gently sloped
away from the dump face toward the natural terrain to enhance dump
truck safety during operation, and also to prevent surface runoff,
during storm events, from cascading over the edge of the piles.
Valley fills usually are started at the upstream end of the valley.
Dumping progresses downstream, with increasing lift thickness. Side-
hill piles are formed by dumping along hillsides or valley slopes so
that the toe of the dump does not cross the drainage. Dumping is
continued along the natural contour and away from the slope. Surfaces
are graded to keep runoff away from the exposed face of the dump. An
open pile may be constructed over relatively flat topography, or as a
ridgeline fill, and can be extended to form a sidehill dump.
Reclamation of waste rock dumps can be done through a number of
approaches. Dump slopes can be armored with competent, durable
hard rock which resembles scree slopes. In this case, these dumps will
typically be left at their natural angle of repose which could be on the
order of 37 degrees. The slopes of waste rock dumps can also be
smoothed with dozers to shallower slope angles than the angle of
repose, for example 2H: 1V. There is considerable controversy about
the reclamation of such slopes through the establishment of vegeta-
tion. One school of thought is that vegetation can not be established
on slopes steeper than 3H: 1V; however, examples exist of successful
revegetation of slopes at 2H: 1V. It is important to control the length of
the slope to reduce the formation of gully erosion.
Major concerns with waste dump construction include operational
stability, as well as long-term stability. Weathering of waste rock fol-
lowing deposition can lead to a change in rock shear strength char-
acteristics and subsequent failure of the structure. Acid rock drainage
(ARD) from waste dumps is a specific concern which must be addressed
at the time of design. ARD can lead to significant environmental
damage if not controlled. ARD from waste rock disposal facilities is by
far the biggest environmental concern in mine waste disposal. This
topic is outside the scope of this book. However, it must be given full
attention during the design of a waste rock disposal facility.
The geotechnical characteristics of waste rock range from that of
 Waste rock 273
rockfill to that of soil. Uhle (1988) summarized and statistically evalu-
ated most of the rockfill data published to date. The shear strength
characteristics of rockfill are typical of granular materials. However,
waste rock demonstrates a much more pronounced curved Mohr
envelope than is typical of sands. The shear strength, under low
normal stress, is therefore, higher than a linear failure envelope will
indicate.
Waste rock disposal facilities for new mines are designed on the
basis of information obtained during the geological exploration. Core
samples of the ore and waste will be available for mineralogical,
metallurgical and geotechnical evaluations. There are very often only
small quantities of core available and it is therefore difficult to perform
sufficient geotechnical testing to confirm the expected shear strength
of the waste. The design engineer must rely on experience in the
selection of design parameters. Engineering geological description of
the core during the exploration program is a very useful tool in
the characterization of the waste. Information such as rock hardness,
mineralogy, rock quality designation (RQD), fracture spacing, and pre-
sence of weathered products are necessary to select appropriate design
parameters .
. It is difficult, if not impossible, to measure the shear strength of
waste rock in the laboratory due to the large sizes of the particles
disposed during operations. Direct shear and triaxial testing of smaller
particles may lead to the selection of overly conservative shear strength
values. Point load testing of intact rock particles, such as cores, can be
used to estimate the shear strength of the waste rock. The following
empirical relationship suggested by Hoek and Bray (1981) can be used
to estimate the shear strength:
t = AoAcr/crc - T)B (12.1)
where: A, and B are the constants defining the shape of the Mohr
failure envelope; the uniaxial compressive strength, crc is given by cr c =
2415 , where Is is the point load strength; and T = !(m - \1m 2 + 4s),
where m and s are dimensionless constants depending on particle
shape.
Waste rock disposal facilities must be designed to be well drained.
Coarse particles deposited at the bottom of the waste rock dump as a
result of natural segregation can form an adequate drain in many
instances. However, if springs are present on the site then specific
underdrains must be installed to intercept these springs and drain
them under controlled conditions. Underdrains consisting of large rock
particles wrapped in geotextile are usually specified. These drains can
also include perforated pipes, however it is good practice to design the
rock drain with sufficient capacity to carry the expected flow.
Since waste rock disposal facilities are constructed by end dumping,
274 Mine waste disposal

they are, therefore, deposited in a relatively loose state. They are

subject to settlement under selfweight, as well as settlement when
structures are placed on them. There are examples of mining structures
being constructed on waste rock disposal facilities. In this case the
waste rock disposal facility must be constructed as an engineered fill.
Placement of the material in 1- to 2-m-thick lifts compacted by multiple
passes of construction traffic will usually result in an adequately com-
pacted structure.
In summary, it can be stated that the geotechnical concerns of waste
rock disposal facilities are related to characterization of the materials in
terms of shear strength and the construction of these facilities for
maximum economy and reliable short and long term stability. Experi-
ence plays a large role in the design of waste rock disposal facilities.

12.3 TAILINGS IMPOUNDMENTS

Tailings are the products remaining after the extraction of metals from
ore by physical and chemical methods. The ore is finely-ground and
slurried to allow the processing. Grain sizes of tailings are typically in
the fine sand and silt range and can contain as much as 80% smaller
than the US No. 200 sieve (0.074mm). Tailings are typically deposited
as a slurry into specially designed impoundments. Special depositional
methods have been developed to enhance the stability of such im-
poundments and to allow for environmentally safe disposal of the
tailings. Vick (1983) describes the production and disposal design of
tailings impoundments in much detail. Van Zyl and Vick (1988) present
papers describing the recent practice of tailings disposal. An extensive
bibliography on tailings disposal is presented by ICOLD (1989).
Tailings are typically deposited in an impoundment formed by an
embankment or dike. The embankment can be constructed of fill
material or, in some instances, of the tailings themselves. Construction
with tailings material involves controlled spigoting of the tailings, i.e.,
the discharge of tailings through several small pipes or spigots, and
allowing the solutions to drain from the solids. In some instances, the
tailings may be cycloned to separate the coarser fraction, which then is
used for embankment construction.
An impoundment can be formed by constructing a single embank-
ment across a valley or a three-sided embankment on the side of a
valley or on sloping ground. On flat ground, an embankment or a
'ring' dike that completely surrounds the deposited tailings materials is
required (Robertson and van Zyl, 1980).
Slurried deposition is the most common and practical means and can
be accomplished by the following methods: subaqueous deposition;
 Tailings impoundments 275

managed deposition; conventional point discharge; and cycloned

deposition. Each is now considered in turn.

Subaqueous deposition
This involves discharge under water in the impoundment. Deposition
is uncontrolled, and the tailings settle as a soft bottom sediment, or are
transported and dispersed over a large area. The resulting low density
material is generally very soft and may require special construction
techniques should a cover be required during closure.

Managed deposition
This method, which has also been referred to as subaerial deposition
(see discussion below) is typically accomplished by discharging the
slurry from one or more points around the perimeter of the impound-
ment area. As the slurry drains toward the low point in the impound-
ment, the solids spread out over the bottom to form beaches or deltas.
The free water drains to a central location, or low point along the
perimeter, where it either is recycled into the processing system,
decanted to a separate containment facility for reuse or evaporation,
or evaporated in place. The point or points of discharge are moved
around the impoundment to allow the exposed solids to dry and
increase in density. Subsequent layers are then deposited, and the
cycle is repeated.

Conventional point discharge

This method is similar to managed deposition, except that no effort is
made to add additional discharge points or to frequently move the
discharge points. As a result, the tailings are usually not as dense
or dry as with managed deposition. Also, with conventional point
discharge, the finer portion of the solids tends to settle out in the same
area and create a soft area that is difficult to access, and subject to
additional settlement/consolidation over time.

Cycloned deposition
This involves separation of the coarser tailings fraction prior to de-
position. This type of deposition creates a zone or pile of coarse solids
relatively free of excess solutions, plus a pool of solutions and wet
fine solids which may not drain readily and, therefore, consolidate.
Operation of a cycloning system requires planning and a significant
amount of management if the system is to function properly. It usually
276 Mine waste disposal

is done, at least in part, to provide coarse tailings for use as fill in

raising the embankment.
The moisture content and density of deposited tailings are functions
of the ore extraction process, method of deposition, grain size dis-
tribution, and specific gravity. Moisture content can vary from 20
to over 60%. For slurried deposition, the material is saturated and
moisture contents typically vary from 40 to 60% and dry densities from
1 to 1.3 g/cm 3 . At the end of deposition the density of a tailings deposit
generally increases with depth. The lower tailings have usually con-
solidated, due to the weight of the upper tailings with time, while the
upper tailings are still in the process of consolidating.
The surface of the tailings will settle after deposition ceases and
solution will drain from the settling deposit. Finite strain consolidation
theory is well suited for evaluating the consolidation of a tailings
deposit during construction as well as after the end of deposition
(Caldwell et a/., 1984).
Where managed, e.g., layered, deposition of tailings has occurred,
the densities may be higher than the slurried deposition, i.e., typically
1.3 to 1.45 g/cm3. As a result there is likely to be less settlement of the
surface of the tailings after closure. With lower moisture contents,
typically 20-40%, and no standing water, the surface of managed
tailings becomes accessible for closure activities in a shorter period of
time than slurried tailings, and usually without the need for special
construction techniques.
It is also possible to 'dry' tailings so that deposition can take place
using 'landfill' techniques. Not all tailings materials are amenable to
this type of treatment as it is dependent on the material characteristics,
especially the presence of clays in the ore. The presence of clay can
make it practically infeasible to 'dry' the tailings, or can result in a
tailings product which still has a high moisture content (Robertson et
a/., 1982).
'Dewatering' or 'drying' of the tailings occurs in the metals recovery
plant using unit process, such as thickening, pressure filtration, vacuum
filtration, and a combination of these. The residual moisture content
is usually at or just below saturation, and generally varies from 15
to 25%, however, for very fine-grained material, higher values can
occur (Robertson et ai., 1982). The dry density of these materials will
generally vary from 1.2 to 1.6 g/cm 3 depending on how much com-
paction occurs when the material is deposited.
Deposition of dried tailings is accomplished by either truck dumping
and spreading with a bulldozer, or by conveyor transport and stacking.
This type of deposition results in a pile with no pools of free-standing
solution. Direct precipitation onto the pile is allowed to drain off and
collect in a separate holding pond. Depending upon the degree of
dewatering, residual seepage of solution can occur as a result of gravity
 Tailings impoundments 277
drainage and consolidation of the tailings. Where the tailings contain a
high proportion of silt- and clay-sized fractions, some form of contain-
ment may be necessary to prevent the tailings from spreading.
Tailings impoundments can be constructed by using tailings to
raise the embankment or by using compacted earth fill as the contain-
ment structure. Tailings impoundments containing toxic materials,
such as cyanided gold tailings, are often constructed as fully-lined
impoundments.
Tailings impoundments raised with tailings can be constructed using
the upstream, centerline, or downstream construction methods (Vick,
1983). Figure 12.1 presents schematics of these construction methods.
For the upstream construction method, tailings deposition can take
place through spigots, i.e., multiple open-ended discharges which will
allow natural segregation of the particles along the beach, so that
coarse particles are deposited close to the embankment, while finer
particles are deposited further down the beach. A beach slope is also
formed which is characteristic of the slurry material and the discharge
rate. Cyclone separation of fine and coarse particles can also be used so
that coarse particles are deposited for embankment construction, while
the finer particles are deposited in the pool area.
Centerline and downstream construction are typically done through
the use of cyclones, where the embankment raises are constructed
with the coarse materials (cyclone underflow), and the fine materials
(cyclone overflow) are deposited behind the embankment. Sufficiently
large volumes of coarse material must be available following cyclone
separation to allow for centerline and downstream construction. The
schematics in Fig. 12.1 clearly show that large volumes of sands
are required to construct centerline and downstream embankments,
especially towards the end of the facility's life.
As described above carefully managed deposition can take place
through the regular rotation of depositional areas, allowing for the
deposition of thin layers, which are allowed to dry and increase in
density before the next layer is deposited. Such thin-layer managed
tailings deposition has been practiced in the South African gold mining
industry since the early part of this century (Ruhmer, 1974). Most
recently, a number of publications have referred to this method as
'subaerial deposition' implying that it is tailings deposition which is
not done under water (Knight and Haile, 1983; Cincilla et al., 1991).
Although this terminology has found its way into the literature, it
does not specifically refer to thin-layer managed deposition. Any
depositional method which allows for deposition of tailings 'under air',
such as cycloned tailings deposition, should also be considered as
subaerial deposition. This author therefore prefers to refer to this
method as thin-layer managed tailings deposition vs. the catch-all term
'subaerial' deposition.
278 Mine waste disposal

Fig. 12.1 Upstream, centerline and downstream construction of tailings

impoundments.

Thin-layer managed tailings deposition has a number of advantages,

however, it is important that the operator of such a tailings impound-
ment must make a firm commitment to managed deposition as it
does require considerable attention. From an operating standpoint,
it is, therefore, more expensive than less managed tailings deposition
methods.
In order to reduce seepage losses to the environment from tailings
impoundments containing toxic materials, a number of alternatives can
be considered. Fully-lined impoundments can be constructed or the
thin-layer managed technique can be combined with underdrained
tailings impoundments to reduce the hydraulic head on the liner, and
thereby reducing the potential for seepage losses.
During the initial stages of tailings deposition, the drain layer
enhances the drainage of the deposited tailings. After the drain layer is
covered with slimes, the tailings permeability above the drain layer
reduces significantly and with time, effectively form a second liner.
Hydraulic conductivities as low as 5 x 10- 7 cm/sec can be obtained in
 Tailings impoundments 279
tailings (depending on particle size and clay mineral content) when
subjected to high overburden stresses. Seepage is removed from the
drain layer during operations, thereby reducing the head on the syn-
thetic liner to very low values. This approach has been successfully
practiced at a number of newly constructed gold tailings impound-
ments in the US.
The sedimentation behavior of tailings along a beach following
deposition has been studied by a number of researchers and has
resulted in the definition of a beach profile as well as the particle size
segregation along the beach. The deposition of particles from the slurry
and, therefore, the segregation along the beach, as well as the beach
profile, are functions of specific gravity of solids, percent solids in
slurry, and discharge rate of the slurry. Melent'ev et al. (1973) propose
a model for the development of a beach, as well as segregation of
particles along the beach.
In 1979, this author, with the help of a translator, translated Chapter
2, Principles of Hydraulic Fill Theory, of the text by Melent'ev et al.
(1973) in a specific project. Subsequently, a number of researchers
have used this translation and have successfully shown that
Melent' ev' s theories are valid and can be applied successfully to
the deposition of gold and platinum tailings (Blight, 1987). Consider-
able effort in this area, combined with seepage analyses, have also
been presented by Abadjiev, (1985).
The profile of a beach is generated by the gravitational sorting of
particle sizes as the slurry flows down the beach. A reduction in
particle size occurs along the beach which results in the reduction of
the hydraulic conductivity of the tailings as a function of distance from
the depositional point (Blight, 1987). A master profile of the beach is
developed which can be expressed as:

(12.2)

where H is the length of beach, Y, the elevation between point

of deposition and pool, X, the distance along beach, and n, the
dimensionless constant dependent on tailings characteristics.
Blight (1987) shows that this expression models the master beach
profile for various tailings materials. Figure 12.2 presents the dimen-
sionless beach profiles for various tailings materials.
Abadjiev (1976, 1985) has suggested the following relationship for
the change of saturated hydraulic conductivity for deposited tailings as
a result of the material segregation along the beach.
k = ae- bH (12.3)
where a and b are constants characteristic of the beach, and H is the
distance along beach from deposition point.
280 Mine waste disposal
1.0

~------Copper (n = 1.4)
0.8
.--------Diamond (n = 1.5)

0.6 ~---- Platinum (n = 2.0)

H/Y ____- - Gold (n = 4.0)

0.4

0.2

O-+----,----,----,----r-~-~

o 0.2 0.4 0.6 0.8 1.0

H/X

Fig. 12.2 Tailings impoundment master beach profile for spigotted tailings,
after Blight (1987).

Further characterization of master beach profiles and depositional

behavior is important for the more accurate modelling of tailings
deposition and therefore the geotechnical characteristics of a tailings
deposit.
Geotechnical concerns of tailings impoundment design center on
static and especially dynamic stability, while the environmental con-
cerns center on seepage of contaminated liquids into the environment,
surface discharge of contaminated liquids, and dusting. Static stability
is determined by the shear strength of the tailings and the presence
and location of a phreatic surface. The potential of a liquefaction failure
of a tailings impoundment must be considered in seismically active
areas. Much has been published about this topic and the interested
reader is referred to the literature (e.g., ICOLD, 1990).
The geotechnical characteristics of tailings depend largely on the
amount of clay minerals in the tailings. If the tailings are free of
clay minerals, then the tailings will behave as a frictional material. A
specific concern during the construction of a tailings impoundment is
the consolidation behavior of the finely-ground slurried materials. If a
 Spent heap leach ore 281

tailings impoundment is constructed rapidly (at a high rate of rise),

then excess pore pressures do not have time to dissipate and can lead
to liquefaction failures. Such failures can occur under static loading
conditions, however, they are obviously more prevalent under dynamic
loading conditions.
The angle of internal friction of non-clayey tailings material under
effective strength conditions is typically in the order of 35 degrees, due
to the angular nature of tailings particles. This value is maintained for
coarse as well as fine tailings, i.e. sands and slimes. Direct shear as
well as triaxial shear tests can be used to obtain shear strength charac-
teristics of tailings. In most cases reconstituted samples must be used
because of the sandy nature of tailings (Chen and Van Zyl, 1988). In
situ testing of tailings impoundments can be done successfully using
piezocone equipment. The results can be used to interpret shear
strength, consolidation as well as permeability parameters (Rust et al.,
1984).
The geotechnical design of a tailings impoundment requires a
thorough understanding of tailings production, depositional behavior,
stability considerations, and environmental design requirements.
Although not discussed in this section, liner design is becoming an
important consideration in tailings impoundment design. In most cases
a single composite liner provides sufficient environmental protection.
Liner design details are discussed elsewhere in this text.

12.4 SPENT HEAP LEACH ORE

Heap leaching is a process where low-grade ores are stacked on a low-

permeability base and irrigated with a lixiviant to extract the metal. In
the case of gold, cyanide is used as a lixiviant and in the case of oxide
copper, sulfuric acid is used as a lixiviant. This method of extraction
was first developed on a commercial scale in the US for gold deposits
in the late 1970s, although copper and uranium heap leaching had
been practiced on a world-wide basis since the earlier part of the 20th
century. Considerable advances have been made over the last decade
in the design of heap leach facilities (Van Zyl et al., 1988).
A large percentage of gold is presently extracted in the US using
heap leach facilities. It is therefore clear that the design of heap leach
facilities is receiving considerable attention. After extraction of the
metal, the spent heap leach ore remains as a waste product. The long-
term disposal of such waste must be carefully considered during the
initial design of the heap leach facility.
Heap leach projects can be developed on permanent pads. For
example, the ore is left on the pad during and after leaching and the
spent ore, therefore, remains on a lined facility. Reusable pads can also
282 Mine waste disposal

be used where the spent ore is removed after metal extraction, and
disposed of at a separate facility. Valley leach facilities are used in steep
terrain to contain the ore. After leaching and rinsing, the ore remains
in its containment and the liner is punctured to allow drainage of
infiltration into the foundation.
Heap leach ores can consist of run-of-mine material, where a blasted
ore is placed on the pad without any further preparation. The ore
can also be prepared, for example, through crushing or crushing and
agglomeration. In the latter cases, the ore particles are typically smaller
than 75 mm and very often smaller than 20 mm. Excessive fines in the
crushed ore can be agglomerated onto the coarse particles by using
Portland cement as a binder and the tumbling action in a rotating drum
(McClelland and Van Zyl, 1988). It is important to maintain uniform
percolation through the heaps, so that the leachate is evenly distributed
and contacts all the sources of the metal in the ore. Application rate of
leachate is typically in the range of 0.1-0.2L1min!m2.
After leaching, the ore must be rinsed so that the heaps can be
reclaimed. In the case of cyanide heap leaching, considerable attention
is presently paid to rinsing of heaps and new methods and results from
research and practice are expected in the next few years. The degrada-
tion of cyanide in heaps is related to the reactive nature of cyanide
which leads to rapid volatilization and oxidation, as well as biological
processes which take place over a longer period of time (Chatwin,
1989). Bacterial degradation of cyanide occurs readily in spent ore
heaps, as well as natural soils.
The spent ore heap is usually considered rinsed when the effluent
from the heap has a weak acid dissociable (WAD) cyanide content of
0.5 mg!l or lower. Although the cyanide level can be reduced, it is
possible that leachate of metals could be of greater concern in the long-
term after closure of a heap leach facility. It is, therefore, important to
reduce long-term infiltration into a reclaimed heap. Final reclamation
may consist of reshaping of the heap to allow placement of a growth
medium and finally revegetation of the heap. In certain extreme climatic
conditions, where high net precipitation is experienced, it may be
necessary to construct a liner over the top of the reshaped heap.
Liner requirements for heap leach facilities differ between the
various states in the US. Single composite liners provide sufficient
environmental protection for permanent pads and reusable pads where
low hydraulic heads are maintained. In the case of valley leach facilities
a double composite liner may be required because of the higher
hydraulic head being maintained in the heap. Similarly, for the ponds
associated with heap leach facilities, it is recommended that double
composite liners be used so that the hydraulic head on the lower
composite liner can be reduced when the leak collection system is
evacuated.
 Liquid waste 283
The slope stability of heaps constructed on clay or geomembrane
liners must be evaluated. Site-specific geometric and material charac-
teristics must be used. If it is assumed that the heap is constructed on
top of an underlying pad, consisting of a geomembrane constructed
onto compacted low permeability soil, then the interface between the
low permeability soil and the geomembrane constitutes the weakest
part of the structure. A block-type failure is the most critical. It can
further be assumed that the heap material, as well as the interface
shear strengths, can be characterized by an angle of internal friction.
The friction angle of a liner interface is dependent on the liner type
as well as the material in contact with the liner. Site specific testing
must always be conducted. However the following general statements
can be made about interface friction angles:
1. more rigid liners, such as high density polyethylene (HOPE) have
lower friction angles than flexible liners, such as very low density
polyethylene (VLOPE) and polyvinylchloride (PVC);
2. textured (surface-roughened) geomembranes have higher friction
angles against soil than smooth geomembranes;
3. the presence of clays against the geomembrane interface results in
lower friction angles than granular materials; and
4. the friction angle of interfaces between two layers of synthetic
materials can be very low, e.g., HOPE and a geotextile.
The design of heap leach facilities is well regulated in the US where
such facilities are found. The geotechnical engineer must use site
specific design criteria to develop an environmentally protective design.
Leakage and stability considerations are foremost in the design of heap
leach facilities for operational and post closure conditions.

12.5 LIQUID WASTE

Liquid waste streams at mines consist of tailings water, leachate from

heap leach facilities, smaller leach streams containing higher concen-
trations of contaminants, and water produced through dewatering of
pits or collection of surface runoff.
The design of liquid waste containments is dependent on the
chemical characteristics of the contained liquid. Sludges often collect in
liquid waste ponds and such sludges can become a concern at the time
of reclamation of the waste.
Closure of liquid waste ponds is usually done through treatment and
disposal of the liquids, such as chemical treatment or evaporation.
Sludge remaining in the bottom of the pond may be removed and
deposited in one of the other mine waste facilities on site, or may have
to be solidified in place through the use of portland cement, or in some
284 Mine waste disposal

cases, may have to be removed and deposited in a hazardous waste

landfill. The pond liners can be folded over and removed and placed in
one of the other mine waste facilities or can be buried in place.

12.6 OTHER MINE WASTES

12.6.1 Coal mining

In the removal of coal, large volumes of overburden and interburden
wastes must be removed. Because of the continuous nature of coal
beds, such mining is done by draglines where the overburden is cast in
the mined-out area. The characteristics of the coal waste are typical of
sedimentary environments consisting of interbedded sandstones and
shales. In coal strip mining, the mine spoil is recontoured and reclaimed
as the mine develops.

12.6.2 Industrial minerals

Industrial minerals include sand and gravel, zeolites, vermiculite, and
other similar minerals. The wastes from such mining operations are
diverse in their characteristics. The mining methods can consist of
mechanical separation of the minerals from waste, most often using
dry processes. Each of these wastes must, therefore, be considered on
a site-specific basis.

REFERENCES

Abadjiev, C.B. (1976) Seepage Through Mill Tailings Dams, Trans. of the 12th
lCOLD, Mexico City, 1, 381-93.
Abadjiev, C.B. (1985) Estimate of the physical characteristics of deposited
tailings in the tailings dam of non ferrous metallurgy, Proc., 11 th Int. Conf.
on Soil Mechanics and Foundation Eng., San Francisco, 3, 1231-4.
ASCE (1979) Current Geotechnical Practice in Mine Waste Disposal, papers
collected by the Committee on Embankment Dams and Slopes of the
Geotech. Eng. Div.
Bishop, A.W., Hutchinson, J.N., Penman, A.D.M. and Evans, H.E. (1969)
Geotechnical investigations into the causes and circumstances of the
disaster of 21st October 1966, in A Selection of Technical Reports Submitted to
the Aberfan Tribunal, Item 1,1-80, H.M.S.O., London.
Blight, G.E. (1987) The concept of the master profile for tailings dam beaches,
in Proc. of the Int. Conf. on Mining and Ind. Waste Management, (eds J.A.
Wates and D. Brink) South African lnst. of Civ. Eng., Johannesburg, South
Africa, pp. 95-100.
Caldwell, J.A., Ferguson, K., Schiffman, R.L. and Van Zyl, D. (1984)
 References 285
Application of Finite Strain Consolidation Theory for Engineering
Design and Environmental Planning of Mine Tailings Impoundments,
Sedimentation/Consolidation Models - Predictions and Validation, (eds R. W.
Yong and F.C Townsend), ASCE, pp. 581-606.
Chatwin, T.D. (1989) Cyanide AttenuationlDegradation in Soil, Resource Recovery
and Conservation Company, Salt Lake City, Utah.
Chen, H.W. and Van Zyl, D. (1988), Shear Strength and Volume Change
Behavior of Copper tailings under Saturated Conditions, in Hydraulic Fill
Structures, (eds D.J.A. van Zyl and S.G. Vick), ASCE.
Cincilla, W.A., Dye, RA. and East, D.R (1991), Nevada Goldfields Aurora
project: a case history of subaerial tailings deposition vs. mechanical
dewatering and disposal, paper presented at the 94th Nat. Western
Mining Conf., Colorado Mining Assoc., Feb. 24-27.
Dobry, R. and Alvarez, L. (1967) Seismic failures of Chilean tailings dams, 1.
Soil Mechanics and Foundations Div., A.S.CE., 93(SM6), 237-60.
EPA (1985), Report to Congress: Wastes from the extraction and beneficiation
of metallic ores, phosphate rock, asbestos, overburden from uranium
mining and oil shale, EPA/530-SW-85-033.
Hoek, E. and Bray, J. (1981), Rock Slope Engineering, revised 3rd edn, The
Institution of Mining and Metallurgy, London.
International Commission on Large Dams (ICOLD) (1989) Bibliography: mine and
industrial tailings dams and dumps, Bulletin 44a.
International Commission on Large Dams (ICOLD) (1990) International
Symposium on Safety and Rehabilitation of Tailings Dams, Sydney, Australia,
May 23.
Knight, RB. and Haile, J.P. (1983), Subaerial Tailings Deposition, 7th Pan-
American Conf. on SM & FE, Vancouver, B.C
Lootens, D.}., Greenslade, W.M. and Barerk, J.M. (eds) (1991) Environmental
Management for the 1990s, Soc. for Mining, Metallurgy, and Exploration,
Inc., Littleton, CO.
McCarter, M.K. (ed.) (1985) Design of Non-Impounding Mine Waste Dumps, Soc.
of Mining Eng., Littleton, CO.
McClelland, G.E. and Van Zyl, D. (1988), Ore preparation: crushing and
agglomeration, Ch. 5. in Introduction to Evaluation, Design and Operation of
Precious Metal Heap Leaching Projects, (eds D.J .A. Van Zyl, I.P.G. Hutchison
and J.E. Kiel), Soc. of Mining Eng., Littleton, CO, pp. 68-91.
Melent'ev, V.A., Kolpashnikov, N.P. and Volnin, B.A. (1973) Hydraulic Fill
Structures, Energy, Moscow.
National Building Research Institute (1959) An investigation into the Stability of
Slime Dams with Particular Reference to the Nature of the Material of their
Construction and the Nature of their Foundation, Pretoria, South Africa.
Robertson, A. MacG. and van Zyl, D.J.A. (1980) Design and construction
options for surface uranium tailings impoundments, Proceedings of the
First International Conference on Uranium Mine Waste Disposal, Soc. of
Mining Engrs, AIME, Chapter 11, pp. 101-119.
Robertson, A. MacG., Fisher, J.W. and Van Zyl, D. (1982) Handling and
disposal of dry uranium tailings, Proc., 5th Symp. on Uranium Mill Tailings
Management, Colorado State Univ., Fort Collins, Colorado, pp. 55-69.
Ruhmer, W.T. (1974), Slimes-dam construction in the gold mines of the Anglo-
American group. 1. South African Institute of Mining and Metallurgy, 74(7),
273-84.
Rust, E., Van Zyl, D. and Follin, S. (1984) Interpretation of piezometer cone
testing of tailings, Proc. of 6th Ann. Symp. on Management of Uranium
286 Mine waste disposal
Tailings, Low-Level Waste, and Haz. Waste, Colorado State Univ., Fort
Collins, Colorado, pp. 627-38.
Smith, D.A. (1987) Mining America: the industry and the environment, 1800-1980,
University Press of Kansas.
Steffen, Robertson and Kirsten (1989) Draft Acid Rock Drainage Technical Guide,
BiTech Publishers, Vancouver.
Uhle, R.J. (1986) A statistical analysis of rockfill data - shear strength deformation
parameters with respect to particle size. MS thesis, Colorado State University,
214pp.
University of California at Berkeley (1988) Mining Waste Study Team, Mining
Waste Study, Final Report, California Water Resources Control Board.
US Department of the Interior (1975) Engineering and Design Manual, Coal
Refuse Disposal Facilities, Mining Enforcement and Safety Administration.
Van Zyl, D. (ed.) (1985) Cyanide and the Environment, 2 vols, Geotechnical
Engineering Program, Colorado State University.
Van Zyl, D.J.A., Hutchison, I.P.G. and Kiel, J.E. (eds) (1988) Introduction to
Evaluation, Design and Operation of Precious Metal Heap Leaching Projects, Soc.
of Mining Eng., Littleton, CO.
Van Zyl, D.J.A. and Vick, S.G. (eds) (1988) Hydraulic Fill Structures, ASCE,
Geot. Special Publication No. 21.
Vick, S.G. (1983) Planning, Design, and Analysis of Tailings Dams, John Wiley and
Sons.
Wahler, W.A. and Associates (1973) Analysis of Coal Refuse Dam Failure, Middle
Fork, Buffalo Creek, Saunders, West Virginia. US Department of the Interior,
Bureau of Mines, 2 vol.
 PART THREE

Rem.ediation
Technologies
 CHAPTER 13

Strategies for remediation

Larry A. Holm

13.1 INTRODUCTION

Within the last several years, waste remediation has become a signifi-
cant area for the application of geotechnical engineering. In particular,
waste remediation efforts at 'uncontrolled' waste sites have drawn
increasing attention from the technical, regulatory, legal, and public
communities. Recognition of potential human health and environ-
mental risks has heightened the efforts on national and international
agendas.
Waste remedial planning and implementation is a complex technical,
regulatory, legal, and public environment in which the geotechnical
engineer applies both general and specific knowledge and experience.
In many ways, waste remedial planning and implementation offer
significant challenge and opportunity for the geotechnical engineer.
Capable application of specific geotechnical technologies (discussed
in other chapters) is dearly needed. Also, the geotechnical engineer
has much to contribute to an overall waste remediation approach by
virtue of understanding decisionmaking under uncertainty. Through
traditional training and experience with natural geologic conditions
and materials, geotechnical engineers possess an understanding of
the thought processes and techniques for dealing with uncertainty.
Although waste remediation embodies many concerns and technologies
that may be new to the geotechnical engineer, the basic understanding
of uncertainty can provide a valuable contribution to the overall plan-
ning and implementation of remedial activities.
By definition, waste 'remediation' deals with corrective efforts on
pre-existing problems. The remedial process for such problems is
fraught with uncertainty about the source, cause, and condition at

Geotechnical Practice for Waste Disposal.

Edited by David E. Daniel.
Published in 1993 by Chapman & Hall, London. ISBN 0 412 35170 6
290 Strategies for remediation

failure, the nature, rate of spread, and extent of the problem con-
ditions, and the potential risks posed to humans and the environment.
The application of new, some relatively unproven, technologies to
these problems further adds to uncertainty. Remedial work, therefore,
often requires a different perspective and technical approach than is
demanded of many traditional engineering endeavors.
Waste remediation efforts are heavily influenced, and in the US often
dictated, by statutory and regulatory compliance. It is increasingly
common that regulatory standards and guidance provide very pre-
scriptive procedural and technical requirements. Furthermore, waste
and environmental regulations are constantly evolving, generally in-
creasing in scope and stringency.
Waste remediation is forensic in character. The litigative potential of
waste remediation activities, at least in the US, often emphasizes legal-
evidential considerations. Such emphasis can expand efforts and divert
attention from the primary goal of determining adequate technical
solutions. In addition, public interest, awareness, and participation in
many remedial planning and implementation efforts require that the
engineer formulate, conduct, and present the rationale and results
of these efforts in a manner that is understandable to an audience
that varies widely in technical background. For success of a project,
remedial strategies and approaches must be formulated with knowl-
edgeable consideration of the many, sometimes conflicting, issues and
goals involved.

13.2 APPROACH TO REMEDIAL PLANNING

AND IMPLEMENTATION

In formulating a successful approach to waste remediation, the geo-

technical engineer should recognize the fundamental differences be-
tween remedial efforts and design of new facilities. Central to the
differences is the high degree of uncertainty associated with remedial
efforts, particularly in terms of definition of the source of contamina-
tion, extent of contamination, and effectiveness of proposed remediation
methodologies. A sound grasp of such uncertainties is essential to
any logical approach to remediation and requires a thorough under-
standing of the nature of uncertainties involved.

13.2.1 Comparison to traditional design and failure evaluation

Geotechnical engineering and, indeed, most engineering and scientific
disciplines function successfully based on accumulated experience; the
applied methodologies are essentially the result of trial and error,
 Approach to remedial planning and implementation 291

however well-founded in scientific principles. In traditional endeavors,

new designs proceed to successful completion through a process that
draws upon, and translates, previous experience to the specific project
at hand. While uncertainties do exist, they are managed by application
of accepted techniques of data gathering (sampling and testing), data
interpretation, analysis, and application of factors of safety. The tech-
niques may vary, but each has its fundamental basis in previous ex-
perience. While the techniques of sampling, testing, interpreting, and
analyzing (including statistical methods) can provide greater insight
into the range or degree of uncertainty, they do not eliminate inherent
uncertainty. Uncertainty is still managed through application of em-
pirical judgment, often in the form of an appropriate conservatism,
application of factors of safety, and/or adherence to accepted design
standards.
When failures occur, back-analyses using traditional design method-
ologies are often found to be lacking. Simply setting the factor of safety
equal to unity may neither predict the failure nor demonstrate the
characteristics evidenced by the failure. The empiricism embodied in
the design methodology often renders the methodology incapable of
adequately modeling the failure. Furthermore, uncertainties in basic
parameters of the analyses allow (and may even encourage) accepting
or 'proving' a failure hypothesis initially perceived to be the correct
one.
Approaches have been developed and applied in geotechnical
engineering that significantly improve the ability to model and under-
stand the causes, mechanisms, and results of failures. These approaches
recognize and attempt to avoid potential pitfalls of such efforts. In any
event, it is seldom possible to understand completely all details of the
cause and result of a failure.
Remedial planning and implementation are most closely comparable
to failure evaluation in the traditional sense; that is, the conditions to
be remediated represent, or are the consequences of, a 'failure'. All of
the limitations and potential pitfalls of traditional failure evaluation
are present in remediation. Furthermore, many more potential
uncertainties are present in waste remediation than in tradtional geo-
technical engineering. Successful remediation demands that the nature
and character of the uncertainty in remediation be understood.

13.2.2 Understanding uncertainties in waste remediation

Uncertainties inherent throughout remedial planning and implementa-
tion can be understood in the context of potential variability and
variations of the conditions requiring remediation, the criteria that
remediation is intended to achieve, and the performance of alternative
or selected technologies.
292 Strategies for remediation
Uncertainty of conditions
Any waste site involves both physical and chemical conditions that will
influence remediation. In a general sense, these include: conditions of
the source(s); contaminant release mechanism(s); fate, transport, and
interactions in the subsurface; and conditions of actual or potential
exposure to receptors. As illustrated in Fig. 13.1, these conditions and
interactions, in even a simplified case, are complex. Although complete
definition of site conditions is impossible, formulation of a conceptual
model in this context can facilitate understanding and communication
of critical conditions requiring remediation and can enhance the under-
standing of general, inherent uncertainties.
It is often useful, or even essential, that the conceptual model of the
site provide distinction between 'on site' and 'offsite' conditions. In
the case of a localized, single source (such as a confined facility), the
distinction may be straightforward. In cases of widespread, multi-source
situations (such as regional groundwater contamination or distributed
airborne contamination), the determination of the 'site' boundary
may be problematic. Though boundary determination may ultimately
be jurisdictional rather than technical, the definition of 'onsite' and
'offsite' can influence the technical approach to remedial planning and
implementation and should be carefully considered.
General understanding of the source(s) of contamination and their
release mechanism(s) may include estimating the location of the source
and the type, total volume, and rate of release of contaminants. At
some waste sites such information is readily available and reasonably
complete. More commonly (particularly for large, old, multi-user sites),

"ONSllE" -------t.-I-I....- - - "<milE" .1

RELEASE CONTAMINATED MIGRATION INTERMEDIATE EXPOSURE
SOURCES MECHANISM MEDIA PATHWAY MEDIA PATHWAY RECEPTORS

Fig. 13.1 Simplified source, media, pathway, receptor diagram.

 Approach to remedial planning and implementation 293

such information is only partially available and may be of questionable

accuracy. Although optimum use should be made of available infor-
mation by reviewing records and interviewing people knowledgeable
of the history of the site, often only rough estimates of the source and
release conditions are possible.
The conditions of the contaminated media at a waste site usually
receive considerable scrutiny during remedial planning. Commonly,
these efforts are referred to as 'determining the nature and extent' of
contamination. The efforts provide information on:
• the types of contaminated media present (e.g., air, soil and sedi-
ments, facilities, surface water, groundwater, and biota);
• the physical characteristics of each medium (e.g., temperature and
wind direction for air, total organic carbon content, cation-exchange
capacity, and index/engineering properties for soils and sediments,
and temperature, pH, hardness, alkalinity, gradients, and flow rates
and quantities for surface water and groundwater);
• the types (both general categories and specific compounds or
elements), concentrations, and characteristics of contaminants
present in each medium; and
• the lateral and vertical extent of each contaminated medium.
The interactions between contaminated media can be important. As
illustrated in Fig. 13.1, interactions may include numerous, complex
pathways for past, ongoing, or potential contaminant migration. The
contaminant fate and transport mechanisms of many of the interactions
may be only partially understood and defy scientific quantification,
thereby allowing only qualitative assessment.
Assessment of background conditions, both physical and chemical,
is often an important part of site characterization. Physical informa-
tion, remote from the site, may be needed to estimate ambient and
boundary conditions for air, surface water, and groundwater. Assess-
ment of chemical background may include estimation of the presence
and spatial distribution of chemical types and concentrations ambient
to the site. For sites in relatively pristine ambient environments,
estimation of background chemical conditions may be straightforward.
More commonly, site locations in areas influenced by urban and
industrial environments pose significant questions regarding the ap-
propriate determination of background conditions.
Since the ultimate goal of any remedial effort is to mitigate impacts
on humans and/or the environment, a reasonable understanding of
exposure and receptor conditions is often critical to remedial planning
and implementation. Exposure conditions include not only the
physical/chemical mechanisms by which contaminants may be avail-
able to receptors, but also the assimilation mechanism (e.g., inhalation,
ingestion, or dermal absorption) and the dose-response characteristics
294 Strategies for remediation
(e.g., type, concentration, quantity, and duration of contaminant
contact). For contaminants with known or suspected health effects,
methodologies for estimating exposure conditions are available (as
discussed in the following section). Significant uncertainty may remain
for individual contaminants for which demonstrated or suspected
health effects are not available and for multiple-contaminant conditions
where synergism may influence cumulative health effects. With respect
to receptor populations, estimates may include the number and type of
human and/or environmental receptors affected by current and future
site-use scenarios. In particular, forecasting future site use may playa
pivotal role in the overall remedial approach.

Uncertainty of criteria
Criteria that influence remedial planning and implementation (as
impediment or impetus) include national, state or provincial, and/or
local statutes, regulations, standards, advisories, or guidance (termed
'regulations' here). Pertinent regulations may include those specific to
waste remediation but also those of more general or environmental
nature. In addition, specific health-related (toxic and carcinogenic) cri-
teria often supplement regulations in application to remedial activities.
Although comprehensive discussion of criteria is beyond the scope
of this chapter, this section attempts to illustrate their importance
in remedial activities and to describe, however generally, potential
influences that may result from the multiplicity, complexity, and
sometimes conflicting character of the waste remediation criteria.
Constant evolution can introduce significant uncertainty as changes in
remedial criteria occur both during and after remedial planning and
implementation.
In general, regulations can be categorized as location-specific,
contaminant-specific, or action/technology-specific. Location-specific
regulations include those related to where the contamination and
remediation occur. Contaminant-specific regulations include those
relating to type, quantity, and/or character of the waste or contamina-
tion to be remediated. Action/technology-specific regulations include
those pertaining to the means and methods of remediation. An im-
portant aspect of regulation categories relates to when in the remedial
process each category of regulation can reasonably be considered. While
location-specific regulations can be brought to bear on the remedial
considerations relatively early (when site identification has been made),
consideration of contaminant-specific regulations may await site char-
acterization information. Likewise, action/technology-specific regula-
tions can only be considered after the concepts, alternatives, and
technologies of the remedial implementation are formulated.
As an example of the regulatory framework for remedial planning
and implementation, the United States Environmental Protection
 Approach to remedial planning and implementation 295

Agency (USEPA, August 1988) has articulated regulatory compliance

criteria for hazardous waste remediation in terms of applicable or
relevant and appropriate requirements (ARARs). In the terminology of
the USEP A, 'applicable' refers to promulgated, legally enforceable laws
and statutes that specifically address waste substances or pollution.
'Relevant and appropriate' (and it is important that it must be both)
refers to promulgated laws and statutes that relate to situations suf-
ficiently similar to the particular waste situation and that are well
suited to the situation. In addition to ARARs, USEPA provides guidance
for consideration of other criteria, phrased 'to be considered (TBC)',
that are nonpromulgated, nonenforceable regulations that are not
strictly 'applicable' nor 'relevant and appropriate' but that may warrant
inclusion based on specific circumstances. While the intent is that
remediation comply with ARARs, USEP A recognizes six conditions of
waiver of ARARs: interim measures, greater risk to health and environ-
ment, technical impracticability, equivalent standard of performance,
inconsistent application of State requirements, and fund balancing. The
US experience clearly demonstrates the myriad of regulations that may
be brought to bear on remedial planning and implementation.
In the US, public participation in remedial planning is mandated by
law (CERCLA, 1980; SARA, 1986; USEPA, October 1988; USEPA, June
1988). The degree of public participation may vary between sites but,
at a minimum, generally includes formal public notification, public
review and comment on documents, public meeting, and written
response to public comment. While public involvement may not
dramatically alter a well-conducted, technically sound remedial effort,
it does require that the engineer formulate, conduct, and present the
rationale and results of these efforts in a manner that is understandable
to an audience that varies widely in technical background.
In addition to specific regulations, criteria are considered in remedial
planning and implementation through preparation of exposure and
risk assessments; typically these include both environmental (eco-
system) and human health considerations. Exposure and risk assess-
ments may influence the degree and methods of remediation, and
indeed, may dictate whether remediation is required at all. They can
also be useful in determining required remedial action monitoring
(prior to, during, and after remedial implementation). The results of
exposure and risk assessments may establish remediation targets that
vary from, and are in addition to, criteria set forth in regulations.
The USEPA has developed guidance for the conduct of Exposure
Assessments (USEPA, 1988) and risk assessments for both environ-
mental receptors (USEPA, 1989) and human health (USEPA, March
1989; USEPA, September 1989). These guidance documents provide
general methods and considerations to be applied in the qualitative
and quantitative evaluation of exposure routes and actual or potential
risk associated with exposure.
296 Strategies for remediation

Uncertainty of technology performance

In concept, potential remedial actions and technologies can be con-
sidered in the context of Fig. 13.1. The overall goal of protection of
human health and the environment can, depending on the conditions
present, be accomplished by removing the sources and/or contaminated
media, eliminating release, migration, and/or exposure pathways, or
removing potential environmental and human receptors. Each has
found application in remediation; often, combinations of actions are
required. In any case, uncertainties in technology performance are
inextricably linked to uncertainties in site conditions.
Many of the actions finding application to waste remediation include
technologies such as removal and containment that are familiar in
traditional geotechnical engineering. In addition, the emphasis on
providing long-term reduction of the mobility, toxicity, and volume of
wastes has increased interest in application of new and innovative
technologies.
In application to waste remediation, traditional control technologies
such as vertical cutoff walls, liners, and cover systems may be needed
to perform under conditions and for purposes that are outside of
traditional experience. For example, while the performance of soil
liners as water barriers is relatively well understood and predictable,
their application as vaporous or liquid waste barriers must satisfy
additional issues and constraints related to performance and perform-
ance prediction.
In addition, traditional technologies may often be subject to con-
straints during implementation that are not present in nonwaste appli-
cations. For example, soil excavation, which may pose little problem
under traditional circumstances, may be subject to significant concerns
relating to vapor emissions and/or contaminated water control during
implementation in waste remediation applications. Such 'secondary'
concerns can represent significant challenge in application of traditional
geotechnical technologies to waste remediation.
The consideration and application of new and innovative technologies
in waste remediation offer considerable challenge to geotechnical
engineers and scientists. Innovative technologies are usually considered
in the context of their ability to modify the chemical or physical char-
acteristics, or otherwise destroy the hazardous component of the
wastes. Because of the potential for permanent remediation, such
technologies warrant careful consideration in remedial planning and
implementation. Uncertainties in the application of innovative tech-
nologies to specific waste situations may include:
• the ability of technology to treat the full range of contaminant types
present, and potential adverse effects on the technology resulting
from the presence of other contaminants;
 Approach to remedial planning and implementation 297

• the character of residuals resulting from the application of the tech-

nology, including technical and/or regulatory requirements for final
disposition;
• the types and character of any by-products of the technology appli-
cation; and
• the performance of the technology in full-scale remedial application
(i.e., scale-up considerations).
While some of the considerations in the application of innovative
technologies go beyond the technical areas of many geotechnical
engineers, most technologies require geotechnical involvement in
assessing the geologic conditions and impacts (for in situ technologies)
and/or handling, preprocessing, and post-processing requirements (for
other technologies). In this context, most of the uncertainties inherent
in traditional geotechnical technologies are present in innovative tech-
nology applications.
In the application of both familiar and innovative technologies, the
ability to predict and monitor performance and the relative accuracy
of such predictions and monitoring can have significant impacts on
remedial planning and implementation.

13.2.3 Requirements of successful remedial approach

Geotechnical engineering has much to contribute to the overall ap-
proach of remedial planning and implementation. While many tech-
nical aspects of traditional geotechnical engineering translate directly
to remedial activities, to be successful, a basic appreciation of many
'new' issues must be incorporated into remedial planning and
implementation.
Any successful approach to remedial planning and implementation
must address a multiplicity of technical and regulatory issues. Such
complexities demand the detailed attention of a coordinated, multi-
disciplinary project team and integration of issues in a way perhaps
unprecedented in other engineering and scientific endeavors.
Because remedial projects address 'failures', an objective and efficient
approach is required for determination of site conditions. Leonards
(1979), although referring to traditional endeavors, articulated an
approach to failure situations, which is also appropriate in approaching
remedial activities.
1. The first step is to conceive all potential mechanisms of failure,
avoiding a priori assessments of which mechanism is more likely to
be the correct one. The first step is important; if the initial listing of
failure mechanisms does not include the true cause, the probability
of its eventual discovery is greatly diminished.
2. Each mechanism is then examined in turn, using all available
298 Strategies for remediation
evidence, seeking to disprove its validity. It is easier to disprove a
hypothesis than to prove it to be correct; moreover, this approach
minimizes the possibility of personal bias entering into the cal-
culations ... even 'fundamentals' can be interpreted in a way that
encourages accepting a hypothesis initially perceived to be the
correct one.
3. The hypothesis that resists all attempts at being disproved is re-
examined with a view towards identifying other conditions that
must prevail if it is to be the only correct one. If a hypothesis is
correct, features of the existing data previously overlooked are now
found to fit into the picture, or simple additional investigations
suggest themselves as a crucial test of validity, or both.
Inherent uncertainties in conditions, criteria, and technology per-
formance must be recognized in formulating a remedial approach.
Regardless of the level of effort expended, uncertainty cannot be
eliminated and can only be managed through an approach that recog-
nizes this fact, strives to understand the certainties, and establishes
and implements an adaptive methodology. The approach must provide
a perspective for flexibility and incorporate methods that can foresee
and logically respond to changes.

13.3 METHODOLOGY FOR REMEDIAL PLANNING

AND IMPLEMENTATION

The process of remedial planning and implementation has evolved as a

general methodology that parallels the study-de sign-build paradigm
familiar in traditional geotechnical, and other, engineering and scientific
efforts. As its basic steps the methodology includes remedial investiga-
tion (RI) and feasibility study (FS), remedial design (RD), and remedial
action (RA).
Figure 13.2 schematically illustrates the typical sequence and flow of
the process. The figure and the following text are intended only to
illustrate the general technical components of the methodology. Specific
technical activities and activities relating to regulatory compliance or
approval are not illustrated; while obviously important, these can only
be determined based on the regulatory situation of the project.

13.3.1 Pre-RIIFS activities

Pre-RIIFS activities include preliminary site inspection and assessment,
compilation and review of existing data, and formulation and planning
of the remedial approach for the site.
The preliminary site inspection and assessment usually involves
 Methodology for remedial planning and implementation 299

1-1'_ _ _ RI/FS _ _----..t.-I._RD ___.t-I'_ _ R A - + - -

REMEDIAL INVESTIGATION
Conduct Field
Investigations
I
Perform
Identify Location- Bench
and/or
~
&Contaminant
Specific Regulations Pilot Tests
PRE-RI FS
Conduct Perform Preliminary
Preliminary Site Risk Assessment
Inspection and H ..,

.
Assessment
Collect & Analyze
Existing Data
Formulate
" ). '0
FEASIBILITY STUDY REMEDIAL REMEDIAL OPERATION/
&Plan Identify & Screen DESIGN ACTION MONITORING
Site-Specific Technologies Analyze
----"
~
Approach Alternatives -A
Assemble Various -,.I Design -V Implement Monitor
Technologies the the the
Compare I< Action Design Action
into Alternatives
Contrast
Identify Action/ Alternatives
Technology-Specific
Regulations

Fig.13.2 Schematic of the remedial process, after US EPA (1988).

reconnaissance of the site and surrounding area, initial contact with

site and regulatory personnel, and obtaining general information about
the site conditions to allow intelligent planning of future activities.
Prior to beginning RIIFS activities, available information and data
pertinent to the site and potential remedial activities should be
gathered, compiled and analyzed, and summarized. Existing informa-
tion and data may be site-specific, such as records of operations and
previous investigations, and/or regional induding surveys of geology,
hydrology, surface soils, and meteorology. The magnitude of potential
costs and effort required to gather new data (during field investigations)
dictates that the best possible use of existing information be made.
Often the amount of existing information far exceeds the amount of
new data that can be obtained even in the most aggressive remedial
investigation; at best, new investigations may only supplement existing
data. While existing data may be of varying quality, thorough review
prior to initiated investigations provides initial characterization of the
site and area to allow efficient planning of investigations and pre-
liminary identification of potential remedial responses.
The formulation of a site model, as illustrated in Fig. l3.1, can be
useful at this stage. While it is seldom possible to define in detail the
sources, media, pathways, and receptors based on existing informa-
tion, formulation of the overall site situation in this context can assist in
focusing investigations on critical conditions and improve the general
understanding of inherent uncertainties. A remedial approach and
plan can then be formulated, based on initial understanding of the site
situation.
300 Strategies for remediation
Situations that pose obvious, substantial, and immediate threat
should be addressed by immediate action. Immediate actions are
usually limited to actions such as removal of containerized concen-
trated wastes, and exposure-control actions such as site access control
or provisions for temporary potable water supply. Immediate actions
need not involve substantial data gathering or alternative evaluation;
typically they proceed to design (often with only limited, confirmatory
data) and directly to implementation.
For large complex sites, the project may be divided into manageable
subprojects (sometimes termed operable units). An operable unit may
be defined as a specific facility (such as a waste pile or lagoon) or
may be media-specific (such as surface water, soils, or groundwater).
Operable units should be defined and selected, based on the potential
for beneficial early and/or separate action, taking into account the
following factors.
• Understanding of conditions i.e., the ability to sufficiently define the
conditions using existing or readily obtainable data and information.
If lengthy data gathering efforts are required in the remedial plan-
ning process, the utility of the operable unit may be diminished.
• Potential for separate action. If potential actions for an operable unit
cannot be planned and implemented separately within an overall
remedial plan for the site, the operable unit may not be practical.
• Scope of the operable unit. The scope of the operable unit should
not be so large (or the potential to dramatically expand so great)
that decisions on early action are discouraged by funding or other
considerations.
Operable units generally follow the same remedial process illustrated
in Fig. 13.2. Properly selected and defined, operable units will allow
the remedial planning and implementation to proceed more rapidly
than may be possible for the site as a whole. When multiple operable
units are employed, the overall remedial planning and implementation
may take on a phased approach, which can provide substantial benefit,
but also must include assessment and evaluation of the combined
effects of the operable units.
Whether for an individual operable unit, or for the site as a whole,
the scoping efforts in the pre-RIfFS include preliminary identification
of potential response actions, preliminary identification of regulatory
requirements, definition of data needs and data quality objectives,
specific planning of field and laboratory investigation tasks, and, if
appropriate, establishing community relations requirements.
Objective identification of the potential range of remedial response
actions can ensure that investigation tasks provide the data necessary
for later detailed evaluation of remedial technologies and alternatives.
Preliminary identification of regulatory requirements ensures that field
 Methodology for remedial planning and implementation 301

and laboratory investigations provide data appropriate for comparison

to standards, criteria, advisories, and guidance that may be brought to
bear on the remediation.
Data needs are identified to define data types, quantity, and quality
necessary to assess general site conditions and regulatory issues, and
to evaluate the applicability of remedial actions and technologies. Data
quality objectives are developed that define specific requirements of
the data, often expressed in terms of accuracy and precision, dictated
by specific data uses (e.g., general site characterization, regulatory
compliance, risk assessment, technical evaluation of technologies, and/
or legal-evidential uses).
These early efforts in the remedial planning and implementation are
essential to efficient and effective project execution. They form the
basis for detailed planning and execution of remedial investigations,
provide preliminary insight into the site model (sources, media, path-
ways, and receptors) and associated uncertainties, and establish essen-
tial overall direction to the remedial planning and implementation
process based on concepts of potential action.

13.3.2 Remedial investigation (RI)

Remedial investigation activities include field and laboratory studies to
gather site data and estimate the conditions present at the site. Inves-
tigations may include:
• inventory of site facilities and surface features;
• subsurface exploration of the vadose zone, soils and geologic strati-
graphy, and hydrogeologic conditions;
• data gathering related to surface water and hydrology; and
• meteorological, demographic, and ecological data gathering.
For each activity, data on the physical and chemical conditions of
each medium (soil, groundwater, surface water and sediments, and
air) may be obtained.
While many of the RI techniques and their applicability, particularly
in subsurface explorations, are familiar in traditional geotechnical
engineering, remedial investigation applications pose additional con-
cerns, such as:
• health and safety of field personnel;
• equipment decontamination (before, during, and after use);
• sample handling, shipment, and chain of custody;
• media cross-contamination; and
• disposal of exploration-derived wastes.
Remedial investigation includes the compilation and interpretation of
field exploration data, and conduct of a preliminary risk assessment to
302 Strategies for remediation

better define the initial site model. Regulations, especially location- and
contaminant-specific, are identified.
As shown in Fig. 13.2 and discussed further in the following section,
remedial investigation activities should overlap with feasibility study
activities. The overlap and implied interaction and iteration is im-
portant in that it focuses the investigation on obtaining those data that
are necessary and sufficient for remedial planning.
Data gaps may be identified, during the RI, based on needs to better
define site conditions or potential technology performance. While data
gaps may warrant phasing of remedial investigations at complex sites,
experienced judgment should be applied in determining sufficiency
of the data for purposes of remedial planning and implementation.
Improved definition of conditions may always be desirable but must be
balanced against potential costs and delays.

13.3.3 Feasibility study (FS)

As shown in Fig. 13.2, the FS and RI overlap and interact to allow
inclusion of investigation findings into the FS considerations, and
feedback to the RI regarding data needs of specific technologies.
The interaction is essential to avoid obtaining data which, although
interesting in defining site conditions, may be of little or no specific
relevance to potential action.
As a first step in the feasibility study, remedial objectives are
formulated. Objectives typically include the broad requirement of pro-
tection of human health and the environment. More specific objectives,
often based on preliminary identification of regulatory and risk criteria,
must be considered in a range. While it may be desirable to firmly
commit to specific objectives, these can seldom be decided prior to the
effectiveness and cost evaluations resulting from the FS.
Feasibility study includes identification of general response actions
and specific technologies and technology options, assembly of tech-
nologies and options into remedial alternatives, and the evaluation and
comparison of assembled alternatives.
An example of the relationship among general response actions,
technologies, and technology options is illustrated in Fig. 13.3. The
heading 'general response actions' denotes general functions of actions;
'no action' and 'institutional controls' often warrant consideration
along with more aggressive actions. Under the heading 'Technology'
there are general types of action for each function, and in the third
column options describe the more specific remedial processes or means
to be employed. The figure does not attempt to list all potential tech-
nologies and options for soil remediation; more are available, some
proprietary, and continually more are being developed. Thorough con-
 Methodology for remedial planning and implementation 303

GENERAL RESPONSE ACTION TECHNOLOGY OPTION

NO ACTION N/A ~A

INSTITUTIONAL CONTROLS f-------I ACCESS AND USE RESTRICTIONSr--- Deed Restrictions, Fencing, Guards

SOIL REMOVAL ~ EXCAVATION ~ Mechanical Excavation

Native Soil, Clay, Synthetic Membranes,

CAPPING Asphalt" Concrete, Sprayed Asphalt,
Portland Cement Concrete, Multilayer Composite

SOIL CONTAINMENT VERTICAL BARRIERS Slurry Wall (Soil-Bentonite, Soil Cement,

Vibraling Beam), Graul Curtain, Sheel Piling

HORIZONTAL BARRIERS Grout Injection Liners (See Capping)

CHEMICAL Neutralization, Oxidation, Chemical Reduction

Waler Leaching, Solvent Extraction, Stabilization,

SOIL TREATMENT PHYSICAL
Encapsulation, Vitrification, Soil Gas Extraction

BIOLOGICAL Aerobic, Anaerobic

THERMAL Incineration, Low-Temperature Thermal Extraction

SOIL DISPOSAL
~ LANDFILL Onsite (See Containment), Offsita Existing Facility,
r - - - Offsile New Facility

Fig. 13.3 Example general response actions, technologies and options - media
soil.

sideration of remedial actions requires that the geotechnical engineer

stay current with remedial technology developments.
While Fig. 13.3 addresses only soil, the complexity, and number of
potential alternative combinations demonstrate the need for careful
screening prior to assembly of alternatives; the complexity is com-
pounded when other media are involved. Technologies and options
are screened based on general understanding of their applicability to
conditions of the site, effectiveness, implementability, and (to a lesser
degree, at this stage) cost. While the technology/option identification
should include a universe of those available, only those technologies
and options having the greatest potential are carried forward for
assembly into alternatives.
Surviving technologies and options are developed and assembled
into remedial action alternatives that address the full range of media
present at the site and that achieve the range of remedial objectives.
Action/technology-specific regulations are identified and considered in
alternative development.
At this point, it is often necessary or desirable to conduct bench-
or pilot-scale testing of technology options to assist in the detailed
evaluations. Proprietary technology options may also require testing
and may pose special challenges in maintaining vendor confidentiality
while still providing objective, independent assessment of performance.
In detailed evaluation, the assembled remedial alternatives are
304 Strategies for remediation

analyzed and compared considering their effectiveness, implement-

ability, and cost, using information on site conditions and technology
performance obtained during the RI and regulatory and risk criteria
identified during the RI and FS.
Effectiveness involves the degree to which overall protection of
human health and the environment and compliance with regulations
are achieved. Effectiveness should consider both short-term (during
implementation) and long-term (during operation) effects and achieve-
ment. In addition, technical and regulatory preference for reduction
in mobility, toxicity, and/or volume may require that these issues
be specifically addressed in detailed evaluation of each assembled
alternative.
Implementability includes both technical and administrative feasibility
of constructing and operating and maintaining the alternative. For
example, implementability considerations may include capacity of the
site to support necessary construction efforts, availability of current
and future disposal facilities, and/or likelihood of obtaining and
renewing required permits.
Costs for each alternative are prepared to include capital, annual
operation and maintenance (O&M), and present worth. While cost
estimates may, with appropriate allowances and contingencies, gener-
ally be expected to be within -30% to +50% for the scope of alter-
native understood at the time of the estimate, uncertainties in site
conditions, criteria, and action performance along with variations in
market conditions, may result in wide disparity between FS-Ievel cost
estimates and actual costs. Such potential disparity may warrant
inclusion of cost-sensitivity analyses and consideration of cost ranges
in comparing alternatives.
The detailed evaluation of alternatives provides the basis for selection
of a remedial alternative for the site. The process of actual selection,
particularly when such a process involves regulatory approval, is
beyond the scope of this chapter. Although alternative selection is not
illustrated in Fig. 13.2, it usually occurs between FS and remedial design
(RO) and may involve formal regulatory and public participation.

13.3.4 Remedial design and remedial action

Remedial design may involve preliminary design and various stages to
completion of final design (USEPA, June 1986). The process of analyses
and design leading to completion of contract documents (plans and
specifications) that are suitable for construction of the remedy is
well understood in traditional engineering. Preliminary design involves
early summary of the engineering parameters (including regulatory
criteria and monitoring) to be addressed by the remedy; draft plans
and specifications should outline the design concepts and process-flow
 Methodology for remedial planning and implementation 305

diagrams. Final design usually proceeds through a series of staged

completion (at points of various percent complete) to allow review and,
if necessary, mid-course corrections. While the process is similar to
traditional engineering design, several issues warrant special consider-
ation in remedial design, including the following.
• Design investigations may be required to confirm or refine previous
estimates of site conditions.
• Bench-, pilot-, and/or demonstration-scale testing may be required
to refine performance estimates or scale-up factors.
• Obtaining site access and normal permits (as well as specific en-
vironmental permits), because of the character and stigma of waste
remediation, may present unique problems.
• Health and safety of personnel (including construction, operation
and maintenance, and community) may influence both the con-
figuration of the design and the methods of implementation.
A discussion of regulatory reviews during design is beyond the
scope of this chapter. The process for interface and regulatory review,
if required, should be established early in the remedial design.
Because of inherent uncertainties and the associated need for modifi-
cations during construction and implementation, contracts should
clearly define procedures for change orders and claims. Careful con-
sideration should be given to the use of performance versus method
specification. Performance specification may often be preferable but
must be consistent with equitable payment and assignment of liability
established in the construction documents.
Procurement for construction services should consider both the
technical requirements and other support needs (such as personnel
health and safety). Highly specialized activities or proprietary tech-
nologies may warrant inclusion of prequalification stipulations. While
openly advertised competitive bids (with contract award based on
price) are common, contracts procured on a negotiated price basis may
be appropriate.
Remedial implementation may include construction or implementa-
tion of the designed remedy, operation and maintenance of the
remedy, and monitoring of remedy performance.

13.3.5 Potential shortcomings of the methodology

The remedial planning and implementation methodology, previously
described, follows the study-de sign-build paradigm similarly applied in
other engineering and scientific endeavors. While the methodology is
basically sound and appropriate, practical difficulties and delays have
been experienced in its application to waste remediation efforts. Often
the difficulties result in failure to move forward in the methodology in
306 Strategies for remediation
the face of the inherent, unavoidable uncertainty associated with waste
remediation projects. The consequence of stalemate can be evidenced
beyond individual projects and the technical community (Wallace,
1987).
A shortcoming of the methodology is the implied separation of the
decision-making process into specific steps (i.e., RIIFS, RD, RA) and
implied finality of the decisions made at each step. In addition, the
methodology may imply a remedial goal is 'reached' and the process
complete, at (or soon after) completion of construction (Brown et al.,
1990).
The combined effect may manifest itself within waste remediation
projects as:
• hesitancy to establish objectives for remediation;
• insistence that additional data be gathered - often this is coupled
with the mistaken impression that additional data will eliminate
uncertainty.
Projects that do move forward in the absence of thorough under-
standing of the inherent uncertainties may experience unanticipated
problems and be without a course for expeditious response. Still more
dire is the potential that unanticipated conditions may continue or
worsen, yet remain unnoticed during the remedial effort.
Potential shortcomings of the methodology warrant recognition and
understanding. The shortcomings can be minimized if supplemented
with a previously established approach for decision-making uncertainty
(Dunnicliff and Deere, 1984; Duplanic and Buckle, 1989; US National
Research Council, 1989; Wallace and Lincoln, 1990; Brown et aI., 1990;
0' Appolonia, 1990).

13.4 APPLICATION OF THE OBSERVATIONAL METHOD

Peck (1969) articulated an approach for dealing with decisionmaking

under the uncertainties inherent in traditional geotechnical engineering.
The approach, phrased the observational method, has been applied
with success in geotechnical engineering in situations that are remark-
ably similar to waste remediation:
Whenever ... some unexpected development has occurred, or
whenever a failure of accident threatens or has already taken
place, an observational procedure may offer the only satisfactory
way out of the difficulties. Under these circumstances perhaps
most engineers would instinctively adopt such a procedure. The
mere observation of events ... often suggests remedial measures
that prove successful. Yet the results are sometimes disappointing
 Application of the observational method 307

and occasionally disastrous because the observations do not con-

stitute part of a well-considered programme encompassing all of
the applicable steps in the complete ... procedure.
(Peck, 1969, p. 173)

13.4.1 General description of the observational method

Peck summarizes the key elements of the complete observational
method in the following ingredients:
(a) Exploration sufficient to establish at least the general nature,
pattern and properties of the deposits, but not necessarily in
detail.
(b) Assessment of the most probable conditions and the most
unfavorable conceivable deviations for these conditions. In this
assessment, geology often plays a major role.
(c) Establishment of the design based on a working hypothesis of
behavior anticipated under the most probable conditions.
(d) Selection of the quantities to be observed as construction pro-
ceeds and calculation of their anticipated values on the basis of
the working hypothesis.
(e) Calculation of the values of the same quantities under the most
unfavorable conditions compatible with the available data con-
cerning the subsurface conditions.
(f) Selection in advance of a course of action or modification of
design for every foreseeable significant deviation of the obser-
vational findings from those predicted on the basis of the
working hypothesis.
(g) Measurement of quantities to be observed and evaluation of
the actual conditions.
(h) Modification of design to suit actual conditions.
(Peck, 1969, p. 173)
Knowledgeable application offers great opportunity for savings of
time and money, and may, in many situations encountered in waste
remediation, offer the only rational way to proceed to confident
remedy.

13.4.2 The observational method in waste remediation

Successful incorporation of the observational method into the method-
ology for waste remediation (section 13.3) requires that the process of
remedial planning and implementation be recognized as a continuum.
That is, the interpretations of conditions, criteria, and performance that
are made in the process are merely working hypotheses. Reassessment
of the hypotheses continues from pre-RIIFS through implementation;
308 Strategies for remediation

throughout the process, additional data needs are assessed based

on their potential to disprove, as suggested by Leonards (1979), the
hypotheses. Decisions are finalized only after the remedy is in place
and sufficient monitoring has been conducted to confirm the hypoth-
eses of conditions and performance.
Figure 13.4 illustrates the remedial process incorporating the obser-
vational method. Compared to Fig. 13.3, the process is similar in its
general sequence and flow. While incorporation of the observational
method into the process maintains basic logic of the study-de sign-build
paradigm, significant yet subtle differences should be noted.
In the remedial investigation, conditions and criteria are interpreted
for both expected conditions and potential deviations. Models (whether
conceptual or analytical) may provide essential insight into the poten-
tial range of conditions to be addressed by remediation. This range
represents the basis for alternative development.
In the feasibility study, technologies and options as well as as-
sembled alternatives are identified and evaluated based on their ability
to address the range conditions and criteria. Models of alternative
performance (whether conceptual or analytical) provide the working
hypothesis for the remediation and, more specifically, provide the
essential performance prediction necessary to guide the selection of
parameters to be monitored. Mark et al. (1989) provide a case study of
the application of the observational method through RIfFS.
Remedial design involves design of the remedy for both expected
conditions and potential deviations, prediction of remedy performance

f-ol·_ _ _ RI / FS .1. 1. - R A - - - - - f -
R D----O.-I-

REMEDIAL INVESTIGATION
REMEDIAL
Conduct Field Evaluate
DESIGN
Investigations Expected
Identify Lacatian- Conditions & Design Action
Deviations
I~ Specific
& Contaminant & Monitoring
Regulations ~ Predict :::) REMEDIAL
ACTION
PRE-RI/FS Perform Bench Performance

r
Perform Preliminary and/or Implement
Conduct Risk Assessment Pilot Tests Determine
Preliminary Site Initial
Course of Design
Inspection and ';> ~ Modifications

L
Assessment ----" Implement
I I
Callect & Analyze I I Design
Existing Data "7 "7
<7 I I Modifications
I I
Formulate FEASIBILITY STUDY I I OPERATION/
I I
& Plan
Site-Specific
Identify & Screen Analyze <~ MONITORING
Technologies Alternatives
Approach
Assemble Various for Expected
Conditions &
r- EVALUA TE O&M
DATA &. MODIFY '-------v
Monitor Action
Performance
Technologies
Deviations DESIGN PER Against
into Alternatives
PREDETERMINED Precictions
Identify Action/ Compare &. COURSE
Technology- Specific Contrast
RegUlations Alternatives

Fig. 13.4 Schematic of the observational method.

 Application of the observational method 309
over the range of conditions and criteria, design of a monitoring system
to measure specific conditions and performance, and predetermination
of a course of action for each potential deviation.
Remedial action involves construction and implementation of the
design for the expected conditions and implementation of the moni-
toring system. It is essential that the data obtained during construction
and implementation be analyzed and compared to predictions in a
timely fashion and that feedback evaluations (as illustrated in Fig. 13.4)
provide the ability for expeditious modification of the action.

13.4.3 Cautions in application of the observational method

Application of the observational method requires that the remedial
process be viewed as a continuum in which action is planned and
implemented with prior consideration for the uncertainties involved.
The complete process demands anticipation of specific unfavorable
conditions, prediction of specific performance, specific monitoring
against each, and predetermination of a course of action to be followed
based on the results of the monitoring. Only with thorough and
knowledgeable incorporation of these elements is the observational
method complete; any lesser approach that implements and then reacts
to the results is not the observational method. The observational
method is only viable if implementation can be modified based on the
results of monitoring. Success of the method requires that the right
quantities be monitored. The potential for progressive failures must
always be considered in that they may not allow sufficient time for
modification based on observations.
While the observational method offers great opportunity for im-
proving the approach and methodologies of waste remediation plan-
ning and implementation, it is not new. Indeed, the adaptation of the
observational method to waste remediation illustrates that there is
nothing new except the history that we never learned. The challenge is
clear to geotechnical engineers who understand the method.
While the ingredients may appear to offer a proceduralized, step-
wise process for project execution, the observational method actually
represents a philosophy more than a methodology. Peck, in the intro-
duction to his 1983 presentation of the Rankine Lecture (Dunnicliff and
Deere, 1984), expressed his concern, which is applicable to the adapta-
tion of the method to waste remediation: ' ... I still feel that my efforts
to formalize the observational method were too contrived, too rigid.'
The observational method is not as difficult as it may appear when
presented in procedures and methodologies intended to 'clarify' the
method. Such efforts may be useful to communicate or illustrate the
method, but they seldom 'clarify' it. We must be careful that the simple
and elegant fundamentals of the method are not clouded or lost com-
pletely in the efforts to convert a 'way of thinking' to a 'way of doing'.
310 Strategies for remediation

REFERENCES

Brown, S.M., Lincoln, D.R. and Wallace, W.A. (1990) Application of the
observational method to hazardous waste engineering, J. Management
Eng., ASCE, 6(4), 479-500.
CERCLA (1980) 96th US Congress Comprehensive Environmental Response,
Compensation, and Liability Act of 1980 (CERCLA), Public Law 96-510,
December 11th.
D' Appolonia, E. (1990) Monitored decisions, twenty-fourth Karl Terzaghi
lecture, J. Geotech. Eng., ASCE, 116(1), 4-34.
Dunnicliff, J. and Deere, D.U. (1984) Judgement in Geotechnical Engineering - The
Professional Legacy of Ralph B. Peck, John Wiley & Sons.
Duplanic, N. and Buckle, G. (1989) Hazardous data explosion, Civ. Eng., Dec.
68-70.
Leonards, G.A. (1979) Discussion of: foundation performance of Tower of Pisa,
J. Geotech. Eng. Dip., ASCE, I05(GTl), 95-105.
Mark, D.L. et al. (1989) Application of the observational method to an operable
unit feasibility study - a case study, Proc. of Superfund '89, Hazardous
Materials Control Research Institute, Silver Spring, Maryland, pp. 436-
442.
Peck, R.B. (1969) Ninth Rankine lecture, advantages and limitations of the
observational method in applied soil mechanics, Geotechnique, 19(2), 171-
87.
SARA (1986) 99th US Congress, Superfund Amendments and Reauthorization Act of
1986 (SARA).
USEPA (June 1986) Superfund Remedial Design and Remedial Action Guidance,
OSWER Directive 9355.0-4A.
USEPA (1988) Superfund Exposure Assessment Manual, OSWER Directive 9285.
5-1.
USEPA (June 1988) Community Relations in Superfund, Interim Version, OSWER
Directive 9230.0-3B.
USEPA (August 8, 1988) CERCLA Compliance With Other Laws Manual,
OSWER Directive 9234.1-01, Draft.
USEP A (October 1988) Guidance for Conducting Remedial Investigations and
Feasibility Studies Under CERCLA, OSWER Directive 9335.3-01, Interim
Final.
USEP A (March 1989) Risk Assessment Guidance For Superfund - Environmental
Evaluation Manual, OSWER Directive 9285.7-01, Interim Final.
USEPA (September 1989) Risk Assessment Guidance For Superfund Volume I:
Human Health Evaluation Manual, OSWER Directive 9285.701a, Interim
Final.
US National Research Council (1989) Commission on Engineering and
Technical Systems, The Geotechnical Board, Geotecimology - Its Impact on
Economic Growth, the Environment, and National Security, National Academy
Press.
Wallace, W.A. (1987) The cleanup of hazardous waste sites in the United
States: positioning for least cost, Proc. World Conf. on Haz. Waste,
Budapest, Hungary, Elsevier Science Publishers, B.V.
Wallace, W.A. and Lincoln, D.R. (1990) How Scientists Make Decisions About
Groundwater and Soil Remediation, National Academy Press.
 CHAPTER 14

Geophysical techniques
for subsurface site
characterization
Richard C. Benson

14.1 INTRODUCTION

Geophysical methods encompass a wide range of surface and down-

hole measurement techniques which provide a means of investigating
subsurface hydrogeologic and geologic conditions. These methods
have also been applied to detecting contaminant plumes and locating
buried waste materials.
The primary factors which limit the accuracy of site characterization
are the limitations of the number of sample points, and the lack
of correct spatial sampling. Geophysical measurements can be made
relatively quickly, thereby providing a means to increase sample den-
sity. Continuous data acquisition along a traverse line can be employed
with certain techniques at speeds up to several kilometers per hour.
Because of the greater sample density, anomalous conditions are more
likely to be detected, resulting in a more accurate characterization of
subsurface conditions.
All of the geophysical methods, like any other means of measure-
ments, have advantages and limitations. There is no single, universally
applicable geophysical method, and some methods are quite site specific
in their performance. Thus, the user must select the method or
methods carefully and understand how they are applied to specific site
conditions and project requirements.
Unlike direct sampling and analysis, such as obtaining a soil or
water sample and sending it to a laboratory, the geophysical methods
provide non-destructive, in situ measurements of physical, electrical or
geochemical properties of the natural or contaminated soil and rock.

Geotechnical Practice for Waste Disposal.

Edited by David E. Daniel.
Published in 1993 by Chapman & Hall, London. ISBN 0 412 35170 6
312 Geophysical techniques for subsurface site characterization

The success of a geophysical method depends on the existence of

a contrast between the measured properties of the target and back-
ground conditions. If there is no measurable contrast in the measured
property, the target will not be detected. Furthermore, if a layer is
sufficiently thin, or if the size of a localized target is sufficiently small
or if it is too deep, it may not be detected.
Geophysical methods are not new; they have been used for many
decades in the exploration for deep oil and gas as well as mineral
exploration. The methods have also been applied to geotechnical
applications and regional water resources development for the last
few decades. The application of the methods to hazardous waste site
investigations are more recent, beginning in the mid to late 1970s.
Applications to the engineering geology, geotechnical application and
hazardous waste sites are somewhat different in their application than
oil, gas and mineral exploration because they are used to obtain
relatively shallow data (typically less than 30 m or so in depth) and
often are required to produce higher resolution.

14.2 BACKGROUND
Often, traditional approaches to subsurface field investigations at
waste disposal sites have been inadequate. Site investigations rely only
upon the use of direct sampling methods such as:
• soil borings for soil and rock samples;
• monitoring wells for gathering hydrogeologic data and water
samples;
• laboratory analysis of soil and water samples to provide a quantitative
assessment of site conditions; and
• extensive interpolation and extrapolation from these points of data.
This approach has evolved over many years in the geotechnical areas
and is commonly considered a standard in which to deal with field
investigations. However, there are numerous pitfalls associated with
this approach, which can result in an incomplete or even erroneous
understanding of site conditions. These pitfalls have been the subject
of numerous papers and conferences over the years (Lysyj, 1983;
Hileman, 1984; Perazzo et al., 1984; Walker and Allen, 1984; Dunbar et
al., 1985). They have also precipitated the tightening of groundwater
monitoring regulations (USEPA, 1985).
The single most critical factor faced in site evaluation work is
accurately characterizing the site's hydrogeology (Benson and Pasley,
1984). With an accurate understanding of a sites hydrogeology, pre-
dicting the location of movements of contaminants, or designing a
clean-up operation would be reasonably straightforward. If all sites
were simple, horizontally stratified geology with uniform properties,
 Background 313

site characterization would be easy, since data from just one boring
would be sufficient to characterize the site. However, in most geologic
settings, this will not be the case. Even at sites whose geology appears
to be uniform there are sometimes subtle variations which can cause
significant errors in site characterization.
In the design of many soil and rock sampling programs and monitor-
ing well networks, the placement of borings and wells has been done
mainly by educated guesswork. The accuracy and effectiveness of
such an approach is heavily dependent upon the assumption that
subsurface conditions are uniform. This approach usually assumes
that regional hydrogeology and groundwater flow (as obtained from
literature) is valid for the site specific setting, and that data from only a
few site specific borings or monitoring wells can be used to characterize
the site. These assumptions are frequently invalid, resulting in non-
representative locations for borings piezometers and monitoring wells
and erroneous generalizations from limited information. To improve
the accuracy of the site investigation, a large number of borings may be
required.

14.2.1 Sample density

A soil or core sample obtained from drilling may be representative of
only a limited area surrounding the hole. Therefore, fractures, cavities,
bedrock channels, sand lenses, and local permeable zones can easily be
missed by borehole programs.
An insight into the number of discrete samples or borings that are
required for accurate site characterization can be obtained by con-
sidering detection probability (Benson and LaFountain, 1984). Fig.
14.1a shows a target area which is 1110 of the total site area. This 'target
area' (whose size and location are usually unknown) could be a waste
burial site, a plume from a chemical spill, or an old sinkhole. Based
upon probability calculations, the number of samples or borings
required to achieve various detection probabilities are shown in Fig.
14.1b. For example, 10 uniformly spaced borings would produce a 10%
probability of detection for a site-target of 10. For smaller targets, such
as a narrow sand lens or fractures, the site-to-target area ratio will
increase significantly, thus 100 to 1000 borings may be required to
achieve a confidence level of 90% making the subsurface investigation
like 'looking for a needle in a haystack'.
It becomes obvious that achieving a good statistical evaluation of
complex hydrogeologic site conditions requires samples or borings to
be placed in a close-order grid, which would reduce the site to 'Swiss
cheese'. In many cases, direct sampling alone is not sufficient to char-
acterize accurately site conditions. This is the primary reason for the
application of the geophysical methods.
314 Geophysical techniques for subsurface site characterization

Area of Target ~
As
Area of Site = As ~ 10
AT

Probability
of Detection AS/At = 10 As/At = 100 As/At = 1000

100 16 160 1600

98 13 130 1300
90 10 100 1000
75 8 80 800
50 5 50 500
40 4 40 400
30 3 30 300

Fig. 14.1 Spatial sampling requirements; (a) illustrates site to larger area ratio
- AslAt = 10 in this case; (b) number of sample points required for various
As/At ratios and probability of detection - this table assumes uniform sample
grid; if a random placement is used, the number of samples must be increased
by a factor of 1.6. Source: Benson and LaFountain (1984).

14.2.2 How geophysical methods are used

To provide a greater volume of measurement

Data obtained from borings or monitoring wells come from a very
localized area and are representative of material conditions at the bore-
hole. Geophysical methods, on the other hand, usually measure a
much larger volume of the subsurface material (Fig. 14.2).

As anomaly detectors
As a result of geophysical measurements being relatively rapid, a larger
number of measurements can be taken, for a given budget. With a
greater number of measurements plus the fact that the measurement
 Background 315

I
I
j6 I
I

Fig. 14.2 Comparison of volumes sampled by geophysical method and a

borehole: (left) a typical geophysical measurement integrates a larger volume of
soil and rock while (right) the volume of soil and rock sampled by drilling is
relatively small.

encompasses a larger volume of subsurface material, the geophysical

methods can detect anomalous areas which may pose potential prob-
lems, and thus are essentially anomaly detectors.
Once an overall characterization of a site has been made using
geophysical methods and anomalous zones identified, drilling and
sampling programmes are made more effective by:
• locating boreholes and monitoring wells to provide samples that are
representative of both anomalous and background conditions;
• minimizing the number of borings, samples, piezometers and
monitoring wells required to characterize accurately a site;
• reducing field investigation time and cost; and
• significantly improving the accuracy of the overall investigation.
This approach yields a much greater confidence in the final results,
with fewer borings or wells, and an overall cost savings. It makes good
sense to minimize the number of monitoring wells at a site while
optimizing the location of those installed. If the wells are located in the
wrong position, they do not provide representative data and a large
amount of relatively useless data would accrue.
Using the geophysical method in a systematic approach, drilling is
no longer being used for hit-or-miss reconnaissance, but is being used
to provide the specific quantitative assessment of subsurface con-
ditions. Boreholes or wells located with this approach may be thought
of as smart holes because they are scientifically placed, for a specific
purpose, in a specific location, based on knowledge of site conditions
from geophysical data (Benson and Pasley, 1984). While smart holes
might sometimes be placed without the use of geophysical methods,
they often can be placed more reliably if the geophysical methods are
incorporated into the subsurface investigation program.
316 Geophysical techniques for subsurface site characterization

If piezometers and wells have already been installed at the site, the
geophysical methods can still provide significant benefits. The location
of existing borings and monitoring wells relative to background, down-
gradient and anomalous site conditions can be assessed, thus providing
a means of evaluating the representativeness of existing data. Then, if
additional borings or wells are needed to fill gaps in the overall site
coverage, they can be placed accurately as smart holes.

To aid in understanding the overall hydrogeologic setting

Assessment of site conditions will often require that an area larger than
the site itself be considered. Contaminant transport by groundwater
and the geohydrologic factors which control groundwater flow, do not
stop at arbitrary site boundaries or property lines. Insight into the
character of local hydrogeologic conditions must be derived from the
knowledge of the broader understanding of the overall hydrogeologic
setting. Omitting an adequate understanding of the overall hydro-
geologic setting will often result in a number of critical gaps in infor-
mation about the site itself. The geophysical methods provide a means
of rapid reconnaissance over larger areas, and often can be employed
to obtain the data to support an understanding of the overall setting, as
well as to provide site specific data.

To provide continuous measurements when possible

Geophysical surveys often involve making measurements of subsurface
properties at discrete points over a site. That is, the instrumentation is
located at a station along a survey line or a grid and measurements are
made one point at a time. The data is then analyzed and plotted to
provide profile lines on contour maps. The degree of detail of such
data is proportional to the number of measurements made. However,
some geophysical methods can provide measurement of subsurface
parameters continuously as the instrumentation moves along the survey
line.
Continuous methods should be employed whenever possible to
maximize the amount of data obtained, to achieve maximum resolution
and to minimize project costs. This approach is particularly necessary
when site conditions are suspected of being highly variable and small
features such as fractures need to be identified.
Although the continuous surface geophysical methods referred to
in this chapter (see section 14.4) are typically limited to a depth of
15 m or less, they are applicable to many site investigations. They can
provide a continuity of subsurface information, not practically obtain-
 Background 317

a) II STATION MEASUREMENTS
(APPROX. 140' APART)

b) CONTINUOUS
MEASUREMENT

Fig. 14.3 Comparison of (a) station and (b) continuous measurements from
the same site. Note: the data was obtained with a truck-mounted EM-34 with a
lO-meter coil spacing. The higher values of electrical conductivity are caused by
moisture within the underlying gypsum rock.

able from station measurements. Continuous surface geophysical

methods can be applied at speeds of I-Skm/hr or more, resulting in a
cost-effective approach for relatively shallow survey work.
The benefits of continuous measurements are seen by comparing
station measurements and continuous measurements shown in Fig.
14.3. The set of data in Fig. 14.3b reveals the highly variable nature of a
site indicated by a continuous measurement technique, while those in
Fig. 14.3a shows the loss of information that can result from making
a limited number of station measurements at the same site and inter-
polating between sample points. Limiting the number of measure-
ments results in a distorted set of data and can lead to errors in
interpretation of site conditions especially when the target size is sig-
nificantly smaller than station spacing.
The data in Fig. 14.3 was obtained by electromagnetic measurements
of subsurface electrical conductivity. The higher conductivity values in
Fig. 14.3b indicate fractures within underlying gypsum rock. These
fractures show up because they are more electrical conductive due
to water content and weathering of the gypsum rock. The fracture
provides a pathway for groundwater.
By running closely-spaced parallel survey lines with continuous
methods, subtle changes in subsurface parameters can often be mapped
and total site (100%) coverage can be achieved if necessary.
318 Geophysical techniques for subsurface site characterization

Fig. 14.4 Modes of geophysical measurements; (a) surface and (b) downhole
geophysics.

Site investigation methods are scale dependent

All site investigation methods, including geophysics, are scale depen-
dent. For example, aerial photography is an effective tool to be used in
obtaining the big picture at a local site investigation, but, it will not
provide any information about site-specific soil conditions at a depth
of 3m. Conversely, a boring will provide detailed information on soil
or rock conditions versus depth, but information from the boring is
only valid for a very limited extent immediately around the borehole.
Surface geophysical measurements (Fig. 14.4a) can be used to determine
detailed soil and rock conditions over a few tens of square meters or
over many square kilometers, while geophysical logging measure-
ments made down a borehole will extend the measurements from the
borehole itself radially to a distance of several centimeters to I-2m
depending upon the method used (Fig. 14.4).
Therefore, the site investigation methods employed must be selected
to suit the particular scale of data needed to meet project requirements
(Benson and Scaife, 1987). A complete subsurface investigation must
include measurements of the big picture (aerial photography), inter-
mediate picture (surface geophysical measurements), and the very local
details (boring, logging and sampling data).

Summary
Geophysical methods:
• usually measure a larger volume than direct sampling;
• are very effective at detecting and mapping anomalies;
• are used to locate or assess representative locations of boreholes,
piezometers, and monitoring wells;
• can be used for reconnaissance measurements to aid in character-
izing the regional setting;
• can be used for gathering both reconnaissance and detailed site
specific data;
 Applications for geophysical measurements 319

• provide a means for obtaining continuous measurements, both on

the surface and downhole; and
• aid in integrating the wide range of scale dependent measurements
from aerial photography to core samples.

14.3 APPLICATIONS FOR GEOPHYSICAL MEASUREMENTS

There are three major areas for the application of geophysical methods
at waste disposal sites. They are:
• assessing hydrogeologic conditions;
• detecting and mapping contaminant plumes; and
• locating and mapping buried wastes and utilities.

14.3.1 Assessing hydrogeologic conditions

Probably the most important task of a site investigation will be assess-
ing the natural hydrogeologic conditions. A variety of geophysical
methods can be used to determine depth to bedrock, degree of
weathering, sand and clay lenses, fracture zones, and buried relic
stream channels (Keys and MacCary, 1976; Benson and Glaccum, 1979;
Benson et al., 1982; Benson and Yuhr, 1987). Accurately understanding
the hydrogeologic conditions and anomalies can make the difference
between success and failure in site characterization, because these fea-
tures will often control groundwater flow and contaminant transport.

14.3.2 Detection and mapping contaminant plumes

A major objective of many site investigations is the detection and
mapping of contaminant plumes, and geophysical methods can be
employed in two ways to solve this problem. Some methods can be
used for the direct detection of contaminants, but in those cases where
the contaminant cannot be detected directly, geophysical methods can
be used to assess detailed hydrogeologic conditions which are control-
ling groundwater flow, such as fractures or a buried channel. Once
the contaminant flow pathways have been identified, direct sampling
methods can be used to further assess conditions (Cartwright and
McComas, 1968; McNeill, 1980; Benson et al., 1982, 1985; Greenhouse
and Monier-Williams, 1985).

14.3.3 Locating and mapping buried wastes and utilities

Geophysical methods can also be used to locate and map the areal
extent and sometimes the depth of buried wastes in trenches and
320 Geophysical techniques for subsurface site characterization
landfills (Benson, et al., 1982). There are also methods which can be
employed to detect buried drums, tanks, and utility lines. In many
cases the trenches associated with pipes and utilities will be of interest
because they form a permeable pathway for contaminant migration.

14.4 SURFACE GEOPHYSICAL METHODS

The surface geophysical methods discussed in this section are

• ground penetrating radar
• electromagnetics
• resistivity
• seismic refraction
• seismic reflection
• micro-gravity
• metal detection
• magnetics.
While these methods are not the only ones which may be em-
ployed, they are presented as the primary methods because they
are regularly used and have been proven effective for waste site
assessments.

14.4.1 Ground penetrating radar

Ground penetrating radar uses high frequency electromagnetic waves
from less than 100 Mhz to 1000 Mhz to acquire subsurface information.
Energy is radiated downward into the ground from a transmitter and is
reflected back to a receiving antenna. The reflected signals are recorded
and produce a continuous cross-sectional picture or profile of shallow
subsurface conditions.
Reflections of the radar wave occur whenever there is a change in
dielectric constant or electrical conductivity between two materials.
Changes in conductivity and in dielectric properties are associated
with natural hydrogeologic conditions such as bedding, cementation,
moisture, clay content, voids, and fractures. Therefore, an interface
between two geologic layers which have a sufficient contrast in electric
properties will show up in the radar profile (Benson and Glaccum,
1979; Benson et al., 1982; Benson and Yuhr, 1987).
The radar record is similar to the view we would get if we observed
the cross-section of soils in a trench or a cross-section of rock at a road
cut. Figure 14.5 shows a radar record (cross-section) of a clean quartz
sand over a clay loam. Note the level of detail that can be obtained in
mapping the top of the clay loam. The numerous small hyperbola in
the upper portion of the record, within the quartz sands, are due to
tree roots or animal burrows.
 Surface geophysical methods 321
Note Numerous Hyperbola

I
2 Meters I

1 FINE
QUARTZ
SAND

11-04..------- Approximately 100 meters -----~.I

l~~~~
Fig. 14.5 Radar profile of quartz sand over clay loam. Note the level of detail
that can be obtained.

The vertical scale of the radar profile is in units of time (nano-

seconds, or 10- 9 sec). The time it takes for an electromagnetic wave to
move down to a reflector and back to the surface (two-way travel time)
is relatively short since the waves are travelling at almost the speed of
light. The time scale then is converted to depth by making measure-
ments or assumptions about the velocity of the waves in the subsurface
materials.
Depth of penetration of the radar wave is highly site-specific. The
method is limited in depth by attenuation due to the higher electrical
conductivity of subsurface materials or scattering. Generally, radar
penetration is greater in coarser, dry, sandy or massive rock, and less
in wet, fine grained clayey (conductive) soils. However, data can be
obtained in saturated materials if the specific conductance of the pore
fluid is sufficiently low. For example, radar has been applied to map
the sediments in fresh water lakes and rivers. While radar penetration
in soil and rock to more than 30 m has been reported, penetration of
5-10 m is more typical. In silts and clays, penetration may be limited to
a meter or less. Yet, in some situations, useful results can be obtained
in silts and clays (Benson and Yuhr, 1990). The water table can be
detected in coarser grained materials but not in finer grained sediments
with a large capillary boundary. Both metallic and nonmetallic buried
pipes and drums can also be detected.
322 Geophysical techniques for subsurface site characterization
The continuous data produced by the radar method offers a number
of advantages over some of the other geophysical methods. Continuous
profiling permits data to be gathered much more rapidly, thereby
providing a large amount of data for a given budget. In some cases,
total site coverage of an area can be obtained. Radar data may be
obtained at speeds up to 5-10km/hr or more. Very high lateral resolu-
tion data can be obtained by towing the antenna by hand at much
slower speeds (less than 1 km/hr).
Radar has the highest resolution of all of the surface geophysical
methods. Vertical resolution of radar data can range from a few
centimeters to about 1 m depending upon the depth and the frequency
used. A variety of antennas can be selected to cover frequencies from
less than 50 Mhz to 2000 Mhz. Lower frequencies provide greater
depths of penetration with lower resolution and the higher frequencies
provide less penetration with higher resolution.
Two types of radar systems are currently available. In the first
system, a separate transmitter and receiver antenna are held in contact
with the ground by either one or two persons at some given distance
apart. The radar data are gathered and then the two antennas picked
up and moved some increment of distance, put in contact with the
ground and a new set of radar data acquired. Each set of data is
recorded on station by station basis and stored in a lap-top computer.
Once the field data has been acquired in this manner, the data is
processed on a lap-top computer, similar to a seismic reflection data.
This station measurement radar system is probably effective in
mapping slightly deeper geologic horizons and horizons with less
reflection coefficient in which the radar data can be improved by
repeated sampling (stacking) and subsequent processing. It is also
usable in areas with brush and trees where a continuous radar antenna
could not be towed. In the second system, the continuously towed
radar system, the antenna is moved continuously across the ground
recording data (real time) producing a graphic record (cross-section).
This system has much higher spatial resolution, both laterally and
vertically since it acquires infinitively more data.
Preliminary field analysis of radar data is possible using the picture-
like record. Despite its simple graphic format, there are many pitfalls in
the interpretation of radar data. There are multiple bands within the
data due to ringing, and these may obscure layers and cause confusion
in interpretation by untrained persons. Overhead reflections from trees
or power lines may also appear on the record when not using shielded
antennas (although this is only a problem with lower frequency
antennas). System and geologic noise can sometimes clutter up the
record, making interpretation difficult. Although radar can be recorded
on magnetic media and processed by computer the necessary infor-
mation can usually be derived from the raw graphic records.
 Surface geophysical methods 323

Approximately 915 meters

Fig. 14.6 Continuous EM profile measurements show a large inorganic plume

(center rear) and considerable natural geological variation.

14.4.2 The electromagnetic and resistivity methods

The electromagnetic (EM) and resistivity methods are similar in the
sense that they both measure the same parameter, but in different
ways. Electrical conductivity values (millisiemens/meter or millimhos/
meter) are the reciprocal of resistivity values (ohm/meter) although
units of ohm-feet are also common in the US. Electrical conductivity (or
resistivity) is a function of the type of soil and rock, its porosity and the
fluids which fill the pore spaces. The specific conductance of the pore
fluids often dominates the measurement. Both methods are applicable
to the assessment of natural hydrogeologic conditions (Griffith and
King, 1969; McNeill, 1980; Telford et al., 1982).
Natural variations in subsurface conductivity, or resistivity, may be
caused by changes in basic soil or rock types, thickness of soil and rock
layers, moisture content, and depth to water table. Localized deposits
of natural organics, clay, sand, gravel, or salt-rich zones will also affect
these values, while structural features such as fractures or voids can
also produce changes in them.
The absolute values of conductivity (or resistivity) for geologic
materials are not necessarily diagnostic in themselves, but their spatial
variations both laterally and with depth, can be significant. It is the
identification of these spatial variations or anomalies which enable the
electrical methods to find potential problem areas rapidly (Fig. 14.6).
Because the specific conductance of the fluids in the pore spaces can
324 Geophysical techniques for subsurface site characterization

340
a)

DISPOSAL SITE

340

~
Shallow Measurements of
Landfill Leachate Plume
(0 to 15 feet deep)

b)

N
240

r
DISPOSAL SITE

Deep Measurements of
Landfill Leachate Plume
(0 to 45 feet deep)

Fig. 14.7 Resistivity map of leachate plume from a landfill. Values are in
ohm-feet; landfill is approx. 1 sq mile. (a) shallow measurements of landfill
leachate plume (0-15 ft deep); (b) deep measurements of landfill leachate
plume (a-45ft deep).

dominate the measurements, detection and mapping of contaminant

plumes can often be accomplished using the electrical methods. Because
in organics in sufficient concentrations are often more electrically con-
ductive than groundwater, both the lateral and vertical extent of an
inorganic plume can be mapped using the electrical methods. Cor-
relation between groundwater chemistry data and results using elec-
trical methods to map inorganics from landfills has been as good as
 Surface geophysical methods 325
0.96 at the 95% confidence level (Benson et ai., 1985). Electrical
methods provide a means of directly mapping the extent of the
inorganic contaminants in situ, obtaining direction of flow, and
estimating concentration gradients (Fig. 14.7). These measurements
can also be used for time-series measurements to obtain plume
dynamics, and thus provide vital information for the modeling of
groundwater flow (Benson et ai., 1988).
If the contaminate plume consists of a mix of organics and inorganics,
such as leachate from a landfill, a first approximation to the distribu-
tion of the organics can be made by using electrical methods to map
the more electrically conductive inorganics (Fig 14.7). Correlation
between groundwater chemistry data for total organic carbon in a
landfill leachate and results using electrical methods has been as good
as 0.85 at the 95% confidence level (Benson et ai., 1985).
In cases where pure organics such as trichloroethylene (TCE) exist,
electrical as well as other geophysical methods can be used to define
permeable pathways or buried channels through which the contami-
nants may migrate. Direct detection of hydrocarbons can sometimes be
accomplished by looking for a conductivity low (resistivity high) asso-
ciated with the organics. The possibility for such an anomaly exists
where large amounts of hydrocarbons have been in place for a long
period of time and there is a sufficient contrast in electrical values
between the natural background values and the hydrocarbons. To date
this approach has had limited success.
Both EM and resistivity methods may be used to obtain data by
'profiling' or 'sounding'. Profiling provides a means of mapping lateral
changes in subsurface electrical conductivity (or resistivity) at a given
depth (Figs. 14.6 and 14.7). Profiling measurements are made by
obtaining data at a number of stations along a survey line, the spacing
between the measurements depending upon the variability of the
setting and upon the lateral resolution desired. At each station along
the profile line, data may be obtained for one depth or a number of
depths depending upon project requirements. It is useful to take at
least two measurements, a shallow one and a deeper one, so that the
influence of the highly variable shallow soils and cultural influences
can be assessed (Fig. 14.7). Profiling is well suited to the delineation
of hydrogeologic anomalies, mapping of contaminant plumes and
location of buried material.
The sounding method provides a means of determining the vertical
changes in electrical conductivity (or resistivity) correlating with soil
and rock layers. In this case, the instrument is located at one location
and measurements made at increasing depths. Interpretation of
sounding data provides the depth, thickness and conductivity (or
resistivity) of subsurface layers with different electrical conductivities
(or resistivities); (Fig. 14.8).
326 Geophysical techniques for subsurface site characterization

Resistivity
Sounding Drillhole
- 0 Meters
Groy Sand
f = 13,000 Tan Sand
-I

Sand
-2
Tan Sand wlClay
p= 6500 Red Sandy Clay
-3

Clayey Sand -4
Dark Tan
Sandy Cloy
-5

Ton Plastic Cloy

-6

p= 18 Gray Plastic Cloy

-7

Massive Clay
-8

-9

~ -10

End of Hole -II

-12

f = 300 -13

Limestone -14

-15

Fig. 14,8 Resistivity geoelectric section showing correlation with a driller's

log. Resistivity values are in ohm-feet.

Electromagnetics
Two types of electromagnetic instrumentation both of which induce
currents into the ground by electromagnetic induction, are in use: time-
domain and frequency-domain. The most common is the frequency-
domain system in which the transmitter is radiating energy at all times,
measuring changes in magnitude of the currents induced within the
ground (McNeill, 1980). The time-domain system, in which the trans-
mitter is cycled on and off, measures changes in the induced currents
within the ground as a function of time.
Because the electromagnetic instruments do not require electrical
contact with the ground, measurements may be made quite rapidly.
Lateral variations in conductivity can be detected and mapped by
profiling. Using commonly available frequency-domain EM instru-
 Surface geophysical methods 327
ments, profiling station measurements may be made to depths ranging
from 1 to 60 m.
Continuous EM profiling data can be obtained to a depth of 15 mm
(Benson et ai., 1982). These continuous measurements significantly
improve lateral resolution for mapping small hydrogeologic features
(Fig. 14.3b). Data can be recorded on an analog strip chart recorder or a
digital data acquisition system. The excellent lateral resolution obtained
from continuous EM profiling data has been used to outline dosely-
spaced burial pits, to reveal the migration of contaminants into the
surrounding soils (Fig. 14.6), or to delineate complex fracture patterns
(Fig. 14.3b; Benson et al., 1982).
In addition to evaluation of natural hydrogeologic conditions and
mapping of contaminant plumes, some of the electromagnetic instru-
mentation can be used to locate trench boundaries, buried wastes and
drums, and metallic utility lines. Frequency-domain EM instruments
provide two outputs consisting of an in-phase component and an
out-of-phase component. The out-of-phase component is used to
measure electrical conductivity and can be used to locate pipes. The in-
phase component is a measure of the magnetic susceptibility, but it can
also be used to detect both ferrous and nonferrous metal. For example,
using the in-phase component, a single 200 L steel drum can be
detected at a depth of about 2 to 3 m.
Vertical variations in conductivity can be determined by sounding.
The instrumentation is placed at one location and measurements are
made at increasing depths by a changing coil spacing and/or coil
orientation. Sounding data can be acquired at depths ranging from 0.75
to 60 m by combining data from a variety of commonly available
frequency-domain EM instruments. The vertical resolution of fre-
quency-domain EM soundings is relatively poor since only a limited
number of measurements are made at a few depths. However, they
do provide a quick means of obtaining limited vertical information.
Time-domain transient EM systems on the other hand are capable of
providing detailed sounding data from depths of 50 m or more than
300m.

Resistivity
As with EM measurements, electrical resistivity measurements are a
function of the type of soil or rock, its porosity, and the fluids which
fill the pore spaces. The method may be used in many of the same
applications as the EM method (Cartwright and McComas, 1968; Griffith
and King, 1969; Zohdy et ai., 1974; Mooney, 1980; Benson et al., 1982;
Telford et al., 1982).
The resistivity method requires that an electrical current be passed
through the ground from a pair of surface electrodes. Both direct
328 Geophysical techniques for subsurface site characterization

current and switched direct current power sources are used. The voltage
resulting on the surface of the ground due to current is measured
between a second pair of electrodes. This requires that metal stakes
be driven into the ground or that non-polarizing (copper-copper
sulfate) electrodes be used. The greater the electrode spacing, the
greater the depth of the measurement. Usually the depth is less than
the spacing between electrodes. There are a number of electrode
geometries that can be used, including the Schlumberger, dipole-dipole
and many more, but the simplest, in terms of geometry is the Wenner
array, which consists of four equally spaced electrodes all in a line. The
apparent resistivity of the soil and rock is calculated based on the
electrode separation, the geometry of the electrode array, the applied
current, and measured voltage.
The resistivity technique may be used for profiling or sounding,
similar to EM measurements. Profiling provides a means of mapping
lateral changes in subsurface electrical properties to a given depth,
and is well suited to the delineation of hydrogeologic anomalies and
mapping of inorganic contaminant plumes (Fig. 14.7).
Sounding measurements provide a means of determining the vertical
changes in subsurface electrical properties. Interpretation of sounding
data provides the depth, thickness and resistivity of subsurface layers.
Data can be interpreted using master curves for 2-3 layers (Orellana
and Mooney, 1966). Computer models are commonly used and may be
used to handle more than 2-3 layers (Mooney, 1980). Sounding data
are used to create a geoelectric section which illustrates changes in the
vertical and lateral resistivity conditions at a site. Figure 14.8 shows a
geoelectric section developed from a single resistivity sounding, along
with a drillers log showing the correlation.
One drawback with resistivity sounding is that the electrode array
requires considerable space. For example, a Wenner array sounding
(with electrodes equally spaced) may require that the spacing between
the electrodes be as much as 3-4 times the depth of interest. There-
fore, a sounding to a depth of 30 m could require an overall array
length (from current electrode to current electrode) of 300 to 400 m. At
many sites this space may not be available.

Comparison of electromagnetic and resistivity measurements

The frequency-domain electromagnetic method is often preferred for
making profiling measurements since it requires less space for a meas-
urement to a given depth. Also because the electromagnetic method
does not require that electrodes be driven into the ground, measure-
ments can be made more rapidly and are not influenced by shallow
geologic noise associated with the electrodes used in resistivity meas-
urements. On the other hand, because resistivity methods provide
 Surface geophysical methods 329
better vertical resolution than the frequency-domain EM method, the
resistivity method is commonly employed for sounding measurements.
When space is limited and deep soundings are needed, there are
advantages to using the time-domain electromagnetics system for
soundings since it requires less space than long resistivity arrays.
Electromagnetic measurements may be affected by buried metal
pipes, metal fences, nearby vehicles, buildings and power lines, etc.
as are resistivity measurements. But resistivity measurements are
often less sensitive to many of these problems, sometimes permitting
resistivity measurements to be made near cultural metal where electro-
magnetic measurements cannot be made.
Electromagnetic and resistivity measurements from the same location
may not agree, due to the difference in the volume of material being
sampled and in the differences in current distribution inherent to the
two methods. Measurements will agree if they are both made over a
uniform media and the instruments have been properly calibrated.

14.4.3 Seismic refraction and reflection

Seismic refraction and reflection techniques are often used to deter-
mine the top of bedrock, to determine depth of water table, and to
assess the continuity of geologic strata, faults and buried bedrock
channels. The refraction method may be used to characterize the type
of rock and degree of weathering based upon the seismic velocity of
the rock. The seismic velocity in rock is related to its density and
hardness. Therefore, characterization of the material on the basis of
seismic velocity can indicate the degree of weathering, and rippability.
Seismic waves are transmitted into the subsurface by a source. These
waves are refracted and reflected when they pass from one soil or rock
type into another which has a different seismic velocity. An array
of geophones placed on the surface measures the travel time of the
seismic waves from the source to the geophones. The refraction and
reflection techniques use the travel times of the waves and the geo-
metry of the source-to-geophone wave paths to model subsurface
conditions. The unit of time is milliseconds (10- 3 seconds). For
most refraction work, the first refracted compressional wave arrivals
(P-waves) are used. For reflection work, the latter arriving reflected
compressional waves are used.
A seismic source, geophones, and a seismograph are required to
make the measurements. The seismic source may be a simple sledge
hammer or other mechanical source with which to strike the ground.
Explosives may be utilized for deeper applications which require
greater energy. Geophones implanted in the surface of the ground
translate the ground vibrations of seismic energy into an electrical
330 Geophysical techniques for subsurface site characterization
signal. The electrical signal is displayed on the seismograph, permit-
ting measurement of the arrival of time of the seismic wave and dis-
playing the wave forms from a number of geophones. Geophone
spacing can be varied from about 1 m to more than 100 m depending
upon the depth of interest and the lateral resolution needed.
Since the seismic refraction and reflection methods measure small
ground vibrations, they are inherently susceptible to vibration noise
from a variety of natural (i.e., wind and waves) and cultural sources
(Le., walking, vehicles and machinery).

Seismic refraction
The refraction method is commonly applied to shallow investigations
up to about 100m (Griffith and King, 1969; Benson et ai., 1982; Telford
et ai., 1982; Haeni, 1986). However, with sufficient energy, surveys to
several hundred meters are possible. Up to three or four layers of soil
and rock can normally be determined, if a sufficient velocity exists
between adjacent layers. A typical refraction line for a shallow inves-
tigation might consist of 12 or 24 geophones set at equal spacings
as close as 1.5-3 m. The refraction survey may require a maximum
source-to-geophone distance 4-5 times the depth of investigation. Sig-
nificantly greater source energy will be required as the depth of inves-
tigation increases. Two inherent limits to the refraction method are its
inability to detect a lower velocity layer beneath a higher velocity
layer and thin layers.
Seismic refraction work can be carried out in a number of ways. The
simplest approach in terms of field and interpretation procedures can
be carried out by creating two separate seismic impulses, one at each
end of the geophone array. The results of travel time measurements
are plotted in a time-distance plot as the first step to the analysis
(Mooney, 1973). The results of this simple measurement provide two
depths and thus, the dip of rock under the array of geophones.
A more detailed refraction survey can be carried out so that depths
are obtained under every geophone (Fig. 14.9). This survey will pro-
duce a detailed profile of the top of rock. Lateral resolution will depend
upon the geophone spacing which might range from 1.5 to 15 m
(Redpath, 1973). The general reciprocal method (Palmer, 1980) will
accommodate varying velocities within each layer, while calculating the
depth under each geophone.

Seismic reflection
By comparison, a seismic reflection survey is capable of much deeper
investigations with less energy than the refraction method. While
reflections have been obtained from depths as shallow as 3 m, the
 Surface geophysical methods 331

700

690

ooo~~g~'~_nd_su_rt~_e__________~

spread velocity
( teeUsecond)

13900 19700 9400 20000 18000 6800 10800 10300 17500 17100
/
8800
\
18900

~ + 00 1+ 00 2 + 00 3 + 00 4 + 00 5 + 00 6 + 00 7 ~ 00 8, 00 9,' 00 10 ~ 00 11 + 00 12+ 00 13 + 00 14' + 00

station location

Fig. 14.9 Profile of top of bedrock from seismic refraction survey. (Depth to
rock determined under each geophone. Geophone spacing is 3m (10ft).)

method is more commonly applied to depths of 15-30 m or greater.

The reflection technique can be used effectively to depths of several
hundred meters or more and can provide relatively detailed geologic
sections (Fig. 14.10).
As in radar reflections, the vertical scale is measured in two-way
travel-time i.e., the time it takes for a wave to travel down to an
interface and back up to the surface again. The time scale must then
be converted to depth making some assumptions regarding seismic
velocity within the strata.
There are two approaches currently used to obtain shallow seismic
reflection data, both of which have only evolved in the mid 1980s: the
common offset method (Hunter et ai., 1982); and the common depth
point (COP) method adapted from the oil industry (Lankston and
Lankston, 1983; Steeples, 1984). The common offset method uses low
cost equipment and software but has some site specific limitations
which are not inherent in the COP method. The common depth point
has fewer site specific limitations, but is more dependent upon
sophisticated hardware and software capabilities. Hardware and soft-
ware for the shallow COP method have just recently become readily
available in the late 1980s.
The shallow high resolution reflection methods discussed here
attempt to utilize the highest frequencies possible (150-600 Hz) to
improve vertical resolution, and relatively closely spaced geophones
(0.3-6 m apart) to provide good lateral resolution. Because of the need
for higher frequencies, attention must be given to selection of a seismic
source and its optimum coupling to soil or rock, as well as geophone
placement.
The reflection method is limited by its ability to transmit energy,
particularly high frequency energy, into the soil and rock. Loose soil
near the surface limits the ability of the soil system to transmit high
 332 Geophysical techniques for subsurface site characterization

a)

b)

...,
:I;
E 20 cr
"'
..J 2
,.:I;
III -<
-< 30...,
l-
a:
,.
II

<
"' 3
I- 40 ~
-<
;t ~
;t ;::
04 SO '"
..J

"'
III F~
:I: 60 ~
I-
c..
"'a

Fig. 14.10 Common offset seismic reflection data showing channel in

bedrock. Source: Hunter et al. (1982).

frequency energy into and out of the rock, limiting the resolution
which can be obtained.

14.4.4 Micro-gravity
Gravity measurements respond to changes in the earth's gravitational
field caused by changes in the density of the soil and rock. By meas-
uring the spatial changes in the gravitational field, variations in sub-
surface geologic conditions can be determined (Griffith and King, 1969;
Telford et al., 1982). There are two basic types of gravity survey: a
regional gravity survey and a local micro-gravity survey. A regional
 Surface geophysical methods 333
.100

..
~
In

.050
Gravity Profile
:E
.E IB+0B 2B+ee 30+00

.."
In
CD 0.0

>
~
> -.as
~ Gravity Low Due to Bedrock Channel
~

-.IS

Fig. 14.11 Micro gravity profile showing bedrock channel. Source: Technos
Inc.

gravity survey which employs widely spaced (a few hundred meters to

a few kilometers) stations is carried out with a standard gravity meter.
Micro-gravity surveys, with closely spaced stations of 1.5-6 m, are
carried out with a very sensitive micro gravimeter. Regional surveys
are used to assess major geologic conditions over many hundreds of
square kilometers. Micro-gravity surveys are used to detect and map
shallow, localized, geologic anomalies such as bedrock channels, frac-
tures, and cavities.
The unit of acceleration used in gravity measurement is the gal. The
earth's normal gravity is 980 gals. Micro-gravity measurements are
sensitive to within a few micro gals (10- 6 gals).
The micro-gravity survey results in a Bouguer anomaly which is the
difference between the observed gravity values and theoretical gravity
values. The Bouguer anomaly is made up of deep-seated effects (the
regional Bouguer anomaly) and shallow effects (the local Bouguer
anomaly). It is the local Bouguer anomaly that is of interest in micro-
gravity work (Fig. 14.11).
A gravimeter is designed to measure extremely small differences in
the gravitational field and is a very delicate instrument. The instrument
is thermostatically controlled to minimize drift caused by temperature
variations. Considerable care must be taken in shipment and general
field use to avoid shock to the instrument. Gravity measurements may
be affected by ground noise (similar to the seismic methods), winds,
and temperature. To compensate for small instrument drift throughout
the day, measurements must be made at a base station every hour or
so, so that drift corrections can be applied to the data. Corrections
must also be made for the constantly changing earth tides, changes in
elevation (to the nearest millimeter), and topography. Gravity data
may be presented as a profile or as a contour map, depending upon
project needs.
334 Geophysical techniques for subsurface site characterization
14.4.5 Metal detection
Metal detectors are commonly used by utility and survey crews for
locating buried pipes, cables, and property stakes. They are also useful
for detecting buried drums and for delineating the boundaries of
trenches containing metallic drums or trash (Fig. 14.12; Benson et al.,
1982). Metal detectors can detect both ferrous metals such as iron and
steel, and non-ferrous metals such as aluminum and copper.
Metal detectors have a relatively short detection range, since the
detector's response is proportional to the cross-section of the farget and
inversely proportional to the sixth power of the distance to the target.
Small metal objects about 10-20 mm in size can be detected at a distance
of approximately 0.5-1 m. Specialized metal detectors will detect
larger objects like standard steel petroleum drums at depths of 1-3 m,
and massive piles of drums may be detected at depths of up to about
3 m. The metal detector is a continuously-sensing instrument and is
used with a sweeping motion while moving forward along a survey
line. It may also be held steady while a traverse line is walked and the
results recorded on a strip-chart recorder. The area of detection of a
metal detector is approximately equal to its coil size or coil spacing
(typically 0.3-1 m). Metal detectors can be affected by nearby metallic
pipes, fences, cars, buildings and in some cases by changes in soil
conditions.

14.4.6 Magnetometry
A magnetometer measures the intensity of the earth's magnetic field.
As with gravity surveys, a magnetic survey can be used to map regional
geologic conditions over large areas. In certain geologic environments,
magnetics can be used to map depth to bedrock, channels, and frac-
tures (Griffith and King, 1969; Breiner, 1973; Telford et al., 1982). The
primary application of magnetic measurements at hazardous waste
sites is in detecting buried drums, tanks, and pipes (Breiner 1973;
Benson et al., 1982). A magnetometer will only respond to ferrous
metals (iron and steel) and will not detect nonferrous metals. The
presence of buried ferrous metals creates a local variation in the
strength of the earth's magnetic field, permitting the detection and
mapping of buried ferrous metal (Fig. 14.13).
Two types of magnetic measurements are commonly made: total
field measurements and gradient measurements. A total field measure-
ment responds to the total magnetic field of the earth, any changes
caused by a. target, natural magnetic, and cultural magnetic noise
(ferrous pipe, fences, buildings and vehicles).
The effectiveness of total field magnetometers can be reduced or
totally inhibited by noise or interference caused by time-variable
 Surface geophysical methods 335

Fig. 14.12 Results of a metal detector survey to locate a burial trench.

Fig. 14.13 Magnetic gradient over a trench with buried drums. The trench is
approx. 20 m by 100 m.

changes in the earth's magnetic field, or spatial variations due to

magnetic minerals in the soil, steel debris, pipes, fences, buildings, and
passing vehicles.
A base station magnetometer can be used to reduce the effects of
natural noise by subtracting the base station values from those of the
search magnetometer. This can minimize any errors due to natural
336 Geophysical techniques for subsurface site characterization

long period changes of the earth's field. Cultural noise, however,

will remain a problem with total field measurements. Many of these
problems can be avoided by use of gradiometer measurements and
proper field techniques.
Gradient measurements are made by a gradiometer, which is simply
two magnetic sensors separated vertically (or horizontally) by 1 to 2 m.
Gradient measurements have some distinct advantages over total field
measurements: they are insensitive to natural changes in the earth's
magnetic field and minimize most cultural effects. Because the response
of a gradiometer is the difference of two total field measurements it
responds only to the local gradient. As a result, it is better able to
locate a relatively small target, such as a buried drum. The disadvantage
of a gradiometer is that it is slightly less sensitive than a total field
instrument.
A total field magnetometer's response is proportional to the mass of
the ferrous target and inversely proportional to the cube of the distance
to the target (such as a drum). A gradiometer response is inversely
proportional to the fourth power of the distance to the target (such as a
drum), making it less sensitive than the total field measurement. While
gradiometers are inherently less sensitive than total field instruments,
they are also much less sensitive to many sources of noise. Under ideal
conditions, a single drum can be detected at depths up to about
6 m with a total field magnetometer (about 3 m with a gradient
magnetometer). Massive piles of drums can be detected at depths up to
15 m or more with a total field magnetometer (about 7 m with a
gradient magnetometer).
A total field or gradient proton procession magnetometer normally
requires the operator to stop to take a measurement, while fluxgate
gradiometers permit the continuous acquisition of data as the magneto-
meter is moved across the site. Continuous coverage is much more
suitable for detailed (high resolution) surveys to identify local targets,
such as drums, and the mapping of areas in which complex anomalies
are expected.

14.5 DOWNHOLE GEOPHYSICAL MEASUREMENTS

One of the most common subsurface investigation techniques is that of

sampling soil and rock at discrete intervals (typically I-2m) as a
boring is advanced. This method provides gross information on sub-
surface lithology but sand lenses, fractures, or other subtle changes
in geology which can affect permeability, can easily go undetected.
Though continuous sampling or coring can improve the description
of geologic conditions, it is costly, time-consuming, and drillers or
 Downhole geophysical measurements 337

geologist's logs are somewhat subjective. Furthermore, 100% recovery

of the sample is rarely achieved.
A number of downhole logging techniques are available for deter-
mining the characteristics of soil, rock, or fluid along the length of a
boring, or monitoring well (Keys and MacCary, 1976). These methods
provide continuous high resolution in situ measurements that are often
more representative of hydrogeologic conditions than are the samples
obtained from borings. An adequate assessment of conditions will
often require that multiple logs be used since each log responds to a
different property of the soil, rock, or fluid. Some of these techniques
will provide measurements from inside plastic or steel casing, and
some will allow measurements to be made in the unsaturated zone as
well as the saturated zone.
Downhole logging measurements can be correlated to the known
geologic strata in one hole and then can be used to identify and
correlate geologic strata in other holes without soil or rock sampling.
Thin layers and subtleties, not readily detected in soil or core samples,
can often be resolved by logging. Logging can significantly improve
one's ability to accurately characterize and correlate strata between
borings by providing high-resolution data, independent of subjective
interpretations of soil and rock type.
A number of soil and rock properties can also be measured in situ.
Values for soil and rock porosity, density, seismic velocity and elastic
moduli can be obtained to facilitate engineering design. Even more
important, is the ability to identify the uniformity or lack of uniformity
of subsurface conditions. Downhole measurements can be used to
identify permeable zones such as sand lenses in glacial tills, weathered
zones, and fractures or cavities in rock. The same measurements are
also effective for identifying impermeable zones such as aquitards, and
help to assess their continuity and integrity.
Monitoring wells which have been in place for years provide the
basis for long-term chemical monitoring. For many of these wells,
neither geologic logs nor installation records are available. Using
downhole techniques, it is often possible to obtain geologic information
and well construction details. In addition, logging may be used to
determine whether a problem exists with well construction and what
type of remedial work, if any, is necessary to correct it.
By running nuclear logs in existing holes with steel or PVC casing,
geologic strata outside the casing can be characterized. Under some
conditions, contaminants outside PVC casing can be characterized by
running electromagnetic induction logs. A downhole television camera
can be used within cased wells, to assess monitoring well conditions or
can be used within an uncased borehole to assess the existence of
fractures.
338 Geophysical techniques for subsurface site characterization
While each log is susceptible to both natural and cultural noise,
borehole diameter and/or well construction will probably be of most
concern. Most logs provide measurements within a radius of 150 to
300mm from the hole. Therefore, as the hole diameter becomes larger,
the measured results become more dominated by drilling and con-
struction aspects.
A description of the commonly used logs is given below. Table 14.1
lists the conditions in which these logs can be used and some inherent
limitations.

14.5.1 Nuclear logs

Natural gamma log

A natural gamma log records the amount of natural gamma radiation
that is emitted by rocks and unconsolidated materials. The chief use of
natural gamma logs is the identification of lithology and stratigraphic
correlation in open or cased holes above and below the water table.
The gamma-emitting radioisotopes normally found in all rocks and
unconsolidated materials are potassium-40 and daughter products of
the uranium and thorium decay series. Because clays and shales con-
centrate these heavy radioactive elements through the processes of
ion exchange and adsorption, the natural gamma activity of clay and
shale-bearing sediments is much higher than that of quartz sands and
carbonates. Therefore, the gamma log, which indicates an increase in
clay or shale content by an increase in counts per second (Fig. 14.14), is
useful for evaluating the presence, variability, and integrity of clays
and shales.
The radius of investigation for the natural gamma log is from about
150 to 300mm from the borehole wall (Keys and MacCary, 1976).

Gamma-gamma (density) log

A gamma-gamma log is used to determine the bulk density of the soil
and rock as well as to identify lithology. The log can be used in open or
cased holes above and below the water table (Fig. 14.15).
The gamma-gamma log is an active probe containing both a radiation
source and a detector. This log provides a response in counts per
second, that is averaged over the distance between the source and
the detector. The count rate is inversely proportional to density and
therefore, the count rate can be used to indicate relative changes in
density. The count rate can be calibrated to provide actual density of
the in situ material. The radius of investigation for the gamma-gamma
log is relatively small (only about 150 mm; Keys and MacCary, 1976).
Table 14.1 General characteristics and use of downhole geophysical log

downhole log parameter measured casing saturated unsaturated radius of effect of hole
(or calculated) measurement diameter and mud
uncased PVC steel
Natural gamma natural gamma yes yes yes yes yes 150-300mm moderate
radiation
Gamma-gamma density yes yes yes yes yes 150mm significant
Neutron porosity below yes yes yes yes yes 150-300mm moderate
water table-
moisture content
above water
table
Induction electrical yes yes no yes yes 750mm negligible
conductivity
Resistivity electrical resistivity yes no no yes no 300mm to 1.5m significant to
minimal
depending upon
probe usea
Single point electrical resistance yes no no yes no near borehole significant
resistance surface
Spontaneous voltaa~- !e~Eonds yes no no yes no near borehole significant
potential (SP) to ISSlml ar surface
minerals and
flow
Temperature temperature yes no no yes no within borehole NA
Fluid conductivity electrical yes no no yes no within borehole NA
conductivity
Flow fluid flow yes no no yes no within borehole NA
Caliper hole diameter yes yes yes yes yes to limit of sensor NA
cased hole ~tcally
diameter . -O.9m

NA: not applicable.

 Geology Natural Gamma
000

Natural Gamma
500

90~
10~
11

12 ----.----

13

15

16

17

20

21

22

20~-----------L~~

Fig. 14.14 Natural gamma logs from two nearby boreholes, 100 ft apart. Note
the characterization and correlation of the shale and limestone units. Source:
Technos Inc.
 Natural Gamma Induction Gamma-Gamma Neutron
Geology 0 500 0 100 0 20,000 0

···,·····,····1···'·····,·····;·····, ..... ;...... ;..... ,......; .....,... , .. '..... ;..... ,.....,......;\;:; ... ,..... ; ; I···········,······

20

30

50

clIff
.
·· ..
.~. . . ..
... .

·· . ..
..... ..
"'m::'m'
..
·· .
. . . .
. . .. .
.
,5 :--0==:
··· . . . . .
••• :. • ••• .: ••••••••••• v .... _ ••••• ; ••••
....
. ... ..
·
·
..
. ..
... ..

···6:·:····:·····:·····:······:···
.. . ...
... ; ... ~:'::::.
· ....
· .'
:·....:......'.....;..... :
"
"

·Bffj"
... ~ ~.

· . . . .

"'hhlli,Jl:il
·· . .. . .
".
..
···:·····:······:·····;·····7····-:·····: .

Fig. 14.15 A suite of logs from within the same borehole, The natural gamma
log provides a means of characterizing the shale. The gamma-gamma log
provides a measure of density and the neutron log provides a measure of
porosity within the shale/limestone units. Source: Technos Inc.
342 Geophysical techniques for subsurface site characterization
Therefore, borehole diameter variations and well construction factors
can affect this log more than other logs.

Neutron-neutron (porosity) log

A neutron-neutron log provides a measure of the moisture content
above the water table and total porosity (based upon the water content
of the rock) below the water table (Fig. 14.15). The log can be run in
open or cased holes, above and below the water table. The neutron-
neutron log is an active probe with both a radiation source and a
detector. It provides a response, in counts per second, that is averaged
over the distance between the source and the detector. The count rate
is inversely proportional to water content and therefore, the count rate
can be used to indicate relative changes in moisture content above the
water table and porosity below the water table. The count rate can be
calibrated to provide actual moisture content or porosity of the in situ
material. Since clays and shales contain bound water, this log will
indicate an increase in moisture or porosity in these materials, since the
log cannot resolve the differences between bound and free water.
Therefore, caution must be used in interpreting this log. The radius of
investigation for the neutron-neutron probe is approximately 150 mm
(up to 300 mm in very porous formations; Keys and MacCary, 1976).
Borehole diameter variations and well construction factors can affect
this log, but not as severely as the density log.

14.5.2 Non-nuclear logs

Induction log
The induction log is an electromagnetic induction tool for measuring
the electrical conductivity of soil or rock in open or PVC-cased bore-
holes above or below the water table (similar to EM measurements
made on the surface). The induction log can be used for identification
of lithology and stratigraphic correlation. Electrical conductivity is a
function of soil and rock type, porosity, permeability, and the fluids
filling the pore spaces. Since the logs response (in millisiemens/meter
or millimhos/meter) will be a function of the specific conductance of
the pore fluids, it is an excellent indicator of inorganic contamination
(Fig. 14.16) and in some cases organic contamination (when organics
are mixed with inorganics, or when thick layers of hydrocarbons are
present). Variations in conductivity may also indicate changes in clay
content, permeability of a formation, or fractures. An induction log
provides data similar to that provided by a resistivity log (because
conductivity is the reciprocal of resistivity). However, the induction log
can be run without electrical contact with the formation, and therefore,
 Downhole geophysical measurements 343

INCREASE IN
CONDUCTIVITY CONDUCTIVITY WATER/POROSITY

....-(counts per second)

(millimhos/meter) (millimhos/meter)
00 0 00
a a
oO
. .~~r-~,- __,--,N
~ 0
o__~~~~~~~.
0 ~

10

ii
CI)
U-
.s 15
.r:.
ii.
CI)
C

20 .--4-----1 20

25 L-1__---1 25

A. EM Induction Log of B. EM Induction Log of C. Neutron-Neutron

Inorganic Contaminants Natural Background (porosity) Log
Conditions

Fig. 14.16 Identifying contaminants and permeable zones using (a) EM

induction log of inorganic contaminants, (b) EM induction log of natural
background conditions and (c) neutron-neutron (porosity) log. Source: Technos
Inc.

it can be used in both the vadose zone and the saturated zone and to
log through PVC casing.
The radius of investigation for the induction log is approximately
0.75 m from the center of the well. Because this log has a much larger
radius of investigation than other logs, it is almost totally insensitive to
borehole and construction effects, and as such it is a good indicator of
the overall soil and rock conditions surrounding the borehole.

Resistivity log
The resistivity log measures the apparent resistivity (ohm-feet or
ohm-meters) of rock and soil within a borehole. Because resistivity is
the reciprocal of conductivity, which is the property measured by an
induction log, the resistivity log responds to and measures the same
properties and features as the induction log. However, because of the
need for electrical contact with the borehole wall, the resistivity log can
only be run in an uncased hole filled with water or drilling fluid.
344 Geophysical techniques for subsurface site characterization
There are a number 6f electrode spacings or geometries which may
be used for resistivity logs. 'Short normal' probes (typically an electrode
spacing of 400 or 450 mm) give good vertical resolution and measures
the apparent resistivity of the formation immediately around the hole.
'Long normal' probes (typically an electrode spacing of 1. 7 m) have less
vertical resolution but measure the apparent resistivity of undisturbed
rock at a larger radius from the hole, similar to the induction log (Keys
and MacCary, 1976).

Resistance log
A resistance log (sometimes referred to as single-point resistance)
measures the resistance (in ohms) of the earth materials lying between
a downhole electrode and a surface electrode. It can only be run in
uncased holes in the saturated zone. The primary uses of resistance
logs are geologic correlation and the identification of fractures or wash-
out zones in resistive rocks. A resistance log should not be confused
with the resistivity log, which provides a quantitative measure of the
material resistivity.
The radius of investigation of the resistance log is quite small. It is in
many cases as strongly affected by conductivity of the borehole fluid as
it is affected by the resistance of the surrounding volume of rock (Keys
and MacCary, 1976).

Spontaneous-potential log
The spontaneous-potential (SP) log measures the natural potential, in
millivolts, developed between the borehole fluid and the surrounding
rock materials. It can only be run in uncased holes within the saturated
zone. The SP voltage consists of two components. The first component
results from electrochemical potential caused by dissimilar minerals or
fluids. The second component is the streaming potential caused by
water moving through a permeable medium.
SP measurements are subject to considerable noise from the elec-
trodes, hydrogeologic conditions and borehole fluids. Even though
these measurements do not provide quantitative results they have a
number of applications, including characterizing lithology, providing
information on the geochemical oxidation reduction conditions, and
providing an indication of fluid flow.
The radius of investigation of the SP log is highly variable (Keys and
MacCary, 1976).

Temperature log
A temperature log provides a continuous record of the temperature of
the borehole fluid immediately surrounding the sensor as it is lowered
 Applications 345
within an open borehole. It is usually run in uncased holes within the
saturated zone. The temperature log will often indicate a zone of
ground water flow within the uncased portion of a borehole. Flow is
indicated when an increase or decrease in water temperature occurs.
Changes in temperature can also be used to monitor leaks in casing
where damage or corrosion has occurred. A temperature log may have
a sensitivity of 0.5 °C or better.

Fluid flow
There are many ways of measuring fluid flow within a borehole (Keys
and MacCary, 1976). The most common uses an impeller-type flow
meter which provides counts per second. It can only be run in uncased
holes within the saturated zone. The count rate can usually be calibrated
to provide results in the desired units.

Fluid conductivity
A fluid conductivity log provides a measurement of the specific con-
ductance of the borehole fluids (micromhos/centimeter). It can only be
run in uncased holes within the saturated zone. If accurate values are
needed, a temperature log must also be run so that corrections can be
made.

Caliper log
A caliper log provides a record of the diameter of an open borehole or
of the inside diameter of a well casing. The log can be run in open
or cased holes, above or below the water table, although it is most
commonly run in open holes. The caliper probe consists of spring-
loaded arms which extend from the logging tool so that they follow the
sides of the borehole or casing.
Caliper logs are utilized to measure hole diameter and to locate
fractures or cavities in an open hole. The caliper log can also be used to
determine well construction details and casing diameter. It can also be
used to reveal any deterioration due to extreme corrosion or accumula-
tion of minerals on the interior of the well casing.

14.6 APPLICATIONS

There is no simple way to select the geophysical method(s) required to

solve a particular problem. Tables 14.2-4 are provided to illustrate how
the geophysical methods may be used to carry out assessment of
hydrogeologic conditions, detecting and mapping contaminants, and
346 Geophysical techniques for subsurface site characterization
locating and mapping buried wastes and utilities. However, simple
tables and rules of thumb often fail when considering specific project
needs and site specific conditions, therefore, the tables presented here
should be used only as a guide.

14.6.1 Assessing hydrogeologic conditions

The first and often the most important task of any site investigation is
the evaluation of the natural hydrogeologic conditions. A description
of the overall hydrogeologic conditions of the setting and site specific
conditions along with any hydrogeologic anomalies is usually required.
Knowledge of the natural anomalies, in relation to the setting and the
site itself ensure that drilling and sampling is done at locations which
will be most likely to yield useful information.
Table 14.2 lists possible applications of surface geophysical methods
and some of their advantages and limitations in evaluating hydro-
geologic conditions. Variations in the shallow natural setting are best
evaluated with ground penetrating radar, which provides the highest
resolution of all of the surface methods. However, depth of penetration
of radar is highly site-specific, and is typically less than 16 m. When
silts and clays are at the surface, penetration may be limited to only
I-2m.
Even with these limitations, ground penetrating radar can often help
solve problems deeper than its sensing range. For example, by looking
for anomalies in shallow marker beds, or observing shallow soil piping,
shallow radar data can be used to predict the presence of cavities and
fractures much beyond its range. Investigation of such near-surface
indicators (NSI) with radar and other methods to evaluate deeper
conditions is a powerful technique (Benson and Yuhr, 1987).
High-resolution seismic reflection can be used with radar to provide
a more complete depth profile. While this seismic reflection method
has less resolution than radar, information can be acquired to depths of
tens or even hundreds of meters. The reflection method is often found
to be ineffective at depths shallower than 7-15 m where radar is most
effective. Therefore, these two methods are quite complementary for
developing detailed geologic profiles. It should be noted, however,
that the cost of seismic work is considerably greater than the cost for a
radar survey.
Seismic refraction and resistivity soundings provide good vertical
information, such as depth to water table and depth to rock though
they are not capable of achieving the lateral and vertical resolution of
radar, or in some cases, the vertical resolution of seismic reflection.
The frequency-domain EM techniques have excellent lateral resolu-
tion in the continuous profiling mode to depths of about 15 m. In
general, the frequency-domain EM are limited in their capability to
Table 14.2 Surface geophysical methods for evaluation of natural hydrogeologic conditions*

method general application continuous depth of penetration major limitations

measurements
Radar profiling and mapping; yes to 30m Penetration limited by soil
highest resolution of any (typically less than 10 m) conditions
method
EM (Frequency- profiling and mapping; yes to 60m Affected by cultural features
domain) very rapid (to 50 feet) (metal fences, pipes,
measurements buildings, vehicles)
EM (time-domain) soundings or profiling and no to few hundred meters Does not provide
mapping measurements shallower
than 150 feet
Resistivity soundings or profiling and no No limit Requires good ground contact
mapping (commonly used to 100 to and long electrode arrays
300m) Integrates a large volume of
subsurface
Affected by cultural features
(metal fences, pipes,
buildings, vehicles)
Seismic refraction profiling and mapping soil no No limit Requires considerable energy
and rock (commonly used to 100 to for deeper surveys
300m) Sensitive to ground vibrations
Seismic reflection profiling and mapping soil no to few hundred meters Shallow surveys, <100 feet are
and rock most critical
Sensitive to ground vibrations
Micro gravity profiling and mapping soil no No limit Very slow, requires extensive
and rock (commonly used to 100 to data reduction
300m) Sensitive to ground vibrations
Magnetics profiling and mapping soil yes No limit Only applicable in certain rock
and rock (commonly used to 100 to environments
300m) Limited by cultural ferrous
metal features

• Applications and comments should only be used as guidelines. In some applications, an alternate method may provide better results.
348 Geophysical techniques for subsurface site characterization

produce vertical detail (sounding data). Yet, these EM methods can

often provide some relative sounding information (i.e., thick versus
thin or shallow versus deep) very quickly and more cost effectively
than resistivity or seismic refraction.
Probably the two best techniques to map lateral variations in soil and
rock, from a speed and resolution point of view, are ground-penetrating
radar and continuous EM measurements. While the radar performance
is highly site specific, the EM technique can be applied in almost any
environment and can often provide deeper information, but with much
less vertical resolution than radar. Continuous EM profiling measure-
ments provide high lateral resolution and can be run at speeds from
1-5km/hr depending on the detail required. The rapid speed at which
EM measurements can be obtained and the option of continuous
profile measurements at depths of up to 15 m makes EM the best
choice for profile work under most situations.
The resistivity method can also be used for profile measurements by
moving the array in small increments to provide data at closely spaced
intervals. This is a slow process relative to an EM survey, and resistivity
data can be affected by near-surface geologic noise at the electrodes.
Sometimes, one method may work and another will fail under a
given set of site conditions. For example, in some cases, resistivity
measurements can be made relatively close to a chain link fence or a
buried pipe line (when they are relatively isolated electrically) where
EM measurements cannot.
In order for any geophysical method to work, there has to be a
contrast in the parameter being measured. The best method is the one
in which the parameters being measured have the greatest contrast and
will be least influenced by site specific conditions and noise. The final
decision must be made on a site-by-site basis.
Once the surface methods have defined the 2-dimensional or even
3-dimensional conditions reasonably well, boring locations can be
selected. These locations should be selected to be representative of
the normal background conditions at the site and to investigate any
anomalies identified by geophysical or other methods identified by
geophysical or other methods.
The drilling program should be designed to provide a means of
accurately characterizing soil and rock conditions to the greatest extent
possible within the budget available. If an adequate downhole logging
program is used, most of the holes can be drilled without extensive
sampling. However, it is always good practice to continuously sample
(or core) and log at least one borehole and then geophysically log this
hole as an on-site 'calibration' hole. Then the other boreholes can often
be drilled and geophysically logged without sampling. This approach
provides a reference for the logging data to compare to site-specific soil
samples or rock core. Then the logs can be used to extrapolate soil and
 Applications 349

rock type and other conditions to nearby boreholes. For example, the
natural gamma logs shown in Fig. 14.14 clearly show the limestone
(low counts) and the shale units (high counts). In this case, correlation
of the stratigraphy from natural gamma logs in adjacent holes is easily
and accurately made.
When the appropriate suite of logs are combined, continuous in situ
logging measurements can be obtained in both the unsaturated and
saturated zones to characterize hydrogeologic conditions. Geologic
formations can be identified and easily correlated from hole to hole.
Relative estimates of clay content, density and porosity can be ob-
tained. Permeable sand lenses and fractures can be identified, as can
impermeable clay and shale zones. In addition, the continuity of
impermeable zones can be assessed. The maximum amount of data
should always be obtained from each hole because borings are few and
costly.
Figure 14.15 shows a suite of natural gamma, density, and porosity
logs from the same hole. The density log shows variable conditions in
the overlying soil, but fairly uniform density within the shale and
limestone units. In contrast the porosity log shows considerable varia-
tion throughout both the shale and limestone units. Without calibra-
tion these logs can be used to indicate relative changes in density and
porosity. By calibrating these logs, quantitative results for density and
porosity may be obtained in some situations.

14.6.2 Detecting and mapping of contaminant plumes

Table 14.3 illustrates how the surface geophysical methods can be
applied to the mapping of contaminant plumes. The fundamental
approach to evaluating the direction of groundwater flow and the
possible extent of a contaminant plume is by determining the hydro-
geologic characteristics of the site (that is, determining the presence of
pathways such as buried channels, fractures, and permeable zones).
'Direct' detection of inorganics (or organics mixed with inorganics)
can be accomplished by the electrical methods (EM and resistivity)
as well as ground penetrating radar, as shown in Table 14.3. When
in organics are present in sufficient amounts, the higher specific con-
ductance of the pore fluids acts as a tracer by which the plume can be
mapped. In cases where inorganic plumes have a very low specific
conductance, or plumes consist of dissolved organics, they will not be
detectable by the electrical methods.
Where suitable penetration is possible, ground penetrating radar can
provide a means for mapping the depth to the top of and lateral extent
of shallow inorganic plumes. However, because of the site-specific
behavior of radar, the EM or resistivity methods are most often used.
Of the two methods, EM measurements are preferred for profile work,
350 Geophysical techniques for subsurface site characterization
Table 14.3 Surface geophysical methods for mapping of contaminant plumes'

Feature to be mapped Comment on approach

Permeable pathways, The fundamental approach to evaluating the direction
bedrock channels, of groundwat ~r flow and the possible extent of a
etc. contaminant plume is by determining the
hydrogeologic characteristics of the site (see Table 14.2
for evaluation of natural hydrogeologic conditions).
Inorganics or mixed When inorganics are present in sufficient concentrations
inorganics and above background or organics are part of such an
organics inorganic plume, they can be detected by the
electrical methods and sometimes radar. The higher
specific conductance of the pore fluids acts as a tracer
by which the plume can be mapped.
Hydrocarbons When sufficient hydrocarbons have been present in the
soil or floating on a shallow water table, for a
sufficient period of time they may sometimes be
detected by the electrical methods or by radar.
Because of their low conductivities (high resistivity)
they may sometimes be detected by the electrical
methods. Due to changes in dielectric constant or
suppression of the capillary zone they may
sometimes by detected by radar (in some situations
where degradation of hydrocarbons is occurring,
conductivity may increase). These applications are
limited and should be treated with caution. A more
reliable approach is to map natural permeable
pathways (see Table 14.2 for evaluation of natural
hydrogeologic conditions and table 14.4 for mapping
of cultural pathways).
Radar Limited applications - may sometimes be used to detect
shallow floaters (0-6 m) to map hydrocarbons in soil.
May detect thickness in some cases.
EM May be applicable to detect low conductivity at some
sites.
Resistivity May be applicable to detect high resistivity at some
sites.

* Applications and comments should be used only as guidelines. In some applications, an

alternate method may provide better results.

particularly where continuous sampling can be employed. Resistivity is

generally preferred for sounding work.
Both resistivity and EM measurements can miss a contaminant
plume if the measurements are in the wrong location. However, rapid
EM profiling by either continuous or station measurements allows
coverage of a site with closely spaced data. It is not unreasonable
 Applications 351
from a cost perspective to have overlapping measurements, therefore
providing total site coverage using the EM profiling method.
Both resistivity and EM are capable of providing vertical sounding
data. The frequency-domain EM method provides a depth of penetra-
tion that is limited to about 60 m and provides less resolution than the
resistivity method since measurements are made only a few depths.
The depth to which resistivity sounding data can be obtained is virtually
unlimited. Depths of one to several hundred meters are obtainable.
However, the long resistivity arrays necessary for deep measurement
may not be practical in many areas due to space restrictions and
cultural factors. Here, the time-domain EM transient systems, which
have a smaller coil size, would be the choice for measurements to
depths from 3 m to a few hundred meters.
In some cases organics can be mapped because they are mixed with
inorganics. Figure 14.6 shows the inorganic plume from a chemical!
drum recycling center which also contains organics. Figure 14.7 shows
the plume from a landfill that contains low levels of organics. The
results in Fig. 14.17 show an excellent comparison of measurements by
EM, resistivity, and organic vapor analyses (OVA) responses obtained
from a mixed plume of organics and in organics confined in a buried
channel. Clearly, if inorganics are present, they should be used as an
easily detectable tracer that will provide a first approximation of where
the organics may be.
Some investigators have suggested that direct detection of shallow,
major hydrocarbon spills can be accomplished by looking for EM
conductivity lows (or resistivity highs) associated with the organics.
Recent spills of petroleum products do not seem to yield a resistivity
high or an EM conductivity low. However, the possibility for an
anomaly exists where the product has been in the ground for some
time, where there is a substantial amount of floating product, and
where the conductivity of the natural soil conditions is high enough (or
resistivity low enough) to provide a reasonable contrast between the
hydrocarbons and the natural soil.
Once the spatial extent of a contaminant has been mapped by sur-
face geophysics and after boreholes have been installed, continuous,
downhole logging can be used to evaluate changes in the vertical
hydraulic conductivity of soil and rock, as well as the distribution of
contaminants. The vertical distribution and concentration of con-
taminants at a site can vary significantly as a result of local changes in
hydraulic conductivity. Because hydraulic conductivity can change by
more than an order of magnitude in less than a meter, it can have a
significant impact on the test results obtained from a monitoring well.
The chemical concentration in a well may be low, average, or high,
depending upon screen length and location. Two downhole logging
techniques particularly well suited for permeability evaluations are the
 400
ELECTROMAGNETlCS
a:
w 300
tu
:::'!
(fJ
0------ 200
I
:::'!
:::i
--.J
100
:::'!

400
RESISTlVITY
a: I
w 300
f-

J~
W
:::'!
(fJ
------ 200
0
I
:::'!
:::i
=:!
:::'! 100 N \

~ I.
~
0 ~.

1000
ORGANIC VAPOR ANAL YZER
z
0 100
:::i

I\A~~
--.J
:::'!
(fJ
------
f-
a: 10
~
D-

) ~
Fig. 14.17 Organic vapor profile over buried channel: (a) electromagnetics, (b)
resistivity and (c) organic vapor analyzer. Note correlation between resistivity
and EM measurements. This is an excellent example of a buried channel
controlling flow and the level of correlation between organic and inorganic
contaminants.
 Applications 353

electromagnetic (EM) induction log (or resistivity logs - only in open

hole) and the neutron (porosity) log. Both of these logs can be run in
an open hole or within an existing PVC-cased well, above or below the
water table (Fig. 14.16).
The EM induction log, shown in Fig. 14.16a, indicates the presence
of inorganic contami.-:.ants which have preferentially migrated within
five discrete zones of increased permeability in the limestone. These
zones are indicated by higher conductivity values and range from 0.5
to 1.5 m in thickness. The presence of high hydraulic conductivity
zones detected by the EM induction method was confirmed by using a
downhole television camera which visually located small cavities and
fractures in each zone. Figure 14.16b shows an EM induction log of
natural conditions taken in a background well. Here the permeable
zones are indicated by the log since there are no contaminants present.
Figure 14.16c shows a neutron log taken in the same background well.
In this log, zones of variable porosity are revealed whether con-
taminants are present or not. Conditions shown by this log are also
representative of conditions at the contaminated well.
An adequate assessment of conditions in a hole often requires that
more than one log be run. At this site, an EM induction log was used
to identify the contaminated zones, and a neutron log was used to
identify zones of increased porosity. Once conditions at a site are
understood, a reliable and representative monitoring well system can
be designed and data from existing monitoring well systems can be
more accurately evaluated.

14.6.3 Location and mapping of buried wastes and utilities

Location and mapping of buried wastes, utilities, drums, and tanks
is a common application of geophysical methods. Table 14.4 lists the
surface geophysical methods applicable to this problem. Locating
buried bulk wastes where no metal is present, can often be accom-
plished by ground-penetrating radar, if soil conditions are suitable.
Often the shallow edges of trenches can be detected even in soil
conditions which provide poor ground-penetrating radar penetration.
Shallow EM tools are also effective for most location problems. When
metals are present, EM conductivity, metal detectors, and magnetom-
eters are the primary choices. Metal detectors and magnetometers are
unaffected by most soil types or by the presence of contaminants.
However, EM measurements are influenced by both variations in soils
and the presence of contaminants.
To locate buried steel drums, the use of metal detectors, mag-
netometers, or the in-phase component of measurements EM are
recommended. All three methods can be used to locate single
Table 14.4 Surface geophysical methods for location and mapping of buried wastes and utilities'

bulk wastes method bulk wastes wlo metals steel drum wlmetals pipes and drums tanks

Radar Very good if soil Very good if soil Good if soil conditions are Very good for metal and
conditions are conditions are appropriate (may non-metal if soil
appropriate appropriate provide depth) conditions are
Sometimes effective to Sometimes effective to appropriate (may
obtain shallow obtain shallow provide depth)
boundaries in poor soil boundaries in poor soil
conditions conditions
EM Excellent to depths less Excellent to depths less Very good (single drum Very good for metal tanks
than 6m than 6m to 2-3m)
Resistivity Good (sounding may Good (sounding may NA NA
provide depth) provide depth)
Seismic refraction Fair (may provide depth) Fair (may provide depth) NA NA
Micro gravity Fair (may provide depth) Fair (may provide depth) NA NA
Metal detector NA Very good (shallow) Very good (shallow) Very good (shallow)
Magnetometer NA Very good (ferrous only; Very good (ferrous only; Very good (ferrous only;
deeper than metal deeper than metal deeper than metal
detector) detector) detector)

* Applications and comments should be used only as guidelines. In some applications, an alternate method may provide better results.
NA: Not applicable
 Applications 355

200 L steel drums as well as large piles of drums within their depth
limitation.
Both the metal detector and the EM will respond to ferrous and non-
ferrous metals, while a magnetometer will respond only to ferrous
metals. Therefore, it is necessary to assess what metals may be present
in order to select the appropriate instrument.
While ground-penetrating radar can be used to find drums, it will
often be unable to detect a single drum if it is not oriented so that
energy is reflected back to the antenna. Furthermore, many natural
and man-made objects may have a radar response similar to that of a
drum.
For small, discrete, critical targets such as a single steel drum, con-
tinuous measurements (on closely spaced lines of about 2m) are
required to provide a reasonable probability of detection. Ground-
penetrating radar, shallow EM measurements, metal detectors, and
certain magnetometers can provide these continuous measurements.
However, there may be cases in which the proximity of other metal
structures may limit the use of EM to locate drums or trenches, making
ground-penetrating radar a clear choice.
Metal detectors and ground-penetrating radar both provide reason-
ably good spatial resolution to pinpoint the location of a target, while
EM and magnetometers do not provide the same target resolution
because the shape of their response curve is broader and often more
complex.
Metal detectors, EM, and total field magnetometers are highly
susceptible to interference from nearby metallic cultural features. Any
of these features can produce an erroneous response that may be
incorrectly interpreted as a subsurface target. Because metal detectors
are relatively short-range devices, they can be operated closer to such
sources of interface, than can most magnetometers.
Measurements made with a total field proton procession magnetom-
eter are susceptible to interference from high magnetic gradients,
natural changes in the earth's magnetic field, and nearby power lines,
whereas fluxgate gradiometers do not suffer from these short-comings.
Seismic refraction, resistivity, magnetic, and gravity techniques may
also be used to locate boundaries of larger trenches and landfills. These
techniques are much slower and will provide less resolution than the
previously described methods. However, they are often the only tech-
niques which can be used to estimate the thickness of a landfill or
trench. It should be noted that interpretation of such data should be
done with caution by experienced personnel.
356 Geophysical techniques for subsurface site characterization

REFERENCES

Benson, R.C and Glaccum, R.A. (1979) Radar surveys for geotechnical site
assessment, in Geophysical Methods in Geotechnical Engineering, Specialty
Session, ASCE, Atlanta, Georgia, pp. 161-78.
Benson, R.C, Glaccum, R.A. and Noel, M.R. (1982) Geophysical Techniques for
Sensing Buried Waste and Waste Migration, EPA - Environmental Monitoring
Systems Laboratory, Las Vegas, Nevada.
Benson, R. C and LaFountain, L.J. (1984) Evaluation of subsidence or collapse
potential due to subsurface cavities, in Sinkholes, Their Geology, Engineering
and Environmental Impact, Proc. of the 1st Multidisciplinary Conf. on Sinkholes,
Sinkhole, Orlando, Florida, pp. 201-15.
Benson, R.C and Pasley, D.C (1984) Ground Water Monitoring: A Practical
Approach for a Major Utility Company, 4th Nat. Symp. and Exposition on
Aquifer Restoration and Ground Water Monitoring, National Water Well
Association.
Benson, R.C and Scaife, J. (1987) Assessment of flow in fractured rock and
karst environments, Proc. 2nd Multidisciplinary Conf. on Sinkholes and the
Environmental Impacts of Karst, Orlando, Florida.
Benson, R.C, Turner, M., Turner, P. and Vogelson, W. (1988) In situ, time-
series measurements for long-term ground-water monitoring, in Ground
Water Contamination: Field Methods, ASTM STP-963 (eds A.G. Collins and
A.I. Johnson), Am. Soc. for Testing and Materials, pp. 58-72.
Benson, R.C, Turner, M., Vogelson, W. and Turner, P. (1985) Correlation
between field geophysical measurements and laboratory water sample
analysis, Proc. Nat. Water Well Association/EPA Conf. on Surface and Borehole
Geophysical Methods in Ground Water Investigations, National Water Well
Association.
Benson, R.C and Yuhr, L. (1987) Assessment and long-term monitoring of
localized subsidence using ground-penetrating radar, Proc. 2nd
Multidisciplinary Conf. on Sinkholes and the Environmental Impact of Karst,
Orlando, Florida.
Benson, R.C and Yuhr, L. (1990) Evaluation of fractures in silts and clay using
ground penetrating radar, presented at the 4th Radar Con£., Denver,
Colorado.
Breiner, S. (1973) Applications Manual for Portable Magnetometers, Geometries,
Sunnyvale, California.
Cartwright, K. and McComas, M. (1968) Geophysical surveys in the vicinity of
sanitary landfills in northeastern Illinois, Ground Water, 6, 23-30.
Dunbar, D., Tuchfeld, H., Siegel, R. and Sterbentz, R. (1985) Ground water
quality anomalies encountered during well construction, sampling and
analysis in the environs of a hazardous waste management facility, Ground
Water Monitoring Review, 5(3), 70-4.
Greenhouse, J.P. and Monier-Williams, M. (1985) Geophysical monitoring of
ground water contamination around waste disposal sites, Ground Water
Monitoring Review, 5(4), 63-9.
Griffith, D.H. and King, R.F. (1969) Applied Geophysics for Engineers and
Geologists, Pergamon Press.
Haeni, P. (1986) Application of Seismic-Refraction Techniques to Hydrologic Studies,
US Geological Survey, Open File Report No. 84-746, Hartford, CT.
Hileman, B. (1984) Water quality uncertainties, Environ. Sci. and Tech., 18(4),
124-6.
Hunter, J.A., Burns, R.A., Good, R.L., MacAulay, H.A. and Cagne, R.M.
 References 357

(1982) Optimum field techniques for bedrock reflection mapping with the
multichannel engineering seismograph, in Current Research, Part B,
Geological Survey of Canada, Paper 82-1 Part B, pp. 125-9.
Keys, W.S. and MacCary, L.M. (1976) Application of borehole geophysics to
water-resources investigations, Techniques of Water-Resources Investigations
of the US Geological Survey, ch El.
Lankston, R. W. and Lankston, M.M. (1983) An Introduction to the Utilization of
the Shallow or Engineering Seismic Reflection Method, Geo-Compu-Graph, Inc.
Lysyj, I. (1983) Indicator methods for post-closure monitoring of ground
waters, in Proc. Nat. Conf. on Management of Uncontrolled Haz. Waste Sites,
Hazardous Materials Control Research Institute, pp. 446-8.
McNeill, J.D. (1980) Electromagnetic resistivity mapping of contaminant
plumes, in Proc. Nat. Conf. on Management of Uncontrolled Haz. Waste Sites,
Washington, DC, pp. 1-6.
Mooney, H.M. (1973) Engineering seismology, in Handbook of Engineering
Geophysics, Vol. 1: Bison Instruments, Minneapolis, Minnesota.
Mooney, H.M. (1980) Electrical resistivity, in Handbook of Engineering Geophysics,
Vol. 2: Bison Instruments, Minneapolis, Minnesota.
Orellana, E. and Mooney, H.M. (1966) Master Tables and Curves for Vertical
Electrical Sounding Over Layered Structures, Interciencia, Madrid, Spain.
Palmer, D. (1980) The Generalized Reciprocal Method of Seismic Refraction
Interpretation (ed. K.B.S. Burke), Dpt. of Geology, Univ. of New
Brunswick, Fredericton, N.B., Canada.
Perazzo, J.A., Dorrler, KC. and Mack, J.P. (1984) Long-term confidence in
ground water monitoring systems, Ground Water Monitoring Review, 4(4),
119-23.
Redpath, B.B. (1973) Seismic Refraction Exploration for Engineering Site
Investigations, Technical Report E-73-4, US Army Engineer Waterways
Experiment Station Explosive Excavation Research Laboratory, Livermore,
CA.
Steeples, Don W. (1984) High resolution seismic reflections at 200 Hz, Oil and
Gas J., 86-92, Dec 3.
Telford, W.M., Geldart, L.P., Sheriff, KE. and Keys, D.A. (1982) Applied
Geophysics, Cambridge University Press.
USEPA (1985) RCRA Ground-Water Monitoring Technical Enforcement
Guidance Document.
Walker, S.E. and Allen, D.C. (1984) Background Ground Water Quality
Monitoring: Temporal Variations, 4th Nat. Symp. and Exposition on Aquifer
Restoration and Ground Water Monitoring, National Water Well
Association.
Zohdy, A. A. , Eaton, G.P. and Mabey, D.R. (1974) Application of surface
geophysics to ground-water investigations, Techniques of Water-Resources
Investigations of the US Geological Survey, ch Dl.
 CHAPTER 15

Soil exploration at
contaminated sites
Charles O. Riggs

15.1 INTRODUCTION

Most of the drilling methods currently used for subsurface environ-

mental exploration are used on a day-to-day basis for geotechnical or
mineral exploration or for groundwater development and are readily
available. It is important to understand that the most economical or the
most readily available drilling method that would be used at a site
for geotechnical sampling or testing, or for mineral exploration, or for
water well drilling may not be the most technically appropriate method
for exploration of contaminated sites. As an example, a particular site
might be most economically drilled and sampled using the straight
hydraulic (mud) rotary method to obtain geotechnical design data or
to install a production water well; however, it might be better to use
hollow-stem augers or casing advancement to sample contaminated
materials, even if auger or casing advancement installation would be
more time-consuming and more expensive.
When contaminated sites are investigated, sampling quality without
cross-contamination is usually the most important consideration. It is
usually the primary goal in subsurface environmental exploration to
advance a borehole and obtain a soil or groundwater sample without
altering the chemistry of the sampled material by the drilling process,
by drilling fluid additives, or by the sampling device and the sampling
mechanics. Environmental exploration of and field installation oversite

Geotechnical Practice for Waste Disposal.

Edited by David E. Daniel.
Published in 1993 by Chapman & Hall, London. ISBN 0 412 35170 6
 Auger drilling 359
personnel (inspectors) should have a thorough understanding of all
generally available drilling methods and techniques, and particularly
the possible effects of the drilling method and technique on sample
quality.

15.2 AUGER DRILLING

Auger drilling with its many technical and economic advantages was
only slowly accepted in geotechnical engineering practice. However,
auger drilling has rapidly evolved as a primary method of subsurface
exploration at contaminated or potentially contaminated sites.

15.2.1 Continuous flight 'solid stem' auger drilling

Continuous flight, 'solid stem', augers (Fig. 15.1) have been used in
geotechnical exploration for many years. They are particularly efficient
and technically effectiv! when drilling relatively shallow exploration
borings in fine grained soils above the groundwater level. However, it
is practically impossible to use solid stem, continuous flight augers
with intermittent soil sampling through a chemically contaminated
ground surface without downwardly transporting contaminants. (The
same problems, probably magnified, occur with the use of 'hand
augers'.)
Continuous flight solid stem augers are usually fabricated with steel
flights welded to a relatively small diameter steel tube. They usually
have 'solid', hexagonal shank and socket, pinned connections (Fig.
15.1, Table 15.1). The flighting is usually 'pitched and timed' to be
practically continuous through each hexagonal shank and socket con-
nection of the auger column. A continuous flight auger column is
advanced with rotary and axial motion from a drill rig. The articulated
auger column is usually comprised of 1.52 m long sections. The auger
column must be completely removed from the borehole to permit a
sample of soil to be taken from beneath the bottom of the hole.

15.2.2 Hollow-stem auger drilling

The hollow-stem auger drilling method was first used for geotechnical
exploration during the 1960s. A hollow-stem auger column has a
continuously open axial stem. Hollow auger tools consist of outer
components and inner components (Fig. 15.2, Table 15.2). When the
borehole is advanced to the depth of sampling, the pilot assembly
and center rods are removed. A sampling tube is inserted through the
hollow axis of the auger column and then pushed or driven into soil
located beneath the base of the string of augers. The sampling tube is
360 Soil exploration at contaminated sites

Fig. 15.1 Continuous flight solid stem auger.

then withdrawn from the hollow stem and the pilot assembly reinserted.
If a monitoring well is being installed, the casing is inserted through
the hollow stem.
Typical hollow auger sections are 1.52 m in length and are available
for monitoring well installations with inside, hollow-stem diameters
ranging from 83 mm to 207 mm (Table 15.2). Most hollow augers
have keyed, box and pin connections for transfer of drilling torque
through the coupling and for easy coupling and uncoupling (Fig. 15.3).
Hollow augers with threaded connections are available, but lubricants
are usually required for the threads. Lubricants may introduce con-
tamination to the soil. If lubricants are used, materials that will not be
misinterpreted as contamination must be used.
Hollow-stem augers have been used to advance soil exploration
 Auger drilling 361
Table 15.1 Typical continuous flight auger sizes

auger series h~xagonal shank a~ger diameter auger head

In. (mm) In. (mm) c~tting diameter
111. (mm)
13 22
1125 16 (21) 8
(67) 3 (76)

1500 11 (29) 21 (64) 3 (76)

8 2
118 (29) 3 (76) 3!8 (79)
11
8
(29) 4 (102) 42 (114)
11 (29) 412 (114) 5 (127)
8
11 (29) 5!2 (140) 6 (152)
8

2000 12 (41) 4 (102) 412 (114)

8
12 (41) 412 (112) 5 (127)
8
12 (41) 5!2 (140) 6 (152)
8
12 (41) 6 (152) ~4 (171)
8
12 (41) 7 (178) 814 (210)
8

2875 12 (41) 728 (200) 9 (229)

8
12 (41) ~8 (225) 10 (254)
8
12 (41) 11 (280) 12 (305)
8

Note: auger diameters are without hardsurfacing and should be considered minimum
manufacturing dimensions.

borings to depths greater than 100 m but are usually used only to
maximum depths of 20-45 m, depending on the density or hardness of
the soil, the depth of the groundwater level, the power available at the
drill spindle and the dead weight of the drill rig.
Groundwater levels and zones of perched water are usually immedi-
ately detected and can be accurately logged during hollow auger drill-
ing. This important observation capability can not be accomplished
during drilling with 'wet' drilling methods. When hollow augers
are used to drill below the groundwater level, water levels are often
controlled within the auger stem to prevent excess seepage stresses
and a 'quick condition' from occurring below the auger head. Seepage
stresses below the auger head can be controlled by injecting drilling
fluids into the hollow augers during drilling. This procedure essentially
provides a straight rotary drilling system requiring only a small amount
of water to maintain the balancing head within the hollow auger.
Under some circumstances the pilot assembly is replaced with a simple
'knock-off' plate on the bottom of the hollow auger head.
362 Soil exploration at contaminated sites

Rod beap
Adapter
DrIve
~ ......."'" Connector

............--HOIIIOIN stem
f!U;¥3r section

C€nterRod

PilOt Asserrbly

AUger Comector
Auger Head

Fig. 15.2 Continuous flight hollow stem auger.

15.3 DIRECT ROTARY DRILLING

Direct (straight) rotary drilling is often called mud rotary drilling;

however, the circulation medium of rotary drilling can be clear water,
some liquid other than water, or air, with or without additives. Core
drilling with liquid or air circulation is a form of direct rotary drilling.
When the 'direct' or 'straight' hydraulic rotary method is used for
subsurface exploration, a bit on the bottom of a drill rod column is
simultaneously rotated and advanced into the soil or rock. Soil or
rock cuttings produced by the drilling process are removed from the
borehole with a fluid that is pumped downward through the drill rod
 Direct rotary drilling 363
Table 15.2 Typical hollow-stem auger sizes

hollow-stem fighting auger head

inside diameter diameter cutting diameter
m. (mm) m. (mm) in. (mm)

*2i (57) ~8 (143) M4 (159)

*2~ (70) rJ8 (156) (J (171)
4
~4 (83) ~8 (168) 7! (184)
4
3;! (95) 7! (181) 7;! (197)
4 8 4
4! (108) 7? (194) 8! (210)
4 8 4
M4 (159) ~8 (244) 10i (260)
8! (207) 11~ (295) 12~ (318)
4

Note: auger fighting diameters should be considered

minimum manufacturing dimensions .
• Seldom used for monitoring well installations.

Fig. 15.3 Joining of hollow stem augers.

column, passing through the bit and then upward within the annulus
of the borehole and the drill rods (Fig. 15.4).
In comparison to auger drilling the complete rotary drilling ptocess
and system is rather complex. The system typically consists of the drill
rods and bit, a mud pit, a suction hose, a recirculation pump, a
pressure hose to a swivel and a rotation spindle which is a part of the
drill rig.
364 Soil exploration at contaminated sites

Fig. 15.4 Direct rotary drilling.

15.3.1 Drilling mud

Drilling mud is the liquid circulation medium of 'hydraulic' rotary
drilling. It usually consists of water and one or more additives.
Commercially available, processed sodium montmorillonite clay
powder (bentonite to a driller) is the most frequently used rotary
drilling fluid additive. The higher quality bentonite additives consist
of sodium montmorillonitic clay powder with sodium carbonate and
polymer additives plus impurities consisting of various other naturally
occurring mineral particles.
Other often used drilling fluid additives include standard bentonite
(without polymers), cellulose polymers, organic polymers, anionic
acrylimide polymers usually in combination with a plyacrylate, sodium
 Direct rotary drilling 365

carbonate (soda ash), potassium chloride (potash), barium sulfate

(barite), and various lost circulation materials such as wood fibers and
shredded paper. Extreme care must be taken to ensure that additives
or leachable components of additives will not introduce any chemicals
that might be misinterpreted as contamination.
A correctly 'designed' drilling fluid with additives can perform some
combination of the following as required:
1. it can provide a viscous medium for improved transportation of
cuttings from the borehole;
2. it can form a membrane seal to prevent excessive loss of water
through the borehole wall;
3. it can react chemically with active clays or shale units to improve bit
performance or to deter swelling of clay or shale units along the
borehole wall; and
4. it can clean and cool the bit.

15.3.2 Viscosity for cuttings transport

Pure water without additives is the 'ideal' drilling fluid but can be
used only under ideal circumstances. For example, when competent
rock, other than a swelling shale is drilled with a core barrel, water
without additives is usually a very adequate drilling fluid. Because the
annulus of a cored hole is usually relatively small (if the drill rods are
size-matched to the core barrel) and the cuttings generated by coring
bits are usually small, up flow velocities adequate for cuttings transport
are obtainable at relatively low fluid pumping rates. The upflow
velocity within the borehole annulus can be computed by dividing the
pumping rate by the section area of the annulus. When boreholes in
rock are cored and when the drill rods are size matched to the core
barrel, e.g. NW rods with NW barrel or when wireline rods are used,
up flow velocities greater than 20 m/min are readily attainable and
usually adequate for cuttings transport without using drilling fluid
additives.
When the soil or rock is drilled for geotechnical exploration or
monitoring well installations using the direct rotary method, both
the borehole annulus and the cuttings are usually larger than with
core drilling. To provide transport capability of the larger cuttings
within the larger annulus without using a pumping rate that will erode
the wall of the borehole, the viscosity of the drilling fluid is usually
increased by adding bentonite powder or a polymer or possibly both.
A quick measure of the viscosity of drilling liquids can be obtained
with a Marsh funnel. The time in seconds for one liter of liquid to flow
by gravity throu&h a standard orifice in the funnel provides a quick
366 Soil exploration at contaminated sites

and easy measure of viscosity. Water has a funnel viscosity of about

26 s. Typical drilling fluids have Marsh funnel viscosities of 36-60 s,
although 35-40 s is often considered ideal. In other words, typical
drilling fluids are often too viscous, but sometimes with good purpose.
When drilling fluids are too viscous, cuttings do not adequately
separate or 'settle out' in the mud pit but are recirculated through the
pump, swivel and drill rods. Recirculation of cuttings can damage
components of the pump and restrict, if not totally block, the recircula-
tion process.

15.3.3 Borehole support

If a boring is drilled using water without additives through a soil or
rock of low hydraulic conductivity, water will not be lost to the ground
and the water in the borehole will exert a pressure on the wall of
the borehole of approximately 2.9 kPa per meter of depth below the
fluid surface. This water pressure, and almost nothing else, keeps the
borehole open. If a boring is drilled into relatively permeable soils,
the water will apply the same magnitude of pressure to the borehole
wall; however, at the same time the water will seep into the soil. When
circulation is stopped, or when the source of water is depleted, the
water level will recede within the borehole, seepage will reverse
from the ground to the borehole, and the borehole walls will usually
collapse. However, a high quality bentonite or some other additive
can be used to form a 'mud cake' on the borehole wall, impeding fluid
loss, thus maintaining the liquid level and the supporting pressure on
the borehole wall. The mud cake will also impede the flow of water
through a well screen and filter pack placed in a borehole rotary drilled
to the complete well depth.
The occurrence of artesian groundwater pressures at depth can cause
the net supporting pressure on the borehole wall to approach zero or
even cause out-flow at the top of the boring. A weighting additive
such as barite (barium sulfate) can be added in controlled amounts to
maintain borehole support or in severe cases prevent artesian flow.
The addition of too much barite, however, can raise fluid densities
excessively high and contribute to hydraulic fracturing of ground.
Barite is usually added to bentonite base mud systems that have an
initial funnel viscosity in excess of 50 seconds in order to maintain the
barite in suspension.

15.3.4 Inhibiting clay activity

The same reaction of water with sodium montmorillonite that produces
a borehole viscosifier or sealer from high quality 'bentonite' powder
can cause a shale stratum along the borehole wall to swell against the
 Direct rotary drilling 367
core barrel or against the drill rods causing the barrel to 'lock up' in the
hole or even 'twist off'. It can also cause a build-up of a ring of soil
cuttings called a 'boot' on the outside of a core barrel or on drill
rods. And, it can contribute to smearing or plugging of the bit. These
undesirable occurrences are often 'inhibited' with the use of polymers
or potash in the drilling fluid. Carboxymethylcelulose (CMC) base
polymers can be used alone as an inhibitor or in combination with
bentonite powder, in which case the bentonite must be the first addi-
tive mixed with the makeup water. Muriated potash can also be
used alone as an inhibitor or in combination with a polyacrylamide
viscosifier, in which case the potash should be first totally dissolved in
the water before adding polymer.

15.3.5 Air rotary drilling

There are acknowledged advantages and disadvantages to the overall
concept of drilling with air. The main advantages are that there is no
cost for air and air is readily available at the drilling site. The main
disadvantage involves the costs of pressurizing (pumping) the air
before using it and cleaning the air before and after using it.
Air rotary drilling for geotechnical exploration is usually limited to
depths of about 75 m and for drilling in soil or rock above or only
a short distance below the groundwater level. Air rotary drilling is
used to much greater depths using large compressors for other
purposes such as groundwater development. An air rotary drilling
system is similar to a hydraulic rotary system. The mud pit, suction
hose and pump are replaced with an air compressor and a dust collector
and usually some kind of air discharge cuttings removal system.
For air rotary drilling, air is the circulation medium. Air provides
hole cleaning and cools the bit but cannot, even with foaming addi-
tives, provide borehole stability. When air is used for straight circula-
tion, a minimum upflow air velocity of about 1200 m/min is often
stipulated for adequate hole cleaning; however, upflow velocities of
much less than 1200 m/min have been successfully used. The adequacy
of a compressor to provide enough air is determined by dividing the
capacity of the compressor by the theoretical, un eroded section area of
the borehole annulus. Often the true annular area becomes oversized
from air blast erosion. Larger air flows are then required to transport
cuttings through the enlarged zone. The use of foaming agents can
reduce the air requirements by as much as one-half.
To increase upflow velocity, the borehole annulus can be reduced by
using larger drill rods but keeping the bit size the same; however, a
minimum radial clearance of about 10 mm is usually maintained for
passage of larger 'chips'.
A 4-9bar compressor is usually adequate when NX, NW or larger
368 Soil exploration at contaminated sites

drill rods are used for drilling typical small diameter geotechnical
exploration borings above the groundwater level to depths of less than
75 m. External flush drill rods should be used to minimize turbulence
and erosion of the borehole annulus.

15.4 CORE DRILLING OF ROCK AND SOIL

For many circumstances, core drilling is the preferred drilling method
for exploration in rock. Core observation allows optimal evaluation of
rock characteristics and placement of sampling devices.
A core barrel is rotated and simultaneously axially advanced with
rotary and axial motion applied through a drill rod column from a
drilling machine. A drilling fluid flushes and cools the bit, transporting
the rock or soil cuttings upward and out of the drilled hole. If the
procedure is successful, the rock or soil at the advancing end of the
core barrel is cut or abraded by the bit into a cylinder that enters and is
protected by the hollow axis of the core barrel (Table 15.3).
This description seems simple enough; however, there is little to be
called simple about core drilling. There are a multitude of variables
that severely complicate the process. The variables are usually more
numerous when core drilling is performed to obtain geotechnical
design parameters in shallow, often weathered or broken rock, than
when core drilling is performed for other purposes such as deep
mineral exploration. The geotechnical investigator or the monitoring
device installer is usually interested in the structural continuity and

Table 15.3 Common coring bit sizes

size bit set reaming shell

designation O.D. I.D. O.D.
mm mm mm
- - - - -

EX 37.7 21.5 37.7

EWD3 37.7 21.2 37.7
AX 48.0 30.1 48.0
AWD4, AWD3 48.0 28.9 48.0
AXCWL, AWC3, AQWL 48.0 27.1 48.0
BX 59.9 42.0 59.9
BWD4, BWD3 59.0 41.0 59.9
BWBWL, BWCE, BQWL 59.9 36.4 59.9
NX 75.7 54.7 75.7
NWD4,NWD3 75.7 52.3 75.7
NXBWL, NWC3, NQWL 75.7 47.6 75.7
NC 92.7 69.5 92.7
HWD4, HWBWL, HWD3 92.7 61.1 92.7
HQWL 96.3 63.5 96.3
CP, PQWL 122.1 93.1 122.1
 Core drilling of rock and soil 369
mass permeability of shallow rock units which are commonly fractured
or jointed and difficult to core drill. At greater depths where rock is
less frequently fractured or jointed, and becomes relatively easy to core
drill, there is usually no further need for geotechnical or monitoring
design information. There is a further complication in that geotechnical
or water well drillers are often rarely required to core drill in rock.

15.4.1 Core drilling systems

Drill rigs that are designed and manufactured specifically for core
drilling can often be used secondarily but usually less efficiently for
other purposes such as auger drilling or straight rotary drilling of soil
or rock. Other drill rigs, however, particularly multipurpose auger-
core-rotary drills, that are designed and manufactured ~pecifically
to efficiently perform drilling other than core drilling are often used
secondarily and sometimes less efficiently for core drilling. The capa-
bility to core a particular rock unit is related to the selection of the bit
and six very important drilling system factors:
1. the range of 'vibration free' rotary velocity available at the spindle of
the drill rig;
2. the capability to control the rate of axial feed of the spindle;
3. the force reaction characteristics of the drill rig and the capability to
control and limit the rock force reaction, therefore, the force on the
bit;
4. the design and condition of the core barrel;
5. the type, size, section length and condition of the drill rods; and
6. when drilling with liquids, the pumping characteristics of the circula-
tion pump including the capability to control the flow of the drilling
fluid - when drilling with air circulation, the performance charac-
teristics of the air compressor are equally significant.
The combined characteristics of the complete system and the
capability of the drill crew all contribute to the performance of the core
barrel and the bit, and the quality of the recovered core.

15.4.2 Conventional core barrels

Several core barrel designs have been standardized by the Diamond
Core Drill Manufacturers Association (DCDMA). However, the most
recent core barrel designs, which usually provide the highest quality
core and the most efficient coring capability, are not DCDMA stan-
dard barrels. It should be understood that unique improvements to
any drilling tool system are usually patented. Such improvements
are seldom 'standardized' until the patents expire, by which time,
invariably, further patented improvements are made.
370 Soil exploration at contaminated sites

15.4.3 Wireline core barrel

Wireline core barrels are similar to conventional core barrels with
the distinguishing exception that the inner tube assembly with core
encased can be removed from the outer tube with an 'overshot' device
lowered into the borehole on a wire line and hoisted to the surface
through open coupling, wireline drill rods. The outer barrel assembly
and the wireline drill rods remain in place in the hole during core
removal. After reinserting the same or another inner tube assembly,
coring can be resumed. A wireline core barrel is very similar to a
conventional core barrel.
There are several proprietary wireline latching and overshot systems.
Most work very well. A latching and overshot system should provide
not only the basic capability of efficiently removing and reinserting the
inner tube assembly but also the driller assurance that the inner tube
assembly is positively latched in the correct position within the head
and the outer tube when and if the inner tube is lowered into place
with a lowering tool.
Drillers often use two inner tube assemblies so that an alternate
inner tube can be lowered or 'dropped' into the core barrel and drill-
ing resumed while the core is being inspected or removed from the
other inner tube.

15.4.4 Split inner tubes

The solid inner tube of conventional and wireline core barrels can
be replaced with a split inner tube that is held together by nylon
reinforced tape wrapped around the barrel in full circle tape slots. After
the split tube is removed from the outer barrel, the tape on the inner
tube is cut and half of the tube is removed to allow inspection and
logging of the core in the configuration that it entered the barrel.
Another type of split inner tube system has an intermediate inner
tube which encases and holds the split tube together. Following a
core run, the split tube and core is hydraulically extruded from the
intermediate inner tube, usually with the same drilling fluid pump that
is used to drill the hole. These barrels are called triple tube core barrels.
Both split inner tubes and triple tubes are available for use with both
wireline and conventional barrels.

15.4.5 Coring bits

The coring bit is the tool that is attached to the bottom of a core barrel
such that if a rotary action, an axial advance and a flushing fluid are
properly applied through the drill rods, the rock (or soil) will be cut
into a cylinder that enters the core barrel.
 Core drilling of rock and soil 371

Bits are manufactured in many sizes and configurations. Accordingly,

the nomenclature of coring bits is quite complex. Bits can be generally
classified as:
1. diamond-surfaced set;
2. diamond-impregnated;
3. carbide insert; and
4. polycrystaline diamond insert.
Most geotechnical and particularly mineral exploration core drilling
is now performed with impregnated diamond bits. However, surface-
set diamond bits are often used and were the first type used for core
drilling. 'Surface-set' means that individual diamonds are set on the
face and sides of the crown of the bit. The nomenclature of coring bits
evolved to describe the details of surface set bits.
For most rock coring situations the use of impregnated bits is much
more economical than the use of surface-set bits, i.e. for drilling in
most hard and abrasive igneous and metamorphic rock units and in the
hardest of sedimentary rocks. The diamonds of impregnated bits are
distributed within the matrix crown. These diamonds are usually the
manufactured type and, therefore, are much smaller than the mined
diamonds that are used for surface set bits.
Some bits have tu'ngsten carbide cutters to core soil or relatively
soft, sedimentary rock. These bits usually have wide kerfs to cut an
'oversized' hole for easier passage of large cuttings. Some harder
sedimentary rocks can be efficiently cored with similar coring bits
that have cutting inserts each of which consists of a thin sheet of
manufactured (polycrystaline) diamond bonded to a carbide insert.

15.4.6 Soil sampling core barrels

Some rock core barrels can be used as soil sampling core barrels if the
appropriate bits are used. In addition there are core barrels that are
specifically designed and identified as soil sampling core barrels. The
two most popular soil sampling barrels are the Denison type and the
Pitcher type.
The Denison type barrel was first used by the Denison (Texas)
District of the US Army Corps of Engineers. It has an inner tube with a
shoe or inner tube extension that can be set approximately flush with
the face of the bit or can be extended variable distances beyond the face
of the bit. The variable inner tube extension was initially accomplished
by varying the length of the shank of the bit. The extensions on
recently manufactured Denison barrels are made by using different
lengths of inner tube shoes. The shoe can be set about 75 mm ahead of
372 Soil exploration at contaminated sites
the face of the bit for sampling the softest materials and flush with the
face of the bit when the soil (or rock) is hard enough to require the
coring action of the bit for penetration. Denison barrels are usually
manufactured with either a 0.6 m or a 1.5 m nominal length barrel.
The Pitcher type barrel was initially manufactured by the Pitcher
Drilling Company in California, and is very similar to the Denison
barrel. The basic difference is that the Pitcher barrel has a coil spring
between the inner tube and the swivel, such that a single length, thin
wall inner tube is automatically advanced variable distances in front of
the face of the bit according to the hardness of the soil being sampled.
When relatively soft ground is encountered during coring, the force of
the spring causes the inner tube to advance ahead of the face of the bit.
When relatively hard ground is encountered, the spring is compressed
and the leading edge of the sample tube is pushed back. It can be
pushed back as far as the face of the bit. Pitcher type barrels are usually
manufactured in 1.5 m or shorter lengths.
Soil sampling core barrels are usually operated at rotary velocities in
the range of 50-150 RPM. This range of rotary velocity minimizes
drilling vibrations and gives the driller better capability to control the
rate of penetration and the circulation fluid pumping rate.
Some of the bits used on soil sampling core barrels can produce
relatively large cuttings. Therefore, larger up flow velocities are required
within the borehole annulus than are usually required for coring rock.
An adequate upflow velocity may not be obtainable when using water
without viscosifier additives because of the often limited capacity of the
fluid pump that is available on an exploration drill rig. For these
circumstances, drilling fluid additives must be used to increase the
viscosity of the drilling fluid. The increased viscosity increases cut-
tings transport capability but decreases settling rates of cuttings in the
recirculation pit. The consequent recirculation of cuttings can cause
plugging of the core barrel.
The usual solution to the cuttings transportation problem is to use as
large a drill rod as available and practical, to use a well designed
fluid recirculation pit of adequate size and to control the drilling fluid
viscosity to allow the settling of cuttings. For many ground conditions,
drilling fluid additives may be required not only to remove the cuttings
but to seal the borehole wall for stability.

15.4.7 Core drilling with air circulation

Core drilling with air is generally limited to coring of soil and sedi-
mentary rocks that can be scratched with a steel knife blade.
When air is used as the principal bit coolant and circulation medium,
large surface set diamonds or carbide type wide kerf bits are usually
used. The use of wide kerf bits is often necessary to prevent a ring of
 Percussion drilling 373
cuttings, a 'boot', from building up on the outside of the core barrel or
on the outside of the drill rocks.
When coring with air, the bit is usually rotated at a rotary velocity
that will provide a circumferential bit velocity of about 10-20 m per
minute. These slow velocities are about 1/4 to 1110 of the velocity that
would be used to core the same rock with liquid circulation. Never-
theless, greater penetration rates than can be attained with liquid
circulation often are the result. It is believed that the absence of liquid
from the surface of the rock eliminates a normal stress effect and allows
the rock to be more easily sheared. An annulus upflow velocity of
about 900-1500 m per minute is usually required for air coring.

15.5 ROTARY PERCUSSION DRILLING

There are two basic types of rotary percussion drills. Rotary percussion
drilling involves:
1. variations of systems that utilize surface operated air-powered or
hydraulically-powered rotary-impact rock drills; and
2. variations of systems that utilize down-the-hole air-operated rotary-
impact hammers.
Air and hydraulic drifters can be used to drill holes in the size range
of about 25-125 mm in diameter. Currently available down-the-hole air
hammers can be used to drill holes in the size range of about 100 mm to
greater than 400 mm in diameter. Down-the-hole air hammers can be
used with most multipurpose or straight rotary drill rigs.
Both down-the-hole hammer systems and rotary impact drifters
employ the same basic drilling mechanism. A bit on the end of an
articulated hollow drill rod column is rotated relatively slowly while
being axially oscillated at a relatively rapid rate. The action of the bit
hammering on the rock face causes the rock to fracture in tensile
rebound following each compressive impact loading. There are two
general types of percussion bits:
1. those with 'X' shaped impact surface (called full face bits); and
2. those with spherical carbide inserts (called button bits).
Under most circumstances, bits with carbide buttons will advance in
rock at a much faster rate than full faced bits of equal diameter.

15.6 PERCUSSION DRILLING

Percussion or cable-tool type drilling, once the principal method of the

water well industry, is seldom used in the US for geotechnical explora-
374 Soil exploration at contaminated sites
tion mainly because of economic factors. Cable-tool drills are some-
times used in geotechnical exploration and water well drilling practice
to advance casing, particularly on environmental exploration projects
where drilling fluids are not permitted. Sometimes a cable-tool drill
and a drill with au gering capability can be used in combination to
advance and case a large diameter borehole for installation of a
groundwater sampling or monitoring device.
A cable-tool drill rig consists of a horizontal walking beam that is
operated off a rotating crank such that the walking beam with attached
cable sheaves is used to raise and drop a heavy string of impact drilling
tools. Each raising and dropping of the tool string provides one impact
of a percussion drilling bit on the rock face at the bottom of the
borehole (or one impact to a casing column).
When sufficient drilling progress has been made, the drilling tool
string is withdrawn and the rock cuttings are removed with a bailer.
The progress of drilling is relatively slow and inefficient with respect of
rotary methods.

15.7 CASING ADVANCEMENT METHODS

For some ground conditions casing advancement (and retraction)

methods may provide the most technically acceptable method of instal-
ling a monitoring well. There are various rotary and rotary percussion
drilling systems for installing casing. Sometimes they work very well
and sometimes the problems during drilling are insurmountable.
Cable-tool drills can also be effectively used to drive and retract casing.
Casing clean out is accomplished with dual tube reverse-air circulation,
or by bailing the hole, or with a rotary or washing process. There
are a very limited number of powerful, 'drill-through' casing drivers
available in North America. Casing advancement methods are most
applicable to the case of monitoring in very deep soils.

15.8 SOIL SAMPLING EQUIPMENT

Soil sampling devices consist of tubes that are pushed or driven into
the soil beneath the bottom of the borehole. To sample with a hollow-
stem auger, the sampling tube is lowered down the center of the
hollow auger and then pushed or driven into the underlying soil. With
other drilling methods, the drilling tool must be removed from the hole
and the sampling tube lowered down the boring on a separate pipe or
wire. In either case, there is a risk that the sampling tube will become
contaminated from soils or liquids that overlie the sampling point.
 Soil sampling equipment 375

Samples may be taken at various frequencies. In traditional geo-

technical subsurface exploration, samples are obtained at an interval of
approximately I-2m, or at a significant change in stratigraphy. For
environmental exploration, continuous sampling or sampling at very
close intervals is common because infrequent sampling can cause the
exploration team to miss a thin contaminated zone or an important
hydrogeologic anomaly.
Soil sampling devices may be grouped into two categories:
1. devices for obtaining high-quality, 'undisturbed' samples; and
2. devices for obtaining disturbed samples.
The degree of disturbance may be evaluated from the area ratio (A R ),
which is defined as follows.
_ cross-sectional area of sampler 100 0/
AR - X /0 (IS. 1)
cross-sectional area of sample

4 4 D 2 _ D2
X 1000;' - x 100% (lS.2)
rrD? °-
0
D? I

4
where Di is the inside diameter of the sampling tube and Do is the
outside diameter. A sampling tube that provides 'undisturbed' samples
should have an area ratio of ,;:;11 %, e.g., as required in ASTM D1S87.
Sampling tubes with larger area ratios may cause significant disturb-
ance of the soil and should not be regarded as undisturbed samplers.
Samplers that do provide 'undisturbed' samples are often called thin-
walled tube samplers or Shelby tube samplers.

15.8.1 Thin-walled tube samplers

In the US, thin-walled tube samplers are readily available in nominal
diameters of SO, 76, and 127 mm, with corresponding wall thicknesses
of 1.24, 1.6S, and 3.0S mm. The usual length is nominally 7S0 mm. The
sharpened sampling end of the tube is usually crimped to a slightly
smaller inside diameter than the rest of the tube to provide inside
clearance that reduces sliding resistance between the inside of the tube
and the soil sample. The inside clearance ratio, CR , is defined as
follows.
CR = Di - De X 100% (lS.3)
De
where Di is the inside diameter of the sampling tube and De is the
inside diameter between machined faces at the sampling end of the
tube. If the sampling end of the tube has not been crimped at all, then
376 Soil exploration at contaminated sites

Di = De and CR = O. More typically, some inside clearance is provided

and CR is between 0.5 and 1.5%. The ASTM practice for undisturbed
sampling (01587) requires that CR = 1%.
Thin-walled sampling tubes are normally pushed into the subsoil to
minimize disturbance. Sometimes the sampler is driven, but driving
may increase disturbance. The recovery ratio is defined as the ratio of
the length of the plug of soil sample retained in the sampling tube
divided by the distance the sampler was pushed or driven into the soil.
Under ideal circumstances, the recovery ratio will equal 1.0, more often
than not the sliding resistance between the sampled material and the
tubing wall will reduce the recovery ratio to much less than 1.0 (soil
beneath the plugged tube is pushed away from the tube).
Once the thin-walled sampling tube has been returned to the sur-
face, it may be handled in one of several ways. One technique is to
extrude the sample in the field and then carefully wrap and seal the
sample in a way that minimizes water loss or (for contaminated soil)
loss of any volatile constituents. Another technique is to seal the ends
of the tube with wax or special end caps to prevent moisture loss. If
volatiles are present, the sample is normally sealed in a way that keeps
air in the tube (termed head space) to an absolute minimum. The
samples are then transported back to the laboratory. If the soil samples
will be subjected to tests aimed at defining in situ properties (e.g.,
shear, consolidation, or hydraulic conductivity tests), the test speci-
mens or sampling tubes containing the specimens must be carefully
packaged and transported in a manner that causes minimal disturbance
(see ASTM 04220).

15.8.2 Disturbed samples

Disturbed samples may be taken for many purposes, e.g., determina-
tion of major soil constituents, analysis of water content, or analysis of
contamination. The advantage of obtaining a disturbed sample is that a
thick-walled sampling tube can be used; the thick-walled tube is robust
and can be used many times before replacement. Also, the tube can
contain small, thin tubes (called liners) on the inside of the thick-
walled sampling tube. The soil is contained in the liners, which are
extremely convenient for storing the samples during transport back to
the laboratory. The liners can be made of inert materials, e.g., stainless
steel, that will not react chemically with the contaminants.
The most common type of sampling device for obtaining disturbed
samples is the split-barrel sampler. A split-barrel sampler is split in
half down the length, but for sampling the two halves are clamped
together at both ends by a sampling shoe and sampling head that
threads into the sampling rod. The split barrel sampler is convenient
 Contaminant control 377
because when the sampler is removed from the borehole, the shoe and
head are unscrewed and then the two halves of the sampling barrel are
separated, revealing the soil sample.
Sampling tubes that obtain disturbed samples come in many lengths
and diameters. One very commonly-used sampler is the standard split-
spoon sampler, which has an inside and outside diameter of 34.9 and
50.8 mm, respectively. The standard split-spoon sampler may be driven
into the soil with a 64 kg hammer to obtain the penetration resistance,
which is known as the standard penetration N value (ASTM 01586).
Larger-diameter samplers (typically about 50-75mm in diameter)
are commonly used, particularly for sampling with liners inside the
sampling barrel. It is becoming common practice to use sampling tubes
with a length of 1.5 m inside hollow-stem augers to sample while the
auger is advanced.

15.8.3 Cleaning sampling equipment

The following guidelines are recommended for sampling contaminated,
or potentially contaminated, soil. As a minimum, sampling equipment
should be washed with a non-phosphate detergent solution and then
rinsed with a control water. If inorganic contaminants are present, a
series of acid-control water rinses may be needed, with the acid serving
as a desorbing agent. If organic contaminants are present, an organic
desorbing agent (such as isopropanol, acetone, or methanol) should be
included in the rinse sequence. The reader may wish to consult ASTM
05088 for further information.

15.9 CONTAMINANT CONTROL

An important consideration in subsurface exploration is control of

contaminant migration within the borehole. Contaminants can be
transported vertically in the borehole, either upward or downward.
Vertical migration of contaminants is undesirable if one is trying to
delineate contaminated and uncontaminated zones.
A most common problem during any drilling process is the transport
of contaminated materials downward in a borehole. It is practically
impossible, without casing the upper part of the hole, to prevent cross
contamination of materials when drilling through contaminated
ground.
The suggested technique for drilling through contaminated mate-
rials into uncontaminated materials is as follows:
1. drill to a depth slightly greater than the depth of suspected
can tamina tion;
378 Soil exploration at contaminated sites

2. set casing to the depth of the base of the borehole; and

3. continue drilling a smaller diameter hole through the casing.
This technique can be slow and expensive, but it provides a reasonably
effective way to minimize downward transport of contaminants. In
some cases, more than one casing may be set, with each successive
casing having a smaller diameter. In addition, one may need to drive
casing while drilling or employ other non-standard techniques.
Contaminants may also be transported upward during drilling. It is
possible to contaminate near-surface materials that might otherwise
not be contaminated. Again, casing may be needed to prevent such
contamination, particularly if the source of contamination is ground
water that might seep laterally into an uncontaminated vadose zone.

15.10 HEALTH AND SAFETY CONSIDERATIONS

Drilling into contaminated materials represents a serious hazard to

the drillers and others near the drilling operation. Health and safety
considerations must be given the highest priority on all projects involv-
ing the possibility that contaminants will be encountered. In general,
the procedure is to assume the worst until data are collected to indicate
that less severe conditions exist. Under worst-case conditions, a
separate air supply is maintained for all the drilling crew and anyone
near the drilling operation, and the highest degree of skin protection is
provided. If conditions are less severe, face masks with respirators may
prove adequate for breathing. In any case, air quality is normally
monitored continuously (e.g., for methane if drilling into old municipal
solid waste or volatile organics if drilling into soil contaminated with
chemical waste) in case an unanticipated and dangerous situation is
encountered.
In the US and most other industrialized countries, extensive training
courses are available and are required for drillers, borehole loggers,
and observers. The problem is not so much in identifying appropriate
health and safety practices as it is in making sure that workers are fully
aware of those practices and are willing to take required precautions
despite the often significant inconvenience and personal discomfort
involved.
 CHAPTER 16

Vapor analysis/extraction
Lyle R. Silica and David L. Jordan

16.1 INTRODUCTION

Interest in the fate, behavior, and remediation of volatile organic

compounds (VOCs) in unsaturated soil, or the vadose zone, has
increased rapidly over the past five years. The US Congress' Office of
Technology Assessment (OTA, 1984) identified VOCs as being one of
the more ubiquitous groups of hazardous chemicals present in con-
taminated ground water nationwide. A major reason for this is the
widespread use of VOCs in the manufacture of pesticides, plastics,
paints, pharmaceuticals, solvents, and textiles. Since volatile con-
stituents of petroleum, as well as a variety of synthetic solvents,
comprise a significant portion of the contamination cases encountered
today, it is no wonder that such interest has been born.
Research and development of technologies directed towards the
identification and clean-up of volatile organics in soil and groundwater
has led to the application of a variety of principles and techniques
to remediation, both off-the-shelf and innovative. Research has been
conducted into the behavior of volatile compounds in the solid, liquid,
and vapor phases of soil and groundwater systems. In addition, com-
puter modeling has been applied in order to understand better the
complex interrelationships of the many factors affecting the migration
and fate of VOCs in the subsurface.
This chapter focuses on the basic principles governing the behavior
and fate of VOCs in the vadose zone, methods for detection and
identification of VOCs, and on remedial technologies relying on vapor-
phase tr.ansport.

Geotechnical Practice for Waste Disposal.

Edited by David E. Daniel.
Published in 1993 by Chapman & Hall, London. ISBN 0 412 35170 6
380 Vapor analysis/extraction

16.2 -VAPOR TRANSPORT IN THE VADOSE ZONE - THEORY

When a liquid VOC is spilled on the soil or leaks from a tank into the
soil, the VOC partitions among the liquid and vapor phases and
becomes dissolved in soil moisture and adsorbed onto the surfaces of
soil minerals and organic matter. The degree of partitioning of the
VOC among these four components will depend on the volatility and
water solubility of the VOC, the soil moisture content, and the type
and amount of soil solids, i.e., the minerals and organic matter.

16.2.1 Partitioning of VOC between pure liquid and soil gas

Partitioning between the pure liquid and soil gas is controlled by the
vapor pressure of the VOC and the temperature (Thibodeaux, 1979). At
equilibrium, the mole fraction of a VOC in the air space above the pure
liquid VOC at a specified temperature is expressed as
pa
Ya =- (16.1)
PT
where Ya is the mole fraction of chemical a, pa is the vapor pressure of
chemical a, and PT is the total pressure in the air space.
Vapor pressures for many VOCs at ambient temperatures are avail-
able in the literature (Perry and Chilton, 1973; Callahan et al., 1979;
Verschueren, 1983; Montgomery and Welkom, 1990). Kerfoot (1991)
has evaluated the effects of temperature on the partitioning which is
discussed at the end of this section.

16.2.2 Partitioning of VOC between soil gas and soil moisture

Partitioning between the VOC vapor in the soil gas and VOC dissolved
in soil moisture may be expressed as, KH , the ratio of its concentration
in each of the two phases.
_ CC
KH- (16.2)
CL
where Cc is the concentration of the VOC in soil gas, and CL is the
dissolved concentration of the VOC in the water phase. At equilibrium,
this ratio is constant for constant temperature and is referred to as
Henry's Law constant (Thibodeaux, 1979).
Henry's Law constant may also be expressed as a function of the
VOC vapor pressure, the concentration of the VOC in water, and
temperature as (Thibodeaux, 1979).

(16.3)
 Vapor transport in the vadose zone - theory 381

where Ma is the gram molecular weight of the VOC, T is the tempera-

ture (in kelvin), and the other parameters are as previously defined.
Dilling (1977) reports values of Henry's Law constant for numerous
chlorinated solvents. Values of KH for selected VOCs are presented
in Table 16.1. Empirically derived values of Henry's Law constants
reported by Dilling (1977), Swallow and Gschwend (1983), and Lappala
and Thompson (1983a) are in reasonable agreement with the calculated
values of KH , keeping in mind the temperature dependence of KH .
Ong and Lion (1991) found through empirical tests that sorption of
VOCs at low vapor concentration «10 mg/I) into soil moisture can
be accounted for by the linear partitioning process as described by
Henry's Law. However, they found that at very low moisture contents
(equivalent to less than eight mono layers of water), the sorption
isotherm became nonlinear, with the observed partitioning into the
water phase being higher than that predicted by Henry's Law. Ong
and Lion (1991) theorized that the increased sorption at low moisture
content may be due to sorption onto the bound water on mineral
surfaces, as opposed to being solely dissolved. To expand on this
theory, the observed increase in VOC sorption by soil moisture over
that predicted by Henry's Law may be attributable to a greatly expand-
ing surface area of water across which partitioning may occur as soil
moisture decreases below a critical value, i.e., a nonlinear relationship

Table 16.1 Reported values of Henry's Law constant, vapor pressure, and
solubility at 25°C for selected chlorinated solvents

chemical solubility vapor Henry's

ill water pressure Law Constant
(ppm) (mmHg) (dimensionless)
~- ------

calculated found
-- -- ---- - - ---- - -- -------

1,1,2,2-Tetrachloroethane 3000 6.5 0.019 NA

1,1,2 -Trichloroethane 4420 23 0.038 NA
1,I-Dichloroethane 8700 82 0.050 0.04*
Tetrachloroethylene 140 18.6 1.2 0.50 t
0.43*
Trichloroeth ylene 1100 74 0.49 0.33 t
trans-Dichloroethylene 6300 326 0.27 NA
cis-Dichloroethylene 3500 206 0.31 NA

NA: not available. Source: Dilling (1977).

* Empirical values reported by Dilling (1977).
t Empirical values will vary from calculated values due to differences in temperature.
t Empirical values reported by Lappala and Thompson (1983a).
382 Vapor analysis /extraction

may exist between soil water-surface area and soil moisture content.
Mineral morphology may also play an important role in the observed
increased sorption rates. The data of Ong and Lion (1991) indicate that
the critical soil moisture content below which vapor-water partitioning
becomes nonlinear varies according to soil type. While the critical
soil moisture for clay minerals such as montmorillonite and kaolinite
appears to be eight monolayers, that for alumina is much lower, as low
as four mono layers of water molecules. However, these low moisture
contents will be limited principally to arid regions and shallow soil
during dry months in humid regions.

16.2.3 Partitioning of VOC between soil moisture and soil solids

Ong and Lion (1991) describe sorption of VOCs onto soil solids from
soil vapor as a two-step linear process for soil moisture above a critical
value. Above that soil moisture content, soil solids will be coated with
layers of water molecules. VOC vapor will partition from the vapor
phase into the liquid water phase. Once in the liquid water, some of
the VOC will be adsorbed onto the soil mineral and organic matter. At
equilibrium; the degree of partitioning between the soil solids and the
soil moisture has been expressed as the linear isotherm (Roberts and
Valocchi, 1981)
5
Ko=- (16.4)
CL
where Ko is the partition coefficient or distribution coefficient (with
units of 13 1m), 5 is the mass of chemical adsorbed per unit dry mass of
soil solids, and CL is the concentration of the chemical in the soil
moisture.
It has been observed that strongly hydrophobic organic chemicals
tend to adsorb more strongly onto the soil solids (Roberts and Valocchi,
1981). Empirical studies by Karickhoff et al. (1979) and Karickhoff (1984)
found that Ko was proportional to the organic carbon content of the
soil as well as the octanol: water partition coefficient (Kow) which
is a measure of the hydrophobicity of an organic chemical. For the
equilibrium condition, this relationship has been expressed by the
following equation (Karickhoff et al., 1979) and the essentially same
relationship found by Rao et al. (1985).
Ko = O.63Kowfoe (16.5)
where Ko is the distribution coefficient of Eq. (16.4), foe is the soil
organic carbon content, and Kow is the octanol: water partition
coefficient.
The amount of carbonaceous matter in the soil is the dominant factor
controlling the extent of adsorption of organic chemicals. Karickhoff
 Vapor transport in the vadose zone - theory 383
Table 16.2 Reported values of octanol: water partition coefficient and calculated
values of partition coefficient for selected chlorinated solvents

chemical octanol: water calculated partition

partition coefficient t,
coefficient* fraction organic carbon

0.001 0.01

1,1,2,2-Tetrachloroethane 2.56 0.23 2.3

1,1-Dichloroethane 1.79 0.04 0.4
Tetrachloroethylene 2.88 0.48 4.8
Trichloroethylene 2.29 0.12 1.2
trans-Dichloroethylene 1.48 0.02 0.2
cis-Dichloroeth ylene 1.48 0.02 0.2

* From Callahan et al. (1979).

tCalculated from Ko = O.63K ow/..o.

et al. (1979) also found that the particle size of the mineral fraction
was important. For example, the partition coefficients for semivolatile
compounds, such as pyrene and methoxychlor, for the sand-sized frac-
tion are approximately 100 times less than their partition coefficients
for the silt- and clay-sized fractions, due primarily to the lower organic
carbon content of the sand (Karickhoff et al., 1979). Table 16.2 presents
data for Kow and calculated values of Ko using Eq. (16.5) for selected
VOCs. From the example calculations of partition coefficients in Table
16.2, these VOCs are not strongly adsorbed onto the soil solids due
to their relatively low octanol: water partition coefficients. In com-
parison, compounds with low volatility and low solubility, such
as pentachlorophenol with a log Kow of 4.74, has a Ko of 35, i.e.,
pentachlorophenol will be preferentially adsorbed to the soil solids by a
factor of 100 to 1000 times greater than the chlorinated solvents listed
in Table 16.2.

16.2.4 Effects of temperature and soil moisture content on

partitioning
Kerfoot (1991) evaluated the theoretical effects of temperature and
pore-water content on the partitioning of VOCs between the soil gas,
soil water, and soil solids, assuming thermodynamic equilibrium con-
ditions. For compounds with a Henry's Law constant of 0.05 (for
example, the chloroethanes), effects of temperature on partitioning are
negligible up to about 40 0c, above which partitioning of the VOC to
the soil-gas phase increases. For VOCs with a Henry's Law constant
384 Vapor analysis/extraction

of 1.0 (such as tetrachloroethene), a relatively constant increase in

partitioning to the gas-phase is predicted by Kerfoot (1991). Based on
his sensitivity analysis, there should be no significant difference in the
relative partitioning between the three phases with respect to tempera-
ture for a 25% water saturation versus a 75% water saturation under
the normal temperature range.
Smith et al. (1990), based on data developed from a site investigation
and laboratory experiments, found that TCE adsorption decreases as
soil moisture content increases from zero to the saturation soil moisture
content (the soil moisture content that is in equilibrium with 100%
humidity). They also found that results of soil and soil-gas sample
analyses indicated that the ratio of the concentration of TCE adsorbed
on the vadose-zone soil to its concentration in the 'soil gas was one to
three orders of magnitude greater than the ratio predicted by equilibrium
models. They concluded that the apparent disequilibrium resulted
from the slow desorption of TCE from the organic matter in the soil
relative to the faster volatilization loss of TCE from the soil system.
Therefore, the assumption of Henry's Law constant and equilibrium
adsorption isotherms may not hold in the field.

16.2.5 Transport of VOC vapor through soil gas

Transport of VOC vapor through soil gas may be under density
gradients, thermal gradients, pressure gradients, or concentration
gradients.

Density gradient
Density-induced convective flow of VOC vapor may be important
where concentrations of denser VOCs are high, especially near the
source. Falta et al. (1989), Sleep and Sykes (1989), and Mendoza and
Frind (1990a; 1990b) concluded that convective flow due to density
gradients may be important. Sleep and Sykes (1989) concluded that
transport due to a density gradient was important when the density
contrast is more than several times the ambient soil-gas density.
In conducting sensitivity analyses of density-dependent VOC vapor
transport, Mendoza and Frind (1990a) concluded that density gradients
would be important at relative densities greater than 1.15.
The importance of density gradients to the transport of VOC vapor is
related to permeability. Falta et al. (1989) and Mendoza and Frind
(1990a) concluded that density-driven transport can be neglected for
permeabilities of less than 10- 11 m 2 (10.1 darcy), i.e., equivalent to a
clean sand (Freeze and Cherry, 1979). Since the effective permeability
of the material to gas is affected by the water content, density-driven
transport should be expected to become less important as the moisture
content increases.
 Vapor transport in the vadose zone - theory 385

Aside from consideration of permeability, density-driven transport of

VOC vapor through the unsaturated zone will be important for regions
close to a source of dense volatile compounds and will be important
early in the history of a spill when free-phase liquids are present.
As the source dissipates and as distance from the source increases,
density-driven transport will become less important. Lin (1990) con-
cluded that density contrast would be an unimportant factor in vapor
transport from contaminated groundwater unless nonaqueous phase
liquid (NAPL) were present above the water table.

Thermal gradient
Thermally-induced convective transport of VOC vapors is generally
limited to the near surface due to the rapid attenuation of temperature
fluctuation with depth. While no research on the thermally-driven
transport of VOC vapors through soil has been reported, research on
transport of water vapor has been. For example, Hillel (1972) stated
that the warming of soil lowers the suction pressure and raises the
vapor pressure of soil water. The resulting thermal gradient induces
water vapor migration from warmer to cooler regions in the soil.
Therefore, the effect of the warming of the soil surface during the day
would be to decrease the upward flux of water vapor.
Thermal gradients in soil that may be significant to soil-gas transport
are generally limited to the upper several feet of the soil column.
Marshall and Holmes (1979) state that the amplitude of thermal fluctua-
tions in soil decreases exponentiaIIy with depth below the surface.
They present the foIIowing empirically derived equation for estimating
the temperature variation as a function of depth and time.

T(z,t) = Ta + Aoe
-(-"-t' sin (2nt
tK ~ -
(n
1:K
)0.52) (16.6)

where T is soil temperature, Ta is the average soil temperature, A o is

the amplitude of the soil surface temperature fluctuation, t is time (day
or year), 1: is the period of the cycle (day or year), and K is the thermal
diffusivity. The thermal diffusivity is assumed constant, although, in
reality, it varies with moisture content.
Soil temperature fluctuates on diurnal and annual cycles. Based on
Eq. (16.6), diurnal soil temperature fluctuations caused by the rising
and setting of the sun are damped out within a short distance below
the soil surface. Figure 16.1 presents the results of applying Eq. (16.6)
for the following conditions:
midsummer with an average soil temperature of 30°C;
a surface soil temperature amplitude of 20 0c; and
a constant average soil moisture content of 20-30% resulting in a
constant thermal diffusivity of 2.0 x 10- 7 m 2 /s.
386 Vapor analysis/extraction

60

0.0 depth
50

40
2: average soil temp
l!!
::I 0.2 mdepth \
~
(I)
30
Q.
E
.lB
·0 20
III

10

0
0 0.2 0.4 0.6 0.8 1.0
time of day (fraction)

Fig. 16.1 Depth of soil temperature fluctuations under ambient diurnal

surface temperature changes for an average midsummer soil temperature of
30°C (77°F) for three soil depths.

These conditions would be common to a large portion of humid tem-

perate climatic regions with vegetated soil or partial shade of the soil
surface (daytime maximum temperatures for either bare or drying and
sparsely vegetated soils in direct sunlight would be expected to be
higher). Figure 16.1 shows the diurnal soil temperature variation at
three depths: 0.0, 0.1, and 0.2m. At a depth of 0.2m, the diurnal soil
temperature variation has been damped out to less than 5% of the
average soil temperature. Thus, in general, diurnal soil temperature
fluctuations are not an important factor in VOC transport through the
vadose zone.
For seasonal soil temperature changes, Fig. 16.2 shows the results
for the soil depths of 0.0, 1.0, and 3.0 m for the same conditions as
Fig. 16.1, except the average annual soil temperature is 17°C and the
amplitude is 13 dc. Over the year, the fluctuation in the average soil
temperature is not damped out until much deeper, i.e., the largest
fluctuation is less than 10% of the average soil temperature at a depth
of 3.0m.
Soil temperature fluctuations may be an important factor on the
partitioning of the VOC between the liquid, gaseous, and adsorbed
phases (Kerfoot, 1991). However, the vertical soil temperature profile
below a depth of 0.2-1.0 m can be considered constant over the short
term of several months. Laterally, where soil surface conditions may
 Vapor transport in the vadose zone - theory 387

60

50

40
2:
~
::::l 0.0 depth
"§ 30
Q)
a.
E
.$ ---...,:,--
1.0 m depth

3.0 m depth
",&mg" \ ' temp

'6
II)
20

10

0
0 0.2 0.4 0.6 0.8 1.0
time of year (fraction)

Fig. 16.2 Depth of soil temperature fluctuations under ambient seasonal

surface temperature changes for an average annual soil temperature of 17°C
(58 OF) for three soil depths.

change from bare soil, to vegetated soil, to shaded soil, to pavement,

localized lateral thermal gradients may exist at these lateral boundaries
for short periods at shallow depths due to differences in the depth
to which the thermal fluctuation penetrates the soil. These localized
regions experiencing thermal gradients will produce insignificant
effects on the soil-gas transport with respect to the typical scale
involved.

Pressure gradients
Barometric pressure effects in the subsurface have been observed by
Turk (1975), Freeze and Cherry (1979), and Nazaroff et al. (1987).
Barometric pressure changes due to the weight of the atmospheric
column overlying the ground surface at any instant will have spatially
uniform effects over large areas, especially for open ground surfaces.
Barometric pressure changes due to the passage of atmospheric highs
and lows will act to depress and raise the water table slightly (Turk,
1975; Freeze and Cherry, 1979). This barometric pressure fluctuation
will cause a slight compression of soil gas during the passage of a high
atmospheric pressure system and a slight expansion of soil gas during
the passage of a low atmospheric pressure system.
The insignificant effect of atmospheric pressure on the unsaturated
388 Vapor analysis/extraction

zone was substantiated by Turk (1975) who found diurnal pressure

fluctuations of up to 60 mm depth in soil. More recently, Kerfoot (1991)
reported that pressure in the unsaturated zone remained stable over an
8-month period to a depth of 2.1 m and concluded that barometric
pn~sure changes have a negligible effect on soil-gas concentrations of
VOCs.
Significant pressure gradients can be established in the unsaturated
zone due to the effects of manmade structures. Nazaroff et al. (1987)
reported increased flow of gaseous contaminants, gasoline vapors, and
radon, resp~ctively, towards basements. Nazaroff et al. (1987) found
that the artificial venting, or depressurization, of houses, such as for
radon gas mitigation, can increase the pressure differentials by 10-20
times natural differentials, and the velocity of radon gas through the
soil towards the basement can approach 1.0 m/h on a transient basis.

Concentration gradients
In general then, except for the region affected by density-induced
transport of concentrated DNAPL vapors, the primary transport
mechanism for VOCs in the unsaturated soil is by diffusion through
the soil gas under concentration gradients. The distribution of VOC
concentration in the soil gas can be modeled by Fick's second law
which in one dimension is expressed by the following equation
(Thibodeaux, 1979).

(16.7)

where C is concentration of the VOC in air, 0 is the diffusion

coefficient, and z is the distance traveled.
Assuming the outer boundary condition is zero concentration, Eq.
(16.7) can also be expressed as (Thibodeaux, 1979)

C(z,t) z ]
= er f c [ --0-5 (16.8)
C(z=o,t=O) 40t"
where C(z,t) is the concentration (as mole fraction) at a distance z and
time t, C(z=o,t=O) is the initial concentration, and erfc is the com-
plimentary error function.
Swallow and Gschwend (1983) conducted limited controlled
laboratory experiments using a glass tank. Although their experimental
design prevented a direct measurement of the concentration of VOCs
in the unsaturated zone, Swallow and Gschwend (1983) concluded that
volatilization can be adequately modeled by Fick's second law.
Lin (1990) also found reasonable agreement between observed VOC
concentrations in soil gas and those predicted by a diffusion-based
model developed by Silka (1986).
 Vapor transport in the vadose zone - theory 389
16.2.6 Diffusion coefficient in soil gas
The estimated diffusion coefficient for VOC vapor in air is 0.43 m 2 /d
(Jury et al., 1983). This single value for D has been used for inter-
mediate molecular weight VOCs based on the similar values obtained
in measurements of the gas diffusion coefficient for numerous organic
chemicals of intermediate molecular weight (Jury et al., 1984; Lin,
1990).
However, the diffusion coefficient in soil gas has been found to be
reduced from that in air by a tortuosity factor which accounts for
decreased cross-sectional area for flow and increased length of the flow
path through soil pores. Jury et al. (1983; 1984) concluded that the
Millington-Quirk tortuosity formula adequately defines the soil-gas
diffusion coefficient. Bruell and Hoag (1986) confirmed the validity of
the Millington-Quirk model in column experiments. This formula is
expressed by 10
Da 3
Dc = - 2 (16.9)
n
where Dc is the diffusion coefficient in soil gas, D is the diffusion
coefficient in air, e.g., 0.43 m 2 /d, a is the volumetric air content of the
soil, and n is the total soil porosity.
Since the VOC vapor may partition between the gas, liquid, and
solid phases, an effective diffusion coefficient can be formulated that
incorporates that partitioning. The removal of VOCs from the soil gas
by partitioning into the soil moisture and soil organic matter results
in a reduction in the apparent diffusion rate, and consequently, the
apparent, or effective, diffusion coefficient. Jury et al. (1983) developed
the following relationship between the diffusion coefficient in soil gas
from Eq. (16.9) and the effective diffusion coefficient
Dc
DE = bKo W
(16.10)
-+-+a
KH KH
where DE is the effective diffusion coefficient in soil gas corrected for
effects of partitioning, Dc is the diffusion coefficient from Eg. (16.9), b
is the bulk dry density of the soil, Ko is the soil partition coefficient
from Eq. (16.5), KH is Henry's Law constant from Eq. (16.3), W is the
volumetric soil moisture content, and a is the volumetric air content
(where n, total porosity, is equal to a + w). Jury et al. (1984) report that
the model for the effective diffusion coefficient expressed in Eq. (16.10)
gives results that are in good agreement with empirically derived
values of DE.
Research completed on the transport of VOC vapor in the unsaturated
zone remains limited primarily to theoretical modeling studies and
empirical observations from field data.
390 Vapor analysis/extraction

Although there has been considerable effort to characterize the

transport and fate of pesticides in soil, the properties of the pesticides
studied differ substantially from those of VOCs, resulting in a con-
siderable difference in their behavior in the unsaturated zone. Research
is still needed to improve the understanding of VOC vapor transport
and fate in unsaturated soil in order to provide improved interpretation
of soil-gas survey results and assist in designing more cost-effective
clean-up programs.

16.3 SOIL-GAS SURVEYS

Soil-gas surveys are becoming more widely accepted as a tool for

the preliminary determination of the extent of soil and ground water
contamination by volatile organic compounds (VOCs). These surveys
are being applied to the investigation of a variety of problems includ-
ing groundwater plume tracking (Glaccum et al., 1983; Lappala and
Thompson, 1983a,b; Marrin, 1984; Voorhees et al., 1984; Malley et al.,
1985; Silka, 1986; Kerfoot, 1987a,b; Marrin and Thompson, 1987; Silka
and Spectre, 1991; Fusillo et al., 1991), leaking tanks and pipelines
(Mehran et al., 1983; Feere and Silka, 1987; Stuart et al., 1989), surface
spills (Silka, 1986; Evans and Thompson, 1986; Marks et al., 1989),
potential releases from landfills, impoundments, and other sources
(Los Angeles County, 1985; Smith et al., 1990; Millison et al., 1990;
Thomsen and Joyner, 1990), and environmental audits and pre-
construction assessments (Silka, 1986; Rizvi and Fleischacker, 1991;
Viellenave and Hickey, 1990).
Since soil contamination by liquid VOCs typically causes vapor
concentrations in soil gas in the immediate vicinity of the source to
increase over several thousand to tens of thousand parts per million,
outward diffusion of the VOC vapor through soil pores results in easily
detectable concentrations of the VOC in soil gas up to tens of meters
away from the source.
Due to the presence of VOCs at many sites of contamination, there
has been increasing interest in the sampling and measurement of
VOCs to determine their extent in soils and ground water. With the
recent development of portable gas chromatographs that rely on a
variety of detectors, quantitative, as well as semiquantitative, field
analysis of VOCs in soil is now possible.
The portable gas chromatograph, a relatively new technology, has
been shown to be especially applicable to the investigation of soil and
ground water contamination through the analysis of shallow soil gas.
Under diffusive transport, VOCs volatilize from ground water and
move upward through the unsaturated zone, ultimately venting to the
atmosphere. It has been shown that the concentration of VOCs in
 Soil-gas surveys 391
samples of shallow soil gas are related to the concentration of VOCs in
ground water (Glaccum et al., 1983; Lappala and Thompson, 1983a and
1984; Swallow and Gschwend, 1983; Marrin, 1984; Silka and Spectre,
1991). Soil-gas surveys are being recognized as a valuable tool, both
alone and in conjunction with other techniques, that can provide data
on the location and extent of soil and ground water contamination and
aid in the design of more detailed ground water studies involving soil
borings and monitoring well networks.
The application of soil-gas monitoring has been found to be use-
ful not only in delineating soil and groundwater contamination at
known contamination sites, but also as a valuable exploratory tool for
identifying unknown sources of contamination. Soil-gas surveys have
been successfully applied to environmental assessments for potential
liabilities related to real estate transactions. In this mode, soil-gas
surveys can be used to assess whether potential subsurface VOC con-
tamination is present. For current owners of industrial plant sites, soil-
gas surveys can be used as an integral part of environmental audits
to identify VOC contamination from previous accidental spills and
leaking tanks and pipelines.
The Los Angeles County Department of Water and Power, under a
cooperative funding agreement with the US Environmental Protection
Agency used soil-gas surveys for the identification of as yet unknown
sources of VOC contamination in the San Fernando ground water
basin.
The successful conduct of a soil-gas survey depends on several
factors, including the size and age of the source, the moisture content
and organic carbon content of the unsaturated zone, and the volatility
and solubility of the VOc. Prior to the conduct of a soil-gas survey,
the effects of these factors should be evaluated in order to optimize
the design of the soil-gas survey. Through review of theory and the
application of computer models, one can evaluate the operational
limitations of soil-gas surveys. This evaluation provides additional
value in allowing improved interpretation of data obtained from a soil-
gas survey.

16.3.1 Summary of soil-gas sampling and analytical methodologies

A number of soil-gas sampling methods have been developed,
including:
1. grab sampling of soil cores for heads pace analysis;
2. surface flux chambers;
3. downhole flux chambers;
4. accumulator devices; and
5. suction ground probes.
392 Vapor analysis/extraction

Diverse analytical protocols also have been developed and used in soil-
gas surveys. For example, the American Petroleum Institute (API,
1985), Silka (1986), Zdeb (1987), Marrin (1988) and Kerfoot (1988)
evaluated soil-gas sampling methodologies, while Steinberg et al. (1990)
compared the responses of the flame ionization detector (FlO) versus
photoionization detector (PID).

Headspace method
The grab sample method with headspace analysis by GC involves the
collection of a core of soil, sealing of the sample with the exclusion of
as much air as possible, and laboratory GC analysis by the headspace
technique for VOCs coming out of the core. This method is simple,
but the soil gas composition is subject to alteration during handling.
Another problem with grab sampling for subsequent laboratory analysis
is that additional time and expense are incurred when results are not
obtained in the field and resampling is required.

Surface flux chamber

The surface flux chamber method involves the installation of an
enclosure on the surface of the ground. Clean, dry air is added to the
chamber at a known rate, and the concentration of VOCs is determined
by a portable GC as the clean air exits the flux chamber. Knowing the
concentration of VOCs in the exiting air and the rate of air flow, the
flux or rate of VOCs emanating from the ground can be calculated.
Although the technique allows rapid determination of the flux of VOCs
through the ground surface, the detection limit of the technique is
necessarily reduced by the dilution effect of the clean air that is intro-
duced (API, 1985). Millison et al. (1990) reported on the use of a
surface flux chamber to assess potential releases of VOCs to the air
for purposes of deriving a hazard ranking score (HRS) for EPA's
Superfund program. While they also employed the suction probe
method, they did not compare the two methods.

Downhole flux chamber

The downhole flux chamber method follows the same principle as the
surface flux chamber, except the chamber is driven or emplaced below
the surface of the ground. This technique also has the basic drawbacks
as the surface flux chamber method, although its detection limit is
improved somewhat by the generally higher concentrations encountered
with increasing depth below the surface. However, the technique is
more labor intensive.
 Soil-gas surveys 393

Surface accumulator method

The accumulator device technique also is referred to as the surface
static collection method and involves the passive sampling of soil gas
by the trapping of VOCs onto an adsorbent contained within a glass
tube (Zdeb, 1987; Kerfoot and Mayer, 1986). A static trap is placed just
below the soil surface and left in the ground for from several hours to
as long as thirty days. The static trap consists of a ferromagnetic wire
coated with activated charcoal that is contained in an inverted test
tube. The trap passively collects diffusing VOCs by adsorption onto the
activated charcoal. The trap is sealed and taken to the laboratory for
analysis of VOCs by Curie point desorption mass spectrometry. The
static collection method provides accurate, time-integrated determina-
tion of gas flux (Malley et aI., 1985). The sampling time can be adjusted
to provide the necessary sensitivity and the technique is simple and
low cost (API, 1985). However, the longer sampling time prolongs the
field study. API (1985) reported uncertainty concerning the retention
efficiency of the activated charcoal.
In addition, the method is limited to shallow soil depths, and vertical
profiling is not feasible. Since the traps are returned to the laboratory
for analysis, the accumulator method does not allow the sampling
design to be altered in the field on the basis of previous results.
Another complication in using the static accumulator traps is that
they will trap any VOC that happens to impinge on the trap. Thus,
extraneous atmospheric sources, such as VOCs from transient auto
emissions, can be caught by the accumulator trap. To control for
this potentially confounding source of VOCs, extra traps must be set
and collected periodically during the course of the total accumulation
period to assess potential effects of transient sources.
Viellenave and Hickey (1990) reported on the use of surface
accumulator traps for an environmental assessment of an office build-
ing. After a one-week sampling period, the results indicated low level
TCE and PCE contamination in the subsurface. Their results illustrate
the potentially low detection limits possible with the accumulator
traps.

Suction ground probe

Ground probe techniques for sampling soil gas involve the insertion of
a tube into the ground to a desired depth, placing a vacuum on the
tube, and analyzing the extracted air for VOCs, typically by field Gc.
API (1985) concluded that the technique is relatively sensitive and,
because the sampling depth can be varied, surface interferences can be
avoided, and sensitivity can be increased if needed. Although API
(1985) considered the ground probe technique labor intensive and not
394 Vapor analysis/extraction
suited to finer grained, clayey soils or where bedrock is near the
surface, the technique has been successfully applied to numerous
sites having a wide range of environmental conditions (e.g. Silka and
Spectre, 1991 and Thomsen and Marrin, 1987).
Thomsen and Joyner (1990) used hand-hammered shallow probes to
1.0 m, hydraulically pressed probes to about 6.0 m, and hallow-stem
auger drilling with a driven probe ahead of the auger down to 36 m to
obtain vertical profiles of VOCs in soil gas. Hammered and hydraulically
driven probes are not good in resistant soil such as coarse gravel or
dense bouldery till. Hand-held electric rotary hammers fitted with
small diameter augers and small hydraulically driven augers have been
used frequently to depths of 6.0m.
Selection of the sampling method depends on the study objectives,
time constraints, and budget. The suction ground probe technique
has been reported in the literature more often than other collec-
tion methods. The following points have been made concerning this
method.
1. In combination with laboratory grade GCs, the method is sensitive
to a wide range of VOCs and allows separation of component VOCs
in soil gas.
2. The method provides rapid turn-around of VOC analyses so that
sampling stations can be adjusted in the field to define better the
extent of soil contamination.
3. In combination with the hand-held rotary hammer method of collec-
tion, the suction ground probe allows sampling in areas of restricted
access.
4. The method allows sampling of soil gas through pavement as well
as open ground.
5. The method is easily adapted to vertical profiling which often is
important for the identification of the VOC source.

16.3.2 Analytical methods and quality control

Soil-gas surveys have been conducted with a full gamut of instruments
from the portable organic vapor analyzer (OVA) equipped with a
flame ionization detector (FlO) and photoionization detector (PIO), to
portable gas chromatographs, to laboratory grade GC or GC/MS instru-
ments. Steinberg et al. (1990) presented a comparison of the response
of field PIO versus FlO and field FlO versus lab GC-FIO. The response
factor for the OVA with FlO changes with hydrocarbon composition
and must be corrected for to make accurate interpretations of field
data. Steinberg et al. (1990) found that the response factor for the FlO
decreases geometrically with carbon number. In their comparison of
the field screening FlO versus lab GC-FIO, they found good agreement
 Soil-gas surveys 395

(correlation of 0.94) between the method with an almost one-to-one

relationship (slope of 0.89).
Marks et al. (1989) made a comparison of PID versus FID for soil-gas
surveys involving petroleum hydrocarbon. They found that the evalua-
tion of a site based only on the concentrations of total hydrocarbons
can result in losing valuable information. Indeed, the total hydrocarbon
reading with an FlO instrument cannot discriminate between the target
compounds of concern and natural methane and light aliphatic hydro-
carbons, which are abundant in nature. Marks et al. (1989) also con-
cluded that, when calibrated with benzene and toluene standards,
both FID and PID GCs will give comparable results, but the total
hydrocarbon response between the two detectors can vary by five to
ten times.
In general, the PIO is more responsive to aromatic and ring com-
pounds. The FlO is a good general detector for hydrocarbons, which
has led to its frequent application in soil-gas surveys, not to mention
the past ready availability of FlO instruments. However, more recently,
the PIO-equipped instruments have become the most commonly used
due to their ease of use.
The electron capture detector (ECO) is the most sensitive detector for
halogenated organics, but can be easily contaminated and difficult to
operate in the field. The argon ionization detector (AID) has been
overlooked for application to soil-gas surveys, although it has sig-
nificant advantages in being very sensitive to a wide range of VOCs
including both aromatic and hydrocarbon compounds.
Because of the poor detection limits and lack of qualitative
chromatographic information available from field FIO-OV As and PIO-
TIPs, field GCs and transportable laboratory-grade GCs are considered
state-of-the-art in soil-gas analyses.
Marks et al. (1990) discussed quality assurance and quality control
concerns for soil-gas surveys. They found that soil-gas surveys have a
100-500% variability for calibrated constituents and up to 1000%
for noncalibrated and total volatile hydrocarbons analyses. They
believe that only a small amount of the variability is derived from the
instrumentation, and that most of the variability is associated with
sampling error and spatial/temporal variability. Marks et al. (1990)
found that variability in sampling depth and variability in vacuum
pressure during soil-gas collection are important factors that are dif-
ficult to come to grips with. In addition, variability can be introduced
through short-circuiting of atmospheric air along the probe shaft.

Computer modeling of soil-gas transport

Although several investigators have developed models for the simula-
tion of the transport of organic chemicals in the soil (Leistra, 1973; Jury
396 Vapor analysis/extraction
et al., 1983, 1984; Rao et al., 1985) these models were limited in their
application to the simulation of the diffusion of VOCs in soil gas. In
general, the previous models were developed for application to the
modeling of pesticide movement and fate in soils. The models are one-
dimensional analytic solutions that do not allow for heterogeneous
soil properties and initial conditions. Also, these models incorporate
transport of the chemical in the liquid phase, as they are intended for
the study of leaching and volatilization of pesticides from soils (Jury et
al., 1983). In the case of the model developed by Rao et al. (1985),
transport by vapor diffusion was omitted.
Silka (1986) described a two-dimensional finite-difference vapor dif-
fusion model and the results of sensitivity analyses of vapor diffusion
through soil gas. Sensitivity analyses demonstrated the importance of
soil moisture and organic matter content in controlling the migration of
VOC vapor through the unsaturated zone.
Sleep and Sykes (1989) developed a two-dimensional finite-element
model which incorporated transport of VOCs in variably saturated
media. Their model included dissolved phase flow and transport, as
well as density-dependent flow and transport of vapor. Mass transfer
was controlled by mass transfer coefficients, rather than traditional
partitioning coefficients. The mass transfer coefficients were varied to
give reasonable results.
Mendoza and McAlary (1990), Mendoza and Frind (1990a,b) devel-
oped a two-dimensional finite-element model which incorporates
diffusion, advection due to density gradients, and advection due to the
vapor mass released by vaporization at the source. For their numerical
experiments, they assume equilibrium partitioning between an oil,
adsorbed, dissolved, and vapor phase.
As an example of a soil-gas model, the two-dimensional vapor dif-
fusion model, 20-DIFF, developed by Silka (1986) is discussed. This
model is a finite-difference, forward difference approximation relative
to time, and is based on Fickian diffusive transport. The model is based
on the following assumptions.
1. Diffusion is described by Fick's Second Law.
2. Partitioning coefficients are linear and system is at equilibrium with
respect to partitioning.
3. The Millington-Quirk tortuosity formula is valid.
4. The soil properties of bulk density and total porosity are
homogeneous.
5. The diffusion coefficient in air is constant for all VOCs.
6. The soil gas is at constant atmospheric pressure.
7. The system is isothermal.
The model allows for heterogeneous initial concentrations with either
constant concentration sources or instantaneous spike sources. The
 Soil-gas surveys 397
1.0~-----------------------------------------------,

0.8

~
~ 0.6
c
0
~
c
Q)
u 0.4
c
0
u

0.2

o+----,---.----~--~--~--~~==~--~--~--~
o 6 12 18 24 30
distance (m)

Fig. 16.3 Comparison of one-dimensional analytical solution of the Fickian

diffusion equation (solid line) to the one-dimensional simulation using the
soil-gas vapor diffusion model 2D-DIFF after 90 days from (solid squares) a
constant concentration source, from Silka (1986).

diffusion coefficients may be varied over the finite-difference grid and

in the x- and y-directions by weighting coefficients. The effects of
partitioning are incorporated by using the effective diffusion coefficient
as defined in Eq. (16.10).

16.3.4 Model verification

The verification of the computational algorithms for the vapor diffusion
model 2D-DIFF has been done by comparisons with computed results
from the one-dimensional analytical solution. The results of this com-
parison are presented in Fig. 16.3 for an elapsed time of 90 days. The
analytical solution is represented by the solid line and the solution
computed by 2D-DIFF is presented as solid squares.

16.3.5 Model validation

Preliminary validation of 2D-DIFF has been performed by Lin (1990)
based on field data from Marrin and Thompson (1987) and Kerfoot
(1987a,b). An adequate validation problem requires spatial information
on the moisture content and organic matter content of the soil, the soil
398 Vapor analysis/extraction

texture, as well as the source characteristics, whether a surface spill or

contaminated ground water. In many instances this information is not
gathered as part of a routine soil-gas survey. Further complications
arise when one considers that field conditions are dynamic, i.e., always
changing. This problem is especially acute for soil moisture content
which, over the large scale, fluctuates seasonally. Also, field data are
not available in sufficient detail to allow description of the spatial
variability of soil conditions resulting in necessary oversimplification of
the physical setting.
Jury (1986) lists several potential problems that must be considered
to carry out successfully a field validation experiment of a vapor dif-
fusion model:
1. lateral and vertical variability of transport and retention parameters
that may introduce heterogeneities and anisotropy that are not
included in the model;
2. macropores, cracks, plant root holes, and animal burrows that may
create discontinuities difficult to account for in the model;
3. time-dependent boundary conditions, such as depth to water table
and seasonally saturated zones that alter the system geometry from
that modeled;
4. problems with validity of measurement techniques for characteriz-
ing field properties; and
5. scale effects that may be important in the field that are not accounted
for in the model, such as localized departure from isothermal condi-
tions or physical heterogeneities.
Striegl (1987) published an account of an apparently successful
modeling of the diffusion of methane from a waste disposal trench in
Illinois using a two-dimensional finite-difference solution of Fickian
diffusion similar to the model described here.
The validation exercise completed by Lin (1990) using 2D-DIFF
provides an example of these difficulties listed by Jury (1986). In the
first validation case, Lin (1990) used data presented by Marrin and
Thompson (1987) from a TeE spill site in Arizona, where the water
table is approximately 30 m deep. Samples were collected through a
probe which was driven ahead of a hollow-stem auger. Included in
these data were soil texture, water content, and air porosity. The soil
organic fraction was assumed to be 0.005 based on a sandy matrix.
Initial vapor concentrations above the water table were calculated
based on measured groundwater concentrations. On the first profile,
the simulated data are, in general, lower than the actual data. A sharp
drop in soil gas concentration across a sandy layer from 3 to 6 m was
both observed in the field and predicted by the model. Results from a
second profile were qualitatively similar where concentrations dropped
across a clay layer, but in general the simulated results were about an
 Soil-gas surveys 399
order of magnitude different than the measured concentrations (Lin,
1990).
In the second case, Lin (1990) used data from Kerfoot (1987a,b) on a
chloroform spill in Nevada, where the water table is 3-4 m below
grade. The soil type is a sandy gravelly loam, with several percent clay.
Samples were collected through a probe hammered into the ground to
a depth of about 30cm. Vertical profiling was also conducted at several
points. Initial vapor concentrations were calculated using measured
groundwater concentrations. In the analysis by Lin (1990), simulated
results matched field observations more closely near the surface, while
they were somewhat elevated above observed levels at deeper depths.
The observed concentration profile with depth exhibited a linearity
usually associated with a steady-state regime. However, the modeled
results did not reach a steady state in a reasonable amount of time. In
this case there may be errors in the input data or some environmental
factor that is not being taken into account by the model (Lin, 1990).
Several possibilities include advection effects and chemical or biological
degradation. In addition, errors may have been introduced during soil-
gas sampling.
Mendoza and Frind (1990a) compared their numerical results with
laboratory experiments. The laboratory apparatus consisted of a 10 X
10 X 3 m sand-filled box instrumented with gas samplers, to which one
liter of TeE was added over 30-cm square area at a depth of 45 cm. In
general, the match between the observed and simulated results is
good, but it is improved by the inclusion of density driven advection in
the numerical model. This effect is more pronounced deeper in the
soil.

16.3.6 Importance of advection

According to Mendoza and Frind (1990a), significant advection of
subsurface vapors may be caused by density gradients due to high
vapor pressures and high molecular weights of chlorinated solvents.
Advection may also be caused by vaporization of free product at the
source. Advection seems to be an important mechanism for dense,
volatile compounds.
However, advection due to density gradients may only be important
near the source. For example, TeE has an ideal gas phase vapor
density of 5.64kg/m3, while air has a density of 1.26kg/m3. An air
mixture saturated with TeE would have a density of 1.46 kg/m3, pro-
ducing an increase in density of only about 15% (Sleep and Sykes,
1989). Ignoring differences in gas phase density leads to underestimates
of dispersion, and hence the spreading of the vapor plume. A saturated
vapor phase will only exist in the vicinity of the free organic liquid, and
thus density differences may only be important near pools of free
400 Vapor analysis/extraction

product. Saturated vapor conditions could also exist in a high-porosity

material with a significant amount of residually saturated product and
a low organic matter and water content, such as sand residually
saturated with TeE. A sand with a fairly even residual saturation of
free product would serve as a significant reservoir for vaporization and
would probably be affected by advection due to density gradients.
In numerical simulations by Sleep and Sykes (1989), incorporation of
advection results in greater horizontal spreading of the vapor plume.
In addition, gas-liquid partitioning reduces the concentration of dis-
solved TCE. The water acts as a sink for TCE vapor and the movement
of the vapor is retarded by gas-liquid partitioning. Concentration
gradients in the aqueous phase are diminished while concentration
gradients in the vapor phase are increased. Simulations involving
diffusion only showed reduced spreading of the vapor and dissolved
plume. Decreased rates of dissolution and volatilization were also
noted, since advection was not available to transport extra mass away
from the source.

16.3.7 Sensitivity analyses

Silka (1988) conducted sensitivity analyses with 2D-DIFF to assist in
designing soil-gas surveys. The objectives were to:
1. estimate the optimum grid spacing for soil-gas surveys;
2. determine what the optimum soil conditions are, especially moisture
content, to maximize the probability of successfully detecting a
surface source from a distance; and
3. determine the level of interference due to upward diffusion of VOC
vapor from underlying contaminated ground water and determine if
the ground water source of VOC vapor may interfere with the
detection of a surface source and vice versa.
The parameters utilized in the sensitivity analyses are presented in
Table 16.3.
Porosity was assumed constant in all runs at a value of 0.4.
Divergence from this value in the field will affect the value of Dc
computed in Eq. (16.9). A change from 0.4 to 0.3 for total porosity
results in Dc increasing by a factor of 1.8, while decreasing porosity
from 0.4 to 0.35 increases Dc by a factor of 1.3. Therefore, the value of
porosity selected will not affect the results significantly when com-
pared to uncertainties in other parameters.
Partitioning of the VOC between the vapor and the soil moisture and
soil solids produces significant retardation of the migration of the
vapor through the soil gas. For the purposes of this sensitivity analysis,
the organic carbon content is assumed zero. However, inspection of
 Soil-gas surveys 401
Table 16.3 Parameters used in modeling diffusion of 1,1,2,2-Tetrachloroethane
and Trichloroethylene through soil-gas using 2D-DIFF.3

Unsaturated zone
Porosity 0.4
Moisture content 0.0,0.04,0.1,0.2
Fraction organic carbon content 0.0, 0.01, 0.03
Temperature 20°C
Bulk density 1350kg/m3
1,1,2,2 -Tetrachloroethane
Molecular weight 168gm/mole
Saturated vapor concentration 6000ppm
Diffusion coefficient in air 0.43m2 /d
Henry's Law coefficient 0.02
Octanol: water partition coefficient 300
Trichloroethylene
Molecular weight 131 gm/mole
Saturated vapor concentration 72 000 ppm
Diffusion coefficient in air 0.43m 2 /d
Henry's Law coefficient 0.4

Eqs. (16.5) and (16.10) shows that even the slightest amount of organic
matter will greatly reduce the effective diffusion coefficient.
Soil moisture content, on the other hand, was varied over four
values (0.0, 0.04, 0.1, and 0.2, equivalent to 0%, 10%, 25%, and 50%
saturation, respectively) of the total pore space filled with water. This
range in soil moisture is typical of those observed for sand (Marshall
and Holmes, 1979).
An additional assumption utilized in the sensitivity analyses of
Silka (1988) is that the soil moisture fluctuates seasonally, a common
occurrence in humid temperate regions. The soil moisture in storage
sustains a net decrease during the summer months when potential
evapotranspiration (ET) exceeds available precipitation. No net down-
ward flux of water occurs to wash VOCs out of the unsaturated zone.
However, during the fall, winter, and spring months, the soil moisture
in storage increases as the potential ET drops off. During these nine
months, a net downward water flux occurs.
Although 2D-DIFF is not designed to simulate the transport of VOC
in the dissolved form in downward percolating water, intuitively, the
addition of clean recharge water to the soil will extract VOC vapor from
the soil gas. The magnitude of the reduction in the concentration of
VOC vapor in the soil gas by dissolution in the incoming recharge
water will depend on the mass of recharging water and the value of the
Henry's Law constant.
402 Vapor analysis/extraction
The net effect of the rainy-season recharge is assumed to essentially
wash out all VOC vapor in the unsaturated zone. Therefore, the
sensitivity analyses are 90 day simulations of the summer months. This
assumption will not be conservative for the arid southwest and will
result in the underestimation of diffusion distances. In the humid
temperate regions, this assumption may overestimate the diffusion
distances. Results of the sensitivity runs based on the parameters in
Table 16.3 are presented in Figs. 16.4 through 16.6.
The sensitivity analyses of 2D-DIFF demonstrate the importance of
soil moisture and organic matter content, as well as the VOC-specific
properties of volatility and solubility, to the transport of VOC vapor in
the soil gas. These results have been applied to the design of soil-gas
surveys in order to optimize grid spacing.
Mendoza and Frind (1990b) performed extensive sensitivity analyses
of their finite-element model. They also choose to neglect partitioning
to the solid phase. A base case which includes the effects of both
advection and diffusion shows a large plume with high vapor and
dissolved concentrations developing rapidly. The density effect of their
generic high molecular weight compound is evident in the sinking of
the plume. Near the plume front, where concentrations in the vapor
phase decrease, the density effects are much less apparent. In the
diffusion-only case, the plume does not move out as fast from the
source without the extra action of density-induced advective forces. In
addition, the plume spreads more uniformly with depth, rather than
sinking towards the water table. Since less mass is sinking through the
soil matrix, more of the source mass is lost to the atmosphere.
Sensitivity analyses for permeability revealed that permeabilities
below which advection becomes the dominant transport process cor-
respond to those of a relatively dry clean coarse sand.
Diffusion will be the dominant transport process in medium sand
and finer-grained matrices. Variations in source vapor density produce
effects similar to those caused by varying the permeability. The ad vec-
tive component of transport increases with increasing density, and vice
versa.
In the pure diffusion case, the mass vaporized as well as the net
mass in the system increases with the source concentration, but is not
affected by variations in permeability. The case is similar when advec-
tive mass flux due to vaporization is included, but increases in source
concentration will cause a correspondingly greater increase in the mass
vaporized and the net mass in the system.
With advection due to density gradients included, an increase in the
source concentration causes very large increases in the mass vaporized
and net mass in the system. Mass retention increases with permeability
due to sinking vapors (Mendoza and Frind, 1990b).
 Soil-gas surveys 403

16.3.8 Soil-gas survey grid spacing

Determination of the appropriate grid spacing for soil-gas surveys
has been left primarily to best judgment of the investigator. Godoy
and Naleid (1990) stated that grid spacings typically are two to four
times the depth to the water table and that 15-30 m grid spacings
are frequently employed for screening large areas. Under ideal,
homogeneous, dry conditions, with no organic matter, soil gas would
be expected to diffuse laterally to a distance equal to the thickness of
the unsaturated zone. In coarse-grained soils, this may hold. In fine-
grained soils, however, a significant capillary fringe may be present
which could diminish the effective depth to the water table. Also,
increasing soil heterogeneities, water content, and organic matter, as
well as age of the source of VOC contamination will decrease the
distance of VOC migration through soil gas.
The simulation of the diffusion of trichloroethylene (TCE) and
1,1,2,2-tetrachloroethane (TET) vapors through the soil away from a
surface spill (Silka, 1986) indicate that a minimum grid spacing of
7-11 m is adequate to allow the detection of a spill in a low organic
soil. For soil that has been dry for a period of several months, such
as during summer droughts, the grid spacing may be increased to
13-16m.
The results of the sensitivity analyses for diffusion of TCE and
TET vapors under varying soil moisture conditions demonstrate the
importance of soil moisture to successfully conducting the soil-gas
survey. Dry soil represents the optimum soil moisture conditions for
the soil-gas survey. Soil moisture is not as critical a factor for detection
of a source of TCE or other more volatile VOC. However, soil moisture
is a critical factor for successfully detecting a source of TET or other less
volatile VOCs and VOCs with small Henry's Law constants.
The computer simulations reviewed indicate that, as the soil moisture
content increases above 0.0%, the rate of diffusion rapidly decreases
(as shown in Fig. 16.4). Thus, the grid spacing used in a soil-gas survey
is critically dependent upon the soil moisture content. Under dry soil
conditions, a grid spacing of up to 30 m may be sufficient to ensure
detection of a source. However, under wet soil conditions, a grid
spacing of less than 10 m may be required. These grid spacings would
decrease for unsaturated zones that are less than 30 m.

16.3.9 Discriminating groundwater versus surface sources

A problem that arises in the course of soil-gas surveys for both detec-
tion of surface sources, but more so, for mapping ground water con-
tamination, is the potential interference caused by VOC vapors from
404 Vapor analysis/extraction

1.0.------------------------------------------------,

0.9 %
curve
moisture
0.8 A 50.0
B 25.0
C 10.0
~ 0.7 D 0.0
~
"
~ 0.6
c
.Q
§ 0.5
E
~ 0.4
c
o
<.>
0.3

0.2

0.1

o+---,---~~-.~~=-~==~r===~--,_--._--~
o 6 12 18 24 30
distance (m)

Fig. 16.4 Distance of diffusion of trichloroethylene under varying soil

moisture content (no soil organic matter) after 90 days from a constant
concentration source, from Silka (1986).

one source confusing the interpretation of the field data concerning the
other source. The interferences are especially problematic in highly
industrialized areas with multiple sources. Where the GC fingerprint
differs, interpretation is made easier. However, where similar com-
pounds are involved in multiple sources, discrimination between the
VOC sources is made more difficult.
Computer simulations were completed to estimate the interference
that may arise from the upward diffusion of VOC vapors emanating
from contaminated ground water. For the modeling, it was assumed
that the highest VOC concentration in ground water underlying the
area of the soil-gas survey was 100 ppb. The depth to the water table
was almost 10 m. This level of ground water contamination could result
in soil-gas concentrations in the upper one meter of soil to be as high
as 150 ppb. Thus, a soil-gas survey for surface sources would not be
able to distinguish surface sources until the vapor concentration due to
the source was above the 150 ppb level.
Conversely, the results of soil-gas surveys for mapping ground water
contamination will be uninterpret~ble in regions near surface sources
of VOC vapor. Inspection of Figs. 16.4 and 16.5, for instance, indicates
that, out to a distance of about 30 m, the concentration of VOC in soil
gas due to the surface source may be greater than 0.001 of the source
 Soil-gas surveys 405
1.0

0.9 Henry's
curve Law
0.8 constant
TET 0.02
~ 0.7 TeE 0.40
r,J
Q
~ 0.6
c
0
~ 0.5
C
OJ
c 0.4
()

0
()

0.3

0.2

0.1

0
0 6 12 18 24 30
distance (m)

Fig. 16.5 Comparison of simulated diffusion of 1,1,2,2-tetrachloroethane

(TET) and trichloroethylene (TCE) in soil with 25% moisture content and no
organic matter, showing the effects of the difference in vapor/water partition
coefficient represented by the Henry's Law constant, from Silka (1986).

concentration. When the saturated vapor concentration for TeE is

72 000 ppm (see Table 16.3), the concentration in soil gas at 30 m may
be as high as 72 ppm. For TET, with a saturated vapor concentration of
6 000 ppm, the concentration in soil gas at a distance of 30 m may be as
high as 6 ppm. The vapor concentration from surface sources can be
sufficient to mask the detection of vapors coming from ground water
contamination when the vapor concentration due to ground water
contamination is of the order of less than 1 ppm - a problem reported
by Marrin (1984).

16.3.10 Delineating surface contamination

Results of sensitivity analyses using the model 20-DIFF have been
presented to assist in designing and interpreting soil-gas surveys for
contaminated soil (Silka, 1986) and leaking underground storage tanks
(Ferre and Silka, 1987). With regard to the use of soil-gas surveys for
identifying shallow contaminated soil (Silka, 1986), the optimum grid
spacing for the soil-gas survey was found to be primarily dependent
upon soil moisture and the value of Henry's Law constant, and, to
a lesser degree, organic matter content. Figure 16.7 illustrates the
406 Vapor analysis/extraction
1.0

0.9 organic
curve
fraction
0.8 A 0.03
B 0.01
0.7 C 0.0
~
~ 0.6
c:
~C 0.5
Q)
u 0.4
c:
0
u
0.3

0.2

0.1

0
0 6 12 18 24 30
distance (m)

Fig. 16.6 Distance of diffusion of 1,1,2,2-tetrachloroethane (TET) for three soil

organic matter contents in dry soil after 90 days from a constant concentration
source, from Silka (1986).

dependence of the effective diffusion coefficient on KH and fraction

of pore volume occupied by water (win). The reduction in DE is
represented by the ratio of DdDc, where Dc is the diffusion coefficient
corrected by the Millington-Quirk tortuosity formula in Eq. (16.9).
Therefore, in dry soil DE IDe is unity.
Optimum conditions for soil-gas surveys are obtained when dry soil
conditions have prevailed prior to the survey. Since moist soil is the
rule, though, especially in the more humid regions, the optimum grid
spacing will generally be less than 30 m. For VOCs with even moderate
values of KH, for example trichloroethylene (TCE) with KH of 0.4 and
1, 1,2,2-tetrachloroethane (TET) with KH of 0.02, the reduction in the
effective diffusion coefficient is sufficient to reduce their distance of
migration. Compared to an ideal VOC that does not partition into soil
water, TCE migrates to only about 60% of the distance, while TET
migrates to only about 20% of the distance of the unretarded ideal
VOC within the same time frame. In homogeneous soils, the maximum
extent of the migration of vapor through soil gas will be limited by
the thickness of the unsaturated zone. However, many soils are
heterogeneous and stratified. Stratified soils can cause greater lateral
migration of the vapor if the upward diffusion is blocked by strata
 Soil-gas surveys 407

1.0

0.9

0.8

0.7

r3 0.6
"
Q
IJ.J
0.5

0.4

0.3

0.2

0.1

00 0.2 0.4 0.6 1.0

% pore volume occupied by water

Fig. 16.7 Dependence of the effective diffusion coefficient, expressed as the

ratio of DEI Dc, on moisture content, expressed as the percentage of total
porosity occupied by water, and Henry's Law constant, KH , from Silka (1988).

containing finer grained soil having a higher moisture content or

higher organic carbon content.

16.3.11 Mapping groundwater plumes

Several observations reported in the literature concerning the interpreta-
tion of soil-gas surveys for groundwater plume mapping have been
investigated using the model 2D-Diff (Silka, 1988). It has been reported
that concentrations of VOCs in soil gas decrease from the source at the
water table to the surface by up to five orders of magnitude (Lapp ala
and Thompson, 1983a). This field observation conforms to that expected
for a diffusion dominated transport system. Since the soil-gas system is
bounded below by an essentially constant concentration source and
above by a constant zero-concentration boundary, i.e., the atmosphere,
the vapor concentration will decrease logarithmically from the water
table to the surface in an ideal, homogeneous soil.
Figures 16.8 and 16.9 show two cases for the distribution of vapor
concentration with distance above the centerline of a plume of con-
taminated ground water where the water table is at a depth of 10 m.
The concentration is presented in dimensionless units. Figure 16.8
408 Vapor analysis/extraction

DE = 0.025m 2 /d
w = 0.06
KH = 0.02
10

I
~ 8 50 100 400 days
.s
Q)
co
==
Q) 6
E
C\l
Q)
g4
~
'5
2

o+-----.----.-----.----~--~
2 3 4 5
log concentration (any units)

Fig. 16.8 voe concentration in soil gas versus time and distance above the
water table under nearly dry (w = 0.06) soil conditions, from Silka (1988).

shows the results for a relatively dry soil having only 20% water
content (a = 0.32, w = 0.08) and a VOC with a KH of 0.02. Figure 16.9
shows the results for a wetter soil with a 50% water content (a = 0.2,
w = 0.2). In both cases, there is greater than three to five orders of
magnitude change in the VOC concentration in soil gas from the water
table to the surface, even when the concentration profile approaches
steady state.
For the dryer soil (Fig. 16.8), the higher effective diffusion coefficient
results in the concentration profile approaching steady-state faster than
the wetter soil case (Fig. 16.9). Thus, the dryer soil conditions will
produce a more responsive concentration profile as compared to the
wetter soil. Shallow soil-gas measurements in the dryer setting will
better reflect the distribution of the VOC in the ground water at that
point in time.
Evans and Thompson (1986) concluded that aerobic biodegradation
of hydrocarbon vapors was the cause of lower than expected con-
centrations, <lOllg/l, in shallow soil gas at depths of less than 1.5 m
when compared to the > 1000 Ilg/1 concentrations in the 1.8-4.3 m
interval. However, they also reported at least one instance when the
concentration gradient reversed and decreased with depth below the
1.8-4.3 m interval.
In order to substantiate the interpretation that biodegradation was
 Soil-gas surveys 409

DE = 0.0023 m2 /d
w = 0.2
KH = 0.02
10

g
Q) 1200 days
::0 8
.El 400

~;;:
6 200
Q)
>
0
.0 100
co
Q)

c 4
(J

.El
'"
'6
2

0+---,---.----.----"1------'
2 3 4 5
log concentration (any units)

Fig. 16.9 VOC concentration in soil gas versus time and distance above the
water table under wet (w = 0.2) soil conditions, from Silka (1988).

occurring, active microbial populations and degradation by-products

should be confirmed in the soil column. It is more likely, though, that
the lower-than-expected concentration in the uppermost 1.5 m was
due to the normal decrease in concentration expected under the con-
centration gradient established by diffusion alone. Inspection of Figs.
16.8 and 16.9 shows that transient concentration profiles can produce a
concentration decrease of greater than three orders of magnitude from
the 1.8-4.3 m interval to the less than 1.8 m interval. Considering that
the concentration decrease due to the diffusion gradient would be
greater in wetter soils, the theory that biodegradation was the cause of
the observed decrease can not be confirmed and is not necessary.
Evans and Thompson (1986) also reported the observation that vapor
concentrations decrease rapidly, by two to three orders of magnitude
just beyond the edge of the ground water contamination zone. Simula-
tions were conducted using 2D-DIFF for an advancing front of a non-
aqueous phase liquid (NAPL) floating on the water table. The model
was set up with a constant saturated vapor concentration of 6 000 ppm
at the NAPL-soil gas interface and an initial concentration of 6 000 ppm
along the left side to represent the downward path of liquid VOC
migration. Two variations were run, one with a vapor diffusion rate
that was faster than the NAPL front velocity, and the second with a
vapor diffusion rate that was slower than the NAPL front velocity.
410 Vapor analysis/extraction
21 ,----.----,-.--,-,----------------,
DE = 0.55 m2 /d
w = 0.0
18

IQ)
15
:c
.l!!
~~ 12
(f)
>-
<1l
c: "0 0
0
0 0
(0
co (0
C')
...J
a.
«
z
Q) 9
>
0
.0
<1l
Q)
()
c:
~ 6
'5

o+-.--+----~.-----~~----~--~--;
60 180 360

o 6 12 18 24 30 36 42
distance of advance (m)

Fig. 16.10 Extent of migration of 0.01 ppm contour for VOC in soil gas from a
vertical spill source and advancing NAPL front floating on the water table
under dry soil conditions, from Silka (1988).

Figures 16.10 and 16.11 show the results in terms of the relative posi-
tions of the NAPL front and the 0.01 ppmv concentration contour. The
results presented in these figures can be applied to the case of only
dissolved VOC in a ground water plume by dividing 0.01 ppm by
6000 ppm, i.e., the contours in the figures would be equivalent to
1.6 x 10- 6 times the concentration of the VOC in the soil gas just
above the water table.
Figure 16.10 indicates that for a source front, i.e., NAPL or dissolved
contaminant plume, that moves slower than the diffusion rate, the
VOC does diffuse beyond the plume edge as observed by Evans and
Thompson (1986), up to a height of 12m above the water table. Above
 Soil-gas surveys 411

21 .---,,-.-rn-----------------------,
DE = 0.0044 m2 /d
w = 0.12
18

g 15
Q)
:0
.s
CD
'lij
:= 12
<::
0
...J
a.. soil-gas front (0.01 ppmv)
«
z
Q)
9
>
0
.0
as
Q)
(.)
<::
as 6 If)
Ci5 >-
as
'5 "0
0 0
C\J C\J
r--
3

o +---~---L._---,L---~----r_~~~~
240 480 600 720 NAPL
front

o 9 18 27 36 45 54
distance of advance (m)

Fig.16.11 Extent of migration of 0.01 ppm contour for VOC in soil-gas from a
vertical spill source and advancing NAPL front floating on the water table
under slightly moist soil conditions, from Silka (1988).

a height of 12 m from the water table, the vapor front falls behind the
liquid front. However, Fig. 16.11 indicates that when the diffusion rate
in soil gas is less than the velocity of the source front, the VOC
distribution in the soil gas will lag behind the source front or edge of
the plume at a much lower height above the water table. In the case
illustrated in Fig. 16.11, the VOC diffusion in the soil gas begins to lag
behind the front at a height of about 4.5 m above the water table. The
lag increases to about 18 m at a height of 9 m above the water table.
Therefore, a soil-gas survey may not always detect VOCs beyond the
edge of the plume, resulting in the underestimation of the extent of the
plume.
412 Vapor analysis/extraction

extent of 1.0 pg/ml

t
groundwater plume

extent of 1.0 ppb

t
TCE soil-gas plume

.....................................
.. ..

highway

extent of 1.0 pg/ml

PCE soil-gas plume
scale
0 __ _ _ 100
meters

Fig. 16.12 Comparison of sOil-gas survey results versus groundwater con-

centrations in total chlorinated species for southern California site, from Silka
and Spectre (1991).

In fact, the lag of the soil-gas plume behind an advancing ground-

water plume was observed in a study of a PCE contamination site
reported by Silka and Spectre (1991). In this case, the site is in the
semi-arid northern Los Angeles County of California. The geology
consists of stratified alluvial silt and sand with some clay with the
water table at a depth of approximately 10 m. Figure 16.12 shows
the comparison of the soil-gas plume measured at a depth of approxi-
mately 1.0 m with the groundwater plume, illustrating the phenomenon
of the lagging soil-gas plume behind the groundwater plume. Figure
16.12 also shows the soil-gas plume is wider than the groundwater
plume, which is due to the fact that the groundwater plume is not
growing sufficiently in the transverse direction to overcome the velocity
of the vapor migration.

Discriminating between ground water and surface sources

A problem that arises in the course of soil-gas surveys for both detec-
tion of surface sources, but more so, for mapping ground water con-
tamination, is the potential interference caused by VOC vapors from
another source. These interferences are especially problematic in highly
industrialized areas with multiple sources. The vapor diffusing laterally
outward from a surface source may be sufficient to prevent the detec-
 Soil-gas surveys 413

tion of vapors emanating from contaminated ground water when the

vapor concentration due to ground water contamination is much less
than that due to the diffusion from the surface source. This potential
problem was recognized by Marrin (1984) and further evaluated by
Silka (1986).
Silka (1986) applied the model 2D-DIFF to estimate the potential
interference from the upward diffusion of VOC vapors emanating from
contaminated ground water. For the modeling, it was assumed that the
highest VOC concentration in ground water underlying the area of the
soil-gas survey was 100 ppb. The depth to the water table was about
10 m. This level of ground water contamination could result in soil-gas
concentrations in the upper 1 m of soil as high as 150 ppb.
In comparison, a surface source of TCE, with a saturated vapor
concentration of 72 000 ppm, could cause a concentration in soil gas at a
distance of 30 m of as high as 72 ppm. Even with a relatively low
saturated vapor concentration, for example TET at 6000 ppm, the con-
centration in soil gas at a distance of 30 m could still be several parts
per million. Vapor concentrations of less than a part per million due
to diffusion from contaminated ground water would be completely
masked by such surface sources. However, more advanced GC finger-
printing of the VOCs would be valuable in deciphering complex site
con.tamination.

16.3.2 Conclusions for soil-gas surveys

VOC vapor migration through the unsaturated zone is controlled

primarily by diffusion. Where vapor is denser than soil gas by more
than 1.15 times, such as near a DNAPL front, and the vadose zone is
coarse grained and relatively dry, density-driven transport can be
important. Vapor diffusion is adequately described by Fick's second
law, and the effects of partitioning between soil-gas and soil moisture
can be adequately explained by Henry's Law. Adsorption of VOCs
onto soil organic matter can be accounted for by the empirical equi-
librium relationships between the octanol-water partition coefficient
and the liquid-solid partition coefficient. However, desorption of VOCs
from soil solids into soil water appears to be kinetically controlled and
can produce disequilibrium conditions.
Design of soil-gas surveys should be developed with an understand-
ing of the potential extent and distribution of contaminants in the
subsurface. Preliminary modeling of the diffusive transport using a
model such as 2D-DIFF can provide useful criteria for designing the
survey and interpretation of subsequent results. Modeling results
demonstrate the importance of soil moisture content for the design of
the soil-gas survey. Optimum conditions for soil-gas surveys occur
414 Vapor analysis/extraction
when lengthy, dry soil conditions have preceded the survey, which
usually occur during July, August, and September over much of the
United States.
Interpretation of soil-gas survey results are hampered by unknown
or poorly defined parameters, such as soil porosity, moisture content,
organic matter content, as well as source history and characteristics. In
general, the variation in soil moisture will have the greatest influence
on the rate of diffusion of VOC vapors through the unsaturated zone,
especially for those VOCs with small values of KH • Slight increases
in soil moisture dramatically reduce the effective diffusion rate and
increase the time required for concentrations in soil gas to approach
steady-state values. Dry soils in arid regions may allow quasi-steady-
state concentration profiles to be approached. However, steady-state
conditions probably are never approached in the humid, temperate
regions where frequent, episodic wet and dry periods occur.

16.4 VAPOR EXTRACTION SYSTEMS

16.4.1 Methods
Vapor extraction systems (VES) can be effective in remediating
unsaturated and saturated soils contaminated by volatile organic com-
pounds (VOCs). The operating principles of YES are straightforward.
Contaminated soil is flushed with fresh air via a vacuum extraction
well, drawing VOCs from the soil. As the contaminant is drawn off,
more VOCs go into the vapor phase as equilibrium is shifted towards
the gas phase, and are again drawn off by the vacuum. This action
serves to remove vapors from soil gas, promote volatilization and mass
reduction of any pure product phase present in the subsurface, force
dissolution of VOCs from soil water, and cause desorption of VOCs
from soil solids. In addition, the increased aeration of the soil may
increase biodegradation of the heavier compounds, which are not as
amenable to vapor extraction. Numerous case studies have shown that
YES is effective in remediating soils contaminated by VOCs such as
gasoline and solvents (Clister and Roberts, 1990). There are a number
of methods of vapor extraction that have been utilized. The simplest of
these is the installation of one or more vapor extraction wells in a
contaminated area with the application of a vacuum to extract VOCs.
There are a number of variations on this theme which may increase the
efficiency of the system.
Plunkett and Simmons (1990) suggest excavating contaminated soils,
placing them in a sealed, above-ground area, and remediating them
using a system of horizontal vapor extraction 'wells' placed in the
contaminated soil pile. Several authors (Pezzullo et al., 1990; Hutton,
 Vapor extraction systems 415

1990) describe the utilization of dual extraction wells for extraction of

contaminated groundwater and soil vapors from the same well. The
groundwater table is lowered due to pumping from the extraction well,
hence creating an induced vadose zone. This induced vadose zone is
then remediated by vacuum extraction.
Air injection has been used in some cases to increase airflow to the
subsurface and volatilization from the groundwater table. Marley
et al. (1990b) demonstrated the use of air injection wells (sparging) to
enhance volatilization in the saturated zone. A pulsed injection system
was used to prevent formation of short-circuiting pathways.
Walter et al. (1990) propose using horizontal 'vapor drains' installed
under a clay cap. The system was designed to operate at low flow rates
to minimize treatment costs for exhaust gases, and to operate over the
life of the clay cap (about 30 years).

16.4.2 Design parameters

The design of a YES system depends on a number of factors. The
physical properties of the contaminant, such as its Henry's Law con-
stant and vapor pressure, are important. In complex mixtures, such as
gasoline, only the more volatile constituents will be removed by vapor
extraction. The semi volatile fractions will remain resistant to removal
by vapor extraction.
A study of site conditions is important in evaluating the possible
effectiveness of YES. According to modeling performed by Johnson et
al. (1990a), the factors which most significantly affect the performance
of a YES are vapor flowrate, contaminant composition, and vapor
flowpath relative to location of the contaminant. They also present
modelling results for estimation of the time required to achieve steady-
state vapor flow, which depends on permeability, radial pressure dis-
tributions, vapor flow rate per unit length of well screen, radial linear
velocity distribution, effects of soil type on vapor flowrates, and water
upwelling caused by vacuum wells. A soil-gas survey is recommended
at the site to define the extent of contamination and the concentrations
present in the subsurface.
Site conditions which are most important include the aerial extent
of VOCs, depth to groundwater, soil type, and soil heterogeneity.
Important soil properties include soil structure, porosity, permeability,
and soil moisture content. The presence and location of soil hetero-
geneities will play an important part in the efficiency of any YES.
Dry, coarse soils, with low organic carbon content will be remediated
faster than moist, fine soils, with high TOe. As the percent moisture
increases, air permeability decreases. In addition, more water available
in the subsurface will cause greater amounts of VOCs to partition into
the water phase. Increased organic carbon content will caUse more
416 Vapor analysis/extraction
VOCs to sorb onto the soil matrix, and hence, lengthening the time
that vapor extraction is required.
Contaminant concentrations in the subsurface can be measured
directly in the field using a portable gas chromatograph, or inferred
from more precise laboratory methods performed on collected soil,
water, or vapor samples. Johnson et al. (1990b) gives the follow-
ing equation to estimate vapor concentrations for a muIticomponent
mixture (such as gasoline) in the subsurface based on compounds
known to be present
C = ~ xiPtMw,i
est L..- RT (16.11)

where: Cest = vapor concentration estimate (mg/L); Xi = mole fraction

of component i; Pt = pure component vapor pressure at temperature
T; Mw,i = molecular weight of component i (mg/mol); R = gas constant
(0.0821 atm/mol K); and T = absolute temperature (K). From CeSI1 one
can calculate the estimated removal rate of VOCs by
(16.12)
where Rest is the estimated removal rate per well and Q is the vapor
flow rate per well. The estimated removal rate can then be compared to
an acceptable removal rate, given by
M est
Racceptable = --
't
(16.13)

where M est is the estimated mass of the spill, and 't is the maximum
acceptable cleanup time (Johnson et al., 1990b). The number of wells
can then be estimated by dividing Racceptable by Rest.
In some cases it may be possible to infer air permeabilities from
measurements of saturated hydraulic conductivity, but this does not
take into account such factors as variable saturation and presence of a
pure product phase. Vacuum pump tests are needed to define air
permeability accurately. Waiter et al. (1990) measured air permeabilities
under field moisture conditions at several different scales both in the
laboratory and in the field. Their results varied over several orders of
magnitude, apparently as a function of the different scales at which the
measurements were made.
Air permeabilities can be measured in the field via pumping tests
where air is drawn from an extraction well, and pressure drawdowns
are measured at several distances from the pumping well. Pressure
drawdown data can be used to estimate air permeability using (Johnson
et al., 1990b)
P'=--
Q J
co -x
~dx (16.14)
4 nm -k - r2,;~ X
-
J.l 4kPat",t
 Vapor extraction systems 417
For (r 2 f;J.!/4kP atm t) < 0.1 Eq. 16.14 can be approximated by

P' = ~ [-0.5772 - In(;;J.! ) + In(t)] (16.15)

4n:m _ atm
J.!
where: P' = gauge pressure measured at distance r and time t; m
stratum thickness; r = radial distance from the pumping well; k = soil
permeability to air flow; J.! = viscosity of air (1.8 x 10- 4 g/cm sec);
f; = air-filled soil void fraction; t = time; Q = volumetric vapor flow
rate from pumping well; and Patm = ambient atmospheric pressure
(1.0atm = 1.013 x 106g/cmsec)
Using Eq. 16.15, it can be shown that a plot of P' versus In(t)
produces a straight line with a slope A and intercept B given by

A=~
k
4n:m -
J.!
2
B = -Q- [ -0.5772 - In ( -r f;J.!
-)] (16.16)
k 4kPatm
4n:m -
J.!
If both Q and m are known, then the equation for slope can be
rearranged and solved for k. If Q or m are unknown, then k must be
solved for using the equations for A and B (Eq. (16.16)).
2
k r f;J.! exp (B
=~ A + 0.5772 ) (16.17)
atm

Q may be difficult to estimate, due to short-circuiting effects. The total

flowrate from the extraction blowers may not be acting over the full
range of the formation, and flow from the ground surface may act to
short-circuit the system.
Marley et al. (1990a) developed analytical and numerical air flow
models to evaluate soil properties based on results of pilot YES opera-
tions for the design of full-scale YES operations. Their numerical solu-
tion allows simulation of soil heterogeneities and flow to partially
penetrating wells. In addition, the model can be used to determine
vapor extraction well spacings, screen intervals, air flow rates, and
system equipment requirements. A steady-state finite-difference air
flow model is calibrated against field results from an air pumping test
to solve for the air permeability.
Walter et al. (1990) developed analytical expressions for computing
drain spacings required to circulate a specified number of pore volumes
in a given time for a given pressure drop. Johnson et al. (1990b)
present a flowchart/decision-tree approach to decide if vapor extraction
418 Vapor analysis/extraction
is practical for a given scenario. The decision to use soil venting is
based upon vapor concentrations, vapor flow conditions to achieve
acceptable removal rates, feasible vapor flow rates, presence of any
residual material after venting, and possible negative side effects of soil
venting.

16.4.3 Radius of influence

The radius of influence of a soil venting well is important to consider in
the design of any YES system. Mutch et al. (1989) found much larger
radii of influence on the field than predicted by theoretical models, and
attribute this to soil heterogeneities, rather than simple anisotropy.
Kuo et al. (1990) define the radius of influence as 'the radial distance
where there is sufficient air flow to reduce the contaminant concentra-
tions below an acceptable level within a pre-specified time frame'. The
radius of influence depends on the characteristics of the vacuum equip-
ment, the physical and chemical characteristics of the contaminant, and
the characteristics of the soils. Kuo et al. (1990) developed a numerical
model to estimate radius of influence and study the effects of changing
various parameters associated with it. They found that the radius of
influence was greatest when the ratio krlkz equaled infinity, where kr is
the permeability in the radial direction, and kz is the permeability in the
vertical direction. The radius of influence is proportional to the ratio
krlkz. Trowbridge and Malot (1990) present some empirical observations
of radius of influence.

16.4.4 Exhaust treatment

There are several major types of treatment currently used to deal
with YES emissions: dispersion, activated carbon, catalytic oxidation,
condensation and liquid extraction, internal combustion and thermal
incineration. The simplest of these is dispersion, where the vapors are
diluted in a dispersion stack. However, this may be problematic in a
regulatory environment. Activated carbon is commonly used where
emissions of VOCs to the atmosphere need to be controlled. Activated
carbon requires monitoring for breakthrough of VOCs, and replace-
ment, which can increase costs. Trowbridge and Malot (1990) dis-
cuss the use of a catalytic oxidizer with a conversion efficiency of
99.8%, which was able to meet stringent local air quality regulations.
Rippberger (1990) discusses an innovative technique whereby all VOC
emissions are burned in an internal combustion engine fitted with a
catalytic converter. In turn, the engine is used to drive the vacuum
extraction equipment. This results in a significant cost savings over
carbon canisters.
 Vapor extraction systems 419
16.4.5 Fractionation of vapors
Various effects have been observed in the operation of YES where the
composition of the contaminant is altered by the removal of the lighter
fractions. Christenson et al. (1989) report that the vapor extraction
seemed to remove more of the chlorinated compounds than the
aromatic compounds. Pezzullo et al. (1990) found that their system
enhanced volatilization of DNAPLs present in fractured bedrock in the
subsurface. Johnson et al. (1990b) found a shift in composition to less
volatile compounds in the first 20 days of vapor extraction, but rela-
tively constant concentrations after that time. They also recommend
using a boiling point curve distribution, inferred from GC data, to help
identify the shift in composition of a complex mixture such as gasoline.
Modeling results from Johnson et al. (1990a) indicate that vapor extrac-
tion removes compounds in the order of their volatilities. Their model
also predicts the loss of soil moisture from vapor extraction, which may
be significant enough to affect soil microbial activity.

16.4.6 Advection vs. diffusion

Walter et al. (1990) used tracer tests to attempt to quantify YES system
performance. Field results indicated a retardation factor of 10-20,
while laboratory column experiments and theoretical calculations
indicated that the retardation factor should be closer to two or less.
This discrepancy was probably due to low permeability layers present
in the field, i.e., similar to the 'dual porosity' in the case of saturated
solute transport.
Heterogeneous soil may contain isolated or dead-end pores and have
variable saturation, and thus not be strongly affected by the YES. Since
advective movement of the VOC in these stagnant areas is negligible,
diffusion of the contaminant will dominate. Diffusion driven transport
is much slower than advective transport, and thus will impede the
efficiency of aVES.
In a soil where some regions allow transport of VOCs by advection
and other regions limit VOC transport to diffusion, the advection
dominated portions of the soil will be quickly flushed of VOCs, while
flushing the diffusion dominated portions will require much longer.
The VOC extraction rate from such a soil will be high in the early
stages of operation, during the advection dominated portion of flush-
ing. Once the advection dominated region has been flushed, the
extraction rate will exhibit considerable tailing, as VOCs slowly diffuse
out of the stagnant regions of the soil into the advection dominated
region. Therefore the extraction rate of a YES in the early stages of
operation should not be extrapolated to the entire flushing period,
since that early observed rate may only reflect the rate of contaminant
420 Vapor analysis/extraction

removal from the advection dominated region of the soiL Thus the
YES process should be quantified with special concern for the potential
importance of diffusion dominated regions to the operation of the YES.

16.4.7 Previous studies

Previous studies include a one-dimensional model developed by Baehr
and Hoag (1987) which describes advection and diffusion of gasoline
vapors in an unsaturated sand column. They found that the rate of
volatilization of organics was high relative to the rate of air flushing in
areas of fairly open porosity. However, they also mention that this may
not hold true in areas where vapor extraction is diffusion limited. Their
results also show that even in an advection dominated system, the
extraction rate will exhibit exponential decay, due to a shift in the
organic phase to less volatile components, as the more volatile com-
ponents are flushed out first.
Bowman (1987) cites empirical results from a YES installed at the
Twin Cities Army Ammunition plant in Minnesota, which displays an
exponential decline in extraction rate (Fig. 16.13). From the figure, it is
apparent that the extraction rate is quite high in the early stages of
operation, but drops off quickly and approaches a steady state value

120

110

100 TRICHLOROETHENE

90

:;- 80
E
0-
S 70
c
0
60
~
C
Q) 50
0
c
0 40
0
1,1,2,2-TETRACHLOROETHANE
30

20
DICHLOROETHENE
10

00 2 4 6 8
time (days)

Fig. 16.13 Monitoring results for a YES installed at the Twin Cities Army
Ammunition plant in Minnesota, from Bowman (1987).
 Vapor extraction systems 421
asymptotically. Even after adding another blower at 120 days, the rate
of extraction remained essentially constant at the asymptotically low
value throughout the remainder of operation. The flattening of the
curve is probably due to both fractionation of VOCs and diffusion
dominated transport.
Woodward and Clyde (Woodward-Clyde Consultants, 1985) also
present empirical data from a contaminated site in Tacoma, Washington
(Fig. 16.14). Their data indicate that the extraction rate reached an
equilibrium value within 90 days over a 260 day test. One possible
explanation of their data is that the rate of air flushing is lower than the
rate of volatilization from soil water and desorption from soil solids,
thus the vapor being extracted was continually being replenished with
VOCs. Therefore, the system remained in the advection dominated
stage, and did not reach the stage where the rate of VOC removal
would be limited by diffusion. The laboratory data for dichloroethylene
(DCE) shows that the rate of extraction drops off significantly after an
initial plateau, and remains fairly constant for three days. This may
represent a transition from the advection dominated to the diffusion
dominated regime.
Mutch et al. (1989) report that total VOC concentrations were in
the range of 100-140 ppmv at startup and declined asymptotically to

700

600

500
>;
ra
~
g400 I-
eD
"@
~ 300
0
E
l!?
200

100

0
0 40 80 120 160 200 240 280
time (days)

Fig. 16.14 Results of a vapor extraction test at the South Tacoma Channel-
Well 12A Superfund Site showing total VOCs over time, from Woodward-
Clyde Consultants (1985).
422 Vapor analysis/extraction
values of 60-70 ppmv. Crow et al. (1987) reported the most significant
declines in concentrations of hydrocarbon vapors in the first few days
of vapor extraction. Several case studies from Trowbridge and Malot
(1990) show a diffusion-dominated regime at late times. Hutton (1990)
compares analytical and pilot results. Pilot results showed an advection-
dominated regime during the first 8-12 weeks, with diffusion dominat-
ing after that point. Their analytical solution fits well during early
(advection-dominated), but not during later, diffusion-dominated time.
They attempted to model the diffusion-dominated regime with an
empirical boundary-layer solution, but with little success.
Numerical modeling results by Silka et al. (1989) of a vapor extraction
system qualitatively agreed with results from several empirical studies.
The proposed model consisted of two zones: an unsaturated, high
porosity region which is advection dominated, and a saturated, low
porosity zone which is diffusion dominated. The response of such a
system to a YES has been shown to consist of an initial period of rapid
vapor extraction, whereby the advective zone is flushed of soil gas. The
advective period is followed by a sharp exponential decrease in the rate
of extraction as the system becomes diffusion dominated.
Thus it is important to consider the diffusion stage when designing
and pilot testing a YES, since predictions of extraction time for a given
remedial action based on the early response of the YES often will be
erroneous.
One other factor should also be considered in designing a YES. In a
highly adsorptive soil, such as a soil with a high organic content, a
large amount of contaminant will be resident in the solid matrix of the
soil. This adsorbed contaminant mass will slowly desorb into the soil
water and then into the soil gas, and keep the soil-gas concentrations
from declining further. This will necessitate further soil-gas extraction.
The amount of contamination in the different phases of the soil should
be measured to gauge the magnitude of these effects. The combined
effects of low-permeability zones (diffusion dominated) and delayed
desorption can be assessed during YES operation by shutting the
system down and monitoring the soil-gas concentrations. A soil with
significant VOC mass contained in low-permeability zones and/or
adsorbed to soil solids will exhibit an increased concentration in the
YES discharge upon restart.

16.4.8 Boundary effects

The ground surface boundary condition is an important parameter to
consider with YES. Numerical simulations by Mendoza and McAlary
(1990) showed that loss of VOCs to the atmosphere could significantly
reduce contaminant concentrations, especially at shallower depths.
Modeling results of Kuo et al. (1990) show that an impermeable
 References 423

boundary over an area being vacuum extracted has little effect. Field
results from Diaz and Abidi (1990) show that air intake wells had little
effect on the performance of the YES, and were subsequently closed.
Sensitivity analyses performed by Mendoza and McAlary (1990b)
showed that the extent of subsurface contamination is sensitive to the
ground surface boundary condition. For a diffusion-only case with an
open ground surface which allowed mass to escape from the sub-
surface, the plume migrated more slowly than a plume beneath
a partially or fully covered ground surface. The boundary effect is
reversed when advection is included, with the plume under the open
ground surface travelling the farthest.
Staes et al. (1992) report on the use of a three-dimensional ground-
water flow model for assisting in the design of a vapor extraction
system. While unavailable for review as of this writing, they probably
converted the normal saturated water permeability to the gas-phase
permeability, using the intrinsic permeability and density and viscosity
of water versus air. The major assumption of the groundwater flow
model being violated with this approach is that the model is based on
the assumption that the fluid is not compressible. The potential impact
of the violation of this assumption is that the model would tend to over-
predict the mass transfer of gas. .

16.4.9 Other YES design considerations

In addition to the considerations of subsurface conditions and their
effects on the operation and performance of the YES, above ground
system design considerations are important. Diehl (1992) reviews the
minimization of emissions from soil vapor extraction (unavailable for
review as of this writing). Depending on jurisdictions, GAC or other
VOC removal may be required for the YES emissions. Innovative
solutions, such as combustion, catalytic converters, and biodegradation
have also been employed with varying degrees of success.
Fuerst and Underwood (1992) discuss vacuum extraction of volatile
and semivolatile compounds at a Superfund site and Lawn (1992)
discusses short-term vapor extraction system design (both unavailable
for review as of this writing).

REFERENCES

API (American Petroleum Institute) (1985) Detection of Hydrocarbons in Ground-

water by Analysis of Shallow Soil Gas/Vapor, API, Washington, DC.
Baehr, A.L. and Hoag, G.E. (1987) A modeling and experimental investigation
of induced venting of gasoline-contaminated soils, in Proc., Conf. on
Environ. and Public Health Effects of Soils Contaminated with Petroleum Products,
30-31 Oct 1987, Univ. of Massachusetts, Amherst, Massachusetts.
424 Vapor analysis/extraction
Bowman, R.S. (1987) Manipulation of the vadose zone to enhance toxic organic
chemical removal, Proc., 2nd Int. Workshop on Behavior of Pollutants in Porous
Media, 14-19 June. Bet Dagan, Israel.
Bruell, CJ. and Hoag, G.E. (1986) The diffusion of gasoline-range hydrocarbon
vapors in porous media, experimental methodologies, in Proc., NWWAI
API Conf. on Petroleum Hydrocarbons and Organic Chemicals in Ground Water -
Prevention, Detection and Restoration, Nat. Water Well Assoc., pp. 420-
43.
Callahan, M.A., Slimak, L.W., Gabel, N.W., May, 1.P., Fowler, CF. et al.
(1979) Water related environmental fate of 129 priority pollutants, vol. II. EP A-
440/4-79-029b, Environmental Protection Agency, Washington, DC
Christenson, D., Loo, W., Relf, M. and Karst, G. (1989) In situ treatment of soil
containing chlorinated solvents, ketones and aromatic organic compounds:
3 case histories, in Proc. of Hazmacon 90, Anaheim, California, Assoc. of
Bay Area Governments, Oakland, CA., pp. 199-209.
Ciister, W.E. and Roberts, B.D. (1990) Operating principles and case histories
of soil vapor extraction systems, in Proc. of the 11th Nat. Conf., Washington,
DC, Haz. Mat. Control Res. Inst., pp. 646-57.
Crow, W.L., Anderson, E.P. and Minugh, E.M. (1987) Subsurface venting of
vapors emanation from hydrocarbon product on ground water, Ground
Water Monitoring Rev., 7(1), 51-7.
Diaz, G.M. and Abidi, S.F. (1990) Vapor extraction - a case history, in Proc. of
Hazmacon 90. Anaheim, California, Assoc. of Bay Area Governments,
Oakland, CA., pp. 247-55.
Diehl, K.M. (1992) Minimizing emission from superfund air strippers and soil
vapor extractors, in Proc. of HMC-South '92 Conf., HMCRI, Greenbelt, MD.
Dilling, W.L. (1977) Interphase transfer processes, II: Evaporation rates of
chloro methanes, ethanes, ethylenes, propanes, and propylenes from
dilute aqueous solutions; comparisons with theoretical predictions, Envron.
Sci. Tech., 11(4), 405-9.
Evans, O.D. and Thompson, G.M. (1986) Field and interpretation techniques
for delineating subsurface petroleum hydrocarbon spills using soil gas
analysis, in Proc., NWWAIAPI Conf. on Petroleum Hydrocarbons and Organic
Chemicals in Ground Water - Prevention, Detection and Restoration, Nat. Water
Well Assoc., pp. 444-55.
Falta, R.W., Javandel, 1., Pruess, K. and Witherspoon, P.A. (1989) Density-
driven flow of gas in the unsaturated zone due to the evaporation of
volatile organic compounds, Water Resources Res., 25(10), 2159-69.
Ferre, P.T. and Silka, L.R. (1987) Application of the soil-gas survey to
underground storage tank monitoring and leak detection, in Proc., Eastern
Regional Ground Water Issues: A Conf., NWWA, Dublin, Ohio, pp. 187-206.
Freeze, R.A. and Cherry, J.A. (1979) Groundwater, Prentice-Hall, Inc.,
Englewood Cliffs, NJ.
Fuerst, D. and Underwood, B. (1992) Vacuum extraction of volatile and
semi volatile compounds at a superfund site, in Proc., HMC-South '92 Conf.,
HMCRI, Greenbelt, MD.
Fusillo, T.V., Keoughan, K.M. and Norwood, J.M. (1991) The use of soil gas
investigations for delineating chlorinated solvent contamination in soils
and ground water, Ground Water Management, Book 7, Nat. Water Well
Assoc., Dublin, Ohio, 163-75.
Glaccum, R., Michael, N., Evans, R. and McMillion, L. (1983) Correlation of
geophysical and organic vapor analyzer data over a conductive plume
containing volatile organics, in Proc., 3rd Nat. Symp. on Aquifer Restoration
 References 425

and Groundwater Monitoring, May 25-27, Columbus, Ohio, Nat. Water Well
Assoc., pp. 421-7.
Godoy, F.E. and Naleid, D.S. (1990) Optimizing the use of soil gas surveys,
Haz. Mat. Control, 3(5), 23-9.
Hillel, D. (1972) Soil and Water, Physical Principles and Processes, Academic Press,
New York.
Hutton, J.H. (1990) Remediation of ground water contaminated with volatile
organic chemicals by in situ aeration: a case study, in Proc., Hazmacon 90,
Anaheim, California, Assoc. of Bay Area Governments, Oakland, CA., pp.
223-34.
Johnson, P.e., Kemblowski, M.W. and Colthart, J.D. (1990a) Quantitative
analysis for the cleanup of hydrocarbon-contaminated soils by in situ soil
venting, Ground Water, 28(3), 413-29.
Johnson, P.e., Stanley, e.c., Kemblowski, M.W., Beyers, D.L. and Colthart,
J.D. (1990b) A practical approach to the design, operation, and monitoring
of in situ soil-venting systems, Ground Water Monitoring Rev., 10(2), 159-
78.
Jury, W.A. (1986) Chemical movement through soil, in Vadose Zone Modelling of
Organic Pollutants, (eds S.c. Hern and S.M. Melancon) Lewis Publishers,
Inc. Chelsea, Michigan, pp. 135-58.
Jury, W.A., Spencer, W.F. and Farmer, W.J. (1983) Behavior assessment model
for trace organics in soil: I. Model description, J. Environ. Qual., 12(4),
558-64.
Jury, W.A., Spencer, W.F. and Farmer, W.J. (1984) Behavior assessment model
for trace organics in soil: IV. Review of experimental evidence, J. Environ.
Qual., 13(4), 580-6.
Karickhoff, S.W. (1984) Organic pollutant sorption in aquatic systems, J.
Hydraulic Eng., 110(6), 707-35.
Karickhoff, S.W., Brown, D.S. and Scott, T.A. (1979) Sorption of hydrophobic
pollutants on natural sediments, Water Res., 13, 241-8.
Kerfoot, H.B. (1987a) Shallow-probe soil-gas sampling for indication of ground-
water contamination by chloroform, lnt. J. Environ. Anal. Chem., 30, 167-
81.
Kerfoot, H.B. (1987b) Field evaluation of three methods of soil-gas
measurement for delineation of groundwater contamination, in Proc., 3rd
Ann. Symp. on Solid Waste Testing and Quality Assurance, 1, Amer. Public
Works Assoc., pp. 1-67 to 1-80.
Kerfoot, H.B. (1988) Is soil-gas analysis an effective means of tracking
contaminant plumes in ground water? What are the limitations of the
technology currently employed? Ground Water Monitoring Rev., 8(2), 54-7.
Kerfoot, H.B. (1991) Subsurface partitioning of volatile organic compounds:
effects of temperature and pore-water content, Ground Water, 29(5), 678-
84.
Kerfoot, H.B. and Mayer, e.L. (1986) The use of industrial hygiene samplers
for soil-gas surveying, Ground Water Monitoring Rev., 1(4), 74-8.
Kuo, J.F., Aieta, E.M. and Yang, P.H. (1990) A two-dimensional model for
estimating radius of influence of a soil venting process, in Proc., Hazmacon
90, Anaheim, California. Assoc. of Bay Area Governments, Oakland, CA.,
pp.235-46.
Lappala, E.G. and Thompson, G.M. (1983a) Detection of ground water
contamination by shallow soil gas sampling in the vadose zone, in Proc. of
the Characterization and Monitoring of the Vadose (Unsaturated) Zone, Dec
8-10, 1983, Las Vegas, Nevada, Nat. Water Well Assoc., pp. 659-79.
426 Vapor analysis/extraction

Lappala, E.G. and Thompson, G.M. (1983b) Detection of ground water

contamination by shallow soil gas sampling in the vadose zone, in Proc.,
Nat. Conf. on Management of Uncontrolled Haz. Waste Sites, Nov 7-9, 1984,
Washington, DC, Haz. Mat. Control Res. Inst., Silver Spring, MD, pp.
20-8.
Lawn, A.M. (1992) Vapor extraction system design for short-term remediation,
in Proc., HMC-South '92 Conj., HMCRI, Greenbelt, MD.
Leistra, M. (1973) Computation models for the transport of pesticides in soil,
Residue Rev., 49, 87-130.
Lin, S. (1990) Soil gas analysis for the investigation of groundwater
contamination by volatile organic chemicals, unpublished Ph.D. thesis,
Univ. of California, Los Angeles.
Los Angeles County (1985) Remedial investigation of the San Fernando Valley
groundwater basin, request for proposal for Consulting Eng. Services, RFP
No. 739, Los Angeles County Dept. of Water and Power.
Malley, M.J., Bath, W.W. and Bongers, L.H. (1985) A case history: surface
static collection and analysis of chlorinated hydrocarbons from
contaminated ground water, in Proc., Conf. on Petroleum Hydrocarbons and
Organic Chemicals in Ground Water: Prevention, Detection, Restoration, Nov
13-15, 1985, Nat. Water Well Assoc.
Marley, M.C, Nangeroni, P.E., Cliff, B.L. and Polonsky, J.D. (1990a) Air flow
modeling for in situ evaluation of soil properties and engineered vapor
extraction systems design, in Proc., Hazmacon 90, Anaheim, California,
Assoc. of Bay Area Governments, Oakland, CA., pp. 208-22.
Marley, M.C, Walsh, M.T. and Nangeroni, P.E. (1990b) Case study on the
application of air sparging as a complementary technology to vapor
extraction at a gasoline spill site in Rhode Island, in Proc., 11th Nat. Conf.,
Washington, DC, Haz. Mat. Control Res. Inst., pp. 636-9.
Marks, B.J., Winsor, T.R. and Eames, M.A. (1989) Interpretation of soil gas
survey results for petroleum hydrocarbons, in Proc., Hazmacon 90,
Anaheim, California, Assoc. of Bay Area Governments, Oakland, CA., pp.
57-67.
Marks, B.J., Baltezore, M.L. and Radke, H. (1990) Quality assurance and
quality control in soil-gas surveys, in Proc., Hazmacon 90, Anaheim,
California, Assoc. of Bay Area Governments, Oakland, CA., pp. 31-41.
Marrin, D.L. (1984) Remote detection and preliminary hazard evaluation of
volatile organic contaminants in ground water, PhD dissertation, Univ. of
Arizona.
Marrin, D.L. (1988) SOil-gas sampling and misinterpretation, Ground Water
Monitoring Rev. J., VIII(2), 51-3.
Marrin, D.L. and Thompson, G.M. (1987) Gaseous behavior of TCE overlying a
contaminated aquifer, Groundwater, 25(1), 21-7.
Marshall, T.J. and Holmes, J.W. (1979) Soil Physics, Cambridge University
Press, New York.
Mehran, M., Nimmons, M.J. and Sirota, E.B. (1983) Delineation of
underground hydrocarbon leaks by organic vapor detection, in Proc., Nat.
Conj. on Management of Uncontrolled Haz. Waste Sites, Oct 31-Nov 2, 1983,
Washington, DC, Haz. Mat. Control Res. Inst., Silver Spring, MD, pp.
94-7.
Mendoza, CA. and Frind, E.O. (1990a) Advective-dispersive transport of
dense organic vapors in the unsaturated zone: 1, model development,
Water Resources Res., 26(3), 379-87.
Mendoza, CA. and Frind, E.O. (1990b) Advective-dispersive transport of
 References 427
dense organic vapors in the unsaturated zone: 2, sensitivity analysis,
Water Resources Res., 26(3), 388-98.
Mendoza, C.A. and McAlary, T.A. (1990) Modeling of ground-water
contamination caused by organic solvent vapors, Ground Water, 28(2),
199-206.
Millison, D., Marcotte, B. and Harris, J. (1990) Applications and Comparison of
Soil Gas, Flux Chamber and Ambient Air Sampling Results to Support
Risk Assessment at a Hazardous Waste Site, in Proc. 11th Nat. Conf.,
Washington, DC, Haz. Mat. Control Res. Inst., pp. 290-6.
Montgomery, J.H. and Welkom, L.M. (1990) Groundwater Chemicals Desk
Reference, Lewis Publishers, Inc., Chelsea, Michigan.
Mutch, R.D., Clarke, A.N., Clarke, J.H. and Wilson, D.J. (1989) In situ vapor
stripping: preliminary results of a field-scale US EPA/Industry Funded
Research Project, in Proc., 10th Nat. Conf., Washington, DC, Haz. Mat.
Control Res. Inst., pp. 562-9.
Nazaroff, W.W., Lewis, S.R., Doyle, S.M., Moed, B.A. and Nero, A.V. (1987)
Experiments on pollutant transport from soil into residential basements by
pressure-driven airflow, Environ. Sci. Tech., 21(5), 459-75.
OTA (1984) Protecting the Nation's Groundwater from Contamination, 1, OTA-O-
233, US Government Printing Office, Washington, DC.
Perry, H.P. and Chilton, C.H. (eds) (1973) Chemical Engineers Handbook, 5th
edn, McGraw-Hill Book Company, New York.
Pezzullo, J.A., Peterson, R.M. and Malot, J.J. (1990) Full-scale remediation at a
superfund site using in situ vacuum extraction and on-site regeneration:
case study - phase I, in Proc., 11th Nat. Conf., Washington, DC, Haz. Mat.
Control Res. lnst., pp. 624-7.
Plunkett, J.B. and Simmons, T.P. (1990) On-site vapor extraction -
demonstrated effectiveness, in Proc., 11th Nat. Conference, Washington,
DC, Haz. Mat. Control Res. Inst., pp. 641-5.
Rao, P.S.c., Hornsby, A.G., Kilcrease, D.P. and Nkedi-Kizza, P. (1985)
Sorption and transport of hydrophobic organic chemicals in aqueous and
mixed solvent systems: Model development and preliminary evaluation, J.
Environ. Qual., 14(3), 376-83.
Rippberger, M.L. (1990) Remediation of gasoline-contaminated groundwater:
spray aeration/internal combustion oxidation, in Proc., 11th Nat. Conf.,
Washington, DC, Haz. Mat. Control Res. Inst., pp. 865-7.
Rizvi, S.s.H. and Fleischacker, S.J. (1991) Determination of VOCs in subsurface
soils using gas sampling: a case study, in Proc. Haz. Mat. Control/Superfund
'91, HMCRI, pp. 479-84.
Roberts, P.V. and Valocchi, A.J. (1981) Principles of Organic Contaminant
Behavior during Artificial Recharge, in Quality of Groundwater, Proc.
Internat. Symp., Studies in Environ. Sci., Vol. 17 (eds van Duijvenbooden,
W., Glasbergen, P. and van Lelyveld, H.). Elsevier Scientific Pub. Co., the
Netherlands, pp. 439-50.
Siegrist, R.L. and Jenssen, P.D. (1990) Evaluation of sampling method effects
on volatile organic compound measurements in contaminated soil,
Environ. Sc. Tech., 24(9), 1387-400.
Silka, L.R. (1986) Simulation of the movement of volatile organic vapor through
the unsaturated zone as it pertains to soil-gas surveys, in Proc., NWWA/
API Conf. on Petroleum Hydrocarbons and Organic Chemicals in Grand
Water - Prevention, Detection and Restoration, Nat. Water Well Assoc.,
pp.204-24.
Silka, L.R. (1988) Simulation of vapor transport through the unsaturated zone -
428 Vapor analysis/extraction

interpretation of soil-gas surveys, Ground Water Monitoring Review J.,

VIII(2), 115-23.
Silka, L.R. et al. (1989) Modeling applications to vapor extraction systems,
presented at the Soil Vapor Extraction Technology Workshop, June 28 &
29, US Environmental Protection Agency Risk Reduction Engineering
Laboratory, Releases Control Branch, Edison, NJ.
Silka, L.R. and Spectre, Y. (1991) Soil-gas survey for mapping chlorinated
solvent plume in groundwater, Ground Water Management, (5), Nat.
Ground Water Assoc., 519-3l.
Sleep, B.E. and Sykes, J.F. (1989) Modeling the transport of volatile organics in
variable saturated media, Water Resources Res., 25(1), 81-92.
Smith, J.A., Chiou, CT., Kammer, J.A. and Kile, D.E. (1990) Effect of soil
moisture on the sorption of trichloroethene vapor to vadose-zone soil at
picatinny Arsenal, New Jersey, Environ. Sc. Techn., 24(5), 676-83.
Staes, E.G., Ratzlaff, S.A. and Stellar, J.R. (1992) Vapor extraction design
using a three-dimensional groundwater flow model, in Proc., HMC-South
'92 Conf., HMCRI, Greenbelt, MD.
Steinberg, S.M., Countess, R.J. and Lin, C.C (1990) Comparison of the
response of VOC analyzers to total hydrocarbon measurements by GC-
FlD-PID, in Proc., Hazmacon 90, Anaheim, California, Assoc. of Bay Area
Governments, Oakland, CA., pp. 42-5l.
Striegl, R.G. (1987) Transport of methane in the unsaturated zone, in Proc.,
Assoc. of Ground Water Scientists and Engineers Program, Focus: Contaminant
Hydrogeology, NWWA Ann. Meeting and Expo, September 14-16.
Stuart, J.D., Lacy, M.J. and Roe, V.D. (1989) Ground water and soil-gas
monitoring at sites of subsurface gasoline contamination in the Northwest,
in Preprinted Extended Abstracts, Div. of Environ. Chern., Am. Chern. Soc.,
pp.489-90.
Swallow, J.A. and Gschwend, P.M. (1983) Volatilization of organic compounds
from unconfined aquifers, in Proc. 3rd Nat. Symp. on Aquifer Restoration and
Groundwater Monitoring, Nat. Water Well Assoc., pp. 327-333.
Thibodeaux, L.J. (1979) Chemodynamics, Environmental Movement of Chemicals in
Air, Water, and Soil, John Wiley & Sons. New York.
Thomsen, K.O. and Joyner, S. (1990) Source Identification and
Characterization Using Areal and Vertical Soil Gas Sampling Techniques,
in Proc. 11th Nat. Conf., Washington, DC, Haz. Mat. Control Res. Inst., pp.
277-84.
Thomson, G.M. and Marrin, D.L. (1987) Soil gas contaminant investigations: a
dynamic approach, Ground Water Monitoring Review, 7(3), 88-93.
Trowbridge, B.E. and Malot, J.J. (1990) Soil remediation and free product
removal using in situ vacuum extraction with catalaytic oxidation, in Proc.
Hazmacon 90, Anaheim, California, Assoc. of Bay Area Governments,
Oakland, CA., pp. 235-46.
Turk, L.J. (1975) Diurnal fluctuations of water tables induced by atmospheric
pressure changes, J. of Hydrology, 26, 1-16.
Verschueren, K. (1983) Handbook of Environmental Data on Organic Chemicals, 2nd
edn, Van Nostrand Reinhold, New York.
Viellenave, J.H. and Hickey, J.C (1990) Use of High Resolution Passive Soil
Gas Analyses to Characterize Sites Contaminated with Unknowns,
Complex Mixtures, and Semi-Volatile Organic Compounds, in Proc. 11th
Nat. Conf., Washington, DC, Haz. Mat. Control Res. Inst., pp. 277-84.
Voorhees, K.J., Hickey, J.C and Klusman, R.W. (1984) Analysis of
 References 429

groundwater contamination by a new static surface trapping/mass

spectrometry technique, Anal. Chem. 56, 2604.
Walter, C.R., Yiannakakis, A., Bentley, H.W., Hauptmann, M. and
Valkengurg, N. (1990) Cas in a hat, in Proc. 11th Nat. ConI, Washington,
DC, Haz. Mat. Control Res. Inst., pp. 557-64.
Woodward-Clyde Consultants (1985) Performance evaluation, pilot scale
installation and operation, soil-gas vapor extraction system, Time Oil Company
site, USEPA Work Assignment No. 74-0N14.1.
Zdeb, T.F. (1987) Multi-depth soil-gas analyses using passive and dynamic
sampling techniques, in Proc., Petroleum Hydrocarbons and Organic Chemicals
in Ground Water: Prevention, Detection, and Restoration, Nat. Water Well
Assoc., Dublin, OH.
 CHAPTER 17

Vertical cutoff walls

Jeffrey C. Evans

17.1 INTRODUCTION

Vertical cutoff walls are installed in the subsurface to control horizontal

movement of groundwater and contaminants. This chapter presents
the objectives and applications for vertical cutoff walls, describes the
types of cutoffs available and presents detailed information to choose,
design and construct the appropriate cutoff wall.
Use of vertical cutoff walls in the subsurface initially began with
groundwater control and structural applications in Europe before 1950
(Xanthakos, 1979). In these applications, vertical cutoff walls were
installed into the subsurface to act as a barrier to horizontal ground-
water flow during excavation. In addition, they were often designed
with reinforced concrete to provide structural stability for the excavation
within the cutoff wall perimeter. Slurry trench cutoff walls have been
used in environmental applications since the 1970s and have come
into widespread usage since the 1980s as a component in the overall
remedial system to control flow of groundwater (Spooner et al., 1984).

17.2 OBJECTIVES

Depending upon the objective of the vertical cutoff wall, different

criteria are considered in the design. For example, if the objective is to
minimize the rate of contaminant transport off-site, it is necessary to
consider contaminant transport through the wall, potential degradation
of the wall, and consequences of inadequate cutoff wall performance.
Contrast this objective to the use of a cutoff wall to minimize the rate
of on-site migration of uncontaminated groundwater from the up-

Geotechnical Practice for Waste Disposal.

Edited by David E. Daniel.
Published in 1993 by Chapman & Hall, London. ISBN 0412351706
 Applications 431

gradient area. In this case, there is a reduced concern about deficiencies

in the cutoff wall and potential degradation due to the flux of uncon-
taminated groundwater through the wall.
In the majority of the situations employing vertical cutoff walls,
environmental control systems include a combination of the techniques
described in this book. If the objective, for example, is to minimize
the rate of contaminant transport off-site, the environmental control
system may include a vertical cutoff wall coupled with a low perme-
ability cover, groundwater withdrawal system, and treatment systems
for the pumpage.
In defining the vertical cutoff wall objectives, it is important to
determine whether the barrier is to act as a ground water barrier with
low hydraulic conductivity or a contaminant transport barrier. For
example, the flux of groundwater through a wall with known hydraulic
conductivity can easily be calculated. However, if the wall is to act as a
contaminant barrier, it is necessary to consider modes of contaminant
transport in addition to advective transport. As described in Chapter 3,
contaminant transport processes include advection and dispersion
(dispersion includes both mechanical dispersion and molecular dif-
fusion). As a result of establishing different objectives, the type of
vertical cutoff wall and the design criteria for the cutoff wall depend
very much upon the defined objectives of that cutoff wall.

17.3 APPLICATIONS

Vertical cutoff walls are traditionally considered for environmental

applications as circumferential barriers; that is, they typically surround
the site (Fig. 17.1). In this system, an inboard subsurface drainage
system and cover over the contaminated area are typically used in
conjunction with the circumferential vertical cutoff wall to protect
offsite groundwater.
Vertical cutoff walls need not be circumferential. They may be used
to effect a vertical barrier for only a portion of the site. For instance, at
the Rocky Mountain Arsenal, the cutoff wall was used down-gradient
of the contamination to improve the effectiveness of a groundwater
withdrawal and treatment system and to permit recharge of treated
water further down-gradient (Campbell and Quintrell, 1985).
Vertical cutoff walls are also employed in up-gradient locations. For
instance, Evans and Hill (1984) describe a vertical cutoff wall that
serves to reduce the rate at which clean, uncontaminated groundwater
flows onto the site from up-gradient locations.
Vertical cutoff walls frequently key into a stratum of naturally low
hydraulic conductivity (Fig. 17.1). A key is not always necessary or cost
effective when contaminated groundwater is being extracted or when
432 Vertical cutoff walls

(a)
cutoff wall

(b)
cutoff wall

Fig. 17.1 Schematic diagram of vertical cutoff wall configuration for typical
site remediation project: (a) plan view and (b) cross-section.

the contaminants are concentrated near the ground surface or floating

on the water table.

17.4 TYPES OF VERTICAL BARRIERS

17.4.1 Vertical cutoff walls of low hydraulic conductivity

Traditionally low permeability vertical cutoff walls are constructed by
installing a vertical barrier into the subsurface. The vertical barrier has
a lower hydraulic conductivity than the associated formation. These
vertical cutoff walls may be divided into several groups based on
construction methods.
Several types of vertical cutoff walls may be constructed using the
slurry trench method of excavation, and hence are called 'slurry trench
cutoff walls'. In the slurry trench method of construction, a vertical
trench is excavated into the subsurface utilizing a slurry, typically of
bentonite and water, to maintain the trench stability. The completed
excavation is then used to form the geometry of the cutoff wall. The
 Feasibility studies 433

completed cutoff wall may consist of soil-bentonite (Millet and Perez,

1981), cement-bentonite (Jefferis, 1981), plastic concrete (Evans et al.,
1987), or structurally reinforced concrete (Boyes, 1975).
In addition to the slurry trench methods of construction for vertical
cutoff walls, there are grouting methods where materials of low
hydraulic conductivity are injected into the subsurface. The method of
injection varies and helps define two sub-classes of vertical cutoff
walls. Grout curtains are vertical cutoff walls constructed using one of
the traditional grouting techniques. Vibrating beam cutoff walls are
constructed by inserting a steel beam into the subsurface using a
vibratory pile driver. The steel beam is modified with grout injection
nozzles and the void space is grouted as the beam is withdrawn
(Leonards et al., 1985).
In addition to grouting and slurry trench techniques, cutoffs can be
constructed using other materials or other composite techniques. For
example, steel sheet pile may be driven to provide a cutoff in the
subsurface. Also, cutoff walls may be constructed with augering tech-
niques, where soils are augered, materials injected through the augers,
and the blend is mixed in situ.

17.4.2 Cutoff walls of high permeability

In addition to vertical cutoff walls which impede groundwater flow by
virtue of their low hydraulic conductivity, a new technique has been
developed to install a vertical cutoff by pumping from within a vertical
drain of high hydraulic conductivity. This technique employs a bio-
degradable drilling mud to maintain the trench stability. The completed
trench is backfilled with a material having a high hydraulic con-
ductivity. Pumping from this highly permeable vertical zone can create
a hydraulic barrier via control of flow patterns.

17.5 FEASIBILITY STUDIES

The first step in the design of a vertical trench cutoff wall is to identify
the design objective(s). For example, the objective may be to provide a
specific hydraulic conductivity to reduce the rate of groundwater flow.
Upon establishment of the vertical cutoff wall objectives, it is then
necessary to determine which types of vertical cutoff wall can meet
these objectives. This can often be accomplished through the use of
groundwater and contaminant transport models. All too often the
cutoff wall is selected as a solution without first establishing the asso-
ciated engineering of cutoff wall objective and analysis.
The next step is to evaluate the hydrogeologic and geotechnical
aspects of the site. A vertical cutoff wall may be the best solution in
434 Vertical cutoff walls

terms of satisfying the project objectives; yet, if the cutoff wall is not
constructible due to boulders in the subsurface or fractured rock, or
unacceptable health or safety hazards posed by contaminated materials,
then it is necessary to look at alternative solutions.
During the feasibility studies it is important to examine both short-
and long-term performance of the cutoff wall. What will its properties
be in terms of hydraulic conductivity, strength and compressibility?
Are these properties satisfactory to meet the objectives of the project?
In the long term, will the wall be durable?
During the design feasibility studies it is important to consider alter-
natives. Although soil-bentonite or cement-bentonite may appear to be
the cutoff wall of choice, it is necessary to consider other alternative
techniques to construct vertical barriers in the subsurface as described
in this chapter. In fact, one alternative may be to not use a vertical
cutoff wall, but to accomplish the objectives using other techniques,
such as increased pumping rates.
Finally, it is necessary to consider cost. If construction conditions are
so difficult and the potential for degradation is so real that the cost of
the successful cutoff wall is not competitive with other techniques,
then one needs to consider other techniques. Generally speaking, soil-
bentonite slurry trench cutoff walls are among the least expensive
techniques available in the US for vertical barriers in the subsurface. In
Europe, cement-bentonite cutoff walls are more prevalent.

17.6 SOIL-BENTONITE SLURRY TRENCH CUTOFF WALLS

17.6.1 General information and technology

Soil-bentonite slurry trench cutoff walls are constructed utilizing the
slurry trench method of excavation. In this method, a trench is exca-
vated into the subsurface and a bentonite-water slurry is used to
maintain trench stability, as shown in the typical section of Fig. 17.2.
The excavation can be completed to depths of IS-20m using a con-
ventional backhoe with an extended stick. For greater depths a drag-
line or clamshell is usually employed. One contractor can excavate to
27 m with a specially modified backhoe (Case International, 1982). The
slurry is typically a bentonite-water mixture consisting of 5% bentonite
and 95% water by weight. Bentonite, a montmorillonitic clay, swells in
the presence of water, imparting a viscous nature to the fluid and
aiding in the formation of a filter cake along the walls of the trench. For
trench stability, the slurry level is maintained at or near the top of the
trench, typically within 1 m. Trench widths vary from 0.6 to 1.5 m, with
0.9 m trench widths being typical. Trench stability results from the
 Feasibility studies 435

Backfill
Mixing Area Trench Spoils

Area of Active

Fig. 17.2 Excavation of trench and placement of soil-bentonite backfill.

fluid pressures of the slurry exceeding the active earth pressures and
the analysis of trench stability is well documented (Nash, 1974).
Once the trench is excavated to the desired depth, the bentonite-
water slurry is replaced with a soil-bentonite backfill having a low
hydraulic conductivity. The backfilled trench forms the completed
vertical cutoff wall. The backfill optimally consists of a mixture of sand,
silt, and clay, and bentonite-water slurry. The soil-bentonite backfill is
placed in the trench at a consistency of high slump (100-150 mm)
concrete. In order to achieve this consistency, the material needs to be
'fluidized' by adding bentonite-water slurry to soil.

17.6.2 Backfill placement

It is conventionally considered that the backfill sloughs forward and
the emplaced backfill results in a homogeneous blend (0' Appolonia,
1980). Research incorporating analytical methods, laboratory model
studies and field observations indicate that backfill/sloughing forward'
may not be the appropriate mechanism by which backfill displaces the
slurry (Evans et al., 1985b). Static equilibrium analysis using rotational
equilibrium methods show that the critical failure surface is at a shallow
depth. Backfill placed on top of the previously placed backfill causes
shallow slope instability near the top of the trench. The mass in move-
ment then slides down the top of the previously placed backfill as a
436 Vertical cutoff walls

Backfill
Slurry Level
Placed Here

\ Bentonite_Wa~ Area of Rotational

Shear Failure
Slurry
'Backfill Movement Downslope
~~-....I

6 to 12
Typical

Fig. 17.3 Movement of soil-bentonite backfill into trench.

viscous fluid as shown on Fig. 17.3. Since backfill slides over previously
placed backfill, entrapment of thin layers of soil-bentonite slurry be-
tween the layers is possible. It is believed that there is little or no
mixing of one backfill batch with the next. Instead, backfill slides
downslope on top of previously placed backfill.

17.6.3 Configuration of wall

During excavation of the trench for soil-bentonite slurry trench cutoff
walls, the bentonite-water slurry within the trench will be level
(horizontal). Thus, when excavating downslope or upslope, the depth
of the slurry in the trench will vary with the site grade. If the surface of
the slurry is not maintained at or near the ground surface, the risk of
trench instability increases. As a result, there is a maximum grade for
cutoff excavation. This limit is normally about 1% although it more
precisely depends upon the results of stability calculations and the
depth of excavation. A deeper excavation results in a longer open
trench between the backfill and the excavation face. Final design studies
must consider the site work necessary to adjust the grade of the cutoff
wall including possible horizontal relocation of the vertical barrier
cutoff wall.
The thickness of the vertical barrier wall depends upon the objectives,
although walls with a width of 750-900 mm are typical. A wall that is
too thin is difficult to excavate and backfill; a wall that is too thick may
increase the costs unnecessarily.
The working surface next to the wall is about 15 m wide to provide
for the backhoe and associated construction traffic. Wider surfaces
allow for easier construction access. The working surface must provide
a stable base to support the excavating equipment as well as any
associated truck traffic. The working surface is normally constructed of
on-site available materials because of the temporary nature of the use.
 Feasibility studies 437

17.6.4 Excavation
Fluid levels are normally maintained within 1 m of the ground surface,
although stability conditions may dictate stricter requirements. Fluid
properties also are related to the stability needs, but fresh slurry has
properties as shown on Table 17.l.
It is noted that excavations in cohesionless soils (sands and gravels)
may experience some 'unraveling' of the slope at the top of the trench.
In contrast, excavations in clay may become unstable due to pre-existing
discontinuities which weaken the clay. In particular, vertical desicca-
tion cracks may result in large chunks of clay falling into the excavated
trench. The presence of tension cracks must be carefully controlled by
limiting traffic along the sides of the trench and minimizing the time
duration of the open trench.
The hydrated slurry filtrate loss of 25 cm3 at a pressure of 689 kPa
for 30 minutes indicate a high quality bentonite. A stabilized Marsh
viscosity of 32-40 seconds provides a fluid material that is pump able
and thin enough to be displaced by the backfill, yet is adequately
viscous to provide trench stability. The density of the slurry should be
sufficient to maintain the trench stability, yet not too high to preclude
backfill displacement of the slurry during placement.

17.6.5 Backfill
Final studies of the backfill mix design include bentonite content,
natural fines content, grain size distribution of the base soil, slump,
backfill hydraulic conductivity and permanence of the soil-bentonite
backfill. Backfill can be prepared using soils excavated from the trench
or using materials from a borrow source. The base soil is mixed with
additional bentonite-water slurry into a homogeneous blend and then
placed in the trench. As a result of the natural variability that may
occur along the length of the trench, the grain size distribution of base
materials from a borrow source may be easier to control and result in a
more uniform soil-bentonite backfill. It is important to include natural
fines in the backfill to reduce the hydraulic conductivity, although a
strong relationship may not exist between fines content and hydraulic
conductivity (Ryan, 1987).
As discussed earlier, the backfill is prepared at a consistency of high
slump concrete, and the proper slump must be determined during
the mix design studies. Laboratory studies are conducted to evaluate
hydraulic conductivity and compatibility prior to construction, and it is
important that these laboratory samples be prepared at the same slump
(and water content) as required in the field. If slump is not controlled
in the laboratory studies, the hydraulic conductivity and compatibility
test data may not be representative of field conditions.
Table 17.1 Recommended properties of bentonite-water slurry

parameter bentonite slurry cement-bentonite slurry

as-mixed during excavation as-mixed during excavation

density (g/cm3) 1.01-1.04 «P backfill -0.25) 1.10-1.24 1.10-1.4

viscosity, apparent (Seconds Marsh) 32-40 38-68 40-45 38-80
filtrate loss, ml <25 100-300
pH 7.5-12 10.5-12 12-13
bentonite content, % by weight 4-7 -6 4-7
other ingredients, % by weight sand, <1 sand, <5 cement, 15-30
 Feasibility studies 439

Laboratory studies of the backfill during final design studies always

include measurement of hydraulic conductivity, and for environmental
applications, compatibility. Both organics and inorganics have been
found to cause increases in bentonite permeability (Alther et al., 1985).
Chemical compatibility is typically evaluated by performing long-term
permeability tests using the contaminants expected (Evans and Fang,
1988).

17.6.6 Technical specifications

In order to prepare the technical specifications for a soil-bentonite
slurry trench cutoff wall, it is first necessary to establish the philosophy
of specification. Some owners take the approach that the specification
should provide guarantees for items over which the contractor may
have little control. For example, the long-term performance of the
cutoff wall may depend upon site and subsurface conditions that the
contractor cannot control. The technical specifications should establish
the reliability and responsibility of the contractor for those things that
can be controlled. In this way, the vertical cutoff will be of high quality
at a competitive cost.
The components of the technical specification for a vertical cutoff
include contractor qualifications, site personnel qualifications, con-
struction requirements, including materials, equipment and methods,
quality control requirements, drawings and the basis of measurement
and payment. The contractor qualifications section of the technical
specifications often requires the bidder to establish experience in con-
structing vertical barriers for environmental applications.
The construction requirements established in the technical specifica-
tion should clearly spell out which materials are and are not permitted,
which equipment is and is not permitted, and what methods will
and will not be permitted. The construction requirements should
be as general as possible to permit contractor innovation, yet be as
specific as possible to ensure that the selected construction method
will not diminish the quality of the completed cutoff wall. The quality
assurance/quality control (QA/QC) requirements of the technical
specification must differentiate between that being provided by the
contractor and that being provided by the owner through the design
engineer. In general, the best philosophy is one which employs the
design engineer's personnel to provide as much quality control as
possible without alleviating the responsibility of the contractor for
building a quality project.
The technical specifications must contain the basis for measurement
and payment. For example, if there is excess slurry loss due to voids in
the subsurface, is this the responsibility of the contractor to provide
the extra bentonite-water slurry, or the owner? A well constructed
440 Vertical cutoff walls

technical specification will delineate slurry loss, which is expected,

i.e., based upon the porosity of the subsurface materials, versus
unexpected based upon information not provided to the contractor,
i.e., the open pipe.

17.6.7 Construction considerations

Grades must be nearly level or discontinuities in grade (steps) must be
considered for the vertical barrier construction. Corners must be care-
fully considered and space allocated for either criss-crossing or for a
gentle radius to avoid backfill 'hang-up' at the corners and increased
potential for backfill entrapment. The depth of the wall, often con-
trolled by hydrogeologic considerations, must be reconsidered with
respect to constructability. For example, shallow excavation during site
preparation to permit a slightly lower grade may allow the entire
project to be excavated utilizing only backhoe, without the need for
specialized equipment required for greater depths. Platform width and
traffic patterns impact the cost of construction. The easier it is for the
contractor to build the job, the lower the price and the higher the
quality.
As part of the final design studies, the continuity of the wall,
the minimum depth of key, the slurry preparation and delivery to
the trench, the disposition of unsuitable excavated materials and the
stability of the trench must be evaluated. Cutoff wall continuity is
typically assured by the method of excavation. The minimum key can
be documented using soundings and sampling of excavated materials.
Slurry mixing is accomplished using mixing ponds with overnight
hydration or high speed colloidal shear mixers. Where space permits,
the option should be provided to the contractor.
The homogeneity, slump, and the engineering properties of the
backfill must be carefully considered so as to minimize slurry entrap-
ment. Backfill mixing is accomplished along the side of the trench, or
in a remote mixing area. The advantage of mixing along the side of the
trench is cost, as no trucking of backfill is typically required for this
alternative. The disadvantages include the need for a larger mixing
area and the difficulty in adding natural fines (or sand and gravels) to
augment the materials excavated from the trench for the base soil. The
advantages of a remote mixing area include increased homogeneity,
uniformity and control, reduced working pad widths needed along the
trench, and finally the ease in which additional base soils can be
brought into the area for mixing.
Slurry entrapment can be minimized by the contractor's method of
construction including the need to ensure that no backfill is dropped
through the slurry in the trench and that the soil-bentonite backfill is at
the proper slump. Further, careful control of the backfill movement
 Feasibility studies 441

around corners must be made to ensure no hanging up at the corners.

The objective of the quality control program is to monitor the con-
struction so that the vertical cutoff wall is built in accordance with the
plans and specifications. It is equally important to verify that the site
and subsurface conditions encountered during construction are con-
sistent with those assumed during the design. It is common practice to
include quality control duties with the contractor's responsibilities. In
so doing, it is possible to address the first issue, which is to ensure that
the wall is properly constructed. However, without a representative of
the design engineer on site during construction, the second objective
cannot be met, that is, assuring that the site and subsurface conditions
are consistent with those assumed in the design. It is recommended
that the quality control responsibilities be carried out by the design
engineer responsible to the owner.
How do the responsibilities of the contractor and engineer differ
on a slurry trench cutoff wall project? First, the contractor should be
required to provide all of the supporting documentation for the quality
of the materials. For example, the data sheets from the bentonite, grain
size distribution analyses of any proposed borrow materials, specifi-
cations of the mixer, and the like. Second, the contractor should be
responsible for providing all of the technical support necessary for the
engineer to complete his quality control testing. For example, when the
engineer needs to sample the mixed backfill, the bulldozer operator
needs to momentarily halt the mixing process to provide the oppor-
tunity for the engineer to obtain his sample. Likewise, excavation must
cease and the excavator must cooperate with the engineer in order to
take soundings at the top of the aquatard and at the bottom of the key.
The contractor must cooperate in other areas as well; in providing
manpower assistance to take readings and to provide testing equip-
ment where requested. It is the engineer's responsibility to conduct on-
site tests for the control of the project. These tests must include
measurements to document the excavation, the backfill and the
materials used, including raw products and mixed products as dis-
cussed below. Thus, working as a team, not in an antagonistic relation-
ship, the contractor and the engineer can work together to provide the
owner with the best possible project at the lowest possible cost.
In order to provide control of the quality of the excavation, there
are several tests and measurements for the engineer to make. First,
soundings need to be obtained at the base of the trench excavation to
determine final grade. Soundings are typically made at 3 m centers
along the trench alignment. Further, for excavation into a key material,
the engineer must be present to inspect visually the excavated material
to assure that the excavation penetrates adequately into the key
material. Testing includes soundings and sampling of the key mate-
rial. Additional monitoring during excavation includes a one-time
442 Vertical cutoff walls

measurement of the backhoe bucket width to ensure that the trench

width will meet the requirements and a measurement of the level of
the working pad and excavation equipment to assure the verticality of
the trench. Finally, testing of the bentonite-water slurry as-mixed and
in the trench is required as discussed below.
Quality control testing of the backfill includes the measurement of
slump to assure that the material is fluid enough to flow into the slurry
trench cutoff wall, yet not so wet that it is too compressible for satis-
factory long-term performance. The slope of the backfill below the
slurry surface, optimally between 6H: 1V and 12H: 1V, is measured
several times a day. In addition to samples for slump testing, samples
are taken periodically for grain size distribution. Finally, samples are
obtained and taken to document the laboratory hydraulic conductivity
of field-mixed samples. It is noted that perfect homogeneity cannot be
achieved under the field mixing conditions. For example, cohesive soils
that are excavated from the underlying aquatard are typically broken
down into clods 25 to 50 mm in diameter (a function of the tracks of
the bulldozer used in blending backfill). Thus, adequate sampling, in
quantity and quality, must be taken to ensure that the requirements of
the project are being met despite this heterogeneity of the backfill. The
methylene blue test may be used in the field to determine the bentonite
content of the soil-bentonite backfill, provided a calibration curve is
developed initially.
Quality control tests are required for the raw materials and the as-
mixed products, specifically the bentonite and bentonite-water slurry.
The bentonite should conform to the project specifications; usually a
slurry grade bentonite with a yield between 90 and 150 barrels. In
addition, the as-mixed bentonite-water slurry must have the proper
Marsh viscosity, filtrate loss and mud density (API, 1984) character-
istics to ensure slurry thin enough to be displaced by the backfill
placement, yet thick enough (and of the appropriate density) to provide
trench stability (Table 17.1). Finally, the unit weight of the as-mixed
slurry is at least 1.02 g/cm3 and is higher in the trench. A maximum
density in the trench or a density differential may be specified. It is
often specified that the bulk density of the slurry in the trench be a
minimum of 0.25 g/cm3 lighter than the backfill which is displacing the
slurry.

17.6.8 Potential failure mechanisms

In order to design and construct properly a soil-bentonite slurry trench
cutoff wall, it is useful to examine potential failure mechanisms. The
failure mechanisms may be classified into two categories: (a) construc-
tion defects; and (b) post construction property changes. The following
discussion addresses these potential failure mechanisms and identifies
 Feasibility studies 443

areas where the engineer and contractor can focus to minimize the
potential for these failure mechanisms (Evans et al., 1985a).
Improperly mixed backfill may contain lumps of unmixed granular
soils and/or pockets of bentonite-water slurry. Either of these hetero-
geneities in the backfill result in zones of higher hydraulic conductivity
within the cutoff wall. Improperly mixed backfill is particularly difficult
to control when the backfill is mixed alongside the trench as compared
to backfill mixed in a remote mix area.
Backfill inhomogeneity may result from slurry entrapment during
placement. For example, if the backfill is too stiff (i.e., the slump is too
low), the backfill may override the slurry and entrap pockets of slurry.
Further, entrapped slurry may result if care is not taken during place-
ment of the backfill and backfill is dropped through the slurry. Finally,
the mechanism of backfill movement inherently risks the entrapment
of thin layers of slurry, although it is expected that these consolidate to
an inconsequential thickness.
Backfill inhomogeneity may also result from trench sediments covered
by the backfill during placement. Trench sediments result from settle-
ment of the more granular fraction from the slurry during periods on
inactivity (i.e., overnight, weekends, holidays). Sediment accumula-
tions need to be checked prior to the start of construction each day.
The coarser fraction in the sediment results in an area of the trench
with higher hydraulic conductivity.
Materials in the trench bottom also result from areas of trench
instability visible from the surface but difficult for the contractor to
remove. However, it is important to enforce the clean out of any cave-
in from trench instability in order to provide a cutoff wall that does not
contain windows. Finally, there may be spalling from the formation,
which is not visible from the surface. Careful sounding of the trench
bottom immediately prior to backfill placement can go a long way to
ensure that backfill spalling does not occur.
Even with a properly constructed soil-bentonite slurry trench cut off
wall without construction defects, changes in the properties of the
soil-bentonite backfill which are time dependent can occur after con-
struction. The time dependent property changes are discussed in the
following paragraphs.
Increases in hydraulic conductivity can occur due to cycles of freezing
and thawing, cycles of wetting and drying, the development of ice
lenses within the backfill or desiccation. A properly designed final
cover will prevent freezing and formation of ice lenses. Recent test
results have shown increases in the hydraulic conductivity in the
portion of the wall at and above the water table (Cooley, 1991).
Soil-bentonite slurry trench cutoff backfill is placed as a saturated
cohesive material with high slump and, as a result, consolidates under
the self-weight loads. Due to friction along the sidewalls of the trench,
444 Vertical cutoff walls

the consolidation settlement results in an unknown state of stress. If

high hydraulic head conditions are imposed (such as beneath the
dam), it is possible that hydraulic fracturing may occur. It has been
found that hydraulic fracturing occurs when the hydrostatic stress
exceeds the mean radial stress (Bjerrum et al., 1972). It is unlikely that
hydraulic fracturing will occur due to drawing down the water table on
the down-gradient side of the cutoff wall.
The most important consideration for the long-term behavior of
the soil-bentonite slurry trench cutoff wall is chemical incompatibility.
Chemical incompatibility may be global as a result of contaminants
dissolved in the ground water, causing hydraulic conductivity increases
in the soil-bentonite. Local incompatibility may result from concen-
trated organics (non-aqueous phase liquids, or NAPLs) floating or
sinking within the formation and attacking localized areas of the cutoff
wall.
When using soil-bentonite slurry trench cutoff walls in waste disposal
applications, it is necessary to control hydraulic conductivity increases
due to chemical attack. Increases in hydraulic conductivity can be
minimized by utilizing a well-graded base soil with a high natural fines
content. In this way, the soil consists of a matrix of progressively
smaller materials which are relatively inert (and not subject to changes
in their diffuse ion layer as a result of changes in pore fluid chemistry).
With a backfill designed in this manner, the hydraulic conductivity will
remain quite low (i.e., about 10- 6 cm/s) even with physiochemical
attack and shrinkage of the bentonite.
It is also important to control settlement of soil-bentonite slurry
trench cutoff walls so as to provide the maximum stress in the cutoff
wall and reduce the potential for hydraulic fracturing. Fortunately,
settlement can be controlled in the same way as control of increases in
hydraulic conductivity through the use of a well-graded base soil with
a high natural fines content.

17.7 CEMENT-BENTONITE SLURRY TRENCH

CUTOFF WALLS

17.7.1 General information and technology

Cement-bentonite slurry trench cutoff walls are constructed using the
slurry trench method of excavation. The difference between cement-
bentonite and soil-bentonite vertical cutoff construction is that the
slurry used in the excavation process is not replaced with a low perme-
ability material as in the soil-bentonite slurry trench cutoff wall tech-
nique. Instead, the slurry, which contains bentonite, cement and water,
is left to harden in place. The hydraulic conductivity of cement-
 Cement-bentonite slurry trench cutoff walls 445
bentonite slurry trench cutoff walls typically varies between 1 X
10- 5 cm/sec and 1 X 10-6 cm/s, as measured in the laboratory on field
mixed samples compared with 1 X 10- 7 cm/s to 1 X 10- 8 cm/s for soil
bentonite.

17.7.2 Design feasibility studies

Since the materials excavated from the trench are not reused for
cement-bentonite slurry trench cutoff walls, the ultimate disposal of
these materials must be considered. The costs and logistics associated
with disposal of the excavation quantities of hazardous waste must be
considered in the feasibility analysis.
The shear strength of cement-bentonite increases with time and
cement content. For typical mixes, the strength of cement-bentonite is
similar to medium to stiff clays. Cement-bentonite backfill is much
stronger and less compressible than soil-bentonite backfill but is more
permeable. Where strength is a particularly important requirement,
additional attention should be focused on this method of vertical cutoff
wall construction.
Recently, progress has been made in developing construction tech-
niques that permit the insertion of geomembrane or polymer sheet
piling in a cement-bentonite wall. As the construction techniques are
perfected and experience with them is gained, this type of design
should see a large increase in use.

17.7.3 Final design studies

Final design studies incorporate most of the items described earlier for
soil-bentonite slurry trench cutoff walls along with additional consider-
ations unique to cement-bentonite. For example, it is important to
study the hydraulic conductivity of cement-bentonite. Experience has
shown that there is little relationship between hydraulic conductivity
and cement content. Hydraulic conductivities <10- 7 cm/s can be
achieved but only with difficulty. A hydraulic conductivity ::::;}O-6 cm/s
is a typical design goal for cement-bentonite walls.

17.7.4 Technical specifications

Bentonite-water slurry used in the preparation of cement-bentonite
slurry is normally prepared using a high speed colloidal shear mixer. In
this mixer, a stabilized Marsh viscosity is obtained for the bentonite-
water slurry in a matter of five minutes or less. The specifications
must cite the requirements for the stabilized Marsh viscosity of the
bentonite-water slurry. The cement is then added and the resulting
cement-bentonite-water slurry is pumped to the trench.
446 Vertical cutoff walls
As a result, at the end of each day, the trench is left filled with
cement-bentonite and partially unexcavated. In the following construc-
tion day, it is important to fully remove the materials at the interface
between the previously placed cement-bentonite and the unexcavated
materials to assure a good key between the materials. This 'cold joint'
is normally satisfactory, assuming a thorough removal of unexcavated
materials from the joint area.

17.7.5 Construction considerations

The construction of a cement-bentonite slurry trench cutoff wall
requires considerably less working area than the construction of a soil-
bentonite slurry trench cutoff wall. There is no need to have large scale
mixing ponds for the hydration of bentonite, nor is there a need to
have a backfill mixing area. Thus the working platform can be relatively
narrow, and the equipment used to mix the cement-bentonite can be
very compact. The bentonite, the cement, and the mixers can all be
efficiently located and the completed cement-bentonite slurry pumped
to the trench. Further, concern about grades may be less significant for
cement-bentonite. Since the trench is constructed incrementally (daily)
and allowed to harden overnight, steps in the elevation of the trench
can be made to permit construction along steeper site grades.

17.7.6 Quality control

The parameters measured for cement-bentonite quality control are
similar to those for water-bentonite slurries. Note that filtrate loss is
not normally measured for the cement-bentonite method of cutoff
construction. This is because filtrate loss, a measurement of the quality
of the bentonite, is greatly increased when the cement is added to the
bentonite-water slurry.
It is important to obtain field mixed cylinders of cement-bentonite for
subsequent laboratory permeability and strength testing. The cylinders
should be cured at 100% humidity for the curing time (typically 28
days) prior to testing.

17.7.7 Failure mechanisms

The potential for construction defects due to backfill inhomogeneity
and slurry entrapment is greatly reduced using cement-bentonite
slurry trench method of construction. Because there is no backfill
mixing or placement and the slurry is left to harden in place, the
potential for construction defects is considerably reduced. The primary
concern is an inadequately mixed batch of cement-bentonite-water
slurry, or improper composition of the slurry. With proper construc-
 Steel sheet piling 447
tion control, however, the ingredients (bentonite, water and cement)
can be added in the proper proportions and the possibility of construc-
tion defects considerably reduced.
Certain post-construction property changes are likewise of less
concern for cement-bentonite than for soil-bentonite. Changes in long-
term performance due to settlement and potential hydraulic fracturing
are reduced due to the hydrated cement structure that develops. How-
ever, chemical incompatibility, both global and local must be con-
sidered. There is less data available on cement-bentonite than on soil-
bentonite. Further, freeze-thaw, desiccation and wet-dry cycles, which
could alter the performance of the wall as previously described, have
similarly not been well studied.
Construction defects unique to cement-bentonite include accurate
batching of mix components, 'first batch' effects, filtrate loss, and cold
joints. The first batch effects result from the fact that the first batch of
the day is prepared in a clean mixer, that is, clean water is added,
bentonite hydrated, and then cement is added. This often results in a
much thicker slurry than in subsequent batches, where a thin film of
bentonite-cement slurry coats the mixer prior to the addition of the
next batch of mix component. It is important to test the materials after
the first batch is prepared to get an accurate representation of what
typically appears in the field.

17.8 STEEL SHEET PILING

Steel sheet piling has found widespread application in conventional

civil engineering construction for reduction of water flow in the sub-
surface. However, for environmental applications, steel sheet piling
has rarely been employed. This is because steel sheet piling is subject
to leakage through the interlocks, and as such, the leakage is higher
than typically accepted for waste containment applications.
In order to reduce seepage through the interlocks, steel sheet piling
has been used in conjunction with cement-bentonite on at least one
occasion. In this project, a cement-bentonite slurry trench cutoff wall
was excavated into the subsurface. Prior to the initial set of the cement-
bentonite slurry, steel sheet piling was inserted into the cutoff wall. In
this way, a composite wall was constructed which had the benefits
of high resistance to hydraulic fracturing under the projected head
conditions, plus the impedance of leakage through the joint by essen-
tially grouting the joints with cement-bentonite. Although this solution
was costly, the limited length of the required cutoff wall made it the
most practical solution.
Until developments are made to reduce the leakage through the
interlocks of steel sheet piling, it is likely that the use of steel sheet
448 Vertical cutoff walls

piling as vertical cutoffs in waste disposal applications will be limited.

It is further noted that the relatively high cost of the steel sheet piling
for a permanent installation provides further disincentive to its use.

17.9 PLASTIC CONCRETE

Recent developments in the study of materials for use in slurry trench

cutoff walls had given rise to the use of plastic concrete in vertical
cutoff wall applications. In a plastic concrete vertical cutoff, the trench
is excavated in panels using the slurry trench method of construction,
as in a soil-bentonite slurry trench cutoff wall. However, the low
hydraulic conductivity of the completed wall is a result of replacement
of the bentonite-water slurry with a plastic concrete material. This
plastic concrete, often known as lean concrete, consists of a mixture of
water, cement, aggregate, and bentonite. The resultant characteristics
provide material that is flowable (as soil-bentonite backfill), but hardens
in place (as cement-bentonite). The resultant plastic concrete wall is
significantly stronger than soil-bentonite and has low hydraulic con-
ductivities desirable for environmental applications.
The hydraulic conductivity of plastic concrete is typically less than
1 x 10- 8 cm/sec. Further, the long-term hydraulic conductivity in
response to permeation with concentrated organic fluids indicates no
degradation in hydraulic conductivity. This is in significant contrast
to the degradation which is demonstrated using soil-bentonite slurry
trench cutoff walls. Thus, for environmental applications, plastic con-
crete appears to provide a combination of low initial hydraulic con-
ductivity and excellent long-term endurance.
Plastic concrete has been used in several applications, but its use is
limited compared with soil-bentonite because of high cost. The cost of
a plastic concrete backfill is greater than that of a soil-bentonite backfill,
particularly where the excavated material in a soil-bentonite trench
can be reused. However, its durability and permanence in specific
applications may outweigh the cost differential, particularly where
the cutoff wall is used as a component in an overall groundwater
remediation program.

17.10 DIAPHRAGM WALLS

Vertical cutoffs can be constructed in the subsurface as diaphragm

walls. A diaphragm wall is typically constructed in panels, as shown
on Fig. 17.4 using a bentonite-water slurry to maintain the excavation
stability. The bentonite-water slurry is displaced by concrete using the
tremie method of displacement. The diaphragm wall has historically
 Vibrating beam cutoff walls 449

been used for dewatering of structures where the completed cutoff wall
is used in the soil retention system. The cutoff wall may then be
incorporated into the overall structure. The structural integrity of the
cutoff wall is then considered in the overall stability of the completed
structure.
Trench excavation is completed utilizing one of several techniques,
including clam shell. After completion of the trench excavation, the
pre-assembled reinforcing cage is lowered into place, and concrete
is tremied to displace the bentonite-water slurry. The strength and
hydraulic conductivity of the final wall is similar to reinforced concrete.
The panels are completed in an alternating sequence and special
precautions are made to ensure a leakproof joint.
Diaphragm walls have been successfully employed on many projects
and visual examination of the completed cutoff walls is possible as
interior excavation proceeds. Problems occasionally develop where a
window in the wall results from improper tremie technique. In most
cases, diaphragm walls provide a positive vertical cutoff to ground-
water flow in the subsurface.
Diaphragm walls have not been applied to environmental appli-
cations due to their high costs.

17.11 VIBRATING BEAM CUTOFF WALLS

Vertical cutoffs can be constructed in the subsurface utilizing the

vibrating beam method of construction. This is not technically a slurry
trench method of construction since a trench is not excavated in the
subsurface utilizing a slurry to maintain the trench stability. In this
application, a beam is modified with grout injection nozzles and con-
nected to a vibratory pile driver. The beam is vibrated into the ground,
creating a slot, which is then filled with grout as the beam is with-
drawn. Grout is also injected as the beam is inserted to act as a
lubricant for beam penetration. Successive penetrations of the beam are
made, and the beam follows the previous footprint and a continuous
wall is constructed. Several studies have been shown to demonstrate
the viability of this technique (Leonards et al., 1985). The grout material
can consist of cement-bentonite material or alternative materials such
as bitumen based asphaltic materials.
The principal advantage of this technology is that it is not necessary
to excavate materials from the subsurface and handle or dispose of the
excavated materials. Since there is no excavation process, health and
safety considerations, and cost, are potentially reduced. The principal
disadvantage comes in the assurance of the integrity of the cutoff wall.
Where walls are to be constructed in shallow depths in relatively loose
materials, the stiffness of the beam goes a long way to ensure the
450 Vertical cutoff walls

Slurry supply pipe

(a)

(c)

Fig. 17.4 Diaphragm cutoff wall: (a) excavate soil and replace with bentonite
slurry; (b) place stop-end tubes and reinforcing steel into fully excavated panel;
(c) pour tremie concrete to displace slurry, remove stop-end tubes, and (d)
different construction phases.

continuity of the cutoff wall. However, where the barrier is very deep,
or where the subsurface conditions vary, the stiffness of the beam can
no longer ensure the continuity of the cutoff wall. Further, even in
loose, granular materials which are initially easily penetrated with the
vibrating beam, the vibration effects tend to densify the materials,
making progressive penetrations successively harder. In this case, it
may be necessary to pre-auger the materials (McLay, 1987).
The hydraulic conductivity of the vibrating beam cutoff wall will be
 Vibrating beam cutoff walls 451

Reinforcing steel

~+++++++411 /'

.. :..::.,' .....', ... :. . .

:
. .: :" .:. ..:: ": :":':' .:. :.: ::. ": "
.... : .. :..: ... ... :.::..:

(b)

, - Panel to be concreted

, ..
,.... ,.r/'.".,.
.., ....

(d)

similar to that of cement-bentonite slurry trench cutoff walls and as a

result may be higher than desirable. Further, since the wall thickness is
related to the thickness of the beam, the thickness of these cutoffs are
generally 150 mm or less.
The vibrating beam has found limited use in environmental appli-
cations. Design feasibility studies typically tip the scale toward this
technique where the cost and hazards of excavating large quantities of
hazardous materials impede the use of one of the more traditional
techniques.
452 Vertical cutoff walls

17.12 DEEP SOIL MIXED CUTOFF WALLS

A technique developed in Japan and recently used in the United States,

known as deep soil mixing, can be employed to construct a vertical
cutoff wall. A series of auger shafts, arranged in line, are used. The
augers, which are modified to include mixing paddles, penetrate the
ground while simultaneously allowing for addition of bentonite-water
slurry. The result is a soil-bentonite cutoff wall mixed in situ. A con-
tinuous wall results from successive overlapping penetrations (Ryan,
1987).
The principal advantage is the reduced health and safety risk asso-
ciated with in situ mixing. The principal disadvantage results from the
engineer's lack of control over backfill properties.

17.13 VERTICAL DRAINS

Recent developments have resulted in the development of permeable

vertical walls that serve as drains. Groundwater or leachate flows
towards the vertical drain where the liquid is collected. From a theor-
etical standpoint, vertical drains can be analyzed as slots and flow to
vertical permeable cutoffs is easily calculated.
The method of construction of the permeable cutoffs is similar to that
for soil-bentonite slurry trench cutoff walls. A trench is first excavated
under the head of a slurry. However, in the case of permeable cutoffs,
the slurry is composed of a biodegradable drilling mud (a biopolymer)
and water. In this way, a thickened mud is used during excavation,
much the same as bentonite-water slurry. The slurry is then displaced
with a permeable material such as sand and gravel. In time, the mud
biodegrades, leaving an open formation. In this way, after the trench is
backfilled and the mud biodegrades, a permeable slot remains in the
subsurface from which groundwater can be pumped (Day, 1991).
The development of this technique provides tremendous advantages
over previous construction techniques for the installation of drain tile
collection systems. Previous to this large excavations with backs lopes
or trench supports were required, including the handwork of laborers
within the trench. Using this technique, the trench support is provided
by the biodegradable slurry, and the process proceeds much more
rapidly, much more safely and at considerably lower cost.

17.14 CAPS FOR CUTOFF WALLS

The construction of a vertical cutoff in the subsurface usually necessi-

tates the subsequent construction of a cutoff wall cap. Most of the
 References 453

barrier materials described are subject to desiccation and freeze-thaw

degradation and may be relatively soft compared with the adjacent
ground. The construction of a cap provides the necessary protection.
The final design studies must consider traffic loading, both parallel
and transverse, and the development of design alternatives. Capping
of cutoff walls can be easily accomplished using geosynthetics,
geogrids, or high-strength geotextiles for reinforcement and geo-
membranes, compacted clay, or geosynthetic clay liners for control of
water movement.

17.15 MONITORING PERFORMANCE

Since the cutoff wall is constructed beneath the head of bentonite-

water slurry in the subsurface, as-built conditions are very difficult
to ascertain. Once installed, the monitoring of vertical cutoff wall
performance is also difficult. Methods to establish the cutoff wall per-
formance include the establishment of background levels of ground-
water and contaminant distribution. Wells can be strategically placed
to monitor changes with time. These wells can monitor performance
during implementation of the system and be used for field pump tests.
In addition, geophysical approaches have been tried.
The most successful method to ensure adequate performance in the
long term is to conduct thorough design studies described earlier and
provide competent field control during construction.

REFERENCES

Alther, G.R., Evans, J.c., Witmer, K.A. and Fang, H.Y. (1985) Inorganic
permeant effects upon bentonite, Hydraulic Barriers in Soil and Rock, ASTM
STP 874, Am. Soc. for Testing and Materials, Philadelphia, pp. 64-74.
American Petroleum Institute (1984) Standard Procedure for Field Testing Drilling
Fluids, API Recommended Practice, API RP 138, 10th edn, Dallas, Texas, June
l.
Bjerrum, L., Nash, J.K.T.L., Kennard, R.M. and Gibson, R.E. (1972) Hydraulic
fracturing in field permeability testing, Geotechnique, 22(2), 319-32.
Boyes, R.G.H. (1975) Structural and Cut-Off Diaphragm Walls, Applied Science
Publishers Ltd., London, England.
Campbell, D.L. and Quintrell, W.N. (1985) Cleanup strategy for Rocky
Mountain Arsenal, Proc., 6th Nat. Conf. on the Management of Uncontrolled
Haz. Waste Sites, HMCRI, November 4-6, Washington, DC, pp. 36-42.
Case International Company (1982) Case Slurry Wall Notebook, Manufacturers
Data, Case International Company, Houston, Texas.
Cooley, B.H. (1991) State of stress in soil bentonite slurry trench cutoff walls,
Honors thesis, Bucknell Univ., Lewisburg, PA.
D' Appolonia, D.J. (1980) Soil-bentonite slurry trench cutoffs, J. Geotech. Eng.
Div., ASCE, 106(4), 399-417.
454 Vertical cutoff walls

Day, S. (1991) Extraction/interception trenches by the bio-polymer slurry

drainage trench technique, Haz. Mat. Control, 4(5), Sept/Oct, 27-3l.
Evans, J.C and Fang, H.Y. (1988) Triaxial permeability and strength testing of
contaminated soils, Advanced Triaxial Testing of Soil and Rock, ASTM STP
977, Am. Soc. for Testing and Materials, Philadelphia, pp. 387-404.
Evans, J.C, Fang, H.Y. and Kugelman, I.J. (1985a) Containment of hazardous
materials with soil-bentonite slurry walls, Proc., 6th Nat. Conf. on the
Management of Uncontrolled Haz. Waste Sites, Washington, DC, Nov, pp.
249-52.
Evans, J.C and Hill, G.H. (1984) Phase I Remediation Studies, Necco Park, Niagara
Falls, NY, Woodward-Clyde Consultants, June 18.
Evans, J.C, Lennon, G.P. and Witmer, K.A. (1985b) Analysis of soil-bentonite
backfill placement in slurry walls, Proc., 6th Nat. Conf. on the Management of
Uncontrolled Haz. Waste Sites, Washington, DC, Nov, 357-6l.
Evans, J.C, Stahl, E.D. and Droof, E. (1987) Plastic Concrete Cutoff Walls,
Geotechnical Practice for Waste Disposal '87, ASCE Geotechnical special publi-
cation (13), June, pp. 462-72.
Jefferis, S.A. (1981) Bentonite-cement slurries for hydraulic cut offs, Proc., 10th
Int. Conf. on Soil Mechanics and Foundation Eng., Stockholm, June 15-19,
A.A. Balkema, Rotterdam, 435-40.
Leonards, G.A., Schmednecht, F., Chameau, J.L. and Diamond, S. (1985) Thin
slurry cutoff walls installed by the vibrated beam method, hydraulic barriers in soil
and rock, ASTM STP 874, (eds A.I. Johnson, R.K. Frobel, N.J. Cavalli and
CB. Peterson) Am. Soc. for Testing and Materials, pp. 34-44.
McLay, D. (1987) Installation of a cement-bentonite slurry wall using the
vibrating beam method - a case history, Proc., 19th Mid-Atlantic Ind. Waste
Conf., Bucknell Univ., Lewisburg, PA, June, 272-82.
Millet, R.A. and Perez, J.Y. (1981) Current USA practice: slurry wall specifi-
cations, J. Geotech. Eng. Div., ASCE, 107(GT8), Aug, 1041-56.
Nash, K.L. (1974) Stability of trenches filled with fluids, J. Canst. Div., ASCE
100(4), 533-42.
Ryan, CR. (1987) Soil-bentonite cutoff walls, Geotechnical Practice for Waste
Disposal '87, (ed. R.D. Woods), ASCE, New York, pp. 182-204.
Spooner, P., et al. (1984) Slurry Trench Construction for Pollution Migration
Control, EP A-540/2-84-00l, US Environmental Protection Agency,
Cincinnati, Ohio.
Xanthakos, P.P. (1979) Slurry Walls, McGraw-Hill Book Co., New York.
 CHAPTER 18

Cover systems
David E. Daniel and Robert M. Koerner

18.1 INTRODUCTION

Final cover systems serve a variety of functions for both new waste
disposal units as well as site remediation projects. The general prin-
ciples of design of cover systems are the same for new disposal units as
well as old waste disposal areas undergoing remediation.
Final cover systems are a critical component in the overall process of
managing liquid and gas movement into and out of the underlying
contaminated material. The main objectives in designing a final cover
system are to separate the buried waste or contaminated material from
the surface environment, to restrict infiltration of water into the waste,
and (in some cases) to control release of gas from the waste. Because of
the variety of waste and site characteristics, each cover design must be
tailored to the specific requirements of each project. It is important for
design engineers to have a thorough understanding of the basic prin-
ciples involved in cover design because the final cover system profile
cannot be generalized for all sites and all waste types.

18.2 BASIC CONCEPTS FOR COVER SYSTEMS

Cover systems may be called upon to perform one or more of the

following functions:
• raise ground surface elevation in low-lying areas;
• minimize the amount of runoff of precipitation;
• promote controlled runoff of whatever precipitation is remaining;
• separate the waste from plants and animals;
• prevent migration of perched leachate out of waste on side slopes;

Geotechnical Practice for Waste Disposal.

Edited by David E. Daniel.
Published in 1993 by Chapman & Hall, London. ISBN 0 412 35170 6
456 Cover systems

• limit infiltration of precipitation into the waste; and

• control release of gas from the waste.
The US Environmental Protection Agency (EPA) considers keeping
water out of the contaminated material to be the prime element in the
final cover (US EPA, 1989).
The cover must perform the intended functions for extended periods
of time. The 'design life' of a cover depends primarily upon the nature
of the waste, the hydrogeology of the site, the length of time that
maintenance of the cover will be provided, and the likelihood of
funds being available for repair of the cover. Unfortunately, too many
designers are forced to design a final cover system for an actively
decomposing and deforming body of waste; a better approach for such
wastes might be to construct a temporary cover and then wait until
substantial decomposition of the waste occurs before attempting to
construct a final cover system. Unfortunately, this latter strategy may
require the passage of many years before most of the decomposition is
complete; special financial arrangements would be required because
the responsible party may not be viable when the time finally comes to
construct the cover. Thus, final covers are usually constructed shortly
after the last placement of waste in new landfills or concurrently with
overall cleanup activities for sites undergoing remediation.
The design of the final cover is complicated by environmental factors
that include:
• temperature extremes, possibly including freeze/thaw to significant
depths;
• cyclic wetting and drying;
• penetration by plant roots, burrowing animals, worms, and insects;
• total and differential settlement caused by compression of the under-
lying waste or foundation soil;
• temporary (or permanent) surcharge loads, e.g., soil stockpiles or
materials storage;
• down-slope slippage or creep;
• vehicular movements on haul roads traversing the cover;
• wind or water erosion;
• deformations caused by earthquakes in seismically-active areas,
• long-term moisture changes caused by water movement into or out
of the underlying waste; and
• alterations caused by gas derived from volatile components of the
waste or decomposition products.
Because of these site-specific environmental stresses and conditions,
the design of a cover system can be very challenging. It is often more
difficult to provide an effective hydraulic barrier layer in a cover system
than in a liner system because the cover system is challenged by
 Details of cover system components 457
unknown and unquantifiable stresses that do not act on liner systems
buried deep beneath the waste.
Most cover systems are composed of multiple components. As
shown in Fig. 18.1, the components of a cover system can be grouped
into five categories:
1. surface layer;
2. protection layer;
3. drainage layer;
4. barrier layer; and
5. gas collection layer.
Not all components are needed for all final covers. For example, a
drainage layer may be needed at a site with large precipitation but not
at an arid site. Similarly, a gas collection layer is a required component
of covers placed over wastes that produce gas but would serve no
useful purpose for waste that does not produce gas.
The range of materials used for the various components of a cover
system are also listed in Fig. 18.1. The materials used to construct the
components of a final cover system can include natural earth materials,
geosynthetics, a blend of natural materials, or a blend of earth and
geosynthetic materials, e.g., geosynthetic clay liners.

18.3 DETAILS OF COVER SYSTEM COMPONENTS

IS.3.1 Surface layer

There are four alternative materials that can be used for the surface
layer:
1. soil (Figs. 18.2a and 18.2b);
2. soil with a geosynthetic erosion control layer placed at the surface
(Fig. 18.3);
3. cobbles (Figs. 18.4a and 18.4b); or
4. paving material (Fig. 18.5).
The advantages and disadvantages of the various materials are ident-
ified below.

Alternate A: uniform protective soil (Fig. 18.2a)

Advantages
1. Construction material is locally available.
2. Special topsoil is not needed.
3. Soil stores water that has infiltrated; plants return much of this
water to the atmosphere via evapotranspiration.
458 Cover systems

.£!£!!!!. .J.!Y"L Primary function(s) Usuaimaterials General considerations

1. Surface Promote vegetative growth Topsoil (humKt site); Surface layer for control of water
layer (most covers); promote cobbles (arid site); and/or wind eroskln is always
evapotranspiration; geosynthetic erosion required
:g prevent erosion control systems
(J)
2. Protection Store water; protect Mixed soils; cobbles Some form of protective layer is

~ layer underlying layers from

intrusion by plants,
animals, and humans;
always required; surface layer
and protective layer may be
combined into a single 'cover
protect barrier tayer from soil'layer
desiccation and freeze!
thaw; maintain stability
3. Drainage Drain away Infitlrating Sands; gravels; Drainage layer is optional;
layer water to minimize geotextiles; geonets; necessary only where excessive
barrier layer contact geocomposites water passes through protection
and to dissipate layer or seepage forces are
seepage forces excessive
4. Barrier Minimize infiltration of Compacted clay liners; Barrier layer is usually required;
layer water into waste geomembranes; may not be needed at extremely
and escape of gas geosynthetic clay arid sites
out of waste "liners
5. Gas Transmit gas to Sand; geotextiles; Required rt waste produces
collection collection points for geonelS excessive quantities of gas
layer removal and/or
cogeneration

Fig. 18.1 Five possible components in a cover system.

4. Soil (if adequately thick) protects liner from freeze/thaw and

wetting/drying.
5. Straightforward construction.

Disadvantages
1. Requires adequate precipitation to support growth of vegetation on
soil.
2. Irrigation could be required during periods of drought to maintain
vegetative cover.
3. Establishing vegetative cover may be difficult if weather conditions
are unfavorable immediately after seeding.
4. Excessive wind or water erosion may occur prior to establishment of
vegetative cover or if vegetation dies.
5. Vegetative cover may be difficult to establish (no topsoil).
6. Little resistance to intrusion by plant roots or burrowing animals.
7. Probably the thickest of all cross-sections adding to possible dif-
ferential settlement.

Assessment
This design has been used where special topsoil is not needed to
support growth of a suitable vegetative cover. This is the simplest of
possible cover soil designs but provides little beyond physical separation
between the ground surface and underlying layers and storage of
water for evapotranspiration. The uniform protective soil layer should
be relatively thick compared to other schemes.
 Details of cover system components 459

(a)

(b)

ww;>«":'~';' .".;,....... .

, pr~Clive soil

Fig. 18.2 (a) Uniform protective soil; (b) topsoil overlying protective soil.

Alternate B: topsoil overlying protective soil (Fig. 18.2b)

Advantages
1. Construction materials are usually available locally.
2. Topsoil and protective soil layer store water that has infiltrated; plants
return much of this water to the atmosphere via evapotranspiration.
3. Plants not only return water to the atmosphere but also reduce soil
erosion and stabilize the topsoil.
4. Thickness of protective soil layer can be adjusted as necessary (e.g.,
to prevent barrier soil from freezing) without affecting the thickness
of topsoil.
5. Plant roots tend to be confined to nutrient-rich topsoil (deep pen-
etration of roots is undesirable).
460 Cover systems

Fig. 18.3 Erosion control layer over topsoil over protective soil in a cover
system.

Disadvantages
1. Requires adequate precipitation to support growth of vegetation on
topsoil (design unsuitable for some arid sites).
2. Irrigation could be required during periods of drought to maintain
vegetative cover.
3. Establishing growth of vegetation may be difficult if weather con-
ditions are unfavorable immediately after seeding.
4. Excessive wind or water erosion may occur prior to establishment of
vegetative cover or if vegetative cover becomes thin.
5. Little resistance to intrusion by burrowing animals.

Assessment
This is the most common soil cover design used in cover systems
and is the minimum required cover soil design established by many
regulatory agencies. The main problems experienced with this design
are maintaining a vegetative growth during drought periods and for-
mation of erosion gullies in the top soil during excessive precipitation
periods. This design can be very effective at sites that receive adequate
rainfall but generally requires maintenance.

Alternate C: erosion control layer over topsoil over protective soil (Fig. 18.3)
Advantages
1. Topsoil and protective soil layer store water that has infiltrated; plants
return much of this water to the atmosphere via evapotranspiration.
2. Plants plus erosion protection layer minimize soil erosion; erosion
 Details of cover system components 461

protection layer minimizes erosion while plant growth is established

and erosion rills and gullies afterward.
3. Thickness of protective soil layer can be adjusted as necessary
without affecting the thickness of topsoil.
4. Plant roots tend to be confined to nutrient-rich topsoil (deep pen-
etration of roots is undesirable).
5. Erosion layer may offer resistance to burrowing animals.
6. Erosion control geosynthetic can be anchored or reinforced to add
stability to steep side slopes.
7. The geosynthetic erosion control material can be preseeded.

Disadvantages
1. Requires adequate precipitation to support growth of vegetation on
topsoil (design unsuitable for some arid sites).
2. Irrigation could be required during periods of drought to maintain
vegetative cover.
3. Establishing growth of vegetation may be difficult if weather con-
ditions are unfavorable immediately after seeding.

Assessment
This design is not nearly as common as the same design without the
geosynthetic erosion control layer. However, many erosion problems
have been observed on covers; the erosion control layer is recom-
mended for many (if not most) sites.

Alternate 0: cobbles over protective soil (Fig. IB.4a)

Advantages
1. No need to establish or maintain vegetative growth (particularly
advantageous for arid sites or sites subject to periodic long droughts).
2. Cobbles provide protection from wind or water erosion; effective-
ness of cobbles not dependent upon maintenance of vegetative
growth.
3. Cobbles may minimize penetration from burrowing animals if rocks
are of adequate size.
4. Cobbles minimize growth of plants with deep roots that could pen-
etrate underlying barrier layers.

Disadvantages
1. Cobbles must be adequately large to resist washout; materials may
not be readily available locally.
2. Minimal evaporation of water from underlying soil (cobbles allow
water to percolate downward but insulate soil from surface
evaporation).
 462 Cover syste
ms

(b)

ted
bles; deeply-roo
y fi ll vO ;ds between cob ed a n d requil-e
n
o w mater;als m a e estabUsh
3. W ;n d -b
l
ce . a y e v e n tu ally becom
a n m
m ain
p la n ttengroWth st e e p slope
s.
fo r
concerns
4. Stability
t
A ss e ss m e n .
blems
Prov;ded pro ve
a ri d sU es
ered o n ly fo
r m protect;
g n ;s US ually cOns;d f W ater e v a poration fro aPProPriate
Th;s de s; o st
ls a n d lack
w;ng an;ma rious, th;s des;gn m a
y be the m o
w ;t h b U rr o
d siteos.t too se
la yr earria re n
osonUe fo

(F;g. 18.4b)
ce and a l deplh
bbles 01 '" 'f o
A II "n a le E. co
Advantages w th (p a rt k
u Ia rl y
ve g ro
Or m a ;n t. ;n vegetati riodic major droUght).
sh
I. N o n e e
d to establi tes o r sUes su b je c " o pe erosh>n; effecnveness
u s fo r a ti d s; ;nd o r Water rowth.
advantageosUrface protect from W n a n c e o f vegetat;ve g
a t upon ma;n te
2. Cobbles t dependent
cobb le s n o
of
 Details of cover system components 463

Fig. 18.5 Paving overlying protective soil in a cover system.

3. Cobbles buried in the cover (biotic barrier) minimize penetration

from burrowing animals and plant roots if rocks are of adequate
size.

Disadvantages
1. Cobbles at surface must be adequately large to resist washout and
cobbles at depth must be very large to stop burrowing animals;
materials may not be readily available locally.
2. Minimal evaporation of water from underlying soil (cobbles at
surface allow water to percolate downward but insulate soil from
surface evaporation).
3. Stability concerns for steep slopes.

Assessment
This design is usually considered only for arid sites requiring maxi-
mum security. A biotic barrier has rarely been used, except for low-
level radioactive waste buried at arid sites. Provided problems with
minimal water evaporation from protective soil layer are not too serious,
this is a viable design for arid sites at which the main concern is uptake
of buried waste constituents by plant roots or burrowing animals. This
cross-section promises to be extremely costly.

Alternate F: paving overlying protective soil (Fig. IS.5)

Advantages
1. Construction materials are usually available locally.
2. May be able to use cover for parking of vehicles.
3. Surface layer also serves at least partly as hydraulic barrier.

Disadvantages
1. Paving will require maintenance.
2. No evaporation or evapotranspiration of water from protective soil.
464 Cover systems

Assessment
This design is only considered for site remediation projects where the
subsurface materials provide stable foundation support and the paving
material can be maintained.
Of the four main alternatives for the surface layer, soil has been
by far the most commonly used material. Soil is generally available
locally at nominal cost, and suitable soil will promote the growth of
vegetation. Roots of plants reinforce the cover soil, restrict the rate of
soil erosion, decrease runoff velocity, and remove water from the soil
via evapotranspiration. Vegetation also creates a leaf cover above the
soil to reduce rainfall impact and decrease wind velocity (Lutton et al.,
1979). Procedures for vegetating cover soils are reviewed by Gilman et
al. (1985).
Guidelines for soil depths are given by USEPA (1989). The main
problems experienced with soil are excessive erosion (especially on
newly-seeded covers or steeply sloping covers) and a need to irrigate
the soil during periods of drought in order to maintain growth of
vegetation. A geosynthetic erosion control layer can be very beneficial
in limiting problems with soil erosion in the period following installa-
tion and can even help to establish vegetative growth with seeds (even
on a time-release basis) embedded in the layer.
Swope (1975) studied 24 landfill covers in Pennsylvania and found
that 33% had slight erosion, 40% moderate erosion, and more than
20% had severe erosion. Johnson and Urie (1985) report that erosion is
made more severe from the installation of hydraulic barriers (which,
without a drainage layer, can produce soaked soil with large amounts
of runoff after heavy rains) and vehicles driving on the surface of the
cover. Many designers underestimate the problem of erosion of the
surface layer prior to establishment of a mature vegetative cover.
Additionally, the problem of stability of the cover soil must be
addressed. A number of failures of covers were reported by Boschuk
(1991). Most of these failures were attributed to fundamental errors
made in assessing the stability of the cover, or in several instances,
lack of an analysis of the stability of the cover. Geogrid or geotextile
reinforcement layers can be incorporated in the cover to increase
factors of safety against slope instability to acceptable levels. Particu-
larly critical interfaces along which slippage would be most likely will
be described later.
In areas where the amount of precipitation is inadequate to support
growth of a vegetative layer, an alternative is to armor the surface with
cobbles. However, cobbles serve as a one-way window for infiltrating
water, i.e., water can infiltrate downward through the cobbles, but
due to lack of plant roots, little water is returned to the surface via
evaporation.
 Details of cover system components 465

Asphaltic concrete surfaces have been used on occasion for the sur-
face layer (Repa et ai., 1987) but only for special remedial applications.

18.3.2 Protection layer

A protection layer (which will generally be some type of local soil) may
serve several functions:
• to store water that has infiltrated into the cover until the water is
later removed by evapotranspiration;
• to physically separate the waste from burrowing animals or plant
roots;
• to minimize the possibility of human intrusion into the waste;
• to protect underlying layers in the cover system from excessive
wetting/drying (which could cause cracking); and
• to protect underlying layers in the cover system from freezing
(which could also cause cracking).
Figures 18.2-5 illustrate the alternative materials for the protective
layer. Local soil is the most commonly used material. Soil can be very
effective in storing water for later evapotranspiration and provides
protection against freezing of underlying materials if the thickness of
the protective layer is sufficiently large. Maps of maximum frost pen-
etration depths are given in various publications, e.g., USEPA (1989).
Cobbles (Fig. 18.4) have occasionally been considered as a barrier to
plant roots and burrowing animals (Cline et al., 1982) but are not
commonly used.
If the protective layer is placed directly on a barrier layer, a plane of
potential slippage exists at the interface between the protective layer
and the barrier layer. The risk of instability will be particularly acute
following periods of heavy rain if no drainage layer is provided beneath
the protective layer. The engineer must ensure that an adequate
factor of safety exists at this and all other interfaces in the cover
system. If the originally-designed cover system does not have an
adequate factor of safety, the steps that are usually taken to increase
the factor of safety include use of different materials (stronger soils or
textured geomembranes), addition of a drainage layer, flattening of
slopes, benches within the slope, or reinforcement of cover soils with
geogrids or high-strength geotextiles.

18.3.3 Drainage layer

A drainage layer is often placed below the protective layer and above
the barrier layer (Fig. 18.1). There are three reasons why a drainage
layer might be desirable:
466 Cover systems

1. to reduce the head of water on the barrier layer, which minimizes

infiltration;
2. to drain the overlying protective layer, which increases the water-
storage capacity of the protective layer; and
3. to reduce pore water pressures in the cover soils, which improves
slope stability.
There are four alternative materials that can be used for the drainage
layer:
1. sand or gravel with either a soil filter (Fig. 18.6a) or a geotextile filter
(Fig. 18.6b);
2. a thick geotextile (Fig. 18.7), which serves as both a drain and a
filter;
3. a geonet with a geotextile filter/separator (Fig. 18.8); and
4. a geocomposite drain (Fig. 18.9), which consists of a polymeric
drainage core and an overlapping geotextile filter/separator.
Selection of the material type is usually based on economics; all the
materials listed above will function adequately if properly designed. If
air space has an economic value (e.g., for a new waste containment
unit), thin, geosynthetic materials rather than a thicker layer of sand or
gravel will usually prove to be the most economical material.
The designer is cautioned to be careful about two details. First, water
must discharge freely from the drainage layer at the base of the cover.
If the outlet plugs or is of inadequate capacity, the toe of the slope will
become saturated, develop excess pore water pressure, and potentially
become unstable. Drainage pipes that commonly surround a site at its
low elevations must have adequate capacity and cannot be allowed to
plug. Conversely, the designer must beware of the possibility that the
drainage layer may drain overlying soils so well that vegetation will
not grow during parts of the year without irrigation. This problem
is minimized by selecting surface and protective soils with adequate
water-holding capacity to maintain vegetative growth.
The advantages and disadvantages of the various alternates are now
considered in some detail.

Alternate A: sand or gravel drain with soil filter (Fig. 18.6a)

Advantages
1. Thickness of earthen materials helps to protect underlying layers
from intrusion, puncture, and temperature extremes such as freeze/
thaw.
2. Long history of use of soil filters and sand or gravel drains in
engineering applications.
 onents
o f COVer system comp 467
Details

..................... .. ......
.
n -....._. _ ..
S--a.-
.. .. ..-
:.: :£ o!

.......; -
r~-~.............."-:..:"-.:~.-::"., ~ - ..:" .
.:"..-::.."-:..:"-....::"..-::.

; : . ; ~ ;
(b)

~;i~ :\ : ~ ~ y ~ ~ : ~ i ~ : ~ f , : 1 ~ i f : ; ; r :
__ • _......_ _......
..::, ••
;l;:~~-;:-:-~:;:-;-;:;:-:-;:;:-;-;~~.--. ....:.l.
~~
) geotexti
le 61t.. .
1t,,; '" (b
h (, ) 'oil 6
d , gmvel d " 'i n wit
S ,n 0
Fig. 18.6
tages aVail,ble
Disadvan ,y n o t be ,e a d ll y
tials m
cavel mate em.
lo caitllayb.le Sand oc g e o n d e c ly ing g e o m
I. SO p o n c to c be n e e d e d
y e , In a y te n d to h io n in g layec m a y ed.
ca v e l d C "nage la
a y li n e c ; a cO s
c ,v e l m a y b e " 'j o ic d
2. G cd of g h c a n lea
e o s y n th e ti ticle SiZe o p e , Whic
bcane oc g s o n m a x im o m par ca te d o w n 'l b il it y .
0 , Ca'tc
ichon
g c a v e l te n d to m ig U c e , a n d 'l o p e in ,t a
s a n d oc a g e pceS S
3. Fines in o f u n c e le a 'e d " " e p
to bUilndtu p
it y can
Assessme , b o t it s a p p li c a b il
ally " 'O n d
is techn"ic" 'c h o n Inatecials
.
n a l d e 'i g n o n
v e n ti o b l e c
l] ,i , COn b y lack o f 'o it
a
b e li m it e d (Fig. 18.6b
)
il h g e n le Xlil, ji lt .,
v e l d m in w
l, B : '" n d 0 , g 'a in g
AdVanAta IIg e
'm s o
p m te ct undedy m
s 1 0 feo
" in a g e layec h e lp h e e z e /t h a w a n d
cavel d o c h 's
1. Tuhniccktu re s a n d 0 , g e " '" ,e extremes s wocks.
Players from ta . m p
g in e e ri n g
i n , in e n
geaVo! d "
y o f O 'e o f s 'n d Oc
is to r
2. L o n g h
468 Cover systems

3. Extensive history of use of geotextile filters to prevent clogging of

granular drainage layers.
4. Geotextile filters are readily available and can be shipped to any
location.
5. Geotextile filter should prevent fines from entering sand or gravel
layer.

Disadvantages
1. Suitable drainage materials may not be available locally.
2. Gravel drainage layer may tend to puncture underlying geomem-
brane or geosynthetic clay liner; a cushioning layer may be needed
or restrictions on maximum particle size of gravel may be required.

Assessment
This conventional design is technically sound. Applicability may be
limited by lack of suitable drainage materials in the local area.

Alternate C: thick geotextile drain (Fig. 18.7)

Advantages
1. Thick geotextiles are readily available and can be shipped to any
location.
2. Specialty geotextiles can be fabricated for site-specific needs, e.g.,
high-denier fibers and composite geotextiles needlepunched
together.
3. Lightweight installation equipment can be used; risk of damaging
underlying liner during installation of drainage layer is minimal.
4. Thick, nonwoven geotextile provides good puncture protection for
underlying barrier layer.
5. Few concerns with stability of drainage layer on steep slopes.
6. Geotextiles can be anchored, providing stability for covers on steep
slopes.

Disadvantages
1. Thin drainage layer does not help very much to protect underlying
materials from freeze/thaw effects.
2. May have inadequate transmissivity to provide drainage in high
precipitation climates.

Assessment
This is a conventional design for a drainage layer that is placed directly
on a geomembrane liner. If suitable drainage materials are not available
locally, this is usually one of the best design alternatives for a drainage
layer. The design can also be used for a drainage layer placed directly
on a compacted clay liner.
 ponents
s o j cover system com 469
Detaz7

xtile drain
Thick geote

1 ~ t f : : ~ : ~ : ; : : k i , { t t t . t 1 ; 1 % ; f ~ i ~ ~ 1 * 1 { *
~i:~N{i . \ ~ ~ ~ ~ ~ age~yer
GeonetdrIDn

18.8)
K eo /ex/ H e filter (Fig.
/h
net dra;n rm
Il l/ """Ie D: Keo
Advantag
es ipped to
a b le a n d can be Sh
adily ava il
lo c a tio n
s .
a n d g o o nets a re re r is k o f dama
ging
a n yo !e x ti Je e se d ;
ent can b
I. Ge U al.
s ta ll a ti on equipm in a g e la yer is minim
ra
llation o f d
e ig h t in
2. U g h tw e r d u ri n g insta
in g li n
underly tages derlying
Disadvan m u c h to protect un
t hecetslp. very
rzed/toheaswn oeff o-
in d ra infraogm e la
fr y
e e
e te rl a c e ; b o n d e d ge
I. mThaterials /geonet in must
e a lo n g gootextile n n s ta b il ity o f bond
l slippa g g-t e d to
2. Potentia re p re fe rr ed b u t lon
n e t) ca n be anchore I

eonets a d to geo
I
texlilesetostgabilG itey . texliles (bonde
o te rlace, if u
nder- I

im p r o v
s u re d . r a n e in I
be as geomemb e. I

a lo ng goonet/ o n th e u p p e r surlac I

o te n ti a l s li p p a g e
o m e m b ra n e
ig h lf o rw a rd a n d the I
3. P g e is n o t s tr a
COntains a
I
s.
lYing liner n g slopes d via laboratory test
I
n e ts o n lo s e I
g o f geo be asses
4. Seamin d connection m u s t
I
I
intend n te
directly
I
Assessme at is plaCed ilable
I
g e la y e r th I
r a draina re n o t ava
l design fo materials a
I

n v e n tio n a d ra in a e e I

This is a C
O f Suitable I

" " n " '> n ln ane liner. I I

on a " I
I
I
I
I
I
I
470 Cover systems

Fig. 18.9 Geocomposite drain in a cover system.

locally, this is often the best design alternative for a drainage layer. The
design can also be used for a drainage layer placed directly on a
compacted clay liner, but a sand or geotextile filter would have to
separate the geonet from the compacted clay liner.

Alternate E: geocomposite drain (Fig. 18.9)

Advantages
1. Geocomposite drainage materials are readily available and can be
shipped to any location.
2. Lightweight installation equipment can be used; risk of damaging
underlying liner during installation of drainage layer is minimized.
3. Can be used on top of a compacted clay liner without need for an
additional separation layer.
4. Connection of geocomposite cores is generally via a male/female
interlock, which is very straightforward.

Disadvantages
1. Thin drainage layer does not help very much to protect underlying
materials from freeze/thaw effects.
2. Potential slippage along geotextile/geocomposite interface. Bonded
geotextiles to geocomposites are preferred, but long-term stability of
bond must be assured.
3. Potential slippage along geocomposite/geomembrane interface, if
underlying liner contains a geomembrane on the upper surface.
Geotextile (bonded to geocomposite) can be anchored to assist
stability.

Assessment
This design has not been used as frequently as a geonet drainage layer
and geotextile filter. Technically, however, the geocomposite drainage
layer can have much greater transmissivity, if required, than the
geonet/geotextile combination.
 Details of cover system components 471

.18.3.4 Barrier layer

The barrier layer is often viewed as the most critical engineered com-
ponent of the final cover system. The barrier layer minimizes percola-
tion of water through the cover system directly by impeding infiltration
through it and indirectly by promoting storage or drainage of water in
the overlying layers and eventual removal of water by runoff, evapo-
transpiration, or internal drainage.
Historically, a compacted clay liner with a thickness of approximately
0.6 m and a design hydraulic conductivity of 1 x 10- 7 cm/s has been
the most commonly used barrier layer. This usage probably reflects the
historic use of clayey materials for civil engineering projects requiring
a layer that impedes water infiltration. Typically, the compacted clay
liner has been used without a geomembrane liner and has been
covered with a relatively thin (150-600 mm) layer of cover soil. For
modern hazardous waste disposal facilities, however, a composite liner
consisting of geomembrane/compacted clay is common in the US, but
otherwise (for MSW facilities) a single compacted clay liner (CCL) is the
dominant barrier material used in cover systems.
It is unfortunate that compacted clay has been so widely and indis-
criminately used in cover systems because there are several problems
which make the long-term performance of a compacted clay liner ques-
tionable as a barrier layer in many cover systems. The problems with
clay liners used in cover systems include:
• clay liners are difficult to compact properly on a gas collection layer
or a soft foundation (e.g., waste);
• compacted clay will tend to desiccate from above and/or below and
crack unless protected adequately;
• compacted clay is vulnerable to damage from freezing and must be
protected from freezing by a suitably thick layer of cover soil;
• differential settlement of underlying compressible waste will crack
compacted clay if tensile strains in the clay become excessive; and
• compacted clay liners are difficult to repair if they are damaged.
In addition, a compacted clay liner is not as effective as a composite
liner in impeding movement of liquid through the liner (Chapter 5).
Some designers have assumed that a compacted clay liner can be
protected from desiccation with a thin layer of cover soil. Montgomery
and Parsons (1989) describe field experiments in which three test plots
were constructed and monitored for three years. A 1.2-m thick layer of
compacted clay was covered with either 150 mm or 450 mm of topsoil at
a temperate site in Wisconsin. After three years, excavations were
made into the compacted clay. The condition of the two plots with 150
or 450mm of top soil was about the same. The upper 200-250mm of
clay was weathered and blocky, cracks up to 12mm wide extended up
472 Cover systems

to 1 m into the clay, roots penetrated up to 250 mm into the clay in a

continuous mat, and some roots extended up to 750 mm into the clay.
Clearly, neither 150 mm nor 450 mm of topsoil was enough to protect
the clay adequately.
Corser and Cranston (1991) describe test plots in which a layer of
compacted clay was covered with:
1. a 600 mm layer of topsoil;
2. an unprotected geomembrane; or
3. a geomembrane overlain by 450 mm of topsoil.
The site was located in a relatively arid part of California. In less than
a year, significant drying and cracking occurred in the plots with soil
cover alone and the geomembrane cover alone, but no significant
desiccation occurred in the clay covered with both a geomembrane and
soil. However, the tests were short-term tests, and long-term drying
may have eventually occurred even with the geomembrane and soil
overburden. Nevertheless, the studies of Montgomery and Parsons
(1989) and Corser and Cranston (1991), taken collectively, illustrate that
the best and perhaps only practical way to protect a compacted clay
liner from desiccation from the surface is to cover the clay liner with
both a geomembrane and a layer of cover soil. To provide less protec-
tion is inappropriate if the designer's intention is for the compacted
clay liner to remain intact and not crack. A geomembrane may not be
necessary if an extremely thick layer of cover soil is used, but the layer
would have to be so thick that for most projects soil alone would be
impractical. Extremely thick layers of protective soil are only practical
for unusual wastes, e.g., radioactive wastes.
Although designers have sometimes considered desiccation from
above in the design of clay liners in cover systems, it appears that few
designers have given careful thought to the potential for the clay liner
to dry from below. Compacted clay liners (CCLs) are usually placed at
water contents wet of optimum and with soil water potentials of 0 to
-1 bar. The equilibrium relative humidity of the air phase in soils with
water potentials >-1 bar is in excess of 99.5%. If the relative humidity
of the gas phase in the underlying material (e.g., soil, final daily cover,
gas collection layer, or the waste itself) is less than 95%, the clay liner
will eventually dry and, depending on the amount of drying, possibly
crack. Even landfill gas may possess the capacity to dry a CCL, and dry
soils or waste materials certainly possess that capability. A geomem-
brane placed beneath the CCL is probably needed if there is potential
for drying from below. However, placement and compaction of CCLs
on top of a geomembrane is tricky; one cannot use extremely heavy
compaction equipment because the geomembrane could be damaged.
Yet another problem with CCLs is potential damage from differential
settlement. It is convenient to define distortion as the differential
 Details of cover system components 473

settlement d that occurs over a distance L (see Fig. lS.10) . Distortion in

a cover stretches the liner and, as a result, produces tensile strains
in the cover components. Murphy and Gilbert (1985) computed the
relationship between distortion and tensile strain for cover components,
and this relationship is shown in Fig. lS.11. Murphy and Gilbert also
summarized available data on the maximum tensile strain that com-
pacted clays can sustain and found that compacted clays cannot sustain
tensile strains larger than 0.1-1 %. From Fig. 18.11, the maximum
distortion !lIL associated with a tensile strain of 0.1-1 % is 0.05-0.l.
Jessberger and Stone (1991) performed centrifuge tests on 35- to 45-
mm-thick prototype clay liners that would be used for cover systems.
(a)

(b)

Fig. 18.10 Definition of distortion = ~/L: (a) before settlement and (b) after
settlement.
1.0

0.8

0.6
...J
<i
0.4

0.2

0 .0
.1 10 100

Tensile strain (%)

Fig. 18.11 Relationship between ~/L distortion and tensile strain in a cover
system, from Murphy and Gilbert (1985) .
474 Cover systems
A 'trap door' beneath the liner was deformed to produce settlement
like that which would be expected in a landfill. Flow rates through the
soil were measured as a function of the distortion in the liner. It was
found that the flow rate through the clay remained low until the
distortion reached ML = 0.1, at which point the flow rate increased
dramatically. Tension cracks were noted in later examinations of the
material. Thus, the experiments reported by Jessberger and Stone
support the findings from Murphy and Gilbert (1985), i.e., eeLs in
cover systems cannot withstand lJ./L distortions greater than approxi-
mately 0.05-0.1 without cracking.
What does this level of distortion mean in practical terms? Suppose
that one observes a circular crater in a landfill cover with a diameter of
6m. What is the maximum settlement at the center of the crater before
significant cracking would be expected in the liner? The horizontal
distance from the edge to center of the crater, L, is 3m, and 0.05-0.1
times L yields a maximum allowable settlement (lJ.) of 0.15-0.3 m. It is
the authors' experience that many, if not alt covers for municipal solid
waste landfills have areas with distortions of this magnitude or larger.
In such cases, the authors would argue that a eeL is an inappropriate
material for a cover system. In defense of eeLs for hazardous waste
disposat cover systems for most modern hazardous waste landfills in
the US probably experience lJ./L distortions less than 0.05 and differ-
ential settlement is not necessarily a problem. Also, for final covers
placed over contaminated soil or other relatively incompressible
materiat differential settlement may not be a problem.
In contrast to the sensitivity of eeLs to cracking caused by differ-
ential settlement, geomembranes can withstand comparatively large
tensile strains, even when stressed three-dimensionally. Indeed, a
geomembrane suffers from few of the problems of a eeL in a cover
system; the geomembrane can withstand large differential settlement
and is not vulnerable to damage from desiccation or freeze-thaw. A
single geomembrane liner is a lot more sensible liner component for
most municipal solid waste cover systems than a single compacted clay
liner.
There are actually seven options for the barrier layer component of a
cover system that a designer should consider. They consist of eeLs,
GMs, GeLs, or combinations thereof:
• a single compacted clay liner, or eeL (Fig. 18.12a);
• a single geomembrane, or GM, liner (Fig. 18.12b);
• a single geosynthetic clay liner, or GeL (Fig. 18.12c);
• a 2-component composite GM/eeL (Fig. 18.12d);
• a 2-component composite GM/GeL (Fig. 18.12e);
• a 3-component composite liner consisting of GM/GeL/GM (Fig.
18.12f); and
 Details of cover system components 475

• a 3-component composite liner consisting of GM/GCL/GM (Fig.

18.12g).
The advantages and disadvantages of each of the seven options are
described below and an assessment of each is given, as well. Further
comments about the most desirable of these seven options are given
later in this chapter.

Alternate A: single compacted clay liner (Fig. 18.12a)

Advantages
1. Convenient construction if clay is available locally.
2. If clays are not locally available, local soils can be blended with
imported, processed clay, e.g., bentonite.
3. Conventional method and material with long history of use.

Disadvantages
1. Clay may dry out from below and crack (dry waste or gas).
2. Clay may dry out from above and crack (arid climate, drought
conditions, or thin cover soil).
3. Tension cracks may form as a result of differential settlement.
4. Freeze/thaw may damage liner (cold regions, thin soil cover).
5. Limited hydraulic effectiveness due to lack of composite action.
6. Clay may be difficult to compact above compressible waste.
7. Clay may not be available locally.
8. Liner is difficult to repair if cracked.
9. Stability concerns for steep slopes.

Assessment
The main problems with this liner are that it almost certainly will be
damaged by wet/dry cycles and (depending on site conditions) may
also be damaged by settlement and/or freeze/thaw. The only way to
prevent wet/dry and freeze/thaw damage is to use an extremely (and
often impractically) thick layer of cover soil. Problems with settlement-
induced cracking and drying from below often make this design
unacceptable no matter how thick the cover soil is. A single compacted
clay liner is not recommended for most cover systems and should only
be considered where there will be a very thick cover soil, no significant
differential settlement, and no drying from below.

Alternate B: single geomembrane liner (Fig. 18.12b)

Advantages
1. Straightforward and rapid installation.
2. Materials are readily available and can be shipped to any location.
 (a)

ion
Gas cOllect
layer

; f ~
I ~ ~il ~t}~ i ~ j ~
f ~
(b)

G'~i'~ ; i\Y;' ;\!*;~1~t:'~ ~ ~ ~l ; : ;~ ~ ! ~ n

Gas COlleclio
layer

(C)

ion
Gas collecl
layer

(d)

n
Gas collectio
laYer
 )
rane (GM
(e) Geomemb

Filter

clron
Gas COl/e
layer

ion
Gas cOllect
layer

(g) s)
ranes (G M
Geomemb
lrc
Geosynthe e l)
Clay liner (G

" d day
t, (a) 'i n g le " 'm p a " ,, ; (d)
en tic d a y li
n
N ,ompon
n ' fo , the b a 'r ie , laY (c) '; n g le geo'ynthe t, g ,o m 'm b 'a n ,1
O p ti o /in,,; compo ,f li n " l
Fig. 18.12 n g le g e o m e m b " n e ted d a y li n ,, ; (e) 'lcompacted d a y li n " l
) 'i o m p '< b ,a n a y
ti n ,, ; (b
geOm'mb,a
n'lc geom'm
oml'O'fte membmn'/g""Ynth
etic d
" 'm p O '; te c clay li n ,, ; (f) ecm . t, g ,o
geo'Ynthber,a
ti ompo,f
r sycst
, ina nadcov(ge)
ggeeoommem 'm b anne"
478 Cover systems

3. Liner not damaged by wet/dry cycles; does not require a thick cover
soil to prevent desiccation of liner.
4. Geomembrane and their seams do no appear to be affected by
freeze/thaw cycling.
5. Some geomembranes can withstand very large differential settlement.
6. Easy to repair.
7. Gas collection penetrations are easy to construct (connection with a
pipe boot).
8. Geomembrane can be anchored if stability is of concern.

Disadvantages
1. Potential slippage at interface between geomembrane and cover soil.
2. Tensile stresses may be mobilized in anchored geomembrane.
3. Vulnerable to puncture.
4. Limited hydraulic effectiveness due to lack of composite action.
5. Type and placement of cover soils is critical to minimize punctures.

Assessment
The main problem with this liner is that a geomembrane can be
punctured during construction and later from accidental intrusion. This
liner is only recommended when no other liner is practical and where
some leakage through the liner can be tolerated.

Alternate C: single geosynthetic clay liner (Fig. 18.12c)

Advantages
1. Straightforward and rapid installation.
2. Materials are readily available and can be shipped to any location.
3. Geosynthetic clay liner possesses some self-healing capability from
minor punctures, desiccation, and freeze/thaw (better in this regard
than a compacted clay liner).
4. Gas collection penetrations are easy to construct.
5. Material is dry when placed and can be constructed in dry con-
ditions without desiccation damage.
6. Geosynthetic clay liner can be easily repaired by patching if
damaged.
7. Geosynthetic clay liner can be anchored if stability is of concern.
8. Geosynthetic clay liner can be reinforced if tensile stresses are of
concern.

Disadvantages
1. Low shear strength of hydrated clay at mid-plane of some geo-
synthetic clay liners.
2. Potential slippage at interface between GeL and underlying filter.
3. Vulnerable to puncture.
 Details of cover system components 479

4. Limited hydraulic effectiveness due to lack of composite action.

5. Placement of cover soil is critical to minimize punctures.
6. Dry GCL will not serve as a barrier to gas migration.
7. Differential settlement can cause clay shifting that could jeopardize
low permeability of liner.
8. Alternating wet/dry cycles may cause shrinkage of GCL, resulting in
leakage until rehydration and swelling reoccurs.

Assessment
Permeability of dry GCL to gas makes this liner alternative unaccept-
able for wastes that produce gas, unless one can be assured that the
clay will be wetted soon after installation and will remain wet there-
after. The geosynthetic clay liner may be punctured during construction
or after installation from accidental intrusion. This liner is only recom-
mended for special cases where some leakage through the liner can be
tolerated.

Alternate D: composite geomembrane/compacted clay liner (Fig. 18.12d)

Advantages
1. Composite liner very effective as a hydraulic barrier.
2. Minimum liner required by some regulatory agencies.
3. Geomembrane protects clay liner from desiccation if cover soil
drys; does not require a thick cover soil to prevent desiccation of
clay.
4. Thick liner cannot be punctured.
5. Geomembrane component can tolerate large differential settlement.
6. Geomembrane can be anchored to provide stability.

Disadvantages
1. Potential slippage along both soil/geomembrane interfaces.
2. Anchored geomembrane can cause tensile stresses to develop in the
geomembrane.
3. Clay may dry from below and crack (dry waste or gas).
4. Tension cracks may form in clay as a result of differential settlement.
5. Freeze/thaw may damage clay liner (cold regions, thin soil cover).
6. Clay may be difficult to compact above compressible waste.
7. Clay may not be available locally.
8. Clay may not be easily repaired if damaged.

Assessment
A geomembrane/compacted clay liner is a common minimum require-
ment by various regulatory agencies for certain categories of waste.
The primary concerns for this type of liner are with the long-term
performance of the compacted clay component. Large differential
480 Cover systems

settlement and drying from below can cause the compacted clay to
crack and increase its permeability. However, if the geomembrane
remains largely intact and the permeability of the compacted clay does
not increase excessively, the composite liner would still be expected to
be reasonably effective.

Alternate E: composite geomembrane/geosynthetic clay liner (Fig. lS.12e)

Advantages
1. Relatively straightforward and rapid construction.
2. Materials are readily available and can be shipped to any location.
3. GCL possess some self-healing capability from minor punctures,
desiccation, and freeze/thaw.
4. GCL is dry when placed and can be constructed in dry conditions
without desiccation damage.
5. Liner can be repaired fairly easily.
6. Composite liner very effective as a hydraulic barrier.
7. Geomembrane component can tolerate large differential settlement.
S. GCL can be anchored if stability is a concern.
9. GCL can be reinforced if tensile stresses are of concern.

Disadvantages
1. Composite action questionable for most GCLs with upper geotextile.
2. Potential slippage along upper and lower surface of geomembrane
and between some types of GCL and filter.
3. Low shear strength of hydrated clay at mid-plane of GCL for some
products.
4. Vulnerable to puncture.
5. Placement of cover soil is critical to minimize punctures.
6. Differential settlement can cause clay deformation that could jeo-
pardize low permeability of GCL in certain locations.

Assessment
This liner system offers redundancy and many practical benefits in
terms of availability of materials and ease of construction. Questions
can be raised about effectiveness of composite action between geo-
membrane and those GCLs with an upper geotextile - this is an area of
research at this time. Provided that issues concerning stability and
puncture are resolved, this system probably offers the best overall
benefit/cost ratio of any liner type for cover systems.
 Details of cover system components 481

Alternate F: composite geomembraneicompacted clay liner/geomembrane

(Fig. 18.12f)
Advantages
1. Composite behavior of GM/CCLIGM extremely effective as a
hydraulic barrier; with two GMs, some cracking of clay from differ-
ential settlement can be tolerated without hurting the overall
hydraulic performance.
2. Geomembranes protect clay liner from desiccation from above or
below; does not require a thick soil cover to prevent desiccation of
the clay.
3. Thick clay liner cannot be punctured.
4. Geomembranes can tolerate large differential settlement.

Disadvantages
1. Difficult construction (especially compaction of clay on GM and
separate anchorage systems for geomembranes).
2. Potential slippage along several interfaces or within CCL (for which
excess pore water pressure cannot dissipate).
3. Freeze/thaw may damage clay liner (cold regions, thin soil cover).
4. Clay may not be available locally.
5. Clay not easily repaired if damaged.
6. Very complicated gas collection penetrations.

Assessment
This is a very conservative design which offers excellent redundancy of
components and, more than any other design involving compacted
clay, provides layers that protect the clay from damage due to desic-
cation. By keeping the clay wet, the protective geomembranes indirectly
help to improve the resistance of the clay to settlement-induced
cracking. However, this design is rarely (if ever) used mainly because
construction is extremely complex and no regulatory agency has
required that a geomembrane be placed below the compacted clay.
This design is only recommended for extremely sophisticated waste
containment facilities where an extraordinary degree of redundancy in
the design is required.

Alternate G: composite geomembrane/geosynthetic clay liner/geomembrane

(Fig. 18.12g)
Advantages
1. Relatively straightforward and rapid construction.
2. Materials are readily available and can be shipped to any location.
3. Composite behavior fairly good, even with imperfect GM/clay
482 Cover systems
contact or some settlement-induced clay deformation, because two
geomembranes are used.
4. Geomembranes protect clay from wet/dry cycles.
5. Geomembranes can tolerate large differential settlement.
6. GeL is dry when placed and will not dry and crack during
installation.
7. Liner can be repaired fairly easily.
8. Anchorage of all components is possible if stability concerns are
present.

Disadvantages
1. Potential slippage along several interfaces and through mid-plane of
some GeLs if GeL becomes fully hydrated.
2. Vulnerable to puncture.
3. Placement of cover soil is critical to minimize punctures.
4. Very complicated gas collection penetrations.

Assessment
This is a very conservative design that offers excellent redundancy of
components and practical advantages such as availability and overall
ease of construction (except gas collection penetrations). Stability can
be assured by anchorage of components and use of high friction
surface on top of upper geomembrane. Provided potential problems
with puncture can be resolved, this liner probably offers the best
combination of extremely low percolation of water through the liner
and practical constructability.

18.3.5 Gas collection layer

The purpose of a gas collection layer is to collect gas for processing or
discharge. Gas flows from the gas collection layer to vents. Flow of gas
may occur passively under the pressure gradient generated naturally or
actively through assistance from a vacuum pump at the surface. A gas
collection layer is only necessary for capping wastes that can produce
gas or volatiles that must be released in a controlled manner.
The gas collection layer must have a high in-plane gas flow rate and
must not become plugged with fine-grained materials. The usual
materials are:
• Sand or gravel with a soil filter (Fig. 18.13a);
• A thick geotextile which serves as both drain and filter (Fig. 18.13b);
• Geonet drain and a geotextile filter (Fig. 18. 13c).
Any of these materials will function well if properly designed, and the
advantages/disadvantages of each is considered below.
 nents
o f cover system compo 483
Details
(a)

(b)

(e)

el d " in
y e " (' J " n d 0 " g,-g'veoheXli/e
I' l d " in 'n d
co1/ection
n ,, , u 'e d in Ihe g od, filte" 'n d (eJ geone
M 'l e stem . d " in 'n
F;g. 1 8 ."fi a cover sygeole"i1e
il IIe,; (bJ
fiwltilerhs,'oin
18. 13a)
w ith 'o d Jiller (Fig.
A: " 'o d n ,
gravel dmin
A II
e s" n n le
De) drain
s in
Advantag n d o r gra
and "
orkf s.
U 'e o f 'o il fi lt e "
inogr yw o
eeri"
1. eLnogning h
484 Cover systems

2. Easy to install perforated collection pipes in sand or gravel gas

collection layer.
3. Base of vent pipes are easy to locate between filter and bottom of
gas collection layer.

Disadvantages
1. Suitable materials may not be readily available locally.
2. Thickness of both components decreases air space within landfill.

Assessment
This is a technically sound design and probably the most common
design when suitable construction materials are available locally. The
major drawback is the loss of air space in the landfill facility.

Alternate B: geotextile drain and filter (Fig. IS.13b)

Advantages
1. Geotextiles are readily available and can be shipped to any location.
2. Virtually no airspace in landfill facility taken by gas collection layer.
3. Provides clean working surface if geosynthetics are used as lower
portion of barrier layer.
4. Rapid, straightforward installation.
5. Specialty geotextiles can be fabricated for site-specific designs, e.g.,
high denier fibers and composite geotextiles needled together.

Disadvantages
Potential slippage between geotextile and overlying geomembrane (if
the barrier layer has a geomembrane on the lower surface of the layer).

Assessment
This design is considered technically equal to designs that employ sand
filter and gravel drainage materials. Selection of natural materials or
geosynthetics is often based upon local availability and cost of natural
materials and construction logistics. In many cases, this would be the
least costly of the gas collection layer alternatives.

Alternate C: geonet drain and geotextile filters (Fig. IS.13c)

Advantages
1. Geotextiles and geonet are readily available and can be shipped to
any location.
2. Geonet has extremely high in-plane gas transmissivity.
3. Geotextiles can be bonded to the geonet using a variety of fabrica-
tion methods.
4. Relatively no loss of airspace in the landfill facility.
 Factors in selection of barrier layer alternatives 485

5. Provides clean working surface if geosynthetics are used as lower

portion of barrier layer.

Disadvantages
1. Potential slippage between geotextile and overlying geomembrane
(if the barrier layer has a geomembrane on the lower surface of the
layer) and between geonet and geotextiles.
2. Continuity of flow within the geonet must be assured across seams
and edges.
3. Somewhat more complex connection to vent pipes than a single,
thick geotextile drain and filter.

Assessment
This design is considered technically equal to designs that employ sand
filter and gravel drainage materials. Selection of natural materials or
geosynthetics is often based upon local availability and cost of natural
materials and construction logistics. The geonet has a higher in-plane
gas transmissivity than a geotextile, which could be advantageous if
gas production rates are extremely large or vent pipes are more widely
spaced than usual.

18.4 FACTORS IN SELECTION OF BARRIER

LAYER ALTERNATIVES

The critical factors that affect selection of a barrier layer design are
climate, the amount of differential settlement to which the cover will be
subjected, the vulnerability of the cover soil to erosion or puncture,
the amount of water percolation through the cover system that can
be tolerated, the need for collection of waste-generated gas, and the
steepness of the slope.
The seven barrier layer alternatives discussed previously and shown
in Fig. 18.12 have been assessed by the authors in light of each of these
factors. For each factor and the condition considered, each liner was
assigned a factor ranging from 1 (not acceptable) to 5 (recommended).
For example, a compacted clay liner was given a 1 (not acceptable) for
cases involving large differential settlement whereas a geomembrane
was assigned a factor of 5 (recommended) because the geomembrane
can withstand large differential settlement. The assessments are sum-
marized in Tables 18.1-6. The authors suggest that designers consider
these assessments carefully for the unique conditions that exist at
a particular site. There may be extenuating circumstances or special
design precautions that have to be taken to mitigate problems that
would normally be present; for this reason, the assessments are for
general guidance only and should not be taken too literally.
Table 18.1 Influence of climate on selection of barrier for cover system

alternate liner component general description of climate at the site

--------

arid cyclic humid

(wet/dry seasons)

A Compacted clay liner 1 1 3

(CCL) desiccation cracks desiccation cracks may get some
desiccation cracks
B Geomembrane liner 5 4 4
(GM)
C Geosynthetic clay liner 3 3 4
(GCL) unacceptable if waste unacceptable if waste
produces gas produces gas
D Composite GM/CCL 2 3 4
long-term desiccation possible desiccation
cracking cracking
E Composite GM/GCL 5 4 5
F Composite GM/CCL/GM 4 4 5
G Composite GM/GCL/GM 5 5 5

Scoring system: 1 = not acceptable; 2 = marginal; 3 = possibly OK; 4 = acceptable; and 5 = recommended.
Table 18.2 Influence of differential settlement on selection of barrier for cover system

alternate liner component amount of differential settlement

major moderate nominal

A Compacted clay liner 1 1 3

(CCL) cracking of clay cracking of clay depends on site
specific factors
B Geomembrane liner 4 5 5
(GM)
C Geosynthetic clay liner 2 3 4
(GCL) lean areas of clay may depends on severity
develop of settlement
D Composite GM/CCL 2 3 4
cracking of clay depends on severity
damage to GM of settlement
E Composite GM/GCL 3 4 5
depends on severity
of settlement
F Composite GM/CCL/GM 3 4 5
cracking of clay
G Composite GM/GCL/GM 4 5 5

Scoring system: 1 = not acceptable; 2 = marginal; 3 = possibly OK; 4 = acceptable; and 5 = recommended.
Table 18.3 Influence of cover soil on selection of barrier for cover system

alternate liner component vulnerablity of cover soil to damage

can be eroded or erosion possible but not can not be eroded or

punctured likely punctured
-------- --------

A Compacted clay liner 1 2 3

(CCL) desiccation of clay desiccation of clay may get some
desiccation cracks
B Geomembrane liner 1 1 3
(GM) UV degradation and UV degradation and depends on other
puncture puncture site specific
factors
C Geosynthetic clay liner 1 1 3
(GCL) puncture puncture depends on other
site specific
factors
D Composite GM/CCL 3 4 4
possible damage to
GM
E Composite GM/GCL 2 3 4
puncture depends on site
specific factors
F Composite GM/CCL/GM 4 5 5
G Composite GM/GCL/GM 4 5 5

Scoring system: 1 = not acceptable; 2 = marginal; 3 = possibly OK; 4 = acceptable; and 5 = recommended.
Table 18.4 Influence of percolation limits on selection of barrier for cover system

alternate liner component allowable water percolation through barrier

------~-~

essentially none very little moderate amount

~~~-~ ----.----~---

A Compacted clay liner 1 2 3

(CCL) not a composite depends on site
specific factors
B Geomembrane liner 1 3 5
(GM) not a composite depends on site
specific factors
C Geosynthetic clay liner 1 2 3
(GCL) not a composite
D Composite GM/CCL 3 4 4
depends on site
specific factors
E Composite GM/GCL 3 4 5
depends on site
specific factors
F Composite GM/CCL/GM 5 5 5
G Composite GM/GCL/GM 5 5 5

Scoring system: 1 = not acceptable; 2 = marginal; 3 = possibly OK; 4 = acceptable; and 5 = recommended.
490 Cover systems
Table 18.5 Influence of gas generation on selection of barrier for cover system

alternate liner component need for a gas collection system

is required not required

-------

A Compacted clay liner 1 1

(CCL) clay drying cracking of clay
B Geomembrane liner 4 5
(GM)
C Geosynthetic clay liner 1 5
(GCL) gas escapes when dry
D Composite GM/CCL 3 5
depends on potential
for clay drying
E Composite GM/GCL 4 5
F Composite GM/CCL/GM 4 5
vent connection detail
complicated
G Composite GM/GCL/GM 4 5
vent connection detail
complicated

Scoring system: 1 = not acceptable; 2 = marginal; 3 = possibly OK; 4 = acceptable; and

5 = recommended.

18.5 GENERAL ASSESSMENT AND COMPARATIVE RANKING

The previous section presented charts of a number of important

landfill cover design factors insofar as the liner, or barrier, system is
concerned. While the rankings of the different liner sections within
each chart are admittedly subjective, it is of interest to now extend
the numerical assessments into a summary and then to contrast the
resulting values to a general cost estimate. A generalized benefit/cost
ratio will be computed for each of the seven alternate liner schemes.
Critically important in such a ranking is the cost of the various schemes.
The following estimated costs will be used .
• Compacted clay liners (CCLs). This is the most difficult element for a
generalized cost estimation. In-place costs of CCLs typically vary
from $5.40 to $21.50 per square meter depending on thickness,
availability, size and type of facility. In extreme conditions (e.g., lack
of clay locally), the cost can be much higher. However, for a 0.6m
thick compacted clay liner of 1 X 10- 7 em/sec hydraulic conductivity,
$7.50 per square meter will be the estimated installed cost .
• Geosynthetic clay liners (GCLs). Of the three commercially available
Table 18.6 Influence of slope inclination on selection of barrier for cover system

alternate liner component steepness of slope

flatter than 6: 1 6:1-3:1 steeper than 3: 1

«9°) (9-18°) (>18°)

A Compacted clay liner 5 4 3

(CCL) depends on clay
properties and
slope angle
B Geomembrane liner 5 5 3
(GM) depends on
material
properties and
slope angle
C Geosynthetic clay liner 4 3 3
(GCL) depends on specific depends on GCL
GCL material material and
slope angle
D Composite GM/CCL 5 3 2
interfacial shear interfacial shear;
possible internal
shear of CCL
E Composite GM/GCL 5 3 2
depends on specific interfacial shear;
GCL material internal shear
for some GCLs
F Composite GM/CCL/GM 5 2 1
interfacial shear; no dissipation
depends on GM of pore water
and CeL properties pressure in clay
G Composite GM/GCL/GM 5 3 2
depends on specific interfacial shear;
GCL material internal shear
for some GCLs

Scoring system: 1 = not acceptable; 2 = marginal; 3 = possibly OK; 4 = acceptable; and 5 = recommended.
Table 18.7 Summary of influence of individual factors on barrier layer for covers
alternate liner climate settlement cover erosion/ allowable pcrcolatiOll gas collection slope
component puncture vulnerability inclinatioll
- - - - - - - ------ ----------- -~--~---~- -_._------ -

arid cyclic humid major mod. nominal major mod. low ess. 1WIle v. little mod. gas 110 gas <9° 9-18° >18°
---- --------~-----~----
- - - - - - - - ----- ----------- -----------------------_.,-----
A CCL 3 3 2 3 2 3 5 4 3
B GM 5 4 4 4 5 5 3 3 5 4 5 5 5 3
C GCL 3 3 4 2 3 4 3 2 3 5 4 3 3
0 GM/CCL 2 3 4 2 3 4 3 4 4 3 4 4 3 5 5 3 2
E GM/GCL 5 4 5 3 4 5 2 3 4 3 4 5 4 5 5 3 2
F GM/CCL/GM 4 4 5 3 4 5 4 5 5 5 5 5 4 5 5 2
G GM/GCL/GM 5 5 5 4 5 5 4 5 5 5 5 5 4 5 5 3 2

Scoring system: 1 = not acceptable; 2 = marginal; 3 = possibly OK (depends on site or product-specific conditions); 4 = acceptable; and 5 = recommended.
 General assessment and comparative ranking 493

GeLs in the US, installed costs vary from $5.40 to 10.80 per square
meter depending on site conditions. A cost of $7.50 per square meter
is assumed .
• Geomembranes (GMs). The range of installed costs for geomem-
branes of the type generally used in cover systems is relatively
narrow in comparison to the other liner materials. For geomem-
branes with good out-of-plane deformation properties and a thick-
ness of approximately 1.0 mm, the installed cost probably averages
about $7.50 per square meter.
Thus, it is seen that each of the three liner components is assumed
to cost approximately $7.50 per square meter installed. Those liner
schemes with two liner elements within them will cost twice those of a
single liner element. Those with three components will cost three times
that of a single component liner.
The benefit/cost ratio was computed by the authors as follows. The
numerical rankings given in Tables 18.1-6 were tabulated into a single
chart (Table 18.7). For each of the seven liner alternates, the numerical
ratings in each horizontal row were summed to obtain an overall
numerical 'benefit' for each alternative. Those liners with the highest
overall ratings for all of the conditions considered scored highest. The
'benefit' number calculated by summing numbers in horizontal rows in
Table 18.7 was then divided by the cost (in dollars per square meter) to
determine the benefit/cost ratio. It is realized that this procedure is
arbitrary, but it should give a good overall picture of which liner
alternatives provide the best performance over a broad range of site
conditions.
Table 18.8 lists results of the benefit/cost computations. The liner
alternatives are grouped into one-, two- or three-layer systems. As a
reminder of the systems: one-layer systems consist of a GM, eeL, or
GeL; two-layer systems consist of GM/eeL and GM/GeL composites;
and three-layer systems consist of GM/eeL/GM and GM/GeL/GM
systems.
From Table 18.8, it can be seen that among the single-layer systems,
the single geomembrane (GM) far outperforms the geosynthetic clay
liner (GeL) in terms of benefit/cost ratio, and that both of these geo-
synthetic materials outperform a compacted clay liner (eeL) in terms
of benefit/cost ratio. The compacted clay liner is the poorest overall
technical choice of any single layer system and is also the scheme with
the poorest benefit/cost ratio that one can utilize as a cover system's
hydraulic barrier. Paradoxically, the eeL is the most widely used and
permitted single liner material in use for current cover systems!
Of the two-layer (composite) systems, the GM/GeL outperforms the
GM/eeL technically and on a benefit/cost ratio basis. This finding is
also important since a composite liner is usually considered to be
494 Cover systems

Table 18.8 Benefit/cost assessment of various liner cross-sections

no of design description overall est. cost benefit!cost ranking
barrier layers alternate benefit* $/m 2 ratio in group
------- ------- -------

One layer A CCL 34 0.065 49 3

B GM 63 0.065 90 1
C GCL 46 0.065 66 2
Two layers D GM/eCL 58 0.13 41 2
E GM/GCL 66 0.13 47 1
Three layers F GM/CCL/GM 72 0.195 34 2
G GM/GCL/GM 77 0.195 37 1

* Determined by summing horizontal rows in Table 1S.7.

a GM/eeL combination. The finding of this analysis is that, for

cover systems, a GM/GeL is preferable to GM/eCL in terms of overall
assessment.
Considering the triple layer systems, the GM/GCL/GM slightly out-
performs a GM/CCL/GM system. However, the difference is relatively
small.
In summary, it is obvious that designers must rethink their current
cover system liner strategy. Compacted clay liners (CCLs) are not the
general answer for the hydraulic barrier in the closure of a solid waste
landfill or site remediation project. With desiccation from above the
liner in arid or seasonally dry climates, drying of the liner from beneath,
and cracking from differential settlement, CCLs cannot be expected to
maintain their initially low hydraulic conductivity.
Conversely, the newer barrier systems, namely geomembranes (GMs)
and geosynthetic clay liners (GCLs), should be seriously considered
for cover systems. Geomembranes, in particular, can afford long-term
barrier protection and undergo large differential settlement if required.
Geosynthetic clay liners offer many of these same attractive features.
Furthermore, the combination of a GM/GeL system results in high
technical marks and provides a backup system which is generally
assumed to be a conservative design strategy.
Lastly, if the authors were pressed to recommend the preferred
barrier system currently available in a 'typical' solid waste landfill
cover (for either a new landfill or an old landfill requiring closure or
remediation), the authors would undoubtedly respond: a single geo-
membrane (GM) liner, or if a composite liner is desired, a geomem-
brane over a geosynthetic clay liner, i.e., a GM/GCL.
 Conclusions 495

18.6 CONCLUSIONS

This chapter has summarized the basic principles employed for design
of final cover systems for new waste containment units and site
remediation projects. The most important conclusions relate to the
barrier layer within the final cover. The main conclusions are as
follows:
• compacted clay liners by themselves are not the general barrier
system of choice;
• geomembranes or geosynthetic clay liners are better overall choices
technically and on the basis of benefit/cost ratio than compacted clay
liners; and
• the barrier system of choice for most municipal solid waste landfill
covers should be a geomembrane or a geomembrane over a geo-
synthetic clay liner (GM/GCL).
Compacted clay may have a place in cover systems, but only for
those limited situations where very small differential settlements are
expected and adequate protection from desiccation and freeze/thaw is
provided.
The conclusions listed above are based upon technical, rather than
regulatory, considerations. The authors would be remiss, however,
if regulatory issues were avoided completely. In this regard, RCRA
covers for hazardous wastes and (by inference) municipal solid waste
require a GM/CCL barrier in the cover. Even for site remediation
projects, RCRA requirements may be applicable or relevant and appro-
priate. For an equivalency evaluation of the typical RCRA cross section,
at least three issues must be addressed.
1. The hydraulic equivalency of the regulation-suggested CCL with
respect to a replacement GCL must be demonstrated. This can be
accomplished using Darcy's formula, but such an equivalency cal-
culation does have assumptions, primarily that of no thin areas nor
punctures in the GCL.
2. Some of the GCLs have an upper geotextile with high in-plane
transmissivity which may compromise the 'intimate contact' of the
clay and geomembrane. Work is actively ongoing by GCL manu-
facturers to decrease lateral flow within the upper geotextile to
values similar to the clay itself.
3. Stability issues of geomembranes and hydrated GCLs must be
addressed using proper geotechnical engineering principles.
With proper design and testing, however, the authors feel that regu-
latory concerns can be overcome with the net result of a superior long-
term barrier system for solid-waste landfills and site remediation
projects than are currently designed and constructed.
496 Cover systems

While it must be emphasized that considerable engineering design

and performance testing is required, the construction quality control
and quality assurance (CQClCQA) issues are also essential. The best of
design schemes and manufactured materials can be completely negated
with poor CQC and the attendant CQA activities. Along with the
necessity of proper placement, positioning and seaming of the GM and
GCL materials, there are also barrier penetrations (such as gas vents
that are required). Prefabricated geomembrane pipe boots (with proper
connections and seams and carefully placed clay for CCLs and GCLs)
are mandatory for the overall success of cover systems as proposed
herein. With proper care by all parties involved, however, we feel that
excellent long-term performance of the geosynthetic barrier systems of
the type recommended in this paper will be realized.

REFERENCES

Boschuk, J. (1991) Landfill covers - an engineering perspective, Geotech. Fabrics

Rep., 9(4), 23-34.
Cline, J.F., Cataldo, D.A., Burton, F.G. and Skiens, W.E. (1982) Biobarriers
used in shallow burial ground stabilization, Nuclear Tech., 58, 150-3.
Corser, P. and Cranston, M. (1991) Observations on long-term performance of
composite clay liners and covers, Proc., Geosynthetics Design and Performance,
Vancouver Geotech. Soc., Vancouver, British Columbia.
Gilman, E.F., Flower, F.B. and Leone, I.D. (1985) Standardized procedures for
planting vegetation on completed sanitary landfills, Waste Management and
Research, 3, 65-80.
Lutton, RJ., Gegan, G.L. and Jones, L.W. (1979) Design and construction
of covers for solid waste landfills, US Environmental Protection Agency,
EPA600/2-79-165, Cincinnati, Ohio.
Jessberger, H.L. and Stone, K.J.L. (1991) Subsidence effects on clay barriers,
Geotechnique, 41(2), 185-194.
Johnson, D.1. and Urie, D.H. (1985) Landfill caps: long-term investments in
need of attention, Waste Management and Research, 3, 143-8.
Montgomery, RJ. and Parsons, L.J. (1989) The Omega Hills final cover test
plot study: three year data summary, presented at the 1989 Ann. Meeting
of the Nat. Solid Waste Management Assoc., Washington, DC.
Murphy, W.L. and Gilbert, P.A. (1985) Settlement and cover subsidence of
hazardous waste landfills, US Environmental Protection Agency, EPA/600/2-
851035, Cincinnati, Ohio.
Repa, E.W., Herrmann, J.G., Tokarski, F. and Eades, R.R (1987) Evaluating
asphalt cap effectiveness at a superfund site, J. Environ. Eng., 113(3),
649-53.
Swope, G.L. (1975) Revegetation of landfill sites, MS thesis, Pennsylvania State
Univ., State Park, Pennsylvania.
US Environmental Protection Agency (1989), Technical Guidance Document, Final
Covers on Hazardous Waste LandfillS and Surface Impoundments, EPA/530-SW-
89-047, Washington, DC.
 CHAPTER 19

Recovery well systems

Bob Kent and Perry Mann

19.1 INTRODUCTION

Plato (c.427 -347 Be) developed a sophisticated set of groundwater

laws. In addition to laws on water conservation and water rights, Plato
developed water pollution regulations, as quoted below.
If anyone deliberately spoils someone else's water supply, whether
spring or reservoir, by poisons or excavations or theft, the injured
party should take his case to the City Wardens and submit his
estimate of the damage in writing. Anyone convicted of fouling
water by magic poisons should, in addition to his fine, purify the
spring or reservoir using whatever methods of purification the
regulations of the expounders prescribe as appropriate to the cir-
cumstances and the individual involved.
(Saunders, 1970)
Although we have recognized the need for protection and remediation
of groundwater resources for over 2000 years, major advances in
remediation technology have only been developed in the last 10 years.
Recovery wells are often used to 'capture' a contaminant plume. Once
the contaminated groundwater is recovered, it is treated and released
in an appropriate manner. Methods for recovering and treating con-
taminated groundwater are reviewed in this chapter.

19.2 SOURCES OF GROUNDWATER CONTAMINATION

Contaminants that pose a risk to groundwater originate from waste

disposal practices, such as leachates from landfills and seepage from
waste lagoons. A second category of contaminant is chemicals that

Geotechnical Practice for Waste Disposal.

Edited by David E. Daniel.
Published in 1993 by Chapman & Hall, London. ISBN 0 412 35170 6
498 Recovery well systems

are generally considered beneficial but when improperly managed

or misapplied, can result in contamination of groundwater supplies.
Gasoline from leaking underground storage tanks and fertilizers or
pesticides applied to lawns and crops are examples of this latter
category.
Industrial plants often are major sources of groundwater contami-
nation. Leakage from a variety of sources within an individual facility
results in the subsurface mixing of different contaminants and increases
the difficulty of designing an efficient recovery and treatment system.

19.3 AQUIFER CHARACTERIZATION

Aquifer characterization is critical for determining the number, spacing,

and pumping rate for recovery wells. Groundwater is in constant
motion, flowing from areas of higher head to lower head. The natural
velocity of groundwater flow ranges from <1 m/yr to >100m/day.
Determination of the direction and rate of groundwater flow in each
contaminated aquifer is critical when designing groundwater recovery
systems. If this is neglected, the spacing or pumping rate of wells may
not be sufficient to intercept all of the contaminant plume.
The maximum rate of contaminated groundwater removal depends
on the transmissivity of an aquifer, whereas the volume that can be
removed by a well is dependent in part, on the storativity of the
aquifer. These hydrologic parameters are the major controlling factors
of the shape and size of the cone of depression, which is used to
determine the 'radius of influence', of a pumping well. Transmissivity
and storativity typically are determined by conducting aquifer pumping
tests in the field. Although other methods exist for estimating these
parameters, actual empirical determinations from field pumping tests
are preferred for the design of recovery well systems. The authors are
familiar with several inadequately designed recovery well systems that
were based on transmissivity values derived from short-term slug tests.
A pumping test of 24 or 48 hours is normally recommended for design
consideration. It should be noted that, for high-yield aquifers that are
grossly contaminated, the cost of storage and treatment of the water
recovered during the pumping test may be prohibitive.

19.4 FLOW TO WELLS

Pumping groundwater from a well results in the lowering of the water

level in the well (or 'drawdown') and a decrease in the groundwater
head of the surrounding water-bearing formation. The greatest amount
of drawdown occurs closest to the pumping well and decreases
 Modeling the well field 499

logarithmically with radial distance from the well. The radius of

influence for a well completed in a confined aquifer typically is larger
than that in an unconfined (water table) aquifer with equal trans-
missivities. Water pumped from a confined aquifer is released from
storage by depressurization and slight expansion of the water and
aquifer matrix (shrinkage of pore space) instantaneously and flows
horizontally to the well. The water pumped from an unconfined
aquifer, however, moves vertically by gravity drainage and results in a
much slower release from storage, emptying of pores, and replace-
ment of water with air. When a single well is pumped, a cone of
depression is formed as the radius of influence caused by the dis-
charging well expands. Theoretically, under ideal aquifer conditions,
this radius expands to infinity. Equilibrium is sometimes reached over
an extended period of time, resulting form recharge and/or discharge
boundaries within the aquifer that prevent further expansion of the
cone. The extent of the radius of influence for a single pumping well is
typically between 10 and 1000 m, depending on the specific hydrologic
conditions.
Well interference occurs when the cones of depression (i.e., the
radii of influence) of two or more wells overlap. Interference between
pumping recovery wells, although usually desirable, must be taken
into account in order to maximize the draw down of the aquifer in
a given area. This interference, if significant, may result in lower
pumping yields for individual wells within a multiple well system. As
such, interference of pumping wells is a critical factor to be considered
in the design of a recovery well system. Through careful design and
planning, the efficiency of a recovery well system for collecting con-
taminants in groundwater can be optimized.

19.5 MODELING THE WELL FIELD

Computer models help simplify the process of determining an optimum

number and spacing of recovery wells. Models can be used to simulate
the distribution of draw down within different arrays of wells com-
pleted within a given aquifer. These models are typically based on the
Theis (1935) nonequilibrium equation, which describes drawdown of
water levels in an ideal confined aquifer around a single pumping well.
The effects of multiple pumping wells are then summed to determine
the draw down at any point in the aquifer at any time.
The Theis solution to the problem of radial flow is based on the
following assumptions:
1. the aquifer is homogeneous, isotropic, of uniform thickness, and of
infinite areal extent;
500 Recovery well systems

2. the discharging well is of infinitesimal diameter and completely

penetrates the aquifer;
3. release of water from storage is instantaneous with the reduction in
pressure due to drawdown;
4. flow to the well is radial and horizontal; and
5. the aquifer is confined (i.e., bound by impermeable strata both
above and below).
As with most cases where models are used, many of the assump-
tions listed above will not be true for the actual aquifer. Hydraulic
characteristics of an aquifer are not the same at every location in the
aquifer and the drawdown predictions made by such models are not
exact; that is, the actual drawdowns observed in the field may differ
from those predicted by the model. In addition, the Theis solution is
applicable to artesian aquifers, and most contamination occurs in
semi-confined or water-table aquifers. As such, the solution derived
normally is too optimistic for water-table conditions.
The optimum number of pumping wells and the optimum pumping
rates may be determined by systematically changing the well spacings
to find a spacing and number of wells that can achieve significant
water-level draw downs, and hence, increase the hydraulic gradient of
the groundwater toward the recovery well throughout a plume without
causing excessive drawdown at anyone well. There are numerous
commercially available models to assist an investigator in finding an
optimum number and spacing of recovery wells.
Figure 19.1 is a conceptual diagram of a recovery well system that
was installed around a leaky lagoon near the leading edge of a con-
taminant plume. If the recovery system were located a significant
distance upgradient from the leading edge of the plume, the natural
flow velocity would need to be included in the computer model to
ensure the capture of the plume. Thomas Prickett's random-walk
model is frequently used to determine the influence of natural flow on
the ability of a recovery well to capture a plume.
An error that is often made in using computer programs to predict
drawdowns and well spacings is to assume that the well is 100%
efficient (no head loss at the well) and that maximum drawdown can
be maintained. In thin water-table aquifers, the saturated thickness will
be relatively small. Because the draw down at a recovery well cannot be
greater than the saturated thickness, the radius of capture possible
for a shallow pumping well decreases with saturated thickness. The
pumping rate predicted to maintain the maximum drawdown cannot
always be maintained, because wells are seldom fully efficient and
efficiency may decrease with time as a result of screen plugging. Thus,
while the computer model may predict that a well can pump at a
certain rate, the well may have significantly greater drawdown due to
 Mass and distribution of contaminants to be removed 501

Abandoned
waste lagoon
Groundwater
... treatment unit
Discharge to
surface water

rr=====jj===::::;r8==~
or reinjection

II .1

'I "
,/

""
'.

Original
~!atic water:.~e~ __

Uncontaminated
aquifer
Uncontaminated
aquifer

Fig. 19.1 Typical well recovery system.

head loss than predicted and reduced pumping rates may be required
to prevent the water level in the recovery well from falling to the pump
intake. With a lower pumping rate, the capture radius will not be as
great as predicted by the models. The result will be less production
than predicted and possibly no capture of portions of the contaminant
plume, resulting in possible off-site migration of contaminants.
Another important factor that is often overlooked in designing well
spacing is the hydraulic gradient. Given the same transmissivity and
storage coefficient, an aquifer with a steeper hydraulic gradient requires
closer spacing of recovery wells to prevent off-site plume migration.

19.6 MASS AND DISTRIBUTION OF CONTAMINANTS

TO BE REMOVED

Before treatment methods can be evaluated, the total mass of con-

taminants to be treated should be known. The mass of contaminants in
groundwater can be estimated by mapping concentrations determined
by chemical analysis. Figure 19.2 indicates the concentration of volatile
organic compounds in groundwater at a site in the northeastern United
States. In this example, the following assumptions were made to
estimate the mass of contaminants requiring treatment:
 ·___ I o
o
30 60
RRH
-11--.--SiStmree9"tt
---r-~--_----:;-;:~~~t --_J meters

LEGEND
/1' ~,o~ .:3i
// ~/ ~\ ~~ 1b5 .2 Monitor well
5 - - Concentration contour (ppm)
I ,
I \ --
\ \
\
, Well no Concentration
(ppm)
I 1 <0.0
\-., 2
3
<0.0
0.5
4 <0.0
5 2.0
, $___ 6 <0.0
7 <0.0
8 13.0
---------.-/~~/- 9 3.0
~ 10 8.0
11 0.3
.2
0--.,. _ 12
13
0.3
<0.0
36 14 4.0
L-------- -- Highway 15 20

Fig. 19.2 Concentration contour of compounds (ppm).

 Cleanup time 503

• the concentration of volatile organic compounds in the aquifer for a

particular area of the site is equal to the average of adjacent contours
(e.g., the concentration in groundwater for the region between the
5- and lO-mg/L contours is assumed to be 7.5 mg/L);
• the aquifer has a uniform concentration of volatile organic com-
pounds with depth equal to the average of adjacent contours (e.g.,
all groundwater beneath the region bounded by the 5- and 10-mg/L
concentration contour would be assumed to have a concentration of
7.5 mg/L for the full depth of the aquifer);
• the porosity of the aquifer is uniform; and
• the saturated thickness is uniform.
The following equation can be used to estimate the mass of con-
taminant in groundwater
L
Mt = 2: M; = A;BnC; (19.1)
;=1

where: M; = mass of contaminant in groundwater for area segment i;

A; = area between concentration contours, area segment i; B =
saturated thickness; n = porosity; C; = average of bounding concentra-
tion contours for area segment i; L = number of area segments to be
summed; and M t = total of mass within all area segments.
In addition to determining the mass of contaminants dissolved in the
aquifer, the mass of contamination in the vadose zone should also be
determined, as well as any free phase that may exist. Ultimate recovered
mass can be compared with predicted mass to assess the effectiveness
of the recovery system.
Concurrent with calculating the mass of contaminants present, the
horizontal and vertical limits of contamination must be determined.
Frequently overlooked in contaminant studies is the vertical com-
ponent of flow. Although most investigations do an adequate job of
determining the horizontal extent of migration, the vertical extent is
not as well defined. Cluster wells or multiple-depth wells are required
to define the plume completely with the deeper wells being screened
below the deepest zone of contamination.

19.7 CLEANUP TIME

Although it may be possible to estimate cleanup times by using com-

puter programs, the length of many cleanup operations is much
greater than predicted. For example, a recovery well system has
recovered more than 140000 cubic meters of contaminated water at a
site in southern Louisiana for which the original plume volume was
calculated to be only 14000 cubic meters. Figure 19.3 demonstrates the
504 Recovery well systems

July
@1982
1000
August
c:
0 @ 1984 July
~ @ 1985
E
Q)
100
()
c: October
0
()
@ 1987
"0
c:
Q)
10
.t: @
a.. March
'§, 1989 September
@ 1989
5

0
0 20 40 60 80 100 120 140
Thousands of cubic meters recovered

Fig. 19.3 Contaminant concentration (logscale) vs. total recovered water.

decreased concentration of contaminants in the recovered groundwater

versus total volume pumped; although there has been a decrease
in concentration, the aquifer is still contaminated. There is very
little documented history of complete cleanup of a contaminated
aquifer. Many recovery operations have been terminated, but most
have not recovered all of the contaminant mass. The recovery systems
have ceased operation because contaminant concentrations have been
reduced to levels acceptable to regulatory authorities. A review of
several case histories indicates that removing 10-25 plume volumes
may be necessary to reduce contaminants to acceptable levels. Many
professionals are beginning to believe that recovery-well systems may
never reduce concentrations of organic contaminants to background
levels. However, because many regulatory authorities have developed
non-degradation policies for groundwater, conflicts are sure to develop
when industries petition to terminate recovery systems. Health-based
risk assessments have been suggested as one method to determine
when to halt recovery operations.
In some recovery systems, contaminant concentrations have been
observed to increase in a recovery well after an extended period of
shutdown. In the case of organic contaminants, this has been attributed
to the desorption of contaminants, predominantly from the clay matrix
in an aquifer, into the water column. This has led to the concept that
recovery wells should be operated in a cyclic manner to maximize
recovery of contaminants. Some dewatering of the aquifer has occurred
 Recovery well design 505
in some water-table aquifers after significant pumping. However,
water and contaminants remain in the dewatered cone of depression.
Cyclic pumping may also be of some benefit in restoring the dewatered
zone in such cases.

19.8 RECOVERY WELL DESIGN

The design of recovery wells depends on the type of aquifer that has
been contaminated and the recovery rate that is required. The recovery
rate must be determined before a recovery well can be properly
designed because the size and type of the pump will be major factors
in determining the diameter of the casing and screen. Recovery wells
range from 40-mm-diameter wells that use pneumatic displacement
pumps that recover about 10 l/day to large-diameter wells equipped
with high-production-rate pumps and large electric motors that can
recover about 10 000 l/min. Examples of typical well completions for
consolidated and unconsolidated aquifers are shown in Fig. 19.4.

19.8.1 Drilling methods

Recovery wells are drilled by the same general drilling techniques that
are used to drill monitoring wells. Shallow recovery wells less than
100 mm in diameter are frequently installed through a hollow-stem
auger (like a monitoring well). In unconsolidated sediments where the
recovery well is greater than 150 mm diameter, rotary drilling methods
are usually employed. When mud rotary is used to complete wells in
unconsolidated sediments, the mud invades the formation and causes
plugging. It is difficult to develop such a well to its full potential.
Although organic drilling fluids have been discouraged for use in
the construction of monitoring wells, they have an advantage over
bentonite-based drilling muds when used for recovery wells because
these organic drilling fluids break down and do not leave residuals that
plug the aquifer pores. In consolidated formations such as sandstones
and limestones, air-rotary drilling is a preferred method of drilling,
especially in formations that have a history of lost circulation.

19.8.2 Materials of construction

Recovery wells are constructed of a variety of materials such as
polyvinyl chloride (PVC), carbon steel, galvanized steel, stainless steel,
and fiberglass. Monitoring well casing and screen are selected to pre-
clude interference with the results of chemical analysis. Recovery well
systems are installed in contaminated zones and precision analytical
results are not usually a concern. For recovery wells, a more important
(a) (b) (c)
Concrete Concrete
pad pad
6 ". <II
t> "'" "' ... I>

7'.\WA\~ .... \'YI/ k'Y/,<9k'-'

'0"' ~

~-.'. ·0,
'I,'
...... ~ut ~
GROUT
"
Grout
~
. .
yX:::~:........ \ '
~ngY
Casing .':J,/ \
Casing Bore \
~.'-; , • .0

:'-'.":
hole ~
~~'-: ~'
. ..
'V 'V 'V
::(\
"C7
~ ~
Bore. Bore·
hole .' hole
'.
1:f1t"=1 0 - . ~ Screen Screen /"
Contaminated Contaminated Contaminated Open bore
~ ~
'.;
zone zone zone hole
.:...: Gravel Gravel
.: pack pack

0'0 '.' ~ ..
~ y\
.0 00 , ,
0. 00 -.:.".;:,
"
. '.
" ~.~Q_--

Fig. 19.4 Typical recovery well completion: unconsolidated aquifer with (a)
straight casing and (b) telescoped completion, and (c) consolidated aquifer with
open hole completion.
 Recovery well design 507
concern is the reaction of contaminated groundwater with the materials
of construction resulting in decreased casing or screen life. PVC casing
and screens may lose their structural integrity if they are placed in
contact with high concentrations of organic contaminants. Wells for the
recovery of free-phase hydrocarbons frequently are constructed of steel
or fiberglass.

19.8.3 Filter pack design

Wells that are screened in unconsolidated aquifers are filter packed
between the screen and the borehole. There are two types of filter
packs: natural pack and artificial pack. A natural pack is produced by
pumping and surging the well and removing the clay and silt matrix
of the aquifer through the screen, leaving the coarser sands as an
envelope around the screen. An aquifer is generally considered suit-
able for use of a natural pack if the aquifer is coarse grained and poorly
sorted (well graded).
Artificial packs are used in recovery wells when the aquifer matrix is
very fine grained or where sand, silt, and gravel are interbedded. The
main advantage of an artificial pack is that the pack material can be
coarser than the formation matrix and, therefore, larger screen slot
sizes can be used. These larger slot sizes provide less opportunity for
blockage or encrustation, either by chemicals or bacteria growth, which
results in decreased well yield.
There is considerable debate over the design of filter packs. Thick-
ness is an important consideration because it dictates the borehole size.
The larger the borehole size, the more cost involved in construction of
the recovery well. Under theoretical conditions filter pack of only a few
millimeters is all that is required. However, because the screen is not
exactly centered, a slightly thicker pack is normally recommended to
prevent formation fines from migrating into the pumping well. Most
recovery-well systems use a filter pack 50-150 mm thick. This requires
drilling a borehole with a diameter 100-300 mm larger than the screen.
If filter pack is too thick, it may be difficult to develop the well properly,
especially if the well was drilled with mud-rotary techniques. Filter
packs in excess of 150 mm are not usually recommended.
Filter packs can be installed using a tremie pipe or they can be
poured into the casing-wellbore annulus if additional filter pack is
required. Although it is not common practice, the filter pack can
consist of different grain sizes placed opposite interbedded strata of
different grain sizes. Most filter packs are not designed to be laminated,
yet many shallow fluvial aquifers consist of interbedded lenticular
strata, resulting in less than perfect gravel packs. For this type of
aquifer, the authors have normally selected a filter-pack size on the
basis of screen slot size rather than the more conventional grain-size
curves of the aquifer matrix.
508 Recovery well systems
19.8.4 Well development
During recovery well drilling, some reduction in hydraulic conductivity
will occur near the well bore and normally cannot be prevented. This
damage to the formation can result in reduced production. Surging and
backwashing will remove some of the damage to the wellbore that
occurred during drilling and will increase the hydraulic conductivity of
the material adjacent to the well bore. The greatest amount of damage
to formations takes place during drilling with mud-rotary rigs because
a mud cake builds up on the formation face during the drilling process.
This mud cake supports the formation and minimizes the loss of
drilling fluids into the formation. However, if this mud cake is not
removed after drilling, the well will produce less water than antici-
pated. Development of the well and removal of the mud cake can
become difficult if the filter pack is too thick or if the aquifer has low
hydraulic conductivity.

19.9 REHABILITATION AND MAINTENANCE OF WELLS

Recovery wells require continued maintenance. Wells completed in

consolidated or fractured aquifers without screens require less re-
habilitation and maintenance than wells completed in unconsolidated
sand formations. A major reason for maintenance is plugging of the
gravel pack or screen resulting in a reduced well yield. Pump failure is
also common in recovery-well systems.
Recovery wells are subject to the same corrosion and plugging
phenomena that occur in water-supply wells. The most common
plugging problems are caused by precipitation of calcium or magnesium
carbonates and iron compounds. Recovery wells can also encounter
maintenance problems in aquifers that are grossly contaminated. Many
organic compounds can undergo chemical reactions that result in the
precipitation or growth of organic sludges. These sludges grow in the
screen or filter pack and are difficult to remove. Generation of carbon
dioxide (C0 2) has been observed in several locations where the organic
contaminant provides a food source to bacteria. This aerobic biological
activity takes place in wells and gravel packs that have been dewatered
and exposed to the atmosphere. CO 2 can cause problems in several
ways: outgassing of CO 2 causes the pH to change and frequently
results in precipitation of carbonates where as the generations of large
quantities of CO 2 can cause cavitation in the pumps.
Several rehabilitation methods that have been developed for the
water-well industry can be used to restore production to a contaminant
recovery well. Acid is most commonly used to clean screens clogged
with calcium carbonate. Hydrochloric acid can be pumped into the well
to dissolve the scale. When acid reacts with carbonate material, CO 2 is
 Pumping system design 509
generated. If acid is used to clean shallow recovery wells, the CO 2 may
cause an eruption in the filter pack surrounding the recovery well or it
may result in acid bubbling back to the ground surface. Acid workover
of shallow recovery wells is generally not recommended. If the aquifer
contains a significant amount of clay, acid can result in the breakdown
of clays and subsequent formation plugging. In these cases, acidifica-
tion of the well may result in the loss of production. In some instances,
production can be re-established in a recovery well using mechanical
screen scrapers. However, in many instances, the cost of attempted
restoration of a recovery well may be more than the installation of
a new well. It is not uncommon for recovery wells in grossly con-
taminated aquifers to be replaced several times before the recovery
operation is complete.
Many recovery wells are completed in unconsolidated sediments
of low hydraulic conductivity, such as silty sand or sandy silt. It is
extremely difficult, if not impossible, to complete a well that does not
produce some sediments. In some cases, the silt and clay are pumped
through the treatment system and decrease the efficiency of the system
or cause it to fail. Silt and sand grains are abrasive and can damage
well screens, pumps, flow meters, and other components of the
system. In some cases, the well becomes full of sediment and must be
cleaned out. The most frequent method of cleaning is to pull the pump
from the well, circulate clean water down the well bore through a drop
pipe, and flush the sediment out.

19.10 PUMPING SYSTEM DESIGN

Two general types of pumping systems are used in the recovery of

contaminated groundwater: double-pump and single-pump systems.
Double-pump systems are used for the recovery of organic compounds
that are not miscible in water, such as straight-chain hydrocarbons.
Most petroleum products fall into this category. It should be noted that
many components of hydrocarbons are soluble in water and, not only
will there be a separate phase of a floating layer of hydrocarbons, but
the water underlying the less-dense phase will contain soluble organic
compounds such as benzene, ethylbenzene, toluene, and xylene.

19.10.1 Double-pump system

Figure 19.5 illustrates a typical double-pump recovery system used for
the recovery of hydrocarbons that are less dense than water, such
as gasoline. In this system, the water table is drawn down by the
groundwater extraction pump and the floating hydrocarbons accu-
mulate in the cone of depression. The hydrocarbons are removed
510 Recovery well systems

-
Fig. 19.5 Typical double pump recovery and air stripper for aquifer
contaminated with floating and dissolved hydrocarbons.

via the hydrocarbon extraction pump and routed to a hydrocarbon

recovery tank. The water that is pumped out by the groundwater
extraction pump contains soluble contaminants such as benzene,
ethylbenzene, toluene, and xylenes that are removed by a water
treatment system.

19.10.2 Single-pump system

The double-pump recovery system was developed to separate the
free-phase hydrocarbon and the groundwater in the recovery well.
The single-pump system recovers both hydrocarbons and water and
delivers a mixture of the two fluids through a single pipe to an above-
ground hydrocarbon-water separator. The separator is used to separate
the hydrocarbon phase, which can be recycled, from the water phase,
which will require treatment. Both electric submersible pumps and
gas-driven systems have been used.
Although leakage of hydrocarbons from underground storage tanks
is a major problem, most instances of major groundwater contami-
nation have resulted from the disposal of wastewater in evaporation
ponds or treatment lagoons. This contaminated water is usually mis-
 Groundwater treatment systems 511

cible with groundwater, and only single-phase-fluid recovery is nec-

essary. Most recovery wells for miscible contaminants are equipped
with either electric submersible pumps in medium-to high-yield aquifers
or small gas or air pumps for low-yield aquifers. Figure 19.6 is a
completion diagram of a recovery well with a submersible pump.

19.11 GROUNDWATER TREATMENT SYSTEMS

After the recovery-well system has been designed and the type and
mass of contaminant(s) have been determined, consideration of the
groundwater treatment facilities can begin. However, it is important
to recognize that the design cannot be completed until key design
parameters have been defined. These parameters are inlet concentra-
tion, desired effluent concentration, flow rate, fate of the residuals
generated by treating the water, and options for disposing of the
treated effluent. Several of the more important considerations are
discussed below.

19.11.1 Inlet concentration

Development of a realistic estimate of the inlet concentration is im-
portant. Underestimating the inlet concentration can lead to a system
that does not effectively treat the groundwater. Overestimating the
inlet concentration can lead to the installation of a system that is more
involved and expensive than it needs to be. A common error that
should be avoided is the 'conservative' estimate that the inlet concen-
tration to the treatment system will be the highest concentration for
each constituent on the site. This implies that the entire site becomes
equivalent to the hot spots for each constituent - a situation that
grossly overstates the treatment requirements.
The average inlet concentration for the treatment system should be
estimated from a material balance based on the flow rates and concen-
trations for each individual recovery well. The flow rates will have
been determined during the development of the recovery well system.
The initial inlet concentrations can be estimated at the location of each
pumping well from the concentration contour, such as that shown in
Figure 19.2. In some cases, it may be economically attractive to design
and operate more than one treatment facility, one for the most con-
centrated area and one for less contaminated areas.

19.11.2 Outlet concentration

The outlet concentration should be established before designing the
treatment fystem because the outlet concentration often influences the
 pRESSuRE ____________ l
GAUGE

TO
THREADED TREATMENT
P LU G ---------_Jd
.~ 9 (f}\ -~~~fLE
WELL SEAL ~ TO CONTROL PANEL

CONCRETE
PAD
I' I
", I

~ ~;'.,: I .-.~~
·.·1 .... •·.
"

.... :{)
:~:~~.
:~
. ·.:.~.I·.
.~·~c· .',':-'
....0",
-:. .. :' .
"; ~~~. .~. ~'.~:
'

:'~~:-' : .....
E-LINE
......
•..• "
:0.;;....'.+----- GROUT
ACCESS PIPE--------f-i:'~·;_··4-_1
.....:.'
.~'•• ;.D
.~:.~ I

......,. ",
~'
" ~: ~
H+-1fr....-.:.,.t---POWER CABLE
'.13 ....
':.:": ~
JO'

CASING ----~
.•,....'-".•;j c .• , .
",
. : ....
"
,: .. -:.f::>
LEVEL CONTROLLERS
(OPTIONAL)
DR 0 P PIPE -------f.o-.''T'";;:+--+~

WATER sz sz
TABLE

-1------ GRAVEL
PACK

SCREEN------~~~-

..
.0
• 0 ~==~~~-----PUMP
o
.
Q ()
Q
.
"

• Q
,
() ~
0
000 .·.o()()~°taQ
() 0 • tit.

Fig. 19.6 Typical submersible pump recovery well system.

 Groundwater treatment systems 513
selection of the treatment technology. The outlet concentration will
usually be heavily influenced by where the effluent will be discharged
and by the potential impact of the individual chemical constituents on
the receiving waters and potential receptors. Health-risk assessments
are sometimes required in order to develop criteria for the outlet
concentrations; however, most effluent limits are established by
regulatory authorities.

19.11.3 Flow rate

The flow rate should be determined by carefully examining the pump
scheme and the hydrogeology of the site. A realistic estimate for the
flow rate should be developed prior to designing the treatment system
because the flow rate determines the size of the treatment equipment,
which, in turn, heavily influences the capital and operating costs. It is
often tempting to specify a wide range for the flow rate; however,
this should be avoided wherever possible because the hardware and
instrumentation needed to effectively treat a highly variable flow rate
involve higher capital and operating costs.

19.11.4 Treatment technologies

Although numerous systems are available for treatment of contami-
nated groundwater, air stripping and carbon adsorption are the most
common treatment for organic contaminants. Carbon and air stripping
are frequently used in series when both volatile and nonvolatile organic
contaminants are present or when a very high removal efficiency is
required.
A detailed assessment of treatment of contaminated groundwater is
beyond the scope of this chapter. However, when designing recovery
well systems, one must have some knowledge of the general treatment
processes available. Selection of treatment technologies is influenced
by the nature of the contaminants present. Groundwater contaminated
by organic compounds can be treated by physical separation or oxida-
tion techniques. Inorganic compounds cannot be oxidized to CO 2 and
H 2 0 and, thus, must be physically separated from the groundwater.
pH adjustment may be used as pretreatment for groundwater
contaminated with either organic or inorganic compounds. For organic
compounds, pH adjustment can either optimize the conditions for
treatment or prevent buildup of scale or bacteria in equipment such as
strippers and carbon absorbers. For inorganic compounds, pH adjust-
ment is used to change a physical or chemical property to enhance
treatment, most often for removal of metals via precipitation.
514 Recovery well systems

Air stripping
The most-common air stripping process used to treat contaminated
water is the packed column or packed tower (Fig. 19.5). The packed
column method of treating groundwater uses a counter current flow
system. Contaminated water is pumped to the top of the stripping
tower and moves downward through the packing under the force of
gravity. Air is forced upward through the column by a blower placed at
the bottom of the tower. Volatile organic compounds are transformed
from a liquid phase to a vapor phase and the volatilized contaminants
then move up and out through an air outlet at the top of the column.
The treated water collects in the bottom and is discharged or reinjected.
In some locations, the air discharged through the top of the tower
is vented directly into the atmosphere. However, many regulatory
authorities prohibit the direct discharge of organic contaminants to the
atmosphere unless the total mass of emissions is very low.
Most manufacturers of air stripper equipment have developed
computer programs to assist the design engineer. Most of these pro-
grams assume that the contaminant concentration remains constant. In
many cases, this is not true and the quality of water from a recovery
well varies with time. If the concentration at the recovery well exceeds
the design input quality, the treatment system may not meet required
discharge standards.
Packed-tower air strippers are not maintenance free. Periodic checks
are required to maintain performance of the system. Frequent prob-
lems include scaling and clogging of the packing material by bacterial
or algal biomass, calcium carbonate, or iron scale similar to that ex-
perienced with well screens. For design of a treatment system, the
recovered water should be analyzed for scaling tendencies in addition
to contaminant concentrations. Pretreatment may be required to
prevent scaling and/or biological fouling.
Air strippers can be designed to be more than 90% efficient at
removing volatile organic compounds. However, more typical removal
efficiencies are indicated in Table 19.1 which contains the analytical
results before and after treatment of groundwater contaminated by a
fuel spill. Although contaminant concentrations were significantly
reduced, additional treatment may be required in order to meet dis-
charge standards. The primary contaminant of concern is benzene,
which normally has a very low allowable discharge concentration limit.
This limit can be attained by connecting an additional stripper in series
or by carbon adsorption of the water after initial air stripping.

Carbon adsorption
Activated-carbon systems are widely used for the treatment of ground-
water contaminated with organic compounds. Recovered water is
 Groundwater treatment systems 515
Table 19.1 Air stripper treatment
date influent concentration effluent concentration
(ppb) (ppb)

benzene ethyl- toluene xylenes benzene ethyl- toluene xylenes

benzene benzene

01119/89 4800 2100 14000 13000 121 <100 300 890

01/25/89 8000 2800 21000 17000 80 49 250 510
02/02/89 3400 2200 15000 16000 100 66 390 630

Flow rate: 6 GPM; tower size: 35.5cm x 5.5 meter; % removal: Benzene, 97-98%;
ethylbenzene, 97-98%; toluene, 97-99%; xylenes, 93-97%.

pumped through granular activated-carbon canisters, where the

organic compounds are adsorbed. Adsorption is a surface phenom-
enon wherein the organic molecules are held in the internal pores of
carbon granules. Activated carbon can only adsorb limited amounts of
contaminants, after which time 'breakthrough' of the contaminants
occurs. As an example, Table 19.2 lists the concentrations of contami-
nant detected in samples of groundwater before and after treatment by
carbon adsorption. It demonstrates that 'breakthrough' times vary with
different contaminants, when the groundwater has a mixture of
contaminants.
Carbon consumption rates have been developed for many com-
pounds and, if the groundwater contains only a few organic contami-
nants at low concentrations, then the anticipated life of the carbon can
usually be estimated (frequently estimated as 10-15% by weight).
However, where numerous organic contaminants are present or the
plume concentrations are not uniform, more intensive evaluations are
necessary to determine consumption rates. Batch test, column test,
or in some cases, pilot tests are suggested. Numerous case histories
exist in which the carbon usage rate was grossly underestimated on the
basis of adsorption capacity data provided by the carbon suppliers.

19.11.5 Fate of residuals from treatment

Treatment of contaminated water usually entails physically separating
the contaminants from the groundwater. The design of the treatment
system is not complete until the fate of the residual contaminants is
defined. In fact, the selection of the treatment technology often is
directed by site-specific restraints or regulatory allowances for the
disposal of the residual contamination. Some of the possible fates
of residuals from air stripping and aqueous-phase activated-carbon
systems are discussed in the following sections.
516 Recovery well systems
Table 19.2 Carbon treatment

date influent concentration effluent concentration

(ppb) (ppb)

benzene ethyl- toluene xylenes benzene ethyl- toluene xylene

benzene benzene

08/28/87 9900 14000 13000 <100 <100 <100

11/05/87 <100 <100 <100
12/10/87 <100 <100 <100
12/14/87 <100 <100 <100
07/05/88 4500 1100 4200 6800 500 <100 <100 <100
07/12/88 2100 <100 1500 700
07/19/88 3300 <100 3500 600

19.11.6 Air emission options

Three potential options are available for controlling vapor emissions
from an air stripper: direct discharge to the atmosphere, incineration/
combustion, and carbon adsorption. Direct discharge is the most
economical alternative. When the level of contaminants in the ground-
water is low, direct discharge frequently is approved by regulatory
authorities. However, when the contaminant level is high, vapor
treatment may be required. Carbon adsorption of vapors is similar
in concept to adsorption of contaminants from water. The greatest
problem with carbon adsorption systems is the disposal or regener-
ation of the carbon. Off-gases from air strippers normally contain low
concentrations of organic contaminants and cannot support a flame. If
treatment is to be by combustion, supplemental fuel is usually re-
quired. Several companies market fume incinerators for use as emis-
sions treatment. If the project is expected to be a short-term operation,
then carbon adsorption may be advantageous, particularly because it
involves a lower capital cost.

19.11.7 Carbon disposal

Several of the treatment options discussed use carbon for either pri-
mary treatment of water or secondary treatment for vapor control.
Once the carbon has reached saturation and no additional adsorption is
possible, the carbon and the adsorbed organic compounds pose a
disposal problem. Disposal options may be limited by regulations
depending on the source and type of contaminant. Carbon frequently
can be returned to the supplier for thermal regeneration and reuse.
 Groundwater treatment systems 517
Other options are off-site incineration, disposal in a landfill, and
possibly on-site regeneration using steam or other methods.

19.11.8 Effluent disposal options

Five major options are available for disposal of treated effluent. Each
method has advantages and limitations. The final option that is selected
frequently is selected on the basis of regulatory considerations rather
than technical merit.

Reinjection
Treated water can be returned to the aquifer after treatment either via
injection wells if the contaminated aquifer is deep or via infiltration
trenches if the aquifer is shallow. If the aquifer is very permeable, then
injection should be explored. However, if the hydraulic conductivity is
low, or if the water table is high, reinjection may result in surface
discharge of the injected fluids and may not be practical. Under those
circumstances, other options may be more advantageous.

Sanitary sewer
Most industrial or urban areas are served by a municipal or regional
sanitary sewer. Most cities or sewer authorities have ordinances that
determine what is suitable for discharge to the sanitary sewer. If dis-
charge is allowed, the sewer authority will specify the allowable quality
and quantity of the effluent and any analytical and reporting require-
ments. Sewer tap fees frequently are required and monthly charges are
assessed on the basis of discharge volume. High discharge fees often
make this option less attractive than others. However, it is the most
trouble-free and practical solution for low discharge volumes in urban
areas.

Surface discharge or evaporation

Treated effluent may sometimes be discharged to a stream, storm
sewer, or drainageway. Permits for this type of discharge are usually
required by regulatory authorities. In many cities, municipal approval
is also required. Permit applications can be quite lengthy and public
hearings frequently are required. Other discharge options include
irrigation, especially in rural areas where land is available. However, if
the treated effluent contains low concentrations of hazardous com-
pounds, securing regulatory approval for irrigation may be impossible.
In these cases, lined evaporation lagoons may be an acceptable option.
518 Recovery well systems

Reuse
If the treatment plant is located at an industrial facility, some water
reuse may be possible. Treated water could be used for plant process
water or cooling water. However, this option generally is not viable for
small industrial facilities.

19.12 RECORD KEEPING

In order to monitor the progress of a recovery well system, good

record-keeping procedures must be developed. The original static
water level, pumping rate, pumping level and contaminant concen-
trations should be determined and recorded for each well. During
initial start-up operations, pumping rates and water levels should be
measured daily. The time between measurements can gradually be
increased to weekly and later, as the system stabilizes, to monthly or
quarterly. Total volume pumped per well and for the total system
should be recorded. A review of these records can detect early prob-
lems before complete component or system failure occurs. Without
adequate records, it is almost impossible to determine the reasons for
decreases in well performance without removing the pump and per-
forming downhole investigations. Chemical analyses of the treatment
system influent and effluent should be performed routinely to deter-
mine the mass of contaminants removed from the groundwater and
the efficiency of the selected treatment option. In addition, monitoring
well samples should be collected and analyzed periodically to measure
the results of the cleanup operation. Without adequate analysis,
updated predictions of cleanup times or treatment cost cannot be
developed.

19.13 CONCLUSIONS

Although groundwater contamination has been recognized as un-

desirable for centuries, heightened environmental awareness in the last
25 years has significantly increased the demand for effective recovery
and treatment systems and technological improvements. In the last 10
years, thousands of groundwater recovery wells have been installed in
the United States. These systems range from small, single-well systems
pumping at low rates to large, multiple-well systems pumping hun-
dreds to thousands of liters per minute. Many recovery/treatment
systems have experienced major operating problems and/or have not
realized expectations. The following elements are responsible for many
of the problems observed:
 References 519
• inadequate determination of aquifer properties resulting in improper
spacing of recovery wells and incorrect predictions of pumping
rates;
• inadequate definition of vertical and horizontal limits of
contamination;
• failure to anticipate chemical and biological reactions that lead to
screen plugging or pump failure;
• failure to recognize the loss of efficiency in recovery wells resulting
in lower well yields and smaller capture radius;
• inadequately defined groundwater chemistry resulting in poor
selection of treatment systems and fouling of air strippers and
carbon beds;
• inadequately defined design parameters resulting in undersized/
oversized treatment systems;
• inadequate budgeting for maintenance of the recovery system;
• failure to maintain records of pumping rates, water levels, and
water quality, resulting in incorrect conclusions concerning system
performance; and
• failure to communicate to the client and regulatory agencies the
limitations of recovery systems, resulting in overally-optimistic
predications of system success.
The lessons learned during the last 10 years of operating recovery
systems indicate that pump-and-treat recovery well systems alone are
generally not sufficient to clean up most aquifers, particularly those
contaminated with organic compounds. However, a recovery well
system can immobilize a plume and prevent further migration of
contaminants. In addition, the bulk of the total mass of contaminants is
usually removed via the recovery system, thus lessening the risk to
human health and the environment. Although recovery well systems
will always be used, additional methods of remediation such as bio-
reclamation soil washing, or soil venting may be needed to reduce
contaminants to acceptable levels.

REFERENCES

Saunders, T.J. (1970) Plato; The Laws, Penguin Books LTD, New York, p. 348.
Theis, C.V. (1935) The relation between the lowering of the piezometric surface
and the rate and duration of discharge of a well using ground-water
storage, Trans. of the Am. Geophysical Union, Washington, DC, 518-24.
 CHAPTER 20

Bioremediation of soils
Raymond C. Loehr

20.1 BACKGROUND

Complex mixtures of chemicals, many defined as hazardous and toxic,

increasingly enter surface soils through spills and unregulated land
disposal. The regulatory and technical challenge is to use cost-effective
control technologies that can treat complex chemical mixtures in con-
taminated soils and thereby reduce the resultant threat to human
health and the environment. Bioremediation can be such a technology.
Bioremediation is a managed, demonstrated active treatment process
that uses microorganisms to degrade and transform organic chemicals
in contaminated soil, sludges and residues. Bioremediation also reduces
the toxicity of the organics and the migration potential of hazardous
constituents in the soils, sludges and residues being treated. As such,
bioremediation processes can be considered as source control, pol-
lution prevention, and risk reduction processes that can reduce or
eliminate groundwater contamination and thereby reduce the need for
costly and long-term groundwater treatment processes.
Bioremediation processes have resulted from the application of
knowledge from microbiology, biochemistry, environmental engineer-
ing and chemical engineering. The fundamentals of biological treat-
ment, as applied to contaminated soil conditions, are the key to the
proper design and operation of the bioremediation processes. In such
processes, established scientific and engineering principles are used to
maintain satisfactory conditions for microbial degradation and loss of
organics.
The advantages of contaminated soil bioremediation processes are
that the processes:
• are used where the problem is located;

Geotechnical Practice for Waste Disposal.

Edited by David E. Daniel.
Published in 1993 by Chapman & Hall, London. ISBN 0412 35170 6
 Background 521

• do not require transporting large quantities of contaminated material

off-site;
• eliminate the problem rather than moving it somewhere else;
• minimize long-term liability;
• are ecologically sound and an extension of natural processes; and
• generally are cost-effective and competitive with other decontami-
nation technologies for organics.
Bioremediation commonly is part of a total system for a site and can be
used with other technologies to remediate a site and lower the overall
cost of site cleanup.
Many bioremediation processes exist. Choosing the appropriate
process is a function of the remediation goals to be achieved, the
physical and chemical characteristics of the material to be treated, the
environmental conditions that are created, the materials handling and
equipment requirements, and the overall economics. The potential
processes include:
• Solid phase processes, such as composting and land treatment; and
• Slurry phase processes, such as liquid solids systems in impound-
ments or bioreactors.
Bioremediation processes have been used in the United States at
Superfund sites. An evaluation of the records of decisions (RODs)
that have been made for Superfund site remediations through 1989
indicates that bioremediation has been recommended for source con-
trol at over twenty sites. The types of bioremediation processes that are
used at these sites are noted in Table 20.1. The chemicals treated at
these sites primarily have been polyaromatic hydrocarbons (PAH),
volatile organics (benzene, toluene, ethylbenzene and xylene which
collectively are noted as BTEX), pentachlorophenol, and phenols.
Other chlorinated and non-chlorinated organiCS also were present at
some sites.

Table 20.1 Bioremediation processes recommended for contaminated soils

and sludges at Superfund sites*

type of process number of times recommended

------------ - - - - - - - - - - - - - - - -

Excavation followed by land treatment 11

Excavation with on-site treatment 3
In situ treatment 4
Lagoon aeration 3
To be determined and other 2

* from review of Superfund RODs issued through 1989 (USEPA, 1990); some RODs
specify multiple remedies.
522 Bioremediation of soils

Both solid phase and slurry phase processes, as discussed in this

chapter, are applicable to surface soils that are contaminated as a
result of spills; pits, ponds and lagoons that have leaked; or improper
chemical disposal practices. The contaminated soils commonly are
unsaturated and exist in the vadose zone above the groundwater. In
situ bioremediation processes also can be considered for the treatment
of contaminated groundwater and deep saturated soils. These latter
processes are discussed in Chapter 21.

20.2 BIOREMEDIATION PERFORMANCE CRITERIA

The performance to be achieved is the key to the feasibility of any

bioremediation process. The selection of performance criteria ('cleanup'
standards) continues to be a subject for discussion and negotiation at
the state and federal level. Soil cleanup criteria are established to
assure protection of human health and the environment. Such pro-
tection is achieved by remediation methods that reduce the mobility,
quantity and toxicity of chemicals in the contaminated soil such that
human and environmental receptors are not threatened by any pathway.
In practice, soil cleanup criteria satisfy the following two goals. The
first is that the concentrations of chemicals left in the soil should be at
levels which protect people from risk due to
• ingestion,
• dermal contact,
• inhalation of fugitive dust, and
• inhalation of gaseous constituents volatilized from the soil.
The second is that the residual chemicals are in a form and at con-
centrations that will not leach and contaminate groundwater.
Acceptable contaminated soil cleanup concentration limits are es-
tablished by the regulatory agencies on a site-specific basis prior to
remediation. Such criteria should take into account site-specific factors
such as contaminant sources, physicochemical conditions, environ-
mental sensitivity, potential for human exposure, and subsequent land
and/or water uses desired.
Risk assessments increasingly are used to identify appropriate
cleanup criteria. An assessment of the risks at a site to be remediated
can determine a target for the remediation effort. Such a risk assess-
ment typically has the components identified in Table 20.2.
The results of a risk assessment help identify:
• the degree of control that is needed;
• the cleanup goals that should be achieved to minimize relative risk
from this site; and
 Bioremediation performance criteria 523
Table 20.2 Major components in a risk assessment to determine soil cleanup
goals

Hazard identification Determine the relative risks resulting from chemicals at a

site:
• select compounds to include;
• identify exposed populations;
• identify sensitive receptors; and
• identify exposure pathways.
Exposure assessment To how much of a compound are people and the
environment exposed? For exposure to occur, the
following must happen.
• release
• contact
• transport
• absorption or intake
Together, these form an exposure pathway. There are
many pathways, such as air, groundwater, surface
water, crop and ingestion.
Toxicity assessment What adverse health effects in humans are potentially
caused by the compounds in question?
Review the threshold and non-threshold effects
potentially caused by the compounds at the
environmental concentration levels.
Risk characterization At the exposures estimated in the exposure assessment,
is there potential for adverse health effects to occur and,
if so, what kind of adverse effects and to what extent?
This step develops a hazard index for threshold effects,
estimates the excess lifetime cancer risk for carcinogens
and estimates risk for other chemicals.

• the type of treatment or remediation that might accomplish the

desired cleanup.
Cleanup goals are established as numerical standards or concentration
limits. Because of the site and risk specific nature of these goals, a
variety of standards or limits have been established at specific sites.
Examples are (USEPA, 1988):
• 10000ppb total petroleum hydrocarbons, 67ppb benzene, 200ppm
toluene, 44 ppm xylene;
• <1 ppm benzene, toluene, ethylbenzene, and xylene (BTEX);
• risk assessment performed at each site;
• <500ppb total hydrocarbons (THC), <100ppb total aromatics;
• to background levels if feasible;
• THe below 100ppm;
• total BTEX 50-100 ppm or lower;
• non-detectable levels of volatiles; and
• <500 ppm BTEX and non-ignitables.
524 Bioremediation of soils

Until a comprehensive database is developed for soil cleanup criteria,

numerical values established as criteria should not be considered as
absolute standards, but rather as guidelines.

20.3 PROCESS FUNDAMENTALS

20.3.1 Overview
The microbial degradation of organic compounds has been recognized
for centuries as a highly efficient and relatively inexpensive method to
treat organic compounds. Biotreatment processes such as composting
for sludges and organic refuse, the activated sludge and trickling filter
processes for wastewater, and anaerobic digestion for manures and
organic sludges have been used for many decades. However, the
application of biological processes to degrade and detoxify industrial
organics in soils is not as common. This is due primarily to a lack of
understanding of:
• the scientific and technical principles involved; and
• how those principles can be used to design and operate bioreme-
diation processes.
This section presents the pertinent principles and factors and discusses
how they can be used and applied to such processes.
Bioremediation of a soil contaminated with organic wastes is ac-
complished by microbial degradation of specific organic compounds.
Organisms in the soil require energy for food and growth. Micro-
organisms obtain this energy through the metabolic degradation of
organic compounds. Thus, bioremediation involves the metabolic
breakdown of organic chemicals as a food and energy source. Gen-
erally, microorganisms that currently exist in soil and water can degrade
the organics in contaminated soils, assuming that the environmental
conditions are suitable and nontoxic conditions exist. Microorganisms
have evolved catalytic systems (enzymes) that degrade naturally
occurring compounds present in the biosphere. However, contami-
nated soils can contain man-made organics that are difficult to degrade.
With such chemicals, the natural microbial enzyme systems may have
to adapt and acclimate to the chemical before degradation can occur.
Acclimation results in an increase in the biodegradation rate of a
chemical after the microbial community is exposed to the chemical for
some period of time. The basic concepts that relate to bioremediation
are:
• bioremediation is a source control, risk reduction and pollution
prevention process;
 Process fundamentals 525
• bioremediation processes reduce the toxicity and migration potential
of organic compounds;
• biodegradation occurs in a wide variety of environments and
both solid phase and slurry phase processes can be used for
bioremediation;
• organic compounds are microbially converted to simpler compounds;
• microorganisms obtain the energy requirements for growth and
maintenance from the compounds they degrade;
• microbial enzymes evolved for the degradation of naturally occur-
ring organics can be acclimated for the bioremediation of many
organics in contaminated soils; and
• suitable environmental conditions are necessary for the successful
bioremediation of contaminated soils.
The factors that affect the performance of bioremediation processes
are noted in Table 20.3.
The microbiological degradation of organics transforms carbon from
the organic to the inorganic state. The transformation of organic carbon
to inorganic carbon (C0 2) is accomplished through enzymatic oxida-
tion, with molecular oxygen involved as a terminal electron acceptor
(aerobic metabolism). This also can occur if the final electron acceptor is
something other than molecular oxygen such as sulfate (50 4 ) or nitrate
(N03 ) (anaerobic metabolism). Aerobic degradation results in the
production of CO 2; anaerobic degradation results in both methane
(CH 4 ) and CO 2 production. Biological treatment produces innocuous
end products - usually CO 2 and microbial biomass.
In the soil, a constituent may not be completely degraded, but
transformed to intermediate product(s). The goal of bioremediation
processes is the detoxification of a parent compound to a product or
product(s) that are no longer hazardous. Thus, degradation may result
in detoxification without complete mineralization.
If bioremediation can occur and is effective, why do some organics,
such as those in contaminated soils, persist in the environment? This is
a reasonable question that is answered by identifying the factors that
affect microorganisms and therefore bioremediation processes. Even
biodegradable organics may persist if adverse factors and non-optimum
conditions exist. The factors that may prevent microbial degradation
and bioremediation include:
• chemical concentrations that are toxic to microorganisms;
• inadequate type or numbers of microorganisms, such as due to toxic
conditions;
• conditions too acid or alkaline;
• lack of nutrients such as nitrogen, phosphorus, potassium, sulfur
or trace elements (many organic chemicals, for example, are not
nutritionally balanced);
526 Bioremediation of soils
Table 20.3 Factors affecting bioremediation processes'

factor comment

Microorganisms Natural organisms are satisfactory,

acclimation may be necessary, suitable
environmental conditions need to be
provided
Toxicity Non-toxic conditions are needed
Available soil water 25-85% of water holding capacity
desirable for solid-phase systems
Oxygen (0 2) Aerobic conditions desired
Electron acceptors Under aerobic conditions, O 2 is the
terminal electron acceRtor; when O 2 is
not available, N03", Fe 3 +, Mn2 + and
S~- can act as electron acceptors
Redox potential (oxidative and Aerobes and facultative aerobes
reduced conditions) predominate at greater than 50 millivolts;
anaerobes at less than 50 millivots
pH 5.5-8.5 for optimum degradation
Nutrients Nitrogen, phosphorus and other
nutrients are needed for microbial
growth
Temperature Degradation rates are affected by
temperature
Water solubility The water solubility of compounds in
contaminated soil can affect degradation
Sorption Many organic compounds are strongly
sorbed to the organic matter in soil
Volatilization Chemicals can be lost by volatilization
Loss rates Total loss rates commonly are reported;
first order rates usually used to describe
losses

• from Huddleston et al. (1986) and Sims et al. (1989) .

• unfavorable moisture conditions (too wet or too dry); and

• lack of oxygen or other electron acceptors.
Bioremediation processes are biological treatment processes that
improve or stimulate the metabolic capabilities of microbial populations
to degrade organic residues. Therefore, it is important to understand
those conditions and reactions so that bioremediation processes can be
successful. With such knowledge, it is possible to modify the non-
optimum conditions so that microbial degradation can occur. For
 Process fundamentals 527
example, if there is inadequate nitrogen or phosphorus, such nutrients
can be added to assure satisfactory microbial degradation. If the
residues are too toxic, addition of other chemicals or uncontaminated
soil may reduce the toxicity to the point that microbial degradation can
occur. If inadequate type or numbers of microorganisms are present,
acclimated organisms can be added.
The factors that affect bioremediation processes were identified in
Table 20.3. The following sections describe the factors that affect the
performance of bioremediation processes.

20.3.2 Microorganisms
Surface soils contain large number of microorganisms that include
aerobic and anaerobic bacteria, fungi, actinomycetes and protozoa
as well as earthworms and higher forms of life. Over one million
organisms can be in one gram of agricultural soil. These organisms are
capable of degrading most natural and synthetic organics that are in a
soil. However, they can accomplish the degradation only if:
• non-toxic conditions exist;
• the organisms have or can develop the enzyme systems capable of
degrading the organic compound; and
• other environmental conditions such as pH, nutrients, oxygen,
temperature, and water are adequate.
The actual degradation that occurs in a bioremediation process is a
result of the mixed microbial population that exists. One group of
microorganisms may partially metabolize a compound and furnish a
suitable substrate for another group of microorganisms. If an organic is
biodegradable and environmental conditions are suitable, the natural
organisms in the soil can adapt to degrade the organic compound.
Specially developed microbial cultures have not been observed to be
needed or successful in bioremediation systems.

20.3.3 Toxicity
Many individuals who consider remediation options for contaminated
soil incorrectly assume that bioremediation is not feasible because toxic
chemicals or hazardous wastes are in the soil. They infer that because
such materials are toxic to living organisms, biological treatment is not
a viable alternative. However, this assumption and inference is not
correct. Many organic chemicals indicated as toxic or in hazardous
wastes can be biodegraded if the contaminated soil is not toxic to
the microorganisms and if other environmental factors, such as pH,
nutrients and oxygen, are suitable.
528 Bioremediation of soils
In considering the feasibility of bioremediation processes, the relative
toxicity of the contaminated soil should be determined. If the soil is
found to be toxic to the soil microorganisms, either steps should be
taken to reduce the toxicity or another remediation process should be
considered.
The usual procedures to quantify toxicity of a chemical are toxicity
assays which measure the effect of the chemical under specified test
conditions. The toxicity of a chemical is proportional to the severity
of the chemical on the monitored response of the test organism(s).
Toxicity assays utilize test species that include rats, fish, invertebrates,
microbes and seeds. The assays may use single or multiple species of
test organisms. Although no single bioassay procedure can provide
a comprehensive toxicity evaluation of a chemical, a valid toxicity
screening test can provide information about the relative toxicity of a
compound and can help predict non-inhibitory chemical application
rates.
Toxicity assays using bacteria as the test organism are rapid, are easy
to use, are cost effective, and use a statistically significant number
of test organisms. One such bacterial toxicity assay method is the
Microtox© assay. This method is relatively simple, rapid and inexpen-
sive. The use of the Microtox© procedure to screen and predict the
treatability potential of waste in soil and of contaminated soil has been
evaluated and found to be satisfactory (Matthews and Bulich, 1984;
Matthews and Hastings, 1987).
The Microtox© system is a standardized toxicity test which utilizes
marine luminescent bacteria (Photobacterium phosphoreum) as indicator
organisms. Bioluminescence of this test organism depends on a com-
plex chain of biochemical reactions involving the luciferin-Iuciferase
system. Chemical inhibition of any of the involved biochemical reac-
tions causes a reduction in bacterial luminescence. The Microtox©
toxicity assessment considers the physiological effect of a toxicant, and
not just mortality.
This method utilizes an instrumental approach in which the indicator
organisms are handled as chemical reagents. Suspensions of about one
million bioluminescent organisms are 'challenged' by addition of serial
dilutions of an aqueous sample. A temperature controlled photometric
device quantifies the light output in each suspension before and after
sample addition. Reduction of light output reflects physiological in-
hibition which indicates presence of toxic constituents in the sample.
Such tests do not provide information on toxicity from a human health
or an environmental standpoint. Rather, they are used as a relative
toxicity screening method for contaminated soil and to identify the
relative toxicity reduction that occurs when chemicals and waste are
managed by bioremediation processes.
 Process fundamentals 529
20.3.4 Water
The presence of water is essential for microbial activity. At low water
concentrations, non-spore forming microorganisms will die or their
concentrations will be reduced greatly. At high water concentrations,
such as at or near saturation, the pores of the soil are filled with water
and diffusion of oxygen from the atmosphere is restricted. Under the
latter situation, anaerobic rather than aerobic conditions will occur
unless oxygen is mechanically added, such as in an aerated slurry
reactor. The water content of soil typically ranges from 15 to 35 volume
percent. At 35%, most soils are water saturated. The water content can
drop below 15% under arid conditions. Soil water content is commonly
expressed as a percentage of soil water holding capacity. A soil water
holding capacity range of 25-100% is the equivalent to a range of
about 7-28 volume percent.
Over the range of about 30-90% of the water holding capacity of the
soil, the moisture content has little effect on biodegradation rates
(Dibble and Bartha, 1979). In solid phase bioremediation systems,
moisture control may be necessary to achieve optimum biodegradation
rates.

20.3.5 Oxygen
Adequate oxygen and aerobic conditions in bioremediation systems are
important to:
• avoid odors produced by anaerobic conditions; and
• produce the most oxidized end-products.
Anaerobic conditions and degradation can occur in bioremediation
systems but should be avoided since anaerobic biodegradation is
slower and less complete and under reduced conditions, most metals
are more water soluble.
For solid phase systems, aerobic conditions are maintained by:
• mechanically mixing the material, such as in composting, or tilling
the mixture, as in land treatment;
• avoiding saturating the mixture with water; and
• maintaining the quantity of degradable material in the mixture such
that the oxygen demand does not exceed the rate that oxygen is
transferred from the atmosphere to the mixture.
For slurry phase bioremediation systems, oxygen is added by
mechanical mixers and aeration systems. These systems also mix the
liquid slurry to keep the particles in suspension and enhance oxygen
transfer. Oxygen transfer in slurry systems usually is less than in dilute
or clean water systems and the aeration equipment should be sized to
530 Bioremediation oj soils

transfer the necessary amount of oxygen under operating and not clean
water conditions. To avoid oxygen limiting conditions, the dissolved
oxygen concentrations in slurry reactors should be maintained above
about 0.5 mg/l.

20.3.6 Electron acceptors and redox potential

In a bioremediation system, microorganisms metabolize organic com-
pounds to obtain biological energy for microbial growth and main-
tenance. In this process, electrons from incompletely oxidized (reduced)
compounds are transferred along respiratory electron transport chains,
energy is captured by the microorganisms, and oxidized end-products
(such as carbon dioxide, CO 2) result. Under aerobic conditions, oxygen
(0 2) is the terminal electron acceptor. When O 2 is not available, nitrate
(NO;-), iron (Fe 3 +), manganese (Mn2+), and sulfate (SO~-) can act
as electron acceptors if the organisms have the appropriate enzyme
systems. The oxidation-reduction potential (redox potential) of a soil
provides a measurement of the electron density of the system. The
redox potential is expressed in millivolts.
Microbial degradation in bioremediation systems removes oxygen
from the soil atmosphere in solid phase systems. If oxygen is not
replenished, the systems become depleted in oxygen, reduced con-
ditions result, and other substances are used as terminal electron
acceptors. To maintain aerobic conditions, oxygen must be added to
the bioremediation systems by the methods noted in section 20.3.5.
The loss of oxygen from the system also can cause a change in
the microbial population. Facultative bacteria, which do use electron
acceptors such as nitrate, sulfate, or oxygen, and anaerobic organisms
become the dominant microbial populations. In bioremediation sys-
tems, the vast majority of the microorganisms are facultative.

20.3.7 pH
The optimum pH for microbial degradation is around neutral, gen-
erally in the range 6-8. However, biodegradation can occur outside
this range although at reduced rates. In bioremediation systems, pH
control rarely is needed unless very acid or alkaline conditions are
encountered. pH control also can be needed if the biological activity
causes a marked change in the pH.

20.3.8 Nutrients
Microbial metabolism and growth requires adequate macro and micro-
nutrients. The soil normally supplies adequate micronutrients such as
trace metals and minerals. However, it cannot be assumed that all soils
 Process fundamentals 531

have adequate macronutrients such as nitrogen (N), phosphorus (P) or

potassium (K). Industrial organics in contaminated soils can be high in
carbonaceous content but low in Nand P.
The need for additional Nand P is controlled by:
• the amount of Nand P in the contaminated soil and the rate that
they are available; and
• the amount and rate at which organic carbon is degraded in the
mixture.
By estimating or measuring these factors, the need for additional N
and P can be determined. A carbon to nitrogen (ClN) ratio of greater
than 35-40 generally indicates inadequate nitrogen. A ClN /P ratio of
about 100: 10: 1 will provide adequate nutrients in a bioremediation
system. If inadequate nitrogen exists, the deficiency can be remedied
by adding a chemical fertilizer such as ammonium sulfate or ammonium
diphosphate. Over time, the need for additional nutrients can decrease
since most of the initially added nutrients remain in the system for
reuse.

20.3.9 Temperature
Biodegradation rates are affected by temperature, generally changing
by a factor of two for a change of 10 DC in the temperature range
of 5-30°C. Low temperature is not lethal to microorganisms but will
drastically reduce biodegradation rates. Biodegradation is essentially
zero at freezing temperatures. Insulated reactors can be used to keep
the reactor temperature 5-10 °C above ambient air temperatures.
In bioremediation systems, temperature control rarely is practiced.
Different systems will operate at different temperatures, however. For
example, solid phase composting type systems may operate at tem-
peratures that are above ambient. The temperature in bioremediation
systems normally are near ambient temperatures and will change
slowly as ambient temperatures change. Abrupt temperature changes
are unlikely.

20.3.10 Water solubility

Microbial degradation is carried out by microbial enzymes. In biore-
mediation systems, the enzymes are not released by the microbial cells
and the substances to be degraded must contact or be transported to
and into the microbial cell. Thus, the water solubility of the chemical is
important since only the water soluble fraction of a chemical is readily
degradable.
Many of the chemicals in contaminated soil are not very water
soluble. This low solubility can reduce the availability of the chemical
532 Bioremediation of soils

for degradation and will be a controlling parameter in bioremediation.

The aqueous solubilities of chemicals commonly found in contaminated
soils are noted in Table 20.4. The aqueous solubility decreases as the
chemical becomes more complex.
Chemicals that have low water solubility generally are soluble in
organic solvents. A commonly used measure of the relative organic
solvent solubility is the octanol-water partition coefficient, Kow, which
is defined as the ratio between the concentrations of the compound in
equal parts of octanol and water. A value greater than 1 indicates that
the compound is more readily soluble in octanol than in water. actanol
models the behavior of other hydrocarbons and is used as a model
nonpolar organic hydrocarbon solvent. Kow values for many chemicals
also are noted in Table 20.4.
The octanol-water partition coefficient is a key parameter in studies
of the environmental fate of chemicals. Kow has been found to be

Table 20.4 Solubility and octanol-water partition cofficients of organic

chemicals'

compound pure compound aqueous

solubility (mglL)

Monocyclic aromatics
Benzene 1787 135
Toluene 515 540
o-Xylene 213 1320
p-Xylene 185 1410
m-Xylene 146 1585
Ethylbenzene 110 1410
Other halogenated compounds
Trichlorethene 1100 260
Pentachlorophenol 14 132000
Polyaromatic hydrocarbons (PAH)
Naphthalene 31 2000
Acenaphthylene 3.93 5500
Acenaphthene 3.42 8300
Fluorene 1.98 15100
Phenanthrene 1.29 28800
Fluoranthene 0.26 79400
Pyrene 0.135 75900
Anthracene 0.066 28200
Benz[ a ]anthracene 0.014 407000
Benzo[ a ]pyrene 0.0038 933000
Chrysene 0.002 407000
Benzo[b]fluoranthene 0.001 3720000
Benzo[k]fluoranthene 0.0003 6920000
Benzo[ghi]perylene 0.00026 17000000
Benz[a ]anthracene 0.014 407000

• adapted from Tetra Tech (1989).

 Process fundamentals 533
related to water solubility, soil and sediment adsorption coefficients
and bioconcentration factors (BCF) (Lyman et al., 1982).
Chemicals with low Kow values, i.e., less than about 10, have high
water solubilities, small adsorption coefficients and small BCF values.
Chemicals with high Kow values, i.e., greater than about 1000, are very
hydrophobic with low water solubilities and high sorption coefficients.
The aqueous solubility (5) of a chemical can be estimated from Kow
values. A number of equations have been developed to correlate these
two parameters. The equations generally take the form of
log 5 = a log Kow + b (20.1)
where a and b are empirical regression equation constants. Examples of
such relationships are noted in Table 20.5.

20.3.11 Sorption
Sorption is another parameter that affects organic chemical degrada-
tion. This occurs because the greater extent a chemical is sorbed to a
soil, the longer it is retained in the soil and the greater amount of time
that is available for degradation. Thus even chemicals that have slow
degradation rates (long half-lives) can undergo satisfactory degradation
in contaminated soil bioremediation systems.
The mobility of a chemical in a soil system can be expressed in terms
of a retardation factor which, as noted in Chapter 3, can be calculated
as
K" X Pb
R = 1 + ~---,c..::. (20.2)
e
Table 20.5 Examples of empirical relationships between aqueous solubility
and Kow values for different chemicals*

equation units chemical classes

of 5 represented

log 5 = -1.37 log Kow + 7.26 IlmollLt mixed classes; aromatics and
chlorinated hydrocarbons well
represented
log 5 = -0.922 log Kow + 4.184 mg/L mixed classes; pesticides well
represented
log 5 = -1.49 log Kow + 7.46 IlmollL mixed classes; several pesticides
log(1/5) = 1.294 log Kow - 0.248 mollL alkenes
log (1/5) = 0.996 log Kow - 0.339 mollL benzene and benzene derivatives
log(1/5) = 1.237 log Kow + 0.248 mollL alkanes

* adapted from Lyman et al. (1982).

t moJlL = moles per 1000 grams of water, i.e., molar solubility.
534 Bioremediation of soils

where R = retardation factor (unitless), Kp = soil-water partition

coefficient (mllg), Pb = dry bulk density of permeable material (g/ml),
and EI = volumetric moisture content (decimal fraction). The partition
coefficient, Kp, also is known as the sorption coefficient of a chemical to
soil, and, as discussed in Chapter 3, the distribution coefficient, Kd . If a
chemical does not interact with the soil, Kp = 0 and the retardation
factor equals 1. Examples of such chemicals are chlorides and nitrates.
Chemicals with high sorption coefficients are more tightly bound to the
soil, are less mobile, and have lower degradation rates.
Adsorption is a surface phenomenon in which matter is extracted
from one phase and concentrated at the surface of the second. Ad-
sorption of a chemical from solution onto a solid occurs as the result of
the lyophobic (solvent-disliking) character of the chemical relative to
the particular solvent, or of a high affinity of the chemical for the solid.
Adsorption is modeled by an isotherm, which is a means of de-
scribing changes in adsorption at constant temperature (see Chapters 2
and 3). Adsorption of many chemicals found in contaminated soils
commonly can be described by a Freundlich isotherm
-1
q = KC" (20.3)

where q = mass of chemical sorbed per mass of soil at equilibrium, K =

adsorption coefficient, C = concentration of chemical in the liquid, and
n = an empirical constant representing adsorption intensity.
The adsorption coefficient, K, can be normalized to the organic
carbon content of the mixture, Kac
(20.4)

where fae is the fraction of organic carbon content of the soil. This
relationship works well for soils with organic carbon contents above
about 0.5%.
Thus, it is possible to express the tendency of a chemical to be
adsorbed in terms of KaCf which is largely independent of the prop-
erties of the soil. Kac may be thought of as the ratio of the amount of
chemical adsorbed per unit weight of organic carbon (oc) in the soil to
the concentration of the chemical in solution at equilibrium.

I1g adsorbedlg organic carbon

(20.5)
Kac = I1g/mL solution

Kac is a constant only if the adsorption isotherms are linear. Generally,

linearity occurs when the equilibrium aqueous phase organic chemical
concentration is below one half of the pure compound aqueous solu-
bility of the chemical.
Kae also has been related to the aqueous solubility, S, and the Kow of a
chemical. The relationships generally take the form of
 Process fundamentals 535

log Koc = e log S + d (20.6)

or
log Koc = e log K>u' + d
where e and d are empirical regression constants. Examples of such
relationships are noted in Table 20.6.

20.3.12 Volatilization
Volatile constituents can be in many soils. Such constituents will be
released to the atmosphere during bioremediation processes. Controls
on volatile losses may be needed as part of the contaminated soil
bioremediation system to meet regulatory requirements.
The occurrence of volatilization in such systems is a function of the
vapor pressure of the chemical and the contact between the chemical in
the soil and the gaseous phase or atmosphere. Bioremediation systems
are mixed and/or aerated to provide adequate oxygen input and aerobic
conditions. As a result, such systems offer excellent opportunities for
volatile compounds to be released.

20.3.13 Chemical loss rates

Results from field systems and laboratory and pilot plant bioreme-
diation systems indicate that the losses of organic compounds can be
described by first-order rate reactions in which the rate of loss of a
chemical is proportional to the chemical concentration.
de
-ke (20.7)
dt

Table 20.6 Examples of empirical relationships between Koc values and other
parameters for different chemicals*

equation chemical classes represented

log Koc = -0.55 log st + 3.64 wide variety, mostly pesticides

log Koc = -0.557 log S+ + 4.277 chlorinated hydrocarbons
log Koc = 0.544 log Kow + 1.377 mostly pesticides
log Koc = O. 937 log Kow - 0.006 aromatics, polynuclear aromatics, triazines and
dinitroaniline herbicides
log Koc = 1.00 log Kow - 0.21 mostly aromatic or polynuclear aromatics

* adapted from Lyman et al. (1982).

t solubility (S) in mg/L.
t solubility (5) in micro-moles/L.
536 Bioremediation of soils

where c = chemical concentration (mass/mass), t = time, k = first-

order rate constant (time-I).
Chemical loss rates usually are discussed in terms of half-life (tl),
i.e., the time required to degrade or lose one-half of the chemic!al
concentration. Mathematically, half-life can be calculated from the
first-order rate constant
0.693
t~ = -k- (20.8)

Examples of half-life values of chemicals in contaminated soil biore-

mediation systems are indicated in Table 20.7.
Most chemical loss rates that are reported represent total chemical
loss and rarely are specific loss mechanisms identified. The losses can
be due to biodegradation, chemical degradation, hydrolysis, photolysis
and volatilization. While it commonly is assumed that microbial de-
gradation is the major loss mechanism, as noted in section 20.3.12,
volatilization also can be a loss mechanism for chemicals with a high
vapor pressure.
Since remediation cleanup goals usually are related to concentration
criteria (section 20.2), total loss rates are useful in determining the
feasibility of bioremediation processes.

Table 20.7 Chemical loss half-life values in con-

taminated soil bioremediation systems*

organic half-life values (tt)

(days) 2
~-~----~~-----

benzene 0.1-1.0
ethylbenzene 6.1
toluene 6.4
o-xylene 93.9
m, p-xylene 8.5-14.7
anthracene 9-53
benzo(a )anthracene 41-231
chrysene 5.5-116
1-methylnaphthalene 12.6
naphthalene 8-30
phenanthrene 23-69
pyrene 10-100

* Data from situations where relatively fresh refinery

wastes were being land treated and evaluated. Loss rates
of consitituents in residuals at closure will be slower than
constituents in fresh wastes.
 Toxicity reduction 537

20.3.14 Summary
Bioremediation processes can treat contaminated soils successfully. The
microorganisms in such systems will grow and metabolize the organics
present if environmental conditions are suitable. Aerobic conditions are
desirable and are accomplished by providing mixing and mechanical
aeration. Nutrients, such as nitrogen and phosphorus, may be needed.
Toxic conditions should not exist and pH and temperature should be
controlled as needed.
Many of the organics in contaminated soil have low water solubility,
which can limit degradation, and are tightly sorbed to the organic
matter in soil. The sorption constants and water solubilities of chemicals
can be estimated from known partition coefficients such as Kow and Koc.
Loss rates, rather than biodegradation rates, are commonly reported.
Such loss rates can be used to determine whether performance goals
can be achieved and the overall feasibility of a bioremediation process.
With an understanding of how the fundamentals described in this
section can relate to and affect the performance of bioremediation
processes, such processes can be technically and economically feasible.

20.4 TOXICITY REDUCTION

When an organic chemical is degraded in a bioremediation process, it

is transformed into other products through biological reactions with or
without complete detoxification and immobilization. Measuring the
loss of the parent compound does not assure that complete detoxi-
fication and immobilization occurs. Intermediate degradation products,
which may be more mobile and/or toxic than the parent compound,
may be generated as the parent compound degrades. Additional
information on the transformation and/or detoxification of a chemical is
necessary to establish that the loss of the parent compound leads to the
complete detoxification of the chemical or waste. Such information can
be obtained using either chemical or bioassay analyses.
Chemical analysis of detoxification products may yield information
about biochemical degradation pathways, but it is time-consuming
and expensive. Bioassays have been used successfully to demonstrate
detoxification of the applied waste in the soil (Matthews and Bulich,
1984; Matthews and Hastings, 1987), and are less expensive and time-
consuming. Such bioassays also have been used as a screening tool to
evaluate the soil treatment potential of a chemical or waste.
The following describes the reduction in toxicity and in chemicals
that occurred in a simulated contaminated soil biotreatment process
(Loehr, 1989; Desappa and Loehr, 1991). The study was done in lab-
oratory soil microcosms using phenol and eight chlorophenols.
538 Bioremediation of soils

The reduction of toxicity that occurred was evaluated by determining

the toxicity of the water soluble fraction (WSF) of the chemical/soil
mixture at the same sampling intervals used to obtain the degradation
data. The chemical compounds that can be extracted with water rep-
resent the potentially leachable fraction of the chemical or any inter-
mediate chemical detoxification products. The WSF of the chemical
poses the greatest threat to groundwater contamination. Hence,
evaluating the loss of the potentially leachable fraction of a chemical is
important.
The concentration of the parent chemical in the WSF also was deter-
mined. This concentration was expressed in terms of quantity of
chemical that was water extractable per kg of the soil. The toxicity
of the WSF was determined using the Microtox© method (Beckman
Instruments, 1982). To put the toxicity reduction data in perspective,
the WSF toxicity reductions, the WSF chemical concentration reduc-
tions and the soil chemical concentration reductions were compared.
The loss of the chemical in the soil and in the WSF and the WSF
toxicity reduction data were evaluated using the first-order kinetic data
(half-lives) that resulted from the microcosm degradation studies.
Figure 20.1 compares the half-life of the chemical in the soil and the
WSF chemical loss for all of the nine chemicals tested. The correlation

30

If)
If) c
0 __
...Jill 20 c
...J>-
«III
u8.
-(I)
c
:E:::
W...J y =-1.4 + 1.47x
:I:,.!.
UCij c r =0.87
...J:I: 10
0
If)

0
0 10 20 30

WSF CHEMICAL LOSS

Half-life (Days)

Fig. 20.1 Comparison of chemical loss in the soil and in the WSF for phenol
and 8 chlorophenols, after Dasapa and Loehr (1991).
 Toxicity reduction 539
shows that, in general, the soil chemical half-life was about 1.5 times
greater than the WSF chemical half-life. For these chemicals, this
indicates that the loss of the chemical in the WSF was faster than the
loss of the chemical in the soil. The correlation also indicated that no
enhanced mobilization of applied chemical occurred as the degradation
and detoxification took place.
Figure 20.2 compares the WSF chemical loss and the WSF toxicity
reduction for all nine tested chemicals. The high correlation suggests
that the WSF toxicity can be attributed to the target chemical concen-
tration in the WSF and that no water extractable toxic intermediate
products were formed. Thus, these chemicals were detoxified in the
soil.
These results indicate that contaminated soil bioremediation pro-
cesses not only reduce the concentration of organic chemicals originally
present, they also reduce the loss of chemicals in the water soluble
fraction in the soil and, based on one bioassay method, reduce the
apparent toxicity of the water soluble fraction. Thus, bioremediation
processes are chemical degradation and toxicity reduction processes.

[] Y = 2.05 + 0.95x
r = 0.90

10 20

WSF CHEMICAL LOSS

Half-Life (Days)

Fig. 20.2 Relationship between WSF chemical loss and toxicity reduction in
the WSF, after Dasapa and Loehr (1991).
540 Bioremediation of soils

20.5 PROCESS DESCRIPTIONS

As noted in section 20.1, bioremediation processes can be categorized

as solid and slurry phase processes. The solid phase processes include:
• surface soil and on-site modifications of land treatment; and
• a modification of the basic compo sting process.
Slurry phase processes include liquid-solids systems that can be used
in existing impoundments or in tanks. The following describes each
process and potential applications.

20.5.1 Solid phase land treatment

Land treatment is a managed technology that involves controlled
application of a waste on the soil surface and/or the incorporation of
the waste or contaminated soil into the upper soil zone. It is not the
indiscriminate dumping of waste on land, and it is not landfilling.
Land treatment technology relies on the dynamic physical, chemical
and biological processes occurring in the soil. As a result, the con-
stituents in the applied wastes are degraded, immobilized or trans-
formed to environmentally acceptable components.
The design and operation of a land treatment facility is based on
sound scientific and engineering principles as well as on extensive
practical field experience. A land treatment site is designed and
operated:
• to maximize waste degradation and immobilization;
• to minimize release of dust and volatile compounds as well as
percolation of water-soluble waste compounds; and
• to control surface water runoff.
Land treatment can be a viable management practice for the treatment
and disposal of hazardous and non-hazardous wastes as well as for
bioremediation. Land treatment has been successfully practiced in all
major climatic regions of the United States, Europe, and Canada under
a wide range of hydrogeologic conditions. In the US, approximately
200 industrial land treatment systems have been in operation, includ-
ing over 100 at petroleum refining facilities.
Both surface soil and on-site land treatment processes are among the
more widely used bioremediation technologies.

Surface soil land treatment bioremediation

This involves:
• keeping contaminated surface soils in place;
 Process descriptions 541

• if needed, adding nutrients to assure adequate biodegradation and

adjusting the pH toward neutral conditions;
• tilling the soil to increase the availability of oxygen and nutrients to
the soil microorganisms; and
• possibly irrigating to assure adequate moisture for microbial
degradation.
The organisms involved in the degradation are indigenous organisms
unless none that can degrade the organics exist in the soil due to prior
toxicity. In such cases, the toxicity needs to be reduced by:
• adding uncontaminated soil; or
• adjusting the environmental conditions and adding acclimated
organisms.
The rate and extent that the organics in the soil are lost will depend
on the combined effect of volatilization, sorption and biological
degradation.
Both degradation and immobilization occur in a land treatment unit
with most of the residual organics and metals remaining in the upper
layers of the soil, essentially in or near the zone mixed by tilling. Figure
20.3 illustrates typical patterns of organics and metals in the soil from a
land treatment unit that had treated petroleum refining wastes and
residues for several decades.
Land treatment units also minimize the leaching of constituents
applied to the unit or in the contaminated soil being bioremediated.
Figure 20.4 illustrates the potential mobility of metals in soils, as
measured by the toxicity characteristic leaching procedure (TCLP), at
various depths at an industry land treatment unit. At well managed
land treatment units, the pattern indicated in Fig. 20.4 is typical, i.e.,
the applied wastes are immobilized and zero to very low concentra-
tions of the constituents in the surface soils are mobile and leach.
Site runoff must be contained and managed at all waste treatment
operations. The runoff water which is collected may be allowed to
evaporate on site, or it may be discharged to a wastewater treatment
system, or may be reapplied to the land treatment site.
Note also that surface soil land treatment systems do not have a
liner.

On-site land treatment systems

An on-site land treatment system does have a liner and is a constructed
unit which contains contaminated soil or clean soil and wastes being
bioremediated and has walls, a drainage system and a leachate collec-
tion system. It generally is above ground and allows more complete
control of the process. An example of such a system is presented in
Fig. 20.5.
 CONCENTRATION (mg/kg dry weight)
100 200 300 400 500

(i)
Q)
.c
(.)

..:. 12
w
U
ita:
~
en 24

3:
o
...J
W
III • Dibenzo (a,h) anthracene
J: 36 • Benzo (a) pyrene
l-
e.. • Chrysene
w
o

48

0-6

(i) 6 - 12
Q)
.c

-
(.)
I: 12 -24

J:
I- 24 - 36
e..
W
0
36-54

54 -72

OIL AND GREASE CONCENTRATION (mg/kg)

Fig. 20.3 Soil cores at an active hazardous waste land treatment site: (a)
average PAH concentrations as a function of depth; (b) oil and grease (mg/kg)
as a function of depth; after Loehr et al. (1990).
 Process descriptions 543

ZINC CONCENTRATION IN TCLP EXTRACT (mg/l)

0 2 3 4
0

12
Cii'
Q)
..c:

-
0
c 24
. .J
(5
(f) 36
U.
0
::I:
I- 48
a.
W
Cl
60

72

Fig. 20.4 Soil cores at a hazardous waste land treatment site: zinc con-
centration in the TCLP extracts; after Loehr et at. (1990).

Trealmenl Zone
System Area Overhead
Leachate CollectIon Syslem
Enclosure (Optional)
SynlheUc Line,
Spray Nozzle Compacted Secondary 5001 Loner

Earth Filled Berm Leachate

10
"
Existing Ground

Fig. 20.5 Schematic of an above ground solid phase bioremediation unit.

An on-site system usually is constructed adjacent to the sailor

wastes requiring bioremediation to minimize transportation costs and
to provide better technical and managerial control of the process.
Except for the fact that the on-site unit is constructed above ground,
the fundamentals and operation of the on-site unit are the same as for
the surface soil unit.
Both surface soil and on-site land treatment systems have been used
successfully to bioremediate:
 544 Bioremediation Of soils
1200
E v: Reapplication

,
a. V
.e: 1000 ~
/ofWaste

Z
0 V- I ~ PlOt 1 I
I0
800
i=
~
Initial Waste ~
Application V-
PIOl2
I
c:
~
Z
600 , /

w ~
U 400
~ V
Z
0 ~ ~
U
200
r;:; ~
-l
0C/) ~ ~
0
~
· 68
dlBl.
·47 -16
V-
o
~[?h.--,
36 139 222 250 294 317 362 398

TIME (Days)
(a)

- 500
E Reapplication
a. 450
.e: /ofWaste
Z
0
400
',/
, I E3o Plot 1
Plot 2
I
350
i=
~ 300
c: r;:;
I- 250
Z
W V
U 200
Z
150
~
0
U ~
-:
100 T
-l
:l
~ :l :o ~l- : I Bll
',/
0
C/)
SO -
0
r::h ,..,n v:n ~ .,..""".

- 47 -16 36 139 222 250 294 317 362 398

TIME (Days)
(b)

Fig. 20.6 Loss of (a) anthracence and (b) benzo(a)antracene, in a field land
treatment soil bioremediation unit.

• contaminated soils at spill sites;

• industrial wastes and residues; and
• soils at surface impoundments and lagoons that are being closed.
Land treatment has been an effective bioremediation process for
contaminated soils. Figure 20.6 illustrates field data that resulted from
the land treatment bioremediation of wood preserving waste contami-
nated soil and sludge. At time zero in these plots, a second application
of the sludge was made and the loss of PAH compounds was moni-
tored over time. The loss rates (t1) of the PAH compounds in these
2
 Process descriptions 545

Table 20.8 Loss rates of PAH compounds from the bioremediation of wood
preserving waste contaminated soil and sludges using land treatment

compound half-life compound half-life

(t1 days) (t1 days)
2 2

naphthalene 95 chrysene 95
acenaphthylene 105 benzo(b )fluoranthene 290
acenaphthene 40 benzo(k)fluoranthene 140
phenanthrene 25 benzo(a )pyrene 100
anthracene 35 dibenzo(a,h)anthracene 260
fluoranthene 40 benzo(g,h,i)perylene 360
pyrene 40 ideno(1,2,3-cd)pyrene 600
benzo( a )anthracene 50

plots are summarized in Table 20.B. The results are indicative of the
type of losses that occur under field situations.

20.5.2 Compo sting

Compo sting as a bioremediation process for contaminated soils and
industrial wastes and residues is similar to the process used for com-
posting of leaves, garbage and food processing residues. The main
difference is that high temperatures rarely are achieved in contami-
nated soil bioremediation composting and the purpose is degradation
and loss of specific organic compounds rather than stabilization to
produce a mulch or soil conditioner. At the completion of a bioreme-
diation compo sting process, the treated material must be disposed of in
an environmentally sound manner. However, both degradation and
immobilization have occurred in this bioremediation process and the
composted material should not cause surface or groundwater problems
at its ultimate disposal site.
The essential elements of compo sting are the same as any biore-
mediation process:
• moisture,
• aeration,
• acclimated organisms,
• satisfactory carbon-nitrogen-phosphorus balance, and
• non-toxic conditions.
The characteristics of contaminated soils or residues considered suit-
able for compo sting are:
• constituents able to be lost by volatilization or degradation, or
immobilized in the system;
546 Bioremediation of soils

Black Plastic
Cover - -........., ,.._---r---- Air Input
"" , . . - - - Water Addition
~--::::-r---;:::r--..."t
Side

Support Posts
Compost
Pile

Grade -"""7'1.--:-::-LU.~.~~.,~."~.~,,
............. ~~ill;im;~~rti:;:~8~"~B~ed of Wood Chips
~

~'" I
71;::::.// I I I J/~,"'.
'Black Plastic Liner
',\ I I I I ..
LJ LJ

Fig. 20.7 Schematic cross-section of a compost type solid phase soil

bioremediation unit.

• a low amount of free liquid so that aerobic conditions can be

maintained;
• a high ratio of inert solids to biodegradable organics; and
• a mixture that can be easily broken up by mechanical turning and/or
is porous to allow air to move through the compo sting solids.
Typical composting systems that can be used for bioremediation are
the windrow, the Beltsville and the in-vessel systems. The windrow
system is an open system, with periodic turning of the compost mix
pile and no forced air. The Beltsville system is an open pile with an air
distribution system under the pile. Air is sucked through the pile from
the atmosphere and exhausted through a blower generally to an air
pollution control system. The in-vessel system occurs in a closed or
unclosed vessel in which mechanical mixing may occur and/or air is
forced through the mixture by blowers. Bulking agents commonly are
added to increase the porosity and assist the flow of air to maintain
aerobic conditions.
For bioremediation composting systems, adequate operational
control as well as control of all emissions, such as leachate and off-
gases, is desirable. In-vessel systems provide such control. An example
of such a system is presented in Fig. 20.7. If gaseous emission is not a
concern, then a windrow or Beltsville system with positive leachate
control can be satisfactory.
Bioremediation compo sting systems commonly are constructed and
operated on-site so as to minimize transportation costs. The systems
can be operated as continuous or batch units. Contaminated soil or
residues are excavated from the site or wastes and residues are added
periodically for treatment and remediation. After meeting specified
clean-up criteria, the remediated mixture may be returned to the site or
disposed of in a landfill or other acceptable location.
 Process descriptions 547

Liquid-Solids
Mixture .. Treated Liquid

. REACTOR
Treated Solids

I
Air/O xyg en
I Offgas
Fig. 20.8 Schematic process flow diagram of a single reactor liquid-solids
bioremediation unit.

Composting has been used to remediate soils contaminated with

diesel fuel and similar petroleum products.

20.5.3 Slurry phase processes

Liquid-solids treatment systems are slurry phase bioremediation
systems operated to maximize mass transfer rates and contact between
contaminants and microorganisms capable of degrading the contami-
nants. Because solids can be treated rapidly in contained reactors,
much less area is needed for this on-site remediation process than for
land treatment. An advantage of liquid-solids systems using tanks is
that regulatory concerns related to the land disposal of hazardous
wastes or to the contamination of groundwater may be eliminated.
Liquid-solids contact treatment is analogous to conventional bio-
logical suspended growth treatment (e.g., activated sludge). These
units are designed to relieve the factors commonly limiting microbial
growth and activity in soil, principally, the availability of carbon
sources, inorganic nutrients and oxygen. To achieve this goal, the
wastes are suspended in a slurry form and mixed. The mixing and
aeration also prevent oxygen transfer limitations. Mixing can be pro-
vided by aeration alone or by aeration and mechanical mixing. Aeration
is provided by floating or submerged aerators, or by compressors
and sparges. Chemicals added to liquid slurry reactors can include
nutrients and neutralizing agents to relieve limitations to microbial
activity.
Liquid-solids systems are a relatively new approach for the reme-
diation of contaminated organic materials. There have been several
applications of these systems for the treatment of wood treating, coal
tar and petroleum wastes. The most frequently used process con-
figuration is a simple batch reactor system (Fig. 20.8). The reactor can
be an above-ground tank or an existing pit or lagoon.
After the bioremediation clean-up goals are achieved in a batch
reactor, the aeration and mixing are discontinued and the solids are
allowed to settle. The treated liquid is decanted and discharged while
the treated solids can be further treated by conventional land treatment
548 Bioremediation of soils

or disposed of on-site. If the process is used in a lagoon or surface

impoundment, it may be possible to decant the liquid and leave the
treated solids in place.
Liquid-solids systems are suitable for contaminated slurries, such as
may be in pits, ponds and lagoons, and contaminated soils which are
too heavily contaminated for direct treatment by conventional land
treatment. The technology will treat polyaromatic hydrocarbons,
naphthalene, phenols, benzene, toluene, xylene and ethylbenzene.
The degree of treatment of anyone chemical is a direct function of its
solubility in water and its rate of biodegradation.
Because of the mixing and aeration in these systems, volatilization
is a significant factor in the removal of organics. Monitoring and
management of the gaseous emissions may be necessary at locations
where air quality concerns exist.
The phenomena which govern the process are mass transfer of
organics from the solid phase to the aqueous phase and biodegradation
of the aqueous phase organics. The design considerations important to
properly apply these systems are:
• physical characteristics of the liquid-solids feed, particularly the
fraction of the mix which is organic, the distribution of organics
within the mixed slurry, and the viscosity and surface tension of the
hydrocarbon phase;
• the nutrient, temperature and oxygen requirements to achieve the
optimal biodegradation;
• solids and hydraulic residence times to perform adequate treatment
and meet the clean-up goals;
• the characteristics of the offgas; and
• the degree of mixing (energy) required to keep the solids in
suspension.
A liquid-solids reactor is one part of a slurry phase treatment system.
Consideration must be given to how the soil or sludge will be removed
from the source and transmitted to the reactor, if pretreatment such as
thickening is desirable, the emission controls that may be needed, and
what steps may be needed to handle and dispose of the treated slurry.
The process options are indicated in Table 20.9.

20.6 SUMMARY

There are several bioremediation systems that can be used for con-
taminated soils and slurries. The benefits of such systems are:
• degradation and immobilization of organics;
• toxicity reduction;
 References 549

Table 20.9 Soil slurry treatment process options

process step process options

Excavation and Transport dredge, earth moving equipment

Pretreatment thicken or dilute
Emission Controls carbon adsorption, biofiltration, combustion,
recirculation of off-gas to slurry reactor
Slurry Treatment batch, semicontinuous, continuous
Post-Treatment land treatment, coagulate/flocculate, thicken, filter
press, solidify, landfill

• source control and pollution prevention; and

• risk reduction.
Both solid phase and slurry phase processes are available and are
being used with such materials. These processes can be used success-
fully if the fundamentals are understood and incorporated into design
and operation. The costs of these systems generally are lower than
alternative remediation methods. Because of the above benefits and the
lower costs, bioremediation systems are an important technology for
remediation and clean-up of specific sites.

REFERENCES

Beckman Instruments, Inc. (1982) Microtox© System Operating Manual,

Beckman Instruments, Inc., Carlsbad, California.
Dasappa, S.M. and Loehr, R.c. (1991) Toxicity reduction in contaminated soil
bioremediation processes, Water Research, 25, 1121-30.
Dibble, J.T. and Bartha, R. (1979) Effect of environmental parameters on the
biodegradation of oil sludge, App. and Environ. Microbial., 37, 729-39.
Huddleston, R.L., Bleckmann, c.A. and Wolfe, J.R. (1986) Land treatment
biological degradation processes, in Land Treatment: A Hazardous Waste
Management Alternative, (eds R.C. Loehr and J.F. Malina, Jr.) Water
Resources SympOSium No. 13, Center for Research in Water Resources,
The Univ. of Texas at Austin, Austin, Texas.
Loehr, R.c. (1989) Treatability Potential for EPA Listed Hazardous Wastes in Soil,
Robert S. Kerr Environmental Research Laboratory, USEPA, Ada,
Oklahoma, PB 89-166-581AS, National Technical Information Service,
Springfield, VA.
Lyman, W.J., Reehl, W.F. and Rosenblatt, D.H. (1982) Handbook of Chemical
Property Estimation Methods, McGraw-Hill Book Company.
Matthews, J.E. and Bulich, A.A. (1984) A toxicity reduction test system to
assist in predicting land treatability of hazardous organic wastes, in
Hazardous and Industrial Solid Waste Testing: 4th Symp., (eds J.K. Petros, Jr.,
W.J. Lacy and R.A. Conway) Philadelphia, ASTM/STP 886.
550 Bioremediation of soils

Matthews, J.E. and Hastings, L. (1987) Evaluation of toxicity test procedure for
screening treatability potential of waste in soil, Toxicity Assessment 2, 265-
81.
Sims, J.L., Sims, R.c. and Matthews, J.E. (1989) Bioremediation of Contaminated
Surface Soils, Robert S. Kerr Environmental Research Laboratory, USEP A,
EP A/600/9-89/073, Ada, Oklahoma.
Tetra Tech, Inc. (1989) MYGRT code version 2.0: an IBM code for simulating
migration of organic and inorganic chemicals in groundwater, EPRI EN-6531,
Final Report, Project 2879-2, Electric Power Research Institute, Palo Alto,
CA.
USEPA (1988) Survey of State Programs Pertaining to Contaminated Soils, Office of
Underground Storage Tanks, Washington, DC.
USEP A (1990) Selected Data on Innovative Treatment Technologies for Superfund
Source Control and Ground Water Remediation, Technology Innovation Office,
Washington, DC, August.
 CHAPTER 21

In situ bioremediation
of groundwater
Gaylen R. Brubaker

21.1 INTRODUCTION

On October 18, 1989 the United States Environmental Protection

Agency (USEPA) issued a memorandum which stated that 'pump-and-
treat' technology is incapable of achieving the agency's groundwater
remediation objectives. The reasons cited for this limitation included:
1. heterogeneity of the contaminated regions;
2. failure to control continued contamination from source zones;
3. inadequate system design; and
4. fundamental limitations of the process to remove occluded and
adsorbed-phase material (Superfund Report, 1989).
As the limitations of the traditional 'pump-and-treat' groundwater
treatment processes have become more broadly recognized (MacKay
and Cherry, 1989; Keely, 1989), scientists and engineers have sought
enhancements and alternatives which would provide more rapid and
reliable remediation of contaminated aquifers. In situ bioremediation is
an increasingly popular groundwater remediation technique, designed
to remove occluded and adsorbed-phase contaminants and, thereby,
provide more aggressive groundwater remediation.
The most common organic contaminants found in groundwater are
chlorinated hydrocarbons (trichloroethene, trichloroethane, dichloro-
ethene), benzene and its derivatives (toluene, ethylbenzene, and
xylene), and polyaromatic hydrocarbons (naphthalene, phenanthrene,
etc.). Since all of these compounds have very low aqueous solubilities
and prefer to be sorbed onto the natural organics in soil, rather than be

Geotechnical Practice for Waste Disposal.

Edited by David E. Daniel.
Published in 1993 by Chapman & Hall, London. ISBN 0 412 35170 6
552 In situ bioremediation of groundwater
dissolved in water, it is very difficult to extract these contaminants
from a soil/groundwater matrix. The aqueous solubilities, octanol-
water partition coefficients, and typical retardation factors for a few of
the more common groundwater contaminants are summarized in Table
21.1.
As the data in this table illustrate, the aqueous solubility of many
common organic contaminants are less than 0.1 % (1000 mg/l) and
many are below 1 mg/l. In addition, these chemicals prefer to bind to

Table 21.1 Physical/chemical properties of selected groundwater contaminants

compound K,W Koc ~ R solubility

(mgll)
-~------ ..

Volatile aromatic hydrocarbons

Benzene 135 85 0.9 5.3 1780
Ethylbenzene 490 309 3.1 16.4 515
Toluene 1412 890 8.9 45.5 140
o-Xylene 589 371 3.7 19.6 175
Non-volatile aromatics hydrocarbons
Acenaphthene 10000 6300 63.0 316 3.42
Acenaphthylene 5000 3150 31.5 159 3.93
Anthracene 28100 17703 177.0 886 0.045
Di-n-butylphthalate 158500 99855 998.6 4993.8 400
Chlorobenzene 692 436 4.4 22.8 500
Chrysene 407000 256410 2564.1 12822 0.0018
p-Cresol 83 52 0.5 3.6 24000
p-Dichlorobenzene 2450 1544 15.4 78 49
Fluoranthene 79400 50022 500.2 2502 0.206
Fluorene 15800 9954 99.5 499 1.69
Naphthalene 1778 1120 11.2 57 31.7
Phenanthrene 28800 18144 181.4 908 1
Pyrene 75800 47754 477.5 2389 0.132
Volatile chlorinated hydrocarbons
Carbon tetrachloride 436 275 2.7 14.7 800
Chloroform 93 59 0.6 3.9 8000
1,1-DCA 62 39 0.4 3.0 5500
1,2-DCA 30 19 0.2 1.9 8690
trans-DCE 30 19 0.2 1.9 800
PCE 398 251 2.5 13.5 150
1,1,1-TCA 150 95 0.9 5.7 4400
1,1,1-TCE 195 123 1.2 7.1 1100
Non-chlorinated solvents
Acetone 0.58 0 0.0 1.0 miscible
Methylethylketone 1.82 1 0.0 1.1 miscible
Ethanol 0.18 0 0.0 1.0 miscible
Ethylene glycol 0.01 0 0.0 1.0 miscible

Koe = organic carbon partition coefficient; Kd = soil water partition coefficient (assumes foe
= 1%); and R = retardation coefficient (assumes j;,e = 1 %).
 Introduction 553

the natural organics within soils rather than to dissolve in water. This
preference is illustrated by Koc and Kd • As a result of these properties,
even after small portions of these molecules dissolve in groundwater,
they move through the matrix at a small fraction of the rate of ground-
water (l/R) because of continuous retardation by the soil. This results
in a very inefficient remediation process for many organic contami-
nants using a pump-and-treat process.
The in situ bioremediation process has been developed to enhance
the performance of the pump-and-treat process by injecting water
soluble nutrients which will stimulate the natural biodegradation of
organic contaminants within the aquifer matrix.
In situ bioremediation involves the introduction of microbial nutrients
(typically ammonia and orthophosphate) and an oxygen source into
a contaminated aquifer to promote the biodegradation of organic
materials within this 'bioactive zone'. Groundwater is withdrawn from
the aquifer to establish hydraulic containment and to increase the
hydraulic gradient, thus increasing the rate at which groundwater
moves through the formation. The captured water is typically treated
to remove contaminants, and then a major portion of the water is
amended with nutrients and reinjected. The basic process is illustrated
in Fig. 21.1.
In order to design successfully an in situ bioremediation process at a
particular site, it is helpful to understand how the general principles of

GROUNDWATER • · 3
TREATMENT NIt. . PO •• 0 2

~-
- -----'

NEW WATER TABLE

ORIGiNAl WATER TABLE

Fig. 21.1 III situ bioremediation of contaminated aquifer.

554 In situ bioremediation of groundwater

microbiology, hydrogeology and chemistry apply to the process. For

example:
• the site must contain bacteria which are capable of degrading the
contaminants of interest;
• the hydrogeology of the site must permit the movement of nutrient-
enriched water through the contaminated region and allow for the
controlled capture of this water; and
• the soil and groundwater must be chemically compatible with the
nutrient sources provided.
The significance of each of these issues will be discussed further in
the following sections. Since metals are not biodegradable, this dis-
cussion will be limited to organic contaminants.
A comprehensive literature review (Lee et al., 1988) and several
review articles (Thomas and Ward, 1989; Wilson et al., 1986) have been
published which provide summaries of in situ bioremediation tech-
nology. However, most of the case histories and more design-oriented
articles have appeared in conference proceedings or other publications
with limited distribution. As a result, it has been difficult to assemble
sufficient information to gain an overall perspective on this emerging
technology. This chapter is intended to fill that gap.
Section 21.2 is a review of early in situ bioremediation projects,
presented as an introduction to the technology and to place the earlier
literature in historical perspective. Advances in oxygen delivery are
discussed in section 21.3. Section 21.4 presents a very brief overview of
some of the relevant principles of microbiology. This is followed, in
section 21.5, by a discussion of the microbial, hydrogeological, and
chemical site characteristics which effect the technical feasibility and
design of the process. This section also includes a discussion of lab-
oratory treatability techniques, as well as design principles used as a
project moves through the design phase. The final portion of the
chapter looks to the future, with a very brief summary of six areas of
research and development that are likely to expand dramatically the
use of in situ bioremediation.

21.2 EARLY IN SITU BIOREMEDIATION PROJECTS

21.2.1 Ambler, Pennsylvania

The first field implementation of an in situ bioremediation process was
initiated in 1972 at a site in Ambler, Pennsylvania (Raymond et al.,
1976). An oil pipeline owned by Sun Oil Company had leaked, spilling
over 490 000 L of high-octane gasoline within the pumping zone of
Ambler's municipal well. Over the next eight months, numerous
 Early in situ bioremediation projects 555
monitoring and recovery wells were installed and almost two-thirds
of the spilled gasoline was recovered from the fractured dolomite
formation. However, further recovery efforts proved unproductive,
leaving an estimated 150000 L of gasoline in the formation.
Since the groundwater pumped from the site contained an average
of 5 ppm of dissolved gasoline, remediation through groundwater
pumping would have required almost 100 years of treatment, even if
the city pumped more than a million liters of water per day. A group of
scientists at Sun's Marcus Hook Pennsylvania Research Center were
called in for consultation. They had been studying the factors which
controlled the microbial degradation of petroleum products under
various conditions and were convinced that the gasoline could be
degraded within an aquifer, if a proper blend of microbial nutrients
could be provided.
A concentrated solution of ammonium sulfate, disodium phosphate,
and monosodium phosphate was periodically introduced into injection
wells at the site. The blend was designed to provide the desired
quantities of ammonia and phosphate, and to buffer the water at a pH
of about 7. In addition, air was introduced into 10 monitoring wells by
using air compressors and gas diffusers. No water was injected in this
first process, but the three recovery wells did draw the nutrients
through the contaminated area. Although there is little direct per-
formance data from this site which conclusively demonstrates the
biodegradation of gasoline, the process did appear to accelerate the
reduction in gasoline concentrations. In addition, two sets of observa-
tions were made which suggested enhanced biodegradative processes.
First, the aeration process increased the average dissolved oxygen
content in the recovered water by 2-3 ppm, and second, the average
population of gasoline-degrading bacteria increased by a factor of 100
in those areas where nutrients were added.

21.2.2 Other early projects

Following the encouraging results of the Ambler project, this process
was used at approximately 20 other sites over the next 10 years, typically
service stations near drinking water aquifers where rapid and complete
cleanup was required. Although it is difficult to gauge the 'success rate'
of these projects using today's standards, it appears that, in most
instances, the concentrations of dissolved contamination and/or
hydrocarbon vapors were reduced faster than would have been ex-
pected using traditional groundwater capture strategies (Du Pont, 1990;
Lee et al., 1988).
By the early 1980s, several other groups had implemented similar
processes for in situ bioremediation. Groundwater contamination from
methylene chloride, acetone, n-butyl alcohol and dimethylaniline was
556 In situ bioremediation of groundwater

the motivation for a combined flushing/in situ strategy in Waldwick,

New Jersey. The contaminated aquifer, located 2-5 m below grade,
was flushed with water, and then the recovered water was treated in a
surface bioreactor. Two trenches were used for water injection and
nine wells for air injection while the recovered water was treated in
aerated bioreactors on the surface, prior to reinjection. The nutrients
from the surface bioreactor, plus the in situ aeration, thus provided the
conditions for subsurface, as well as surface biodegradation (Jhaveria
and Mazzacca, 1985).
The literature also contains other case histories in which surface
biological processes were combined with in situ soil flushing processes.
One example of this strategy was employed at an ethylene glycol spill
at the Naval Air Engineering Center in Lakehurst, New Jersey (Flatham
and Caplan, 1986). Approximately 15 000 L of cooling water containing
25% ethylene glycol leaked from a lined surface lagoon. An area
measuring 14 m x 55 m was identified that had groundwater con-
tamination in excess of 1400 ppm. Bench-scale studies showed that
these concentrations were not toxic and that indigenous bacteria
from the site could degrade the contamination, under proper pH and
nutrient conditions.
A series of recovery wells was used to withdraw contaminated
groundwater from the site. This water was treated in an activated
sludge bioreactor which was mobilized on-site. The treated water then
passed through a settling chamber and was reinjected, through five
reinjection wells, to flush more contamination from the site and pro-
vide nutrients and dissolved oxygen for subsurface biodegradation.
This strategy has also been used for gasoline, formaldehyde and mixed
solvent spills, though in most of these projects, no special effort was
made to provide additional oxygen to the subsurface environment.
Although oxygen addition is not important when flushing contami-
nants such as alcohols or ketones, which are easily extracted from soil,
it is a critical component of the in situ biodegradation process when the
contaminants of concern are mostly sorbed within the soil matrix. As a
result, this author will reserve the term 'in situ bioremediation' to
processes where the rate of biodegradation is being directly stimulated
in the subsurface through the addition of an oxygen source (or alter-
native electron acceptor), as well as nutrients and/or bacteria.

21.3 ADVANCES IN OXYGEN DELIVERY

21.3.1 Laboratory studies

In 1982, the American Petroleum Institute (API) funded a study to
enhance the oxygen delivery for in situ bioremediation processes (Texas
 Advances in oxygen delivery 557
Research Institute, 1982). The first phase of this project, a literature
review and analysis, suggested that hydrogen peroxide was the pre-
ferred alternative. Hydrogen peroxide is used as an oxidizing agent in
many industrial applications and is often used as a backup source of
oxygen for industrial waste treatment systems. Since one kilogram of
hydrogen peroxide decomposes to form nearly one-half kilogram of
oxygen, and hydrogen peroxide is totally miscible with water, it is
an environmentally attractive alternative. These benefits seemed to
outweigh its potential toxicity to bacteria, the need to store and handle
peroxide in specialized equipment, and its cost.
A second phase of the study examined the toxicity of various con-
centrations of peroxide to gasoline-degrading bacteria. These studies
showed that, although small populations of bacteria were sensitive to
as little as 100 ppm of peroxide, more mature cultures could tolerate
over 1000 ppm.
FMC Corporation, a producer of hydrogen peroxide, became in-
terested in this new potential application of peroxide in 1983 (Brown
et al., 1984). FMC performed a series of column studies to examine the
degradation of gasoline in the presence of various concentrations and
sources of dissolved oxygen (DO). The test conditions included water
saturated with air (8ppm DO) and oxygen (40ppm DO), as well as
water amended with 250 ppm (118 ppm DO equivalents) and 400ppm
(200ppm DO equivalents) of hydrogen peroxide. Several important
observations were reported.
• Oxygen consumption was rapid under all test conditions.
• After two weeks, the number of total and gasoline-degrading
bacteria increased as the oxygen delivery increased. The 200 ppm
DO columns had greater than 100 times more bacteria than those
which received 8 ppm.
• The loss in gasoline, accounted for through biodegradation, was
proportional to the mass of oxygen supplied.
FMC then proceeded to inject a solution of hydrogen peroxide (the
concentration was not reported) into a monitoring well at an active
bioremediation site for 22 hours. Observations were made of the
changes in dissolved oxygen in a second monitoring well 8 m down-
gradient. Significant increases in DO were observed within 10 hours,
increasing to 15 ppm over the next 72 hours. Although they did not
measure microbial activity in this instance, transport of dissolved
oxygen from peroxide was demonstrated.

21.3.2 API field demonstration

Based on the increasing evidence that hydrogen peroxide would pro-
vide a significant advantage over simple aeration of aquifers, the
558 In situ bioremediation of groundwater
American Petroleum Institute (API) funded a field demonstration of
the technology to improve documentation of basic design and per-
formance issues (API, 1987; Brown and Norris, 1986; IT Corporation,
1986).
The site of interest was a petroleum products terminal in northern
Indiana. In 1979, a major leak resulted in percolation of unleaded
gasoline through the base of the bermed tank enclosure and down
through sandy soil, until it reached the water table 7m below grade.
Three recovery wells were installed, soon after the spill, and the phase-
separated product was collected. However, the soil below the original
zone of contamination, and near the water table downgradient, con-
tained significant petroleum contamination. The groundwater varied
seasonally between about 7 and 8 m below grade and had a relatively
flat surface.
The test area was designed to consist of two concentric triangular
injection galleries, centered around an existing recovery well. The
inner gallery was designed to receive water enriched with nutrients
and hydrogen peroxide, while the outer gallery was designed to
receive clean water only, thus isolating the test area hydraulically from
the surrounding contamination. The interior and exterior galleries were
constructed about 1.5 m deep and measured approximately 14m and
27 m on a side, respectively. The pumping wells in the center of the
treatment area were capable of yielding over 280 IImin (Fig. 21.2).
The site soils consisted of glacially-deposited sands and gravels.
Although the site was initially identified as 'relatively homogeneous
fine to medium grained sands', this assessment was amended, as
initial coring data was gathered to generate baseline data for the bio-
degradation process. These borings indicated significant variability in
the texture, ranging from tight, fine sands to loose, coarse gravels. In
addition, several discontinuous layers of fine sands containing higher
silt fractions were identified in the upper portions of the aquifer,
roughly 1.5-2 m below grade.
The presence of the silt zone directly below the infiltration galleries
had not been anticipated in the original design. As a result, the rate of
groundwater injection was limited to 40-60 Llmin, as opposed to the
200-300 Llmin range originally planned. In addition, precipitation of
minerals (iron and calcium phosphates) appeared to further reduce the
rate of groundwater injection. After several months of operation, the
injection design was altered to increase the amount of water (and thus
nutrients) that could be added to the site, thus increasing the rate of
biodegradation that would be observed during the test. First, nutrients
were added to both the outer and the inner injection galleries, and
then six injection wells were installed (two along the side of each outer
triangle). As a result of these modifications, a total of 2.9 million liters
of water containing 7700 kg of hydrogen peroxide, 4700 kg of am-
 Advances in oxygen delivery 559

• •
00 o

o MONITORING WELL
o
• INJECTION WELL

Fig. 21.2 Plan view of API biodegradation field demonstration.

monium chloride and 5500 kg of sodium phosphates and a variety of

micronutrients were injected over the 24-week test.
The hydrogen peroxide was added continuously, initially at 100 ppm
and then slowly increasing to 500 ppm, while the ammonium salts and
phosphates were added in concentrated batches, two to three times per
week. Throughout the test, groundwater from 15 observation wells
within the site was analyzed weekly for nutrients, total and gasoline-
degrading bacteria, dissolved oxygen, and pH. In addition, extensive
soil coring was performed.
Soil samples were collected at 13 locations at the beginning of the
test, 6 locations at week 5, 8 locations at week 16, and 19 locations at
the end of the study. In the first three corings, samples were collected
at three separate depths; a fourth depth was added for the final round
of sampling. Samples were analyzed for petroleum hydrocarbon, total
heterotrophic bacteria, gasoline-degrading bacteria, and nutrients. The
conditions within the inner gallery, prior to and during the test, are
summarized below. These data shown in Table 21.2, representative of
a large mass of data from this project, provided the most convincing,
documented support for in situ biodegradation to date.
560 In situ bioremediation of groundwater
Table 21.2 Conditions within inner gallery at treatment site

before treatment post treatment

ammonia (ppm) <1 100-250

orthophosphate (ppm) <1 100-250
heterotrophic bacteria (cfu/g*) 1.2 x 103 1.4 x 107
gasoline-degrading bacteria (cfu/g*) 2.0 X 102 1.1 X 106
petroleum hydrocarbon (kgs) 1900 ± 380 540 ± 230

* Colony forming units: a standard method of counting bacteria per gram of soil.

The existence of areas which had received only nutrients during a

portion of the test, and the existence of areas of low hydraulic con-
ductivity in the southern portion of the site, generated a large amount
of data allowing for a number of other important observations.
• Areas which received the most nutrients (and oxygen) showed the
highest increase in bacterial populations and the greatest decrease in
petroleum hydrocarbon.
• Even though degradation was documented throughout the site, the
concentration of dissolved oxygen in the monitoring wells remained
low «2 ppm) throughout the project.
• Average concentrations of petroleum hydrocarbon in the inner
gallery decreased from 5373 to 1546 ppm over the 24-week project.
• Over 5400 kg of gasoline were degraded throughout the treatment
area during the test. This suggests that each kilogram of degraded
gasoline required 1.5 kilograms of oxygen.
Because of budget limitations, the API was not able to continue the test
to determine what levels of remediation could be achieved, what
quantity of dissolved oxygen would be required to complete the bio-
degradation process, or whether the rate of biodegradation would slow
at lower levels of contamination. However, this study did demonstrate
the tremendous potential of hydrogen peroxide as a means to stimulate
aerobic biodegradation in contaminated aquifers.

21.4 PRINCIPLES OF MICROBIAL PROCESSES

Although microbial processes have long been recognized as important

in subsurface environments (Back, 1989), microbiologists have only
recently recognized the abundance and versatility of subsurface
microbes in microbial degradation. This section provides a brief sum-
mary of the general nature of subsurface bacteria which are important
in bioremediation processes. Many other types of microbial species
 Principles of microbial processes 561

exist in the subsurface, but discussion of these species is beyond the

scope of this chapter.
Bacteria can adapt to survive under a wide range of environmental
conditions, but they are most active in moist soils with a pH in the
range 6.5-8. Although microbial activity is accelerated at warmer
temperatures, metabolic rates in soil are usually limited by the avail-
ability of carbon or oxygen, and temperature is rarely a significant
factor in controlling subsurface biodegradation. In spite of these
'preferred conditions' for microbial activity, viable bacteria, both
aerobic and anaerobic, have been detected more than 100 m below land
surface (Fliermans and Balkwill, 1989).

21.4.1 Characteristics of aerobic bacteria

Aerobic bacteria are single cell organisms, about the size of clay par-
ticles (0.5-3 J.1m), which exist in a variety of shapes, including rods,
spheres, and chains. Aerobic bacteria grow by converting organic
compounds into cell mass and carbon dioxide, and by converting
oxygen to water vapor. Since the cells are made up of a variety of
proteins, lipids, nucleic acids, and other complex constituents. Bacteria
also need nitrogen, phosphorus, sulfur, and trace amounts of metals to
grow. The non-water portion of a bacteria typically contains about 53%
carbon, 12% nitrogen, 3% phosphorus, 1% sulfur, 20% oxygen, 7%
hydrogen, and 7% ash, for a total mass of 10- 12 per cell.
Due to the scarcity of available carbon and nutrient sources in most
subsurface environments, these bacteria appear to live in a somewhat
dormant state most of the time. However, when a suitable carbon
source becomes available (as after a gasoline spill), the bacteria have
the ability to multiply very rapidly, often doubling every 45 minutes.
Pristine soils in shallow unconfined aquifers commonly contain 100-
1000 aerobic bacteria per gram of soil, but this level may increase to 105
or more within one week of a contamination event. When this occurs,
microbial activity is generally limited by the availability of dissolved
oxygen. This increased oxygen demand can be observed by monitoring
the dissolved oxygen levels in a contaminated aquifer. It is common for
pristine, shallow groundwater to contain 4-6ppm of dissolved oxygen
(DO), while water contaminated by a biodegradable material will
generally contain less than 0.5 ppm. (Commercial test kits are now
available which allow rapid measurement of the dissolved oxygen
levels in water.) Depressed levels near a suspected contaminant plume
is evidence that the natural bacteria present at the site are degrading
the material but that the rate is oxygen limited.
The bacteria present within an aquifer are a consortium of many
different genera and species, each with slightly different biodegrada-
tion capabilities. Jamison and co-workers isolated 32 distinct cultures of
562 In situ bioremediation of groundwater

bacteria from one site contaminated with gasoline, all of which could
grow on gasoline as a sole carbon source (Jamison et al., 1976). These
included ten different species of the Nocardia genus, four Acinetobacter
and eight Pseudomonas species, as well as several other bacteria. These
workers then tested the ability of the various culture isolates to de-
grade various constituents of gasoline. They found that, although the
complete consortium of bacteria could degrade all of the components of
gasoline, there were several constituents which were not degraded by
any of the pure isolates and no individual isolate degraded all con-
stituents. In general, however, they found that Nocardia were primarily
responsible for paraffin degradation and that the Pseudomonas cultures
were responsible for aromatic degradation.

21.4.2 The nature of biodegradation

Essentially all commercial bioremediation processes involve accel-
erating aerobic biodegradation of a contaminant, where the contami-
nant acts as a carbon source and a source of energy for the bacteria.
Bacteria convert the carbon source (benzene for example) to either cell
mass or to carbon dioxide and water. In the process, the bacteria gain
energy, which allows the cell to function. The overall biodegradation
process can be expressed through simple chemical formulae, such as
those shown below. Equation (21.1) shows the conversion of benzene
and oxygen to cell mass and water, while Eq. (21.1) shows the final
products of carbon dioxide and water.
Oxidation to cell mass
C6H6 + 2.50 2 + NH3 ~ C 5 H 7 0 2 N + CO 2 + HzO (21.1)
Complete mineralization
C5 H 7 0 Z + 50z ~ 5C02 + 2HzO + NH3 (21.2)
These equations illustrate that aerobic oxidation of hydrocarbons
requires large quantities of oxygen, in this instance 3.1 kilogram of
oxygen per kilogram of contaminant. Of course, the process is much
more complex, in that bacteria actually perform a series of smaller
reactions, each of which involves a particular microbial enzyme.
The enzymatic processes are well understood for petroleum hydro-
carbons and most common non-chlorinated chemicals. Hydrocarbons
such as octane, for example, are oxidized first to octanol, then to
octylaldehyde and octanoic acid (Fig. 21.3) (Sufiita, 1985). Octanoic acid
is then converted to hexanoic acid, and a two-carbon segment becomes
part of the tricarboxylic cycle of the bacteria. The process is then
repeated, until the compound is completely degraded (Fig. 21.4).
Branched and cyclic compounds tend to degrade more slowly than
unbranched compounds because they do not 'fit' as easily into several
 r-
Principles of microbial processes 563
CH 3 - (CH 2 )3 - CH 2 - CH 2 - CH 2 - CH 3

\;0,

CH 3 - (CH 2 )3 - CH2- CH2 - CH2 - CH 20H

Fig. 21.3 Terminal oxidation of n-octane.

of the enzymes responsible for initial oxidation. The octane example

also illustrates steps in the pathway where NAD (nicotinamide adenine
dinucleotide) is converted to NADH (nicotinamide adenine dinucleo-
tide dihydride) or where FAD is converted to FADH. These com-
pounds are one means for the cell to store energy, which can then be
used to perform other cell functions, at some later time, and return to
the NAD or FAD form.
The biodegradation pathway for benzene in illustrated in Fig. 21.5.
Similar enzymatic pathways are used for many substituted aromatic
compounds, though the presence of chlorine, nitro, and some other
substituents can deactivate the ring or prevent certain enzymatic
reactions, reducing the rates of degradation.
Some of the enzymes which are responsible for degradation reac-
tions are quite specific and will only react with chemicals of specific
geometry. The enzyme responsible for the initial oxidation of octane to
octanol, for example, prefers to react with terminal carbons (those at
the end of a chain). As a result, cyclic hydrocarbons, which contain no
terminal carbons, are degraded much more slowly than straight chain
hydrocarbons of similar length.
Since the relative specificity of enzymes is frequently not known, it
is very difficult to predict the relative rates of degradation of mixtures
of compounds, even when the rates for the pure compounds are
known. In one instance, an easily degraded compound will stimulate
the production of an enzyme system needed to degrade a 'difficult'
564 In situ bioremediation of groundwater

o
CH3 - (CH 2 )3- CH 2 - CH 2- CH 2- C" - OH

r~CoA
o
"
r __
CH 3 - (CH 2 ) 3 - CH 2 - CH 2 - CH 2 - C - CoA

FAD

,~FADH
o
CH 3 - (CH 2 ) 3 - CH 2 - CH = CH -
"
C - CoA

o\I

C~ - (CH 2 ) 3 - CH 2 - CHOH - CH 2 - C - CoA

r-:o::~H
II
a II
CH 3 - (CH 2 )3 - CH 2 - C - CH 2 - C - CoA

r~CoA o
\I \I

CH 3 - (CH 2 )3 - CH 2 - C - CoA + CH 3 - C - CoA

Fig. 21.4 p-oxidation of n-octane.

compound, thus accelerating the degradation of this more recalcitrant

compound. In another instance, the recalcitrant compound will retard
all microbial activity, while in a third instance, the bacteria will degrade
only the easier material until it is depleted and then, gradually, begin
to degrade the second carbon source. As a result of the unpredictability
of this process, some type of treatability study is usually required
during the design phase of any biodegradation project, which involves
an unfamiliar combination of constituents.
Two general characteristics of the aerobic biodegradation process are
especially important to note, when treating potentially hazardous
compounds .
• The first oxidation step is usually the rate limiting step. After a
compound becomes incorporated into the cell, it is usually degraded
to its final end products quickly, with little accumulation of inter-
mediate products.
 Principles of microbial processes 565

1°""'"
~COOH

1
1
1 1
Mineralization

Fig. 21.5 Biodegradation of benzene.

• Contaminants are quickly transformed into the same intermediates

which result from the degradation of 'natural' microbial substrates.
In those cases where the bacteria are able to make only limited changes
to the substrate, the process is referred to as biotransformation. In
those instances when bacteria convert the substrate to carbon dioxide
and water (and other inorganic materials), the process is called miner-
alization. Although both can be considered biodegradation, mineral-
ization of the contaminant is the desired process.
Some aerobic bacteria have the ability to use nitrate as an electron
acceptor when molecular oxygen is not available. The process is called
denitrification, and bacteria with this ability are called denitrifiers. In
areas where neither oxygen nor nitrate is available, other types of
bacteria may dominate. These bacteria can use sulfate (sulfate reducers),
acetate (fermentative), or carbon dioxide (methanogenic) as electron
acceptors. However, these bacteria are only active when the soil matrix
is in a very reduced condition.
Recent studies have suggested that many of the contaminants of
566 In situ bioremediation of groundwater

concern are degradable anaerobically through either denitrification

or methanogenesis. Since these degradation pathways are not well
understood and are not yet practiced commercially, their discussion
will be deferred to a later section. However, two cautions are worth
emphasizing for anaerobic processes. Under some conditions, an-
aerobic processes will create sulfide or methane as by-products of
degradation. Secondly, anaerobic processes tend to be much slower
than aerobic processes, such that intermediates can accumulate. These
degradation intermediates for some anaerobic processes are more toxic
than the starting compound. The degradation of trichloroethylene
to dichloroethylene and vinyl chloride is the most commonly cited
example of this phenomena. These properties of anaerobic processes
can lead to major operational problems, if they have not been con-
sidered in the process design.

21.5 EVALUATION, DESIGN, AND IMPLEMENTATION

Although the design, predictability, and versatility of in situ biore-

mediation processes have improved significantly in the past 15 years,
the majority of commercial projects are still petroleum hydrocarbons
(gasoline, kerosene, fuel oil, diesel fuel, and jet fuel) in permeable
aquifers (Yaniga et ai., 1985; Yaniga and Smith, 1986; Bell and Byrd,
1988; Brown et ai., 1988). This section will provide an overview of the
contaminant, hydrogeological, and nutrient interaction issues which
are typically considered in evaluating the suitability of a site for in situ
bioremediation and in designing a remediation process.

21.5.1 Contaminant characteristics

Although the microbial environment affects the rate of biodegradation
of a contaminant in a particular location, it is clear that certain chemical
constituents are inherently more accessible to microbial degradation
than others. The aqueous solubility, molecular size, and types of
chemical bonds within a particular chemical structure affect the ability
of a bacteria to metabolize the molecule and obtain energy from the
process. Several literature reviews have been published which sum-
marize the general 'biodegradability' of the types of compounds which
are most frequently of concern (Alexander, 1981; Babea and Vaishnav,
1987). In addition, at least one reference provides a guide to the con-
centrations at which various contaminants may also be toxic to organ-
isms (Verschueren, 1983). Collectively this literature can be used to
provide the reader with an indication of the biodegradability of a
particular compound. A brief summary of general trends of aerobic
degradability is provided below (Brubaker, 1989). For the purposes of
 Evaluation, design, and implementation 567
this analysis, an easily degradable material is one for which common
bacteria can mineralize the compound (or mixture) of interest, with
relatively short acclimation times, and can use the material as a sole
carbon source.
• Simple hydrocarbons and light petroleum distillates, such as
gasoline, kerosene, diesel, jet fuel, and light mineral oils are gen-
erally degradable. Their rate of degradation decreases with increas-
ing molecular weight and decreasing solubility. Increased branching
and cyclic structures also slow the degradation process.
• Aromatic hydrocarbons with up to two rings (including benzene,
toluene, xylene, ethylbenzene, and naphthalene) are readily de-
gradable. The rate of degradation of larger polyaromatic hydro-
carbons (greater than three rings) decreases as size increases and
solubility decreases.
• Alcohols, amines, esters, carboxylic acids, and nitriles are usually
degradable, but these compounds are often toxic to unacclimated
bacteria at high concentrations. Nitrobenzenes and ethers are
usually more slowly degraded.
• Chlorinated hydrocarbons (both straight chain and aromatic) become
increasingly difficult to degrade as the degree of chlorine sub-
stitution increases. Polychlorinated biphenyls (PCBs) and other
polychlorinated hydrocarbons (chloroform, carbon tetrachloride,
tetrachloroethylene, trichloroethylene, and dichloroethylenes) are
not readily degraded aerobically.
• Pesticides are another very complex set of environmentally per-
sistent compounds. In general, those which are found at hazardous
waste sites (DDT, Lindane, Aldrin, Chlordane etc.) are not readily
biodegraded.
The degradation of viscous organics materials, like number 6 fuel oil,
creosote, and refinery wastes is often controlled by their physical
condition in the soil/water matrix. When they exist as small droplets of
oil, occluded within the pores of a soil, there is very little exposed
surface area for degradation, and the biodegradation process will be
inhibited.
Although it is difficult to quantify the importance of contaminant
distribution to project feasibility, project success requires movement of
nutrient-enriched water through those areas of the site which contain
the highest concentrations of contamination. Sites which contain a
few point sources of contamination, whether a lagoon or a leaking
tank, can generally be treated fairly reliably with an in situ treatment
method. However, in situ bioremediation strategies for sites which
contain multiple and undefined sources of contamination become
much more difficult to design and operate in a predictable fashion.
The probability of successful remediation is definitely increased by a
568 In situ bioremediation of groundwater

thorough understanding of the sources and transport mechanisms for

the contaminants.

21.5.2 Hydrogeology
In most in situ bioremediation projects, mass transport of oxygen and
nutrients is the largest single factor affecting the duration and efficiency
of the project. Although the nature of the oxygen and nutrient sources
are significant issues in this design, the design and operation of the
groundwater recovery and injection system is of primary importance.
Since the demand for oxygen far exceeds that of other nutrients, and
microbial activity is essentially instantaneous relative to groundwater
movement, the delivery of dissolved oxygen is the rate-limiting step.
The objectives of the groundwater system are to establish hydraulic
containment of dissolved phase contamination and to transport nu-
trients to the area of nutrient demand. (For very insoluble organics,
such as four and five ring polyaromatic hydrocarbons, dissolution and
microbial degradation may actually also be controlling factors in the
process rate.)
Since both the soils and the contaminant distribution at a site are
typically non-homogeneous, the initial determination of feasibility
and cost will usually depend on an estimate of the total volume of
degradable material, the oxygen demand of this area, the concentration
of DO which can be delivered in groundwater, and the rate at which
groundwater can be injected into the area of concern. These estimates
are illustrated in the following subsection.
As in all groundwater treatment processes, hydraulic conductivity,
saturated thickness and the natural groundwater gradient are im-
portant parameters. These parameters must be defined in order to
determine the rate at which groundwater can be withdrawn from a
well, the average groundwater velocity, and the capture zone defined
by a specific pumping rate (Keely and Tsang, 1983). In a similar fash-
ion, the thickness and hydraulic conductivity of the unsaturated zone
will affect the amount of water which can be injected into a soil, as well
as the sphere of influence of this injection point. The influence of
injection points must overlap to provide nutrients to the entire con-
taminated zone, in the same manner that the capture zones must
overlap for complete containment.
Successful design of this groundwater recovery and reinjection
system requires a detailed understanding of the hydrogeology of the
site. Although preliminary estimates of hydraulic conductivity can be
made based on soil descriptions, grain size analysis, or slug tests, a
pump test (preferably in combination with reinjection) should always
be performed whilst the final design is being established.
A three-dimensional understanding of contaminant distribution is
 Evaluation, design, and implementation 569

also required, so that the oxygen delivery is concentrated in the areas

of highest oxygen demand. General indicator parameters, such as
total organic carbon (TOC), chemical oxygen demand (COD), or total
petroleum hydrocarbon constituents (TPHC), are much preferred over
compound-specific analyses for these calculations. It is also important
to be aware of other sources of oxygen demand (buried plant or animal
wastes, septic or sewer systems, non-priority pollutant constituents)
which will also be degraded during a bioremediation process, con-
suming oxygen in the process.
Since the concentration of contaminants often decreases exponen-
tially with distance from a point source, it is often beneficial to con-
centrate injection point(s) near areas of high oxygen demand. This
strategy will often allow the injected groundwater to follow the same
migration paths as the contaminants. In larger sites, it may be nec-
essary to divide the site into treatment cells, to allow for more efficient
nutrient transport. In some instances this will result in cleaning up
80% of the site very rapidly and then concentrating on the primary
source area for a longer treatment time. This possibility should be
considered when designing the water treatment system and other
process components, as it may create opportunities to stagger treat-
ment in areas of low contamination and, thus, lower the overall project
cost.
When designing a simple groundwater capture or pump-and-treat
process, hydrogeologists are typically not concerned with the time
required for the contaminant to migrate through the soil to the re-
covery point. However, for in situ bioremediation projects, the process
efficiency is highly dependent on how long it takes for the nutrients to
reach the areas where they are required by the bacteria. Although the
sensitivity to this issue depends on the ability of the nutrients to move
through the soil, peroxide stability, and oxygen demand, in general, it
is often desirable to create a groundwater injectionlrecovery system
which will deliver nutrients to all areas of the contaminated area within
60-90 days of the time of injection.
Of course, how far groundwater can move in 60 days is dependent
on the permeability and porosity of the aquifer and the amount of
hydraulic gradient created by the injection and recovery system. In a
medium sand aquifer with a hydraulic conductivity of 5 x 10- 4 mis, an
effective porosity of 0.35 and an induced gradient of 0.04, the average
fluid velocity is 5 m per day between the point of injection and the
point of recovery or treatment. As a result, groundwater remedia-
tion processes are very difficult to implement for soils with hydraulic
conductivities less than about 10- 6 m/s.
In addition, unconfined aquifers which are very thin allow very little
drawdown, thus limiting the aquifer yield and the area of influence of
a well. This can make a design difficult, unless a recovery trench can be
570 In situ bioremediation of groundwater
used. Likewise, aquifers which are very close to the surface limit the
amount of groundwater injection that can occur without flooding. This
too, limits the hydraulic head that can be induced and the influence of
a particular injection well.
Computer modeling is a very powerful tool in the design of the
injection and recovery strategy for an in situ bioremediation process.
Computer simulations help to evaluate whether the aquifer charac-
teristics measured are internally consistent, and are useful in predicting
the pumping rates, flow velocities, and capture zones which will result
from various injection/recovery strategies. It is important to investigate
the sensitivity of the system to various assumptions and parameters
before finalizing the design, and then to be conservative in designing
the injection wells. Since injection wells are notorious for losing their
efficiency with time, the system should deliver nutrients to all areas of
the site, even when the efficiency of a well is only 70%. Conservative
design will also help to compensate for the non-homogeneity within a
site. Maintenance of wells should also be planned on a routine basis.
In evaluating the computer simulations and examining design
options, the drawdown at recovery wells should be minimized to the
extent possible to avoid drawing the nutrients below the zone of
highest contamination. The desire to minimize drawdown must be
balanced against the conflicting objective of optimizing the hydraulic
gradient and groundwater flow rates. The use of recovery trenches or a
series of small recovery wells, in place of a few larger wells, will aid in
this objective. Likewise, injection trenches or multiple injection wells
are often used as a point of groundwater injection. In general, it should
be assumed that the amount of groundwater that can be injected into
an aquifer will be about 75% of the amount which can be withdrawn
while maintaining good hydrogeologic control. The remaining 25% will
need to be discharged off-site.

21.5.3 Oxygen demand and oxygen supply

The concentration of oxygen in injected water and the total volume of
water that is injected become important factors in predicting the time
required for the biodegradation of an easily degraded constituent. For
example, if an aquifer were contaminated with diesel fuel over an area
of 10 m x 20 m to a depth of 2 m at an average concentration of
200 ppm, (milligram of contaminant per kilogram of soil), this aquifer
would contain 141 kg of contamination (assuming a bulk density of
1760 kg/m3). Since about three kilograms of oxygen are required to
mineralize one kilogram of hydrocarbon, this fuel creates an artificial
oxygen demand of about 423 kg.
If this aquifer were a medium sand (hydraulic conductivity of
10- 4 mis, and effective porosity of 0.35) and the natural gradient across
 Evaluation, design, and implementation 571

the contaminated region were 1 %, then the average velocity of water

into the contaminated zone (10 m wide by 2 m deep) would be 0.0247 ml
day. The daily volume of water moving into the area is the flux times
the area of flow (10m x 2m) or, 1.73m3 of water per day. Since
shallow groundwater typically contains about 5 ppm of dissolved
oxygen, and water weighs about 1000kg/m3, this natural flow of
groundwater will deliver 0.00S7kg of DO per day. At this rate, nature
will deliver the oxygen required to mineralize the hydrocarbons present
in about 130 years.
If a groundwater recovery trench were installed, clean water re-
injected upgradient to increase the hydraulic gradient to 4%, and the
reinjected water saturated with dissolved oxygen (S ppm), the system
would now deliver 0.055kg of dissolved oxygen per day, reducing
the treatment time to 21 years. (Of course, some of the constituents
would be extracted from the site during this period.) By adding
hydrogen peroxide as an oxygen source at 300 ppm (the equivalent of
150ppm of DO), it would be possible to deliver 1.04 kg of DO per day
and reduce the project time at this contaminated region to less than
two years. (The actual calculation suggests seven months, but, in
practice, one would not operate at the full 300 ppm for the entire
treatment period.) This type of simple calculation provides a reason-
able estimate of the time required for in situ bioremediation of a readily
degraded contaminant.

21.5.4 Soil and groundwater chemistry

Microbial degradation processes are controlled by the presence of
appropriate bacteria, the suitability of the environment for microbial
degradation and the general chemical structure of the contaminants of
interest. Unfortunately, very little information about the first two areas
is typically obtained during initial site investigations. However, some
information can often be obtained by examining the general ground-
water parameters.
An examination of ammonia, chloride, phosphate, dissolved oxygen,
conductivity, and pH data can provide considerable insight into the
microbial environment. Groundwater pH is perhaps the most basic
characteristic, since soil bacteria prefer pH in the range of 6.5-S.
Elevated concentrations of ammonia, chloride, or sulfide can be an
indication of septic or sanitary contamination or a highly anaerobic
zone. This could indicate a large oxygen demand from other readily
degradable materials, which would significantly impact cost and
timing. There are several other chemistry issues which do not affect
microbial health but may affect feasibility of the process by affecting
nutrient transport.
Although simple in concept, the process of transporting an electron
572 In situ bioremediation of groundwater
acceptor (typically hydrogen peroxide) through soils can present
challenging problems at some sites. In 'ideal sites', hydrogen peroxide
can provide a very effective means of delivering 150-250 ppm of dis-
solved oxygen in injected water. Hydrogen peroxide is infinitely
soluble in water and decomposes in the presence of soil, such that
300 ppm of hydrogen peroxide produces about 150 ppm of dissolved
oxygen. The main design concern in using peroxide is controlling the
rate of this catalytic decomposition.
In those soils which contain high concentrations of natural organics,
iron, nickel, and copper, the catalytic decomposition of peroxide can be
so rapid that the groundwater quickly becomes supersaturated with
oxygen. This can create gas blockage of the formation, loss of oxygen,
and poor transport of DO to remote areas of the site. Although methods
exist for conditioning soils to minimize these effects, this phenomenon
can limit the concentration at which the peroxide can be used from
500 ppm, a common concentration in clean sands, to as little as 100 ppm.
Nitrogen and phosphorus are usually required in small amounts to
support increased bacterial growth, while trace minerals are readily
available in soil. These requirements are usually satisfied through
the periodic addition of nutrient formulations containing a blend of
ammonia and phosphate maintaining 2-10 ppm of each in the ground-
water. However, ammonia and phosphate transport can be difficult at
some sites, as these materials can be retarded and consumed while
passing through soils. Ammonia can be lost through ion exchange with
other cations in clays, and soils and phosphate can interact with
cations, especially calcium and magnesium. Although these processes
do not make the nutrients unavailable, they do limit the distance that
nutrients can be moved effectively through the formation and/or
impact the formulation of nutrients that are most effective. In addition,
the general introduction of these nutrients, with associated ions, might
lead to dispersion or swelling of clay particles.
For the purposes of an initial feasibility assessment, the interactions
of phosphate with calcium and magnesium (both of which form in-
soluble precipitates with orthophosphate) are probably the only trans-
port issues which can be assessed. Both cations inhibit phosphate
transport at high concentrations. Heavy metals, such as cadmium,
chromium, arsenic, and mercury, are toxic to bacteria in certain forms
and concentrations. The presence of these materials, especially if they
extend beyond a localized area, should be identified as an area for
further analysis when evaluating any microbial-based process.

21.5.5 Laboratory treatability and design studies

The extent and complexity of laboratory treatability and design studies
is very site specific. When working with common mixtures of con-
 Evaluation, design, and implementation 573
taminants (for example, gasoline) where the biodegradability is well
known, it is probably not necessary to confirm biodegradation, but it
may be desirable to evaluate the benefit of adding various levels of
nutrients or of adjusting the pH. Since the bacteria responsible for
biodegradation have adapted to the environmental conditions in the
site soil, and the soil is also a source of nutrients, biodegradation
studies should be performed in the presence of soil from the site. This
is commonly achieved by creating a series of replicate samples, con-
taining a slurry of soil and groundwater, and then varying the treat-
ment conditions. Conditions tested usually include a microbial inhibited
control (poisoned with mercuric chloride or sodium azide), a nutrient
control (no added nutrients), plus several sets with various nutrients or
additives. In some cases an anoxic control (no oxygen) is also included
in this series. For the experiment to be meaningful it is critical to
estimate the maximum possible oxygen demand of the system. The
oxygen addition schedule must then be designed to satisfy this need.
Samples are sacrificed, after some appropriate period of time, and
the entire contents of the vessel are extracted and analyzed for the
contaminants of concern. The size and type of sample bottles and
the time between samples is heavily dependent on the nature of the
systems being tested. For gasoline or diesel fuel, a series of 'volatile
organic analysis' bottles might be used to contain volatiles. Analyses
would occur at regularly spaced intervals for 2-4 weeks. In this case,
peroxide or oxygen would need to be injected regularly into the bottle
to maintain the dissolved oxygen level. If, however, the contamination
consisted of non-volatile compounds, like pesticides or polynuclear
aromatic compounds, open mouth bottles might be used and the
required incubation may be up to three months.
In general, biodegradation studies are helpful in predicting the
relative rate at which various constituents will degrade and the maxi-
mum level of remediation which can be predicted. The actual rates in
the field are likely to be limited by oxygen transport, desorption, and
other mechanical factors, which are difficult to simulate in the lab-
oratory. Many researchers in the area have attempted to use column
studies to mimic field conditions. However, because of the variability
inherent in soil studies and the inability to sample the soil matrix of a
column without sacrificing the entire column, these studies require
multiple sets of replicate columns in order to effect the desired repro-
ducibility. As a result, most commercial design/treatability studies are
performed in either static or slurried jars, rather than in columns.
Peroxide stability has become a topic of increasing concern in the
past few years, since the rate at which peroxide decomposes upon
contact with soil can be a major design factor for an in situ bioreme-
diation project. In principle, a peroxide stability test consists of making
up a slurry of hydrogen peroxide (typically 500 ppm) and soil, and then
574 In situ bioremediation of groundwater
analyzing aliquots of the solution after I, 2, 4, 8, and 24 hours to
determine the concentration of remaining peroxide (Brubaker and
Crockett, 1986). One recent presentation (Lawes, 1990) suggests that
the results of this experiment are very sensitive to the details of how
the experiment is performed (soil-to-water ratio, size of flask, rate of
stirring, etc.). Since the test is generally useful only as a means of
comparing the stability of various soils within a site, or between sites
with which the engineer or scientist has operating experience, there is
probably no 'best' method, as long as consistency is maintained.
Nutrient interaction with soil and/or groundwater is often evaluated
as a means of determining the ease of transport through the soil
matrix. These tests are often fairly simple. One example is exposing
a solution of a nutrient-amended groundwater to site soil and then
analyzing the water for the nutrients, to measure adsorption (Brubaker
and Crockett, 1986). Some firms prefer to pass a solution of the nu-
trient through a column of soil and then analyze the effluent for
breakthrough. Either method will provide an indication of the amount
of ion exchange and/or adsorption of the nutrients with the soil. The
column study is a little more elaborate and provides more detailed
information.

21.5.6 Full-scale design and implementation

There are many site specific features which affect the ultimate design
and implementation of a full-scale in situ bioremediation project. Many
of these relate directly to the physical constraints of the site which limit
the location of wells and/or trenches, groundwater treatment systems,
and associated plumbing. In addition, the remediation objectives for
sites vary, both in terms of the level of remediation required, but also
relative to the rate of remediation. (It seldom makes economic sense,
for example, to use in situ bioremediation to remediate an area until the
source zones have been removed.)
Besides the design features outlined below, issues of water balance
and disposal are often important. Since roughly 75% of the ground-
water recovered from a site will typically be reinjected, it is generally
advisable to remove the dissolved contamination prior to reinjection,
even if the water which is discharged offsite does not require treat-
ment. Reinjection of dissolved constituents can increase the possibility
of microbial growth at the injection wells and it increases the oxygen
demand in the aquifer. Any of the standard groundwater treatments
can be used, but it is important to recognize that air strippers, carbon
beds, and recovery wells may be more prone to biological fouling than
normal, because of the nutrients added. (Fouling in injection wells is
actuallly fairly uncommon when peroxide is used as an oxygen source.)
Since hydrogeologic control is a critical aspect of any in situ biore-
mediation project, it is common to operate the injection and recovery
 Ongoing research and development 575

system for several days to weeks, prior to nutrient injection, to stabilize

the water-flow rates. The addition of nutrients provides an opportunity
to use chloride (present in the ammonium chloride) as a conservative
tracer to confirm nutrient movement, flow rates, and capture. Hydrogen
peroxide addition is typically initiated 3-10 days after the nutrients,
starting at 100 ppm and then increasing to 300-600 ppm. Hydrogen
peroxide is typically delivered onsite as either a 35% or 50% solution
and is metered into the groundwater for dilution. As a concentrated
solution, peroxide is corrosive to skin and can decompose violently, if
contaminated with metals. Therefore, this material should only be
handled according to manufacturer's directions and in equipment
designed and designated for its use.
The amount of monitoring required at a site is dependent on the
size of the site and how rapidly concentrations are likely to change.
Groundwater monitoring is usually performed weekly to monthly for
pH, DO, nutrients, bacteria, and contaminants. Most of these par-
ameters can be monitored easily on-site using handy field test kits or
meters. However, since the concentration of dissolved contamination
usually increases during the initial phases of nutrient addition, the best
means of verifying process performance is through soil coring. Since
contaminant distribution is irregular, the decrease in concentration is
usually not detectable until the process is about one-third complete,
and even then, the most interpretable data will be collected if the cores
used for monitoring are grouped to allow comparison of samples
collected in the 'same location', before and after treatment.
As the process proceeds, the observed concentration of constituents
in the groundwater often remain high until the remediation is nearly
complete, and the mass of sorbed contamination is almost depleted.
During this time the nitrogen and phosphate will also have moved
further and faster than the dissolved oxygen. As the DO monitored
downgradient of the contaminated area increases, the peroxide level
can be decreased until the point where the oxygen demand of the
aquifer has been satisfied. At the point where the natural oxygen
supply to the aquifer is sufficient to sustain the continued biological
degradation of the contaminants present, the addition of more nu-
trients or oxygen has no apparent benefit. Although this point may not
necessarily reflect 'site closure', in most cases continued containment
and/or monitoring will provide as much benefit as continued active
nutrient addition.

21.6 ONGOING RESEARCH AND DEVELOPMENT

During the past five years, researchers and practitioners from a wide
range of disciplines have contributed dramatically to our understand-
ing of in situ bioremediation, asking new questions and providing new
576 In situ bioremediation of groundwater
insight regarding the versatility, reliability, and optimization of this
process (McCarty, 1987; Wilson and Ward, 1987; Thomas and Ward,
1989). This section contains a brief summary of a few of the more
promising and innovative areas of development.

21.6.1 Biodegradation using denitrification

As recently as seven years ago, it was generally accepted that aromatic
rings which lacked oxygen substitution were not biodegradable by
denitrifying bacteria (Bouwer and McCarty, 1983). Recent research is
challenging those assumptions and examining the possibility of using
nitrate as an electron acceptor to enhance the biodegradation of organic
contaminants. Although nitrate is often itself a groundwater contami-
nant, it is much cheaper than hydrogen peroxide and is more easily
transported through soils. The ability to use this material as an oxygen
source could significantly reduce the cost and complexity of the pro-
cess. Denitrification also has advantages over other anaerobic processes
in that this process uses the same enzymatic pathways as the aerobic
process. This is important because these pathways are well-known and
there are no undesirable by-products.
In the mid-1980s, Swiss workers (Kuhn et al., 1985; Zeyer et al., 1986;
Kuhn et al., 1988) reported the biodegradation of m-xylene by deni-
trifying bacteria which had been isolated from denitrifying river sedi-
ment and fed xylene continuously for several months. In a series of
carefully controlled column studies, including experiments with 14C
labelled material, these workers observed the degradation of m-xylene
and several other substituted benzenes. However, these bacteria were
unable to degrade benzene, naphthalene, methylcyclohexane or 1,3-
dimethylcyclohexane. Related research by a Canadian group (Major
et al., 1988) demonstrated the biotransformation of benzene, toluene
and xylene in laboratory studies using a sandy aquifer material, and
a recent study (Mihelic, 1988) describes the biodegradation of naph-
thalene, naphthol, and acenaphthalene under denitrifying conditions.
Field activities have been limited to one controlled field study, which
failed to confirm degradation rates in excess of those expected from
aerobic processes (Berry-Spark and Barker, 1987), and two attempts at
field-scale remediation. The first involved a sandy aquifer contami-
nated with mineral oil. Water containing 300 ppm of nitrate was in-
troduced into this contaminated zone for approximately 18 months,
resulting in dramatic decrease in dissolved levels of both aromatic
and aliphatic constituents (Geldner, 1987; Werner, 1985). The second
project involved the injection of 45 ppm nitrate to an aquifer containing
gasoline. Although data collected to date indicate a decrease in the
levels of petroleum constituents at this site, it is difficult to gain any
useful insight into the process from the limited information available
(Sheehan et al., 1988). The US EPA is in the process of performing
 Ongoing research and development 577
much more elaborate evaluation of this technology at the Traverse City
site (Hutchins, 1991). Laboratory data from this project confirm the
presence of biodegradation of alkylated benzenes by denitrifying bac-
teria but benzene was not degraded under the conditions tested.

21.6.2 Methanogenic biodegradation

Although it has long been recognized that many aromatic compounds
could be biodegraded under methanogenic conditions (Kuhn and
Suflita, 1989), it is only very recently that laboratory studies under
controlled conditions (Grbic-Galic and Vogel, 1987) have confirmed this
observation. In addition, anecdotal field observations (Hayman et al.,
1988) suggest that there may be techniques of enhancing the anaerobic
degradation of many organic constituents. This phenomenon is being
examined in more detail at the EPA Traverse City project where zones
of methanogenesis and of aerobic biodegradation were detected
(Wilson and Ward, 1987).

21.6.3 Modelling the in situ bioremediation process

There have been increasing efforts to model both natural biodegrada-
tion processes and active bioremediation in the subsurface. These
models are helpful in identifying the critical parameters which affect
microbial activity and may some day be used in combination with risk
assessments to define 'acceptable' levels of remediation at sites near
sensitive receptors. BIOPLUME II was developed for the EPA, based
on several years of study at the United Creosote site in Conroe, Texas,
and it is now being applied to an aviation gasoline spill in Traverse
City, Michigan (Rifai et ai., 1987; Rifai et ai., 1989). The current version
has been 'packaged' in a manner which makes it fairly 'user friendly',
though it is not yet in wide-spread use. Most of the other models being
developed are more research oriented and not intended for commercial
use (Widdowson et ai., 1988; Doyle and Piotrowski, 1989; Tim and
Mostaghimi, 1989).
There are several research groups (Cunningham et ai., 1988; Rittmann
et al., 1988; Alexander and Scow, 1989; Chiang et ai., 1989; Swindoll
et al., 1988) looking at the kinetics and detailed mechanisms of bio-
degradation on a more molecular level, examining the role of biofilms
under the types of environments found in aquifers. This research has
the potential to lead to very important insights into the factors which
control the biodegradation process.

21.6.4 Chlorinated hydrocarbons

Chlorinated solvents such as trichloroethylene, trichloroethane, di-
chloroethylene, and associated compounds are almost ubiquitous at
578 In situ bioremediation of groundwater
industrial sites in many urban areas. Their presence and persistence
results from widespread use as degreasers and cleaning solvents, their
moderate solubility and volatility, their density (since they are denser
than water they migrate down through aquifers), and their chemical
and microbial stability (Vogel et al., 1987). The carbon chlorine bond
is a very 'stable' bond, and it appears that microorganisms do not
generate much usable energy from the oxidation of this bond. As a
result, even those bacteria able to biologically oxidize these compounds
require another food source for energy.
During the past ten years, many researchers have been actively
studying the chemical and microbial transformations of this class of
compounds in subsurface environments (Fogel et al., 1986; Wilson
et al., 1987; Nelson et al., 1988). Methanogenic and aerobic processes
have been studied under a broad range of laboratory conditions, and
an extensive EPA-funded field study has been conducted to develop
and evaluate techniques for implementing in situ bioremediation of
these materials in a real aquifer (Semprini et al., 1988). The results of
this work can be summarized as follows.
• Halogenated aliphatic compounds undergo a limited number of
chemical and biological reactions in natural settings. These include
hydrolysis, dehydrohalogenation, and oxidation/reduction reactions.
• The rate of microbial oxidation decreases as the number of chlorine
atoms increases. The rate of reductive dehalogenation increases as
the number of chlorine atoms increases.
• Aerobic in situ processes are likely to require the addition of a readily
degradable carbon source, which will stimulate the microbial popu-
lation of interest (for example, methane or toluene), and sufficient
oxygen to degrade both the original contaminants and the added
carbon. The ratio of added carbon to contaminants is likely to be in
the range of 30: 1.
Based on the complexity of this process, both from a microbial and
from an engineering perspective, it is doubtful that the in situ bio-
remediation of highly chlorinated solvents will become 'standard'
technology in the next 3-5 years.

21.6.5 PAHs and other less soluble contaminants

Over 90% of all of the commercial experience with in situ bioremedia-
tion involves the biodegradation of contaminants which are soluble at
the part per million level and whose remediation levels are defined in
terms of part per billion of dissolved constituents. As projects progress
into groundwater corrective action, an increasing number of sites will
be examined to determine how to remediate soluble polyaromatics
hydrocarbons (PAHs) to low part per billion or even part per trillion
 Ongoing research and development 579
levels. In situ bioremediation of contaminants which have very low
solubility may evolve into a very different process than that which
exists today. Two of the unique aspects of this problem are sum-
marized below .
• The rate of biodegradation of these compounds appear limited by
its rate of dissolution even in well mixed systems. The rate of
degradation of compounds which partition strongly to soil and
desorb slowly may be dependent on these physical-chemical prop-
erties, in addition to nutrient transport.
• Current research suggests that the degradation of PAHs usually
slows or ceases at levels of 10-50 ppb. The factors which control this
phenomena are not well understood, nor are ways to continue
remediation beyond this level.
Although a variety of groups (Scow et al., 1986; Swindoll et al., 1988;
Borden et al., 1989; Doyle and Piotrowski, 1989; Smith et al., 1989;
Wang et al., 1990) are beginning to probe these and related issues,
much more experience is required with these compounds before
remediation of 'heavy' organics to sub-part per million levels will be
routine.

21.6.6 In situ bioremediation in the unsaturated zone

Even though this chapter has focused on the bioremediation within the
saturated portion of contaminated sites, it is important to recognize
that contamination above the water table, in the unsaturated zone, is
also of concern. Residual contamination in these areas can continue to
contaminate groundwater as precipitation percolates through this
region or as the water table fluctuates. Although unsaturated soils
are often easier to excavate and treat than saturated soils, the inacces-
sibility of soils at many sites has created a need for a better under-
standing of biodegradation in unsaturated soils.
The introduction of vapor extraction techniques has led to an in-
creased awareness of factors which control the fate and transport of
organic vapors in unsaturated soils. One of the important insights from
this process has been the increased awareness that aerobic biodegra-
dation within the unsaturated zone can lead to oxygen depletion and
carbon dioxide inhibition, even in shallow soils. Monitoring at the EPA
Traverse City site indicated oxygen concentrations of less than 3% and
carbon dioxide levels in excess of 10% at depths greater than 2 m
(Kampbell, 1990).
The critical factors for vadose zone biodegradation appear to be the
same as those involved in land treatment (see Chapter 20) and the in
situ bioremediation of aquifers, with some subtle differences. Oxygen
exchange is very important and is usually achieved by using the
580 In situ bioremediation of groundwater

techniques that are developing in the vapor extraction area, drawing

air through the unsaturated zone (Dineen et al., 1990; Hinchie and
Miller, 1992). Moisture is probably the next most important factor and
must be maintained within a level that keeps the soils moist without
seriously inhibiting air flow. Ammonia-nitrogen and phosphate are
probably the next factors to be considered. Percolation of nutrient-
enriched water, either through closely spaced injection points or
through surface irrigation have both been used, as has gaseous
ammonia. Although the amount of detailed study in this area still trails
that of saturated zone treatment and surficial land treatment, this
technique will certainly be used increasingly in the future.

REFERENCES

Alexander, M. (1981) Biodegradation of chemicals of environmental concern,

Science 211, 132-8.
Alexander, M. and Scow, K.W. (1989) Kinetics of biodegradation in soil,
Reactions and Movement of Organic Chemicals in Soils, Madison, WI, 243-69.
American Petroleum Institute (API) (1987) Field study of enhanced subsurface
biodegradation of hydrocarbons using hydrogen peroxide as an oxygen source,
Publication No. 4448, Washington.
Babea, L. and Vaishnav, D.O. (1987) Prediction of biodegradability of selected
organic compounds, J. Indust. Microbiol. 2, 107-15.
Back, W. (1989) Early concepts of the role of microorganisms in hydrogeology,
Ground Water 27, 618-22.
Bell, R.A. and Byrd, R.E. (1988) Case history: gasoline spill in fractured
bedrock addressed with bioreclamation, Proc., HazTech Int. Conj.,
Cleveland, HazTech Int., Bellevue, WA, 2C-122-52.
Berry-Spark, K. and Barker, J.F. (1987) Nitrate remediation of gasoline
contaminated groundwater: results of a controlled field experiment, in
Proc., NWWAIAPI Conj. on Petroleum Hydrocarbons and Organic Chemicals in
Groundwater: Prevention, Detection and Restoration, NWW A, Dublin, OH,
pp. 127-44.
Borden, KC., Lee, M.D., Thomas, J.M., Bedient, P.B. and Ward, H.C. (1989)
In situ measurement and numerical simulation of oxygen limited
biotransformation, Ground Water Monitoring Rev., Winter, 83-91.
Bouwer, E.J. and McCarty, P.L. (1983) Transformation of halogenated organic
compounds under denitrification conditions, Appl. Environ. Microbiol., 45,
1295-9.
Brown, R.A., Matson, C. and Benazon, N. (1988) The treatment of
groundwater contaminated with petroleum hydrocarbons using in situ
bioreciamation, in Proc., HazTech Int. Conj., Cleveland, Haztech
International, Bellevue, WA, pp. 2C-153-79.
Brown, R.A. and Norris, R.D. (1986) Field demonstration of enhanced
bioreciamation, in Proc., 6th Nat. Symp. on Aquifer Restoration and
Groundwater Monitoring, NWWA, Dublin, OH, p. 421.
Brown, R.A., Norris, R.D. and Raymond, R.L. (1984) Oxygen transport in
contaminated aquifers, in Proc., Petroleum Hydrocarbons and Organic
Chemicals in Ground Water, NWWA, Dublin, OH, pp. 421-41.
 References 581

Brubaker, G.R (1989) Screening criteria for in situ bioreclamation of

contaminated aquifers, in Proc., Haz. Wastes and Haz. Mat., HMCRI, New
Orleans: pp. 319-21.
Brubaker, G.R. and Crockett, L.L. (1986) In situ aquifer remediation using
enhanced bioreclamation, in Proc., Hazmat 86, Atlantic City, Tower Conf.
Management, Glen Ellyn, IL.
Chiang, c.Y., Salanitro, J.P., Chai, E.Y., Colthart, J.D. and Klein, c.L. (1989)
Aerobic biodegradation of benzene, toluene, and xylene in a sandy aquifer
- data analysis, and computer modeling, Ground Water, 27, 823-34.
Cunningham, A.B., Characklis, W.G. and Bouwer, E.F. (1988) Influence of
microbial transformation on the in situ Biodegradation of groundwater
contaminants, in Proc., Petroleum Hydrocarbons and Organic Chemicals in
Ground Water, NWW A, Dublin, OH, pp. 669-84.
Dineen, D., Slater, J.P., Hicks, P. and Holland, L.D. (1990) In situ biological
remediation of petroleum hydrocarbons in unsaturated soils, in Petroleum
Contaminated Soils, Volume 3 (eds P.T. Kostecki and E.J. Calabrese), Lewis
Publishers, Chelsea, MI, pp. 177-93.
Dupont Biosystems (1990) Marketing Literature.
Doyle, R. and Piotrowski, M. (1989) In situ bioremediation of wood-treating
compounds in groundwater, in Proc., 6th Nat. Conf. on Haz. Wastes and
Haz. Materials, HMCRI, Silver Spring, MD, pp. 308-14.
Flathman, P.E. and Caplan, J.A. (1986) Cleanup of contaminated soils and
groundwater using biological techniques, in Proc., 3rd Nat. Conf. on Haz.
Wastes and Haz. Materials, HMCRI, Silver Spring, MD.
Fliermans, C.B. and Balkwill, D.L. (1989) Microbial life in deep terrestrial
subsurfaces, Bioscience, 39(6), 370-7.
Fogel, M.M., Taddeo, A.R and Fogel, S. (1986) Biodegradation of chlorinated
ethenes by a methane-utilizing mixed culture, Appl. Environ. Microbiol., 51,
720-4.
Geldner, P. (1987) Stimulated in situ biodegradation of aromatic hydrocarbons,
in Proc., 2nd Int. Conf. on New Frontiers for Haz. Waste Management,
Pittsburgh, PA, pp. 247-57.
Grbic-Galic, D. and Vogel, T.M. (1987) Transformation of toluene and benzene
by mixed methanogenic cultures, Appl. Environ. Microbiol., 53, 254-60.
Hayman, J.W., Adams, RB. and McNally, J.H. (1988) Anaerobic
biodegradation of hydrocarbons in confined soils beneath busy places - a
unique method of methane control, in Proc., Petroleum Hydrocarbons and
Organic Chemicals in Ground Water, NWW A, Dublin, OH, pp. 383-96.
Hinchie, RE. and Miller, R.N. (1992) Bioventing for In Situ Remediation of
Petroleum Hydrocarbons, in Bioventing and Vapor Extraction: Uses and
Applications in Remediation Operations, Air and Waste Management
Association, pp. 39-48.
Hutchins, S.R and Wilson, J.T. (1991) Laboratory and field studies in a fuel-
contaminated aquifer under denitrifying conditions in, In Situ Bio-
reclamation, Butterworth-Heineman, London, pp. 157-72.
IT Corporation (1986) In Situ Bioreclamation: A Case History - Gasoline
Contamination in Northern Indiana.
Jamison, V.W., Raymond, R.L. and Hudson, J.O. (1976) Biodegradation of
high octane gasoline, in Proc., 3rd Int. Biodegradation Symp., Applied
Science Publishers, Englewood, NJ, pp. 187-96.
Jhaveria, V. and Mazzacca, A.J. (1985) Bio-reclamation of ground and ground
water by in situ biodegradation: a case history, in Proc., 6th Nat. Conf. on
Uncontrolled Haz. Sites, Washington, DC, pp. 242-52.
582 In situ bioremediation of groundwater
Kampbell, D. (1990) Simplified soil gas sensing techniques for plume mapping
and remediation monitoring, in Petroleum Contaminated Soils, Lewis
Publishers, Chelsea, MI, 3.
Keely, J.F. (1989) Performance evaluations of pump-and-treat remediations, EPA/540/
489/006.
Keely, J.F. and Tsang, CF. (1983) Velocity plots and capture zones of pumping
centers for groundwater investigations, Ground Water, Nov/Dec, 701-13.
Kuhn, E.P., Colberg, P.J., Schnoor, J.L., Wanner, 0., Zehnder, A.J.B. and
Schwarzenbach, R.P. (1985) Microbial transformation of substituted
benzenes during infiltration of river water to groundwater and laboratory
column studies, Environ. Sci. Techno!., 19, 961-8.
Kuhn, E.P., Zeyer, L Eicher, P. and Schwarzenbach, KP. (1988) Anaerobic
degradation of alkylated benzenes in denitrifying laboratory aquifer
columns, App!. Environ. Microbio!., 54, 490-6.
Kuhn, E.P. and Sufiita, J.M. (1989) Dehalogenation of pesticides by anaerobic
microorganisms in soil and groundwater - a review, in Reactions and
Movement of Organic Chemicals in Soils, Soil Sc. Soc. of Amer., Special
Publication No. 22, Madison, Wisconsin. pp. 111-80.
Lawes, B. (1990) Soil induced hydrogen peroxide decomposition, in Petroleum
Contaminated Soils, Lewis Publishers, Chelsea, MI, 3, ch 19.
Lee, M.D., Thomas, J.M., Borden, KC, Bedient, P.B., Ward CH. and Wilson,
J.T. (1988) Biorestoration of aquifers contaminated with organic com-
pounds, CRC Critical Reviews in Environ. Control, 18, 29-89.
MacKay, D.M. and Cherry, J.A. (1989) Groundwater contamination: pump-
and-treat remediation, Environ. Sci. Techno!., 23, 630-6.
Major, D.W., Mayfield, CI. and Barker, J.F. (1988) Biotransformation of
benzene by denitrification in aquifer sand, Ground Water, 26, 16-22.
McCarty, P.L. (1987) Bioengineering issues related to in situ remediation of
contaminated soils and groundwater, presented at Reducing Risks from
Environmental Chemicals through Biotechnology, Univ. of Washington, Seattle,
Washington.
Mihelic, J.K (1988) Microbial Degradation of Polycyclic Aromatic Hydrocarbons
Under Denitrification Conditions in Soil-Water Suspensions. PhD thesis,
Carnegie Mellon University, Pittsburgh.
Nelson, M.J.K., Montgomery, S.O. and Pritchard, P.H. (1988) Trichloro-
ethylene metabolism by microorganisms that degrade aromatic com-
pounds, App!. Environ. Microbio!., 54, 604-6.
Raymond, KL., Jamison, V.W. and Hudson, J.O. (1976) Biodegradation of
high octane gasoline, in AIChE Symp. Series, 73, 390-404.
Rifai, H.S., Bedient, P.B. and Wilson, J.T. (1989) BIOPLUME model for
contaminant transport affected by oxygen limited biodegradation, Environ. Res.
Brief, EPA/600/M019.
Rifai, H.S., Bedient, P.B., Borden, KC and Haasbeek, J.F. (1987) BIOPLUME
II - Computer Model of Two-Dimensional Transport under the Influence of
Oxygen Limited Biodegradation in Ground Water, User's Manual, Version 1.0,
Rice Univ., Houston, TX.
Rittmann, B.E., Valochcchi, A.J., Odencrantz, J.E. and Bae, W. (1988) In Situ
Bioreclamation of Contaminated Groundwater, HWRIC Report 031, Illinois
State Water Survey, Savoy, IL.
Scow, K.M., Simkins, S. and Alexander, M. (1986) Kinetics of mineralization of
organic compounds at low concentrations in soil, App!. Environ. Microbiol.,
51, 1028-35.
 References 583

Semprini, L., Roberts, P.V., Hopkins, G.D. and McKay, D.M. (1988) A Field
Evaluation of In-Situ Biodegradation for Aquifer Restoration, EPA Project
Summary 600/S2-87/096.
Sheehan, P.J., Schneiter, R.W., Mohr, T.K.G. and Gersberg, R.M. (1988)
Bioreclamation of gasoline contaminated groundwater without oxygen
addition, in Proceed., 2nd Nat. Outdoor Action Conf. on Aquifer Restoration,
Ground Water Monitoring and Geophysical Methods, Las Vegas, pp. 183-99.
Smith, J.R., Nakles, D.V., Sherman, D.F., Neuhauser, E.F. and Loehr, R.C
(1989) Environmental Fate Mechanisms Influencing Biological Degradation
of Coal-Tar Derived Polynuclear Aromatic Hydrocarbons in Soil Systems.
Presented at The 3rd Int. Conf. on New Frontiers for Haz. Waste Management,
Pittsburgh, PA.
Suflita, J .M. (1985) Microbial principles of in situ remediation of aquifers, in
Proc. of HazPro 85, Pudvan Publishers, Northbrook, IL.
Swindoll, CM., Aelion, CM., Dobbins, D.C, Jiang, 0., Long, S.L. and
Pfaender, F.K. (1988) Aerobic biodegradation on natural and xenobiotic
organic compounds by subsurface microbial communities, Environ. Toxico!.
Chern., 7, 291-9.
Texas Research Institute (1982) Feasibility Studies on the Use of Hydrogen
Peroxide to Enhance Microbial Degradation of Gasoline, prepared for API,
Washington, DC
Thomas, J.M. and Ward, CH. (1989) In situ biorestoration of organic
contaminants in the subsurface, Environ. Sci. Techno!., 23, 760-6.
Tim, U.S. and Mostaghimi, S. (1989) Modeling Transport of a Degradable
Chemical and its Metabolites in the Unsaturated Zone. Ground Water, 27,
672-81.
Verschueren, K. (ed.) (1983) Handbook of Environmental Data on Organic
Chemicals, 2nd edn, Van Nostrand Reinhold, NY.
Vogel, T.M., Criddle, CS. and McCarty, P.L. (1987) Transformations of
halogenated aliphatic compounds, Environ. Sci. Technol., 21, 722-36.
Wang, Y.S., Subba-Rao, R.V. and DuPont, R.R. (1990) Personal
communication.
Werner, P. (1985) A new way for the decontamination of polluted aquifers by
biodegradation, Water Supply, 3, 41-7.
Widdowson, M.A., Molz, F.J. and Benefield, L.D. (1988) Modeling multiple
organic contaminant transport and biotransformation under aerobic and
anaerobic (denitrifying) conditions in the subsurface, in Proc. of Petroleum
Hydrocarbons and Organic Chemicals in Ground Water, NWWA, Dublin, OH,
pp. 397-416.
Wilson, J.T., Leach, L.E., Henson, M. and Jones, J.N. (1986) In situ
biorestoration as a ground water remediation technique, Ground Water
Monitoring Rev., Fall, 56-64.
Wilson, J.T., Fogel, S. and Roberts, P.V. (1987) Biological treatment of
trichloroethylene in situ, in Detection Control and Renovation of Contaminated
Ground Water, ASCE, New York.
Wilson, J.T. and Ward, CH. (1987) Opportunities for Bioreclamation of
Aquifers Contaminated with Petroleum Hydrocarbons. Develop. Indust.
Microbio!., 27, 109-116.
Yaniga, P.M., Matson, C and Demko, D.J. (1985) Restoration of water quality
in a multiaquifer system via in situ biodegradation of the organic
contaminants, in Proc., 5th Nat. Symp. on Aquifer Restoration and
Groundwater Monitoring, NWWA, Dublin, OH, pp. 510-8.
584 In situ bioremediation of groundwater
Yaniga, P.M. and Smith, W. (1986) Aquifer restoration via accelerated in situ
biodegradation of organic contaminants, in Proc.: Management of
Uncontrolled Haz. Waste Sites, Washington, pp. 333-8.
Zeyer, J., Kuhn, E.P. and Schwarzenbach, R.P. (1986) Rapid microbial
mineralization of toluene and 1,3-dimethylbenzene in the absence of
molecular oxygen, Appl. Environ. Microbiol., 52, 944-7.
 CHAPTER 22

Soil washing
Paul B. Trost

22.1 INTRODUCTION

Increasing costs, long-term liabilities, and bans on land disposal makes

disposal in hazardous waste landfills increasingly unattractive. The
high cost per tonne of incineration is further stimulating the rapid
development and application of new and innovative on-site detoxi-
fication/volume reduction techniques.
Soil washing is a technique applicable to volume reduction/on-site
detoxification, and can be defined as the removal and concentration of
organic and/or inorganic contaminants from excavated soils. Typically,
soil washing has developed along three parallel avenues of:
1. particle size fractionation;
2. aqueous-based systems using some form of mechanical and/or
chemical method to liberate and concentrate the contaminants; and
3. counter current decantation (CCD) employing solvents for organic
contaminants and acidslbases or chelating agents for inorganic
contaminants.

22.1.1 Particle size fractionation

Particle size fractionation is based on the premise that contaminant
distribution preferentially occurs in the fine fraction, e.g., <741lm or
-200 Mesh. In this small size range, clays and finely divided organics
predominantly occur. To apply this method, tumblers, trommels, high
pressure washing, sonification, hydrocyclones, or vibratory screens -
in conjunction with sieve screens - are used to separate the fine
fractions from the coarse fractions and thus achieve a volume reduction
and concentration of contaminants.

Geotechnical Practice for Waste Disposal.

Edited by David E. Daniel.
Published in 1993 by Chapman & Hall, London. ISBN 0412351706
 586 Soil washing

·:ONTAMiNANT
':O~LEC:IO~ Cieer.
v\G~er " i Dewctered Soil
~Q!-,
wcec,'(J
---'-------- ~1 U~etl'rned to
Excavation)
recesscf'"/ "d Surge Tcnk
CLer8" Material Baic .... :e
Feed C;ec-. ~ewatered Sail Wasr, Water

5J ~PD 8.5 ,pO 35.4 T:JJ

BC% SOlids 65% Sol;ds
i% Organ:c 5.9% C~gof'ic
5oi"as 40 ~pc. 4 T;:::JD 36 TP~
:v'c!e r 1 J ""PD 4 jPD '9.4 TPJ 13.4 TPD
0.5 ,PD nil

Fig. 22.1 Generalised flowsheet and material balance for 50 TPD soil wash
plant. Source: MTARRI Golden, Colorado.

22.1.2 Aqueous-based soil wash systems

Aqueous~based soil wash systems may or may not employ particle
sizing techniques but generally utilize methods such as froth flotation*
to concentrate the contaminants in only 5-15% of the original volume.
This concentration is achieved by the use of surfactants, alkaline
agents, and/or minor amounts of solvents introduced into an aqueous
slurry containing the sized or crushed soil. The slurry is 'conditioned'
in a pre~mix step to liberate the contaminants from the soil mineral
particles then pumped to the froth flotation cells to achieve separation I
concentration of contaminants into the froth. The cleaned soil is
dewatered and recycled to its original excavation site. Please see the
flowsheet in Fig. 22.1.

22.1.3 Counter current decantation (CCO)

CCO is a commonly applied technique in the chemical and minerals
processing business. Basically, a series (1-7) of thickeners is utilized

*Froth flotation is utilized by the mineral processing industry and should not be con-
fused with dissolved air flotation (DAF).
 History/status of soil washing 587
HX

- - - - Solution :vemert

Clean
~So;1
V.etal
(Teils)
Bearing
Solution

Sol;ds Movemen~ - - - _

Fig. 22.2 Schematic flowsheet: metal removal by counter current decantation.

Source: MT ARRI Golden, Colorado.

with the contaminated soil introduced into tank 1 allowed to settle and
then pumped from tank 1 to tank 2, and on to tank 3, etc. The solvent,
acidlbase, orchelating agent is pumped into the last tank (e.g., tank 3),
allowed to mix with the soil then the liquid overflow containing con-
taminant and solvent is pumped to tank 2 and that overflow to tank 1
(Fig. 22.2). Thus, the equilibrium constant is shifted, driving the
contaminants out of the soil and into the liquid phase.
A modification of this CCO system is a solvent extraction process
similar to a small oil refinery where the contaminants are removed in
vessels/towers utilizing unique solvents and/or critical fluids.

22.2 HISTORY/STATUS OF SOIL WASHING

Soil washing in the US has been investigated by the EPA at their

Edison, New Jersey, facility since approximately 1982. Their operation
consisted of high pressure water jets to achieve soil matrix breakup
followed by screening and aqueous-based counter current decantation
(CCO). Various companies in the US have conducted extensive bench-
scale, and/or pilot-scale tests on different soil wash systems. Although
no full-scale cleanup has been achieved as of late 1989, at least one
is anticipated to be initiated in 1990. Numerous companies are simul-
taneously pursuing the particle sizing, aqueous-based systems, CCO
systems and/or solvent extraction methods.
In Europe, full-scale cleanups using soil washing are presently
occurring. Typically, these soil wash systems are fixed or semi-per-
manent installations capable of treating 200-600 tonnes/day at costs of
$100-200/tonne. Particle sizing is generally an integral part of their
systems together with aqueous-based washing employing either froth
588 Soil washing

flotation, ultrasonic vibration, low frequency vibration, solid-liquid

separation, hydrocyclones and/or flocculation.

22.3 REGULATORY AND COMMUNITY ACCEPTANCE

The USEP A is actively promoting the continued development of

soil washing technologies, and recent consent orders which specify
cleanup levels and methods have included soil washing. A trend
appears to be developing which, under these negotiated consent
orders, the cleanup levels are slightly rising. Thus, if this trend con-
tinues, the market for soil wash may increase. If cleanup levels re-
quired a 99.99% removal rate of the contaminant, the market for soil
washing would be limited. However, if contaminant removal rates are
lowered to the 92-99% range, the market for soil washing will expand
significantly. Soil washing operates as a volume reduction system with
the contaminants going off site for final disposal/incineration.

22.4 FACTORS INHIBITING SOIL WASH TREATMENT

One main problem exists for soil wash treatment of contaminated soils,
i.e., the wide variability of conditions encountered at most contami-
nated sites. Variability occurs both in the type of contaminants and
the nature of the soil mineralogy. This wide variability requires that
significant bench-scale testing be completed prior to ascertaining if
a particular soil wash technology is applicable and can achieve the
cleanup level goals at a competitive price. Because of this needed
bench-scale testing combined with the relatively high contaminant
removal rates required, soil washing may be more applicable to larger
volume contaminated sites as opposed to smaller volume sites; thus,
allowing more resources (time and money) for bench-scale testing to
achieve these low cleanup levels. A likely breakpoint between soil
washing vs. haul and dispose and/or incinerate appears to be about
2500m3 .

22.4.1 Contaminant variability

Contaminant variability at most sites requiring remediation appears to
be the norm rather than the exception. For example, at one Superfund
site the main contaminant was polychlorinated biphenols (PCBs), both
in the soil and in the groundwater. However, since transformers were
filled with a mixture of PCBs and mineral oil, there is far more mass of
mineral oil than PCBs. Also, this same site was previously a coal tar
gas manufacturing site; thus, the PCB-rich mineral oil has migrated
 Factors inhibiting soil wash treatment 589

Table 22.1 Organic contaminants at superfund

sites

compound average order oj

concentration occurrence
(mglkg)
~~-~--- -~~

TCE 103 highest

toluene 1121 I
benzene 17 I
chloroform <1 I
PCBs NA I
1,1,1-TCA NA I
PCE 540 I
phenol 511 I
xylene 8400 I
ethylbenzene 3200 I
PAH NA lowest

Notes: concentrations may range from 1-100000mg/kg;

NA = not available. Source: Esposito (1988).

through soils previously contaminated with polynuclear aromatic

hydrocarbons (PAHs) such as chrysene and benzo(a)pyrene, and
probably cyanide (CN). Cleanup must include all these chemicals.
Soil washing, if applied, must therefore simultaneously meet clean-
up criteria for all of these contaminants, which have very different
chemical activities.
Table 22.1 shows the order of occurrence and average concentration
of the most commonly occurring organic contaminants at Superfund
sites in the US. Typically, more than one of these organic contaminants
is present at every site.
Although low in ranking, a significant cleanup target has been PAHs
associated with past wood-treating operations (e.g., creosote) and
coal gasification sites. Semi-volatile organic contaminants of signi-
ficant concern are benzo(a)pyrene, dibenzo(a,h)anthracene, benzo(a)-
anthracene, chrysene, anthracene, and naphthalene. Cleanup levels for
these compounds may be 1 mg/kg or less with their total not to exceed
lOmg/kg.
Table 22.2 shows the average concentration of inorganic contami-
nants in Superfund-type soils and their relative order of occurrence.
Not only is soil washing hampered by the wide variability of con-
taminants, but also by the mixture of both organic and inorganic
contaminants. This mixture is shown in Table 22.3.
Soil washing of organic and inorganic contaminated soils usually
requires combining two operational processes. Organic contaminant
removal may be accomplished by froth flotation or a solvent extraction
process; however, inorganic contaminant removal will require a CCO
590 Soil washing
Table 22.2 Inorganic contaminants at superfund
sites

elements average order of

concentration occurrence
(mg/kg)

Pb 3100 highest
Cr 370 I
Zn 5000 I
Cd 180 I
As 90 I
Cu 2100 I
Hg NA I
CN 54 I
Ni 200 lowest

Notes: concentrations may range from <1-100000mg/kg;

NA = not available. Source: Esposito (1988).

Table 22.3 Sites with mixed contaminants

historical usage of organic contaminant inorganic

contaminated sites contaminant
----------- ---------- - - - - - - - - -

Wood treating sites PAH, PCP Cu, Cr, As

Coal gas manufacturing sites PAH CN
Pesticide/herbicide DDT, Endrin, Aldrin Dieldrin, As
man ufacturing etc., Arsenious acid
Electronic manufacturing TCE,PCE Cu
Oil refinery sites BETX*, PAHs, Crude oil Cr, Pb, V

* BETX = benzene, ethylbenzene, toluene, xylene.

circuit or other leach circuit having acidslbases/chelating agents to

solubilize the metals followed by precipitation of the removed metals
or a significant volume of finer rejection.
The need for two operational processes will therefore increase the
cost and may seriously impact any economic advantage of soil wash-
ing. This economic impact will be most seriously felt for on-site treat-
ment at smaller sites, e.g., <10000 tonnes where bench-scale testing,
process optimization, and mob/demob costs will have a proportion-
ately larger cost/tonne than for larger sites.

22.4.2 Mineralogic variability

Esposito (1988) cited PEl Associates' 1987 study of 151 contaminated
sites in the US to show that contaminated soil is more likely to be
composed of clay to sandy-clay soil as shown in Table 22.4.
 Distribution of contaminants in soils 591

Table 22.4 Soil types at superfund sites

soil type % of occurrence

Sandy-clay 30
Clay 10
Sandy 20
No information 40

Source: Esposito (1988).

Thus, the soil wash system should be capable of handling relatively

high percentages of fine-grained to clay-sized fractions, otherwise the
system may have limited applicability and thus limited economic
advantage due to high disposal costs of the contaminant-rich soil
fraction.

22.5 DISTRIBUTION OF CONTAMINANTS IN SOILS

Generally contaminants are more concentrated in the finer fractions of

the soil particles than in the coarser fraction. However, a significant
amount of contaminants may still be present in the coarse particles,
which also must be removed to achieve the cleanup goals. For example,
Table 22.5 shows the distribution of pentachlorophenol (PCP), an
organic contaminant, as a function of particle size.
Removal of the -4.5 f..lm size fraction by screening would result in a
weight loss of only 6.5% of the total feed weight but would achieve
removal of 65.5% of the contained PCP and, secondarily, would allow
higher throughput of the remaining soil through a soil washing process
by decreasing the settling and thickening times of the residual washed
soil. Thus, screening for this soil is a very attractive alternative based
on the projected economics of total treatment.
Table 22.6 shows distribution of selected inorganic contaminants as a
function of soil particle size for an example site. If the soil in Table 22.6
were screened with the -lO!lm size fraction removed, the weight to
be disposed of would be 31.2% of the original feed weight and would
contain 68.6% of the total arsenic present. Unfortunately, the Pb
removal would not be as effective as the As and, thus, additional
treatment would be required.
Contrarily, Table 22.7 shows that the distribution of PCBs in a
quartz-rich but fine-grained soil are not particularly concentrated in the
finer soil fraction and therefore very little advantage would be gained
by screening and removal of the fine fraction since removal of 51.4% of
the total soil weight would only achieve 73.8% removal of the PCBs.
For the clay-size fractions, sorption of inorganic cations is well
documented (Chapters 2 and 3); clays also will sorb many organic
592 Soil washing
Table 22.5 Example of distribution of pentachloro-
phenol as a function of soil size fractions

size fraction (Ilm) wt% of % distribution

fraction PCP

+45 89.5 28.5

-45 to +4.5 4.0 6.0
-4.5 6.5 65.5

Table 22.6 Example of distribution of inorganic contaminants as a function of

grain size

size wt% As wt% Cd wt% Fe wt% Pb wt%

fraction fraction dist. dist. dist. dist.
--------------

+30 Mesh 8.9 800 6 60 7.2 27000 9.7 500 6.7

-30 to +325 41.5 340 11.7 33 17.8 8900 14.4 220 13.3
-325 to +2OIl 12.1 670 6.9 54 8.7 12000 5.8 680 12.3
-20 to +10 7.3 1100 6.8 110 10.7 18000 5.2 1300 14.1
-10 to +5 4.2 2000 7.1 160 9.0 34000 5.7 1500 9.4
-5 27.0 2700 61.5 130 46.7 55000 59.2 1100 44.2

Table 22.7 Distribution of PCBs in soil fractions

size fraction (11m) wt% of % distribution

fraction of PCB

+44 (coarse silt) 48.6 26.2

+20 to-44 40.8 55.0
+10 to -20 7.5 12.2
-10 3.1 6.6

compounds. For the larger sized particles, contaminant sorption may

be partially explained by any or all of the following mechanisms:
1. sorption of cations by iron or manganese oxides coating individual
mineral grains;
2. sorption of organic contaminants by iron and/or manganese oxides
which have previously sorbed naturally occurring humates;
3. sorption of organic contaminants by naturally occurring humic acids
which are coating individual mineral grains;
4. flocculation/precipitation; and
5. combination of any or all of the above.
Removal of contaminants from these larger sized particles is thus
a prime requirement of soil washing techniques. For example, the
 Application of soil wash processes 593
MT ARRI process utilizes an aqueous system with froth flotation.
Generally, higher removal rates of organic contaminants have been
observed in a basic system as opposed to a neutral or acid system. One
possible explanation would be the increased solubility of humic acids
in higher pH systems, thus breaking the humate-contaminant bond
and/or solubilizing the humate and contaminant off the mineral grain.

22.6 APPLICATION OF SOIL WASH PROCESSES

As previously stated, the biggest problem with application of soil

washing processes is the diversity of the contaminants combined with
the diversity of the mineralogy of the soil. Thus, prudent cpplication of
any soil washing system will generally be in four stages: bench-scale
testing; process engineering; field pilot testing; and full-scale operation.

Phase I - bench-scale testing

Objective:
To conduct bench-scale testing on 100-1000 gm of contaminated soil to
achieve the maximum removal of contaminants and/or meet cleanup
levels while simultaneously producing the minimum quantities of
reject material - containing the contaminant - which requires final
disposal by incineration or burial.

Time:
2-12 weeks

Phase II - process engineering

Objective:
Based on the technical success of the bench-scale testing, conduct
process engineering/optimization studies and tests to identify changes
or modifications required by a particular soil wash process to treat a
mineralogically unique soil.

Time:
1-24 weeks

Phase III - field pilot testing

Objective:
For larger size cleanups, e.g., >10000m 3 , conduct pilot testing on site
to determine operating parameters, final cleanup costs, and establish
that cleanup levels can be achieved in an environmentally safe manner,
594 Soil washing
e.g., with minimal air emissions, spill control, process water treatment,
and at an acceptable price.

Time:
1-4 weeks

Phase IV - full-scale operation

Objective:
Conduct the full-scale remediation and replace the cleaned soil back
into the original excavation site.

Time:
1-4 years

The above four phases would typically be applied to projects having

greater than 10 000 m 3 and with diverse contaminants present. How-
ever, for smaller size projects where the contaminants can be easily
removed, e.g., volatile organic contaminants, the bench-scale testing,
and process engineering steps should require minimal time and cost
and a pilot test may not be required. This assumes the soil washing
process to be applied has been sufficiently field tested so general
operating parameters are already defined.
Europeans in Germany and the Netherlands have installed fixed soil
washing facilities. Thus, no pilot test is needed and minimal process
engineering testing/design is required for fixed time facilities since
their operating parameters have been well defined.

22.7 DESCRIPTION OF PROCESSES

22.7.1 Operations using sizing, high pressure water washing

and additives

EPA mobile extraction system

Process description
The USEPA has conducted both bench-scale and field testing of their
soil washing process. Generally, their soil washing process relied on
screening and removal of the >25mm sized material followed by rotary
drum washing of the <25 mm material. The <2 mm size fraction was
subjected to counter current washing with water and in some cases
additives. If cleaned, the <2 mm was filterpressed and deemed capable
of being returned to the excavation site; if contaminant removal was
insufficient, the soil was disposed of by land filling and/or incineration.
The >2mm soil particles were subjected to high pressure water wash-
 Description of processes 595

ing in the rotary drum with a 'water knife' and then discharged as
clean.
Waste streams are process water and generally the <2mm sized
fractions. Process water was passed through activated carbon prior to
discharge, whereas the <2 mm fraction generally had to be subjected to
subsequent treatment, e.g., solidification or incineration.

Application/results
Testing was conducted on both organic and inorganic contaminated
soils. According to a paper published by Scholz and Milanowski (1983),
for large particles (2-12.7 mm) removal rates were: phenol - 99.1 %,
As 2 0 3 - 52.1%, and PCB - 28.6% after 120 minutes of washing. For
the <2 mm sized fractions, the counter current extraction produced
removal rates of: phenol - 77.8-98.6%, As 2 0 3 - 85% (at pH 1), and
PCB - 37.5% (using Tween 80 surfactant).

Advantages/disadvantages
High processing rates can be achieved by using the rotary drum with
high pressure washing. Unfortunately, residual contaminant levels in
both the coarse and fine-grained size fractions appear to be relatively
high and thus may not meet cleanup standards for many sites. The
system was designed to be mobile and occupied three standard size
flatbed trailers.

Stage of development
Bench-scale tests and pilot-scale field operations have been conducted.
No full-scale cleanup was attempted.

Heijmans Millieutchniek BV (Nunno et al., 1989)

Process description
This soil washing process is based primarily on particle sizing under
the assumption the majority of the contaminants are concentrated in
the finer fractions. All <631lm sized particles are screened out and
rejected, whereas the >631lm particles are washed with detergents
and oxidants, the contaminants removed by precipitation and/or
flocculation.

Areas of application
Cyanides, heavy metals, mineral oil, and kerosine where the fines
«63 11m) are less than 20-30% by weight.

Results
Table 22.8 shows contaminant removal rates as reported by Nunno et
al. (1989).
596 Soil washing

Table 22.8 Soil washing results for the Heijmans Millieutchniek BV process

soil type contaminant before (mg/kg) after (mg/kg)

Silt/sand total CN 250-500 10-15
Cr 43-45 11-15
Ni 250-890 40-70
Zn 460-720 140-200
Coarse sand kerosine 5000-7000 80-120
Coarse sand PNA 250-400 0.5-10
Silt/fine sand mineral oil 3000-8000 90-120

Source: Nunno et al. (1989).

Advantages/disadvantages
May have a high reject rate for finer-grained soils. Since the reject
material must be further treated or disposed of, treatment costs per
total tonne treated will far exceed the reported operating costs.

Stage of development
Operating full-scale as a fixed-base facility.

HWZ Bodemsanering, Amersfont, Netherlands

Process description
According to Nunno et al. (1989) the HWZ process is very similar to the
Heijmans process described above.

Areas of application
As described above.

Results
See Table 22.9.

Advantages/disadvantages
Treatment of fine-grained soils appears to be a major problem; also,
removal rates for both organic and inorganic contaminants may not be
high enough to meet US standards.

Stage of development
Operating full-scale.

Harbaurer GmbH. Berlin, West Germany

Process description
Although this process also uses particle sizing, it rejects only the
<151lm clay-sized particles. Reportedly (Nunno et al., 1989), low
 Description of processes 597
Table 22.9 Soil washing results for HWZ process

contaminant before (mglkg) after (mglkg)

--~ ------- ----- ~-----~--

PNA 100-150 15-20

CN 100-200 5-15
Chlorinated
hydrocarbons 20-30 <1
Pb 300 75-125

Table 22.10 Soil wash results Harbaurer process

contaminant before (mglkg) after (mglkg)

Total organics 5403 201

Phenol 115 7
Chlorinated organics 90.3 NO
PCBs 3.2 0.5

frequency vibration with detergents and/or other extractants sig-

nificantly aids in desorption of contaminants from the remaining finer-
grained particles. Water is treated by flotation, air stripping, ion
exchange, and finally activated carbon.

Areas of application
Organic contaminants, phenols, PNA, PCBs.

Results
See Table 22.10.

Advantages/disadvantages
Lower overall operating costs are achieved by only rejecting the
<15 ~m material. This soil washing process was the only one reported
by Nunno et al. (1989) that showed PCB removal rates, and hence may
be applicable for PCB contaminated soils.

Stage of development
Full-scale operation at a fixed facility.

22.7.2 Operations using water washing and froth flotation

MTARRI process
Description
Solid feed from the site of excavation is classified by a screening,
crushing or other means, and then fed to a mixing tank. Typically, all
598 Soil washing

Table 22.11 Waste-Tech bench-scale test results

contaminant before after

(mglkg) (mglkg)
--------- -------- - - -

Gasoline 54400 29
Benzene + toluene 38 ND
Chlorinated hydrocarbons 40-100 ND
e.g., TCE, PCE
PNA 6-170 0.04-1.0
PCB 470 6.1
Cr 2590 104
Zn 3540 60
Hg 157 <0.8
As 12100 <5.4

fines are retained. At the mIxmg tank chemical reagents of caustic

and possibly surfactants are added together with recycled water. The
slurry, being approximately 25% solids is then pumped to the froth
flotation cells.
Residence time in the flotation cells is dependent on the nature of
the contaminant and the soil type. At the flotation cells additional
reagents are added as needed and the froth, containing the organic
and a portion of the inorganic contaminants, is skimmed off. Solids
progress continuously through the system with residence time per cell
adjusted by means of weir gates. Solids are then fed to a thickener
and if desired a belt filter press. Separated water is recycled to the
mixing tank with bleed off water amenable to standard water treat-
ment processes. If the metal content is sufficiently high, a subsequent
counter current decantation (CCO) process can be employed to remove
the residual inorganic contaminants.

Applications
The process has been bench-scale tested on sludges, creosote con-
taminated soils, gasoline and diesel contaminated soils, and soils
having both volatile and semi-volatile contaminants, PCBs, As, Cr, and
Zn. Recent field testing of a mobile 50-tonne-per-day soil washing unit
yielded removal rates from kerosene-contaminated soils that exceeded
anticipated results from bench-scale tests.

Results
See Table 22.11.

Advantages/disadvantages
By retaining most of the feed material and thus minimizing rejecting
the fines, overall operating costs can be significantly reduced. If the
 Description of processes 599

Table 22.12 Heidemij Uitvoering operating results

contaminant before (mg/kg) after (mg/kg)

---------- - --- -- ~----

PNA 19 0.34
Chlorinated
hydrocarbons 5.3 0.4
Pesticides 650 14.4
Oils, hydrocarbons >1000 65
Toluene, benzene 3000-18000 20
Hg 11900 110
As 67 1.5

cleanup levels achieved during bench-scale testing can be consistently

achieved in the field, many of the US cleanup standards can be met.
Disposal of the contaminant-rich froth equal to 5-15% of the original
feed weight will be the other major expense. Typically, process water is
recycled, thus disposal costs for water treatment is relatively low.
Flotation cells were purchased with covers thus allowing air emission
standards to be met. Treatment of high clay soils has been successfully
demonstrated at the bench-scale.

Stage of development
Bench-scale testing and minor pilot-scale testing.

Heidemij Uitvoering, BV's Hertogenbosch, The Netherlands

Process description
Soils are initially screened to remove the <50 11m size fraction and
hopefully a significant amount of contaminants. After screening, froth
flotation is utilized to achieve separation and concentration of the
contaminants from the mineral particles. The slurry is then settled, and
washed according to Nunno et al. (1989).

Areas of application
eN, heavy metals, oils, pesticides. The system was not applicable to
PCB removal.

Results
See Table 22.12.

Advantages/disadvantages
As with other froth flotation soil washing systems, quite high con-
taminant removal rates may be achieved. The main disadvantage is the
apparent lack of certain PCB and PNA removal capabilities.
600 Soil washing

Table 22.13 Soil washing results for Biotrol process (bench scale)

soil type contaminant before (mglkg) after (mglkg)

- - - - --- -------- - - - - - -

Silty sands Pentachlorophenol 218 22

Phenanthrene 102 1
Fluoranthrene 16 0.5
Benzo(a)pyrene 2.7 0.002
Pyrene 16.7 0.36

Stage of development
Field operational.

Biotrol
Process description
Biotrol screens the feed soils to obtain a feed for soil washing. Screen-
ing rejects >12 mm material and the clay-sized material. The > 12 mm
material is treated by aqueous scrubbing, whereas the clay fraction is
rejected to be disposed of by incineration or other treatment process.
The -12 mm down to clay-size material is then soil washed with
surfactants in a counter current system using froth flotation. Con-
centrated contaminants produced in the froth are then also disposed of
by further treatment, e.g., incineration.

Areas of application
Pentachlorophenol and creosote-contaminated soils have been treated
by this process.

Results
See Table 22.13.

Advantages/disadvantages
May have a high reject rate for finer-grained soils, thus total treatment
costs may significantly increase. Cleanup levels appear to be low
enough to meet many US standards.

Stage of development
A 250 kg per hour demonstration unit has been built and field tested.

22.7.3 Operations using solvents and/or solvent extraction

BEST process
A unique solvent extraction process (originally developed for organic
extraction of a sewage sludge) has been modified, developed, and
 Description of processes 601

field tested for hazardous waste. This solvent extraction process uses
triethylamine which is soluble in oil and water below 18°C and only
in oil above 54 0c. A two phase separation is then achieved and
the triethylamine recovered for recycling by distillation. Solids are
separated in a centrifuge and then dried. The process can operate at
100-1000 tonne per day. A 100 tonne per day skid-mounted plant is
transportable to the site with four truck loads.

Applications
This process is amenable to separator sludges, other toxic organic
sludges, and wood treating sludges from creosote. Additionally, the
process has been tested on PCB-contaminated soils and sediments.

Results
According to Best, the process will remove 99+% of the organic con-
taminants and has been bench scale tested to remove 99% of PCB oil
from a soil. Field testing has been completed on refinery sludges and
the processed soils passed required leachability tests.

Advantages/disadvantages
The process is mobile but resembles a mini-refinery and therefore
has limited mobility. However, for large remediation projects, e.g.,
20 000 m 3 , the system may be applicable. No metal removal is achieved,
therefore requiring possible further treatment.

Stage of development
A mobile, field-operational unit has been employed in cleanups.

CF systems
Removal of unwanted chemical compounds by extraction with critical
fluids is commonly practiced in the food and chemical industry.
For example, the removal of caffeine from coffee beans is currently
accomplished by using carbon dioxide as a critical fluid. This tech-
nology has recently been transferred to the hazardous waste business.
Critical fluids commonly employed in hazardous waste treatment are
CO 2, propane or other light hydrocarbon gases. Their advantages
are low cost, low toxicity and ease of separation from the extracted
contaminants.

Description
Sludges are slurried to a conditioner tank where a homogeneous
slurry mixture is produced and then fed to the solvent extraction unit.
Recycled solvent also enters the extractor. From the extractor the slurry
is pumped to a separator where the contaminant and critical fluid is
602 Soil washing
initially separated from the solids and water. A second solid/liquid
separator, involving settling and filtration, produces a solids stream
and a water stream. Solvent can be recovered from the water at this
stage according to CFS. Solvents and organic contaminants are fed to a
solvent recovery still where the solvents are recovered, compressed
and recycled to the extractor with the heat from the recompression
used to vaporize the solvent. The organic contaminant is then recovered.

Areas of application
The process has apparently been tested on separator sludges and oil
contaminated soil. Presumably the technique would be applicable to
both volatile and semi-volatile contaminants.

Results
According to CFS literature, removal efficiencies of organic con-
taminants from the solids range from 87-98% (as a percentage of
feed). However, the metal concentrations (as a percentage of feed)
were found to increase due to weight removal of water and organic
contaminant. Purgeable organics (benzene, toluene, xylene) showed a
95-99+% removal and the solids passed the TCLP test. Semi-volatiles
such as naphthalene and phenanthrene showed a 99+% removal and
also passed the TCLP test. Recent testing on PCB-contaminated soils
was reportedly encouraging.

Advantages/disadvantages
High contaminant removal rates may meet US cleanup level require-
ments for many sites. This technology may be especially suitable for
emulsion treatment.

Results
As with the BEST system, removal of inorganic contaminants is not
generally accomplished simultaneously with removal of the organic
contaminants, hence a second step to remove metal (inorganic) con-
taminants is required, such as fixation.

Stage of development
A pilot-scale unit is currently being field tested.

ENSR process
Description
Soil is sized to <12 mm and then fed to a mixer where solvent and
other chemical reagents are added to form a 33-50% (by weight)
slurry. This slurry is pumped to a series of 3-5 counter current extrac-
 References 603
Table 22.14 ENSR bench scale results

contaminant concentratioll concentration

before treatmellt after treatmellt
(mg/kg) (mg/kg)

PCB 1944 1
PCB 3300 non-detectable
PCB 1147 22

tion cells to achieve contaminant removal. Cleaned soil is dried and

residual solvent removed. The contaminated solvent is distilled to
recover and recycle the solvent, whereas the still bottoms - rich In
contaminant - are sent for off-site disposal/destruction.

Application
Applicable to most organic contaminants and PCBs.

Results
See Table 22.14.

Advantages/disadvantages
High moisture content in soils may affect processing rates and/or
cleanup levels attainable. However, that problem could be solved by
predrying of the soil feed. Cleanup levels for PCBs are well within the
standards set by the USEPA and therefore this process may be both
technically and economically viable for PCB removal. As with other
solvent extraction processes, the method does not address metal con-
taminant removal if a mixed organic/inorganic waste is present.

Stage of development
Bench-scale.

REFERENCES

Esposito, P. (1988) Characterization of RCRA/CERCLA sites with contaminated

soil, presented at EPA Workshop on Extractive Treatment of Excavated
Soil, Dec 1-2, Edison, NJ.
Nunno, T.J., Hyman, J.A. and Pheiffer, T. (1989) Part 7: European approaches
to site remediation, HMC Magazille, Sept/Oct, 38-46.
Scholz, R. and Milanowski, J. (1983) Mobile system for extracting spilled hazardous
materials from excavated soils. EPA-600/S2-83-100, US Environmental
Protection Agency.
 PART FOUR

Monitoring
 CHAPTER 23

Monitoring wells
Bob Kent and Mark P. Hemingway

23.1 INTRODUCTION

23.1.1 Background
The installation of ground-water monitoring wells has developed over
the past decade into a highly specialized activity, quite distinct from
the fields of water supply well and geotechnical drilling from which
it was derived. Stringent regulatory demands, sensitive analytical
techniques, and exotic construction materials all require the monitoring
well designer to comprehend a broad spectrum of technologies and
practices.
The purpose of a monitoring well is, ultimately, to obtain a representa-
tive sample or measurement from the ground water within a particular
hydrogeologic stratum or group of strata. This sample or measurement
may be used to ascertain the presence or absence of introduced con-
taminants, as well as the concentrations of particular constituents.
While this objective is conceptually simple, difficulty arises from the
fact that, as with Heisenberg's electron, the act of observing alters that
which is being observed.
The installation of a well is a process which is traumatic from the
perspective of the aquifer. Traces of hydrocarbon fuels and lubricants
may be introduced into the subsurface from the drilling rig hydraulics,
packings, and exhaust, even when no leaks from these sources are
visible. Drilling fluids mix with and alter the natural waters in the
subsurface, and also produce a cake of foreign material on the borehole
wall. Fragments of metal from the drill bit or auger flights are eroded
and remain in the borehole. Any materials used to construct the well
must interact to some degree with the water entering the well from the

Geotechnical Practice jor Waste Disposal.

Edited by David E. Daniel.
Published in 1993 by Chapman & Hall, London. ISBN 0 412 35170 6
608 Monitoring wells

surrounding soils. Under the existing state of the science and engineer-
ing of well installation, these influences are unavoidable - they can be
minimized, but not entirely eliminated.
From a realistic standpoint, then, the question is not whether the act
of ground-water monitoring will affect the ground water, but rather
the permissible extent of these effects. Given this, the objective
of any designer of monitoring wells must be to determine the true
purpose of the proposed well or well system, and to select practical
and economical installation practices and protocols which address that
purpose. This should include as large a contingency as possible for
difficulties and problems which may arise during the installation
process.
Design under such a philosophy is more than simply a rote obedience
to a set of guidelines. It requires some pragmatic awareness of the
potential for problems and inaccuracies in the field practices utilized.
Installing a monitoring well is typically a difficult and expensive
process. There may be no second chance to accomplish a given task
following an initial failure. Sound monitoring well design should there-
fore use methods which are more than simply possible; they must have
an intrinsically high probability of success. The purpose of this chapter
is to acquaint the reader with the practices of monitoring well design,
installation, and sampling, with an emphasis on the realities involved
in these activities.

23.1.2 Components of a monitoring well

Although well designs will vary widely depending upon purpose and
installation setting, most wells have broadly similar configurations and
components. The arrangement and nomenclature of these components
for an idealized well are illustrated in Fig. 23.1a. The terminology used
for particular components is subject to wide personal and regional
variation, so the terminology used here may not correspond perfectly
with that used by others.
A typical well consists most fundamentally of a length of pipe
extending from the ground surface into the water-bearing stratum
of interest. This penetration is usually accomplished by means of a
borehole, which may be advanced using a variety of drilling methods.
Within the targeted stratum, the pipe is perforated in some manner to
permit the entry of ground water for sampling. The perforated section
of the pipe is termed the well screen, and the unperforated section
extending to the surface is termed the riser. Some wells are equipped
with a short length of unperforated pipe below the screen called a
sediment sump, for the collection of sediment which enters the well.
The sump, screen, and riser may be collectively termed the well
assembly or well string, as distinguished from the term 'well', which
 Introduction 609
(a) (b)
Well plug _--,,""_Well plug
LockIng cover lockIng cover -
ProtectIve caSIng -

. ""~ tr.::~:~
.:....;.
Uppermost . .
wate<' bcanng . .

. ~~~~~~~r: ~;l~.. . Well screen

~ ~i; : ~ .~:~.~:e~::.
..
monItored

----,..- ----------- :.}:.

'111~iL
~:,..: -------- ..
:::(~'. -----------------

~ ~~~ ~~~~~;!~ump
~==~:=:~: ·,'r;:.:,,:,.: :~~~ :=:=~~t;~~::-:
borehole
- - - - - - - - - - - - _____ . Tocal depth - - - __________ ______ Total depth
----:...---:...-- -------:...-:..- --------:..-_-_ 01
--:...--------~--------------------------~-:::....-:
:-:-:~:~:-:-:-:-:-:-:~-:::-:~_::of borehole
Fig. 23.1 Components of an idealized monitoring well: (a) single-cased and
(b) multiple-cased.

includes both the well assembly, the borehole, and annular materials,
as well as surficial structures.
The annulus surrounding the screen is typically backfilled with
granular material, such as silica sand, to form a filter pack. The primary
purposes of this filter pack are to minimize the amount of fine sedi-
ment which enters the well with the ground-water, and to minimize
collapse of the borehole surrounding the screen. The annulus be-
tween the filter pack and the surface is sealed using a variety of low
permeability materials, such as bentonite clay and Portland cement.
This reduces the entry of surficial waters or waters from non-targeted
water-bearing strata into the well screen.
At the surface, the riser may be protected against vandalism and
vehicular damage by a metal protective casing or a manhole. A surface
seal of concrete is typically placed around this protective structure to
prevent surface drainage into the borehole.
610 Monitoring wells

For wells with multiple casings (illustrated in Fig. 23.1b), the well
installation process is preceded by the installation of a non-perforated
pipe string, which is sealed in place. This pipe string, termed a casing,
allows the isolation of shallow water-bearing zones which are undesir-
able due to contamination or other factors prior to well installation.
After the casing is installed, the borehole for the well is advanced and
the well installed through the casing. If only one casing is utilized, it is
typically termed the surface casing.

23.2 MONITORING WELL PLACEMENT

Most guidance suggests a minimum of four monitoring wells per

potential source of ground-water contaminants. Three of these wells
would be placed downgradient of the potential source, and one would
be placed upgradient (USEPA, 1986). The upgradient well permits the
assessment of the background character of the site's ground water,
and the downgradient wells permit the detection of any contaminant
plumes emanating from the source. This configuration would be the
simplest possible, and may be adequate where hydrologic conditions
and contaminant movement are very straightforward.
The proper placement of wells at most sites cannot be dealt with in
such a simple manner. The depth to water, characteristics of flow, and
pattern of aquifers and aquitards in both the saturated and unsaturated
zone all affect the movement of contaminants. Aquifers of high
hydraulic conductivity, for example, tend to promote the formation of
narrow, elongated contaminant plumes, while broader plumes tend to
develop within aquifers of lower hydraulic conductivity. The optimal
placement of monitoring wells requires, therefore, both a basic under-
standing of the site hydrogeology, and of the possible migration paths
of the suspected contaminant plume. Information on the latter topic is
provided in examinations of the movement of immiscible and dissolved
phase constituents through the subsurface performed by Schwille
(1984, 1989), Mackay et al. (1985), Hinchee and Reisinger (1987), and
Mercer and Cohen (1990).
One example of the influence of site-specific hydrologic conditions
on the movement of contaminants was documented by Haug et al.
(1989). The subject site was underlain predominantly by low hydraulic
conductivity soils, but also included several narrow, elongated sand
bodies. These sand strata behaved as preferential pathways for con-
taminant movement due to their higher hydraulic conductivity, but
could have easily been missed by a cursory subsurface investigation
due to their low width. An effective ground-water monitoring program
at the site had to account for the presence of these features.
A second example would result from the presence of a low hydraulic
 Monitoring well design 611

SURFICIAL SOURCE

SURFICIAL CLAYS ~~~~~'74~~(LL../.+-,-~

SAND

WATER TABLE

Fig. 23.2 Conceptual view of the potential effect of a shallow low hydraulic
conductivity layer on the downward migration of contaminants through the
unsaturated zone. Impact to the water table aquifer directly below the
contaminant source is diverted to another area.

conductivity layer above the water table. Such a layer could alter the
downward movement of contaminants from a surficial source (Fig.
23.2). Perched water above this confining layer might experience
substantial impact, with little or no contamination detectable in the
ground water from below the water table.
Given the number of factors which can playa controlling role in the
movement of hydrocarbons through the subsurface, the selection of
proper well locations can be a complex process, especially for sites
within complicated hydrogeologic settings. For such a setting, many of
the issues affecting well location cannot be resolved until a part of the
investigative process is complete. For such settings, a phased approach
to subsurface investigations is most appropriate. Such an approach
requires an initial well installation phase, which establishes the basic
hydrogeologic parameters of the site. The data from this phase is then
utilized to select additional well locations, and to refine the monitoring
well design.

23.3 MONITORING WELL DESIGN

23.3.1 Monitoring well materials

Monitoring well riser and screen are available in a host of materials.
These include stainless and carbon steel, fluoropolymers such as
polytetrafluoroethene (PTFE), fiberglass, thermoplastics such as poly-
vinyl chloride (PVC) and polypropylene, and aluminum. The relative
advantages of these materials are summarized in Table 23.1. PVC is
612 Monitoring wells

perhaps the most commonly used monitoring well material, due to its
economy, strength, availability, and general inertness.
Most of these materials utilize threaded connections, or joints, to join
the lengths of pipe into a continuous well assembly. This is a depar-
ture from early monitoring well installation practices, when wells
were often constructed of glued-joint PVc. This practice was almost
universally discontinued when analyses from several investigations
indicated that the organic constituents of the cement used were bleed-
ing into the water in the well, diminishing the usefulness of the wells
(Fig. 23.3). Although glued-joint PVC may still be appropriate for use
when monitoring for constituents such as metals or inorganic salts, the
early problems with this type of joint have created an aversion to its
use.
The thread design on most brands of riser and screen varies con-
siderably from manufacturer to manufacturer. As a result, one brand of
monitoring well pipe is typically not compatible with another. The
quality of the threading and pipe is also highly variable from one
manufacturer to another, which reduces the value of many low cost
brands. The economy disappears when the joints fit together poorly,
resulting in a difficult installation and possible losses is well integrity.
The designer should use caution, therefore, when purchasing from an
unfamiliar manufacturer.
Two factors should govern the material of construction used. These
are material strength and chemical reactivity. The first parameter,
structural strength, is of greatest significance when designing deep
monitoring wells (>30 m). Under such circumstances, stresses upon
the well material become considerable. The crush and bursting
resistance, tensile strength, and joint strength of the material selected
must all be sufficient to endure these stresses, or well failure will
occur during or after installation. The importance of material strength
became obvious when PTFE well materials first appeared. The low
tensile strength and low coefficient of friction led to failures of the
threaded joints, as threads flexed and slid across one another during
installation.
A more common structural problem is faced when using PVC riser
on deep (>30 m) wells in conjunction with annular seals consisting of
cement or bentonite cement grouts. The hydration of the Portland
cement in these grouts can generate considerable heat, which can
soften the PVc, reducing its crush resistance. This, in turn can permit
deformation of the well riser from the differential pressure on the
outside riser. To moderate both the heat generation and the differential
pressure, cement grouts should be emplaced around PVC riser in
increments of approximately 20- 25 m, allowing each increment to
cure before emplacing the next. When necessary, the riser may also be
kept full of potable water, which absorbs heat and helps balance the
Table 23.1 Relative advantage of the commonly utilized monitoring well pipe construction materials

material polyvinyl chloride stainless steel carbon steel polytetrafluoroethylene

Chemical Generally good, Very good Very good, except for Very good, susceptible to
resistance susceptible to attack by corrosion attack only by certain
some separate-phase exotic fluorinated
hydrocarbons, especially compounds
ketones, esters, and
aromatics
Inertness Moderately good, some Very good Moderately good, although Moderately good, some
sorption effects corrosion enhances sorption effects
sorption effects
Corrosion Very good Very good, although may Susceptible to attack in Very good
resistance corrode and leach aggressive soils and
chromium in highly waters
acidic waters
Strength Weaker, less rigid than High strength at all High strength at all High impact strength, but
steel materials, soft at temperatures temperatures low tensile strength and
high temperatures, wear resistance compared
brittle at low to other plastics used for
well materials. Screen
slots may compress
closed with time
Cost Low Moderate for riser Low to moderate High
High for screens
Other Quality highly variable Much heavier than Much heavier than plastics Poor availability
Availability good plastics
614 Monitoring wells

(a) ,..--_ _ _ _ _ _ _ _ _--, (b)

Fig. 23.3 Gas chromatographic scans of water samples from: (a) groundwater
from a monitoring well constructed from glued-joint PVC pipe; and (b) distilled
water in which a sample of glued PVC was immersed for several hours. The
major peaks on the scans are 2-butanone, toluene, and tetrahydrofuran.

pressure differential. The disadvantage of the latter technique is

that the water lost to the aquifer through the well must typically be
recovered, lengthening the development effort.
The second characteristic of importance, chemical reactivity, reflects
the potential for interaction between the materials of well construction
and the ground water, natural dissolved constituents, and dissolved
and separate-phase contaminants. Specific concerns are:
• alteration of natural or contaminant chemistry by the well materials,
modifying the results of ground-water analyses;
• degradation of the well materials by the ground-water, such as
accelerated corrosion of carbon steels under aggressive ground-
water conditions; and
• degradation of the well materials by free-phase or high concentra-
tion dissolved phase organic constituents, such as the dissolution of
PVC by some organic liquids.
The first of these concerns, the alteration of ground-water sample
chemistry by the well material, typically receives the greatest attention.
A study by Parker et al. (1990) indicates that, for organic contaminants,
stainless steel affected constituent concentration the least. Interestingly,
PVC performed better in this regard than PTFE. The critical point,
 Monitoring well design 615
however, is that PVC did not modify the concentrations of any of the
tested constituents by more than 8% in a 24 hour period. Since the first
step in properly sampling ground water is the purging of the well,
the sampled water should be brought into first contact with the well
materials shortly prior to sampling. While the sorption/desorption
of constituents in wells within actual aquifers is substantially more
complex than the tested laboratory microcosms, these findings do
suggest PVC is adequate for ground-water monitoring purposes.

23.3.2 Monitoring well diameter

Although well materials exist in a wide array of sizes, from 10 mm to
in excess of 600 mm in diameter, practicality and standardization of
practice have narrowed the selection for most monitoring wells to
either 50-mm or 100-mm nominal diameters. Larger sizes are less fre-
quently used, since the increased production resulting from their
greater open surface screened area is not usually of critical importance
for ground-water sampling.
Narrowing the selection to these two sizes, however, still leaves the
monitoring system designer with a series of decisions. The selection of
a 50-mm diameter well lowers material costs and increases the ease of
well installation, but the lower open surface area and smaller size
render well development in poor aquifers difficult relative to the devel-
opment of 100-mm diameter wells. Gass (1988) discusses the preferability
of 100-mm wells in fine-grained strata to minimize sample turbidity. In
addition, if a pump is to be used for well development, purging,
sampling, or for conversion of the well from monitoring to ground-
water recovery, the use of 50-mm wells restricts the pump options
available. Very few types of electrical submersibles are sufficiently
small to be used in wells of that diameter.
Well string materials (i.e., riser and screen) for 100-mm wells, on the
other hand, are typically 2-4 times more expensive per unit length
than 50-mm diameter materials. If relatively costly materials of con-
struction such as PTFE or stainless steel are required, this factor can
substantially inflate the cost of installation. In addition, for hollow-
stem auger drilling and installation, 100-mm well strings require the
use of an oversized auger for through-the-stem installation. As will be
discussed later, the augers most commonly used in the environmental
drilling industry have stem inner diameters (IDs) of approximately
75-100 mm. The installation of a 100-mm diameter well (usually 113-
mm outer diameter (00)) requires the use of an auger with a stem 10
of at least 165 mm. Both drilling and sampling are more difficult with
these larger augers, slowing production and further increasing installa-
tion costs.
For most purposes, however, the selection of 100-mm diameter well
616 Monitoring wells

materials accrues advantages that offset these increased difficulties and

costs. The larger surface area of these materials permits much more
rapid and effective well development, especially in poor aquifers such
as silty sands, silts, and clays, than with the use of 50-mm diameter
wells. If dedicated electrical submersible pumps are required, readily
available and economical units can be utilized, rather than specialized
units constructed for the smaller diameter wells.

23.3.3 Screen length and placement

The screen length of a monitoring well should be governed by two
factors: the thickness of the stratum of interest, and the potential
vertical distribution of contaminants within that stratum. It is not
reasonable, in most cases, to expect any uniformity in this vertical
distribution, given the number of factors which influence the vertical
movement of contaminants. A dense, high total dissolved solids (TDS)
leachate plume may sink readily through an aquifer (Fig. 23.4a). In
contrast, dissolved phase plumes in high flow velocity aquifers may
stream along the upper surface for hundreds of meters with little
vertical mixing (23.4b). Inhomogeneities in hydraulic conductivity
caused by horizontal stratification could also restrict downward move-
ment of contaminants. As with well placement, some degree of predic-
tive capability may be necessary to properly place monitoring well
screens.
After assessing these factors, the monitoring system designer must
select a screen length which will minimize the vertical extent of
saturated soils which will be monitored. Many hydrogeologic settings
may be addressed by screening the entire extent of saturation within
the targeted stratum. Prime examples include:
• the targeted stratum is relatively thin «3 m);
• there is reason to suspect a generally uniform distribution of con-
taminants within the stratum; and
• the purpose of the well is to determine presence versus absence of
contaminants, rather than precise concentrations.
The down side of this method, and the fact which restricts its applica-
tion, is an averaging effect which occurs during sampling. The mixture
of water from uncontaminated portions of the aquifer with impacted
water will dilute constituent concentrations in samples from the
monitoring well. This effect will be enhanced if the more highly pro-
ductive (in terms of yield) portions of the targeted stratum are not
contaminated. The dilution in such cases could entirely mask low
levels of contamination.
One commonly utilized alternative to screening the entire targeted
stratum is the use of monitoring well clusters. These consist of two or
 Monitoring well design 617
SURFICIAL SOURCE
DOWN-GRADIENT
MONITORING WELL

',: .. ,', .

DENSE LEACHA
GROUND-W A TER
PLUME
FLOW DIRECTION
(SPECIFIC GRAVITY» 1)

SURFICIAL SOURCE

... . ..

DISSOLVED PHASE PLUM E"

GROUND-W A TER
FLOW DIRECTION
(HIGH TRANSMISSIVITY AQUIFER)

Fig. 23.4 Conceptual view of the potential movement: (a) of a dense, high-
TDS plume through the saturated zone; and (b) of a dissolved-phase
hydrocarbon plume in an aquifer of high ground-water flow velocity aquifer.
Note that in (a) the illustrated downgradient well would not detect the
presence of the plume (scenario based on an actual occurrence).

more wells at a single location, with screened intervals set at staggered

depths (Fig. 23.5). This permits the assessment of relatively discrete
portions of the targeted stratum during sampling. Although this type
of installation has been performed with multiple well assemblies within
a single borehole, the potential for a poor installation using this
method is high. Multiple boreholes, with single well assemblies within
each are currently more commonly utilized for cluster installation. The
618 Monitoring wells

INTERMEDIA TE
SHALLOW DEEP

------------- ------------

Fig. 23.5 Construction of an idealized monitoring well cluster.

primary disadvantages to the use of monitoring well clusters include

the following:
• increased installation, sampling, and analytical costs, due to the
increased number of wells and samples necessary; and
• for some locations within crowded facilities, it may be difficult to
place multiple wells within a single locale.
If the potential detection of free phase hydrocarbons is one purpose of
the monitoring well, this will also affect the placement of the well
screen. As discussed in section 23.2, free-phase hydrocarbons tend to
accumulate at the upper surface of the capillary fringe. This is true both
 Monitoring well design 619
for those hydrocarbons with specific gravities less than that of water
(referred to as light non-aqueous phase hydrocarbons (LNAPLs), or
floaters), and, in many cases, for those more dense than water (dense
non-aqueous phase hydrocarbons (DNAPLs), or sinkers). A screen
placement across the upper surface of the capillary fringe is required
for detection of these phases. For cases where DNAPLs have overcome
the surface tension effects resisting their entry to water-saturated soils,
placement of the screen across the base of the lowest portion of the
targeted stratum is appropriate.
Although both these approaches seem relatively straightforward
from a conceptual standpoint, the realities involved are not as co-
operative. For hydrocarbon phases above the capillary fringe, it is first
necessary to locate the upper surface of this fringe. This may not
be possible during drilling, especially if mud rotary techniques are
utilized, and so must often be estimated from water levels in nearby
wells. In addition, once the capillary fringe is located, it is subject to
temporal fluctuations related to seasonal water level changes. A well
functional for floating hydrocarbon detection in December may there-
fore be useless for this purpose in June. As with well placement, it is
necessary to have some understanding of site hydrologic conditions,
and the possible temporal changes in those conditions, before selecting
screen placements. As discussed, this may require the use of a phased
approach to monitoring well installation.

23.3.4 Screen type

Well screens have been manufactured using a number of methods and
opening styles, including torch- and saw-cut slots, louvered slots,
down-hole perforated. For monitoring applications, however, two
styles have emerged as predominant. The first is the mill-slotted,
which consists of a section of pipe with controlled-width slots cut
through the pipe wall. These slots are typically oriented perpendicular
to the long axis of the pipe, and are cut using banks of precision saws
in a milling operation. This type of screen, shown in Fig. 23.6a, is
commonly available in PVc, and is also manufactured in stainless
steel, polyethylene, PTFE, and other materials.
The second commonly used screen type is continuously-slotted (also
referred to as wire-wrapped). This construction consists of a wire with
a triangular cross-section, which is wound around a cylindrical base of
rods such that an open space of controlled width is maintained be-
tween windings (see Fig. 23.6b). As shown in the figure, the triangular
wire permits the entry without plugging of sediments too fine for the
screen to exclude. These screens are manufactured in stainless steel,
PVC, and fiberglass.
 . . . . .--0
a

RECTANGULAR
SCREEN WIRE

V-SHAPED
SCREEN WIRE

Fig. 23.6 (a) Mill-slotted well screen . (b) Continuously-slotted (wire-wrapped

screen). The cross-sectional detail of sediment grains passing through the
screen illustrates the effect of the triangular profile of the screen wires. Any
sediment fine enough to pass through the screen slot will enter the screen
freely without clogging. A screen constructed of rectangular profile wire, in
contrast, can be readily plugged by grains of elongate or irregular shape.
 Monitoring well design 621
Although more costly than the mill-slotted version, continuously-
slotted screens are often preferred due to their larger open area per
unit length, as well as their high tensile strength. Although the larger
surface area is typically considered to enhance well yield, it should be
considered that this will only occur when the screen open area is less
than that of the surrounding soils or the filter pack. Once the open area
of a well screen has reached that of these soils or the pack, the factor
limiting flow into the well is the open area of the soils or pack, and not
that of the screen. Most poorly graded (uniform) sand filter packs will
have open areas in the range of 10-15%. If this can be attained by a
mill-slotted screen, the use of a continuously-slotted type may not
improve well performance (Clark, 1988). Additionally, if the purpose of
the well does not require maximum well yield, mill-slotted screens may
be adequate.
The slot width of the screen must be based upon the grain size
characteristics of the targeted strata alone, for naturally packed wells,
and upon the characteristics of both the stratum and the filter pack, for
artificially filter packed wells. These concepts, and their bearing on
screen slot size, will be discussed in more detail in section 23.3.6.

23.3.5 Sediment sumps

Many designers are now incorporating a sediment sump, also called a
silt trap or sediment trap, into monitoring well constructions. This
feature is a modification of a common water well practice, and consists
simply of a blank (unperforated) section of pipe placed below the base
of the screen (Fig. 23.7). Its purpose is to provide a catchtrap for fine
sand and silt which bypasses the filter pack and screen and settles out
within the well, a phenomenon known as 'silting up'. This sediment
collects within the sump rather than within the screen, and there-
fore does not reduce the functional screened length of the well. This
minimizes the need for periodic cleanouts of the screen to maintain the
well's effectiveness.
Sumps may be a useful addition when it is anticipated that a large
volume of fine sediment will be entering the well. This would include
wells set in silt or clay strata, or installations for which difficulty in
properly emplacing an artificial filter pack is anticipated. Gass (1988)
recommends the use of a sump in fine-grained aquifers as a means of
avoiding the great difficulty in producing low turbidity water from
such low-producing strata.
In general, however, sumps are unnecessary except for situations
where very turbid waters may be expected for the life of the well. If a
filter pack can be successfully emplaced, therefore, a sump may be
superfluous. Since a sump adds length to the well string below the
622 Monitoring wells

==f.;:i;~~r-- WELL SCREEN

BOREHOLE

SEDIMENT SUMP
FILTER PACK

Fig. 23.7 Construction of a sediment sump.

base of the screen, it is necessary to add additional length to the

borehole (or 'overdrill') in order to place the screen at the desired
interval. For at least one commonly encountered hydrogeologic scenario,
where the screen is placed at the base of an aquifer underlain by a thin
confining layer, overdrilling for sump placement can potentially breach
the lower confining layer.
If a sump is required, lengths of 0.6-1.6 m are adequate for most
applications. Since most monitoring wells are pumped no more than
four or five times per year, this length should be sufficient for several
years' accumulation of sediment. Some wells which have longer
screens (typically greater than 6 m) and more frequent usage may
require 3-m sump lengths. Sumps on PVC wells may be field-fabricated
using a piece of blank pipe sealed at the lower end with a slip-fit cap,
which is secured with rivets or screws. Steel sumps are typically sealed
at the lower end by a circular plate welded in place. Alternatively,
sumps may now be obtained pre-fabricated from many monitoring well
materials suppliers.
If a dedicated electric submersible pump is to be placed in a well
equipped with a sediment sump, it is best to avoid placing the pump
within the sump. This precludes the flow of water around the motor
on the base of the pump, which results in insufficient motor cooling
and shortened pump life.
 Monitoring well design 623
23.3.6 Filter packs
Most of the discussion within this section will be pertinent to what are
termed artificial filter packs; that is, filter packs comprised of material
introduced into the borehole. These comprise the majority of filter
packs utilized in monitoring wells today. An alternative worthy of
mention, but which is often forgotten or neglected, is the natural filter
pack. For this approach, the hydrogeologic unit being screened is
permitted to collapse around the well screen. This type of filter pack is
not appropriate when this targeted stratum consists largely of fine sand
or smaller grain sizes. When this stratum is comprised of sand and
gravel, however, with little or no fines to enter the screen and create
turbidity, there are distinct advantages to the use of a natural filter
pack. First, it eliminates one material which must be introduced into
the well, which helps to minimize the impact of the well upon the
subsurface environment, as well as the cost and effort of well installa-
tion. Second, for mud rotary installations, the collapse of the borehole
during natural packing will disrupt the sheath of drilling mud which
forms during drilling on the borehole's inner surface. This simplifies
and enhances well development.
Although natural filter packs are appropriate in certain circum-
stances, they do not currently find wide application. The following
discussions of filter pack selection and installation will therefore reflect
practices related to artificial filter packs.

Gradation
Many monitoring wells are installed in strata which would not normally
be considered aquifers; that is, in sand units which have high fines
contents and/or abundant layers of fine sediments, or in silt or clay
strata. This has become especially prevalent in the wake of regulatory
guidelines or requirements which mandate monitoring of the upper-
most water-bearing zones beneath the site of investigation. In such
cases, a filter pack (less correctly referred to as a sand or gravel pack) is
the preferred method of controlling the entry of fine sand, silt, and/or
clay into the well screen. As discussed in section 23.1.2, such a filter
pack is constructed by backfilling the annular space surrounding the
screen with a granular, relatively inert material. In most cases, this
consists of a clean, silica sand.
Filter pack selection techniques developed for water supply well
applications are often suggested for monitoring well design. The dif-
ference in purpose for which these two types of wells are intended
makes this practice inappropriate. For water supply wells, the filter
pack must be carefully selected and emplaced to minimize resistance to
water flow into the well. This may entail the analysis of grain size
624 Monitoring wells

trends within the targeted stratum during the drilling process, and the
selection of a variety of sand sizes for the filter pack. These filter pack
sands are then emplaced to match the layering or other grain size
variations present within the stratum. This technique was developed to
maximize well efficiency and yield in the highly productive sands in
which many water supply wells are installed.
For monitoring wells, in contrast, yield is not the primary considera-
tion in well performance. The function of the monitoring well filter
pack, therefore, is not identical to that of the water supply well filter
pack. Monitoring well filter packs are intended primarily to:
• control the entry of fines into the well, so as to minimize the turbidity
of ground-water samples; and
• provide structural support for the borehole in which the well is
installed, and for the overlying annular sealing materials.
Neither of these purposes requires perfect compatibility between the
grain-size gradation of the filter pack and that of the aquifer.
Despite this, monitoring well specifications or regulations often
require the selection of a filter pack based upon detailed grain size
analyses of the aquifer. As stated, the purpose of monitoring well filter
packs makes such a degree of compatibility unnecessary. In addition, it
is not practicable to attain this degree of compatibility, given the nature
of the typical monitoring well installation process. Most monitoring
well installations are performed as rapid, short-term operations, with
materials purchased in advance and ready at the site. This expedites
the installation of multiple well systems. The use of water supply
well filter pack selection techniques would require a more deliberate
approach, with filter pack material purchase delayed until the comple-
tion of the monitoring well borehole. This approach would dramatically
increase both the cost and difficulty of well installation, without any
significant resulting benefit.
It is more efficient in terms of material and installation practice to
select a material which broadly suits the general characteristics of the
aquifer, and is compatible with the screen slot selection. Several guide-
lines for this selection process are present in the literature, e.g. Driscoll
(1986), and Gass (1988). In contrast to the well-graded filter packs
used for water wells, these publications recommend the selection of a
relatively uniform monitoring well filter pack, with the uniformity
coefficient (C u ) for the filter pack ranging from 1 to 3. The uniformity
coefficient, Cur is equivalent to the ratio of the 60% finer grain size
(0 60 ) of the material to the 10% finer grain size (OlD)' Driscoll
recommends selection of a filter pack material with a 50% finer (0 50 )
two times larger than that of the finest portion of surrounding soils.
This produces a uniform filter pack only slightly coarser than the finest
portion of the surrounding native material. Gass prefers a greater
 Monitoring well design 625
contrast between the natural soils and the filter pack, recommending
that the D30 of the filter pack be 3-4 times larger than that of the finest
portion of the surrounding soils. The slot size of the well screen should
then be selected such that no more than 10% of the filter pack material
can pass through the slots.
In addition to this portion of the filter pack, some designers
recommend the placement of approximately 0.3 m of much finer sand
(e.g., 'flour sand') at the upper portion of the pack. This secondary
filter pack is intended to supplement the bentonite seal (this seal is
discussed in detail in section 23.3.7) in precluding the entry of grout-
related fluids into the screened interval of the well.

Filter pack materials and nomenclature

Although 'filter sands' are commercially available, ordinary silica blast
sand is generally suitable for monitoring well filter pack construction,
provided it is selected with some discrimination. Quality of blast sand
may vary considerably from supplier to supplier, but is generally con-
sistent for any given supplier through time. As with other monitoring
well materials, then, it is best, when possible, to rely on a familiar
supplier who has met specifications consistently in the past.
The main variable between brands is the mineral composition of the
sand, since this is naturally dependent on the deposit from which the
sand is quarried. It is best to identify one or two brands within a region
which are composed of at least 90% quartz grains. This will assure the
well installer of a filter pack composition in excess of 95% Si0 2 . Such
sands contain a minimum of the more soluble metals and other cations
present in feldspars and accessory minerals such as micas, magnetite,
and hornblende.
In addition to a minimum of potentially unstable minerals, it is
important to have confidence in the cleanliness of the filter pack
materials prior to their installation. Most sands are water-washed and
thoroughly dried prior to sale, minimizing the potential for importing
chemical contaminants. If there are specific reasons to suspect the filter
pack materials of impurity, it may be advisable to collect a sample of
the sand for leachate analysis. In most instances, however, this is
probably an excess of caution.
Sands are normally given a product number or name by the supplier
or manufacturer based upon their mean grain size and grading (sorting).
There is no consistency in the product numbering systems utilized by
different producers - they are specific to the company producing the
sand. These numbers are, therefore, only minimally descriptive, and
their usage should be avoided in well specifications and construction
details. Instead, sands should be described on the basis of a mesh
number, or sieve number range. This is a number which delineates the
626 Monitoring wells

Table 23.2 Opening sizes for US sieves

US standard opening size opening size

sieve numbers (mm) (microns)

4 4.699 4699
6 3.327 3327
8 2.362 2362
10 2.000 2000
12 1.651 1651
14 1.400 1400
16 1.180 1180
18 1.000 1000
20 0.850 850
25 0.710 710
30 0.600 600
35 0.500 500
40 0.425 425
50 0.295 295
60 0.250 250
70 0.210 210
80 0.175 175
100 0.147 147
140 0.105 105
200 0.074 74

two US sieve numbers which bound the upper and lower grain size
ranges of the blast sand. For example, describing a sand as a 16-40
Mesh indicates that approximately 95 weight percent of that sand falls
between US sieve numbers 16 and 40 (see Table 23.2 for a listing of
opening sizes for US sieves). These mesh numbers may be obtained by
contacting the sand manufacturer and requesting details on the grada-
tion of the subject sand.

Filter pack emplacement

Two methods are commonly used for filter pack emplacement. The
most commonly used is to pour the sand into the annular space of the
borehole at the surface, allowing it to settle into the space surrounding
the screen. The sand is properly poured both slowly and steadily, with
frequent soundings of the sand level with a weighted measure. These
measures will minimize the possibility of bridging, which is the term
used for sand and other materials lodging in the annular space above
the depth intended (Fig. 23.8). Bridging can seriously impair proper
filter pack emplacement, even to the extent of compromising the well.
This method may be effectively utilized in conjunction with hollow
stem auger drilling techniques. This is accomplished by pouring the
sand into the stem of the augers as the augers are removed from the
 Monitoring well design 627

," ,
..'
BRIDGED FILTER - - - - - .....:
PACK SAND

I~--I---- CENTRALIZER

- f - - - - WELL SCREEN

Fig. 23.8 An illustration of bridging, which occurs when a material such as

sand or bentonite pellets lodge in the borehole above the depth for which they
were targeted. If the bridge cannot be dislodged, the monitoring well
construction may have to be aborted.

borehole in increments of less than 300cm. This type of installation is

very useful in situations when auger removal would result in excessive
borehole collapse. Care must be utilized, however, to avoid accumula-
tions of sand within the auger stem of more than approximately 300 cm,
since this can wedge the well riser or screen inside the stem. In
addition, the relatively small diameter of the auger stem reduces the
working area available for sand passage, which in turn enhances the
possibility of bridging.
To use this method for installations within mud rotary drilled
boreholes, it is usually necessary to thin the viscous drilling mud
within the borehole by mixing potable water. This is performed in
order to induce the filter pack materials to settle through the annulus at
a rate which is not prohibitively slow. Filter pack emplacement should
commence as quickly as possible after this thinning is performed, how-
ever, since the thinning of the mud enhances the possibility of bore-
hole collapse. The volume of water used for the thinning procedure
628 Monitoring wells
should be kept to a minimum, to minimize both disturbance of the
borehole ('wash outs'), and fluid losses to the surrounding soils which
will extend development time.
The second option for the installation of filter pack materials in a
mud rotary drilled borehole is emplacement through tremie pipe. For
this method, a string of pipe is lowered into the annulus to the desired
filter pack depth interval. The drilling fluid in the borehole is thinned
using potable water until a slurry of filter pack sand and water can be
induced to flow down the tremie pipe. As this slurry of sand and water
exits the lower end of the tremie pipe, the sand settles out into the
bottom of the borehole, forming the filter pack. This method is often
regarded as necessary to counteract the tendency of the silica sands to
form 'fining upward' sequences in the filter pack as they settle through
the long column of fluid in the borehole.
For a properly selected monitoring well filter pack material, this
method is rarely necessary. As mentioned earlier in this section, filter
packs for monitoring wells should be uniform in gradation, that is,
they should consist of a narrow range of grain sizes. This uniformity
will minimize the potential impact of any fining upward sequences
formed during filter pack emplacement. Further, given the purposes
for a monitoring well filter pack - precluding the entry of fines and
supporting the borehole and annular seal - any fining upward
sequences which do occur will not impair its function. Since the tremie
method results in the introduction of additional water to the borehole,
which must then be recovered during development, and also requires
additional time and effort to perform during installation, its use is not
recommended.

Other options
In addition to the emplacement of granular filter packs into the annulus
of the monitoring well, there are filtering mechanisms which may be
installed as a part of the well assembly. These include the use of:
• well screens which have been wrapped in filter fabric; and
• prepacks, which are well screens which are manufactured of two
nested screened pipes, with a silica sand filter pack between the
screens. This entire assembly is installed as a single piece.
Several problems exist with the use of filter fabrics. They tend to clog
easily, and as a result can severely restrict the entry of ground water
into the well. In addition, many fabrics may restrict the passage of
separate-phase hydrocarbons.
Pre packs are typically a more effective approach, but also have short-
comings. Because they are constructed of two nested screens, the outer
diameter (00) of the prepack is larger than that of a single screen
 Monitoring well design 629
of identical inner diameter (10). This larger 00 may complicate
installations through an auger stem, or in other restricted settings. The
thickness of the sand in a prepack is much lower than that within an
annular filter pack, and may be less effective in the removal of some
types of turbidity. The cost of prepacks is typically higher than that
of a single screen and filter pack material purchased separately. If
difficulties in filter pack installation are anticipated, however, pre packs
are an alternative worth evaluating.

23.3.7 Annular seals

After the filter pack has been emplaced, it is necessary to fill the
remaining annular space with a material of low hydraulic conductivity
to serve as an annular seal. This seal is typically emplaced as a liquid,
referred to as grout, which is usually comprised of either Portland
cement, a slurry of bentonite clay, or Portland cement with an admix-
ture of bentonite. The purpose of the annular seal is to prevent the
exchange of fluids between the screened interval and higher strata
and/or the surface. Without this precaution, cross contamination
between aquifers could occur, limiting the possibility of obtaining
a representative sample from the stratum or interval of interest.
An improperly sealed well can actually enhance the spread of con-
taminants, by providing ground water from a contaminated aquifer
access to less impacted strata.
The annular seal should be continuous with a seal at the surface,
which is often referred to as a sanitary seal or surface seal. The surface
seal should be designed to prevent the pooling of surface waters at
the wellhead (the surface expression of the well). It should, more
importantly, preclude the entry of these surface waters into the well
and borehole. In addition, it should be durable enough to withstand
stresses associated with weather and incidental impacts. Most surface
seals are constructed of concrete.

Bentonite seals
To prevent the grout and grout-derived fluids from entering the filter
pack of the monitoring well, it is standard practice to place a seal of
bentonite directly upon the upper surface of the filter pack. Since
cement-based grouts are a strong source of hydroxyl ions and free
lime, this helps to minimize the impact of annular grouts on the
ground-water chemistry of the targeted stratum. This seal should be
emplaced as a solid, rather than a slurry, and should have a minimum
thickness of 0.5-2 m.
Where stratigraphy permits, this seal may be placed to extend up
into the nearest overlying aquitard. This helps to prevent the migration
630 Monitoring wells

'0","".
TARGETED
AOUIFER • .•

IotIGRATIOIII OF HIGH pH FLUIDS .

FROM GROUT IIIITO WELL
, .

Fig. 23.9 Bentonite seal placement. The seal placement on the left permits
high-pH fluids from the Portland cement grout to bypass the seal through the
aquifer, and enter the well. By extending the seal upward another meter, right,
the grout fluids may be better isolated from the well screen.

of high pH fluids from the grout around the seal via the aquifer
soils (Fig. 23.9). This design modification is recommended in low
hydraulic conductivity aquifers, or when 50-mm diameter wells are
being installed. Under these two circumstances, it is often difficult to
flush enough water through the filter pack and well to remove any
grout-derived caustic fluids. An extension of this technique sometimes
used when installing shallow wells is to entirely eliminate the use of
liquid grouts, and extend the solid bentonite seal from the filter pack to
the surface.
The most commonly utilized form of bentonite used for these seals is
bentonite pellets. These are highly compressed spheres of finely-ground
bentonite clay (a high-grade sodium or calcium montmorillonite),
which are supplied in several sizes up to 12-mm diameter. Pellets of
'pure' bentonite, without admixtures, should be utilized. Upon contact
with water, these hard dry pellets hydrate very rapidly, swelling and
becoming cohesive. Within a few minutes, they can form a water-
resistant seal. While this rapid hydration and cohesive character
makes them very effective as a sealant material, it can also render their
emplacement more difficult. After entering ground water or drilling
fluid standing in a borehole, the peBets have a tendency to adhere to
any obstacle they encounter, as well as to clump together. As a result,
pellets tend to bridge (Fig. 23.8) at any constriction within the annular
space.
 Monitoring well design 631
This effect is quite important when design specifications call for
emplacement of pellets through the stem of hollow stem augers. When
the diameter of the well is relatively close to that of the auger stem,
attempts to emplace pellets through the auger stem have a low success
rate. The working area between the well and the inside of the stem in
these cases is simply not sufficient to prevent the bridging of the sticky
wet pellets. When this occurs, the well and/or augers must usually be
removed to eliminate the bridge, and the well installation procedure
returns to its starting point.
This effect can be directly addressed through the use of larger stem
diameter augers, since this increases the open area available for pellet
passage. One example would be the use of 165-mm stem diameter
augers for the installation of a 50-mm well, rather than the 90-mm stem
diameter augers which are more commonly used for this purpose. This
practice will, incidentally, also decrease the risk of bridging during
filter pack emplacement.
Another possible solution is the use of 'timed release' pellets, which
have a slow-dissolving coating to delay the initial hydration period.
This is not a complete solution to the problem, however, since the
coating only reduces, and does not eliminate, the pellets' tendency to
bridge within the augers. If this type of pellet is used, the manufac-
turer should be contacted to provide a chemical characterization of the
coating, in order to ensure the absence of interferences with suspected
ground-water constituents.
The question of hydration time for pellets is still a point of conten-
tion. While field testing of pellets (not the 'timed release' type) by
the authors has indicated the formation of satisfactory seal within
only a few moments of contact with water, many specifications for the
installation of monitoring wells mandate hydration periods of one to
several hours. While this period may be required for complete pellet
hydration, such complete hydration is not necessary to produce a
coherent seal. Once the points of contact between pellets have expanded
sufficiently to eliminate the interconnection of pores between the
pellets, the seal is sufficient to preclude the passage of grout-related
fluids.
In addition to pellets, bentonite chips also find frequent application
in bentonite seals. These chips are simply gravel-sized fragments of
unground and untreated bentonite, which are graded based upon their
size. Bentonite chips are compositionally identical to bentonite pellets,
and are typically obtainable at approximately 25% of pellet cost. Since
the chips have not been ground to a powder, they hydrate more slowly
than pellets, reducing the likelihood of bridging in problem installa-
tions (such as through auger stems). The downside of this slower
hydration and sealing is the greater time required between the emplace-
ment of the chips and that of the grout. An absolute minimum of 30
632 Monitoring wells

minutes should be allowed for seal formation; even longer periods are
preferable. This is in contrast to the 10-15 minutes required for pellet
hydration.

Grouts
A grout typically consists of a slurry of cement and/or clays in water,
although other materials are used for special applications. The require-
ments of a grout material usually include its hydration and solidification
following emplacement, and the formation of a relatively permanent,
low hydraulic conductivity seal.

Portland cement-based grouts

Neat cement grouts consist of Portland cement in water, with no
admixtures of bentonite or other additives. Admixtures of sand or pea
gravel to these cement grouts result in formation of a type of concrete,
adding some structural strength to the grout, but having little effect on
temperature, hydration time, or shrinkage.
More commonly used in the monitoring well installation practice
are cement grouts with admixtures of 2-5% powdered bentonite by
weight. This weight percentage is expressed as
weight dry bentonite
Percent bentonite = . h d x 100%
welg t ry cement
Bentonite admixtures slow the development of compressive strength in
the grout, and reduce the final strength of the solid grout (Smith, 1976;
Dowell Schlumberger, 1984). Additional effects include an increase in
viscosity, a decrease in thickening time and grout density, and reduc-
tions in fluid losses from the emplaced grout. The higher viscosity
results in a better displacement of drilling mud from the well's annulus,
resulting in a more complete final seal. The reduction in fluid losses
may help reduce shrinkage of the grout after emplacement (Driscoll,
1986; Lehr et al., 1988), as well as minimizing the loss of constituents
which might affect ground-water chemistry. The reduction in shrinkage
occurs only for bentonite mixtures up to 6%. Above this percentage,
shrinkage may actually be exacerbated (Driscoll, 1986).
The mixture of bentonite with Portland cement should begin with
the hydration of the bentonite alone, without the Portland cement.
This permits relatively complete hydration of the bentonite. This
hydration is not as complete in the presence of Portland cement,
because it is inhibited to some degree by Ca++ ions released into the
slurry by the cement (Dowell Schlumberger, 1984). After the bentonite
is thoroughly mixed, the addition of Portland cement may proceed.
Most monitoring well specifications call for the mixture of 22-28
liters of water per 43-kg sack of Portland cement, with approximately 4
 Monitoring well design 633
liters per sack additional for each percent bentonite added to the grout.
These specifications are typically based on the standard practices of
specialty cementing companies, such as oil well cementing opera-
tions. While these water-to-cement ratios are feasible using the mixing
equipment available to this type of firm, the simpler, smaller scale
equipment utilized by most monitoring well installers is typically not
capable of meeting this mix specification. The grout simply becomes
too viscous to pump during the mixing process. In deference to the
realities of field installation, then, the designer should be prepared to
consider a less stringent water-to-cement ratio.

Bentonite grouts
The use of bentonite grouts for monitoring well installation has gained
some degree of acceptance during the past decade. Such grouts, which
consist of high percentages of bentonite in a water slurry, do offer
some distinct advantages. The thixotropic nature of bentonite-water
slurries permits the pumping of the grouts for extended periods after
mixing, which reduces the potential penalty of delays in grouting. The
bentonite grouts generate far fewer calcium and hydroxyl ions than
Portland cement, reducing the potential for high pH problems from
grout leachates. In addition, the potential for problems with PVC pipe
softening due to Portland cement heat of hydration is eliminated with
bentonite grouts. Finally, the field mixing of these grouts is typically
faster and easier than that of the Portland cement-based variety.
These grouts are not appropriate for many installations, however.
The plastic and thixotropic nature of bentonite which eases and
extends grout pumping also eliminates most of the mechanical strength
of the bentonite grouts. They cannot, therefore, be used in well
designs requiring a strong, non-plastic solid grout column. Since the
water is less strongly bound in the chemical structure of bentonite
slurries than in that of Portland cements, fluid loss and attendant
shrinkage is always a possibility. This effect is probably of concern
mainly in arid-zone settings, especially in strong capillarity soils such
as clays and silts.
If bentonite grouts are utilized, the design specifications should
delineate the permissibility of additives. Many of these grouts utilize
organic polymers to improve pump ability, or metal oxides to improve
yield. If these constituents could result in interferences with monitored
constituents, their use should be barred in the well installation
specifications.

Grout emplacement
Grouts are typically emplaced through a string of pipe lowered into the
annular space, referred to as a tremie pipe. The tremie, which is
typically 25- to 50-mm diameter, is suspended in the annular space
634 Monitoring wells

" '+ - - + - -I - --TREMIE PIPE

RISER

GROUT DISPLACING DRILLING

FLUIDS UP THE ANNULUS

INTERVAL OF POOR
GROUT INTEGRITY ----l~~ UNDISPLACED DRILLING FLUIDS
DUE TO INCOMPLETE
DISPLACEMENT OF
DRILLING FLUIDS

1'·:':":*- - BENTONITE SEAL

Fig.23.10 Conceptual view of the formation of a zone of possible low annular

seal integrity as a result of incomplete displacement of annular fluids by a
pumped grout.

with its lower end approximately I-2m above the top of annular
bentonite seal. The 1-2 m gap minimizes the possibility that the grout
stream will wash out and displace part of the bentonite seal. Grout is
pumped into the tremie at the surface, to discharge into lower portion
of the annulus. In theory, the grout entirely fills the lower portion of
the annular space, moving upward and displacing drilling fluid or
natural waters in a plug.
While this method is certainly capable of providing a generally
adequate seal, its potential short-comings should also be recognized.
Cementing research in oil field drilling and well installation has long
recognized the difficulty in emplacing a complete cement seal within a
fluid-filled borehole. The main problem arises from the incomplete
displacement of the fluid in the annulus by the incoming grout. If this
fluid has a viscosity substantially greater than that of water, such as is
the case for most drilling fluids, or for a mixture of auger cuttings and
ground water, grouts will tend to stream or channel rather than plug
flow. This effect would be strongest nearest the point of discharge from
the tremie, so the grout seal would tend to be least complete near its
base (Fig. 23.10).
While the scope of this problem has never been determined for
monitoring wells, the implication is clear: the lower portion of an
annular seal should not be relied upon as a perfect barrier to fluid
migration. No solution to the potential problem is presently available,
given the small scale of most monitoring well installations. It may be
 Drilling methods 635
possible to reduce this channeling effect by using as large a tremie pipe
diameter and as high a pumping rate as possible. The best approach,
however, is to emplace the annular bentonite seal properly, so that
minor gaps in the grout seal will not impair the function of the well.
Monitoring well installations often require the installation of grout
into dry annular volumes, i.e. those which contain no standing liquids,
such as ground water or drilling mud. In these situations, there is no
disadvantage to introducing the grout into the annulus at the surface,
without the use of tremie. Being liquid, the grout will effectively fill the
annular volume.

23.4 DRILLING METHODS

Several drilling methods have been used for the installation of monitor-
ing wells. These methods were discussed in detail in Chapter 15 and
are reviewed here in terms of monitoring well installation. The most
common methods of well installation are air rotary, direct mud rotary,
reverse mud rotary, hollow stem augers, solid stem augers, and
cable tool. The drilling method selected for well installation at a par-
ticular site must be suitable for the subsurface conditions likely to be
encountered. In addition, it must be amenable to the particulars of well
installation included in the design. One of the best methods of evaluat-
ing these factors is to contact consultants, drilling companies, and
other agencies with experience in drilling within the subject area.
These persons can often indicate which methods have historically been
utilized with the greatest rate of successful installations.
In the following discussion, the focus will be placed upon the two
methods most commonly used for monitoring well installations: hollow
stem auger and direct mud rotary.

23.4.1 Hollow stem auger drilling

This method has attained a favored status in environmental drilling,
especially among most of the regulatory agencies involved in ground-
water monitoring, for several reasons. First, the absence of drilling
fluids minimizes the potential impact of the drilling operation on the
aquifer hydrogeochemistry. Second, the hollow stem of the auger
typically permits convenient sampling as the borehole is advanced.
The use of soil coring barrels, also called continuous samplers, even
permits the collection of a relatively continuous soil sample as the
augers are advanced. The hollow stem also permits the augers to serve
as a temporary casing during well installation. In poorly consolidated
formations which would collapse if unsupported, the stem may
636 Monitoring wells
therefore permit the installation of the well pipe, the filter pack, and
the annular seal before the augers are withdrawn.
Sampling using hollow stem augers may be complicated by the
presence in the subsurface of saturated, unconsolidated sands (also
known as running, flowing, or heaving sands). These sands typically
enter and fill the lower part of the auger stem, often to an extent of 1 m
or more. Once within the stem, these sands can be difficult to remove,
and their presence can preclude both sampling and well installation.
If these sands are known to be a subsurface condition, the use of
alternative drilling methods may be preferable.
Although advances in drilling rig torque capacities and drilling tech-
nology are extending the range of depths and conditions within auger
capabilities, auger drilling methods are still not appropriate for every
setting. Case histories of augers being advanced to lOO-m depths,
for example, are not unknown, but the number of practitioners and
rigs capable of performing these feats are still severely limited. As
suggested in the introduction to this chapter, it is critical to recognize
the difference between practices which are possible, and those with a
high probability of success. Augers, as with other methods, should be
utilized for situations well within their capabilities in order to minimize
the risk of a failed or poor installation.

23.4.2 Direct mud rotary

Well installation is conceptually straightforward within boreholes
advanced using mud rotary methods. When the total depth of the
borehole is attained, the mud density is adjusted to ensure that the
borehole will remain stable upon removal of the bit. The drill string is
then removed, and the riser and screen of the well assembly lowered
into the hole to the desired depth. Depending on the density of the
drilling fluid in the borehole and the material used for well construc-
tion, the well assembly may have to be filled with clean water to
counteract its buoyancy, and permit it to sink through the drilling
mud. This is typically necessary when non-metallic well materials,
such as PVC or PTFE, are utilized. The filter pack and annular seal are
then introduced into the annulus of the borehole between the well and
the borehole.
As with hollow stem augers, there are real, pragmatic limitations to
the capabilities of this drilling method. Specifications which fail to
recognize these limitations decrease the probability of successfully
installing a well.
One example of a commonly applied design specification which
may not recognize the physical limitations of the drilling method is
the elimination of additives to the drilling mud. These specifications
typically permit only potable water, with no bentonite, polymers, etc.
 Drilling methods 637

admixtures. The motivation is to remove any possible interferences

between these additives and the suspected ground-water contaminants,
as well as other impacts to the aquifer and ground-water character
which are attendant to the use of these additives. In addition, the
formation of a bentonite 'mud cake' on the borehole wall is hypo-
thetically eliminated.
While the approach seems logical, its value decreases if it limits the
capabilities of the drilling method so severely that drilling or well
installation is compromised. Many aquifer materials, such as coarse
sands or gravels, may not be drillable using neat water muds, since
the lifting capability of the fluid returning up the hole is strongly
dependent upon its viscosity. In addition, borehole stability in un-
consolidated materials is typically tenuous in boreholes drilled in
this fashion. Finally, one function of these additives is to minimize
fluid losses from the borehole into the targeted stratum. Without this
effect, the volume of potable water introduced into the subsurface, and
the dilution effects which accompany this, may increase dramatically.
One side effect of these fluid losses is the formation of an even thicker
mud cake, consisting of natural clays, than might have been formed if
higher quality clays such as bentonites had been used.
As with other methods, the well designer is forced to weigh both
sides of the issue. In this case, this would involve balancing the
possibility of chemical interference or hydraulic conductivity reduction
from the bentonite against the possibility of a borehole which has
collapsed around the well string, rendering proper filter packing and
sealing impossible, or of large fluid losses into the aquifer.

23.4.3 Direct mud rotary vs. hollow stem augers

The relative advantages of each method for monitoring well installation
are summarized and compared in Table 23.3. Briefly, hollow stem
augers are favored because of their minimal impact on the aquifer and
natural waters. Shallow soil sampling and well installation are rela-
tively easy using this method, especially in the absence of flowing
sands. The absence of introduced fluids permits a relatively accurate
assessment of the depth to water.
Mud rotary is usable in a wider variety of geologic settings and
depth ranges than augers. In addition, well installation is often much
simpler, since the installer has the entire borehole for placement of
annular materials, rather than just the auger stem. Mud rotary does
require the introduction of some foriegn liquids into the aquifer, which
may be undesirable for many investigations. Some types of geophysical
logging can assist in the determination of depth to water, but the
degree of certainty is much lower than that possible with auger
drilling.
Table 23.3 Relative advantages of mud rotary vs. hollow stem auger drilling
methods for monitoring well installation

criteria hollow stem auger direct mud rotary

-----------

Materials which Unconsolidated to poorly Unconsolidated to

can be drilled consolidated materials consolidated materials
Drilling depth Generally limited to less No strict depth limitations
capabilities than SOm
Soil sampling Rapid and simple through Slower for most sysems
the auger stem than with augers
Possible to continuously Typically requires bit
core while drilling removal to collect sample
Ground water level Can be measured during Cannot be readily measured
measurement drilling until well installed
Ground water Can be sampled while Cannot be readily measured
sampling drilling until well installed
Geophysical Certain logs (e.g. natural Almost all logs can be run
logging gamma) can be run in borehole
through the auger stem
Well installation Can be performed through Difficult in collapsible
the augers for collapsible formations, but borehole
formations more accessible than for
installations through
augers
Rapidity, mobility, Rapid setup and Setup and teardown
access, footprint teardown, good access generally slower, access
to remote locations, not as good, and footprint
generally smaller larger than for
footprint than comparably-sized auger
comparably-sized mud rigs
rotary rigs
Aquifer Minimization of fluid Fluid interactions result in
interferences addition minimizes greater impacts than
impact to aquifer augering
geochemistry If managed, impact is
typically still within
acceptable limits
Development time Typically short, less than Short to long, dependent
for direct mud rotary upon fluid losses,
aquifer productivity, etc.
Cost per unit Typically low, less than for Variable, dependent upon
drilled direct mud rotary, when site specific conditions
applied appropriately Generally higher than auger
costs in comparable
settings
 Monitoring well development 639
Table 23.3 Continued

criteria hollow stem auger direct mud rotary

---~~~--~ -~---~~~----~---

Equipment Typically easy and rapid, Typically moderate level of

decontamination less difficult than for effort, more complex than
direct mud rotary for auger drilling
Cuttings control Cuttings mostly solids, Cuttings mixture of liquids
control straightforward and solids, control more
complex

Note: Comparisons are based on typical equipment and conditions. Individual site and
drilling conditions may modify the favorability of one method versus another based on
any of the above criteria.

23.5 MONITORING WELL DEVELOPMENT

As mentioned in this chapter's introduction, the drilling and well

installation process is unavoidably traumatic from the perspective of
the aquifer. The shearing and churning of soils, the suspension of
natural and introduced clays and silts, and the introduction of non-
aquifer fluids, all combine to damage the natural structure of the soils
in the vicinity of the well. This damage is often, to some degree,
irreparable. While the use of sound drilling procedures and reasonable
caution can minimize these effects, well installation should always
include post-installation efforts to repair the subsurface impacts of
drilling. These efforts are termed well development.

23.5.1 Primary purposes of monitoring well development

Drilling fluid removal

During mud rotary installations, the borehole is kept full of a water-
based mud to a level at or above surrounding grade. As a result, the
head within the borehole is almost always greater than that within the
subsurface soils, resulting in drilling fluid movement into the aquifer.
These introduced fluids alter the ground-water chemistry surrounding
the borehole by displacing and mixing with natural waters. This effect
is also present, although to a lesser degree, when water must be added
during an installation using hollow-stem augers. This is sometimes
necessary to counteract the effect of flowing sands (section 23.4.1) or to
dislodge an annular bridge (section 23.3.6).
One purpose of development in such cases is to remove the intro-
duced water from the aquifer, so that subsequent sampling is more
representative of the natural waters than of the drilling fluids. A logical
starting point for this effort is to calculate the volume of fluid lost to the
640 Monitoring wells

subsurface during drilling. This is done by maintaining a fluid balance

during the drilling process. This balance tracks the volume of fluid
used to advance the borehole and install the well, less any storage
within the borehole and at the surface. In most cases, only losses to the
targeted stratum are of interest, and not those which occur within
overlying strata. This may not be true if the overlying strata are moni-
tored by other monitoring wells at the same location (i.e., shallower
wells within a monitoring well cluster).
The volume of fluid lost to the targeted stratum is the minimum
volume of fluid which typically should be removed during develop-
ment. Dilution of this lost fluid within natural waters, however, usually
requires the removal of a larger volume to properly address the impact
of the lost fluid. This volume may be several times that of actual fluid
loss, and is dependent upon a range of site-specific characteristics.
Pragmatically, then, there is no way to accurately predict how much
water the development process must remove to compensate for these
fluid losses. This uncertainty is usually addressed by the field measure-
ment of some basic physical and chemical characteristics of the waters
removed during development. Drilling fluids tend to have elevated pH
(typically 9 or above) and specific conductances. These parameters, as
well as temperature, should be measured at intervals during develop-
ment. Development is typically accepted as sufficient when these
parameters stabilize to less than 10% variance for several measurements.

Mud cake removal

Drilling fluids are utilized to reduce fluid loss from a borehole by
forming a thin, low hydraulic conductivity coating, typically called a
mud or filter cake, on the borehole surface. This cake remains com-
pletely or partially intact following the well installation process. Once
installation is complete, such a cake forms a low hydraulic conductivity
sheath around the filter pack, and so may dramatically reduce the yield
of the well.
Hollow-stem auger installations are not necessarily immune to the
deleterious effects of a mud cake. If sufficient clays are encountered
during the drilling process, these cohesive soils may be smeared into
the equivalent of a mud cake by the rotation and downward motion of
the augers (Fig. 23.11).
Although yield is typically not a primary consideration in monitoring
well installation, large losses in well efficiency may impair its useful-
ness. If a strongly affected well is to be used for aquifer testing, for
example, such as slug and pump tests, it may not accurately represent
the hydrologic character of the tested strata. For a well placed within a
low hydraulic conductivity stratum, the presence of a mud cake can
greatly slow recovery by the well to the pre-sampling purge. This can
 Monitoring well development 641

-+-""""""'d,.-- HOLLOW STEM AUGER

CLA Y STRATUM -=-_-_

--------
..... . . :',', ,',

. :.' .

SAND STRATUM

. ... .
'
'.' .
',' .
DETAIL

Fig. 23.11 Conceptual view of the formation of a 'mud cake' during auger
drilling. Clays from the upper layer can be smeared across the surface of the
lower portion of the borehole, creating a low permeability sheath around the
borehole.

extend the time required to collect groundwater samples, in some cases

by days.
Mud cake problems should be addressed promptly following the
installation of the well. Hydrated clays, including those which
comprise the bentonite used in most drilling fluids, tend to form an
increasingly organized and coherent structure with increasing time. For
this reason, the difficulty in removing a mud cake increases if the cake
is left undisturbed following the drilling process. If left in place for
several days or weeks, therefore, a mud cake can cause irreparable
damage to the well's capabilities.

Turbidity reduction
As mentioned in section 23.3.6, filter pack installation is often imperfect
due to bridging or borehole instability. If these problems occur, the
physical result may be gaps in the filter pack, which will be occupied
by natural sediments. If the aquifer contains significant percentages of
silt or very fine sand, these fines will enter the well with the ground-
water through these 'breaches' in the filter pack, resulting in high
642 Monitoring wells

levels of turbidity. Due to interferences between these suspended sedi-

ments and certain analytical methods, this may result in water samples
from the subject well being unsuitable for testing. Although turbidity
can be eliminated after the fact by sample filtration, the practice of
filtration is not universally considered acceptable. It is preferable,
therefore, to minimize or eliminate the turbidity at the point of entry of
ground water into the well. This may be accomplished, in some cases,
by development of the well.
One approach is to remove sufficient water from the well to flush all
of the fines capable of passing the screen into the well, where they can
be collected and removed. This produces, in effect, a natural filter pack
where the breach in the artificial filter pack exists. For situations where
fines contents exceed 5-10%, however, or where these breaches are
numerous or extensive, it may not be possible to implement this
approach.
Under such circumstances, it is advisable to try an alternative
approach. The turbid water within the well should be removed using
a method which agitates the well and its surroundings as little as
possible. Submersible pumping (preferably with a continuous, rather
than a surging action) or double-tube airlifting are two possibilities.
Once the turbidity levels drop, the well can be purged and sampled.
The clarity of the water is not stable, however, since any agitation of
the well will result in renewed entry of fines, and therefore turbid
water, into the well.

23.5.2 Methods of development for monitoring wells

A large number of development options exist, with the required
hardware ranging from sophisticated prefabricated tools to implements
which may be constructed from readily available materials. In spite
of this diversity, most can be classified as causing groundwater flow
which is either unidirectional (water withdrawal) or bidirectional
(surging) relative to the well.
Unidirectional methods, such as submersible pumping and airlifting,
move water from the surrounding stratum into the well from the
aquifer. These methods transfer a minimum of energy to the filter pack
and surrounding soils, and so are not very useful for disruption of the
mud cake. These methods are useful in removing large volumes of
water from the well, in order to recover lost drilling fluids or flush out
fine sediment which will pass through the slots in the well screen.
Bidirectional methods surge the well, moving water into the well
from the surrounding soils, and back into those soils from the well
in a reciprocating manner. Commonly used bidirectional methods
include surging with a bailer or surge block, and the modification
of unidirectional methods such as pumping to permit periods of
 Dedicated purging and sampling systems 643
backwash. One example of the latter would be the use of airlifting
techniques to lift an air-water column almost to the surface, then
shutting off the air feed to the system and allowing the column to fall
back into the well. Repeated use of this technique can produce an
effective surging action. The effectiveness of bidirectional methods is
typically enhanced when performed on IOO-mm diameter wells, since
the larger open area of these wells permits more of the energy of the
surge to reach the filter pack and surrounding soils.

23.6 DEDICATED PURGING AND SAMPLING SYSTEMS

Piror to the sampling of a monitoring well, the water standing in the

well must be removed, permitting 'fresh' water from the aquifer to
enter the well. This procedure, called 'purging', is intended to remove
water which may have been subjected to chemical change due to
extended contact with the unnatural conditions and materials within
the well. Purge quantities are typically expressed in terms of casing
volumes, with a casing volume consisting of the volume of water
present within the riser and screen at the time of sampling. Purge
quantities of 3-5 casing volumes are commonly specified, although
many specifications now also call for the measurement of pH, specific
conductance and temperature of purged waters at regular increments.
Under these conditions, the purge is considered to be complete only
when these parameters stabilize.
In order to minimize the need for equipment transportation and
decontamination during sampling, purging systems which are dedicated
to a single monitoring well are widely used. Such a system remains
within the well continuously, and is used only for that well. Some
types may be used for sample collection, as well as purging. Such
dedicated purging systems are available in a range of types and degree
of complexity. Some of the more commonly utilized varieties are dis-
cussed in the following sections.

23.6.1 Bailers
Bailers have long been the workhorses of groundwater sampling, due
to their simplicity, low cost, and the wide range of bailer materials
available. A bailer consists fundamentally of a length of pipe or tubing
with a one-way valve fixed to the lower end. Bailers are available in
most well construction materials, including stainless steel, PVC, and
PTFE. When lowered on a rope or cable into the water within a well,
the valve unseals, permitting the bailer to fill. As the bailer is raised
from the water, however, the valve again seats and seals, retaining the
water within. The bailer is raised to the surface and dumped, and the
644 Monitoring wells
process repeated. The rope used is typically a relatively inert synthetic
fiber, such as nylon or polypropylene.
While very convenient for shallow wells, they are not recommended
for use in purging and sampling wells in which the depth to water is in
excess of approximately 30 m. Beyond that depth, the mass of rope
becomes unwieldy, the purging rate slows due to the time required for
each interaction, and any small leaks in the bailer valve (a common
phenomenon) result in the loss of most of the water load during the
upward trip.

23.6.2 Pneumatic pumps

Pneumatic pumps are operated by air or gas pressure, which is used to
displace a volume of fluid upward through a discharge line. Being air-
driven, these pumps require a source of pressurized gas or air at the
well-head. Samples are collected at the surface point of discharge. This
type of pump includes bladder, piston, and other styles, with some
variability in performance.
There are some advantages to the use of these systems for monitor-
ing. They can be constructed of materials such as stainless steel and
PTFE, which are widely considered to be relatively non-reactive with
groundwater (section 23.3.1). They are widely accepted for sample
collection for volatile organic chemical analysis, and may be used in 50-
mm diameter wells. Finally, they can be left relatively unattended for
several hours at a time for slow purging wells. Such systems are,
however, relatively expensive, and pumping rates for most styles are
generally low. Depending upon their construction, some types are
limited by air pressure requirements to depths less than approximately
75m.

23.6.3 Submersible electric pumps

These pumps consist of an electric motor and pump assembly, which
are suspended within the well at the lower end of a string of discharge
pipe or tubing. A source of electricity is required at the wellhead. This
type of pump has the broadest range of applicability of the systems
commonly used within monitoring wells. Models are available for flow
rates ranging from a few liters per minute to several hundred liters per
minute, and for depths to several hundred meters. Submersible pumps
for well diameters as small as 50 mm are currently available. Costs are
typically lower than for pneumatic systems.
Submersible pumps are currently not widely accepted for the collec-
tion of groundwater samples for volatile organic chemical analysis.
This is due to concern regarding the off-gassing of these volatile con-
stituents which may occur as the water passes through the pump's
 Dedicated purging and sampling systems 645

MAIN PUMP DISCHARGE

(MINIMAL CONTROL, MAXIMAL AERATION)

~==s=~~/..~ ....,-,. ?"--,

"-~:'\
'--.,,~
~

CONTROL VALVES

PUMP DISCHARGE PIPE

t'-. --1/4" SAMPLING DISCHARGE
(MAXIMAL CONTROL, MINIMAL AERATION)

~VOC SAMPLING VIAL

WELL RISER
1 f-----GROUND SURFACE

,L--=l-:t----- SUBMERSIBLE PUMP

Fig. 23.12 Construction of a wellhead sampling tap to permit the collection of

samples for VOC analysis from a submersible pump discharge.

rapidly rotating impeller system. It is probable, however, that the loss

of volatiles attributed to the passage through the pump is actually due
to aeration at the point of surface discharge. Since pumping rates from
submersible pumps tend to be relatively high, the collection of a non-
aerated sample can be quite difficult. This effect, however, can be
counteracted by the proper construction of a wellhead sampling port
for the collection of such samples, as illustrated in Fig. 23.12. Such a
sampling port should be of small diameter, and be placed upstream of
the main discharge. This permits sample collection prior to the aeration
which occurs at the pump discharge. If sampling is being performed in
order to satisfy regulatory requirements, however, the acceptability of
this method to the overseeing regulatory body should be confirmed in
advance of system installation.

23.6.4 Manual pumps

These are dedicated systems which are operated manually, and so
require no electrical or air supply. One type of manual system uses two
646 Monitoring wells
strings of PVC pipe, one nested within the other, and both placed
within the monitoring well. When the inner pipe is moved up and
down, the action of a plunger and valve system at its base moves water
up the annular space between the two pump pipes. The water is then
discharged from a spout on the larger pipe. Flow rates up to 201lmin
are feasible with these units. These pumps are not universally accepted
for sample collection for volatile organic chemical analysis, due to
concerns regarding aeration during the rise through the pump, and
within the pump near the point of discharge.
A second style of manual pump consists simply of a foot valve on a
length of flexible tubing extending into the well. These tubing or
inertial pumps are operated by rapidly moving the tubing up and
down within the well in a reciprocating manner. Since water can move
into the tubing only on the downstroke, due to the action of the foot
valve, the energy of the reciprocating movement gradually impels the
water up the tubing. Discharge occurs at the top of the tubing. Flow
rates of up to 1611min are possible, and operable depths extend to
approximately 45 m for manual operation. The depth range may be
extended to greater than 60 m by the use of a motorized assist to
provide the reciprocating motion. Materials of construction include a
variety of plastics, as well as PTFE. These pumps are probably more
widely accepted for volatile organic chemical sample collection than the
nested pipe pump discussed above, since the opportunity for water
contact with air is limited within the small diameter tubing of the
inertial pump.

23.7 MONITORING WELL SURFACE COMPLETIONS

After the well has been installed, a protective structure must be placed
around the riser at the ground surface. This structure, termed a surface
completion, is to protect the well against accidental impacts, tamper-
ing, or vandalism. Surface completions typically fall into one of two
types.
• Flush completions This type is utilized in active, high traffic areas,
where a high-profile completion would be an obstruction to site use.
These completions are low in physical profile, usually extending no
more than several centimeters above surrounding grade .
• Above grade completion This type is utilized in less active areas, or in
areas where a more visible completion is helpful, such as in dense
vegetation. these completions extend well above grade, typically by
0.6-1.2m.
Flush completions consist of a sturdy steel or concrete manhole or
meter box set in concrete around the top of the riser. The riser is cut off
 .1--.-+---- MANHOLE
CONCRETE SLEEVE

:~"-----SAND FOR
DRAINAGE

WELL RISER --r,,£,.!,,4

/-A----GROUT SEAL

BARRIER OR
GUARD POST

HASP

·. PLATE

CONCRETE
.. DRAIN
H---LOCKING COVER

Ir:1+---SAND (OPTIONAL)
HOLE

··..
/ CONCRETE PAD

. ..
. ....
,

.,
1[,.oI'.4I-':!--- WELL RISER

'/7''i--- GROUT SEAL

Fig. 23.13 Construction of a surface completion: (a) flush, using a manhole,

and (b) above ground, using a locking cover and protective casing.
648 Monitoring wells
just below grade prior to setting the manhole, and the manhole itself is
set slightly above grade to promote surface drainage away from the
well (Fig. 23.13a). When constructing this type of completion, the
manhole cover should never be in contact with the upper end of
the well riser or any cap on the riser. If the lid is resting on the riser
when a vehicle passes over the completion, damage to both the lid and
well may result.
Flush completions are not recommended for use in areas where
substantial surface water ponding may occur. They are also not well-
suited for grassy or brushy areas, where plant growth may make the
wells difficult to locate.
For above-ground completions, the riser of the well is terminated
well above grade. A protective casing, usually constructed of steel, is
placed over the riser and concreted in place. This casing is equipped
with a lockable cap or lid to prevent tampering or vandalism. A square
pad of concrete is constructed around the base of the protective casing
(Fig. 23. 13b). If the risk of damage from site activity is high, addi-
tional protection is afforded by the placement of steel barrier posts
at regular intervals around the well. These surface completions are
typically painted in bright colors to enhance their visibility.
Regardless of which style of surface completion is utilized, the
protection of the well should always include a cap or plug at the
upper end of the riser. This precludes the entry of condensation, rust,
and dirt which- may collect within the manhole or protective casing.
Available styles of monitoring well caps or plugs include the following:
• slip caps which slide over the top of the riser and are held in place
by friction;
• compression plugs which fit into the upper end of the riser (a wing
nut on the upper side of the plug permits the compression of the
plug, which forces a rubber bushing against the inner wall of the
riser - this seals the well and holds the plug firmly in place); and
• threaded caps or plugs which thread directly onto a threaded fitting
at the top of the riser - this type is useful only if the riser terminates
in a threaded fitting at the desired completion type, or can be
custom threaded in the field.

23.8 EQUIPMENT DECONTAMINATION

In association with ground-water monitoring installation practice,

decontamination, or 'decon', of equipment is performed for three
reasons:
• equipment entering a site may be deconned to remove contaminants
acquired at previously visited sites, or during travel;
 Equipment decontamination 649

• equipment may be deconned between each utilization, to prevent

the transport of contaminants between wells or boreholes, or be-
tween depths within a single borehole; or
• equipment exiting a site may be deconned to prevent the transport
of contaminants off-site.
The most commonly utilized decontamination technique for large
equipment consists of a high pressure (greater than 5 MPa), high tem-
perature (greater than 90°C) hot water or steam wash. This method, if
performed in a thorough manner, will remove residues of oils and
greases, caked mud, and dust and other loose particulates. For smaller
equipment, such as that used for soil sampling, this method may be
supplemented or replaced with solvent and/or detergent washes and
deionized water rinses. Clean equipment should be covered with
plastic sheeting for transport or storage.
Although a valid practice, decon is often carried to levels at which
the additional protection accrued does not justify the effort and cost
expended. Attempts to remove trace amounts of grease during de-
contamination, for example, may be fruitless if those same oils and
greases are known to be generated by the operation of the drilling rig.
Wrapping the well and riser in a sealing material for transport to the
installation site will not protect against airborne contaminants if those
same contaminants are present when the pipe is unwrapped at the
installation site.
As mentioned during the introduction to this chapter, the installa-
tion of a monitoring well must be accompanied by some disruption of
the subsurface, both physical and chemical. Rather than relying on
an extreme level of decontamination for 'perfect' protection from con-
taminant introduction for installations of high protocol, it is more
productive to utilize reasonable caution and sound installation tech-
niques. Even if minute amounts of foreign constituents are intro-
duced during installation, a properly installed and sampled well will be
purged several times during development and the pre-sample purge.
These purges should further reduce the probability of impact from any
introduction of trace constituents.

REFERENCES

Clark, L. (1988) The Field Guide to Water Wells and Boreholes, Open University
Press, John Wiley and Sons, New York, New York.
Dowell Schlumberger (1984) Cementing Technology, Nova Communications,
Ltd., London, England.
Driscoll, F.G. (ed.) (1986). Groundwater and Wells, 2nd edn, Johnson Division,
St. Paul, Minnesota.
650 Monitoring wells

Gass, T. (1988) Monitoring Well Filter Pack and Screen Slot Selection: A
Reassessment of Design Parameters. Water Well f., June, pp. 30-2.
Haug, A., Petrini, R.H., Grisak, G.E. and Klahsen, K. (1989) Application of
geostatistical methods to assess positions and spacings of groundwater
monitoring wells, in Proc., Conf. on Petroleum Hydrocarbons and Organic
Chemicals in Ground Water: Prevention, Detection and Restoration, NWWA,
Dublin, Ohio, pp. 159-74.
Hinchee, R.E. and Reisinger, H.J. (1987) A practical application of multiphase
transport theory to groundwater contamination problems, Ground Water
Monitoring Rev., Winter 1987.
Lehr, J., Hurlburt, S., Gallagher, B. and Voytek, J. (1988) Design and
Construction of Water Wells, NWWA, Van Nostrand Reinhold Company,
New York, New York.
Mackay D.M., Roberts, P.V. and Cherry, J.A. (1985) Transport of Organic
Contaminants in Groundwater, Environ. Sc. and Tech. 19(5), 384-92.
Mercer, J.W. and Cohen, R.M. (1990) Transport of immiscible fluid in the
subsurface, J. Contaminant Hydrology, 6, 107-63.
Parker, L.V., Hewitt, A.D. and Jenkins, T.F. (1990) Influence of Casing
Materials on Trace-Level Chemicals in Well Water. Ground Water
Monitoring Rev., Spring, pp. 146-56.
Schwille, F. (1984) Migration of organic fluids immiscible with water in the
unsaturated zone, in Pollutants in Porous Media: The Unsaturated Zone
Between the Soil Surface and Groundwater, (eds B. Yaron, G. Dagon, and J.
Goldschmid) Springer-Verlag, New York, pp. 27-48.
Schwille, F. (1989) Dense Chlorinated Solvents in Porous and Fractured Media: Model
Experiments, Lewis Publishers, Ann Arbor, Michigan.
Smith, D.W. (1976) Cementing, Amer. Inst. of Mining, Metallurgical, and
Petroleum Eng., Inc., Dallas, Texas.
USEPA (1986) RCRA Ground-water Monitoring Technical Enforcement Guidance
Document, OSWER-9950.1, US Government Printing Office, Washington,
DC.
 CHAPTER 24

Vadose zone monitoring

Lorne G. Everett

24.1 INTRODUCTION

Dr. Oscar E. Meinzer, in his tome entitled, Hydrology, which was first
presented by the National Research Council in 1942, referred to the
vadose zone as 'no man's land'. From 1942 through the formation of
the US Environmental Protection Agency (EPA) in 1970, the vadose
zone was fundamentally ignored. The first major national groundwater
monitoring contract was awarded by EPA in 1972, on a sole-source
basis, to General Electric's 'TEMPO' in Santa Barbara, California. In the
years 1974-76, it became clear to the research team that groundwater
investigations could not 'black box' the vadose zone. This initial effort
resulted in a IS-step groundwater monitoring methodology published
by Todd et al. (1976) and Everett (1980). Fundamental to this national
strategy was an appreciation of contaminant transport through various
parts of the vadose zone.
With the passage of the Resource Conservation and Reclamation Act
(RCRA) of 1976, the first federal requirement for vadose zone monitor-
ing was introduced. Specifically, soil core monitoring and soil pore-
liquid monitoring were required as a part of an unsaturated zone
monitoring strategy for hazardous waste land treatment units. Although
RCRA passed in 1976, the first guidance document did not appear
until ten years later (USEP A, 1986). Current state and federal legisla-
tion in the US, especially related to solid waste disposal sites and
underground storage tank sites, are now heavily directed towards
vadose zone monitoring.
Philosophically and practically, our approach to groundwater moni-
toring of hazardous waste sites has been flawed. In effect, designing a
groundwater monitoring strategy to detect potential contamination

Geotechnical Practice for Waste Disposal.

Edited by David E. Daniel.
Published in 1993 by Chapman & Hall, London. ISBN 0 412 35170 6
652 Vadose zone monitoring

WELL
• WATER

I ( 'e""ll"'"
I .:) 0 .''''e'-
0 0 01 r .. o- Pft .... Zone
0 LNAPL
0 AIR
o 0 0 • 01 (Wo,., 0"" 1..,1

I r~~~
• 0 0 0 • 0
'HG G
0 TIl,.. - Pfto •• Zone
00 a 00.
• 0 0 • 0 00 0
(Wo'.', L.NAPL on" AId
-OiL
• 0
-.-;:;~~~
3 0
o 0 00 0 00

°. ,..
Two-PlIo .. Zone
0 0 Q0. 00

....
• • Q. 0 0 (WO,., 01111 L.NAPLJ

... .....
(.NAPL

••• ,
00 0 0
0
·_L~· ~
WATER ••• On. - Pfto .. Zone
(Wole,1

Fig. 24.1 Conceptual representation of a distribution of LNAPL, water, and

air in a porous medium and monitoring well.

from a hazardous waste site is akin to a health monitoring strategy

which tells you when the patient is dead. For the most part, when
contamination, such as a light non-aqueous phase liquid (LNAPL) (Fig.
24.1) or a dense non-aqueous phase liquid (DNAPL), has got through
a thick vadose zone, the remediation costs can be staggering. The
philosophy of monitoring closed hazardous waste sites exclusively
through the use of monitoring wells is an even further indication of
a defective monitoring strategy. Currently, the EPA is evaluating
amendments to RCRA and CERCLA which will introduce early alert or
vadose zone monitoring strategies as a part of a closure program.
Prior to a discussion of vadose zone monitoring techniques, it is
imperative to realize a fundamental appreciation of storage and flux in
the vadose zone. The physical properties of the vadose zone associated
with storage of water include:
1. total thickness;
2. porosity;
3. bulk density;
4. water content;
5. soil/water characteristics;
6. field capacity (specific retention);
7. specific yield; and
8. fillable porosity.
Technical information on each of these properties may be found in
reference works by Davis and de Wiest (1966); Childs (1969); Hillel
(1971); Colley et al. (1972); Bouwer (1978); Brakensiek et al. (1979);
Freeze and Cherry (1979); Everett et al. (1984).
 Vadose zone description 653
The flux of water through the soil and lower vadose zone underlying
a source of contamination is staggeringly complex. Elements con-
tributing to the complexity of flow include variations in the state of
water saturation and spacial variations in the physical and hydraulic
properties of the vadose zone. The difficulties in attempting to describe
water movement in the vadose zone preclude presenting exact tech-
niques for estimating travel time of water and water borne pollutants
through this region. A number of indirect methods have been described
by Everett et al. (1984) covering:
1. infiltration at the land surface;
2. unsaturated flow in the vadose zone; and
3. flow in the saturated regions of the vadose zone.
This chapter on vadose zone monitoring is directed towards devel-
oping a geotechnical appreciation of the vadose zone through a dis-
cussion of soil moisture/tension relationships, soil pore-liquid moni-
toring, and some indirect soil moisture monitoring techniques.

24.2 VADOSE ZONE DESCRIPTION

The hydrogeologic profile extending from the ground surface to the

upper surface of the principal water-bearing formation is called the
'vadose zone'. As pointed out by Davis and de Wiest (1966), Bouwer
(1978), Driscoll (1986), USEPA (1986), and Everett et al. (1988), the term
'vadose zone' is preferable to the often-used term 'unsaturated zone'
because saturated regions or a wide distribution of tension-saturated
regions are frequently present in the vadose zone. Davis and DeWiest
(1966) subdivided the vadose zone into three regions, designated as:
the soil zone, the intermediate vadose zone, and the capillary fringe
(Fig. 24.2).

Soil zone
The term 'soil' is often used incorrectly to refer to the loose and
discontinuous blanket of decayed rock debris overlying solid bedrock.
This generic material is more appropriately referred to as the 'regolith'.
Soil, on the other hand, is only the very uppermost part of the regolith
where chemical and physical weathering are the most active. The
movement of water in the soil zone occurs mainly as unsaturated flow
caused by infiltration, percolation, redistribution, and evaporation
(Klute, 1965). The depth of the soil water zone typically varies from
1-10 m. Within these soils, horizons of heavy clay may develop which
can create shallow perched water tables (Everett, 1980; Everett et al.
1988). The water is held in soils by adhesive and cohesive forces in
 654 Vadose zone monitoring

Soil water

Vadose
zone
f Vadose
water
Inter-
mediate
vadose
zone

Water
- -
, ~
table LCaPiliary
water
Interstitial
water
Phreatic
Phreatic water
zone
(groundwater,
Sub-
suriace
water

Water in
unconnected
pores
i
I
I
I
I Water only In
chemical
combination
With rock

Fig. 24.2 Classification of subsurface water (Davis and de Wiest, 1966).

addition to body forces which are often called the 'force of gravity'.
This attraction of the water molecules for each other and the attraction
of the water molecule to the soil surface results in a thin film along the
surface of each soil particle. When the water-holding capacity of the
capillary forces is exceeded, water begins to percolate downward under
the force of gravity.

Intermediate vadose zone

As the weathered materials of the regolith merge with the underlying
deposits, which are often unweathered, a transition is made to the
intermediate vadose zone. In some regions, this zone may be practically
non-existent with the soil zone merging directly into the bedrock.
 Vadose zone description 655

The intermediate vadose zone in alluvial basins, however, may be

hundreds of meters thick, in which case the zone is unlikely to be
laid down in a uniform pattern. In fact, it is more likely that these
sedimentary basins may contain micro- or macro-lenses of silts and
clays which can be interbedded with gravels. Water in the intermediate
vadose zone may exist primarily in the unsaturated state and some
regions may receive very little inflow from above. Occasionally, perched
groundwater may develop in the interfacial deposits of regions con-
taining varying textures. Such perching may be hydraulically con-
nected to ephemeral or perennial stream channels, wherein temporary
or permanent perched water tables may develop.

Capillary fringe
The capillary fringe is the base of the unsaturated zone that merges
with underlying saturated deposits of the principal water bearing
formation. This zone is not characterized as much by the nature of
geologic materials as by the presence of water under conditions of
saturation or near saturation. Studies by Luthin and Day (1955) and
Kraijenhoff van deLeur (1962) have shown that both the hydraulic
conductivity and flux may remain high for some vertical distance in the
capillary fringe, depending on the nature of the materials. In general,
the thickness of the capillary fringe is greater in fine materials than in
coarse deposits.
An understanding of the height of the capillary fringe is important
for several reasons. Clearly, any capillary fringe which extends up into
the unprotected foundations of buildings will result in continuous
mildew problems, resulting in esthetic and structural damage (Fig.
24.3). In addition, investigations related to petroleum hydrocarbon
spills and leaks have shown that the water table will act as a barrier
to the free product light non-aqueous phase liquids. Controversy
exists, however, as to the extent of hydrocarbon movement through
the capillary fringe. One school of thought implies that the capillary
fringe acts as a barrier to the free hydrocarbon product. A second
school of thought indicates that the concept of a capillary fringe is
not consistent with the interpretation of hydrocarbon models. Current
federal legislation related to monitoring underground storage tanks,
however, dictates that the screened interval must extend 1.5 m above
the seasonal high water table to account for capillary activity. It should
be pointed out, however, that capillary rise which is determined by
grain size distribution is indicative of a minimum capillary fringe,
whereas, infiltration/percolation of water from the surface could result
in a maximum capillary rise on top of which hydrocarbon contamina-
tion may exist. The estimation of minimum and maximum capillary rise
measurements is poorly documented in the literature.
656 Vadose zone monitoring

SILT CLAY

1000

o HOLTZ/KOVACS (1981)
McCARTHY (1977)

o HANSBO (1975)

{:;, LANE/WASHBURN (1946)

100

w
..J
<I:
U
C/)

~
0
..J
~

W
C/)
10
c:
>-
c:
<I:
..J
..J
Cl..
<I:
U

1.0 0.1 0.01 0.001 0.000 1

EFFECTIVE PORE DIAMETER (mm) LOG SCALE

Fig. 24.3 Height of capillary rise.

Vadose zone flow regimes

The classical concept of infiltration depicts a distinct, somewhat
uniform wetting front which slowly advances in a Darcian flow regime
after a precipitation event. This wetting front infiltration concept was
formally presented by Bodman and Coleman in 1944. The second flow
regime is sometimes referred to as flow down macropore, non-capillary
 Soil pore-liquid monitoring for hazardous waste sites 657
flow, subsurface storm flow, channel flow, and other descriptive
names, but most often referred to as macro pore flow. Unfortunately,
this is a poor choice of terms since macropores typically are dis-
continuous, and therefore could not support the flows associated with
them. A more accurate term should be secondary porosity or fracture
flow. For a mathematical comparison of unsaturated Darcian flow and
flow-through secondary porosity, see Hem and Mellencon (1986) and
Mercer et al. (1983).

24.3 SOIL PORE-LIQUID MONITORING

FOR HAZARDOUS WASTE SITES

Generally, we distinguish three ranges of water saturation between the

limits of 0% and 100%. Figure 24.4 shows water in a water wet
granular soil such as sand. At a very low saturation water forms rings,
called pendular rings, around grain contact points. The air/water
interface takes on the shape of a saddle under these conditions.
As seen in Fig. 24.4a, the water saturation is rather low, and the
rings are isolated and do not form a continuous water phase. Although
a very thin film of water, several molecules thick, which does not
behave as ordinary liquid water, due to the very strong forces of
attraction between the water and solid molecules, does remain on the
solid surface, practically no pressure can be transmitted through it
from one ring to the next. These conditions are typically too dry to
use field tensiometers effectively. Figure 24.4b shows a pendular ring
between two spheres. For this case, it is possible to relate the volume
of the ring to the radius of the curvature of the air/water interface. The
latter, in turn, is related, to the difference in pressures in the air and
water across it.
As the amount of water saturation increases, the pendular rings
expand until a continuous water phase is formed. The saturation at
which this occurs is called equilibrium water saturation. Above this
equilibrium saturation, the saturation is called funicular and flow of
water is possible (Fig. 24.4c). Both the water and the air phases are
continuous. As the water saturation increases, a situation develops in
which the air (non-wetting phase) is no longer a continuous phase; it
breaks into individual bubbles lodged in the larger pores (Fig. 24.4d).
The air is then said to be in a state of insular saturation. These bubbles
of air can only move if a pressure difference, sufficient to squeeze it
through a capillary size restriction, is applied across it in the water.
Obviously, if the air can escape from the void space (or is dissolved in
the water) we have complete water saturation.
Classically, moisture in the vadose zone system is broken down into
three components: gravity, pellicular, and hygroscopic waters. Gravity
658 Vadose zone monitoring

---, --I~-"
(c)
(a)

c:::J _ (b)

air water

(d)

Fig. 24.4 Water and air saturation states.

water is the moisture in a soil which can be drained by gravitational

forces; pellicular water is the moisture in a soil which cannot be drained
by gravity forces, but can be lost to the atmosphere through natural
evaporation; and hygroscopic water is the moisture which will never
be lost through the above natural processes. This distribution of soil
moisture can be mathematically approximated by a step-wise distribu-
tion composed of three components (Mercer et al., 1983). Figure 24.5
shows a comparison of the step-wise wetness distribution and its
corresponding original wetness distribution.

24.3.1 Soil moisture/tension relationships

Unlike water in a bucket, free, unlimited access to water does not exist
in the soil. Soil water or, as it is frequently called, 'soil moisture', is
stored in the small 'capillary' spaces between the soil particles and on
the surfaces of the soil particles. The water is attracted to the soil
particles, and tends to adhere to the soil. The smaller the capillary
spaces between the particles, the greater the sticking force. For this
reason, it is harder to get moisture out of fine clay soils than it is from
the larger pores in sandy soils, even if the percentage of moisture in
the soil, by weight, is the same.
Figure 24.6 shows the results of careful research work done with
special extractors. As described by the Soilmoisture Equipment
Corporation (1983), the graph shows the relationship of the percent of
moisture in a soil to the pressure required to remove the moisture from
the soil. These are called moisture retention curves. The pressure is
 Soil pore-liquid monitoring for hazardous waste sites 659
Wetness
•
Or----y---------,--------------------,---------~

Gravity water
deficit

~j
Natural wetness
distribution curve ~ __

Step-wise wetness _____________

distribution curve
Bottom of trench cap

t_--I·If4-__
.. ...

Component of porosity
component of porosity
~..
for pellicular water - -..
I I
-c-o-m-p-o-n-en-t-o-f-p-o-ro-s-ity~·
for gravity water
for hygroscopic water

Fig. 24.5 Comparison of a schematic natural wetness distribution curve and

its corresponding step-wise wetness distribution curve, after Mercer et al.
(1983).

measured in bars* which is a unit of pressure in the metric system.

Figure 24.6 clearly points out that two factors are involved in determin-
ing ease of water sampling: moisture content; and soil type.
Moisture in unsaturated soil is always held at suctions or pressures
below atmospheric pressure. To remove the moisture, one must be
able to develop a negative pressure or vacuum to pull the moisture
away from around the soil particles. For this reason, we speak of 'soil
suction'. In wet soils, the soil suction is low, and the soil moisture can
be removed rather easily. In dry soils, the soil suction is high, and it is
difficult to remove the soil moisture.
Given two soils (one clay and one sand) with identical moisture
contents, it will be more difficult to extract water from the finer soil
(clay) because water is held more strongly in very small capillary
spaces in clays.
Another fact, brought out by the graphs on Fig. 24.6, is that silty clay
soil with 30% moisture, if placed in contact with a sandy soil with only

* By definition, a bar is a unit of pressure equal to 106 dyne/cm2. It is equivalent to

100 kPa (Kilopascals) or approximately 1 atmosphere, or 750 mm of mercury, or 1020 cm
of water.
660 Vadose zone monitoring
50

40
E
Cl
'CD
~ Chino silty clay
~
:s.
~
::l
Cii
'0
E
'#.
10 Pachappa fine sandy loam
Hanford sand

00 0.8 2
Soil suction (bars)

Fig. 24.6 Moisture retention curves - three soil types, after Soilmoisture
Equipment Corporation (1983).

10% moisture will actually suck moisture out of the sandy soil until the
moisture content in the sandy soil is only 5%. This is due to the greater
soil tension in the fine clay texture.

24.3.2 Soil pore-liquid monitoring

The sampling of soil pore-liquid was reported in the literature in the
early 1900s when Briggs and McCall (1904) described a porous ceramic
cup which they termed an 'artificial root'. As shown in Fig. 24.7,
different soils are capable of yielding different levels of water. At
saturation, the volumetric water content is equivalent to the soil porosity,
while in contrast, the unsaturated zone is usually found to have a soil
moisture content less than saturation. For example, the specific reten-
tion curve on Fig. 24.7 depicts the percentage of water retained in
previously saturated soils of varying texture after gravity drainage has
occurred. Lysimeters are used to sample pore-liquids in unsaturated
media because pore-liquid will not readily enter an open cavity at
pressures less than atmospheric (the Richard's outflow principle).
Lysimeters are made up of a body tube and a porous segment. When
placed in the soil, the pores become an extension of the pore space
of the soil. Consequently, the water content of the soil and porous
segment become equilibrated at the existing soil-water pressure. By
applying a vacuum to the interior of the porous segment such that
the pressure is slightly less inside the lysimeter than in the soil,
flow occurs into the lysimeter. The sample is pumped to the surface,
 Soil pore-liquid monitoring for hazardous waste sites 661
% by volume
I\) I\)
o o o (}1

Clay and silt

Sandy clay

Fine sand

Fine sand

Medium
sand

Coarse
sand
--------
Coarse
sand

Gravelly
sand
(J)
"0
Fine <D
0
gravel 3i
0
Medium
gravel ~
~
0
::J
Medium
gravel

Coarse
gravel

Coarse
gravel

Boulders

Fig. 24.7 Variation of porosity, specific yield, and specific retention with grain
size, after Scott and Scalmanini (1978).

permitting laboratory determination of the quality of the soil pore-

liquids.
It should be recognized, however, that situations may occur where
the flow velocities in the unsaturated zone are higher than empirically
demonstrated by Darcy's Law. As a result, the wetting front will not be
uniform and most of the flow will occur through macropores. This type
of gravity flow in highly structured soils will not be sampled effectively
by suction lysimeters. The most promising technique for sampling soil
662 Vadose zone monitoring
pore-liquid in highly structured soils is pan lysimeters (e.g., free
drainage samplers).

24.3.3 Pore-liquid sampling equipment

Well and open cavities cannot be used to collect solution flowing in the
unsaturated zone under suction (negative pressures). The sampling
devices for such unsaturated media are thus called suction samplers or
lysimeters. Everett and Wilson (1983) provide an in-depth evalua-
tion of the majority of unsaturated zone monitoring equipment. Law
Engineering and Testing Company (1982) provides a description of
some of the available suction lysimeters.
Suction lysimeters may be subdivided into three categories:
1. vacuum operated soil-water samplers;
2. vacuum-pressure samplers; and
3. high-pressure vacuum samplers with check valves.
Soil-water samplers generally consist of a ceramic cup mounted on the
end of a small-diameter PVC tube, similar to a tensiometer. The upper
end of the PVC tubing projects above the soils surface. A rubber
stopper and outlet tubing are inserted into the upper end. Vacuum is
applied to the system and soil water moves into the cup. To extract a
sample, a small-diameter tube is inserted within the outlet tubing and
extended to the base of the cup. The small-diameter tubing is con-
nected to a sample-collection flask. A vacuum is applied via a hand
vacuum-pressure pump and the sample is sucked into the collection
flask. These units are generally used to sample to depths up to 2 m
from the land surface. Consequently, they are used primarily to
monitor the near-surface movement of pollutants from irrigation return
flow.
To extract samples from depths greater than the suction lift of water
(about 8 m), a second type of unit is available, the so-called vacuum-
pressure lysimeter. These units were developed by Parizek and Lane
(1970) for sampling the deep movement of pollutants from a land
disposal project in Pennsylvania. The design of the Parizek and Lane
sampler is shown in Fig. 24.8. The body tube of the unit is about
600mm long, holding about 11 of sample. Two polyethylene lines are
forced through a two-hole rubber stopper sealed into a body tube. One
line extends to the base of the ceramic cup as shown and the other
terminates a short distance below the rubber stopper. The longer line
connects to a sample bottle and the shorter line connects to a vacuum-
pressure pump. All lines and connections are sealed.
In operation, a vacuum is applied to the system (the longer tube to
the sample bottle is clamped shut at this time). When sufficient time
has been allowed for the unit to fill with solution, the vacuum is
 Soil pore-liquid monitoring for hazardous waste sites 663

Two-way pump

Vacuum port
and gauge Sample bottle

Pressure Discharge tube

vacuum in
.... L.JULJ. . .
Bentonite .::\: .::}
3" .. ,
16 tube ' ..

Plastic pipe
24" long

6" hole
with tamped
silica sand
backfill

Porous cup

Bentonite
..............
..............
.. . . . . . . . . . . ...
Fig. 24.8 Vacuum-pressure sampler (Parizek and Lane, 1970).

released and the clamp on the outlet line is opened. Air pressure is
then applied to the system, forcing the sample into the collection flask.
A basic problem with this unit is that when air pressure is applied,
some of the solution in the cup may be forced back through the cup
into the surrounding pore-water system. Consequently, this type of
pressure-vacuum system is recommended for depths only up to about
15 m below land surface. In addition to the monitoring effort of Parizek
and Lane, these units have been used extensively to sample leachate
movement in the vadose zone underlying sanitary landfills.
Wood (1973) reported on a modified version of the design of Parizek
and Lane. Wood's design, the third suction sampler to be discussed in
this subsection overcomes the main problem of the simple pressure-
vacuum system; namely, that solution is forced out of the cup during
application of pressure. A sketch of the sampler is shown in Fig. 24.9.
The cup ensemble is divided into lower and upper chambers. The two
chambers are isolated except for a connecting tube with a check valve.
A sample delivery tube extends from the base of the upper chamber to
664 Vadose zone monitoring

Vacuum-air pressure line

-t-----H- Upper check valve

I-I:::",..-L-----Jt-- Sample discharge line

-~H- Upper chamber

Vl<Tn\i~~t-- Lower check valve

+----H-Tubing

Lower chamber
Suction cup

Fig_ 24.9 'Hi/pressure-vacuum soil-water sampler'. Courtesy of Soilmoisture

Equipment Corporation (1978).

the surface. This tube also contains a check valve. A second shorter
tube terminating at the top of the sampler is used to deliver vacuum or
pressure. In operation, when a vacuum is applied to the system, it
extends to the cup through the open one-way check valve. The second
check valve in the delivery tube is shut. The sample is delivered into
the upper chamber, which is about 1 1 in capacity. To deliver the
sample to the surface, the vacuum is released and pressure (generally
of nitrogen gas) is applied to the shorter tube. The one-way valve to
the cup is shut and the one-way valve in the delivery tube is opened.
Sample is then forced to the surface. High pressures can be applied
with this unit without danger of damaging the cup. Consequently, this
sampler can be used to depths of about 100 m below land surface
(Soilmoisture Equipment Corporation, 1978). These samplers have
been widely used in solid waste landfills.

24.3.4 Porous segments in lysimeters

The vadose (unsaturated) zone consists of a mixture of soil particles,
water that is held on the surface of the particles and in small capillary
 Soil pore-liquid monitoring for hazardous waste sites 665

spaces between the particles, and interconnecting air passages that are
open to the atmosphere at the soil surface. Removing moisture for
chemical analysis from the vadose zone requires the use of special
porous materials. Simply exerting a suction on an open tube inserted
into the vadose zone will not remove moisture since the interconnecting
air passages in the soil will result only in the flow of air into the
evacuated tube. However, by using a porous cup sealed to the end of
the tube, samples can be removed by suction, providing the diameter
of the individual pores in the porous cup do not exceed a critical value.
If the porous cup is fabricated from a hydrophilic material, such as
ceramic, water will fill the pores of the cup completely. The water bonds
to the porous ceramic and cannot be removed from the pores unless
the air pressure differential across the wall of the cup reaches a critical
value which is related to the pore size. If the porous cup is fabricated
from a hydrophobic material such as PTFE (polytetrafluorethylene),
water will fill the pores of the cup but the bonding of the water to the
hydrophobic material will be less.
The air pressure required to force air through a porous cup which
has been thoroughly wetted with water is called the 'bubbling pressure'
or 'air entry value'. The smaller the pores in the cup the higher this
pressure will be. The relation of the pore size to the bubbling pressure
or air entry value is defined by the equation

D = 30Y (24.1)
P
where D is the pore diameter measured in microns, P is the bubbling
pressure or air entry value measured in millimeters of mercury, and Y
is the surface tension of water measured in dynes/cm.
In order to build a soil water sampling device which can be used
successfully in the vadose zone to withdraw moisture from the soil, the
device must incorporate a porous cup which has pores so small that
the air in the soil, under atmospheric pressure, cannot enter even
though a full vacuum is created within the sampler. Under these
conditions, water from the capillary spaces in the soil will flow through
the pores in the porous cup and into the sampler but air will not enter
(Fig. 24.10).
With respect to Eq. (24.1), the maximum size of the pores thatwill
permit this action is as follows. At 20 e the surface tension of water is
0

72 dynes/cm. The maximum air pressure is 1 atmosphere or 760 mm of

mercury. In accordance with the equation, the maximum pore size in
the porous cup would be D = (30)(72)/(76) = 2.8Ilm. The pore size of
ceramic cups is between 2 and 311m. If the pores of the wetted sampler
cup do not exceed 2.81lm in diameter, then a full vacuum can be
maintained within the sampler and the water films in the pores of the
porous cup will not break down. If the pore size of the cup is twice
this amount, namely 5.7Ilm, then the maximum vacuum that can
 666 Vadose zone monitoring
OutsIde of Iyslmeter cup wall

A Soil/silica flour 1 . ----

I Inside of Iyslmeter cup wall

Ceramic/PTFE cup

Pore size { ceramic - 2- 3 ml~rons

PTFE - 70- 90 mIcrons

B Atmospheric Wetted (pores full)

pressure (1 bar) - - atmospheric pressure (1 bar)

C Atmospheric
- - Lyslmeter suction (2 cb)
pressure (1 bar)

D Atmospheric
- - Lysimeter suction (4 cb)
pressure (1 bar)

Lysimeter suction
E Air entry
(bupbling pressure) ---~~~~-. > 233 cb' ceramic
> 7 cb PTFE

Fig. 24.10 Diagrammatic view of lysimeter cup wall.

* Vacuum > lOOcb not possible
be pulled within the sampler is 380 mm of mercury or 50% of an
atmosphere . Likewise, if the pore size is twice again as large, namely
ll.4llm in diameter, then the maximum vacuum that can be pulled
without the cup leaking air is 190mm of mercury or 25% of an atmos-
phere. Since the majority of the pores used in PTFE suction lysimeters
are about 40l1m, the bubbling pressure is only a few per cent of an
atmosphere.
Where porous materials are being used in air-water systems, such as
in suction lysimeters, the most direct method of evaluating the pore
size of the material is through the use of air pressure. By thoroughly
wetting the porous material and then exposing one side of it to increas-
ingly higher air pressure values, with the other side under water, one
can readily observe when the air pressure becomes high enough to
enter the pores and cause bubbling on the opposite side. The specific
 Soil pore-liquid monitoring jar hazardous waste sites 667
air pressure at which this bubbling occurs is a direct measurement of
the pore size as defined by the above formula and indicates directly
the effectiveness of the porous materials to withstand air pressure dif-
ferentials when in use. Evaluating pore size distribution by the mercury
intrusion method or other means does not give direct information as to
how the porous material will perform in the air-water system in which
it is used. PTFE pores are generally round and symmetrical, while
ceramic pores are of various ragged shapes. The strength of the water
meniscuses in the individual pores are a function of pore shape as well
as overall size, and for this reason an accurate measurement of the
pressure at which the meniscus will break down and allow air to enter
can only be made accurately by direct measurement of the bubbling
pressure or air entry value of the wetted porous material.

24.3.5 Lysimeter failure confirmation

In the event that a sample cannot be retrieved from an installed suction
lysimeter under conditions where the operator knows that the soil
suction levels should be low enough to obtain a sample, such as
after a major rainfall event, specific procedures should be followed.
Adjacent to a suction lysimeter that appears to have failed, a soil
suction determination must be made to determine if the available soil
moisture is high enough to obtain a sample. Soil suctions are deter-
mined using tensiometers. Tensiometers are commercially available
and are produced with various designs and lengths.
A tensiometer consists of a tube with a porous ceramic tip on the
bottom, a vacuum gauge near the top, and a sealing cap. When it is
filled with water and inserted into the soil, water can move into and
out of the tensiometer through the connecting pores in the tip. As the
soil dries and water moves out of the tensiometer, it creates a vacuum
inside the tensiometer, which is indicated on the gauge. When the
vacuum created equals the 'soil suction', water stops flowing out of
the tensiometer. The dial gauge reading is then a direct measure of the
force required to move the water from the soil. If the soil dries further,
additional water moves out until a higher vacuum level is reached.
When moisture is added to the soil, the reverse process takes place.
Moisture from the soil moves back into the tensiometer through the
porous tip until the vacuum level is reduced to equal the lower soil
suction value, then water movement stops. If enough water is added to
the soil so that it is completely saturated, the gauge reading on the
tensiometer will drop to zero. Because water can move back and forth
through the pores in the porous ceramic tip, the gauge reading is
always in balance with the soil suction.
The effective operational range for suction lysimeters is between
saturation and 60 centibars of suction as determined by the tensiometer.
668 Vadose zone monitoring
Above 60 centibars of suction, a ceramic lysimeter will operate (Everett
et aI., 1986). However, the flow rates will be so low that effectively one
cannot get a sample. If the tensiometer readings are between 0 and 60
centibars of suction, the suction lysimeter should obtain a sample. If
no sample is obtained under these soil suction ranges, the suction
lysimeter will be deemed to have failed and should be excavated or
abandoned.

24.4 INDIRECT MONITORING DEVICES

Monitoring units classically are grouped into direct sampling or

measurement units and indirect measurement units. Soil suction
samplers measure soil salinity directly, whereas measurements using
the four-probe method must be converted to salinity by a calibration
relationship. Similarly, tensiometers measure matric potential directly,
whereas the output from heat dissipation sensors must be converted to
matric potential using a calibration curve. Numerous types of vadose
zone indirect measurement techniques are available which include
the EC probe, salinity sensors, gravimetric measurements, neutron
moderation, gamma ray attenuation, hygrometer/cycrometers, heat
dissipation sensors, resistor/capacitor type sensors, remote sensing,
heat pulse methods, etc. Each of the above indirect vadose zone
monitoring techniques are extensively described by Everett et al. (1984).
The most promising of the indirect vadose zone monitoring tech-
niques is called neutron moderation. Neutron moderation has been
shown to be one of the most accurate and precise indirect predictors of
soil moisture distribution. As a result, neutron moderation is rapidly
gaining acceptance as a real-time, minimal cost indirect technique for
measuring dissolved contaminant migration in the vadose zone.

Principle of the neutron moderation method

The principle of neutron moderation depends on two properties
related to the introduction of neutrons with matter: scattering and
capture (Gardner, 1965). High-energy neutrons emitted from a radio-
active source may be slowed down or thermalized by collisions with
atomic nuclei. The statistical probability for such thermalization
depends on the scattering 'cross-section' of various nuclei. The scatter-
ing cross-section of hydrogen causes a greater thermalizing affect on
fast neutrons than many elements commonly found in soils. This forms
the basis for detecting the concentration of water in soil (van Bave!,
1963). The second property of interest in the neutron moderation
method is capture of slow neutrons by elements present in the soil
with the release of other nuclear properties or energy. Cadmium and
 Indirect monitoring devices 669
boron have extremely large capture cross-sections compared with
hydrogen. The property of energy released during capture serves as a
means of detecting the concentration of slow neutrons.
In operation, when a source of fast neutrons is lowered into a soil
through a suitable well or casing, a cloud of thermalized neutrons is
established (Gardner, 1965). This cloud reflects the moderating effect of
scattering cross-sections of nuclei in the soil mass on fast neutrons. If a
suitable calibration is made to isolate the moderating effects of soil
nuclei other than hydrogen, changes in the volume of the thermalized
cloud will reflect changes in water content. In general, the wetter the
soil, the smaller the volume of thermalized neutrons. A detector that
relies on capture of thermalized neutrons is used in conjunction with
suitable electronic circuitry to measure the water volume.
Instrumentation used to measure the water content of neutron
moderation requires three principal components:
1. a source of fast neutrons, i.e., 50 millicuries of americium beryllium;
2. a detector of slow neutrons, i.e., boron detector; and
3. an instrument to determine the count rate from the detection
equipment (Fig. 24.11).

Field implementation
Access tubes for neutron moisture logging may be constructed of
seamless steel, aluminum, or PVc. In addition, access tubes can either
use closed or perforated casing. Although the use of PVC casing slightly
dampens the signal, more than sufficient thermalized neutrons can be
detected to measure soil moisture over all ranges of concern (Keller
et al., 1989).
The application of neutron moderation at landfill sites is gaining
acceptance. In addition, the cost advantages of neutron moderation are
resulting in widespread application of the technology as a part of
long-term post-closure monitoring strategies for closed landfills and
contamination sites. Neutron probes cost less than $5000, provide real-
time digital readout, have zero analysis cost, and have a lifetime which
is substantially longer than most landfill or contamination sites.
With respect to landfills and contamination sites, the neutron
moderation access tubes can either be installed vertically, on an angle,
or horizontally beneath landfill cell expansion areas. Since neutron
moderation operates over the range between zero matric potential
(water saturation) and 90 bars of soil water tension, the neutron probe
can be utilized under all conditions. Although the neutron probe was
originally designed to measure indirectly soil moisture distribution,
recent patents have expanded the application of neutron moderation to
include soil gas analysis. In effect, patents held by Toxic Technology,
670 Vadose zone monitoring

r-~---- Single-conductor
cable to logging
equipment

Land surface

'.: ~rmal neutron

. / ....!J'" 1<-1 00 eV

Electronics Fast ~
and neutron ~
power supply 1Qs eV ~
. 1 MeV Hydrogen
atom

Neutron moderation

Detector

) - Neutrons

Source sub

When beryllium is
bombarded with a
particles from
americium, neutrons
are emitted

Average energy loss :'

per collision: .~.,---~.,....,..,---~'.'.' Neutron capture
hydrogen 63% .
oxygen 12%

Fig. 24.11 Equipment and principles of neutron moisture logging.

Inc. of Costa Mesa, California utilize a perforated hydroprobe access

casing wherein the neutron probe identifies the location of the con-
tamination. Subsequently, a soil gas sample is taken adjacent to the
depth identified by the neutron probe and analyzed using a field GC to
confirm the presence of volatile organic contaminants. This approach is
particularly attractive because the neutron probe is used as a least-cost
screening tool, wherein confirmation of dissolved organic contamina-
tion can be obtained through soil gas analysis. Clearly, therefore, the
installation of neutron probe access tubes below landfill liner sumps,
beneath V-shaped landfill cells, or through any depth associated with
 Indirect monitoring devices 671

contamination sites, provide a very accurate, and precise method of

measuring not only soil moisture migration, but also dissolved con-
taminant transport.

Underground storage tank monitoring

One example of the application of neutron moderation to underground
storage tanks is provided to show the multiple application of direct and
indirect monitoring techniques. When a release takes place from a
petroleum hydrocarbon underground storage tank, fluids will begin to
pass through the vadose zone. This release results in an absorbed
phase in the vadose zone, sometimes called residual oil saturation, a
vapor phase, a dissolved phase in soil moisture, and a free product
phase. The neutron moderation technique has been shown to defini-
tively identify each of these phases. In addition, the application of
neutron moderation satisfies tank integrity testing, tank and pipeline
monitoring, risk analysis for adsorbed and dissolved phase, dis-
tinguishes between spills and leaks, and greatly enhanced passive and
active remediation.
The neutron probe is a 50 mm by approximately 250 mm long probe
which conveniently passes through a 50 mm subsurface casing which
may be either closed or perforated. If perforated casing is used, the
process is protected by a US Patent held by Toxic Technology, Inc.
Based on Richard's principle of unsaturated flow, fluids will not move
from a small pore to a large pore under unsaturated conditions, and, as
such, water will not move into the perforated casing in the vadose
zone. Hundreds of gasoline stations have been monitored using per-
forated casing and not one incident of fluids moving into the casing
have been detected, thereby totally supporting Richard's principle. In
effect, the neutron moderation technique uses the neutron probe to
give an exact vertical distribution of soil moisture (dissolved phase in
the vadose zone or absorbed phase of hydrocarbon in the vadose zone)
(Fig. 24.12). Since the neutron probe reads hydrogen atoms from either
water or hydrocarbons (or, for that matter, from any other source of
hydrogen) a field gas chromatograph (GC) is used to confirm the
presence of volatile hydrocarbons.
A typical installation for monitoring an underground storage tank
involves four slotted PVC casings located in the corners of a tank
excavation. When the access holes are drilled, a continuous core is
obtained to determine the exact background condition. Regardless of
background conditions, any deviation in terms of any of the flow
conditions described above is precisely picked up by the neutron probe
and confirmed by the Gc. The time savings, the cost advantages, and
the multiple application of the technique has enormous attraction to an
underground storage tank program.
672 Vadose zone monitoring

Shell Oil (Manchester and Denker)

8000-r------------------------------------~

7000 _____ Soil at residual oil

/' saturation

'E Water
::J
0 6000 table
()

c
e
"5 Soil at natural
(I) 5000
z condition

4000

3000
I
0 10 20 30 40
Depth (tt)
----<::>----- 615, 10/15
- - - . - - - 615,10/23

Fig. 24.12 Vertical profile of vadose zone contamination based upon neutron
moderation.

Doctoral candidates at the University of California at Santa Barbara

have demonstrated through artificial leakage conditions that a rate of
O.ll1hr can be determined using the neutron probe system. Other
researchers have routinely demonstrated that the technique can satisfy
the 0.2 lIhr leak detection capability typically required in the US. If an
anomaly is identified, successive runs with a tighter timeframe, i.e.,
measured in hours, can demonstrate whether dissolved product or free
product is moving down through the excavation. This rate of flow for
both the dissolved product, adsorbed product, and free product con-
stitutes the basis for a definitive risk analysis. By tracking the move-
ment of the contaminant front. one can easily determine the difference
between a spill or a leak.
With respect to remediation, the perforated access holes offer
immediate opportunities for minimal cost techniques. For example,
small spills and leaks, which have been repaired, can be quickly
remediated through the use of soil venting techniques using a blower
system on the existing perforated access tubes. In addition, the
opportunity to use an inlet air tube and to drive the volatiles towards
the other three outlet perforated tubes can accelerate the process.
For spills and leaks of larger proportion, further active approaches
can include the use of the perforated access holes to introduce air,
nutrients, soil moisture, bacteria, pH controls, etc., all of which can be
 References 673

easily introduced through the use of the perforated casing to accelerate

either an active or a passive bioremediation program.
A major advantage of the technique is gaining widespread applica-
tion in California relative to passive hydrocarbon remediation. In
particular, the neutron probe provides a definitive risk analysis relative
to whether the dissolved phase or free product is moving. The use of
packers allows the taking of soil gas samples directly adjacent to the
lens of contamination. In particular, the neutron probe provides a
definitive risk analysis relative to whether the dissolved phase or free
product is moving. The use of packers allows the taking of soil gas
samples directly adjacent to the lens of contamination. The analysis of
the gas distribution in the vapors obtained utilizes the ratio of carbon
dioxide to oxygen to demonstrate passive bioremediation. The gas
analysis also demonstrates the aerobic decomposition of the hydro-
carbon, however, if the system should turn, anaerobic, air can easily be
introduced through the perforated casing. The elevated levels of CO 2
relative to O 2 demonstrates that passive bioremediation is taking
place and since neither the dissolved phase nor the adsorbed phase is
moving, the contaminated soil can be left in place.
The application appears to be a combination of proven techniques
which satisfy tank integrity testing, pipeline testing, leak detection
monitoring, active remediation programs, and more importantly,
allows for inplace passive remediation strategies. Since the radio-
active source will certainly outlive the hydrocarbon contamination, the
real time, zero analysis cost screening, is a major advantage to the
technique.

REFERENCES

Bodman, G.B. and Coleman, E.A. (1944) Moisture and Energy Conditions
During Downward Entry of Water into Soils, Soil Sci. Soc. Amer. Proc., 8,
116-22.
Bouwer, H. (1978) Groundwater Hydrology, McGraw-Hill, New York.
Brakensiek, D.L., Osborn, H.B. and Rawls, W.J. (1979) Field Manual for Research
in Agricultural Hydrology, Agricultural Handbook No. 224, Sciences and
Education Adm., US Dept. of Agric.
Briggs, L.J. and McCall, A.G. (1904) An artificial root for inducing capillary
movement of soil moisture, Science 20, 566-9.
Childs, E.C. (1969) An Introduction to the Physical Basis of Soil Water Phenomena,
Wiley Interscience, New York.
Cooley, R.L., Harsh, J.F. and Lewis, D.C. (1972) Principles of Ground-Water
Hydrology, Hydrologic Engineering Methods for Water Resources
Development, Vol. 10, Hydrologic Engineering Center, US Army Corps of
Engineers, Davis, California.
Davis, S.N. and De Wiest, R.J.M. (1966) Hydrogeology, John Wiley and Sons,
New York.
674 Vadose zone monitoring
Driscoll, F.G. (1986) Groundwater and Wells, (2nd edn), Johnson Division, St.
Paul, Minnesota.
Everett, L.G. (1980) Groundwater Monitoring, General Electric Co. Technology
Marketing Operations, Schenectady, New York.
Everett, L.G., McMillion, L.G. and Eccles, L.A. (1988) Suction lysimeter
operation at hazardous waste sites, in Groundwater Contamination - Field
Methods, (eds A.G. Collins and A.I. Johnson), ASTM, STP-963,
Philadelphia, Pennsylvania.
Everett, L.G. and Wilson, L.G. (1983) Unsaturated Zone Monitoring at Hazardous
Waste Land Treatment Units, Guidance Manual to EPA, Office of Solid
Waste, Washington, DC
Everett, L.G., Wilson, L.G., and Hoylman, E.W. (1984) Vadose Zone Monitoring
Jor Hazardous Waste Sites, Noyes Data Corp., Park Ridge, New Jersey.
Everett, L.G., Wilson, L.G. and McMillion, L.G. (1986) Suction Lysimeter
Operation at Hazardous Waste Sites, ASTM, Cocoa Beach, Florida.
Freeze, RA. and Cherry, J.A. (1979) Groundwater, Prentice-Hall, Englewood
Cliffs, New Jersey, 604pp.
Gardner, W.H. (1965) Water Content, in Methods oj Soil Analyses, (ed. CA.
Black), Agronomy No.9, Amer. Soc. Agron., Madison, Wisconsin, pp.
82-125.
Hern, S.C and Mellencon, S.M. (1986) Vadose Zone Modeling oj Organic
Pollutants, Lewis Publishers, Chelsea, Michigan.
Hillel, D. (1971) Soil and Water Physical Principles and Processes, Academic Press,
New York.
Keller, B.R, Everett, L.G. and Marks, RJ. (1989) Ejects oj Access Tube Material
and Grout on Neutron Probe Measurements in the Vadose Zone. Ground Water
Monitoring Review, NWWA, Fall Issue, Dublin, Ohio.
Klute, A. (1965) Laboratory measurement of hydraulic conductivity of
unsaturated soil, in Methods oj Soil Analyses, (ed. C.A. Black), Agronomy
No.9, Amer. Soc. Agron., Madison, Wisconsin, pp. 253-61.
Kraijenhoff van deLeur, D.A. (1962) Some Effects of the Unsaturated Zone on
Nonsteady Free-Surface Groundwater Flow as Studied in a Scaled
Granular Model, J. Geophys. Res., 67(11), 4347-62.
Law Engineering and Testing Co. (1982) Lysimeter evaluation, report to
American Petroleum Institute.
Luthin, J.M. and Day, P.R (1955) Lateral Flow Above a Sloping Water Table,
Soil Sci. Soc. Amer. Proc., 19, 406-10.
Meinzer, O.E. (1942) Groundwater, in Hydrology, (ed. Oscar E. Meinzer),
Dover Publications, New York, 385-477.
Mercer, J.W., Rao, P.S. and Wendell Marine, I. (1983) Role oj the Unsaturated
Zone in Radioactive and Hazardous Waste Disposal, Ann Arbor Science
Publishers, Ann Arbor, Michigan.
Parizek, RR and Lane, B.E. (1970) Soil-water sampling using pan and deep
pressure-vacuum lysimeters, J. oj Hydrology, 11, 1-21.
Scott, V.H. and Scalmanini, ].c. (1978) Water Wells and Pumps: Their Design,
Construction, Operation and Maintenance, Bull. 1889, Div. Agric. Sci., U. of
California.
Soilmoisture Equipment Corporation (1978) Operating instructions for the
Model 1900 Soil Water Sampler, Santa Barbara, California.
Soilmoisture Equipment Corporation (1983) Internal memo on soil tension,
Santa Barbara, California.
Todd, D.K., Tinlin, R.M., Schmidt, K.D. and Everett, L.G. (1976) Monitoring
Groundwater Quality: Monitoring Methodology, EPA-600/4-76-026, USEPA,
 References 675
Environmental Monitoring and Support Laboratory, Las Vegas, Nevada.
US Environmental Protection Agency (1986) Permit Guidance Manual on
Unsaturated Zone Monitoring for Hazardous Waste Land Treatment Units, EP AI
530-SW-86-040, OSW&ER, Washington, DC.
van Savel, C.H.M. (1963) Neutron Scattering Measurement of Soil Moisture:
Development and Current Status, Proc. Int. Symp. Humidity and Moisture,
Washington, DC., 171-84.
Wood, W.W. (1973) A technique using porous cups for water sampling at any
depth in the unsaturated zone, Water Resources Res., 9(2), 486-8.
 Index

Above-ground landfill 101 Below-ground landfill 101

Adsorption isotherm 41-3 Bentonite 434-5
Advection 33-4 Biodegradation 562-6
dispersion equation 40-1 Biogeochemical testing 26
VOC 399-400 Bioremediation
Aerobic performance criteria 522-4
bacteria 561-2 processes 526
biodegradation 408 Biotransformation 565
decomposition 118-19 Bodied solvent 178
oxidation 562 Borehole support 366
Air Breakthrough curve 47
permeability 416 Buried waste 353-5
rotary drilling 367-8
stripping 514 Caliper log 345
Anaerobic Capillary
acid decomposition 119 fringe 655
bacteria 116 zone 69
decomposition 116-18 Caps for cutoff walls 452-3
Aqueous-based soil wash systems Carbon adsorption 514-15
586 Case history
Aqueous solubility 533 chemical transport 61-2
Aquifer in situ bioremediation 554-6
alluvial 85 leachate production 133-4
confined 82 soil vapor extraction 420-3
fractured 89 soil washing 587-8
homogeneity 79-80 water balance in landfills 222-8
types 81-2 Casing advancement 374
unconfined 82 Cellulose 115
Aquitard 90 Cement-bentonite wall 444-7
Area ratio 375 Chemical loss rates 535-6
Auger Chlorinated hydrocarbons 577-8
drilling 359-63 Chromium 17-19
sizes 361, 363 Clay liner 139-50, 479
Bentofix 151-5
Bailers 643-4 Bentomat 151-9
Barometric pressure effects 387 chemical attack 149-51
Batch test 22-4 Claymax 151-9
678 Index

Clay liner (contd) Cyclone 275-6

compaction requirements 139-40,
146-8 Daily cover 97
construction 143-6 Deep soil mixed cutoff walls 452
diffusion through 56 Degree of saturation 68
freeze-thaw effects 145-6 Denison barrel 371-2
Gundseal151-9 Denitrification 565, 576
in cover 475 Density gradient of VOC 384-5
materials 140-4 Diffusion 34-5
protection 145 coefficient in soil gas 389-90
quality control 145-6 Diaphragm walls 448-9
soil Differential settlement 472-4
compaction 144-5 Discharge 81
placement 144 Dispersivity 38-9
processing 143-4 Distribution coefficient 23, 382
strength 254-5 DNAPL
test pads 148-9 definition 13
Cleaning sampling equipment 377 vapors 388
Climate 234 Double-pump system 509-10
Coal 284 Downhole
Cobbles 457,461-2,465 flux chamber 392
Column tests 24-6 geophysical measurement 336-46
Compaction requirements 139 Drainage length 236-8
Compactive energy 139-41 Drilling
Compounds 379-423 fluid removal 639-40
Composite liner 105-9 mud 364-5
Compo sting 545-6
Concentration gradients 388
Cone of depression 499 Effective porosity 50, 68
Contact adhesive seams 178 Effective stress 246-8
Contaminant plumes (geophysical Effluent disposal options 517-18
detection) 349-53 Electromagnetic methods 323-9
Core Environmental geotechnology history
barrels 369-70 1-8
drilling 368-73 Ethylene interpolymer alloy 168
Coring Evapotranspiration 220-1
bit sizes 368 Exhaust treatment 418
bits 370-1 Extrusion welding 177
Counter current decantation 586
Coupled flow processes 36 Fermentative bacteria 123
Cover system 104 Field capacity 74
barrier layer 471-96 Fick's law
basic concepts 455-6 first 35
components 457-8 second 46
drainage layer 465-70 Filter
gas collection layer 482-5 clogging 200-1
protection layer 465 design 199-202
surface layer 457-65 pack 507
Cyanide 18-19 Flame ionization detector 394
 Index 679

Flexible membrane liner, see 478,480

Geomembrane liner applications 158-60
Fluid conductivity 345 comparison with compacted clay 160
Flux hydraulic conductivity 155-8
advective mass 34 overlap 152
diffusive mass 35 shear strength 158-60
dispersive 38 types 151-2
example 55-6 Geotextile
total mass 39 drain 468
water 33 filter 467
Fractionation of vapors 419 Gravimeter 333
Fracture porosity 68 Grid spacing 403
Freundlich Ground penetrating radar 320-2
equation 23-4 Ground water
isotherm 43 definition 67
flow regimes 80-90
Gamma-gamma (density) log 338 treatment systems 511-18
Gas Ground water monitoring 607-50
characteristics 124-8 phased approach 27-30
chromatograph 390 see also Monitoring well
collection 208-12
production 124-5 Hazardous waste 126
vents 211 definition 9-10
Geochemical attenuation 16-20 generation rates 7-8
capacity 19-20 Head
cation exchange 17 elevation 69
clay 17 hydraulic 69
examples 17-20 pressure 69
laboratory testing 20-6 Headspace method 392
Geochemical processes 16-17 Health and safety considerations 378
Geocomposite 191 Heap leach 281-3
drain 470 HELP model 215-22
Geomembrane liner 475, 479 Hemicellulose 116
biological resistance 172 Henry's Law 380-1
chemical resistance 171-2 Hollow-stem auger 359-63, 635-6
construction quality control 176-82 Hot
design 172-6 air method 177
durability 179, 183 shoe 177
history 164-6 wedge 177
leakage fraction 238-9 Hydraulic
polymer types 168-9 conductivity 71, 83
stress cracking 171 gradient 77
tensile properties 170-1 Hydrodynamic dispersion 39
test methods 170-6 Hydrogen peroxide 557
thickness 170 Hydrostratigraphic unit
thermal properties 172 aquifer 73
uses 166-7 aquifuge 73
Geonet 191, 203, 469 aquitard 73
Geosynthetic clay liner 105, 151-60, confining unit 73
680 Index

Ice-contact deposits. 86 Methane 115-23

Impoundment '4-6 gas collection and removal system
Induction log 342-3· 208-12
Infiltrability 74 production 128-31
Infiltration 71, 218-20 phase 120-3
capacity 74 yield 124-5
rate 74 Methanogenic biodegradation 577
Inorganic contaminants 590 Methanogens 116
Inside clearance ratio 375 Microbial processes 560-6
Interface friction 171 Micro-gravity 332-3
Interfacial shear 109, 246 Microbiological degradation 525
Intergranular 67 Microorganism 527
flow 84-7 Microtox 528
Intimate contact 107 Mine waste 269-71
Intrinsic permeability 71 Moisture characteristic curve
Isotherm 22-4 definition 75
Isotropy 79-80 hysteresis 76
Moisture
Land treatment 540-9 content 68
Landfill retention curves 659
disposal unit 110 Monitoring well
operations 110-11 annular seal 629-35
microbiology 115-23 bentonite seals 629-32
Langmuir isotherm 43 components 608-10
Lateral drainage design 611-35
layer 233 development 639-43
percolation 235 diameter 615
Leachate 99 drilling methods 635-9
analysis 126 equipment decontamination 648-9
characteristics 125-6 filter pack 623-9
concentration 60 grouts 632-5
definition 14 materials 611-15
flux 61 placement 610-11
recirculation 111 sampling 643-6
recycle 129 screen 616-21
Leak detection system 202-5 sediment sump 621-2
Liner leakage fraction 238-9 surface completions 646-8
Liner requirements 103-4 Mud
LNAPLs, definition 13 cake removal 640-1
Local flow system 80 rotary drilling 636-7
Lysimeter 662-9 Municipal refuse
composition 114-15
Magnetometry 334-6 methane potential 115
Manhole 199
Manning formula 196-7 N value 377
Manual pumps 645-6 NAPLs definition 13
Mechanical dispersion 36-9, 51 Natural clay liner 4-6, 137-9
Melt bonding 177 Natural soil liner 102
Metal detection 334 Neutron moderation method 668-9
 Index 681
Neutron-neutron log 342 effective 68, 83
porosity log 342 fracture 68
NIMBY 100 general 67
Non-aqueous-phase liquids, intergranular 67
definition 13 primary 67
Non-nuclear logs 342-5 secondary 67, 88
Non-reactive solutes 33 Portable gas chromatographs 394
Nuclear logs 338-42 Potentiometric surface 82
Nutrients 530 Precipitation reactions 17
Pressure gradients of VOC 387-8
Pressure/vacuum lysimeter 26-7
Observational method 306-9
Primary leachate collection system 188
Octanol, water partition coefficient
removal 194-204
383,532
Primary porosity 67
Organic
Pumping system design 509-11
carbon 45
carbon partition coefficient 45
Radioactive chemical 45
compound adsorption 45
Radius of influence 418
contaminants 589
Reactive solutes 41
vapor analyzer 394
Recharge 71-2, 81
Outwash 86
Record keeping 518
Oxygen
Records of decisions 521
delivery 556
Recovery
demand 570-1
ratio 376
supply 570-1
well 505-8
Redox potential 530
Particle size fractionation 585-6 Refuse
Partition coefficient 41, 382 composition 251
Partitioning of VOC 380-3 decomposition 131-3
Paving 463 Regional flow system 80
material 457 Regulatory requirements 8-10
Percussion drilling 373-4 Remedial
Permittivity 193 action 298,304-5
Photoionization detector 394 design 298,304-5
Pipe 195-7 feasibility study 298, 302-4
risers 199 investigation 298,301-2
Pitcher barrel 372 planning 290-8
Plastic concrete 448 Resistance log 344
Pneumatic pumps 644 Resistivity log 343-4
Polyethylene Resistivity methods 323-9
chlorinated 168 Retardation 48
chlorosulfonated 168 factor 40-1
high density 168, 283 Richards Equation 75
very low density 168, 283 Risk assessment 523
Polyvinyl chloride 168, 283 Rock
Pore-liquid monitoring 660-2 definition 66
Pore-liquid sampling equipment fractured 87
662-4 sedimentary 87
Porosity Rotary drilling 362-8
682 Index

Rotary percussion drilling 373 Solid stem auger 359

Runoff coefficient 216 Solid waste
characteristics 12-13
Sample density 313-14 definition 11
Sanitary landfill 97-8 Solvent
Seaming of geomembrane adhesive seams 178
bodied 179 seams 178
extrusion Sorption 533-5
fillet 179 Specific discharge 78
flat 179 Specific storage 78
hot air 179 Split
hot wedge 179 barrel sampler 376
solvent 179 inner tubes 370
adhesive 179 spoon sampler 377
Secondary leachate collection system Spontaneous-potential log 344
188 Stability analysis
Secondary leachate collection and charts 257-9
removal system 202-5 cover 262-5
Secondary porosity 67 methods 254-66
Seepage velocity 33-4, 50, 78 seismic effects 265-6
Seismic Standard penetration test 377
reflection 329-32 Steel sheet piling 447-8
refraction 329-32 Strength of refuse 250-3
Selenium 17-19 Submersible electric pumps 644-5
Sequential batch-contact testing 20-2 Suction 69
Shear strength of waste 252-3 ground probe 393-4
Shelby tube 375 Surface
Single-pump system 510-11 accumulator method 393
Siting of landfills 99-100 flux chamber 392
Slope length 236
Slope stability failure Tailings impoundments 274-81
criterion 248-50 TCLP 10-11
modes 245-6 Temperature log 344
Slurry Thermal
phase processes 547-9 fusion 177
trench 434-7 gradient of VOC 385-7
wall construction 440-2,446 Thornthwaite method 216-17
Soil Thin-walled tube samplers 375-6
bentonite backfill 434-7 Till 86
definition 66 Topsoil 231-3, 459
gas surveys 390-414 Tortuosity 35
sampling Total stress 246-8
core barrels 371-2 Toxici ty 527-9
equipment 374-7 Toxicity characteristic leaching
wash procedure 10-11
application 593-4 Transit time 34, 51
processes 594-603 determination 60-1
treatment 589-91 Transmissivity 192
zone 653 Transport processes
 Index 683

advection 33-4 transport 380-90

coupled flow 36, 39 Vegetation 230-1
diffusion 34-5 Vegetative growth 222
mechanical dispersion 36-8 Vertical cutoff walls
Transport of voe vapor 384-90 applications 431-2
Treatability studies 572-4 types 432-3
Trophic groups 116 Vertical drains 452
Turbidity reduction 641-2 Vibrating beam cutoff walls 449-51
voe (volatile organic compound) 379
Ultrasonic seaming 178 Volatilization 535
Uncertainties in waste remediation
291-7 Waste rock 271-4
Underground storage tank Water
monitoring 671-3 balance calculations 215-22
Utilities 353-5 table 82
Well
Vadose zone field 499-501
description 653-7 development 508
flow regimes 656-7 interference 499
Vapor recovery system 501
extraction 414-23 Wireline core barrel 370