Lecture Notes in Civil Engineering

Ranjith Dissanayake
Priyan Mendis Editors

ICSBE 2018
Proceedings of the 9th International
Conference on Sustainable Built
Environment
Lecture Notes in Civil Engineering

Volume 44

Series Editors
Marco di Prisco, Politecnico di Milano, Milano, Italy
Sheng-Hong Chen, School of Water Resources and Hydropower Engineering,
Wuhan University, Wuhan, China
Ioannis Vayas, Institute of Steel Structures, National Technical University of
Athens, Athens, Greece
Sanjay Kumar Shukla, School of Engineering, Edith Cowan University, Joondalup,
WA, Australia
Anuj Sharma, Iowa State University, Ames, IA, USA
Nagesh Kumar, Department of Civil Engineering, Indian Institute of Science
Bangalore, Bangalore, Karnataka, India
Chien Ming Wang, School of Civil Engineering, The University of Queensland,
Brisbane, QLD, Australia
Lecture Notes in Civil Engineering (LNCE) publishes the latest developments in
Civil Engineering - quickly, informally and in top quality. Though original research
reported in proceedings and post-proceedings represents the core of LNCE, edited
volumes of exceptionally high quality and interest may also be considered for
publication. Volumes published in LNCE embrace all aspects and subfields of, as
well as new challenges in, Civil Engineering. Topics in the series include:
• Construction and Structural Mechanics
• Building Materials
• Concrete, Steel and Timber Structures
• Geotechnical Engineering
• Earthquake Engineering
• Coastal Engineering
• Hydraulics, Hydrology and Water Resources Engineering
• Environmental Engineering and Sustainability
• Structural Health and Monitoring
• Surveying and Geographical Information Systems
• Heating, Ventilation and Air Conditioning (HVAC)
• Transportation and Traffic
• Risk Analysis
• Safety and Security
To submit a proposal or request further information, please contact the appropriate
Springer Editor:
– Mr. Pierpaolo Riva at [email protected] (Europe and Americas);
– Ms. Swati Meherishi at [email protected] (India);
– Ms. Li Shen at [email protected] (China);
– Dr. Loyola D’Silva at [email protected] (S-E Asia and Australia/NZ).

Indexed by Scopus

More information about this series at http://www.springer.com/series/15087

Ranjith Dissanayake Priyan Mendis
•

Editors

ICSBE 2018
Proceedings of the 9th International
Conference on Sustainable Built Environment

123
Editors
Ranjith Dissanayake Priyan Mendis
Department of Civil Engineering Department of Infrastructure Engineering
University of Peradeniya University of Melbourne
Kandy, Sri Lanka Parkville, VIC, Australia

ISSN 2366-2557 ISSN 2366-2565 (electronic)

Lecture Notes in Civil Engineering
ISBN 978-981-13-9748-6 ISBN 978-981-13-9749-3 (eBook)
https://doi.org/10.1007/978-981-13-9749-3
© Springer Nature Singapore Pte Ltd. 2020
This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part
of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations,
recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission
or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar
methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are exempt from
the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the
authors or the editors give a warranty, expressed or implied, with respect to the material contained
herein or for any errors or omissions that may have been made. The publisher remains neutral with regard
to jurisdictional claims in published maps and institutional affiliations.

This Springer imprint is published by the registered company Springer Nature Singapore Pte Ltd.
The registered company address is: 152 Beach Road, #21-01/04 Gateway East, Singapore 189721,
Singapore
Preface

It is with great pleasure that we present the proceedings of the 9th International
Conference on Sustainable Built Environment (ICSBE) 2018. This is the ninth
consecutively organized conference following a series of International Conference
since 2010, keeping its tradition of adhering to engineering excellence.
Taking a step forward from the last seven events, the coverage of specialty areas
in this conference has been diversified. This book contains the manuscripts of
research work from many different sub-specialties. We expect that all these
manuscripts will be presented at parallel sessions from 14 to 15 December 2018.
We would like to express our appreciation to all keynote lecturers for their
invaluable contribution to the development of a sustainable world. We are also very
grateful to the authors for contributing research papers of high quality. The
manuscripts in this proceeding book have been reviewed by a panel of academic
and professional experts who have vast expertise in their respective fields. The
enormous work carried out by these reviewers is gratefully appreciated as well. We
are also pleased to acknowledge the advice and assistance provided by the members
of the international advisory committee and members of the editorial committee
along with many others who volunteered to assist to make this very significant
event a success. Furthermore, we acknowledge the financial sponsorship provided
by many organizations that has been extremely supportive towards the success of
this international conference.
It is the earnest wish of the editors that this proceeding book would be used by
the research community and practicing engineers who are directly or indirectly
involved in studies related to sustainability.

December 2018 Ranjith Dissanayake

Priyan Mendis
Editorial Committee

v
vi Preface
Contents

Climate Change and Sustainable Transportation

Important Aspects of Evacuation Planning for the Coastal
Communities in Sri Lanka . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
G. P. Jayasiri, O. P. C. Randil, G. M. C. A. Perera, C. S. A. Siriwardana,
P. B. R. Dissanayake, and C. S. Bandara
Application of GPS/GIS Based Travel Mode Detection Method
for Energy Efficient Transportation Sector . . . . . . . . . . . . . . . . . . . . . . 11
G. W. H. A. Chandrasiri, K. A. G. G. K. Kumarasinghe,
and H. K. Nandalal
Selection of Optimum Junction Operation Strategy
for Gatambe Intersection Using VISSIM Simulation . . . . . . . . . . . . . . . 22
K. L. D. Maduranga, R. G. N. Yasamali, I. M. S. Sathyaprasad,
and H. U. Weerakoon
Gap Assessment of Warning and Dissemination Process of Early
Warning System in Coastal Areas of Sri Lanka . . . . . . . . . . . . . . . . . . . 36
H. M. S. S. Hippola, E. M. S. D. Jayasooriya, G. P. Jayasiri,
Chameera Randil, Chamal Perera, K. K. K. Sylva, A. K. Kulathunga,
C. S. Bandara, C. S. A. Siriwardena, and P. B. R. Dissanayake
Gaps in the Evacuation of People with Special Needs During
Coastal Disasters in Sri Lanka . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
E. M. S. D. Jayasooriya, H. M. S. S. Hippola, C. S. Bandara,
A. K. Kulathunga, K. K. K. Sylva, P. B. R. Dissanayake, G. P. Jayasiri,
Chameera Randil, Chamal Perera, and C. S. A. Siriwardana

Community Environment and the Ecosystem

Multifunctional Landscapes for Urban Flood Management – A Study
with Reference to the Kotte Wetland . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
D. N. Denipitiya and S. Udalamaththa

vii
viii Contents

Single and Binary Adsorption of Heavy Metal Ions from Aqueous

Solutions Using Coirpith as the Adsorbent . . . . . . . . . . . . . . . . . . . . . . . 75
E. N. Ranasinghe and B. M. W. P. K. Amarasinghe
Sustainable Autarky of Food-Energy-Water (Safe-Water) . . . . . . . . . . . 91
N. Khandan, D. Tchinda, S. M. Henkanatte-Gedera,
I. S. A. Abeysiriwardana-Arachchige, H. M. K. Delanka-Pedige,
S. P. Munasinghe-Arachchige, and Y. Zhang
Study of Evidences on Chronic Kidney Disease Due to Unknown
Etiology Based on Environmental, Social, Economic and Health
Patterns of Selected Population . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
S. M. Rajapakshe, S. K. Weragoda, T. Kawakami,
and W. B. M. L. I. Weerasekara
Algal Wastewater Treatment: Study of Polyculture in Fed-Batch
Mode in a Winter Climate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 107
I. S. A. Abeysiriwardana-Arachchige, W. Chapman,
and N. Nirmalakhandan
Comparative Study of Spatial and Temporal Variation of Drought
Using Remotely Sensed Data - A Case Study for Kirindi Oya Basin . . . 116
S. U. Hendawitharana, M. K. D. D. Priyasad, and R. L. H. L. Rajapakse
Potential for Formation of Trihalomethane in Diverted
and Non-diverted Areas of Mahaweli River in Sri Lanka . . . . . . . . . . . 131
H. R. L. C. Bandara, W. B. M. L. I. Weerasekara, and S. K. Weragoda
Studying About the Manganese Phase Diagram When Manganese
Is Reacting with Calcium Hypochlorite . . . . . . . . . . . . . . . . . . . . . . . . . 137
S. A. Abeysinghe, S. K. Weragoda, and W. B. M. L. I. Weerasekara
Watershed Characterization of Demodara Reservoir with Special
Attention on Soluble Manganese Concentration . . . . . . . . . . . . . . . . . . . 147
W. G. Chathura Sanjeewa, R. A. C. B. Ranaweera, Ruwan Liyanage,
W. B. M. M. W. Wijekoon, and G. P. R. Ireshika
Application of Safety Plans for Rainwater Harvesting Systems
in Sri Lanka . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 162
K. A. I. L. Rathnayake, S. A. C. Upeka, K. A. M. Kularathne,
P. G. K. B. Ranathunga, T. D. Jayasinghe, W. B. M. L. I. Weerasekara,
and S. K. Weragoda
Hardness and Fluoride Removal Efficiency of Groundwater:
Use of Physiochemical Methods in Water Purification
at Anuradhapura District Dry Zone in Sri Lanka . . . . . . . . . . . . . . . . . 171
K. M. S. Kumara, K. A. M. Kularathne, P. G. K. B. Ranathunga,
S. A. C. Upeka, D. C. Herath, and S. K. Weragoda
Contents ix

Land-Use Planning Model: A Case Study on Transit-Oriented

Development (TOD) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
G. A. M. A. S. B. Udapitiya, R. M. T. Rathnayake, and K. K. K. Sylva
Effective Sizing of Natural Systems for Grey Water Treatment
Using Locally Derived Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193
S. Karunaratne
Sustainability via ERP and BIM Integration . . . . . . . . . . . . . . . . . . . . . 202
F. S. T. Hewavitharana and A. A. D. A. J. Perera
A Comparative Study of Speed and Flow Measurements Methods
as Applied to Four Lane Dual Carriageway Roads . . . . . . . . . . . . . . . . 211
W. U. A. Lowe, H. S. A. Mendis, and I. M. S. Sathyaprasad

Green Technology and Energy

A Stochastic Differential Equation Based Wind Speed
Forecasting Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227
H. M. D. P. Bandarathilake and G. W. R. M. R. Palamakumbura
Experimental Investigation to Achieve Minimum Surface
Roughness in Wire EDM Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 239
K. P. M. Priyankara and G. I. P. Perera
Eco-Friendly Porous Carbon Materials for Wastewater Treatment . . . . 252
C. J. Thambiliyagodage, V. Y. Cooray, I. N. Perera,
and R. D. Wijesekera
Identification of Joint Efficiencies in 13 mm Finger Jointed
Timber Species Used in Sri Lanka . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261
C. K. Muthumala, Sudhira De Silva, K. K. I. U. Arunakumara,
and P. L. A. G. Alwis
Experimental Investigation of Top Hinged Flap Type Wave Energy
Device with a Bottom Gap – Preliminary Experimentations . . . . . . . . . 268
L. U. Bakmeedeniya and S. D. G. S. P. Gunawardane
Increasing the Efficiency of a Dye-Sensitized Solid-State Solar Cell
by Iodine Elimination Process in Hole Conductor Material . . . . . . . . . . 282
A. W. M. V. Ekanayake, G. R. A. Kumara, R. M. G. Rajapaksa,
and A. Pallegedara
Monitoring Volatile Fatty Acids in an Anaerobic Process
with Microbial Fuel Cell . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 288
H. M. K. S. B. Thennakoon, S. M. W. T. P. K. T. Ariyarathne,
and M. Danthurebandara
x Contents

Embodied Carbon Reduction Strategies for Buildings . . . . . . . . . . . . . . 295

L. M. T. Kumari, U. Kulatunga, N. Madusanka, and N. Jayasena

Sustainable Building and Infrastructure

Lessons to Learn and Challenges to Address in Resilient
Historic Towns . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 311
M. De Fino, E. Cantatore, A. Scioti, and F. Fatiguso
Pedestrian Wind Comfort Study Using Computational Fluid
Dynamic (CFD) Simulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 323
Sarah Fernando, Shiromal Fernando, and Priyan Mendis
Performance of Paving Materials in Outdoor Landscaping . . . . . . . . . . 340
G. H. Galabada and R. U. Halwatura
Determining Criticality of Performance Indicators
for a Construction Company . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 353
K. K. Tripathi, K. N. Jha, and A. K. Jain
Seepage in Iranamadu Dam and Its Influence on the Stability . . . . . . . . 365
T. Venuja and L. C. Kurukulasuriya
Application of Principle of Network Connectivity in Creation
of Sustainable Urban Form: Case of Negombo Water Fronts . . . . . . . . 378
D. S. N. Jayakody
Development of a Post-tentioned Masonry Retaining Wall System . . . . . 394
W. W. P. K. Perera
Roads for Sustainable Cities – What Is Wrong with Us? . . . . . . . . . . . . 411
K. S. Weerasekera
A Sustainable Green Façade Retrofit for An Iconic Built Structure:
A Case Study at SAINTGITS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 419
Jacob Thottathil Varghese, Sajan Thomas, and Joselin Herbert
Embankment Construction on Power Blender Cement Mixed Soft
Soil Treated Ground for CKE Bypass . . . . . . . . . . . . . . . . . . . . . . . . . . 433
N. Vasanthan, H. Ohashi, F. Obara, S. D. L. Tilak, M. C. W. Dissanayake,
and A. Karunawardena
Reinforced Concrete Bridge Deck Repair with Titanium NSM . . . . . . . 447
S. L. Platt, K. A. Harries, and M. J. McCabe
Physical Behavior of CFRP Retrofitted Reinforced Concrete
Slab-Column Connections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 458
R. H. M. Dissanayaka, M. A. L. Silva, L. P. G. Magallagoda,
and J. C. P. H. Gamage
Contents xi

Review on Fire Performance of Cellular Lightweight Concrete . . . . . . . 470

I. R. Upasiri, K. M. C. Konthesingha, K. Poologanathan,
S. M. A. Nanayakkara, and B. Nagaratnam
Numerical Modelling of Re-cracking Behaviour in Retrofitted
Crack Stop Holes in Steel Structures . . . . . . . . . . . . . . . . . . . . . . . . . . . 479
S. Abeygunasekera, J. C. P. H. Gamage, and S. Fawzia
Comparison of Properties of Masonry Walls Constructed
with Commonly Used Bricks and Non-standard Large Bricks . . . . . . . . 489
V. L. W. Seneviratne and D. Nanayakkara
Sustainable Maintenance of Steel Truss Bridges:
A Conceptual Framework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 499
Sudath C. Siriwardane

Sustainable Manufacturing
Properties of Portland Blast Furnace Slag Cements Affecting
Their Autogenous Shrinkage Behaviors . . . . . . . . . . . . . . . . . . . . . . . . . 513
S. Suntharalingam and Y. Takahashi
Fire Performance of CFRP Strengthened Steel I Beams Cured
at Elevated Temperature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 526
E. R. K. Chandrathilaka and J. C. P. H. Gamage
Treatment of Wastewater with High Nitrogen Loading by Moving
Bed Biofilm Reactor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 538
H. K. D. T. Karunarathne, L. K. U. De Silva, and K. B. S. N. Jinadasa
Experimental and Analytical Study of Residual Yield and Tensile
Strengths in Partially Corroded Structural Steel . . . . . . . . . . . . . . . . . . 550
P. A. K. Karunananda, K. A. D. D. T. Kariyapperuma, R. P. Weeraratne,
and H. D. M. Sanjeewani
Mathematical Modeling of the Hydrolysis Phase in Anaerobic
Digestion of Solid Waste . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 559
D. I. U. Edirisinghe, S. M. W. T. P. K. Ariyarathna, and C. S. Kalpage
Design of a Robotic Firefighting Machine . . . . . . . . . . . . . . . . . . . . . . . 567
R. Munzara, S. M. Ganyani, and T. Mushiri
Design of a Portable and Sustainable Bicycle Frame . . . . . . . . . . . . . . . 581
T. Mushiri, R. Gumbo, and C. Mbohwa
Design for Manufacture and Assembly of an Automated Dish
Washing Machine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 595
Tawanda Mushiri, Rutendo F. Solomon, and Charles Mbohwa
xii Contents

Integration of BIM and Construction Supply Chain Through

Supply Chain Management; An Information Flow Model . . . . . . . . . . . 604
A. P. Rathnasinghe, M. K. C. S. Wijewickrama, U. Kulatunga,
and H. S. Jayasena
Removal of Heavy Metals from Industrial Wastewater
Through Minerals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 615
U. A. S. L. Muthukalum, C. A. Gunathilake, and C. S. Kalpage
Author Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 633
Climate Change and Sustainable
Transportation
 Important Aspects of Evacuation Planning
for the Coastal Communities in Sri Lanka

G. P. Jayasiri1(&), O. P. C. Randil1, G. M. C. A. Perera1,

C. S. A. Siriwardana1, P. B. R. Dissanayake2, and C. S. Bandara2
1
University of Moratuwa, Moratuwa, Sri Lanka
[email protected]
2
University of Peradeniya, Peradeniya, Sri Lanka

Abstract. Sri Lanka is a tropical island in the Indian Ocean which has suffered
from various disasters in the recent past. Flood events which occurred in the last
three consecutive years clearly shows that the country is affected from climate
change. During a disaster event, evacuation of vulnerable and affected com-
munities is necessary to reduce the number of deaths. In the Sri Lankan per-
spective, most of the recent disasters severely damaged the coastal belt of the
country. This study was carried out to evaluate the different aspects of evacu-
ation planning with respect to the coastal districts in Sri Lanka. Some of the
aspects considered are early warning dissemination, evacuation routes, shelters,
drills and training and the effects of having a family vehicle, disabled people,
neighbours and domestic animals. During this study, an online questionnaire
survey was distributed after carrying out an in-depth literature survey to gather
data. The lack of trust in authorities who disseminate early warnings and limited
knowledge on evacuation routes and shelters in the residential areas are some of
the key issues which were identified during this study. Furthermore, it was
identified that social media plays a pivotal role in disseminating the early
warnings in the urban areas. Increasing the number of awareness campaigns
related to evacuation and using social media wisely are some of the recom-
mendations made, which will build the coastal community resilience via
proactive measures.

Keywords: Evacuation planning
Resilience
Early warning

1 Introduction

Extreme weather events and natural hazards have been identified as the top two global
risks in terms of likelihood (WEF 2018). When looking at Sri Lanka, it is visible that
this statement is highly applicable considering the frequent disasters after the 2004
Indian Ocean Tsunami. During 2016, 2017 and 2018, the country witnessed extreme
rains which triggered flooding in several urban areas. The number of affected people
from the 2018 floods were around 150,000 which shows a significant decrease com-
pared to the events in 2016 and 2017 (DMC 2018a, b). The 2017 flood event which

© Springer Nature Singapore Pte Ltd. 2020

R. Dissanayake and P. Mendis (Eds.): ICSBE 2018, LNCE 44, pp. 3–10, 2020.
https://doi.org/10.1007/978-981-13-9749-3_1
4 G. P. Jayasiri et al.

damaged both the infrastructure and communities of the densely populated districts in
Sri Lanka raised the number of affected families up to 630,000 (MoNPEA 2017a, b).
These three consecutive events show that Sri Lanka should undertake quick resilience
measures in order to reduce the number of affected people. Evacuation planning is an
aspect which should be given due recognition when building community resilience. Most
of vulnerable communities face several problems during the evacuation process after an
early warning is disseminated. There is always a set of individuals who will not respond to
the EW messages and evacuation orders following the guidelines, but will act on their
own (Sakamoto et al. 2016). Hence the respective authorities, in the Sri Lankan case, the
tri-forces sometimes must persuade them. The lack of awareness and trust in authorities,
unawareness on evacuation plans, shelters and shortest routes, minor experience in drills
and training, increasing number of deaths and deficiencies in capacity has led to the need
to carry out this research to find out the challenges and barriers in evacuation planning to
provide recommendations to reduce the disaster risk of the coastal communities. The
study is focused mainly on the evacuation scenarios of floods and Tsunamis since the
coastal belt of the island is more vulnerable to these hazards.

2 Methodology

To evaluate the different aspects in evacuation planning for the coastal communities in
Sri Lanka, initially, an in-depth literature review was carried out to identify the existing
status by referring to several plans and programmes prepared by various national and
Non-Governmental Organizations (NGOs). In addition, online databases like JSTOR,
Google scholar ad ScienceDirect were used to search for peer-reviewed articles across
the world. Combinations of keywords: evacuation planning, coastal hazards, emergency
management, disaster resilience and climate change were employed in the searches. The
preliminary materials were screened further to obtain a shortlist of materials relevant to
the research. These were used for snowball sampling to find further literature. Finally,
after a discussion with the Sri Lankan academics involved in the Disaster Management
sector, three main themes, namely; early warning dissemination, evacuation routes,
shelters, drills and training and the effect of having a family vehicle, disabled people,
neighbours and domestic animals were selected to identify gaps.
Then, an online questionnaire survey was distributed to gather the community
perception on evacuation planning in coastal communities. The Survey was distributed
via social media and emails. This was done to gather the opinions of the people who
have frequent internet access in the urban areas. Thematic content analysis method was
used to analyse the data under the above aspects. Finally, after the analysis, several
recommendations were provided to address the gaps and challenges identified in the
above themes.
 Important Aspects of Evacuation Planning for the Coastal Communities 5

3 Results and Discussion

3.1 General Information
From the online survey, 97 filled questionnaires were collected which led to the
assessment of the existing problems and challenges in evacuation planning. Most
responses were from the younger generation (age group: 15–30 years) as they are the
demographic that uses social media mostly.
Out of the 14 coastal districts, information was collected from 9 districts. Most of
the data is concentrated in the Western province and the Southern province as the
conclusions will be more focused on those two provinces. Most of the people who are
responded were from the engineering and higher education sectors. 34% were female
respondents. More than 70% of the families consist of 4 to 5 members and at least one
person has a mobile phone. Furthermore, when looking at the education level of the
families who responded; more than 90% of them have at least one member who has
diploma level or university level education. Facebook is the most commonly used
social network by the respondents. According to Fig. 1, when looking at the public
transport facilities and road conditions in the respective areas, the respondents were
satisfied with them.

Fig. 1. Road condition and quality of public transport according to respondents

Looking at the general opinions about hazards, more than 40% of the respondents
think that storms, floods and lightning strikes will affect their area while 15% were
confident that their area is hazard free. This fact confirms that most of the respondents
are aware that they are vulnerable to floods. When rating the trust placed on the
authorities towards the notices issued by the authorities to evacuate for disasters more
than 70% rate it below 3, on a scale of 1 to 5 (1 being worse and 5 being excellent).
This is a critical issue which needs to be addressed, which shows that communities do
not take evacuation notices seriously.
6 G. P. Jayasiri et al.

3.2 Early Warning Dissemination

The means of receiving Early Warnings (EW) also plays a pivotal role during evac-
uation. At present, Sri Lanka has a Multi-Hazard Early Warning Dissemination System
and an Early Warning Coordination Network from the national level down to the grass
root level (DMC 2015). In addition, the National Emergency Operations Plan (NEOP)
has identified the role of the media and the military for an effective EW dissemination
(DMC 2015). Analysis of the online survey shows that most respondents get early
warnings via Facebook, TV and Mobile phone SMSs as shown in Fig. 2.

Fig. 2. Way of getting EW messages

During a disaster, sometimes, access to the internet might not be possible. Fur-
thermore, a statistical analysis was done to identify the correlations between the above
responses. According to the Shapiro – Wilk normality analysis done to find a correlation
between the respondent’s district and social media, all the selected variables have a
significance less than 0.05, which depicts that the data significantly variates from a
normal distribution. From the analysis, it was visible that there aren’t any correlations
between the respondent’s districts and preferred way of getting EW messages. But it was
visible that there are weak positive correlations between Facebook, WhatsApp, Twitter
and other social media. This shows that a selected individual respondent has a slight
similarity on their opinion of social media. Furthermore, WhatsApp and media channels
have a weak negative correlation which shows a tendency in the respondents to prefer
either one of them to get EWs. In addition, from the District and Social Network cross
tabulation, it was observed that more than 80% of the respondents either use Facebook
or WhatsApp as a social media network. Hence, this shows that when using social media
to disseminate EW messages Facebook and WhatsApp can be more effective. The
Facebook page which was launched recently and is managed by the Disaster Man-
agement Centre shows that they have undertaken this initiative (DMC 2018b).
 Important Aspects of Evacuation Planning for the Coastal Communities 7

Maintaining the trust on authorities who give evacuation instruction is a challenge

(CEDMA 2017). When analysing the comments made by the respondents there were
some serious responses regarding the EW mechanisms. Some believe that announce-
ments by local authorities are reliable as those are issued right prior to the disaster, but
the predictions issued by Department of Meteorology (DoM) are inaccurate. During a
recent study, it was noted that DoM needs advanced technology to do local area
modelling which is needed to produce accurate results. (Jayasiri et al. 2018).
Untimed EW and unnecessary exaggeration were also noted as critical issues. Some
people felt that it is very important to respond to disaster warnings from the respective
authorities even though the level of trust is low, because otherwise, the loss will be
severe. Furthermore, one respondent said that Sri Lanka has improved surveillance and
other facilities regarding Disaster Management by introducing new communication
methods. It was also noted that the authorities may be paying special attention to areas
near the capital as well.
During the recent monsoon rains, several early warnings were disseminated via
social media which helped the authorities inform people about the upcoming disaster
risk. Respondents have also mentioned that conveying early warnings via Mobile
Phones and door to door methods are very effective while the use of newspapers to
distribute early warnings was not as effective. In addition, around 70% of the
respondents are capable of passing the early warning messages they get to more than 10
people.

3.3 Evacuation Routes, Shelters, Drills and Training

According to NEOP, Administrative heads of provincial, district, divisional and GN
levels, DMC coordinators and fire service departments of local authorities are some of
the responsible stakeholders for evacuation (DMC 2015). In addition, Non-
Governmental Organizations like Red Cross and Sarvodaya also come to aid during
the evacuation. Apart from the national level programmes, awareness and knowledge
on evacuation planning should be incorporated to the school level as well. More
participation of school children in local evacuation planning is necessary to increase
their preparedness (Vásquez et al. 2018). At present, the country has developed
guidelines for the school’s disaster safety which identifies the importance of evacuation
planning as well (DoE 2008). In the post-disaster recovery plans, the importance of
using schools as evacuation shelters was also identified (MoNPEA 2017b).
During the analysis, it was noted that 74% of the respondents have not undergone
any evacuation drills. Most of them were from Colombo, Galle, Gampaha, Kalutara
and Matara districts. 26% of the respondents who said that they have experience of
drills noted that it was mostly Tsunami evacuation drills, fire drills at their workplaces
and first aid drills.
Furthermore, almost every sector responded to having at least 10% of respondents
with evacuation drill experience as in Fig. 3.
8 G. P. Jayasiri et al.

Fig. 3. Respondents experience in evacuation drills

Knowing the safest and fastest ways to the evacuation shelters helps the residents to
get out of their houses and reach a more convenient place in time. In addition, iden-
tification of the safest place in the relevant area which can provide shelter is important.
During the survey, 37% of the people said that they know the evacuation routes while
others were either not sure or thought there aren’t any. Out of the 37% who said they
know about the evacuation routes, 74% firmly said that there are no evacuation signs
along the routes. Due to the significant damage caused by the 2004 Indian Ocean
Tsunami, most of the evacuation signs in the country are for Tsunamis. But for other
disasters, there are hardly any.
In addition, only 18% of the respondents know the evacuation shelters in the area.
This is a significantly a low value which the authorities should take immediate action to
remedy by conducting awareness programmes. Religious places, schools and hospitals
can be taken as evacuation shelters since most of them are on a higher ground.
According to the survey, respondents said that it will take around 5–10 min to reach the
nearest shelter in their areas. All of the 18% who said they know the evacuation shelter;
were confident that they know the shortest routes as well.
Furthermore, 60% of the respondents said that they don’t have any personal
experience of a disaster. Most of the remaining 40% have experiences related to the
Tsunami, strong winds and floods. More than 70% of them rated their evacuation
experience by giving a value greater than 3 within a scale of 1–5 where 1 being poor
and 5 being excellent.

3.4 Effect of Having a Family Vehicle, Disabled People, Neighbours

and Domestic Animals
Sometimes, having a family vehicle helps a speedy evacuation for residents except
during a flood. 81% of the respondents said that they have a family vehicle and most of
them were cars which have the capacity to carry around 5 five passengers. In com-
parison to the total number of family members of the respondents, only 60% of them
have the capability of carrying their entire family at once during an evacuation by their
vehicle.
 Important Aspects of Evacuation Planning for the Coastal Communities 9

During an evacuation, protective care of disabled people is necessary, supported by

advance preparation (Owens et al. 2013). Only 7% of the respondents had disabled
people and most of them need assistance in walking.
During an evacuation, looking after relatives and neighbours sometimes becomes a
priority rather than evacuating your own self. 80% of the respondents said that they will
put an extra effort to take their neighbours with them if they are not capable. 6% of the
respondents believed that they do not have enough space in their vehicles to carry them
while 12% believe that they are capable to evacuate by themselves.
People who have pets and other domestic animals (people with farms) tend to care
for them during an emergency. During this survey, the importance of this issue was
analysed and out of the respondents 47% had domestic animals and 78% mentioned
that they will put an extra effort to carry their pets during an evacuation while the others
believe that the animals can survive by themselves.

4 Conclusions and Recommendations

The analysis of the information led to several conclusions regarding the evacuation
planning for coastal communities in Sri Lanka. Even though Sri Lanka has a properly
established EW mechanism, the lack of trust in authorities who disseminate early
warnings is an issue. Social media plays an important role when informing the general
public about evacuation notices. DMC and other relevant authorities should consider
this fact and should use social media wisely for EW processes. If the authorities can
develop a mechanism where only vulnerable people get an evacuation notice to their
mobile phones as an SMS, that would be a perfect scenario. Limited knowledge and
awareness on evacuation routes and shelters in residential areas are some of the key
issues which were identified during this study. Local authorities should undertake
necessary actions to establish sign boards which will guide people to evacuation
shelters quickly. In addition, they can conduct smaller-scale awareness campaigns and
scenario-based evacuation drills. Most people who live in towns have a family vehicle
which helps them evacuate quickly. For those who need transportation assistance, local
authorities, with the help of NGOs, can arrange those facilities during an evacuation
process. Even though we should always care about our neighbours and other vulnerable
people we have to make sure that we are safe first.
Proper evacuation notices, well-established guidelines and clear signboards to
evacuation shelters will always help affected communities evacuate safely which will
reduce the number of deaths during a disaster.

Acknowledgement. This research was conducted in collaboration with the Faculty of Engi-
neering of University of Moratuwa and University of Peradeniya with the financial aid from
CABARET (Capacity Building in Asia for Resilience Education), a project of the European
Union’s Erasmus+ programme – Key action 2 – Capacity building in the field of higher
education.
10 G. P. Jayasiri et al.

Disclaimer. The European Commission support for the production of this publication does not
constitute an endorsement of the contents, which reflects the views only of the authors, and the
Commission cannot be held responsible for any use, which may be made of the information
contained therein.

References
Center for Excellence in Disaster Management and Humanitarian Assistance (CEDMA) (2017)
Sri Lanka disaster management handbook
Department of Education (DoE) (2008) National guidelines for school disaster safety. http://
www.preventionweb.net/files/9252_srinatguilines.pdf
Disaster Management Centre (DMC) (2015) National emergency operations plan, vol I
Disaster Management Centre (DMC) (2018a) Summary situation report - southwest monsoon
Disaster Management Centre (DMC) (2018b) DMC Sri Lanka. https://www.facebook.com/
dmc.gov.lk/
Jayasiri GP, Siriwardena CSA, Hettiarachchi SSL, Dissanayake PBR, Bandara CS (2018)
Important aspects in building community resilience of Coastal Districts in Sri Lanka. In:
International conference on disaster management
Ministry of National Policies and Economic Affairs (MoNPEA) (2017a) Floods and landslides.
Integrated risk assesment
Ministry of National Policies and Economic Affairs (MoNPEA) (2017b) Post-disaster recovery
plan
Owens JK, Warner Stidham A, Owens EL (2013) Disaster evacuation for persons with special
needs: a content analysis of information on YouTube. Appl Nurs Res 26(4):273–275
Sakamoto M, Kuri M, Iguchi M, Maki N (2016) Disaster governance in disaster management
planning—analysis of the evacuation planning process for Kuchinoerabujima Volcano
Eruption. J Nat Disaster Sci 37(2):105–117
Vásquez A, Marinkovic K, Bernales M, León J, González J, Castro S (2018) Children’s views on
evacuation drills and school preparedness: mapping experiences and unfolding perspectives.
Int J Disaster Risk Reduction 28(June 2017):165–175
World Economic Forum (WEF) (2018) The global risks report. ISBN 10.978-1-944835-15-6
 Application of GPS/GIS Based Travel Mode
Detection Method for Energy Efficient
Transportation Sector

G. W. H. A. Chandrasiri(&), K. A. G. G. K. Kumarasinghe,
and H. K. Nandalal

Faculty of Engineering, University of Peradeniya, Peradeniya, Sri Lanka

[email protected]

Abstract. The planning of efficient transportation systems is a central topic of

concern as it saves energy and contributes to economic development both at
national and household level. The development and optimization of highway
and transportation engineering models such as transportation networks, parking
facilities and as well as reduction of traffic congestion etc., require information
on travel modes. To overcome the burdens for the respondent and the other
drawbacks of the conventional methods related to travel mode detection, com-
bination of Global Positioning System (GPS) and Geographical Information
System (GIS) were tested in this research. After validating the accuracy of the
smartphone based ‘My Tracks’ GPS mobile application, GPS travel route with a
manually entered travel mode diary was used to collect data from randomly
selected days and students in two faculties in the University of Peradeniya. Out
of 125 datasets, 99 accurate datasets were selected for the analysis using
‘ArcGIS’ software coupling with ‘STATA’ statistical software. Rule-Based
algorithms and heuristic based methods are two main approaches that were used
to detect four travel modes (Walk, Bus, Motor Vehicles and Motor Bicycles).
Differentiation of Bus mode from the motor vehicles (except motor bicycles)
was the most challenging part of the research and a country specific novel
method was introduced in this research. Travel mode detection results were
compared with relevant correctly completed travel mode diaries to validate the
methodology. The overall travel mode detection accuracy of the new method is
96% for the accurate data sets compared to conventional travel mode diary
method. The success rate of collecting an accurate data set is 79%. The GPS/GIS
based travel mode detection could be applied in energy efficient development in
the transportation sector. In future research, the bus mode detection method
invented in this research could be used to detect driver behaviour.

Keywords: Travel mode
GPS
GIS
My tracks
University of Peradeniya

1 Introduction

An efficient transportation system is crucial for the development of a country. An

energy efficient transportation system contributes to the development of the economy
both at national and household level and the sustainability of a green environment and

© Springer Nature Singapore Pte Ltd. 2020

R. Dissanayake and P. Mendis (Eds.): ICSBE 2018, LNCE 44, pp. 11–21, 2020.
https://doi.org/10.1007/978-981-13-9749-3_2
12 G. W. H. A. Chandrasiri et al.

ecosystem. The detection of travel modes plays a key role while designing energy
efficient transportation system. Household travel surveys have been used as a tool since
1970’s to collect data to build transportation models such as transportation networks
and parking facilities (Zong et al. 2016). The Travel diary method is one of the
conventional methods to carry out travel surveys. It has many drawbacks as it is a more
time consuming and respondent burdening practice. Consequently, the use of Global
Positioning System (GPS) devices for collecting travel mode data became popular in
the recent years. It reduces the respondent burden while researcher uses the respondent
as a passive data collector. Though the GPS records accurate location, time, elevation
information, it cannot record details like travel modes and trip purposes. Therefore, in
this study Geographical Information System (GIS) was used with GPS data to detect
travel modes. Gong and Chen (2011) have studied GPS/GIS method for travel mode
detection in New York City using GPS devices. At present, smartphones have GPS
facility and therefore accuracy and applicability of GPS facility in the smartphones has
become an alternative to GPS devices-based travel mode detection method. However,
information on use of smartphones-based GPS in travel mode detection is scarce.
Therefore, a study has been designed to investigate the applicability of GPS/GIS
method using mobile GPS applications for travel mode detection in the University of
Peradeniya. The objectives of the study are; (i) to detect the travel mode using smart
phone-based GPS applications (ii) to compare the results of modal shares (iii) to
evaluate the success rate of smart phone-based travel mode detection compared to that
of travel diary-based travel mode detection method (iv) identify and compare the travel
modes of students in the Faculties of Engineering and Science using novel smartphone-
based travel mode detection method.

2 Background Information on GPS Based Travel Surveys

GPS based travel surveys have many advantages over traditional paper-based travel
surveys or travel diary method. Because, in traditional paper-based surveys, respon-
dents have to fill many pages manually. It is reported that respondents often forget to
enter the trip segment and it is a time-consuming procedure. In GPS based travel
surveys, respondents are acting as passive data collectors and therefore respondent
burden is comparatively lower. The GPS method shows the traveller’s exact route and
because of the accuracy of the data, sample size can be reduced. Moreover, travel
speeds can be calculated for the particular route at particular time period. It is a good
indicator to assess the level of service for a given transportation network. Due to the
easiness of the method, length of the survey can be extended from traditional single day
survey to multiple day survey. Accuracy can be validated for the daily routine of the
same respondent through this multiple survey. Though the advantages of the GPS
based travel survey are higher compared to traditional methods, it has few drawbacks
like any other method. Since GPS traces the respondent’s exact route, it minimizes the
privacy concern. Signal loss and degradations in shopping malls, tunnels, and urban
canyons are the other drawback in this method as well as the fact that GPS cannot
directly identify travel mode and trip purposes. In this research, the main focus is on
combining GPS and GIS methods to give a solution for travel mode detection.
 Application of GPS/GIS Based Travel Mode Detection Method 13

3 Literature Review

Shalaby and Chung (2005) developed a trip reconstruction software tool that auto-
matically identifies the travelled road links and modes used by the respondent using
GPS traces. Conventional GIS map matching algorithm was adopted in this method to
identify road links while rule based algorithms were used to identify four travel modes
(walk, bicycle, bus and passenger car). GPS travel data and the multimodal trans-
portation network model of downtown Toronto has been used for the evaluation of the
developed software tool. It is reported that said method correctly matched 79% of all
links travelled and 92% of all trip modes. However, they only allowed for one-purpose
trips and maximum of two possible modes per trip.
Zong et al. (2016) developed a hybrid procedure for travel mode detection and it
was presented using large-scale origin-destination survey conducted in Beijing in 2010.
They determined the subway mode separately from other modes and used two methods
to detect the subway mode. It is reported that 53.8% success rate for multinomial logit
model while 97.8% success was for GIS algorithm. Results revealed that the GPS/GIS
method provides higher accuracy for subway modes. Nested logit model was employed
to determine the modes of Walking, Bicycle, Car and Bus. Combined success rate of
the hybrid procedure was 86%.
Gong and Chen (2011) developed a GIS algorithm to automatically process data
that was traced from GPS based travel surveys. They have detected five travel modes
(Walk, Car, Bus, Subway and commuter rail) and those detected results from GIS
algorithm were compared against conventional travel diaries. They have conducted two
small GPS surveys and success rates of that combined survey was 82.6%. In their
research, they allowed an unlimited number of mode transfers within a trip. Further-
more, they built connectivity into their multimodal transportation network in GIS to
find the right link for matching the GPS traces, or interpolate the underground subway
mode. In their multi modal transportation network, the success rate for the subway
mode detection equals to 65.6%.
Clifford et al. (2008) used set of developed heuristic rules to determine both travel
modes and trip purposes. Heuristics based on speed and route of travel, as well as some
demographic information. In this study, four different travel modes (walk, bicycle,
private vehicle and public transport) have been considered. They compared results with
respective travel diaries.

4 Methodology

4.1 Study Area

The study area is University of Peradeniya, a state university in Sri Lanka. It has nine
faculties and about 11000 students.
14 G. W. H. A. Chandrasiri et al.

4.2 GPS Data Collection

Instead of handheld GPS loggers, smartphones with GPS facility were used for this
research. Several GPS tracking applications such as My tracks, Geo tracker, Track Me,
GPS Trip recorder were tested and ‘My tracks’ and ‘Geo tracker’ were selected as GPS
tracking applications based on the performance of said two applications. It was noticed
that internet facility is needed for tracking paths in Geo tracker occasionally and GPS
signal losses were higher than the My tracks application. Finally, ‘My tracks’ was
selected as the best tracking application. Accuracy of the mobile GPS application was
checked against a GPS receiver called etrex10. Obviously, the traced data points for a
given route was low in the mobile phone compared to the GPS receiver. But we
checked the accurate data points that were traced from the mobile phone. After three
consecutive days of pilot travel survey with two different mobile phones, we found that
the average accuracy of the traced data points of two mobiles phones were 88.6% and
93% respectively. Since the accuracy of the ‘My Tracks’ application is greater than
75%, it was selected for the study. It was noticed that the signal losses were occurred in
both etrex10 GPS receiver and the two GPS enabled mobiles phones in shopping malls
and the tunnels.

4.3 GIS Data Collection

In Sri Lanka, local agencies do not have transportation networks as GIS layers. As a
result, to get an initial idea whether the traced GPS data points are correct or not, first
travel routes were exported to the Google earth. Also, bus halts were identified using
Google earth as an indicator to separate bus mode from the other modes.

4.4 Socio Demographic Data Collection

Peoples’ mode choices are changed with their socio demographic information like age,
gender, vehicle ownership, location of residence etc. Therefore, before selecting par-
ticipants, questionnaire-based survey was conducted for total student population to
collect socio demographic data.

4.5 Sample Selection

The target population could be stratified by three main attributes, namely, (i) Respon-
dent type, (ii) Location of stay/residence wise and (iii) faculty wise. By respondent
category, population can be classified as students, academic staff and non-academic
staff. By location of stay/residence category, population can be classified as within the
university and out of the university. By faculty, population can be classified into nine
different faculties. The target group of this investigation was limited to the students of
the Faculties of Engineering and Science. The reasons for selecting those faculties to
this research was that the Faculty of Engineering is isolated from other faculties as well
as it is the faculty of the research team while the Faculty of Science is inside the
University premises and there is a higher number of students in that faculty. Academic
and non-academic staff were not selected for this sample because of their busy life
 Application of GPS/GIS Based Travel Mode Detection Method 15

style. Under location of stay category both students in hostels and residential students
were selected. The final sample of 100 students was selected using stratified random
sampling technique from both Science and Engineering faculties. Of these 100 stu-
dents, 79 students were selected for data collection as others have shown problems in
data recording using mobile application.

4.6 Key Steps in Methodology

4.6.1 GPX Data
In ‘My Tracks’, the GPS routes are saved as GPX files. In that file there is no velocity
data indicated directly. Velocity is the main parameter to separate travel modes. So that
first GPX file was exported to the ArcGIS and route was projected from WGS 1984 to
Kandawala, Sri Lanka grid. Example for a projected route is presented as Fig. 1.
Then X, Y coordinates were created in the attribute table. After that, attribute table was
exported to the MS Excel.

Fig. 1. Example for a projected route

4.6.2 Erroneous Data

After exporting the attribute table to the MS Excel, erroneous data were filtered by
elevation. Erroneous elevation data were represented as −9999. The accuracy level of
the data set was calculated using the Eq. 1 given below and data sets with accuracy
more than 80% were selected to detect travel modes. Table 1 represents the accuracy
results for ten datasets.

ða  bÞ  100
Accuracy ¼ ð1Þ
a

Where,
a = Traced data points
b = Erroneous data points
16 G. W. H. A. Chandrasiri et al.

Table 1. Accuracy results for ten datasets

Dataset Traced data point Erroneous data points Accuracy %
1 1409 97 93.1
2 489 145 70.3
3 1615 123 92.4
4 752 299 60.2
5 3403 18 99.5
6 1168 19 98.4
7 485 14 97.1
8 1267 11 99.1
9 343 201 41.4
10 954 4 99.6

4.6.3 Velocity Calculation

X and Y coordinates for each and every point in the route were calculated in the
ArcGIS. After exporting the attribution table with those coordinates, consecutive point
velocities can be calculated in MS Excel.

4.6.4 Travel Modes Calibration in STATA Statistical Software

STATA statistical software is used to detect travel modes. Commands were generated
according to local speed guide lines and set of data sets. Set of travel survey routes
from ‘My Tracks’ were used, because though the local speed limits are indicated, actual
level of service may be changed with the corresponding study area.
Data sets with calculated point velocities generated in Excel sheets were exported
to STATA statistical software. Primarily, data points were categorized into three modes
considering the velocities. Velocity ranges and corresponding travel modes are as
below.

Travel mode ¼ 0; Standstill; if Velocity  0:2 km=h

Travel mode ¼ 1; Walk; if 0:2 km=h\Velocity  8 km=h
Travel mode ¼ 2; Vehicle; if 8 km=h\Velocity  70 km=h

Then the data set was further categorized into 300 s intervals and average velocities
were taken in each 300 s intervals. Using those average velocities Vehicle mode was
categorized as below.

Travel mode ¼ 3; Bus; if 8 km=h\Average velocity  30 km=h

Travel mode ¼ 4; Motor Vehicle; if Average velocity [ 30 km=h

Most frequent travel mode for the points in the 300 s interval was used as the travel
mode in that 300 s interval. Then detected travel modes for each 300 s intervals were
plotted against time. Mode changes can then be identified from that plot. But the
drawback in the plotting method is that sometimes due to traffic congestions motor
 Application of GPS/GIS Based Travel Mode Detection Method 17

vehicles can be categorised from time to time in the bus category as presented in Fig. 2.
Therefore, to differentiate motor vehicles from bus category, a bus halt layer was
created in Google earth around the University of Peradeniya area as presented in Fig. 3
and points representing standstill, walk and vehicle in each dataset were separated in
ArcGIS. Then, that bus halt layer was exported to Arc GIS and 75 m buffer zones were
created around the bus halt. After that, standstill and walk points inside the buffer zones
were clipped as shown in Fig. 4. If more than 90% of standstill and walk points near
the bus halts coincide with the buffer zone, it was filtered as a Bus Mode. Filled
questionnaires were used to differentiate motor bicycles from the motor vehicles.
Finally, one day mode changes in each respondent were tabulated.

Fig. 2. Travel mode vs. time

Fig. 3. Bus halt layer around Kandy area

18 G. W. H. A. Chandrasiri et al.

Fig. 4. Clipped standstill and walk points

5 Results and Discussions

5.1 Travel Mode Detection
Analysis of data revealed that travel modes such as walk, bus, motor vehicle and motor
bicycles could be detected using the smart phone-based GPS application. It was found
that about 31% of the respondents use only one travel mode - for example walk mode –
in their daily travel. About 61% of respondents use 2 travel modes and 6% use 3 travel
modes in their daily travel (Table 2). The majority of respondents use Walk travel
mode plus any one of other travel modes in their daily travelling. But their mode
selections can be changed with several factors like distance, time, salary, level of
service of the transportation networks etc. In this study we mainly focused on the
applicability of smartphone-based method and the accuracy of the travel mode
detection.

Table 2. Number of travel modes used by respondents

No. of modes Freq. Percent Cum.
1 19 31.67 31.67
2 37 61.67 93.33
3 4 6.67 100
Total 60 100
 Application of GPS/GIS Based Travel Mode Detection Method 19

Table 3. Number of respondents and their selection of modes

Travel mode No of respondent
Walk 79
Bus 39
Motor vehicle 13
Motor bicycles 8

The result in Table 3 indicates that all the respondents had walked in their daily
routines. Among students in the Faculties of Engineering and Science, it was found that
the bus mode usage in the Faculty of Science is higher than that of Engineering
(Fig. 5). Common reason for that may be the distance from the location of stay to the
Faculty of Engineering is short compared to the Faulty of Science.

No of Respondents vs travel
modes
40

20

0
Walk Bus Motor Motor
Vehicle Bicycle

Faculty of Engineering Faculty of Science

Fig. 5. Comparison of travel modes used by responded of two faculties

5.2 Modal Shares

Modal share percentage per day was identified for each respondent.
Figure 6 shows the results of the comparison of modal shares. It indicates that walk
and bus combination is the highest share followed by walk and walk with motor
vehicle combination.
20 G. W. H. A. Chandrasiri et al.

Fig. 6. Modal share of 60 students in both faculties

5.3 Success Rate of Travel Mode Detection Using Smart Phone-Based

GPS Application
The travel modes detected using smartphone-based GPS/GIS method was compared
with the travel diaries of relevant responded. It was found that out of 79 dataset results
76 were matched with the relevant travel diary. Therefore, the success rate of detecting
travel modes in this research is 96%.
Among routes with more than 80% accurate data sets, overall travel mode detection
accuracy is 96% and the success rate of collecting accurate dataset is 79%.

6 Conclusion

The new method of smartphone-based GIS/GPS application could be used to detect

travel modes accurately. About 61% of the study population use two travel modes in
their daily travel. The new method has a success rate of 96% compared to the con-
ventional travel diary method. In this study location - University of Peradeniya - the
high share associated with the walk mode is because more than half of the respondents
are from university residences. Due to lack of owned vehicles, students had used buses
to complete their trips. Overall, this new method is accurate, faster and less respondent
burden and hence it can be used to detect travel modes in an institute or a city. Those
results can be used to develop transportation facilities (Access roads, Parking, Shuttle
services, etc.) in that area.
Travel planning using smart phone-based GPS/GIS application helps develop
energy efficient transport facilities and would therefore contribute to sustainable
management of green energy, environment and ecosystem services.

Acknowledgements. The authors wish to thank respondents who were willing to do the daily
travel survey, evaluation panel for their guidance and comments to complete this research
successfully.
 Application of GPS/GIS Based Travel Mode Detection Method 21

References
Clifford E, Zhang J, Stopher P (2008) Determining trip information using GPS data. Institute of
Transport and Logistics Studies, Sydney
Zong F, Yuan Y, Liu J, Bai Y, He Y (2016) Identifying travel mode with GPS data. Transp Plan
Technol 40(2):242–255
Gong H, Chen C (2011) A GPS/GIS method for travel mode detection in New York City.
Comput Environ Urban Syst 36(2):131–139
Shalaby A, Hwan Chung E (2005) A trip reconstruction tool for GPS-based personal travel
surveys. Transp Plan Technol 29(1):381–401
 Selection of Optimum Junction Operation
Strategy for Gatambe Intersection
Using VISSIM Simulation

K. L. D. Maduranga1(&), R. G. N. Yasamali1, I. M. S. Sathyaprasad1,

and H. U. Weerakoon2
1
Department of Civil Engineering, University of Peradeniya,
Peradeniya, Sri Lanka
[email protected]
2
Central Engineering Services (Pvt) Limited,
Digana Villege, Rajawella, Sri Lanka

Abstract. At present, traffic congestion is a major issue in the Kandy city area.
It is observed that most of the congestion occurs at junctions and intersections.
One of the most congested locations in the Kandy road network is the Gatambe
junction. Under this research, the aim is to simulate the Peradeniya – Gatambe
corridor using VISSIM simulation software and identify the optimum junction
improvement strategy for the Gatambe intersection. Gatambe intersection is a
four-way intersection that connects the Colombo – Kandy road (A1), William
Gopallawa Mw (AB42) and Gannoruwa road (B365) directly. At present, the
Gatambe intersection operates as a roundabout and at peak hours it is controlled
manually by traffic police. Because of the heavy upstream traffic congestion near
the Gatambe intersection, queues propagate upstream of the junction. Although
many solutions including those needing heavy capital investment have been
proposed to reduce congestion at Getambe junction, it is important that all
possible solutions are tested on a simulation model before any physical
implementation. In this study VISSIM is used as the simulation model. VISSIM
is a microscopic simulation model that can predict network performance at the
operational level, optimize control systems, and predict their behaviour before
implementation. For simulation the necessary geometric and traffic data were
collected as the first step. Then, the existing situation was simulated using
VISSIM and the results validated with the observed survey queue length results.
After that, the model was used to simulate three improvement strategies; sig-
nalization without geometric improvement, signalization with geometric
improvement and grade separation. According to the delay results, the benefit of
each alternative was estimated and a cost benefit analysis was conducted to
identify optimum junction improvement strategy.

Keywords: Microscopic simulation
PTV VISSIM

Optimum junction operation

© Springer Nature Singapore Pte Ltd. 2020

R. Dissanayake and P. Mendis (Eds.): ICSBE 2018, LNCE 44, pp. 22–35, 2020.
https://doi.org/10.1007/978-981-13-9749-3_3
 Selection of Optimum Junction Operation Strategy for Gatambe Intersection 23

1 Introduction

Traffic congestion is a major problem in many cities all around the world. Motorists
and passengers waste their time on the streets due to heavy traffic congestion and it
increases air pollution. In Sri Lanka this is a common issue in many cities. Kandy city
is severely impacted with the lack of road facilities. There are many identified diffi-
culties in expanding the existing road network in Kandy due to geographical aspects
and historical value. Kandy is located in a mountainous terrain 116 km away from
Colombo in the central province of Sri Lanka. It is recognized as the second busiest city
of the country. The Kandy city area road network consists of a few approaches and the
western approach is one of its main approaches. The road link, which directly connects
Kandy and Colombo, belongs to the western approach. In this research, the aim is to
simulate the road network to get an idea about the behaviour of traffic in the western
corridor. Data on traffic behaviour is obtained through a simulation model. When
considering the simulation, it is important to understand the concept, which is a set of
variables that contain sufficient information to describe the evaluation of the system
over a time period. Traffic simulation models are useful for both microscopic and
macroscopic perspectives. This simulation is done by VISSIM and it is a microscopic
simulation model. It is concerned with the individual elements of transportation sys-
tems, such as individual traveler behaviour and individual vehicle dynamics.

1.1 Microscopic Simulation

Transportation models can be generally classified into microscopic, mesoscopic,
macroscopic models. Microscopic models study individual elements of transportation
systems, such as individual vehicle dynamics and individual traveler behaviour.
Mesoscopic models analyse transportation elements in small groups, within which
elements are considered homogeneous. Macroscopic models deal with aggregated
characteristics of transportation elements, such as aggregated traffic flow dynamics and
zonal-level travel demand analysis. Microscopic simulation is related to small scale
operations and it is designed to run actual traffic operations as well. There are functions
based on the number of behavioural models such as car following model, lane changing
model and gap acceptance model. Microscopic simulation is best suited to evaluate
operational and management strategies to address congestion.

1.2 VISSIM
VISSIM is a part of the PVT software and it provides microscopic simulation methods.
It helps to assess and solve a wide range of transportation problems. VISSIM primarily
contains scientifically approved models for car –following, lane-changing and pedes-
trian moments. VISSIM consists of road users, public transport, pedestrians and their
interactions with each other. This network is used by cars, HGVs, buses, trains and
trams as well as cyclists and pedestrians. A wide selection of evaluations are available
for online and offline analysis. VISSIM is capable of animated visualization in 2D or
3D and it fills the space between technical expertise and non – technical audience.
24 K. L. D. Maduranga et al.

VISSIM can be used to analyse traffic of various alternative road designs which include
both urban and highway configurations.

1.3 Importance of Finding a Strategic Solution for Gatambe Intersection

Gatambe intersection is located at 7.270577 latitude and 80.6047636 longitude and is a
four-way intersection. This is connected with Colombo – Kandy road (A1), William
Gopallawa Mw (AB42) and Gannoruwa road (B365) directly. The Gatambe inter-
section is within 111–112 km. Though the Gatambe intersection is a roundabout
nowadays it is operated manually at peak hours using traffic police. Because of the high
congestion near Gatambe intersection the queue propagates both upstream and
downstream areas. So it has become a major issue in present.
Strategic solutions means a long-term solution. The solution should have the ability
to adjust with different variations of the parameters and situations. The importance of
finding a strategic solution for the Gatambe intersection is mainly to reduce the traffic
congestion. The Peradeniya Botanical Gardens are located near the Gatambe inter-
section and it has a high influence on the traffic congestion. Hence, the propagation of
the queue is rapidly increasing around it. There are two bridges near the Gatambe
intersection and it also influences congestion. If the road network has been blocked
near the Gatambe intersection, it will make sense to the congestion near Gannoruwa
Bridge, Railway crossing at William Gopallawa Mw, Dangolla junction and towards
the old Kandy road. Due to the heavy congestion, the travelers have to suffer from
delay, high fuel consumption and CO2 emission. There are many proposed solutions
for the reduction of congestion near the Gatambe intersection. It is essential to test the
most suitable solution through a simulation software before construction. That will help
to reduce expenses and save time by selecting most suitable solution (Fig. 1).

Fig. 1. Gatambe junction

 Selection of Optimum Junction Operation Strategy for Gatambe Intersection 25

2 Literature Review

The literature review was divided into two main topics. They are researches based on
VISSIM simulation models and based on selection of optimum junction improvement
strategy.

2.1 VISSIM Based Simulation Models

Dong (2013) analysed the road network of CBD in Beijing, China by using a VISSIM
based simulation and it was used to analyse the traffic influences in future years. Here,
four different traffic organization plans are compared. Since the driving behaviour in
Beijing are significantly different from the USA or Europe, behavioural studies are
conducted to calibrate parameters before building up the model. This paper uses travel
time, travel speed, queue length and delays as evaluation indicators to conduct the
comparisons. They used O-D matrices and some default values in VISSIM as input.
Natalia and Leszek (2017) intended to develop a simulation model to analyse
impact of heavy transport generated by the port to the city traffic. They analysed the
traffic conditions of individual and collective traffic. They considered peak hours when
collecting the data. They mainly focused on traffic patterns, changing of time varying;
pedestrian behaviour and impact of signalized systems. They used VISSIM simulation
software to simulate the model and analyse the results. The results were presented
based on time of travel and delay.
Heuristic algorithm based application tool is used to calibrate the microscopic
simulation parameters (Mohammad Rayeeduk Kalam 2017). This described that sim-
ulation parameters in VISSIM are based on three heuristic optimization algorithms.
This research explained that VISSIM calibration offers four objective functions such as
speed, flow, delay and multi-objective criteria and VISSIM has the ability to calibrate
all urban or rural driving behaviours. The results show that compared to other algo-
rithms, simultaneous perturbation stochastic approximation generally obtained an
acceptable set of parameters within a short period.
Papageorgiou (2006) conducted a research on a microscopic simulation model for
traffic management. It mainly presents the microscopic simulation model development
of a major highly congested traffic network in Cyprus. This paper explains about traffic
flow theory and how a traffic flow simulation model is developed for an effective
transport planning.
Kun and Lei (2007) presented a microscopic traffic emission simulation platform by
using VISSIM and the study shows that vehicles’ emissions are strongly dependant on
vehicles operation modes(acceleration). Further, it shows that optimization of signal
timing plans can improve the traffic operation as well as reduce the traffic emission.
Papageorgiou (2006) conducted a computer based approach that goes through
various stages which include problem identification, model objectives, model devel-
opment, model calibration, model validation, scenario preparation, simulation experi-
ments and simulated results evaluation. The proposed approach is applied in the case of
developing a microscopic traffic simulation model.
Brugmann et al. (2014) intended to develop a simulation model for real –world
microscopic traffic simulation. This model consists many elements such as road
26 K. L. D. Maduranga et al.

network data, several subject specific model families for vehicle types and turning
elements. VISSIM is a simulation tool for the design of traffic actuated control systems,
mainly used to analyse traffic of both urban and highway configurations.

2.2 Selection of the Optimum Junction Improvement Strategy

Yun and Ji (2013) studied the difference between the stop sign intersection and yield
sign intersection from the aspects of delay analysis. The traffic efficiency of the yield
sign intersection and the stop sign intersection has been explained by using VISSIM.
Huanyun and Fei (2013) presented a development of traffic safety evaluation
method based on simulated conflicts at signalized intersections. In this study, simulated
conflicts were used to pre-evaluate the safety performance of signalized intersections.
Simulated conflicts were generated by the VISSIM simulation. It was found that the
safety performance of this intersection was improved after reducing the speed limit.
Zhixia and Madhav (2014) intended a modelling reservation based autonomous
intersection control in VISSIM. This model was introduced and implemented in
VISSIM. Comparison was done with the results obtained. The comparison was done
within optimized signalized control, significantly reduce delays, intersection capacity
and with lower volume to –capacity ratios under various traffic demand conditions.
Lee Vien and Mohammed Bally (2014) presented microscopic simulations on the
design and operational performance of diverging diamond interchange. Results from
traffic impact studies indicated that most of the existing intersections in urban areas
have already exceeded the capacity. This paper elaborates the microscopic simulation
model developed using the VISSIM software to investigate the design and operational
performance of a diverging diamond interchange for different traffic scenarios (Fig. 2).

Fig. 2. Flow chart of the methodology

 Selection of Optimum Junction Operation Strategy for Gatambe Intersection 27

3 Methodology
3.1 Data Collection
A survey was conducted from 6.00 A.M to 8 P.M during a weekday and the below
mentioned activities were conducted to collect the necessary data to identify the
problem.
• Classified Turning Movement Count at 4 junctions. (CTMC)
• Queue length measurements of each road.
• Rail gate operation survey (Fig. 3).

Fig. 3. Layout of the conducted survey

Geometric data was collected to draw the roundabout. Following figure shows the
dimensions of the existing roundabout. In addition to this, all necessary dimensions of
the road network were measured for input in the simulation software (Fig. 4).
28 K. L. D. Maduranga et al.

Fig. 4. Dimensions of roundabout

Through classified turning movement count peak hour and off peak hour was
identified (Fig. 5).

Fig. 6. Turning movements at the junction

Peak hour – 07:15–08:15

Off peak hour – 14:00–15:00
Following figure shows the turning movements of the roundabout during the peak
hour (Fig. 6).
 Selection of Optimum Junction Operation Strategy for Gatambe Intersection 29

Fig. 5. 14 h total entry vehicle flow

4 Alternative Solutions

To find the optimum junction improvement strategy, different options were tested using
VISSIM simulation software. The alternative solutions were;
1. Option A: Roundabout
2. Option B: Signalization without geometric improvement
3. Option C: Signalization with geometric improvement - 3 phase and 4 phase
4. Option D: Grade separation
From each option, the average delay per vehicle during peak and off peak hour was
taken to find the optimum results with comparing the benefit cost ratio.

4.1 Option A
The first option was the existing situation, Roundabout option. To run the VISSIM
program we needed to draw the roundabout with the road network in the VISSIM
interface including correct geometric data. Google maps was used as the VISSIM
interface and then road network was drawn in the VISSIM interface, traffic data was fed
to the model considering the vehicle composition. Following figure shows the simu-
lated model in VISSIM interface (Fig. 7).

Fig. 7. Existing condition

30 K. L. D. Maduranga et al.

This is the reference model for the simulation. The model was validated using the
simulated queue length results and with actual queue lengths which were taken from
the conducted survey.

4.2 Option B
Under this option, four phase signalization was used without the geometric improve-
ment. First the four phase signal time was calculated. Then validated road network
model was used and the signal time was input to the model (Figs. 8 and 9).
Cycle time – 217 s

Fig. 8. Four phase signal time

Fig. 9. 4 – phase signalization without geometric improvement

4.3 Option C
Under the geometric improvement, an additional lane is provided for the right turn
movement of the Peradeniya road and Gannoruwa entry road is widened to a three lane
road. The length of the additional lane of the Peradeniya road was given by considering
 Selection of Optimum Junction Operation Strategy for Gatambe Intersection 31

maximum queue length which can be accommodated and with the maximum queue
length results which were taken from the simulation of the existing condition. The
additional lane of the Gannoruwa road is provided along the bridge (Figs. 10, 11 and 12).
3- Phase signalization
Cycle time – 120 s

Fig. 10. Three phase signal time

Fig. 11. 3 – phase signalization with geometric improvement

4 – Phase signalization
Cycle time – 217 s

Fig. 12. 4 – phase signalization with geometric improvement

32 K. L. D. Maduranga et al.

4.4 Option D
A conceptual proposal to grade separate movements PW and WP is considered as
option D. When considering the 14-h flow and the peak hour flow, it is observed that
there is no predominant movement. However, percentage turning movements of the
William Gopallawa Mawatha to Peradeniya road (WP) and Peradeniya road to the
William Gopallawa Mawatha (PW) sums up to 28% of total. This conceptual proposal
incorporates grade separation of PW and WP over other movements and also the
railway (Fig. 13).

Fig. 13. Grade Separation

Design details of grade separator

• Carriageway width (3.65 m * 2) = 7.3 m
• Median Width = 0.8 m
• Shoulder Width (0.3 m * 2) = 0.6 m
• Total Width of the Flyover = 8.7 m
• Gradient = 0.06
• Clearance Height = 7.5 m

5 Economic Analysis

5.1 Outputs from VISSIM Simulation

The results observed after running VISSIM simulation for the four options A, B, C and
D are given in the Table 1.
By only looking the delay results of these alternatives, it is difficult to get the
optimum results due to high cost of some options. Because of that, carrying a Benefit
Cost analysis was a better way to find the optimum solution. Following table shows the
Benefit Cost analysis which was done for each option.
 Selection of Optimum Junction Operation Strategy for Gatambe Intersection 33

Table 1. Average individual delay per vehicle during peak and off peak hour from simulation
Average individual vehicle delay/(s)
Option Peak hour Off peak hour
A 268 157
B 212 138
C – 3 phase 88 64
C – 4 phase 176 114
D 22 13

5.2 Estimation of Benefits

According to the standards practice suggested for the assessment of the transportation
section public investments, following are the identified benefits,
1. Travel time saving
2. Reduction of VOC
3. Reduction of emission
4. Improved safety
For the Benefit, only the travel time saving is considered to compare the alterna-
tives. Assuming that night travel time saving is not a benefit to the alternative options,
the benefit was calculated considering 3 peak hours and 9 off-peak hours per day. To
calculate the benefit considered CCPI 2018 values based on CCPI 1999 values
(Table 2).

Table 2. Estimated benefit for each alternative

Option Benefit/(Rs. Million)
Option A –
Option B 63.48
Option C – 3 phase 251.56
Option C – 4 phase 123.38
Option D 370.90

5.3 Estimation of Cost

Following cost components are considered in the economic analysis
1. Land acquisition cost
2. Construction cost
3. Maintenance cost
Construction costs were taken from the BOQ provided by the SCDP and land
acquisition cost is calculated considering the land price and land area to be acquired.
Summarized cost for four option are as follow (Table 3).
34 K. L. D. Maduranga et al.

Table 3. Estimated cost of each alternative

Option Land acquisition Construction cost/ Signalization cost/ Total cost/
cost/(Rs. million) (Rs. million) (Rs. million) (Rs. million)
A – – –
B – 50 30 80
C–3 70 114 30 214
phase
C–4 70 114 30 214
phase
D 60 513 – 573

Finally, the present value for the benefit and the cost were calculated to take the
benefit to cost ratio for each option assuming the design life of 20 years and discount
rate 10%. Maintenance cost for flyover is Rs 5 Million and for rest of the options
maintenance cost is Rs 4 Million.

Table 4. Benefit – cost analysis for the alternative options

Option
A B C D E
3 Phase 4 Phase
Benefit/Rs. million – 63.48 251.56 123.38 370.90 240.97
PV of benefit/Rs. million – 540.44 1912.98 938.24 2513.97 1832.45
Cost/Rs. million – 80 214 214 573 269
PV of cost/Rs. million – 103.15 221.66 221.66 520.22 274.96
Benefit/cost – 5.24 8.63 4.23 4.83 6.66

6 Conclusion

The purpose of this study is to review various options that can be implemented to
reduce the traffic congestion at the Gatambe intersection and find the optimum solution.
Table 1 shows the average individual delay per vehicle during peak and off peak hour
which were taken from the simulation. It shows that all the alternative options have low
individual delay per vehicle than the existing situation. Through this simulation, the
effect of each method can be implemented before construction. Therefore, this simu-
lation method is very useful for all the practical situations. However, by considering
only results of individual delay per vehicle, the best solution cannot be found.
A Benefit –Cost analysis was done to get the best solution which gives the least delay
for a low cost. To get the benefit, travel time saving was considered.
From Table 4, the maximum benefit to cost ratio is given by the option C, 3 – phase
signalization with geometric improvement. This option consists of a 3 – phase sig-
nalization system with geometric improvement which has an additional right turn lane
for the Peradeniya road and the Gannoruwa entry road is widened to a three lane road.
 Selection of Optimum Junction Operation Strategy for Gatambe Intersection 35

Although the travel time saving of the flyover is maximum, due to the high cost it
got a lower value in the benefit-cost ratio. Also, when considering the 4 – phase
signalization without geometric improvement, it gives a considerable delay. Therefore,
the benefit is less. By analysing these benefit-cost ratios, the geometric improvement
with a 3 – phase signalization is the optimum solution.

Acknowledgements. The authors wish to express their sincere thanks to Eng. (Mrs) T.N.
Siyambalapitiya, Director (Planning), RDA, Engineering staff of planning branch of RDA,
Traffic survey team of RDA and all the members in the survey team for the great support in
conducting the survey successfully. A special thank goes to Eng. Manjula Kularathne of CMC,
Mr. Chathura and Mr Nadeeka Jayasooriya of University of Moratuwa for their advice on
VISSIM simulation software.

References
Brugmann J, Schreckenberg M, Luther W (2014) A variable simulation model for real-world
microscopic traffic simulations. Simul Model Pract Theory 48:58–92
Dong L, Xiaokuan Y, Chao G (2013) VISSIM - based simulation analysis on road network of
CBD in Beijing, China. Procedia Soc Behav Sci 96(2013):461–472
Huanyun Z, Fei H (2013) Development of traffic safety evaluation method based on simulated
conflicts at signalized intersections. Procedia Soc Behav Sci 96:881–885
Kun C, Lei Y (2007) Microscopic traffic-emission simulation and case study for evaluation of
traffic control strategies. J Transp Syst Eng Inf Technol 7(1):93–100
Lee Vien L, Mohamed Bally M (2014) Microscpic simulation on the design and operational
performance of diverging diamond interchange. Transp Res Procedia 6:198–212
Mohammad Rayeedul Kalam S et al (2017) Heuristic algorithm based application tool to
calibrate microscopic simulation parameters. traffic and transportation engineering (English
edition)
Natalia S, Leszek S (2017) The use of a simulation model to analyze the impact of heavy
transport generated by the port to the city traffic. J KONBiN 42:25–42
Papageorgiou GN (2006) Towards a microscopic simulation model for traffic management: a
computer-based approach. IFAC Proc Volumes 39:403–411
Weerakoon H, Weeraman T (2017) Use of microscopic traffic simulation model as a test bench to
assess improvement options for Gatambe Intersection. S.N., Kandy
Yun M, Ji J (2013) Delay analysis of stop sign intersection and yield sign intersection based on
VISSIM. Procedia-Soc Behav Sci 96:2024–2031
Zhandong Z, Shaohui C, Yanquan Y, Aixiu H (2016) VISSIM simulation based expressway exit
control modes research. Procedia Eng 137:738–746
Zhixia L, Madhav V (2014) Modelling reservation - based autonomous intersection control in
VISSIM. Veh - Highway Autom Committee Peer 2381:81–90
 Gap Assessment of Warning
and Dissemination Process of Early Warning
System in Coastal Areas of Sri Lanka

H. M. S. S. Hippola1(&), E. M. S. D. Jayasooriya1,
G. P. Jayasiri2, Chameera Randil2, Chamal Perera2,
K. K. K. Sylva1, A. K. Kulathunga1, C. S. Bandara1,
C. S. A. Siriwardena2, and P. B. R. Dissanayake1
1
University of Peradeniya, Peradeniya, Sri Lanka
[email protected]
2
University of Moratuwa, Moratuwa, Sri Lanka

Abstract. Multi Hazard Early Warning Systems (MHEWS) issue warnings to

help communities safely evacuate from hazardous areas. Studies showed that the
response of coastal communities to warnings issued by the existing MHEW
system in Asian countries including Sri Lanka is insufficient. In this study, the
current context and gaps of MHEW system related to the “warning and dis-
semination” process in the coastal regions of Sri Lanka were identified using a
proper research methodology. Survey was carried out through oral question-
naires in selected coastal areas of Sri Lanka. By interpreting the output of the
interviews, along with a deep literature review, the current context and gaps of
the MHEW systems related to warning and dissemination in coastal areas of Sri
Lanka were listed out. From the identified problems, key areas which need
special attention of the Disaster Management Centre, were highlighted. Among
them, introducing a better Nation-wide Emergency Communication system
using more advanced technologies to directly disseminate warnings to vulner-
able communities instantly, automating possible components of the warning
dissemination process, increasing the number of early warning towers,
increasing the human resource in the disaster management centre and estab-
lishing a specialized media department at the DMC to coordinate with the media
were prominent. In addition, the necessity of improving response capability was
highlighted.

Keywords: Disasters
MHEW
Warning and dissemination

1 Introduction

Sri Lanka, being an island situated in the Indian Ocean, is prone to coastal hazards.
Among those hazards, floods, cyclones, thunder storms and coastal erosions occur at
more frequent intervals, a tsunami had occurred once in the recent past while there are
risks of earthquakes as well. Sri Lanka, already ravaged by the floods in 2003, was then
hit by the monstrous Indian Ocean tsunami in 2004 which engulfed approximately
35,000 Sri Lankan lives. These conversions of coastal hazards in to massive disasters

© Springer Nature Singapore Pte Ltd. 2020

R. Dissanayake and P. Mendis (Eds.): ICSBE 2018, LNCE 44, pp. 36–44, 2020.
https://doi.org/10.1007/978-981-13-9749-3_4
 Gap Assessment of Warning and Dissemination Process 37

led the government of Sri Lanka to strengthen legislature and institutional arrangements
for disaster risk management. An efficient Multi Hazard Early Warning (MHEW)
system was required urgently to reduce the risk of destruction of lives and property.
Subsequently, Sri Lanka enacted the Disaster Management Act No. 13 in 2005. Under
this Act, Disaster Management Centre (DMC) of The Ministry of Disaster Manage-
ment, Sri Lanka was established with the responsibility of preparing and implementing
of a National Disaster Management Plan and a National Emergency Response Plan.
Despite these legislative and institutional reformations, Sri Lanka has experienced
several coastal hazards with considerable damage since 2005 (Disaster Management
Centre 2016). Flood events were reported every year affecting 19 districts on average.
Each flood event has caused considerable loss of lives with the flood events in 2008
and 2011 being those with highest number of death tolls, 44 and 69 respectively. In
addition, high winds have caused a considerable loss of lives in the years 2011 and
2013. Furthermore, the most recent southwest monsoon weather condition has caused
22 deaths. It is to be noted that these numbers are quite high for hazards such as floods
and high winds and the consequences are much more severe if another tsunami hap-
pens. These facts reflect either the inefficiency of MHEW or the reluctance of the
vulnerable communities to respond to early warnings. Therefore it is timely to identify
gaps of the existing MHEW system in Sri Lanka and make the required improvements.
Furthermore it is equally vital to address and remedy the problem of communities being
reluctant to obey early warnings.
The United Nations International Strategy of Disaster Reduction (UNISDR 2006)
points out four key elements of a people centred early warning systems. They are Risk
Knowledge, Monitoring and Warning Service, Dissemination and Communication and
Response Capability. Risk knowledge involves assessing the dynamic nature of haz-
ards and vulnerabilities due to climate change and change in demography. Monitoring
and Warning services should consist of a sound scientific basis to continuously predict
and forecast hazards, timely and accurately while dissemination and communication
processes should ensure receipt of clear, meaningful messages containing critical and
useful information for vulnerable communities, with sufficient lead time to evacuate
safely. The final key element focuses on the response capability of vulnerable com-
munities. These communities should understand their risk situation and act accordingly
obeying and respecting the early warnings issued.
The whole MHEW process can be discretized in to the aforementioned key ele-
ments to study the current context and gaps. Hettiarachchi et al. (2017) have investi-
gated the efficiency and effectiveness of the existing disaster management frameworks
in Sri Lanka in a more general sense. This paper presents a discussion that focuses on
the current context and gaps of the Dissemination and Communication processes of
existing MHEW in Sri Lanka under the subtopics, Communication Channels, Warning
Messages and Coordination with Media and Military. Furthermore the reasons for the
poor response from the vulnerable communities for early warnings issued are discussed
towards the end. Dutta et al. (2014) have done a similar study to assess gaps and
strengthen early warning System in Cambodia.
38 H. M. S. S. Hippola et al.

2 Existing Warning and Dissemination System in Sri Lanka

An efficient early warning system should focus to generate and disseminate timely and
meaningful warning information to enable individuals, communities and organizations
threatened by hazards to take necessary prepared measures and act appropriately with
sufficient time to reduce the possibility of harm or losses. The process of delivering
early warnings, originated from the technical agencies, to end users, involves several
pre-planned steps. A schematic diagram of the warning and dissemination system of Sri
Lanka is given in Fig. 1.

Fig. 1. Multi hazard warning dissemination system (Extracted from the DMC official website)

At a time of an impending disaster, both local and international technical agencies

analyse the data they receive, to predict the occurrence of the concerned disaster. It is
their responsibility to forecast the scale and issue warning alerts for different types of
hazards. In Sri Lanka, local technical agencies involved are the Department of Mete-
orology (DOM), National Building Research Organization (NBRO), The National
Aquatic Resources Research and Development Agency (NARA), Department of Irri-
gation (DOI), Geological Survey and Mines Bureau (GSMB), Coast Conservation and
Coastal Resource Management Department and Marine Environment Protection
 Gap Assessment of Warning and Dissemination Process 39

Authority (MEPA). The DOM generally monitors the occurrence of tropical cyclones
and tsunami while the DOI monitors floods. The international technical agencies that
assist the DMC are also listed in Fig. 1. It is the responsibility of these local and
international technical agencies to inform the Director General of DMC and Emergency
Operating Centre (EOC) if a disaster is likely to happen. It should be highlighted that
Probability of Detection (POD) and Probability of False Alarm (PFA) are decisive
factors which governs the credibility of these early warnings issued.
Once the DMC is informed regarding an impending disaster, the MHEW dis-
semination unit of the DMC initiates the process of warning dissemination, after an
analysis of the situation. They communicate with EOC that operates for 24 h every day
which will later coordinates the whole warning dissemination process with the armed
forces (army, navy, air force, coast guard and police), media and relevant government
and non-government agencies. The outputs of the MHEW dissemination unit of the
DMC are radio/television news, short message system (SMS), fax, internet, satellite,
police and military communication, early warning towers and other media. It is
important to note that efficiency, technology of these communication channels and
outputs directly affect the speed of warning dissemination.
It is the responsibility of the EOC to disseminate the warnings to administrative
divisions such as provincial councils, district secretariats, divisional secretariats and
local authorities. In parallel to these communications, the EOC disseminates the early
warnings to disaster management committees at district, divisional and village levels.
This is followed by the receipt of early warnings by the vulnerable community through
speaker systems, bells/sirens, messengers and riders/cyclers. Police stations and the
Grama Niladhari (village level administrative officers of the government) usually assist
the warning dissemination process at the village level.

3 Methodology

The gap assessment of the warning dissemination process and assessing the response
capability for coastal hazards was done by conducting separate interviews with disaster
management officials, academics who are engaged in disaster related research and
vulnerable communities in coastal areas of Sri Lanka. The technique used to select
respondents in vulnerable communities was the snow ball sampling method. The oral
interviews carried out on professionals and vulnerable people were based on carefully
prepared questionnaires. The meetings were conducted with the aim of understanding
warning and dissemination process and key elements governing response capability.
The flow chart in Fig. 2 demonstrate the methodology.
40 H. M. S. S. Hippola et al.

Fig. 2. Methodology

4 Results and Discussions

This section presents the current context, gaps and suggestions for improvement of the
warning dissemination process for coastal communities in Sri Lanka under the sub-
topics communication channels, coordination, warning messages and response capa-
bility of vulnerable communities.

4.1 Communication Channels

Communication of the early warnings generated by local and international technical
agencies is currently carried out in four layers namely, National, District, Divisional
and Grama Niladhari (village) levels. At the National level, Early Warning
(EW) messages are disseminated to emergency response committees that are then
passed to their relevant organizations. The District Disaster Management Centre Units
(DDMCU) act as focal points of communicating EW messages to district, divisional
and Grama Niladhari levels. It is the responsibility of the DDCU to disseminate EWs to
 Gap Assessment of Warning and Dissemination Process 41

District Secretariat, Divisional Secretariat, stake holder agencies and relevant political
authorities. In the meantime, the warnings are disseminated to the vulnerable com-
munities through local authorities. At Grama Niladhari level, EW messages are dis-
seminated to the vulnerable communities through last mile communication tools. The
communication methods used in each of these levels are presented in Table 1. It is to be
noted that a common set of communication methods are used by technical agencies,
national level and district levels while last mile communication methods are used at the
village level.

Table 1. Communication methods used in warning dissemination process

Technical agencies/national and district levels Village level
Police and military communication Police vehicles
Satellite and radio communications Public address system
Intra government network Sirens
Fax Temple and church bells
Internet Riders/messengers
Early warning towers Early warning towers
Media Media
Telephones, SMS Telephones, SMS

The responses from the professionals suggest that overall communication network
can be rated as moderate. However, some of the professionals have specifically rated
the communication of EOC, with administrative levels and media as poor. An efficient
warning dissemination process demands a seamless flow of warnings. Drawbacks in a
few links might act as bottle necks to drastically reduce the efficiency of warning
dissemination process. The following paragraphs discuss the possible gaps in the
existing communication channels.
Most of the professionals rated the standard and conditions of the equipment as
moderate and agreed that the equipment are maintained at frequent intervals. Among
these equipment, early warning towers were identified as the most preferred method of
warning receiving. The general public in coastal communities confirmed that the early
warning towers are maintained and tested in frequent time intervals. However, they
have pointed out that the number of existing early warning towers are insufficient to
cover all the regions; thus there is a need of increasing number of early warning towers.
Majority of the professionals were not satisfied about the speed of communication
and technologies used. It is to be noted that the speed of communication can be
improved drastically by establishing a better Nation-wide Emergency Communication
system with new technologies which can directly convey warnings from National level
to end users. In addition, the conventional equipment can be updated and maintained at
better standards to disseminate the warnings effectively. Furthermore, almost all pro-
fessionals suggested that updating databases is as equally important as much as the
standard of equipment.
42 H. M. S. S. Hippola et al.

Majority of the professionals pointed out that the lack of staff, is a major concern in
disaster management institutions. It is true that in the current system, a considerable
human involvement is required when disseminating a warning from national level to
end users. Therefore the current system requires either the enrolment of new staff or
introducing new technologies to automate possible sections of the warning dissemi-
nation process. In addition, enrolment of new staff can be encouraged by increasing
salaries and other benefits.

4.2 Coordination of Emergency Operation Centre with Media

and Military
At a time of an impending disaster, the DMC provides accurate warnings to the media
through appointed media spokespersons. The media will disseminate the warning
messages to last mile vulnerable communities. Therefore, it is the responsibility of
DMC to ensure an accurate press release providing the required warning messages to
media within the shortest possible time. Majority of the professionals had a moderate
feeling regarding the coordination between the EOC and the media. However, some of
the professionals pointed out that the media tends to act on their own will and thus
reduce the efficiency of warning dissemination.
To address this issue, some professionals have suggested to establish a specialized
media department at the DMC to build up a better understanding with the media. This
specialized coordination will allow the DMC and the media to be efficiently prepared
for a warning dissemination situation depending on the type of hazard and vulnerable
community.
The military forces consists of the Army, Navy, Air Force and the Police. These
forces have the inherent capability to disseminate early warnings to vulnerable com-
munities and direct target population to safe locations. Therefore the DMC should
disseminate the warning messages to the forces with maximum possible accuracy.
Almost all of the professionals have rated the coordination between EOC and Military
forces as good.

4.3 Warning Messages

Warning messages received by the vulnerable communities should convey the specific
nature of the threat and its impacts. The UNISDR (2009) presents that warning mes-
sages should be tailored to the specific needs of those at risk, should be geographically
specific, should in cooperate the understanding of the values, concerns and interests of
those who will need to take action, should be well recognized and understood, should
be consistent over time, and should include follow up actions.
Majority of the professionals rate the condition of warning messages as moderate.
They point out that the current warning dissemination system has previously prepared
warning messages with procedures to update them. However, the existing warning
messages should be improved drastically to meet the standards suggested in the
UNISDR (2006). The reduced quality of warning messages was reflected in the recent
flood events where twenty people died. Therefore it is mandatory to create new
 Gap Assessment of Warning and Dissemination Process 43

meaningful warning messages accounting for the type of disaster, vulnerable com-
munity, communication method etc.

4.4 Response Capability

From the above analysis, it is evident that the warning and dissemination processes
implemented in Sri Lanka are satisfactory despite the improvements suggested.
However, the death tolls in recent hazardous events suggests that there is a serious issue
when disseminating early warnings to vulnerable communities. If the warning dis-
semination process stands satisfactory, the problem lies with the response capabilities
of the vulnerable communities. The interviews conducted with the coastal communities
pointed out several root causes for this poor response capability. The UNISDR (2009)
presents a check list to improve the response capability of vulnerable communities.
This check list includes respecting warnings, disaster preparedness and established
response plans, assessed and strengthened community response capacity and enhanced
public awareness and education. This check list can be used to propose suggestions in
order to improve the response capability of Sri Lankan coastal communities.
The business oriented life style of people in coastal communities has led to their
lack of knowledge on disasters. This lack of knowledge was reflected in 2004 tsunami
incident, where people went towards the sea when water level reduced prior to the huge
waves. The surveys revealed that most of the people are unaware about the decision
makers, how warnings are disseminated, simple information on hazards, vulnerabilities
etc. This unawareness may lead to situations where people ignore warning messages
and become victims of the disaster. Another aspect of this unawareness is, most of the
people preferred to respond only to bells/sirens, messengers and early warning towers
rather than warning messages from SMS, internet and media. To address this issue of
not knowing, public awareness programs should be conducted highlighting the MHEW
process, a general knowledge about disasters, reliable early warning sources and
methods of responding to early warnings. In addition, the unawareness can be further
reduced by introducing new academic curriculums with disaster management courses
from the school level to the university level.
False alarms is another key factor which reduces the response of vulnerable
communities for early warnings. One false alarm by the authorities will drastically
reduce the credibility of the sources. For example, most of the people in Benthota area
experienced false alarm before the storm event of November 2017 and faced the
consequences of not obeying the next warning where the storm actually happened. Due
to these facts, false alarms should be minimized as much as possible. However, this
breach of trust can be restored by educating the general public on the real scenario of
forecasting and early warnings.
The Sri Lankan culture is built around the concept of “home” where people live
throughout their lives. The interviews with people in coastal communities exposed that
some people show reluctance to evacuate leaving houses, personal belongings, vehicles
and agricultural lands for the sake of one warning. Furthermore, this reluctance can
increase due to the absence of a specific place to evacuate. To avoid these miscon-
ceptions, the awareness of people needs to be raised through education and awareness
programs.
44 H. M. S. S. Hippola et al.

5 Conclusions

The paper proposes several improvements to the warning and dissemination process
under the sections communication channels, warning messages and co-ordination with
media and military forces. The main improvements proposed to the communication
channels are, introduce a better Nation-wide Emergency Communication system using
more advanced technologies to directly disseminate warnings to vulnerable commu-
nities instantly, automating possible components of warning dissemination process,
improving the maintenance of communication tools, increasing the number of early
warning towers and increasing the human resources in the disaster management centre.
In addition, the possibility of establishing a specialized media department at the DMC
to coordinate with the media was highlighted. Furthermore, it was proposed to make
new meaningful warning messages, accounting for the type of disaster, vulnerable
community and communication method. In the next stage of the methodology, it was
identified that the moderate warning and dissemination system existing in Sri Lanka
becomes useless when the vulnerable communities lack response capability. Finally,
education and awareness programs were identified as a remedy to improve the response
capability of coastal communities in Sri Lanka.

Acknowledgements. This study was conducted with the financial aid from CABARET
(Capacity Building in Asia for Resilience Education), a project of the European Union’s Eras-
mus + Programme - Key action 2 - Capacity building in the field of higher education.

References
Disaster Management Centre (2016) Impacts of disasters in Sri Lanka. https://app.adpc.net/sites/
default/files/public/publications/attachments/Impacts%20of%20Disasters%20in%20Sri%20L
anka%202016.pdf
Dutta R, Basnayake S, Ahmed AK (2014) Assessing gaps and strengthening early warning
system to manage disasters in Cambodia. J Integr Disaster Risk Manag 172 and 173
Hettiarachchi SSL, Siriwardena CS, Jayasiri GP (2017) Investigation of efficiency and
effectiveness of the existing disaster management frameworks in Sri Lanka. In: 7th
international conference on building resilience; using scientific knowledge to inform policy
and practice in disaster risk reduction
Ministry of disaster Management - Sri Lanka (2013) Progress Report 2013. http://www.dmc.gov.
lk/index.php?option=com_content&view=article&id=75&Itemid=161&lang=en
UN-ISDR Developing Early Warning Systems: A Check list (2006) The third international
conference on early warning. https://www.unisdr.org/2006/ppew/info-resources/ewc3/
checklist/English.pdf
UN-ISDR Terminology on Disaster Risk Reduction (2009). http://www.unisdr.org/files/7817_
UNISDRTerminologyEnglish.pdf
 Gaps in the Evacuation of People with Special
Needs During Coastal Disasters in Sri Lanka

E. M. S. D. Jayasooriya1(&), H. M. S. S. Hippola1, C. S. Bandara1,

A. K. Kulathunga1, K. K. K. Sylva1, P. B. R. Dissanayake1,
G. P. Jayasiri2, Chameera Randil2, Chamal Perera2,
and C. S. A. Siriwardana2
1
University of Peradeniya, Peradeniya, Sri Lanka
[email protected]
2
University of Moratuwa, Moratuwa, Sri Lanka

Abstract. Coastal hazards taken place all over the world have ended up with
major devastations such as loss of life, property damage and environmental
degradation more often. Sri Lanka, being an island in the Indian ocean is much
vulnerable to coastal hazards; the unprecedented human disaster that caused by
the Asian Tsunami was one of the worst examples. Present studies show that
there are lot of programs carried out by the government and non-government
organizations to drag down the adversity towards the coastal communities in the
country who are much vulnerable to coastal disasters, focusing all the segments
of basically social, economic and environmental aspects. Evacuation of coastal
communities safely in disastrous situations is one of the main requirements
within the disaster recovery framework. Having effective multi-hazard early
warning (MHEW) systems and procedures in place in the coastal regions of the
country helps to reduce the disaster risk. Although there are policies and
practices for the evacuation of coastal communities including MHEW systems,
the present study shows that the attention and preparedness towards the evac-
uation of people with special needs in the coastal communities is more or less
omitted by the responsible parties. Moreover, when disasters strike, people with
special needs meet various discriminations in terms of evacuation to a shelter or
relief points and more often are overlooked from the emergency preparedness,
response and recovery stages. Therefore, the present research focuses on
determining the gaps in the evacuation of people with special needs in the
context of MHEW systems during coastal hazards in Sri Lanka. Open interviews
and telephone interviews were conducted in few major coastal cities and
understood the likelihood of the obstacles, issues and the perception towards the
evacuation of people with special needs. A questionnaire was developed
according to the open interviews and fine-tuned in three sandpit events
employing various local and foreign professionals in academics. The ques-
tionnaires were distributed online and the responses were analyzed. Eventually,
a substantial number of gaps were identified in supporting the prevailing sys-
tems in Sri Lanka and was arrived at conclusions to impact the government
through the findings in order to make the evacuation of people with special
needs during coastal disasters more efficient.

© Springer Nature Singapore Pte Ltd. 2020

R. Dissanayake and P. Mendis (Eds.): ICSBE 2018, LNCE 44, pp. 45–53, 2020.
https://doi.org/10.1007/978-981-13-9749-3_5
46 E. M. S. D. Jayasooriya et al.

Keywords: Coastal hazards
Evacuation
People with special needs

Multi hazard early warning systems

1 Introduction

The ongoing studies from the National Council for Persons with Disabilities in Sri
Lanka professes a deficiency of research regarding the guidelines and understanding
about the preparedness in the early stage before a disaster, emergency response stage
during a disaster and recovery stage for people with special needs; pointedly people
who are living with disabilities. When it comes to world scenarios, disasters such as
Asian Tsunami 2004 jeopardise the past approaches such as “wait-for-help” in an
evacuation of people with mobility deficiencies. Asian Tsunami that brought severe
damage to Sri Lanka even showed that the above method resulted in lot of loss of lives
due to its irregularities. In the Sendai Framework for Disaster Risk Reduction, people
with special needs have been addresses more or less and the responsibilities of the
stakeholders towards them (Stough and Kang 2015). When it comes to children,
Children with special needs seem expelled from disaster risk reduction (DRR) efforts
thus leading to enhanced problems related to physical, psychological, and educational
growth and development starting from the early stage of their lives (Ronah et al. 2015).
According to Nobuyuki et al. (2006), the mortality has been observed among females,
elderly and children in Sri Lanka due to the Asian Tsunami 2004. Moreover, related
issues and obstacles of how to facilitate and help people with disabilities during an
emergency disastrous situation, how to deal and support to each other, and help policy
and emergency planners, have been disregarded and unnoticed (Alexander et al. 2012).
The content of literature and research on the aforementioned context is trivial compared
to the need and demand (Parr 1987) and latest efforts to reiterate it have been neither
numerous nor copious (Fjord and Manderson 2009).
The ultimate goal of this project is to ameliorate and promote efforts and plans for
an enhanced preparedness delivered by government and non-government institutions
related to people with special needs during every stage of a disaster such as before,
during and after the disaster. The aforementioned objective was polished through
observing the severity and the occurrence of coastal disasters, livelihood of the people
with special needs, emergency needs, government cooperation and other relevant
community conditions.
The research team collected and assessed data related to ongoing issues of people
with disabilities and their domiciliary environment when a disastrous situation strikes.
Withal, the research team also assembled data from key government officers such as
“Grama Niladhari” and divisional secretariats, also comprising emergency planners,
 Gaps in the Evacuation of People with Special Needs During Coastal Disasters 47

multi service providers, and other respective government and non-government bodies.
People were gathered at the regional offices and the discussions were carried out. The
aim of this data collection was to congregate people with special needs and the
organizations related to disaster management to build up strong relationship and review
the issues and matters related to preparedness and emergency evacuation in a more
productive and lively manner.
The current study also aims at sharing the findings with the emergency planners,
managers and decision makers who directly or indirectly work for the disaster risk
reduction and making efforts to include people with special needs to consider as a
particularly and admittedly vulnerable group during coastal disasters.

1.1 Disability and Its Implications in Sri Lanka

According to the National Policy on Disability, the legitimate interpretation of dis-
ability in Sri Lanka which was stated in the protection of the rights of persons with
disabilities which was originated by the Sri Lankan Parliament in 1996 is given below.
A “person with disability means any person who, as a result of any deficiency in his
physical or mental capabilities, whether congenital or not, is unable by himself to
ensure for himself, wholly or partly, the necessities of life” (De Silva et al. 2007). This
definition seems comprehensive, including most aspects of medical, social and eco-
nomic factors binding with people with disabilities. However, there are numerous
definitions and implications for persons with disabilities in various acts and curricu-
lums contemporary prevailing in Sri Lanka. Furthermore, people with disabilities and
special needs are considered as a vulnerable group in this all above context (Hem-
ingway and Priestley 2006) and still this so-called vulnerability is barely even con-
templated or said. Nevertheless, according to the literature available it appears that
there is a smidgen of actions related to the disaster risk reduction in the context of
disability. According to the census and statistics in Sri Lanka in the year 2001, there are
a significant number of people with special needs living in the country including the
areas prone to coastal disasters (Ministry of Social Welfare 2003). Table 1 was
extracted from the national report of census and statistics and it shows the statistical
illustrations of the different disabilities in the areas we concerned for the present study.
It is important to note that Table 1 includes only a few districts that we covered our
scope of work.
48 E. M. S. D. Jayasooriya et al.

Table 1. The census and statistics of six major near coastal districts in Sri Lanka in the year of
2001, according to type of disability, number and rate per 10,000 population
Type of disability, number and rate per 10,000 population
District
Hearing/Speaki Disability in Disability in Other Physical Mental
and sex Total Seeing
ng Hands Legs Disability Disability

No Rate No Rate No Rate No Rate No Rate No Rate No Rate

Total(18 Districts)
Both 274711 162.9 69096 41 73343 43.5 48131 28.5 90576 53.7 13275 7.9 69026 40.9
Male 158446 189.9 35419 42.5 40584 48.6 31070 37.2 56529 67.8 7344 8.8 37181 44.6
Female 116265 136.4 33677 39.5 32759 38.4 17061 20 34047 40 5931 7 31845 37.4
Colombo
Both 27430 122.8 5998 26.8 6414 28.7 4854 21.7 9168 41 1219 5.5 8730 39.1
Male 15355 135.4 2997 26.4 3507 30.9 3208 28.3 5655 49.9 698 6.2 4391 38.7
Female 12075 109.8 3001 27.3 2907 26.4 1646 15 3513 31.9 521 4.7 4339 39.4
Kalutara
Both 18431 173.7 4690 44.2 5043 47.5 3272 30.8 6209 58.5 808 7.6 4809 45.3
Male 10471 201.1 2303 44.2 2757 52.9 2096 40.3 3818 73.3 431 8.3 2605 50
Female 7960 147.4 2387 44.2 2286 42.3 1176 21.8 2391 44.3 377 7 2204 40.8
Galle
Both 17873 180.4 4722 47.7 5051 51 3051 30.8 5762 58.2 666 6.7 4981 50.3
Male 9796 204.3 2235 46.6 2727 56.9 1888 39.4 3389 70.7 346 7.2 2585 53.9
Female 8077 158 2487 48.7 2324 45.5 1163 22.8 2373 46.4 320 6.3 2396 46.9
Matara
Both 15255 200.4 3773 49.6 4320 56.7 2535 33.3 4764 62.6 681 8.9 4379 57.5
Male 8486 231 1849 50.3 2373 64.6 1574 42.8 2772 75.4 336 9.1 2343 63.8
Female 6769 171.9 1924 48.9 1947 49.4 961 24.4 1992 50.6 345 8.8 2036 51.7
Hambanthota
Both 12736 242.4 3682 70.1 3287 62.6 2079 39.6 3745 71.3 628 12 3391 64.5
Male 7181 274.8 1852 70.9 1836 70.3 1272 48.7 2203 84.3 317 12.1 1882 72
Female 5555 210.3 1830 69.3 1451 54.9 807 30.6 1542 58.4 311 11.8 1509 57.1
Ampara
Both 8462 143.6 2339 39.7 2367 40.2 1289 21.9 2511 42.6 324 5.5 1670 28.3
Male 4973 168.4 1247 42.2 1298 44 839 28.4 1623 55 191 6.5 946 32
Female 3489 118.7 1092 37.1 1069 36.4 450 15.3 888 30.2 133 4.5 724 24.6

2 Methodology and Study Plan

According to the above explained content, the motif of the current study is to find out
some of the gaps in the currently prevailing system in Sri Lanka related to people with
special needs and their emergency evacuation during a coastal disaster. It was per-
formed with regards to the various hazards occur, disability and its’ implications,
livelihood of the people with special need, evacuation routes and shelters, different
emergency needs and also the level of local government cooperation before and during
a disaster.
 Gaps in the Evacuation of People with Special Needs During Coastal Disasters 49

2.1 Data Collection: Survey with People with Special Needs and Their
Relatives, Local Public and Government and Non-Government
Agencies
The approach was consisted with telephone calls, open and closed interviews and also
an online survey. Basically, the evacuation of people with special needs by the normal
people was considered while determining the administrative gaps and existing
methodologies. The interviews were done in 6 major near coastal districts. (Colombo,
Kaluthara, Galle, Matara, Hambanthota, Ampara). A part of the interviews was collected
through telephone survey of the people with special needs and as well as normal public.
The other part of the responses was collected through an online questionnaire survey of
persons with reference to the evacuation of people with special needs who live in the
community. The questionnaire was developed according to the open interviews and
fine-tuned in three sandpit events employing various local and foreign professionals in
academics. The questionnaires were distributed online and the responses were analyzed.

3 Results and Discussion

The responses were obtained from people with special needs, normal people in being
with the perception of evacuation of people with special needs and a qualitative response
from professionals were gathered in order to visualize and determine the capacity of the
main topic. Half of the responses from the normal people reveals that they have someone
with special needs near their vicinity and more than sixty percent of the respondents
expressed that they put an extra effort to evacuate that sort of person while risking their
own lives. It seemed that the majority of people tends to help the people with special
needs in any kind of emergency which can be highly evaluated as a great attitude in the
community. The rest has mentioned that basically they do not have the capacity to do so
since they are more concerned about their own evacuation. But findings about priorities
stated that nearly eighty percent of people tend to prioritize their family evacuation.

3.1 Key Findings on Different Types of Hazards and Severity

From the telephone conversations and open interviews conducted, it was identified that
the most common coastal disasters are coastal flood, sea level rise, storms and heavy rains
and extreme wind forced- waves. Although Tsunami has once occurred and has brought
lot of damage, it is not considered as a frequent disaster. As reported storms have high
frequency of occurring. Although Asian Tsunami destroyed almost everything belong to
them, that kind of higher devastative Tsunami has low frequency of occurring.

3.2 Key Findings on Early Warning Systems

Whatever to achieve, the early warnings should be reached to the last mile so that
people can react to the disasters easily to an extent where it is possible to the maximum.
It was found that main early warning systems currently prevailing are SMS alert,
Tsunami early warning towers, television/radio, door to door, telephone/internet and
50 E. M. S. D. Jayasooriya et al.

police warnings. But the most common system was TV/Radio that was more than 80%.
Some people complained that they totally have not received any kind of early warnings.
When it comes to the effectiveness, majority of respondents nearly 60% have men-
tioned that it was approximately fifty percent effective.

3.3 Key Findings on Evacuations and Disabilities

As per the findings majority of people that have evacuated someone with special needs
during a coastal disaster reported that the incident was with majorly pregnant woman.
Nevertheless, there were significant amount of people with waling disability as well.
Few were with other disabilities. Almost everyone in this category mentioned that they
faced difficulties in providing mobility assistance like transportation. The other con-
siderable fact was that a mentioned above large number of people have not received the
early warnings before the Asian Tsunami 2004 and even for the disasters occurred later.

3.4 Key Findings on Trainings/Drills

According to the survey, approximately over 85% of normal people have undergone
some sort of a training drill after the Asian Tsunami 2004 but not before. The drills
have been composed with self-evacuation and as well evacuation of people with special
needs. As per the telephone discussions and open interviews conducted with the people
with special needs it is reported that they rarely had training or any kind of drills
regarding this. This fact might reveal the low attention towards those people in disasters
as the government and administrations mostly believe that the people with special
needs are never left behind by the normal people. Suggestions for having a drill
affirmed that there were few major opinions such as swimming training, first aid
training, trainings on specific ways of carrying for safety of people with special needs
and trainings on finding a shelter. There were lot of individual various and dynamic
opinions as it depends on the mentality and the knowledge of the respondent. However,
a majority of people over 80% mentioned that the training drills were held but they
have not been informed it prior and they could not participate. So, the proper coor-
dination and lack of responsibilities of the administrations were identified. Some of the
respondents pointed out that the drills are not regularly or periodically happening but it
is just only after a certain disaster.

3.5 Key Findings on Shelter

According to the respondents, it appears that nearly 60% of people know about all the
shelters that they might transfer during a disaster. As they told, after the Tsunami 2004
the sign boards and maps of evacuation paths and shortest distances to the evacuation
shelters were established but not maintained. Nowadays they are hardly seen. Some-
how over 50% of respondents know about their specific evacuation shelters. What that
means is in almost every administrative division (Grama Niladhari divisions) disaster
management centre has specified site specific evacuation shelters and routes. As per the
findings, over 70% of people go to the nearest temples because traditionally the temples
have been constructed at places where it has a higher elevation related to the particular
 Gaps in the Evacuation of People with Special Needs During Coastal Disasters 51

area. Approximately 20% of people use schools as their evacuation shelters while the
rest use some other specified location or some people do not have that kind of place due
to various reasons. It is important to note that the findings about the facilities of those
shelters strongly disclosed that the places do not have proper utilities and facilities.
Nearly 90% of respondents firmly mentioned about the lack of facilities and utilities for
the people who has special needs. The major facts they addressed are mentioned below.
As per the findings, there are no specifically built in evacuation shelters for even
normal people instead of people with special needs rather than a temple or public
school. Since they are not built or constructed upon that purpose the facilities might be
inadequate obviously. Majority of the people mentioned about the lack of space,
sanitary facilities, gender bias irregularities and more often ethnic bias complications
are also there. It is a very paramount discussion to bring forward that according to the
respondents there have been a tendency for sexual harassments and the incidents have
been reported several times. Because of the difficulties of the people with special needs
they more often face these kinds of situations. The government and respective
administrations should attempt this case immediately and at the same time the attitudes
of the people should be questioned.

3.6 Key Findings Related to Emergency Needs

According to the telephone conversations and the open interviews conducted the
emergency needs of those people who have special needs were identified for a certain
extent. Self-care materials and equipment, dependency on utility services (water,
electricity, telephone etc.), removing household debris and evacuation, transport,
evacuation assistance, resupply of very essential goods, asking for help and assistance,
necessity of a guide dog or self-evacuation plan and local government cooperation. The
respondents stated that the most important emergency need is transport and mobility.
At the same time, they were asked whether they are given or provided those emergency
needs at the time of a disaster by the government or any sort of responsible authorities.
The respondent’s mentioned that they are only given food and first aid more often and
even that provisions are not equalized among the community because of several socio
cultural and ethnic irregularities. Eventually, suggestions were asked about the emer-
gency needs. Majority of people mentioned that the needs of infants and pregnant
women are highly conspicuous.

3.7 Key Findings Related to Local Government Cooperation

and the Improvement
Over 80% of people stated firmly that the cooperation of the local government is not
sufficient due to several reasons. The major reasons that can be emphasised are as
follows. Lack of household preparedness of the government administrations, lack of
proper statistics about the people with special needs, funding, lack of pre-planned
evacuation procedures conducted by the government lack of information about the
disasters and new technology. There were several other opinions depending on the
various perspectives. In order to improve the local system few improvement sugges-
tions could be highlighted according to the respondents. Majority of people have
52 E. M. S. D. Jayasooriya et al.

suggested that it is important to gain understanding of the present hazards and some
sort of technical and general awareness is needed. It is a learnt lesson that when
Tsunami 2004 occurred, none of the people in coastal areas knew that this was a
Tsunami and the severity of the oncoming disaster. Furthermore, estimating the
capacity of the people who have special needs was also considered important by the
majority of respondents. Since the level of need and assistance differ depending on the
capacity of the people it would be a proper procedure to identify specific people for
specific need and assistance. Creating a personal care network consisting of at least 3
people for each site was also proposed by significant number of respondents. So that
special care can be addressed for the people who need help. Moreover, wearing a label
or bracelet in order to identify those type people was another major suggestion since
during those kinds of an event people might understand that the particular person with
the label or bracelet needs help and they can attend. All above suggestions should come
with proper planning and preparedness.
On the other hand, there is a school of thought that is composed with government
officers and volunteers, who has devoted their lives during these types of coastal
disasters to safeguard and protect the lives of the community, pointedly people with
special needs. Despite of the ardent followers who think that an extra attention towards
people with special needs is not an imperative task during a coastal or any disaster, it
can be seen that a vigorous force is developing through the social media and academic
research supporting the aforementioned issues of the people with special needs.

4 Conclusions

It is evident from the findings that there is an evoking interest towards the establish-
ment of proper policies and frameworks for the people with special needs by extending
the current disaster management frameworks by cooperating existing context of people
with special needs with regards to the coastal disasters. The cooperation and association
of agencies and institutions working for people with special are still substantive
questions when integrating people with special needs into all the stages of a disaster
such as planning and preparedness, response and recovery. It is increasingly necessary
to establish adequate subsidies to cater planning and preparedness, response and
recovery for persons with special needs. Moreover, an island wide disaster planning
programme with proper statistics of people with special needs and equally caring for
those people without any boundaries is a must.

Acknowledgements. This research was conducted in collaboration with the University of

Peradeniya, Faculty of Engineering and the financial aid from CABARET (Capacity Building in
Asia for Resilience Education), a project of the European Union’s Erasmus + programme – Key
action 2 – Capacity building in the field of higher education.
 Gaps in the Evacuation of People with Special Needs During Coastal Disasters 53

References
Alexander D, Gaillard JC, Wisner B (2012) Disability and disaster. The Routledge handbook of
hazards and disaster risk reduction. Routledge, London/New York, pp 413–423
De Silva WI, Amarabandu WP, Gunasekara HR (2007) Disability amongst the elderly in Sri
Lanka: research studies series, Number 1. Institute for Health Policy, Colombo: Department
of Census and Statistics, Charachteristics of the disabled Persons, Census of Population and
housing 2001, Sri Lanka
Fjord L, Manderson L (2009) Anthropological perspectives on disasters and disability: an
introduction. Hum Organ 68(1):64–72
Hemingway L, Priestley M (2006) Natural hazards, human vulnerability and disabling societies:
a disaster for disabled people? Rev. Disabil. Stud. 2(3):57–67
Ministry of Social Welfare (2003) National Policy on Disability for Sri Lanka
Nishikiori N, Abe T, Costa DG, Dharmaratne SD, Kunii O, Moji K (2006) Who died as a result
of the Tsunami? – risk factors of mortality among internally displaced persons in Sri Lanka: a
retrospective cohort analysis. BMC Public Health 6(1):73
Parr AR (1987) Disasters and disabled persons: an examination of the safety needs of a neglected
minority. Disasters 11(2):148–159
Ronoh S, Gaillard JC, Marlowe J (2015) Children with disabilities and disaster risk reduction: a
review. Int J Disaster Risk Sci 6(1):38–48
Report of the Online Forum on Disabled and other Vulnerable People in Natural Disasters (2006)
World Bank Disability & Development Team (Human Development Network), World Bank
Institute’s Multimedia Center, Center for International Rehabilitation, Leonard Cheshire
Centre of Conflict Recovery, University College London
Stough LM, Kang D (2015) The Sendai framework for disaster risk reduction and persons with
disabilities. Int J Disaster Risk Sci 6(2):140–149
Community Environment and the
Ecosystem
 Multifunctional Landscapes for Urban Flood
Management – A Study with Reference
to the Kotte Wetland

D. N. Denipitiya(&) and S. Udalamaththa

Department of Architecture, University of Moratuwa, Moratuwa, Sri Lanka

[email protected]

Abstract. Climate change, combined with the lack of drainage caused by

urbanization contributes greatly to flooding in urban areas. With the intensifi-
cation of urban floods in recent years posing a challenge to the development of
cities, research is being carried out to focus on providing solutions for this
global issue. Early research has concluded that the loss of wetlands which are
natural flood detention areas has a notable effect on aggravating floods. Hence, it
is evident that emphasis needs to be given to shifting to a landscape architectural
approach of blue and green infrastructure, focusing on a multifunctional solution
of integrating the landscape elements of land, wetland and water with envi-
ronmental, social and economic benefits. Due to the socio-economical signifi-
cance of the Colombo city, flooding is considered a significant hinderance for
the city’s development. Several projects carried out around the Kotte marsh
focusing on the Diyawanna Oya in Sri Jayawardenepura Kotte have attempted to
address this issue through multifunctional landscape developments. This
research investigates the effectiveness of these multifunctional spaces for flood
management. The literature review identifies multifunctional landscape as the
most significant solution with several principles and concepts under research.
Through this, a theoretical framework is formed to measure the effectiveness of
selected multifunctional landscapes. According to this study, multifunctional
landscapes emphasize on sustainable water management in parallel to social
responsiveness with the aim of making cities resilient to floods. The research
was carried out through field studies, photographic surveys and mapping. It was
identified that the landscape developments around the Kotte wetland included
landscape elements which used materials and technology to create effective
multifunctional spaces. However, it was concluded that the continuous decrease
in natural flood detention would affect the large-scale flood resilience of the
landscape in the long run.

Keywords: Urban flood management
Landscape architecture
Wetland

Multifunctional landscapes
Resilient cities
Colombo

1 Introduction

Floods are caused by high runoff rates generated due to intense rainfalls, causing a
surface outflow in areas where no water surface occurs (Zevenbergen et al. 2010).
Urban flooding is associated with inadequate urban planning and the effects of climate
© Springer Nature Singapore Pte Ltd. 2020
R. Dissanayake and P. Mendis (Eds.): ICSBE 2018, LNCE 44, pp. 57–74, 2020.
https://doi.org/10.1007/978-981-13-9749-3_6
58 D. N. Denipitiya and S. Udalamaththa

change acting on the lack of drainage in an urban area (Zevenbergen et al. 2008). Cities
are considered more exposed to the impacts of floods due to the high population and
economic and commercial activities that are hindered.
Urban flood management as an entirety, responds to a threat caused due to change
in land use patterns which affect traditional water management methods. The need to
address sustainable development and urban resilience in the modern day has brought
about the research and implementation of integrated solutions for water management
such as multifunctional landscapes with blue-green infrastructure (Bacchin et al. 2014;
Rezende et al. 2011; Kato and Ahern 2009; Cesar et al. 2007). It focusses on a holistic
and landscape architectural approach of rethinking stormwater infrastructure while
actively contributing to the quality of urban life, creating multifunctional urban spaces,
with many additional social, economic and ecological benefits.
Globally, studies have been done and measures taken to address urban floods
through landscape developments. The main aim of this study is to establish the
importance of implementing multifunctional solutions in Colombo as the sustainable
method of mitigating floods, by considering the effectiveness of the solutions currently
implemented in the Kotte Wetland around the Diyawanna Oya.

2 Literature Review

2.1 Urban Flood Management and Flood Control Measures

As defined by Dekay and O’Brien (2001), “A city is a human ecosystem set in a
landscape. Cities have historically had a transformative effect in their local hydrologic
systems” (p. 20). The urbanization process of a city promotes the removal of original
vegetation coverage, increases imperviousness and promotes the occupation of flood
plains (Miguez et al. 2007). Urban flood management does not necessarily mean the
absolute prevention of floods. It is a concept that goes hand in hand with stormwater
management which is an integral part of any community. However, due to the impacts
of urbanization, the unpredictable nature of floods and restrictions in space in urban
areas in the future, the management for future floods calls for new sustainable and
integrated approaches (Jha et al. 2012).
Increasing infiltration through improved permeability in urban areas or
detention/retention of stormwater at source can reduce flood risk (Jha et al. 2012).
Moreover, the natural landscapes such as wetlands naturally serve the purpose of
storing flood water, acting as both detention and retention basins (Brody et al. 2007).

2.2 Importance of Wetlands

As a significant result of urbanization, the increase in impervious surfaces contributes
greatly to flood occurrences and there is a positive correlation between impervious
surfaces and frequency of flood events (Brody et al. 2008; Cheng et al. 2014; Leopold
1968). One of the most significant ways in which impervious surfaces can aggravate
flooding is through the alteration or elimination of naturally occurring wetlands (Brody
et al. 2008).
 Multifunctional Landscapes for Urban Flood Management 59

Wetland is identified as an ecosystem service that helps to control flooding and

erosion, replenish the ground water table and enhance biodiversity (Niering 1970).
Several early research done on wetlands and flooding demonstrated the ability of
wetlands to store water (Cheng et al. 2014; Brody et al. 2008, 2007). Wetlands not only
provide the ecological infrastructure for watershed systems but also offer natural flood
mitigation by maintaining a properly functioning water cycle (Brody et al. 2008). Lack
of knowledge and awareness on the importance of wetlands has resulted in their
significant reduction.
While it is likely that the rate of flooding will increase due to large developments and
impervious surfaces, results of research prove that wetland alteration in a small-scale has
a greater impact on flooding in the long run (Brody et al. 2007). The results of the research
conducted by Brody et al. (2008) confirm that “the alteration of naturally occurring
wetlands is the most important built environment indicator of flood damage” (p. 13).

2.3 Landscape Architectural Approach

It is now evident that due to development of cities, there is a great increase in the impacts
on the natural environment and to collectively achieve both economic growth and
environmental protection, sustainable management through landscape needs to be
implemented. This includes not only addressing water quantity and flow issues but
integrating society and enhancing the character of the area while managing floods (Cheng
et al. 2014). At present, there is a challenge to move away from traditional approaches
toward a more sustainable approach. For this purpose, “landscape infrastructure is being
explored as a concept that expands the traditional planning and design strategies towards
the realization of a multifunctional system” (Bacchin et al. 2014, p. 2).
It is also evident that existing drainage infrastructure, built to manage urban runoff,
will become obsolete in the near future due to climate and urbanization uncertainties
(Bacchin et al. 2014). Hence, water management systems need to be designed to be
adaptable to change and possibly with a multi purpose use. Furthermore, in order to
move away from the common notion that cities and floods cannot co-exist, a shift in
perception and creative design and planning of cities is required. As an example, open
spaces become multifunctional to convey and store flood water during rainfall events
(Turner 2016; Bacchin et al. 2014).

2.4 Multifunctional Landscapes

As the term implies, multifunctionality is the idea that a single space can have multiple
uses in order to benefit both people and their environment (Macfarlane 2007). In dense
urbanized environments, it is not always easy to find suitable areas for new stormwater
infrastructure. In such situations a more effective option is the use of multifunctional
landscapes where city revitalization and flood control can go hand in hand (Rezende
et al. 2011; Miguez et al. 2007). Hence, a multipurpose landscape not only manages the
quantity of water but also focuses more on the social integration and aesthetic quality of
the landscape (Wong and Eadie 2000). This forms a parallel argument to that of Van
Herk et al. (2013) who state that we should move from acting against water, to living
with it. Further research provides a direct connection to multifunctionality of space as
60 D. N. Denipitiya and S. Udalamaththa

Liao (2012) states, “the theory of urban resilience to floods challenges the conventional
wisdom that cities cannot live without flood control” (p. 01).
Among several scholarly interpretations of multifunctional landscapes, the basic
idea is that a given area should fulfil different functions and objectives. It is to treat the
problem at its source by making use of the urban landscape and recovering natural flow
conditions while enhancing urban space through an integrated design of built and
natural environment (Miguez et al. 2007).
As Lawson (2014) states “Most of the time we are not under flood conditions - so
why not design the landscape to be multifunctional and accrue benefits?” (p. 51).
Further, Kato and Ahern (2009) argue that multifunctional landscapes including eco-
logical greenways and mixed use urban developments, which provide ecological,
recreational, cultural, aesthetic or other purposes are “an important and necessary
component of sustainable landscapes” (p. 1). However, according to a study done by
Cesar et al. (2007) on the use of public squares as flood control measures, one major
challenge that had to be overcome was to develop public acceptance and convince the
local community of its importance.

2.5 Principles of Urban Flood Management for Multifunctionality

Several principles for urban flood management have emerged through studies and
implementation of ideas. They take into consideration the theories of urban resilience
and multifunctionality as well as the enhancement of the landscape.

2.5.1 Best Management Practices

Best Management Practices (BMP) is a flood management principle associated with the
concept of green infrastructure. BMP include both structural strategies and non-
structural strategies to manage stormwater runoff (Cheng et al. 2014). They focus on
providing effective sustainable solutions for water management while ensuring better
community appreciation. The structural strategies include;
• Rain Gardens
• Vegetated Swales
• Pervious pavement
• Filtration-Storage system
• Detention Basin

2.5.2 East Asian Water Management Principles

In the East Asian context, the research done by Luo et al. (2015) provides compre-
hensive details on the Japanese and Chinese flood management systems. In Singapore,
for the purpose of water management, they follow a holistic and strategic plan which
follows the vision of converting engineered concrete drains, canals and reservoirs into
recreational (multifunctional) channels, and manage them to supply water and mitigate
floods (Hua and Wen 2011). Further, many countries such as Singapore and China,
address the need for outdoor spaces for social interaction through water front devel-
opment. This is carried out in a multifunctional manner where these developments also
help the city adapt to flood situations while creating visually pleasing landscapes.
 Multifunctional Landscapes for Urban Flood Management 61

• Integrated river development

• Green dams and embankments
• Social integration
• Urban regeneration

2.5.3 Blue-Green Cities

Constructed water channels in cities carry away urban rainfall, which nature can do
more aesthetically and at no cost. Dekay and O’Brien (2001) express, “Imagine a ‘blue-
way network’ where streams meander through our neighbourhoods, carrying away
runoff and slowing its flow, providing rich natural habitat…” (p. 22). This principle,
known as the concept of ‘Blue-Green Cities’, focuses mainly on recreating the natural
water cycle and contributing to the aesthetic development and amenity of the city. It
creates a bond between water management and green infrastructure (Turner 2016;
O’Donnell 2016). Blue-Green infrastructure aims to reduce the amount of water that is
directed towards artificial drainage systems through natural means such as infiltration,
interception, transpiration and temporary or long-term storage. Hence working with
nature to control water at source (Lawson 2014).
• Multifunctional spaces
• Stormwater retention/detention
• Living with floods

2.6 Theoretical Framework

As a summary of the literature review, to determine the degree of multifunctionality the
following framework has been identified to analyze the selected case study (Fig. 1).

Fig. 1. Theoretical framework

62 D. N. Denipitiya and S. Udalamaththa

3 Methodology

The study adopted a mixed method which is twofold; a literature survey and case study.
Concepts and arguments related to the integration of multifunctionality for flood
management were identified through a literature survey. The factors which assist in
determining the degree of multifunctionality of flood responsive landscapes are sum-
marized and applied to analyze the case study. The public open space developments in
the Kotte wetland around the Diyawanna Oya, Sri Jayawardenepura Kotte in lieu of
flood mitigation and landscape development were identified for further study.
Data was collected through existing literature, mapping, on-site observations,
documentation and interviews. Both quantitative and qualitative methods are used to
analyze the data. However, the effectiveness of multifunctional landscapes is discussed
qualitatively.

3.1 Limitations
The analysis only focuses on analyzing the multifunctionality of the landscape. The
study does not include a quantitative analysis of flood such as water quantity and rate of
runoff and does not discuss in detail the ecological details such as the diversity of flora
and fauna, and water quality. The social survey was also limited to 30 respondents.

4 Case Study

The city of Colombo situated in the wet zone of Sri Lanka and experiences rain year-
round. Flooding in Colombo has worsened in recent years primarily due to the city’s
inability to absorb water as the natural flood storage capacity has been reduced by
inadequate planning and wetland encroachment. The most recent major floods in the
year 1992 recorded the highest ever rainfall (493 mm) followed by the years 2010 and
2016.
As a result of several studies, five wetlands including the Kotte Wetland and a
“green belt” around the Diyawanna water body were established as flood detention
areas in Colombo (CEA and Euroconsult 1995). The recent developments in the green
belt aimed to preserve the flood detention area through sustainable and social landscape
developments, which are considered significant for flood mitigation in Colombo
(Fig. 2).

4.1 Kotte Wetland

Colombo is historically a large area of wetland and natural landscapes. As a result of
imbalanced development of Colombo and mismanagement of water related landscapes,
these wetlands have now depleted. Diyawanna Oya and the Kotte wetlands are some of
the last remaining extensive areas of wetland in Colombo having the capacity to
mitigate floods. The green belt of wetlands around the Oya act as temporary storage
basins for flood water. According to the present land use, the area of wetland has
reduced to approximately 230 Ha. Hence, for the purpose of conserving the remaining
 Multifunctional Landscapes for Urban Flood Management 63

Green belt
. . .... . Catchment area

Fig. 2. Colombo flood detention areas (Source: CEA and Euroconsult 1995)

wetlands and flood detention areas, the local authorities have taken measures to prevent
further landfill and encroachment. This has been implemented through multifunctional
landscape developments in the area (Fig. 3).
Diyatha Uyana
The present location of Diyatha Uyana has undergone several landscape changes in the
past. It was initially a natural wetland area, then infilled to construct a golf course. In
2010, this area was developed as a public recreational space while serving purpose of a
flood detention area. It includes a waterfront development, recreational park and sales
area, and acts as the center for local horticultural sales.
Diyawanna Oya Linear Park
With the development of the city for improved flood retention due to the flood of 2010,
the waterfront area around Diyawanna Oya facing the parliament was developed as a
linear park with jogging tracks while increasing the retention capacity of the Oya. This
was developed to serve the purpose of flood retention, recreation and health benefits for
the community.
Beddagana Wetland Park
This park is located close to the parliament complex, bordering the Diyawanna Oya. It
is a wetland protected within a wildlife sanctuary and classified as a Wetland Protection
Zone due to its high bio-diversity and bird habitats, home to many endemic birds and
other faunal species. This area was developed as a wetland park, enhancing the natural
environment and bio diversity through recreational and educational facilities, while
maintaining and preserving it for flood retention.
64 D. N. Denipitiya and S. Udalamaththa

DIYATHA
UYANA

BEDDAGANA
WETLAND
PARK

DIYAWANNA
OYA LINEAR
PARK

Fig. 3. Map of study area and selected landscape developments

5 Data Presentation and Analysis

Landscape Architectural solutions for flood management could have subjective out-
comes. Hence, determining the multifunctionality of the selected parks takes into
consideration these variations and provides a qualitative outcome.

5.1 Wetland Footprint Variation

Sri Jayawardenepura Kotte was home to more than a 1000 Ha of wetland before the
administrative capital was shifted here. The dredging of the wetlands for the con-
struction of the parliament complex and surrounding lake in the years 1979–1983
resulted in reclamation of about 272 Ha of natural flood retention area. Since then this
area has seen a rapid decrease in wetlands (CEA and Euroconsult 1995) (Fig. 4 and
Table 1).
It is evident that within a period of 17 years, the amount of marshes and agricultural
fields which contribute to natural flood retention have decreased by almost half.
Although there is a significant increase in the area of water bodies, due to the flood
management and development activities that took place after 2010, this increase does
not adequately compensate for the loss of wetlands as evident through the recent flood
 Multifunctional Landscapes for Urban Flood Management 65

Fig. 4. Graphical illustration of land use

Table 1. Land use variation in the study area

1999 (Ha) % 2007 (Ha) % 2016 (Ha) %
Wetlands 391 21 314 16.9 233 12.5
Water bodies 91 4.9 107 5.8 145 7.8
Reclaimed land 44 2.4 31 1.6 30 1.6
Built area 1334 71.7 1408 75.7 1451 78.1
Total 1860 100 1860 100 1860 100

event of 2016. The built area as expected shows an increase. However, the rate of
increasing built area has decreased.
Although many steps were taken during the last few years to conserve what is left
of the natural wetlands, for the many benefits we gain from this ecosystem and also to
manage the increasing flood risk, the loss of wetlands have not decreased. The
developments that took place have somewhat succeeded in preserving the flood
detention in the public open spaces. However, a larger area has been dredged without
preserving what is left of the wetland or creating man-made wetlands to compensate for
the lost wetland area. Hence, there is an imbalance between open water and marsh
wetlands when considering their flood retention capacity.

5.2 Functions and Activities

This factor will be discussed separately for each park.
Diyatha Uyana
Table 2 indicate the activities in this public open space. Most functions in this public
space have been designed to incorporate supplementary functions. This is a good
application of multifunctional landscapes (Miguez et al. 2007). Most of these sec-
ondary functions cater to social requirements, hence, it is evident that the local com-
munity has been benefitted through this development.
66 D. N. Denipitiya and S. Udalamaththa

Table 2. Functions at Diyatha Uyana

Main function / Elements Supplementary functions

Water retention lake Rowing track
Sea plane track
Retention ponds and constructed wetlands Landscape feature
Jogging track Scenic walkway
Street art
Water front seating and lawn Inundation space for major floods
Children’s play area

Vehicle parking Public performance space

Children’s play space
Food outlets and seating
Weekly sales outlets Social gathering space
Seating area
Boat ride

Diyawanna Oya Linear Park

This park was developed on the idea of increasing the water retention capacity of the
lake and a water front development to provide recreation and health benefits to the
community (Table 3).

Table 3. Functions at the Linear Park

Main function / Elements Supplementary functions

Jogging track along Diyawanna Oya Inundation area for high water level of
the lake
Waterfront seating Flood inundation area / Water deten-
tion
Parking Flood inundation area / Water deten-
tion
Open play space Exercising space
Activities related to paddy farming
Food vendors
Paddy fields Water detention
Natural landscape
Jogging/cycle track along paddy fields
Exercising space
Dry weather children’s play ground Flood water detention space

Canal for excess runoff drainage Landscape feature

 Multifunctional Landscapes for Urban Flood Management 67

Beddagana Wetland Park

The Beddagana wetland park functions more as a nature park for observation and
education and does not include a wide variety of activities (Table 4).

Table 4. Functions at Beddagana Wetland Park

Main function / Elements Supplementary functions

Jogging path
Boardwalk
Bird watching
Dry weather play ground Flood water detention space
Bird watching
Outdoor learning Outdoor seating
Nature trail Scenic walkway
Open viewing decks
Vehicle park
Canal as physical barrier Landscape feature

The main activities that have been integrated into the design concerns involvement
of the public in effectively conserving the wetland. This ensures that the natural wet-
land is protected through multifunctionality combining both the hydrological/
ecological benefits of wetland and social awareness and appreciation.
Social Responsiveness
The social responsiveness to the public landscapes of Diyatha Uyana, Diyawanna Oya
linear park and Beddagana wetland park were studied through a questionnaire mainly
by the frequency and reasons of visit, and overall perception of the landscape (Table 5).

Table 5. Frequency of visit by respondents

Public landscapes No of respondents
Frequently Occasionally Total
Diyatha Uyana 8 22 30
Diyawanna Oya linear park 8 14 22
Beddagana wetland park 2 13 15

Diyatha Uyana is a park with diverse activities that cater to a wider range of users
and it is clearly the more popular public space. It is evident that these multifunctional
landscapes have been accepted by the community as 87% of the respondents stated that
they prefer spending time in public open spaces (Fig. 5).
68 D. N. Denipitiya and S. Udalamaththa

Other 7%

Play area 23%

Natural landscape features 27%

Eating facilities 17%

Commercial activities 50%

Jogging/Exercise 77%
Spend time with family
/friends 53%

0% 20% 40% 60% 80% 100%

Fig. 5. Reason for visiting public open spaces

5.3 Landscape Elements and Technology

This factor will also be discussed separately for each park.
Diyatha Uyana
The landscape elements and material used in the design complement its functions in
most instances and contribute to create flood resilience (Table 6).

Table 6. Materials and landscape elements used at Diyatha Uyana

Landscape elements / uses Materials / technology used

Walking/jogging track Sand paved, tar bitumen, concrete paving
stone
At a higher elevation to water level of the
lake
Seating area Maintained lawn
Closer to the water level of the lake
Open viewing deck Timber with gaps, railway sleepers
Retention ponds Grass slope (non-vegetated)
Constructed wetlands Curb stone, grass slope
Dredged lake Grass, steep slope
Parking and commercial area Concrete paving
Children’s play area Sand

There are three main walkways designed at different elevations (Fig. 6). The sur-
rounding lawn and seating area along the edge of the Oya is at a lower level -
 Multifunctional Landscapes for Urban Flood Management 69

approximately 1 m above the mean water level. In the event of a high precipitation, this
area will get inundated creating a new high-water level (Fig. 7). It is similar to the use
of a green dam/embankment for natural flood control (Luo et al. 2015). This multi-
functional use of the open space enables floods to occur without causing great
inconvenience in this particular area.

Fig. 6. Different levels of walkways and multifunctional embankments

Fig. 7. Seating area inundated during the Fig. 8. Retention pond surrounded by well-
flood of 2016 maintained lawn

Most materials used for walkways are not permeable and will further contribute to
the increasing impervious surfaces in the city, which will cause further surface runoff.
This water will potentially runoff into the lake or the constructed wetlands and con-
tribute to unnecessary increase in the water level of the Diyawanna Oya.
The adjacent part of the Diyawanna lake has been dredged to increase the water
retention capacity but there are also problems of silting. The retention ponds and
constructed wetlands directly contribute to flood management (Fig. 8). However, they
are regularly maintained thereby, disturbing ecological succession.
70 D. N. Denipitiya and S. Udalamaththa

Fig. 9. Main element of the design – the Fig. 10. Stormwater canal integrated into the
Jogging track landscape

Diyawanna Oya Linear Park

Most activities and elements incorporated in this development are designed to be
multifunctional in events of high precipitation (Table 7).

Table 7. Materials and landscape elements used at Diyawanna Linear Park

Landscape elements / uses Materials / technology used

Walking/jogging track Sand paved, compacted earth
Intermediate level between water level
and road
Seating area Maintained lawn
Retention ponds Grass, steep slope
Carpark Concrete paving
Open play space Compacted earth
Dry weather children’s play area Sand, natural ground
Canals Grass slopes
Drains for runoff from roads Concrete
Direct water to nearest waterbody

Figure 9 shows the main element in this area, the jogging track. Most part of it
including the carpark has been infilled. They are at a lower level than the main road and
act as a high-water level for the lake in the event of a flood. In this case, the main road
acts as the embankment for the city, and the jogging track carpark and most seating
areas become part of the detention basin.
The paddy fields in the vicinity act as natural flood detention areas. However, due
to the significant absence of natural or constructed wetland or any other BMPs, it is
possible that excess water will continuously flow downstream and not be detained at
 Multifunctional Landscapes for Urban Flood Management 71

source. The dry weather playground incorporates multifunctionality. It is surrounded

on one side by a stormwater canal and on the other by an existing retention pond
(Fig. 10). Hence, in the instance of high precipitation, this playground will be inun-
dated converting it to a flood detention landscape. The blue green principles are
implemented in the design through the use of open drains leading to natural water
bodies.
Beddagana Wetland Park
The Beddagana park focuses on conserving the natural wetland to maintain flood
detention capacity (Table 8).

Table 8. Materials and landscape elements used at Beddagana Wetland Park

Landscape elements / uses Materials / technology used

Access walkway/jogging track Compacted earth
Closer to wetland level, lower level than the main
road
Carpark Concrete paving
Raised walkway Timber with gaps
Minimum impact on wetland
Dry weather play ground Grass lawn
Maintains natural edge
Bird watching tower Steel columns
Minimum impact on wetland
Bund roads/ internal pathways Compacted earth, gravel
Open viewing decks Timber with gaps

This wetland is subject to periodic flooding lasting for 3–7 days (Uni Consultancy
Services 2011). Hence it has been designed without obstructing the existing drainage
paths and has ensured the least impact to the wetland so that the flood retention and
detention capacity is not compromised. The jogging track, internal nature trails and
existing bund roads which have been integrated into the design are of compacted earth
or are gravel-paved, thereby encouraging ground water absorption. The raised timber
walkways ensure minimum impact on the natural landscape. Similar treatment has also
been given to the other built structures.
The dry weather playground as well as most part of the wetland acts as a detention
basin for flood water at source. However, the area for parking has taken up a significant
area of wetland. Although they are essential for a public space, BMPs such as per-
meable paving and vegetated swales could have been introduced.
72 D. N. Denipitiya and S. Udalamaththa

6 Conclusion

Based on the factors to assess the multifunctionality of the landscape developments in

the Kotte wetland and Diyawanna Oya, the research attempted to determine if the three
selected parks were multifunctional in terms of its ability to resist urban flooding.
All three landscape developments made use of multifunctionality in their designs,
incorporating principles of urban flood management, integrating multifunctional spaces
rather than single-use spaces and multifunctional elements. The use of technology
similar to embankments is an effective method to create resilient landscapes. In the
Diyatha Uyana and Linear park, there is a concern that the dredging of large areas for
lakes will not detain the expected amount of water. There is a possibility of down-
stream flow due to the loss of natural vegetation and wetland on the edges of the water
bodies. Further, these developments have been implemented for a 10-year flood return
period and will need to be constantly improved and maintained. Hence, they may not
be effective in the case of a 50-year flood but resilient to minor floods. The Beddagana
park, however, is a good solution conserving valuable wetland that could be encroa-
ched in the future. This landscape design effectively integrates the natural hydrological
functions of the area to ensure an effective and flood resilient landscape design.
By analyzing the social responsiveness for multifunctional open space develop-
ments, it is evident that people interact with these public open spaces and that it
addresses public recreation, the health and well-being. However, the community needs
to be given awareness on the objectives of such developments in order to have a better
a sense of ownership of public landscapes. Taking into consideration an overall per-
spective, these landscapes effectively address the multifunctional use of the landscape
for flood mitigation as well as social integration.
The landscape developments analyzed in the study area all aim to conserve the
natural flood retention capacity, but in most cases have focused on increasing the water
retention capacity of the Diyawanna Oya. The use of constructed vegetated wetlands
would have been a more effective water management system for flood detention.
Although an effort has been taken to maintain the water retention capacity of the natural
environment, it has not reached a satisfactory level for flood management due to the
continuous loss of natural wetland as proved by the findings.
Further, no planning solution has been given for the inundation of residential areas
caused by urbanization and development of property. To mitigate this, multifunctional
use and BMPs such as vegetated swales, rain gardens and filtration-storage systems to
detain water at source should be further encouraged in private developments. This will
not only help in flood management but will integrate the community in the process
(Fig. 11).
With further studies in this field, future developments for sustainable flood man-
agement by means of multifunctional landscapes will have the potential to successfully
manage any degree of floods, thereby establishing flood resilient cities.
 Multifunctional Landscapes for Urban Flood Management 73

Fig. 11. Use of BMP’s in residential areas to integrate the community in flood mitigation

References
Bacchin TK et al (2014) Green-blue multifunctional infrastructure: An urban landscape system
design new approach. S.N., Sarawak, pp 1–8
Brody SD, Highfield WE, Ryu HC, Spanel-Weber L (2007) Examining the relationship between
wetland alteration and watershed flooding in Texas and Florida. Nat Hazards 40(2):413–428
Brody SD et al (2008) Identifying the impact of the built environment on flood damage in Texas.
Disasters 32(1):1–18
CEA and Euroconsult (1995) Wetland conservation project - Colombo flood detention areas.
CEA, Colombo
Cesar F et al (2007) Comparison of different multifunctional landscapes approaches for flood
control in developing countries. NOVATECH, pp 83–90
Cheng C, Brabec EA, Yang Y-CE, Ryan RL (2014) Rethinking stormwater management in a
changin world: effects of detention for flooding hazard mitigation under climate change
scenarios in the Charles river watershed. In: Li M (ed) Landscape research record, pp 214–
228
Dekay M, O’Brien M (2001) Gray city, green city. In: Forum for applied research and public
policy, pp 19–27
Hua D, Wen LY (2011) Managing sotrmwater in urbanised Singapore for flood control. PUB,
Tokyo
Jha AK, Bloch R, Lamond J (2012) Cities and flooding: a guide to integrated urban flood risk
management for the 21st century. The World Bank, Washington, DC
Kato S, Ahern J (2009) Multifunctional landscapes as a basis for sustainable landscape
devlopment. J Jpn Inst Landscape Architect 72(5):799–804
Lawson E (2014) Managing urban flood risk: The blue-green appraoch. http://www.
bluegreencities.ac.uk/bluegreencities/documents/publicsciencelecture-dec2014-bgc.pdf
Leopold LB (1968) Hydrology for urban planning: a guidebook on the hydrologic effects of
urban land use. USGS Circular
Liao K (2012) A theory on urban resilience to floods—a basis for alternative planning practices.
Ecol Soc 17(4). http://www.jstor.org/stable/26269244
74 D. N. Denipitiya and S. Udalamaththa

Luo P, He B, Takara K, Xiong YE, Nover D, Duan W, Fukushi K (2015) Historical assessment
of Chinese and Japanese flood management policies and implications for managing future
floods. Environ Sci Policy 48:265–277. https://doi.org/10.1016/j.envsci.2014.12.015
Macfarlane R (2007) Multifunctional landscapes: conceptual and planning issues for the
countryside. In: Benson JF, Roe M (eds) Landscape and sustainability, 2nd edn. Taylor &
Francis Group, New York, pp 138–166
Miguez MG, Mascarenhas FC, Magalhaes LP (2007) Multifunctional landscapes for urban flood
control in developing countries. Int J Sustain Dev Plan 2(2):153–166
Niering WA (1970) The ecology of wetlands in urban areas. In: Dansereau P (ed) Challenge for
survival: land, air, & water for man in Megalopolis. Columbia University Press, New York
O’Donnell E (2016) University of Nottingham blogs/blue green cities. http://www.blogs.
nottingham.ac.uk
Rezende OM, Carneiro PR, Miguez MG (2011) Sustainable stormwater management: use of
multifunctional landscapes in urban drainage for flood control. S.N., Porto Alegre, pp 1–8
Turner TF (2016) How landscape change causes flooding in towns and cities. http://www.
landscapearchitecture.org.uk
Uni Consultancy Services (2011) Environmental screening report for establishment of
Beddagana and Kotte ramparts parks
Van Herk S, Rijke J, Zevenbergen C, Ashley R (2013) Understanding the transition to integrated
flood risk management in the Netherlands. Environ Innov Soc Trans 15:84–100
Wong THF, Eadie ML (2000) Water sensitive urban design - a paradigm shift in urban design.
In: 10th world water congress
Zevenbergen C et al (2010) Urban flood management. Taylor & Francis Group, Florida
Zevenbergen C, Veerbeek W, Gersonius B, Ven Herk S (2008) Challenges in urban flood
management: travelling across spatial and temporal scales. J Flood Risk Manag 1:81–88
 Single and Binary Adsorption of Heavy Metal
Ions from Aqueous Solutions Using Coirpith
as the Adsorbent

E. N. Ranasinghe and B. M. W. P. K. Amarasinghe(&)

Department of Chemical and Process Engineering, University of Moratuwa,

Moratuwa, Sri Lanka
[email protected]

Abstract. The adsorption of nickel and zinc ions onto coir pith was studied
using batch experiments at 30 °C and 40 °C. Adsorption kinetics and equilib-
rium isotherms were obtained for both single component adsorption and binary
adsorption. The kinetics of single system and binary system adsorption of nickel
and zinc was studied using Lagergren pseudo first order model, pseudo second
order model and the intraparticle diffusion model. The experimental data
showed all of the systems could be described using pseudo second order model.
In the single system the coir pith had a higher capacity for zinc than nickel. In
both the nickel and zinc single systems the adsorption decreased when tem-
perature increased from 30 °C to 40 °C. In the binary system at 30 °C the coir
pith had a larger capacity for zinc adsorption then nickel but at 40 °C coir pith
had a larger capacity for nickel than zinc. Equilibrium data were satisfactorily
fitted to Langmuir and Freundlich isotherms. When comparing the adsorption
capacity, qmax, the experimental data showed the binary system had a lesser
capacity (qmax 13.6 mg g−1) than the single systems nickel (qmax 27 mg g−1)
and zinc (qmax 22.7 mg g−1) under similar conditions. This suggests the com-
petitive nature of the systems decreases the overall adsorption.

Keywords: Binary adsorption
Heavy metal
Nickel
Zinc
Coir pith

1 Introduction

Heavy metals, such as zinc, cadmium, nickel and lead occur naturally in very low
concentrations in both soil and water as dissolved metals ions and in their native forms.
However due to increased industrial and agricultural activity and poor management of
waste produced there has been an increase in heavy metals introduced into the envi-
ronment. This can lead to many harmful effects to aquatic environments, plants and
human health as in many developing countries such as India and Sri Lanka. In reviews
by Wijayawardhana et al. (2016) and (Adikaram and Pitawala 2017), it was found that
the reported levels of heavy metals found in Sri Lankan soil were within the safe levels
for daily activity. However as heavy metals do not degrade overtime they tend to build
up in water and soil, especially in coastal regions.
The toxicity of the heavy metals depend both on the chemistry of the metal and the
concentration in water and soil. Nickel and zinc are harmless in low concentrations but

© Springer Nature Singapore Pte Ltd. 2020

R. Dissanayake and P. Mendis (Eds.): ICSBE 2018, LNCE 44, pp. 75–90, 2020.
https://doi.org/10.1007/978-981-13-9749-3_7
76 E. N. Ranasinghe and B. M. W. P. K. Amarasinghe

the presence of these chemicals in fertilizers, pesticides and galvanized items have lead
to higher concentrations in natural water sources such as rivers and nearby wells. One
method of removing these heavy metals which is appealing due to its easy of appli-
cation is adsorption, where the pollutant is adsorbed onto the adsorbent and then
filtered out with the solid adsorbent. The efficacy of the adsorbent depends on the
surface chemistry and structure of the adsorbent. Ideally the adsorbent should be both
cheap and readily available. A cheap alternative is to use agricultural waste such as rice
husks, vegetable peel, straw or coir pith as bioadsorbents. In a study by Amarasinghe
(2007) coir pith and tea waste was shown to have a higher capacity for metals such as
lead compared to activated carbon.
In a review by Nassrulla et al. (2016) it was shown that lignin can be a good
nonspecific adsorbent to remove heavy metal pollutants from waste water. (Ho and
McKay 1999) showed peat was an effective adsorbent in the removal of nickel and
copper ions from aqueous solutions. In a study by Amarasinghe (2011) showed that
coir pith can be used to remove heavy metals lead and cadmium from waste water.
Based on previous research and the low cost and availability, for this study coir pith
was chosen as the adsorbent. Coir pith is made up of mostly lignin and cellulose, which
are organic polymers which contain many polar surface functional groups such as
esters, ethers and carbonyl groups.
Nickel and zinc are common heavy metals used in the electroplating industry as
well as being present in many fertilizers and pesticides used in agriculture. Nickel and
zinc often occur together both naturally and in industrial applications. Zinc and nickel
have similar atomic mass, ionic size and density. Studies of the competitive adsorption
of similarly sized metal ions by (Sdiri et al. 2014) which compared zinc and copper and
Shaheen et al. (2012) which compared copper, nickel and zinc showed that small
differences in ionic size can result in large differences in adsorption capacity and
preference.
This study will focus on the adsorption of nickel and zinc in both single system and
binary system environments. The adsorption isotherm will be studied at 30 °C and the
kinetics of adsorption will be studied at 30 °C and 40 °C. From the experimental
analysis the change in the kinetics of the adsorption at the two different temperatures as
well as the mechanism of adsorption of nickel and zinc will be deduced.

2 Methods and Materials

2.1 Preparation of the Adsorbent

Coir pith was obtained from Sampath Coir Mill, Singakkuliya, Sri Lanka and dried in
the oven using stainless steel trays at 60 °C for 12 h. It was sieved using a sieve shaker
to obtain particles with a size distribution between 335 lm and 1 mm. The coir pith
was washed with distilled water and then dried in the oven at 60 °C for 12 h.
 Single and Binary Adsorption of Heavy Metal Ions 77

2.2 Characterization of the Adsorbent

To determine the surface functional groups, Fourier transform infrared spectrometer
(FTIR) analyses were performed. The coir pith was imaged using scanning electron
microscopy (SEM) at 500 magnification.

2.3 Batch Adsorption Test

The adsorbate stock solution of 1000 mg L−1 Ni(II) and Zn(II) were prepared from
analytical grade nickel nitrate Ni(NO3)2.6H2O and zinc nitrate Zn(NO3)2.6H2O (sup-
plied by Sigma-Aldrich chemicals, Germany) respectively in distilled water.

2.3.1 Adsorption Isotherm

For the isotherm study, 500 ml solutions of Ni, Zn or a mixture of both in a 1:1 ratio
with initial concentrations of 10 mg L−1, 20 mg L−1, 50 mg L−1, 100 mg L−1 and
200 mg L−1 were made by diluting stock solutions of 1000 mg L−1 using distilled
water. 0.75 g of the coir pith adsorbent was added to the solution and stirred using an
agitator system at a speed of 135 rpm at 30 °C for 120 min. The sample concentrations
were determined using inductively coupled plasma mass spectrometry (ICP-MS).

2.3.2 Adsorption Kinetics

For the study of kinetics over different temperatures 500 ml solutions of nickel or zinc
or a mixture of both in a 1:1 ratio with total initial concentrations of 50 mg L−1. 0.75 g
of the coir pith adsorbent was added to the solution and stirred constantly. Samples
were taken out at times of 1 min, 2 min, 5 min, 10 min, 30 min, and 120 min. The
experiments were carried out at 30 °C and 40 °C. The concentrations of the samples
were determined using ICP-MS.

2.4 Analytical Models

2.4.1 Isotherm Models
Several adsorption isotherm models have been developed, and are widely used, to
describe adsorption monolayer formation.
Langnuir isotherm model assumes a complete monolayer coverage of the solid
adsorbent (Langmuir 1916) and is given by Eq. 1, where qe is the equilibrium
adsorption capacity, qmax is the maximum adsorption capacity corresponding to com-
plete coverage of available sites, Ce the equilibrium concentration of metal ions in
solution, kL is the Langmuir isotherm constant (L/mg).

qmax kL Ce
qe ¼ ð1Þ
1 þ kL Ce

Equation 1 can be rearranged to the linear form expressed in Eq. 2.

Ce 1 Ce
¼ þ ð2Þ
qe kL qmax qmax
78 E. N. Ranasinghe and B. M. W. P. K. Amarasinghe

The Freundlich isotherm is an empirical equation which takes into account

heterogeneous adsorption (Freundlich 1906a, b). The Freundlich isotherm equation is
described in Eq. 4 where qe is the equilibrium adsorption capacity, Ce is the equilib-
rium concentration, kF is the Freundlich constant which is the relative adsorption
capacity of the adsorbent, n is the Freundlich constant which indicates adsorption
intensity.
1
qe ¼ kF
Cen ð3Þ

Equation 3 can be rearranged to the linear form expressed in Eq. 4. The isotherm
can be obtained by plotting log qe against log Ce.

1
logqe ¼ logkF þ logCe ð4Þ
n

Extended Langmuir model is used to analyse the isotherm of the binary system.
This model takes into consideration the competitive forces that occur during adsorption
(Singh et al. 2017). qe;Ni is the equilibrium adsorption capacity of nickel at equilibrium,
Ce;Ni is the equilibrium concentration of nickel, kL;Ni is the Langmuir constant for
nickel, qe;Zn is equilibrium adsorption capacity of zinc at equilibrium, Ce;Zn is the
equilibrium concentration of zinc, kL;Zn is the Langmuir constant for zinc, and qmax is
the maximum capacity for the binary system.


qmax KL;Ni Ce;Ni
qe;Ni ¼
ð5Þ
1 þ KL;Ni Ce;Ni þ KL;Zn Ce;Zn


qmax KL;Zn Ce;Zn
qe;Zn ¼
ð6Þ
1 þ KL;Ni Ce;Ni þ KL;Zn Ce;Zn

2.4.2 Kinetic Models and Thermodynamic Analysis

Lagergren pseudo-first model has been used by many researches to study adsorption
where the sorbate interaction was not significant, and usually modelled only the initial
adsorption accurately (Kumar et al. 2005). The pseudo-second order model was shown
to be accurate in modelling the whole adsorption process.
Lagergren pseudo first order which was proposed by Lageregen (1898), describes
physisorption and can be derived from the integrated rate law (Eq. 8) where k1 is the rate
constant of first order sorption, qe is the amount of solute adsorbed at equilibrium (mg
g−1), qt is amount of solute adsorbed on the surface of the sorbent at any time t (mg g−1).

dqt
¼ k 1 ð qe  qt Þ ð7Þ
dt
 Single and Binary Adsorption of Heavy Metal Ions 79

Equation 7 can be rearranged to a linearized form which is more commonly used.

Log(qe ) and k1 can be determined from the intercept and slope of the plot of log
(qe  qt ) against t.

k1
logðqe  qt Þ ¼ logðqe Þ  t ð8Þ
2:303

Lagergren pseudo second order model, which was proposed by Ho (1995) can be
used to study processes where the adsorption reaction is the rate determining step. The
solute and the adsorbent undergo a chemical reaction i.e. chemisorptions takes place.
Lagergren pseudo second order can be derived from the integrated rate law (Eq. 10)
where k2 is the rate constant of adsorption (g mg−1 min−1), qe the amount of divalent
metal ions adsorbed at equilibrium (mg g−1), and qt is the amount of divalent metal
ions on the surface of the adsorbent at any time, t (mg g−1).

dqt
¼ k 2 ð qe  q t Þ 2 ð9Þ
dt

The integrated form of Eq. 9 can be rearranged to its linear for Eq. 10. The values
of k2 and qe can be determined from the intercept and slope from the plot of qtt against t.

t 1 1
¼ 2
þ t ð10Þ
qt k 2 qe qe

Intraparticle diffusion model (Cheung et al. 2007) was used to analyse the
adsorption and involves the transportation of the solute from the aqueous phase to the
solid surface and then a diffusion of the solute into the pores of the solid. This equation
is described in Eq. 11 where qt is the amount adsorbed (mg g−1) at time t (min). The
rate constant kid is directly evaluated from the slope of the regression line. The intercept
c provides information of the thickness of the boundary layer, the larger the intercept
the larger the boundary effect.
1
qt ¼ kid
t2 þ c ð11Þ

The thermodynamic potential of the adsorption systems at two different tempera-

tures were determined using the value of Kd calculated using Eq. 12 and then substituted
into Eq. 13 to calculate DG (J mol−1) which is the change in free energy, R is the gas
constant (8.314 J mol−1 K−1) and T is the temperature (K) (Sharifipour et al. 2015)
qe
Kd ¼ ð12Þ
Ce

DG ¼ RTlnKd ð13Þ
80 E. N. Ranasinghe and B. M. W. P. K. Amarasinghe

3 Results and Discussion

3.1 Characterization of Adsorbent
The surface morphology was examined by scanning electron microscopy (SEM) is
shown in Fig. 1. Coir pith contains pores of various sizes with diameters between 25 to
50 µm. The low density and porous structure of coir pith allows a large surface of the
adsorption of heavy metals.

Fig. 1. SEM image of coir pith

Fig. 2. FTIR spectra of (a) coir pith (b) Ni(II) loaded coir pith (c) Zn(II) loaded coir pith (d) Ni
(II) and Zn(II) loaded coir pith

Fourier transform infrared spectra (FTIR) was used to measure the change in
vibrational frequency in the functional groups of the adsorbent. The adsorbent was
scanned in the wavelength range of 4000–8000 cm−1. Figure 2 shows the FTIR
spectrum of (a) coir pith before adsorbent, (b) the spectrum of coir pith after nickel
adsorption, (c) the spectrum of coir pith after zinc adsorption, (d) the spectrum of coir
 Single and Binary Adsorption of Heavy Metal Ions 81

pith after nickel and zinc binary adsorption. The FTIR analysis shows the presence of
surface functional groups at their corresponding wavelengths; -OH (3000–3500 cm−1),
C-H (2900 cm−1), C = O (1600 cm−1), and C-O (1200–100 cm−1). These peaks cor-
respond to the polar groups of the lignin and cellulose of which coir pith is made of.
The intensity of the OH peak decreases when adsorption takes place, suggesting that
during adsorption the OH bond was broken and the metal ion chemically bonds to the
functional group via the oxygen atom. The most significant decrease was after zinc
adsorption suggesting that coir pith has a greater capacity for zinc than nickel. The
amount of adsorption in the binary system is comparable to that of the nickel single
system, and therefore less than the average of the nickel and zinc single systems. This
suggests that due to the competitive nature of the binary system, adsorption is
suppressed.

3.2 Adsorption Isotherms for Ni and Zn

The adsorption isotherm of single system nickel, single system zinc and total binary
system was obtained by plotting adsorption capacity at equilibrium, qe against equi-
librium concentration, Ce as shown in Fig. 3. The adsorption isotherm of nickel shows
a mostly linear relationship between qe and Ce. The isotherm of zinc shows, qe
increases as the solution concentration increases. The shape of the zinc isotherm
suggests there is an exponential relationship between qe and Ce. The difference in the
isotherms is possibly because the nickel ion is smaller than zinc which therefore allows
the adsorbent to absorb more nickel in its pores than zinc. At higher initial concen-
trations the adsorption in the binary system is much less than that of the single systems.
The competitive nature of the binary system results in less adsorption and possibly less
diffusion into the cavities of the adsorbent.

60
50
40 Zinc
qe /mg g-1

Zinc
30
Nickel
20
Nickel
10
Binary
0
0 100 200 Binary

Ce/ mg L-1

Fig. 3. Adsorption isotherms of single system nickel and zinc with trend lines deduced by fitting
the data into the Freundlich model and the adsorption isotherm of the binary isotherm with the
trend line deduced by fitting the data into the Langmuir isotherm model.
82 E. N. Ranasinghe and B. M. W. P. K. Amarasinghe

8 Nickel
7
Nickel
qe/ mg g-1 6
Zinc
5
Zinc
4
3
2
0 50 100

Ce/ mg L-1

Fig. 4. Adsorption isotherms of the nickel and zinc components of the binary adsorption system.
The regression lines were drawn by fitting the data into the Langmuir model.

The adsorption isotherm of the nickel and zinc components of the binary system
shown in Fig. 4 is a type 1 isotherm where there is a formation of a monolayer
followed by desorption at higher concentrations.

Table 1. Isotherm parameters for the single system and binary adsorption of nickel and zinc
solutions onto 0.75 g of coir pith at 30 °C and 40 °C
Isotherm model Single Binary system
system
Parameter Ni Zn Ni Zn Total
Langmuir
qmax/mg g−1 27 22.7 13.5
kL/L mg−1 0.03 0.14 0.38
R2 0.82 0.83 0.90
Freundlich
kF 0.10 3.17 6.95
1/n 1.32 0.53 0.13
R2 0.98 0.97 0.57
Extended Langmuir
qmax/mg g−1 13.6
kL/L mg−1 0.304 0.345
R2 0.50 0.54

The data from the nickel single system, zinc single system and binary adsorption at
30 °C was fitted to the Lagmuir, and Freundlich isotherm models. The Langmuir
parameters qmax and kL were calculated using Eq. 2. The Freundlich parameters kF and
1/n were calculated using Eq. 4 and the results are shown in Table 1. The maximum
 Single and Binary Adsorption of Heavy Metal Ions 83

adsorption capacity corresponding to complete monolayer coverage and saturation of

all available adsorption sites, qmax, of nickel was 27 mg g−1, and qmax of zinc was
22.7 mg g−1. The Langmuir constant, kL, which is related to the free energy of
adsorption is lower for nickel than for zinc being 0.03 L mg−1 and 0.14 L mg−1
respectively. This suggests that coir pith has a greater capacity for nickel than for zinc.
The Freundlich parameter, kF, which describes relative adsorption capacity, shows
that that of zinc is greater than that of nickel being 3.17 and 0.1 respectively. This
suggests that zinc has a greater affinity for the adsorbent than nickel. The Freundlich
parameter 1/n which is the adsorption intensity, is greater for nickel than for zinc being
1.32 and 0.53 respectively. As both values are greater than 0 the processes are con-
firmed to be chemisorptions. As 1/n of zinc is closer to 0, the curvature of the isotherm
is greater and the process can be interpreted as being more heterogeneous.
In order to calculate qmax using the Freundlich model, the experiment needs to be
carried out using a fixed initial concentration and varied dosages of adsorbent and then
analysed using Eq. 14 suggested by Halsey (1952). The Freundlich expression is an
exponential equation and therefore assumes that as the solution concentration increases,
the concentration of the solute on the adsorbent increases. The value of kF is larger for
zinc but the value for 1/n is larger for nickel. At a lower initial concentrations Co the
qmax of zinc would be higher than nickel while a larger Co would show a larger qmax for
nickel.
qmax
kF ¼ 1 ð14Þ
Con

The data from the total binary system at 30 °C was fitted to the Langmuir and
Freundlich isotherm models while the nickel and zinc components of the binary
adsorption were analysed using the extended Langmuir isotherm model. When fit into
the Langmuir model, it fit well with a R2 value of 0.90 while the qmax value is
13.5 mg g−1. When the binary data was fit into the Freundlich model, the R2 value was
very low at 0.57 which means that the model is not suitable for this system. As the
binary system fit the Langmuir model best, the adsorption in the binary system most
likely forms a monolayer with very little diffusion into the adsorbent. The extended
Langmuir isotherm model calculates the parameters pertaining to nickel and zinc
simultaneously. The qmax for the whole system which includes both nickel and zinc
adsorption was 13.6 mg g−1, which is very similar to the Langmuir. The R2 value is 0.5
and 0.54 for nickel and zinc respectively. This low correlation coefficient suggests that
the extended Langmuir model is unsuitable for this system. Since the Langmuir model
shows a better fit for the data and is simpler to model it would be the better model to
use to analyse this system.

3.3 Kinetic Study of Adsorption of Nickel and Zinc Single System

and Binary Nickel and Zinc Adsorption
Both single system and binary adsorption of nickel and zinc were studied at 30 °C and
40 °C by plotting the mass of metal ion adsorbed onto the adsorbent at time t, qt, was
against time. The data from the adsorption of single systems and binary system nickel
84 E. N. Ranasinghe and B. M. W. P. K. Amarasinghe

45
30 °C
40

qt / mg g-1
35 30 °C
30
25 40 °C
20 40 °C
15
10
5
0 50 100
t / min

Fig. 5. Kinetics study of nickel solution of initial concentration 50 mg L−1 onto 0.75 of coir pith
at 30 °C and 40 °C. The regression lines were drawn by fitting the data to Lagergren pseudo
second order model.

19
17
qt/ mg g-1

15
30 °C
13
11 30 °C
9 40 °C
7 40 °C
5
0 50 100
time / min

Fig. 6. Kinetic study of zinc solution of initial concentration 50 mg L−1 at 30 °C and 40 °C. The
regression lines were drawn by fitting the data to the Lagergren pseudo second order model.

and zinc at 30 °C and 40 °C were analysed using Lagergren first order, Lagergren
second order models and intraparticle diffusion model to study the mechanism of
adsorption. The kinetic parameters qe,1 and k1 were calculated using Eq. 8, qe,2 and k2
were calculated using Eq. 10. The diffusion parameters kid and c were calculated by
fitting the data into the intraparticle diffusion model described in Eq. 11. The calculated
parameters for single and binary systems were summarised in Tables 2 and 3
respectively. Figure 5 shows the qt against time for the adsorption of nickel at 30 °C
 Single and Binary Adsorption of Heavy Metal Ions 85

20 30 °C
30 °C
15
40 °C
qt/ mg g-1 40 °C
10

5
0 50 100
t / min

Fig. 7. Kinetic study of binary zinc and nickel solution of initial concentration 50 mg L−1 at 30 °C
and 40 °C. The regression lines were drawn by fitting the data to Lagergren pseudo second order
model

Table 2. Kinetic parameters for the adsorption of single system nickel and single system zinc at
30 °C and 40 °C from a solution of initial concentration 50 mg L−1 onto 0.75 g of coir pith.
Kinetic model Single system
Parameter Ni Zn
30 °C 40 °C 30 °C 40 °C
Lagergren pseudo first order
qe,1/mg g−1 – – 13.19 5.16
k1/min−1 – – 0.096 0.105
R2 – – 0.91 0.57
Lagergren pseudo second order
qe,2/mg g−1 8.62 6.62 18.52 16.98
k2/g mg−1 min−1 −0.052 −0.036 0.026 0.082
R2 0.99 0.99 0.98 0.99
qe experimental 8.67 6.69 17.95 16.63
Intraparticle diffusion
kid/mg g−1 min−1/2 −0.09 −0.39 2.44 0.83
c/mg g−1 10.78 12.72 6.18 12.04
R2 0.06 0.29 0.86 0.29

and 40 °C. This shows that as temperature of the solution increases, the rate as well as
the amount of metal ion adsorbed increases initially. This is possibly due to the
expansion of the solid adsorbent and increase in kinetic energy as the temperature
increases.
86 E. N. Ranasinghe and B. M. W. P. K. Amarasinghe

Table 3. Kinetic parameters for the adsorption of binary system nickel and zinc at 30 °C and
40 °C from a solution of initial concentration 50 mg L−1 onto 0.75 g of coir pith.
Kinetic model Binary system
Parameter Ni Zn Total
30 °C 40 °C 30 °C 40 °C 30 °C 40 °C
Lagergren pseudo first order
qe,1/mg g−1 0.51 2.66 1.19 – 1.75 1.72
k1/min−1 −0.014 0.04 −0.0299 – −0.023 0.023
R2 0.01 0.69 0.54 – 0.34 0.21
Lagergren pseudo second order
qe,2/mg g−1 6.67 7.44 7.35 3.98 14.08 11.12
k2/gmg−1 min−1 0.234 0.059 0.04 −0.077 0.038 0.15
R2 0.99 0.99 0.99 0.98 0.98 0.96
qeexperimental 6.67 7.34 7.33 4.01 14 11.35
Intraparticle diffusion
kid/mg g−1 min−1/2 −0.001 0.274 0.103 −0.026 0.102 0.248
c/mg g−1 6.42 4.55 5.45 5.44 11.87 9.997
R2 0 0.68 0.15 0 0.06 0.16

However, as the system reaches equilibrium desorption occurs. Figure 6 which

shows qt against time for the adsorption of zinc shows that the rate increases with
temperature but the amount of zinc adsorbed decreases when the temperature increases
from 30 °C to 40 °C. Figure 7 shows the amount of metal ion adsorbed, qt, against
time for the binary system of nickel and zinc. The overall adsorption decreases as the
temperature increases from 30 °C to 40 °C. For the binary adsorption at 40 °C there is
initially more adsorption than at 30 °C but there is a significant amount of desorption
that occurs as it reaches equilibrium. As shown in Table 3, the equilibrium adsorption
capacity, qe, of nickel in the binary system increases from 6.67 mg g−1 to 7.34 mg
g−1while qe of zinc decreases from 7.33 mg g−1 to 4.01 mg g−1. The data of the single
system adsorption of nickel was fit into the Lagergren pseudo first order model. Due to
the equilibrium adsorption capacity, qe, being much smaller than preceding qt values it
was not possible to use this model. The adsorption data for zinc adsorption at 30 °C fit
well into the Lagergren pseudo first order model with a R2 of 0.91. However the
calculated qe,1 value 13.19 mg g−1 was much smaller than the experimentally obtained
value of qe 18.52 mg g−1. The adsorption capacity decreases as temperature increases
from 13.19 to 5.16 mg g−1, this suggests that due to the increased kinetic energy of the
system, it overcomes the attraction between zinc and the adsorbent leading to more
desorption. The same adsorption data was analysed using the Lagergren pseudo second
order model. The calculated qe,2 values for both nickel and zinc at 30 °C and 40 °C,
more closely matched the experimentally obtained values. The correlation coefficient,
R2, for all four systems were between 0.98 and 0.99. The values of qe,2 for nickel
shows the adsorption capacity decreased as temperature increased and reduced from
8.62 mg g−1 to 6.62 mg g−1. Similarly for zinc, the capacity decreases from
18.52 mg g−1 to 16.98 mg g−1. This is likely due to the increase in kinetic energy of
 Single and Binary Adsorption of Heavy Metal Ions 87

the system at higher temperatures promoting desorption. Overall, the adsorption of

nickel was about half that of zinc, this is likely because nickel ion is smaller and with a
higher charge density than the zinc ion, so it would prefer the more polar environment
of water whereas zinc would more easily bind to the polar coir pith functional groups.
Therefore, the amount of zinc adsorbed is greater than nickel under similar conditions.
The kinetic data for the single system adsorption was fit into the intraparticle diffusion
model as shown in Table 2. The R2 values were all very low except for the zinc
adsorption at 30 °C. This indicates that the intraparticle diffusion was not a rate
determining step for the nickel adsorption or zinc adsorption at 40 °C. As the isotherm
data also suggests the adsorption is primarily dependent on the solution concentration,
it is likely that the rate determining step in nickel adsorption is the chemisorptions to
the surface. The intraparticle diffusion in the zinc system at 30 °C was slow enough
that it affects the rate.

Table 4. Thermodynamic parameters for the single and binary systems of nickel and zinc at
30 °C and 40 °C
Parameters Single Binary
Ni Zn Ni Zn Total
−1
Kd/L g 30 °C 0.234 0.780 0.444 0.524 0.483
40 °C 0.167 0.665 0.524 0.211 0.343
DG/J mol−1 30 °C 3672 625 2043 1629 1835
40 °C 4674 1062 1683 4056 2782

Based on the FTIR spectra and kinetic data, it can be deduced that adsorption
involves a chemical reaction at the –OH groups of the functional groups on the coir pith
surface. As the adsorption in the binary system is competitive it can be presumed that
nickel and zinc bind to the surface at the same sites. The mechanism of adsorption is
likely a complexation of the metal ion and the surface functional groups as illustrated in
Fig. 8 where M can represent either nickel or zinc ions.

Fig. 8. Mechanism of adsorption of the metal ion onto the coir pith surface
88 E. N. Ranasinghe and B. M. W. P. K. Amarasinghe

The equilibrium constant Kd, was calculated using Eq. 12 and the Gibbs free
energy, DG, which was calculated using Eq. 13, was summarised in Table 4. DG for
nickel becomes more positive from 3672 J mol−1 K−1 to 4674 J mol−1 K−1 as the
temperature increases meaning the reaction becomes less spontaneous. DG increases
from 625 J mol−1 K−1 to 1061 J mol−1 K−1 for the zinc adsorption. When comparing
the DG of the adsorption of zinc and nickel, the zinc adsorption was shown to be more
spontaneous. The total adsorption in the binary system at 30 °C is 14 mg g−1 which is
larger than the average adsorption of the single systems at 30 °C. The percentage of
nickel adsorption is greater than in the single system while the zinc adsorption is less
than the adsorption in the single system. The total binary adsorption at 40 °C is
11.35 mg g−1, which is about the average of the adsorption in the nickel and zinc
single system under similar conditions. However the relative proportions of nickel
adsorbed was greater than zinc. This suggests that the adsorptions of nickel and zinc are
not independent of each other and the presence of competitive ions suppresses
adsorption of zinc but encourages adsorption of nickel. The binary adsorption data was
studied using the same single system models. When the binary adsorption data was fit
into the Lagergren pseudo first order model, the calculated qe values for the total binary
system at 30 °C and 40 °C was 1.75 mg g−1 and 1.72 mg g−1 respectfully varied
greatly from the experimentally obtained qe value, 14 mg g−1 and 11.12 mg g−1. The
correlation coefficients, R2, of the data, fit to the pseudo first order model for nickel and
zinc where 0.34 and 0.21 at 30 °C and 40 °C. This shows that Lagergren pseudo first
order model is not suitable to analyse the binary adsorption of nickel and zinc. The data
was also fit to Largergren pseudo second order model. The qe value calculated using
this model at 30 °C and 40 °C were 14.01 mg g−1 and 11.12 mg g−1 which were very
similar to the experimentally obtained values 14 mg g−1 and 11.35 mg g−1. The R2
values are 0.98 and 0.96 at 30 °C and 40 °C. This shows the adsorption occurring is
primarily chemisorptions and that this model can be used to analyse the binary
adsorption of nickel and zinc onto coir pith. Similar to the single system adsorption, the
adsorption of nickel was less than the adsorption of zinc at 30 °C. When the tem-
perature was increased to 40 °C the adsorption of nickel increases from 6.67 mg g−1 to
7.35 mg g−1and the adsorption of zinc decreased from 7.34 mg g−1 to 3.98 mg g−1.
Unlike in the single system, there is a greater decrease in the amount of zinc adsorbed
at 40 °C compared to zinc adsorbed at 30 °C. While there is more zinc adsorption than
nickel adsorption at 30 °C the difference between the two amounts is much less than in
the single system and at 40 °C there is also more nickel adsorption than zinc
adsorption. The binary adsorption data was fit to the intraparticle diffusion models.
However, the R2 values were very small, making it very unlikely that the model is
suitable to be used to analyse this data. This, along with the isotherm data, suggests that
there is little to no diffusion of the metal ions into the pores of the coir pith in the binary
system. The DG value of the total binary adsorption at 30 °C was 1835 J mol−1 K−1
and increases to 2782 J mol−1 K−1 as the temperature increases to 40 °C. Therefore the
adsorption gets less spontaneous as temperature increases.
 Single and Binary Adsorption of Heavy Metal Ions 89

4 Conclusion

The data from both the single systems were fit into the Langmuir, and Freundlich
isotherm models. Nickel and zinc single systems fit both models satisfactorily. How-
ever, they fit the Freundlich model the better, suggesting the adsorption surface is
heterogeneous. The isotherm data from the binary system was fit to the Langmuir,
Freundlich and extended Langmuir models. However they only fit only the Langmuir
model satisfactorily, suggesting a homogenous adsorption surface. The kinetic data for
single and binary system adsorption fit Lagergren pseudo second order model best,
suggesting the adsorption is mostly chemisorption. Single system zinc adsorption and
single system nickel adsorption decreased as the temperature did. In the binary system,
the adsorption capacity was less than the average of the single system under similar
conditions, where nickel percentage of nickel adsorption is more compared to the
single system, suggesting the competitive nature of the binary system promotes the
adsorption of nickel while it demotes zinc adsorption.

Acknowledgements. The authors wish to acknowledge the financial support from the
University of Moratuwa, Senate Research Committee for the grant SRC/ST/2017/31.

References
Adikaram AMNM, Pitawala HMTGA (2017) An overview of heavy metal contamination in
coastal sediments of Sri Lanka. JTFE 7(2):1–9
Amarasinghe BMWPK (2007) Comparison of adsorption characteristics of waste biomass
materials for the removal of Pb ions from industrial effluents. Eng: J Inst Eng Sri Lanka 40
(4):167–174
Amarasinghe BMWPK (2011) Lead and cadmium removal from aqueous medium using coir pith
as adsorbent: batch fixed bed column studies. JTFE 1(1):36–47
Cheung WH, Szeto YS, McKay G (2007) Intraparticle diffusion processes during acid dye
adsorption onto chitosan. BioresourTechnol 98:2897–2904
Freundlich H (1906a) Über die adsorption in lösungen (Adsorption in solution). J Chem Phys
57:384–470
Freundlich HMF (1906b) Über die adsorption in lösungen. Z Phys Chem 57:385–470
Halsey GD (1952) The role of surface heterogeneity. Adv Catal 4:259–269
Ho YS (1995) Adsorption of heavy metals from waste streams by peat. Ph.D. thesis, University
of Birmingham, Birmingham, UK
Ho YS, McKay G (1999) Competitive sorption of copper and nickel ions from aqueous solution
using peat. Adsorption 5:409–417
Kumar KV, Ramamurthi V, Sivanesan S (2005) Modeling the mechanism involved during the
sorption of methylene blue onto fly ash. J Colloid Interface Sci 284(1):14–21
Lagergren S (1898) About the theory of so-called adsorption of soluble substances. Kungliga
Svenska Vetenskapsakademiens. Handlingar 244:1–39
Langmuir (1916) The constitution and fundamental properties of solids andliquids. J Am Chem
Soc 38:2221–2295
Nasrullah A, Bhat AH, Isa MH (2016) Lignin: a sustainable biosorbent for heavy metal
adsorption from wastewater, a review. In: AIP conference proceedings
90 E. N. Ranasinghe and B. M. W. P. K. Amarasinghe

Sdiri AT, Higashi T, Jamoussi F (2014) Adsorption of copper and zinc onto natural clay in single
and binary systems. Int J Environ Sci Technol 11:1081–1092
Shaheen SM, Derbalah AS, Moghanm FS (2012) Removal of heavy metals from aqueous
solution by zeolite in competitive sorption system. IJESD 3(4):362–367
Sharifipour F, Hojati S, Landi A, Faz Cano A (2015) Kinetics and thermodynamics of lead
adsorption from aqueous solutions onto Iranian Sepiolite and Zeolite. Int J Environ Res 9
(3):1001–1010
Singhe N, Balomajumder C (2017) Equilibrium isotherm and kinetic studies for the
simulataneous removal of phenol and cyanide by use of S. Odorifera (MTCC 5700)
immobilized on coconut shell activated carbon. Appl Water Sci 7:3241–3255
Wijayawardhana D, Herath V, Weerasinghe A (2016) Heavy metal pollution in sri lanka with
special reference to agriculture: a review of current research evidence. Rajarata Univ J 4
(1):52–66
 Sustainable Autarky of Food-Energy-Water
(Safe-Water)

N. Khandan(&), D. Tchinda, S. M. Henkanatte-Gedera,

I. S. A. Abeysiriwardana-Arachchige, H. M. K. Delanka-Pedige,
S. P. Munasinghe-Arachchige, and Y. Zhang

New Mexico State University,

PO Box 30001 MSC 3CE, Las Cruces, NM 88003, USA
[email protected]

Abstract. Urban wastewaters (UWW) are now being recognized as a resource,

rich in nutrients and energy, rather than a waste stream that has to be treated and
disposed of at the input of considerable energy and associated environmental
emissions. This paper presents an algal-based approach for treating UWWs to
the mandated discharge levels in a single step and at the same time, recover net
energy, nutrients, and water from UWWs for use in the food-energy-water
sectors. The proposed approach entails cultivation of energy-rich algal biomass
in primary-settled UWW, followed by hydrothermal liquefaction of the resulting
algal biomass to recover its energy- and nutrient-contents. Results from a pilot
scale field demonstration project are presented to demonstrate the feasibility of
this approach. The pilot scale study conducted at a local wastewater treatment
plant confirmed that the discharge standards for biochemical oxygen demand
(BOD), nitrogen, and phosphorous could be met with a batch process time of 3
days. Recovery experiments showed that more than 20% of the ammoniacal
nitrogen (N-NH3) and more than 90% of the phosphorous (P) in the primary-
settled wastewater could be recovered as struvite by the proposed process.

Keywords: Algal wastewater treatment
Energy recovery
Nutrient recovery

Mixotrophic algal growth
Pilot scale demonstration

1 Introduction

With nearly 70% of the world population already concentrated in towns and cities, and
with continuing trends in population growth and sprawl of metropolitan areas, provi-
sion of the basic human needs of food, energy, and water to the inhabitants is seen as a
major challenge. Being net consumers of food, energy, and water, current urban
infrastructure that has continued to rely on obsolete technologies to provide these basic
needs is not sustainable. Development of a modern infrastructure that can recover and
recycle energy and materials within the food-energy-water (F-E-W) sectors at the
urban-agriculture interface could be an option to improve the sustainability of both
urban and agricultural systems.
In this paper, we report on algal-based, photosynthetically oxygenated waste-to-
energy recovery (POWER) system for recovering energy, water, and nutrients from

© Springer Nature Singapore Pte Ltd. 2020

R. Dissanayake and P. Mendis (Eds.): ICSBE 2018, LNCE 44, pp. 91–98, 2020.
https://doi.org/10.1007/978-981-13-9749-3_8
92 N. Khandan et al.

urban wastewaters (UWW) that can be utilized in the food production in the agriculture
sector. The POWER system affords waste treatment, net energy production, and
recovery of irrigation-quality water and high-purity crop fertilizers. The strain used in
the POWER system is an extremophile, Galdieria sulphuraria, capable of mixotrophic
growth. Optimal growth conditions for Galdieria sulphuraria are a temperature range
of 45–55 °C and a pH range of 1–4. Mixotrophic metabolism of Galdieria sulphuraria
enables single-step removal of carbon (BOD) and nutrients (nitrogen, N and phos-
phates, P) from UWWs. This mixotrophic algal system is capable of incorporating all
the C, N, and P in UWW into energy- and nutrient-rich biomass without any energy
input and without any loss of C as CO2 as in the current activated sludge systems. The
extreme culture conditions of above-ambient temperature and acidic pH levels in the
POWER system contribute to simultaneous deactivation of pathogens and viruses.
Additionally, solar energy captured via photosynthesis is incorporated into the biomass.
The energy- and nutrient-rich biomass generated in the POWER system is then
hydrothermally processed to yield biocrude, biochar, and an aqueous phase with sol-
ubilized N and P in concentrated form as by-products for recovery and use as fertil-
izers. This paper includes details of the POWER system and results from a pilot scale
study conducted under field conditions at an urban wastewater treatment plant.

2 POWER System

The first step in the POWER system entails cultivation of Galdieria sulphuraria,
5587.1 (GS) in primary-settled UWW. Due to its metabolic versatility, GS is capable of
removing dissolved organic carbon (measured as biochemical oxygen demand, BOD),
and nutrients (quantified as ammoniacal nitrogen and phosphates) in primary-settled
wastewater in a single reactor, without any external energy demand. The algal biomass
cultivated in this step incorporates most of the organic carbon, nitrogen, and phos-
phates contained in the wastewater. In addition, this biomass embodies most of the
energy content of the wastewater captured by the heterotrophic mode as well as solar
energy captured by photosynthesis in the autotrophic mode.
Subsequent steps in this system entail harvesting of the energy-rich biomass cul-
tivated in the first step followed by hydrothermal liquefaction (HTL) at moderate
temperatures (<350 ºC) and pressures (<20 MPa). In the HTL process, lipid-, protein-,
and carbohydrate-contents of the algal biomass are converted to four product streams:
an energy-rich oil phase; an N-rich aqueous phase; a P-rich solid residue; and waste off-
gases. These products of HTL afford efficient recovery of the energy- and nutrient-
contents of the algal biomass in the third step.
The oil-phase in the aqueous phase is separated by gravity settling and processed
further to yield biocrude. The P-rich solid phase, referred to as biochar, is separated
from the product mix by filtration. Particulate form of phosphorous in the biochar is
then extracted by leaching under basic conditions. The biochar leachate and the oil-
extracted aqueous phase are then mixed together and dosed with magnesium chloride
to precipitate struvite crystals for use as crop fertilizers.
 Sustainable Autarky of Food-Energy-Water (Safe-Water) 93

2.1 Pilot Scale System

A pilot scale version of the POWER system fed with primary-settled wastewater has
been deployed at a local wastewater treatment plant. The bioreactors of the POWER
system were seeded with G. sulphuraria CCMEE 5587.1 obtained from Culture
Collection of Microorganisms from Extreme Environments (University of Oregon).
Since the optimal pH for G. sulphuraria is 1 to 4, the initial pH of the culture volume in
the bioreactors was adjusted to 2.5 by adding 10N of H2SO4.
The bioreactors are of enclosed design, fabricated of clear polyethylene and fitted
with a motor-driven paddlewheel to enable circulation and mixing of its contents. The
active volume of the bioreactors was 700L and the culture depth was maintained at
20 cm. The headspace of the bioreactor was continuously fed with 2% carbon dioxide.
The bioreactors were fed with primary effluent and evaluated first in batch mode and
then in fed-batch mode. Samples were collected daily and analysed for ammoniacal
nitrogen (NH3-N) and phosphates (PO4); and every two days, for BOD5 (Fig. 1).

Fig. 1. Schematic of the pilot scale system and field installation at a wastewater treatment plant.

3 Results

3.1 Removal of BOD, N, and P

Results from tests on two identical bioreactors of the POWER system (R1 and R2)
conducted over a period of 120 days in 25 fed-batch cycles are summarized in Fig. 2.
As can be noted from Fig. 2, the bioreactors were able to reduce BOD5, NH3-N, and
PO4 to near-discharge standards in 2 days, irrespective of the variations in the initial
concentrations and ambient conditions. The high degree of reproducibility of the
system is also illustrated in Fig. 2.
94 N. Khandan et al.

200
100

Concentration [mg/L]

10

R1 R2 R1 R2 R1 R2 R1 R2 R1 R2 R1 R2
0.2
Day 0 Day 2 Day 0 Day 2 Day 0 Day 2
BOD5 NH3-N PO4

Fig. 2. Reduction of BOD, NH3-N, and PO4 in 2 days in 25 fed-batch cycles run over 120 days.
Dashed line shows the respective discharge standards.

3.2 Removal of Pathogens

Fecal coliform count in the primary effluent was reduced instantly in the bioreactors to
non-detectable level (<1 CFU/mL). Total coliform count was reduced to non-
detectable levels (<1 CFU/mL) within one day of treatment. Total bacteria popula-
tion was reduced by 1.68 logs in the algal system at the end of treatment. Enterococcus
faecalis and Escherichia coli copy numbers were reduced to non-detectable levels
(<2 copies/mL) resulting in 3.85 and 5.44 log reductions, respectively (Delanka-Pedige
2018). Of special note, Pseudomonas aeruginosa, Salmonella enterica and Staphylo-
coccus aureus were not detected in the primary effluent.
Based on a mechanistic study undertaken to discern the key factors responsible for the
high bacterial inactivation noted in this algal system, it was concluded that the inacti-
vation could be attributed primarily to the low pH conditions maintained in the reactor.
Additionally, the synergistic effect of high dissolved oxygen levels (5.1 to 7.8 mg/L;
average = 6.2 ± 0.53 mg/L) during sunlight hours with the potential to produce reactive
oxygen species could also have contributed to bacterial inactivation. Even though the
enclosed reactor design enabled above-ambient culture temperatures during the test
period (27 to 46 ˚C; average = 34.9 ± 4.3 ˚C), temperatures below 45 °C were found to
be insignificant in inactivating bacteria (Munasinghe-Arachchige 2018).
 Sustainable Autarky of Food-Energy-Water (Safe-Water) 95

3.3 Recovery of Energy

A core technology embedded in the POWER system is hydrothermal liquefaction of
the algal biomass that enables recovery of energy and nutrients for reuse. The yields of
the four products of HTL (biocrude, biochar, aqueous phase, and off-gases) are func-
tions of the HTL process temperature. Given the energy content of the wastewater (6.3–
7.6 kJ/L) the net energy recovery from the integrated process is dependent primarily on
the energy input to the HTL process to maintain the HTL process temperature on one
hand, and the resulting yield of biocrude, on the other.
Simulation models and experimental results were combined to estimate the energy
input for heating the feed to the HTL process temperature, energy that could be
recovered from the biocrude and biochar, and the net energy yield at various HTL
process temperatures. Results of this analysis are summarized in Fig. 3 that show that
the optimal HTL process temperature is 180 ºC, if energy could be extracted from
biochar as well. However, technologies for efficient recovery of energy from biochar
are not common. With biocrude as the sole energy product, 300 ºC is seen as the
optimal HTL process temperature. Considering the energy content of UWW (6.3 to
7.6 kJ/L), the net energy output estimated here including biochar is 4.5 kJ/L, which
translates to energy recovery efficiency of 60%.

Net Energy output

Thermal energy input
Mechanical energy input
Energy output from biocrude

Energy output from biochar

330

300
HTL temperature [ºC]

275

250

225

200

180

-15 -10 -5 0 5 10 15 20 25 30
Energy [MW]

Fig. 3. Energy input to, output from, and net energy from hydrothermal liquefaction process as
a function of process temperatures.
96 N. Khandan et al.

3.4 Recovery of Nutrients

The efficiency of recovery of nutrients from the algal biomass is a function of the HTL
process temperature. Concentrations of total N and total P in the aqueous resulting from
hydrothermal liquefaction of the warm, cold, and mixed cultures of GS cultivated in
primary settled wastewater were measured These results indicated that typical total N
averaged 1,450 mg/L and typical total P averaged 70 mg/L. These high concentrations
enable efficient recovery of N and P for beneficial use. Available P in the biochar
samples was typically less than 50 mg P/kg. However, total phosphorous in the biochar
exceeded 50,000 mg/kg (or more than 5% by weight).
It was hypothesized that the non-available P could be extracted from the biochar by
leaching under acidic or basic conditions. Experiments were conducted to assess the
leaching efficiency at pH = 7, pH = 1.5, and pH = 12 to establish the optimum con-
ditions to extract P into the available form. Based on the results of the leaching
experiments it was concluded that pH > 12 would be optimal for complete extraction
of P from biochar.
Magnesium in the form of magnesium chloride was selected to recover N and P as
struvite. Struvite was rapidly precipitated upon addition of magnesium chloride to the
mixture of the biocrude-extracted aqueous phase and the biochar-leachate. Chemical
analysis of the algal struvite precipitated in this manner indicated that its composition
was close to that of pure struvite.
A mass balance was constructed across the entire system to assess the efficiency of
recovery of nutrients by the POWER system. From the results of a typical overall mass
balance summary shown in Fig. 4, N-recovery as struvite can be seen as 16% and P-
recovery as 89%. This analysis showed that the supernatant from the struvite precip-
itation was rich in N; about 71% of the N in the primary-settled waste remained
unrecovered. One option for utilizing this resource is to recycle it to the algal culti-
vation step to boost biomass production. Future research needs to be directed towards
developing technologies for recovering N from this by-product stream.

3.5 POWER System in F-E-W Sectors

Results of long-term evaluation of the POWER system suggests that the system has the
potential to render wastewater treatment a sustainable operation. It can conserve the
energy consumed in the traditional aerobic wastewater treatment practice and the
energy consumed in the manufacture of fertilizers. It can minimize environmental
emissions associated with the current wastewater treatment practice and recover
valuable resources for reuse and thus conserve the limited natural resources. Figure 5
shows a conceptual system for applying the POWER system at the urban-agriculture
interface. By recovering irrigation-quality water, net energy, and fertilizers from the
wastewaters from urban and agricultural sources for reuse overall sustainability could
be improved.
 Sustainable Autarky of Food-Energy-Water (Safe-Water) 97

Fig. 4. Overall mass balance on ammoniacal nitrogen (N, kg/d) and phosphates (P, kg/day) for a
flow rate of 1 MGD.

Food Water
C, N & P

N&P

Agricultural Energy Urban

systems systems
C, N & P
Wastes Wastes
[C, N & P] [C, N & P]

POWER
CO2 Reclaimed water system
Recovered
N&P
Recovered energy

Fig. 5. Application of POWER system at the urban-agriculture interface to recover irrigation-

quality water, net energy, and fertilizers.
98 N. Khandan et al.

4 Conclusions

Feasibility of the POWER system in treating primary-settled urban wastewater and

recovering net energy and nutrients from it was demonstrated at pilot scale. Long-term,
stable operation of the system in a reproducible manner has been demonstrated. The
algal cultivation-hydrothermal liquefaction-struvite precipitation pathway embedded in
the POWER system has the potential to afford sustainable wastewater treatment.
A conceptual scheme to apply the POWER system at the urban-agriculture interface is
proposed to recover and recycle resources within the food-water-energy sectors.

Acknowledgements. This study was supported in part by the National Science Foundation
Engineering Research Center for Reinventing the Nation’s Urban Water Infrastructure (ReNU-
WIt), award EEC 1028968; the National Science Foundation Energize New Mexico EPSCoR
award IIA-1301346; College of Engineering at New Mexico State University, and the Ed &
Harold Foreman Endowed Chair.

References
Henkanatte-Gedera SM et al (2017) Removal of dissolved organic carbon and nutrients from
urban wastewaters by Galdieria sulphuraria: laboratory to field scale demonstration. Algal Res
24:450–456
Selvaratnam T et al (2014) Evaluation of a thermo-tolerant acidophilic alga, Galdieria
sulphuraria, for nutrient removal from urban wastewaters. Bioresour Technol 156:395–399
Delanka-Pedige HMK (2018) Algal-based wastewater treatment: assessment of pathogen
removal and bacterial community profile of effluent. MS thesis, New Mexico State University,
Las Cruces
Munasinghe-Arachchige SP (2018) Algal-based wastewater treatment: factors influencing
bacterial inactivation and potential for coliform regrowth, MS thesis, New Mexico State
University, Las Cruces
Selvaratnam T et al (2015) Feasibility of algal systems for sustainable wastewater treatment.
Renew Energy 82:71–76
Henkanatte-Gedera SM et al (2015) Algal-based, single-step treatment of urban wastewaters.
Biores Technol 189:273–278
 Study of Evidences on Chronic Kidney Disease
Due to Unknown Etiology Based
on Environmental, Social, Economic
and Health Patterns of Selected Population
A Case Study Done at Wilgamuwa Area

S. M. Rajapakshe1(&), S. K. Weragoda1, T. Kawakami2,

and W. B. M. L. I. Weerasekara1
1
Water Safety Plan Advisory Unit, National Water Supply and Drainage Board,
Katugastota, Sri Lanka
[email protected]
2
Toyama Prefectural University, Imizu, Japan

Abstract. Chronic Kidney Disease (CKD) is becoming a major rising public

health problem in Sri Lanka. In addition, CKD has a very complicated inter-
relationship with other diseases. Accordingly, a community-based study was
conducted to identify the evidences of the prevalence of CKD and correlate the
relationship between drinking water and causes for CKD. Further, a study was
conducted to identify the impact on lifestyle due to CKD issue. The study
particularly focused on the Wilgamuwa Divisional Secretariat area where the
burden of CKD is pronounced and the underlying cause is still unknown. The
sample of the study was selected representing high and low prevalence of CKD
which included both patients and non-patients. A questionnaire survey was
conducted to gather data, mainly focused on areas like drinking water sources,
social, environmental, health, economic, genetic and cultivation practices.
Drinking water sources from the dug wells and deep tube wells used by both
patients and non-patients was analysed for physical and chemical parameters
including heavy metals. Water quality parameters like pH, conductivity, tur-
bidity, oxidation- reduction potential and dissolved Oxygen were analysed
onsite. Heavy metals were analysed from Inductively Coupled Plasma Mass
Spectroscopy (ICP-MS) and anions was analysed from Ion Chromatography
(IC). Locations of water sources and patients and non-patients were traced by
GPS and a groundwater quality and patient’s distribution map was prepared
using GIS and a study of the patient distribution and causative factors interre-
lating the water quality was conducted. Based on all the water quality param-
eters and questionnaire survey results in the study, some of the dug wells in the
area were identified as doubtful whereas some were suitable for drinking water.
It can be concluded that there is no relationship with Hardiness and Fluoride
with CKDU and most of the water quality parameters did not exceed the SLS
drinking water standards SLS 614:2013.

© Springer Nature Singapore Pte Ltd. 2020

R. Dissanayake and P. Mendis (Eds.): ICSBE 2018, LNCE 44, pp. 99–106, 2020.
https://doi.org/10.1007/978-981-13-9749-3_9
100 S. M. Rajapakshe et al.

1 Introduction

Chronic Kidney Disease (CKD) has become a serious health issue and is increasing at
an alarming rate today in Sri Lanka. It is a major health care problem, especially in the
dry zone. It is observed that CKD is caused by multiple factors, linked in combination.
Over one thousand people were reported as dead due to CKD and more than 60,000
patients have registered at renal clinics of several government hospitals in the dry zone
of the island (Weragoda et al. 2013).
CKD is most prevalent among male farmers in the 40 to 60 age group engaged in
rice cultivation, although lately, it has also been detected among women, children, and
even cattle (Centre for Science and Environment 2012). There is scientific consensus
that this fatal disease in Sri Lanka is not related to known causes commonly identified
with kidney disease, such as, diabetes, hypertension, obesity, or other factors such as
snake bites (Ministry of Healthcare and Nutrition, Epidemiology Unit, Sri Lanka and
World Health Organization 2009). After year 2000, there was a rapid increment of the
CKD patients in Sri Lanka and from those, majority were in the dry zone, especially
Anuradhapura, Polonnaruwa, Kurunegala and Matale districts. There was a big impact
of the disease on social, economic, political, and environmental backgrounds of the
country. Identifying causal factors is very essential to answering this problem. So,
number of researches are conducted based on different disciplines. Groundwater is the
main drinking water resource and more than 90% of the drinking water requirements
for the rural communities are obtained by shallow and deep wells.
The selected study area of Wilgamuwa is situated in the Matale District in the
Central Province of Sri Lanka. Topographically, the western part of the study area is
hilly while the eastern part is flat and undulating. The Mahaweli Ganga is the main
surface water body of the region that flows along the direction from south to north. In
addition, irrigation tanks, channels, and minor streams are located in the study area
(Fig. 2).
Most of the irrigation network of the area is based on water diverted from the
Mahaweli River. Geologically, Migmatitic gneiss is the main rock types in this area and
the overburden is mainly composed of a weathered profile. Annual precipitation of the
area is around 1750 mm and the study area belongs to the intermediate climatic zone.
The main recharging source of both shallow and deep groundwater of the area is
from local rainfall. However, in some areas groundwater sources recharge from local
rainfall as well as irrigated water. So, contaminated groundwater in the shallow dug
wells in the areas may be used by the people of the villages (Figs. 1 and 3).
Study of Evidences on Chronic Kidney Disease Due to Unknown Etiology 101

Fig. 1. CKDu patient distribution map Wilgamuwa GS division

102 S. M. Rajapakshe et al.

Fig. 2. Location map of the study area

Fig. 3. Elevation map of the study area

2 Materials and Methods

This community-based study was conducted to identify the causes for the prevalence of
CKDu and to evaluate the relationship between drinking water quality and causes for
CKD. The stratified sampling collection method was applied and samples collected
 Study of Evidences on Chronic Kidney Disease Due to Unknown Etiology 103

from the selected 11 Grama Niladhari Divisions (GND) representing high and low
prevalence of CKDu which included both patients and non-patients.
A questionnaire survey was conducted to gather data, mainly focused on areas like
drinking water sources, social, environmental, health, clinical, economic, genetic and
occupational practices. Chemical (Cd, As, Pb, F, Mg, Mn, Ca, Hardness) and physical
parameters (onsite analysis was conducted for pH, conductivity, turbidity, oxidation-
reduction potential (ORP) and dissolved oxygen) were analysed for drinking water
sources from the dug wells and deep tube wells used by both patients and non-patients.

Fig. 4. (a) pH variation of patients (b) pH variation of non-patients (c) Fluoride variation of
patients (d) Fluoride variation of non-patients (e) EC variation of patients (f) EC variation of non-
patients (g) Calcium variation of patients (h) Calcium variation of non-patients (i) Arsenic
variation of patients (j) Arsenic variation of patients (k) Maganese variation of patients
(l) Maganese variation of non-patients
104 S. M. Rajapakshe et al.

Fig. 4. (continued)
 Study of Evidences on Chronic Kidney Disease Due to Unknown Etiology 105

Fig. 4. (continued)

Heavy metals were analysed from Inductively Coupled Plasma Mass Spectroscopy
(ICP-MS).
Locations of water sources, patient and non-patient, were traced by GPS and
groundwater quality and patient distribution map using GIS was prepared and the
patient distribution and causative factors interrelating the water quality was studied.
Descriptive statistical results are shown in graphical form comparing the chemical
parameters with recommended levels stipulated by Sri Lankan standards (Fig. 4).

3 Results

Based on all the water quality parameters and the questionnaire survey results, some of
the dug wells were identified as doubtful whereas some were suitable for drinking
water and also, significant difference was not observed in CKD prevalent and non-
prevalent areas.
Many of those affected by CKDu are young to middle aged male farmers who work
on rice farms (paddy fields) and other agricultural commodity cultivations, where they
are presumably exposed to unusually high levels of environmental hazards compared to
their non-farming peers.
In both patient and non-patients, high Fluoride, Manganese and Magnesium con-
centration were observed while Calcium concentration is within the limits.
Male farmers 25–65 years of age are the most affected and die of this disease in
greatest numbers; they also frequently experience unprotected exposure to agro-
chemicals and other toxins in their harsh work environment, and also engage in high-
risk behaviour in comparison to women.
On top of regularly getting dehydrated while at work and drinking less water, one
of the plausible reasons male farmers are more exposed to potential toxins is that many
have disregarded the age-old habit of taking clean water from home to the farmlands.
They have adapted an easier and a dangerous habit of drinking polluted water directly
from nearby streams and from the paddy fields while, women and children continue to
106 S. M. Rajapakshe et al.

drink water from relatively protected shallow wells and deep tube wells near their
homes. Those affected most are male farmers who work in harsh climatic conditions, in
flat lands with poor drainage; they do not use protective gear when handling and
spraying toxic agrochemicals, nor while working in farming land.
These farmers frequently get dehydrated during work in hot, and harsh climatic
conditions, and drink too little water (and some drink contaminated water directly from
paddy fields). In addition, the affected areas experience unpredictable weather patterns
with prolonged dry spells. All CKDu-affected areas in the world are predominantly
agricultural regions with agro-based economies; do not have a centrally purified pipe
borne water supply, have less than optimal access to medical care, and are economi-
cally relatively poor. These chemicals included Cd, As, Pb, F, and Ca which previous
researches have been suggested as causative factors for CKD.
It was seen that in the GND divisions where there is high prevalence of CKD, most
of the agricultural practices and drinking water sources were unsatisfactory. Some of
the dug wells were contaminated with heavy metals. But still, the water quality
parameters of some patients’ locations were satisfactory as a drinking water source.
Therefore, a further detailed study on all other factors is needed.

4 Discussion

It was seen that in the GND divisions where there is high prevalence of CKD, most of
the agricultural practices and drinking water sources were unsatisfactory. Some of the
dug wells were contaminated with heavy metals. But still, the water quality parameters
of some patients’ locations were satisfactory as a drinking water source. Therefore,
further detailed study on all other factors is needed.

5 Conclusion

According to the samples analysed, it can be concluded that most of the water quality
parameters do not exceed the SLS drinking water standards. But considerably, Man-
ganese, Fluoride, Magnesium exceeds the SLS drinking water standards in both patient
and non-patients.
The aetiology of CKDu in of Sri Lanka is probably multifactorial, involving one or
more environmental factors and a possible genetic predisposition in vulnerable
populations.

References
Centre for Science and Environment (2012) Environmental contamination and its association
with Chronic Kidney Disease of Unknown Aetiology in North Central Region of Sri Lanka
Ministry of Healthcare and Nutrition, Epidemiology Unit, Sri Lanka and World Health
Organization (2009) Chronic kidney disease of unknown aetiology (CKDu). National Research
Programme for the Chronic Kidney Disease of Unknown Aetiology (CKDu) in Sri Lanka
Weragoda S, Kawakami T, Motoyama A, Kodithukakku S (2013) Is groundwater in Dry Zone,
Sri Lanka safe to drink? Sri Lanka Collaborative Research
 Algal Wastewater Treatment:
Study of Polyculture in Fed-Batch
Mode in a Winter Climate

I. S. A. Abeysiriwardana-Arachchige, W. Chapman,
and N. Nirmalakhandan(&)

Civil Engineering Department, New Mexico State University,

Las Cruces, NM 88003, USA
[email protected]

Abstract. Since the 1950s, reactors using algal-bacterial polycultures have

been engineered by researchers to treat wastewater. The symbiotic relationship
between algae and bacteria accommodates energy-efficient wastewater treatment
by eliminating artificial aeration. Here, we investigated the applicability of a
novel algal-based system using two polycultures to treat primary-settled
municipal wastewater under field conditions in a winter climate (average daily
temperature: 4–16.5 °C). The two polycultures were cultured in two closed
700L bioreactors deployed at a local wastewater treatment plant in Las Cruces,
NM, United States. The Ammoniacal nitrogen and phosphate concentrations and
OD750 of the two reactors were analyzed daily. The BOD5 in the reactors were
analyzed once in two days. The samples from the two reactors were analyzed
over 44 days from January to March in 2018. The culture in reactor 1 attained
average ammoniacal nitrogen, phosphate, and BOD5 removals of 99.3%, 70.1%,
and 70.0% respectively in a single step within five days. The corresponding
efficiencies for reactor 2 were 44.3%, 47.5%, and 50.7%. The average batch
processing times for reactor 1 and reactor 2 were 1.02 and 4.02 days respec-
tively. This paper confirms the ability of both the polycultures to treat primary-
settled wastewater in a single step.

Keywords: Algal wastewater treatment
Phosphorus removal

Nitrogen removal
Polyculture

1 Introduction

Wastewater treatment utilities have continued to rely on energy-intensive and unsus-

tainable technologies for removing nutrients (nitrogen and phosphorus) from
wastewaters to the required discharge standards. Researchers have been looking for
alternative wastewater treatment technologies to conserve energy and recover nutrients
from wastewaters to render the process sustainable and cost-effective [1]. On realizing
the capability of algae to uptake nutrients while growing photosynthetically, several
studies have evaluated the feasibility of using algae in wastewater treatment systems.
Since the premise to use an algae-bacteria mixed polyculture for wastewater treatment
was tested in 1950s [2], many studies have been conducted to develop engineered

© Springer Nature Singapore Pte Ltd. 2020

R. Dissanayake and P. Mendis (Eds.): ICSBE 2018, LNCE 44, pp. 107–115, 2020.
https://doi.org/10.1007/978-981-13-9749-3_10
108 I. S. A. Abeysiriwardana-Arachchige et al.

systems with algae-bacteria cultures to treat industrial and domestic wastewaters [3–6].
The symbiotic relationship between algae and bacteria has made energy efficient
wastewater treatment possible in solar-powered photobioreactors [1].
Microalgae and bacteria have been shown to coexist in culture while mutually
benefiting each other. With the presence of sunlight, microalgae photosynthetically
produce oxygen which is used by aerobic bacteria to biodegrade organic matter. The
algae then consume the carbon dioxide released by bacterial respiration whereby
eliminating artificial aeration [1, 6]. More recently, the ability of algal species such as
Galdieria sulphuraria [7, 8] to mixotrophically metabolize organic carbon and nutrients
in a single step to meet the respective discharge standards has been recognized. Since
most of the ammoniacal nitrogen and phosphorus in wastewater accumulate in algal
cells [2, 6], such a mixotrophic system can achieve efficient and economical wastewater
treatment, while generating biomass that can be processed downstream to recover the
nutrients as crop fertilizers [6] or recover energy in the form of biocrude [9, 10].
An algae-based, Photosynthetically Oxygenated Waste to Energy Recovery
(POWER) system has been developed at New Mexico State University (NMSU) for
energy-efficient wastewater treatment and resource recovery. An axenic culture capable
of mixotrophic metabolism, G. sulphuraria, used in this system has demonstrated
superior carbon and nutrient removal where the culture temperatures are near its
optimal range of 27–46 °C [7, 8]. Since the temperature drops in winter, G. sulphuraria
loses its treatment efficiency. Hence, finding an adequate culture replacement for winter
had been a long-standing quest. The studies that have been done so far on urban
wastewater (UWW) treatment with microalgae in winter climate are very limited.
Therefore, a polyculture that could sustain the cold weather was selected assuming it
could treat urban wastewater as efficient as by G. sulphuraria during summer.
This study evaluated the feasibility of two polycultures to accommodate winter
treatment in the POWER system. The two polycultures were assumed to consist of two
different algal-bacterial compositions. The specific objective of the study was to
investigate the removal efficiency of dissolved organic carbon and nutrients by the two
polycultures in a winter climate.

2 Materials and Methods

2.1 Microalgae Strain and Cultivation Methods

Polycultures obtained from the Fabian Garcia Research Facility (New Mexico State
University) were grown in two pilot scale photobioreactors deployed at a local
wastewater treatment facility, Las Cruces, NM, USA. The bioreactor was fed with
primary-settled wastewater. Unlike previous studies on the cultivation of G. sulphu-
raria, the addition of growth mediums of any kind or enrichment of headspace with
carbon dioxide was not done during this study. The initial pH in the reactors was not
chemically adjusted.
 Algal Wastewater Treatment 109

2.2 Equipment Used and Fed-Batch Cultivation

The experiments were conducted in two identical enclosed photobioreactors (R1 and
R2) fabricated from a thin clear polymer allowing for light penetration and distribution.
The enclosed design enabled the prevention of evaporation and contamination, and
conserved heat inside the reactor. The culture depth was maintained at 20 cm and
circulation and mixing in the reactor were controlled by a motor-driven paddle wheel.
The reactor was operated in a fed-batch mode, with a working volume of 700L,
which comprised of 400L of primary effluent and 300L of culture. Each fed-batch cycle
was terminated when the reactor reached the stipulated discharge standards for all the
three primary pollutants: ammoniacal nitrogen - 10 mg/L; phosphate - 1 mg/L; and
BOD5 - 30 mg/L. The paddle wheel was then switched off for 24 h to settle the
biomass, and 400L of the resulting supernatant was discharged and replenished with a
fresh batch of 400L of primary-settled wastewater. These fed-batch experiments were
conducted continuously for 44 days from January through March in 2018.
Instead of direct measurement of the dry density, the biomass concentration in the
reactor was analyzed daily in terms of optical density at 750 nm wavelength (OD 750)
with a HACH DR 6000 spectrophotometer. Ammoniacal nitrogen and phosphate
concentrations were also analyzed daily using the same equipment following the sal-
icylate method 8155 and the phosver3 method 8048 respectively. Dissolved BOD5 was
measured every two days with the standard method 5210 B. Furthermore, the tem-
peratures in each reactor and in the field were recorded at an interval of 15 min using
Onset HOBO data loggers (UA-002-64).

2.3 Summary of Operation Conditions

The temporal variation of daily average temperatures is illustrated in Fig. 1. The
average daily temperature in the reactor and in the atmosphere ranged between 8–21 °C
and 4–16.5 °C respectively. The concentrations of ammoniacal nitrogen, phosphate,
and BOD5 in the primary effluent fed to the bioreactors ranged between 19.8–
40.3 mg/L, 0.6–2.1 mg/L, and 39.5–70 mg/L respectively.

Fig. 1. Temporal variation of average daily temperature

110 I. S. A. Abeysiriwardana-Arachchige et al.

3 Results and Discussion

3.1 Growth, PH, and Reactor Stability
The temporal variation of biomass growth is illustrated in Fig. 2. The two reactors R1
and R2 maintained growth rates of 0.016 ± 0.001 and 0.014 ± 0.001 OD 750 nm
day−1 respectively. The hypothesis test (H0: ß1 = ß2 vs. Ha: ß1 6¼ ß2) to check equality
of the two slopes concluded that there is no statistically significant difference between
growth rates at a significance level of 0.05 (P ’ 0.88). However, both the reactors in
the current study seemed to experience a lag phase compared to the previous results on
the POWER system [11] with G. sulphuraria during summer time.

Fig. 2. Variation of optical density at 750 nm of the reactors with time. Vertical lines indicate
cycle separation.

As the cultures used in this experiment endured the cold climate, stable operation of
both reactors was possible. Unlike in the case of G. sulphuraria, the photobioreactors
were neither supplemented with CO2 nor acidified. The pH variation in the two setups
during the study period is shown in Fig. 3. As can be seen, the pH of reactor 2 was
slightly higher than that of reactor 1; however, the overall pH increased in both reactors
as the treatment cycle progressed. As explained by the following equation, the increase
of pH can be ascribed to the uptake of photosynthetic carbon by microalgae from the
surrounding which promotes the release of hydroxyl ions to the medium [12, 13].

HCO
3 $ CO2 þ OH

ð1Þ
 Algal Wastewater Treatment 111

Fig. 3. Variation of pH in the reactors with time. Vertical lines indicate cycle separation.

3.2 Nutrient Removal

The temporal variation of the ammoniacal nitrogen profile during the study period is
illustrated in Fig. 4. As seen in the plot, both reactors achieved the discharge standard
of 10 mg/L at the end of the operation cycle. For the 8 cycles, the average volumetric
nitrogen removal in reactor 1 was 5.34 ± 2.97 mg L−1 day−1 with 99.3% removal. The
corresponding parameters for reactor 2 were 1.78 ± 1.03 mg L−1 day−1 and 44.3%
respectively.
The temporal profile of phosphate concentration in the two reactors along with the
allowable discharge limit (1 mg/L) is shown in Fig. 4. The average volumetric phos-
phate removal considering 8 operation cycles for reactor 1 was 0.11 ± 0.09 mg L−1
day−1 with 70.1% removal. The corresponding values for reactor 2 were
0.07 ± 0.06 mg L−1 day−1 and 47.5% respectively. The lower removal rates could be
due to the lower average phosphate concentration observed during winter (1.26 mg/L)
compared to that in summer (3.74 mg/L). Furthermore, the instantaneous phosphate
removal rates on the initial day immediately after feeding primary effluent to reactor 1
and reactor 2 were as high as 0.92 ± 0.54 mg L−1 day−1, 0.80 ± 0.66 mg L−1 day−1
respectively. The phosphate concentration of primary wastewater dropped instanta-
neously by 67.8% and 52.6% on average after been mixed with the preadapted cultures
in R1 and R2. This drop was calculated considering the difference of phosphate con-
centrations in the primary-settled wastewater and in the reactor on day 0. Apart from
the dilution effect, this pronounced drop of phosphorus concentration can also be
attributed to adsorption of phosphorus to the surface of cells and the surface of
bioreactors [12]. Furthermore, the subsequent rise of phosphorus concentration in the
reactors could partially be due to the release of phosphorus by ruptured cells as
described by [12]. This phenomenon is common with old cultures, therefore, the
subsequent rise of phosphorus concentrations in reactor one during the last two cycles
can be explained [14] (Fig. 5).
112 I. S. A. Abeysiriwardana-Arachchige et al.

Fig. 4. Variation of NH4-N concentration with time. Vertical lines indicate cycle separation.

Fig. 5. Variation of phosphate concentration with time. Vertical lines indicate cycle separation

3.3 Carbon Removal

The temporal profile of organic carbon during the study period is illustrated in Fig. 6.
As can be noted, both reactors achieved the permissible discharge level by the end of
the cycle time. Considering the 8 operation cycles, the average volumetric BOD
removal for reactor 1 was 12.93 ± 6.53 mg L−1 day−1 with 70.0% removal; the
corresponding values for reactor 2 were 6.77 ± 3.51 mg L−1 day−1 and 50.7%.
The required minimum cycle time to achieve all three discharge standards for each
cycle is demonstrated in Fig. 7. The average cycle time in reactor 2 (4.02 days)
appeared nearly 4-fold higher compared with that in reactor 1 (1.02 days)
 Algal Wastewater Treatment 113

Fig. 6. Variation of BOD concentration with time. Vertical lines indicate cycle separation.

Reactor 1 Reactor 2

7
6
5
Time [Days]

4
3
2
1
0
1 2 3 4 5 6 7 8
Cycle

Fig. 7. Required minimum cycle time considering all three discharge standards

4 Comparison of Previous Studies

The availability of literature for studies conducted with algal-bacterial cultures during
cold climates (5–15 °C) is limited due to the widely experienced performance limita-
tions of such systems in extreme environments. A study conducted by [15] using
cyanobacterial strains isolated from the Arctic and Antarctic environments reported
maximum nitrogen and phosphorus removal efficiencies of 2.3 mg NL−1d−1 and 0.07–
0.15 mg PL−1d−1 at 15 °C under laboratory conditions. The maximum nitrogen
removal rate reported in the above study is higher compared to the average nitrogen
removal rate in reactor 2 (1.78 mg L−1 day−1). However, reactor 1 showed a higher
average removal rate (5.43 mg L−1 day−1) under field conditions (average culture
temperature: 8.45–21.25 °C). The average phosphate removal rates were however
similar to the highest phosphorus removal rates reported by [15].
[16] reported the average batch processing times for experiments conducted in
warmer temperatures (culture temperature: 24–36 °C) with G. sulphuraria Soos strain,
G. sulphuraria 5587.1 and, mixed (a mixture of 5587.1 & Soos) as 4.62 days, 3.87
114 I. S. A. Abeysiriwardana-Arachchige et al.

days and 4.16 days respectively. The batch processing time reported in reactor 2 was
similar to those reported in the above study. However, the culture in reactor 1 managed
to reach the discharge standards in lesser time (1.02 days) despite the cold climate.

5 Conclusion

The study evaluated the suitability of two polycultures to serve in the POWER system
during cold temperatures. A summary of the research findings from January to March is
documented below:
• Average removal efficiencies of BOD5, ammoniacal nitrogen, and phosphate
respectively by 70.0%, 99.3% and 70.1% by reactor 1 and 50.7%, 44.3% and 47.5%
by reactor 2.
• Both reactors remained stable despite the cold weather. reactor 1 and 2 reached the
mandated discharge standards respectively in 1.02 days and 4.02 days.
• The initial phosphate concentration in the POWER system drops rapidly. This could
both be due to the dilution effect and surface adsorption.
• Carbon dioxide and nutrients were not supplied to support the growth. Therefore,
winter operation of the POWER system with the polycultures can be more eco-
nomical even compared to the summer operation with G. sulphuraria.
Based on the results, a polyculture with a similar composition as in reactor 1 is
recommended for winter operations of the POWER system.

References
1. de Godos I, Blanco S, García-Encina PA, Becares E, Muñoz R (2009) Long-term operation
of high rate algal ponds for the bioremediation of piggery wastewaters at high loading rates.
Bioresour Technol 100:4332–4339. https://doi.org/10.1016/J.BIORTECH.2009.04.016
2. Oswald WJ, Gotaas HB, Golueke CG, Kellen WR, Gloyna EF, Hermann ER (1957) Algae
in waste treatment [with discussion]. Sewage Ind Wastes 29:437–457. http://www.jstor.org/
stable/25033322
3. Muñoz R, Guieysse B (2006) Algal–bacterial processes for the treatment of hazardous
contaminants: a review. Water Res 40:2799–2815. https://doi.org/10.1016/J.WATRES.
2006.06.011
4. García J, Mujeriego R, Hernández-Mariné M (2000) High rate algal pond operating
strategies for urban wastewater nitrogen removal. J Appl Phycol 12:331–339. https://doi.org/
10.1023/A:1008146421368
5. Bordel S, Guieysse B, Muñoz R (2009) Mechanistic model for the reclamation of industrial
wastewaters using algal–bacterial photobioreactors. Environ Sci Technol 43:3200–3207.
https://doi.org/10.1021/es802156e
6. Su Y, Mennerich A, Urban B (2011) Municipal wastewater treatment and biomass
accumulation with a wastewater-born and settleable algal-bacterial culture. Water Res
45:3351–3358. https://doi.org/10.1016/J.WATRES.2011.03.046
 Algal Wastewater Treatment 115

7. Selvaratnam T, Henkanatte-Gedera SM, Muppaneni T, Nirmalakhandan N, Deng S,

Lammers PJ (2016) Maximizing recovery of energy and nutrients from urban wastewaters.
Energy 104:16–23. https://doi.org/10.1016/j.energy.2016.03.102
8. Henkanatte-Gedera SM, Selvaratnam T, Karbakhshravari M, Myint M, Nirmalakhandan N,
Van Voorhies W, Lammers PJ (2017) Removal of dissolved organic carbon and nutrients
from urban wastewaters by Galdieria sulphuraria: laboratory to field scale demonstration.
Algal Res 24:450–456. https://doi.org/10.1016/J.ALGAL.2016.08.001
9. Mehrabadi A, Craggs R, Farid MM (2015) Wastewater treatment high rate algal ponds
(WWT HRAP) for low-cost biofuel production. Bioresour Technol 184:202–214. https://doi.
org/10.1016/J.BIORTECH.2014.11.004
10. Park JBK, Craggs RJ, Shilton AN (2013) Enhancing biomass energy yield from pilot-scale
high rate algal ponds with recycling. Water Res 47:4422–4432. https://doi.org/10.1016/J.
WATRES.2013.04.001
11. Tchinda D, Henkanatte-Gedera SM, Abeysiriwardana-Arachchige ISA, Delanka-Pedige
HMK, Munasinghe-Arachchige SP, Zhang Y, Nirmalakhandan N (2019) Single-step
treatment of primary effluent by Galdieria sulphuraria: removal of biochemical oxygen
demand, nutrients, and pathogens. Algal Res 42:101578. https://doi.org/10.1016/J.ALGAL.
2019.101578
12. Martínez M, Sánchez S, Jiménez F, El Yousfi L, Muñoz L (2000) Nitrogen and Phosphorus
removal from urban wastewater by the microalga Scenedesmus obliquus. Bioresour Technol
73:263–272. https://doi.org/10.1016/S0960-8524(99)00121-2
13. Axelsson L (1988) Changes in pH as a measure of photosynthesis by marine macroalgae.
Mar Biol 294:287–294. https://doi.org/10.1007/BF00391314
14. Droop M (1975) The nutrient status of algal cells in batch culture. J Mar Biol Assoc UK55
(3):541–555. https://doi.org/10.1017/S0025315400017240
15. Chevalier P, Proulx D, Lessard P, Vincent WF, de la Noüe J (2000) Nitrogen and phosphorus
removal by high latitude mat-forming cyanobacteria for potential use in tertiary wastewater
treatment. J Appl Phycol 12:105–112. https://doi.org/10.1023/A:1008168128654
16. Henkanatte-Gedara S (2018) Urban wastewater treatment by Galdieria sulphuraria: single
step process for dissolved organics and nutrient removal. New Mexico State University
 Comparative Study of Spatial and Temporal
Variation of Drought Using Remotely Sensed
Data - A Case Study for Kirindi Oya Basin

S. U. Hendawitharana1(&), M. K. D. D. Priyasad2,
and R. L. H. L. Rajapakse1
1
Department of Civil Engineering,
University of Moratuwa, Katubedda, Sri Lanka
[email protected]
2
Department of Computer Science and Engineering,
University of Moratuwa, Katubedda, Sri Lanka

Abstract. Droughts have turned out to be a standout amongst the most dis-
cussed natural hazards in the recent past due to its adverse impacts on
humankind. The effects of drought differ from region to region and time to time.
Unlike, the other cataclysmic events, a drought is a mind-boggling phenomenon
due to its longer propagation time and the effects being long-lasting. This study
focuses on analysing the spatial and temporal variations of the drought in the
Kirindi Oya basin which lies in the intermediate, dry and arid climatological
zones. Daily remotely sensed precipitation data for a span of 37 years from
1979–2017 has been used with a spatial resolution of 0.1°. Several drought
indices, namely Standardized Precipitation Index (SPI), Deciles, Percent of
Normal, Rainfall Anomaly Index and Z index have been considered. A com-
parative analysis of the applicability of each drought index along with the
validation by ground reality is presented.
The results of the study imply the importance of conducting a study at the
sub-basin scale for the river basins lying in several vastly different climato-
logical regions. The Kirindi Oya basin that is considered, lies in three different
climatological regions, where the upper Kirindi Oya subbasin is in the inter-
mediate zone, the lower Kirindi Oya in the arid zone and the rest in the dry zone.
The results indicate that the effects of drought can vary even within a river basin
and a sub basin scale analysis of drought gives a comparatively accurate
interpretation of the variation of drought within the basin. Furthermore, the
recent findings confirms that the drought prediction at a sub basin scale is more
effective for river basins which lie in several climate zones.

Keywords: Drought indices
Remotely sensed data

Spatial and temporal analysis

1 Introduction

Drought can be identified as a natural phenomenon which propagates gradually and

unlike other natural disasters, its effects can last for an extended time (Wilhite 2000;
Shahabfar and Eitzinger 2013). The unavailability of a solitary definition for droughts

© Springer Nature Singapore Pte Ltd. 2020

R. Dissanayake and P. Mendis (Eds.): ICSBE 2018, LNCE 44, pp. 116–130, 2020.
https://doi.org/10.1007/978-981-13-9749-3_11
 Comparative Study of Spatial and Temporal Variation of Drought 117

has made decision making using drought-related studies more complex (Wilhite and
Glantz 1985; Wilhite 2000). For the simplicity of analysis, droughts have been cate-
gorized into four main types namely; meteorological, hydrological, agricultural and
socio-economic (Bayissa et al. 2018). All three former drought types occur following a
meteorological drought and an extended meteorological drought can lead to the latter
drought type (Zargar et al. 2011; Shah and Mishra 2015).
Spatial and temporal studies of droughts can be done using drought indices. The
drought index is a measure of drought and for each drought type, a different set of
drought indices have been developed. There are numerous drought indices developed
as a measure of different drought types and its appropriateness may differ depending
upon the geography, information accessibility and several other factors (Hayes 2017).
Drought indices provide a qualitative indication of droughts, where the results would
suggest the severity of the drought (Richard and Heim 2002; Svoboda and Fuchs
2016). However, the impacts from the same severity class can vary from region to
region and is dependent upon the climate conditions of the region under consideration.
Even though precipitation is the main input for most drought indices, it cannot be taken
directly as a measure of drought, as the amount of rainfall differs spatially (Svoboda
and Fuchs 2016). Thus, the drought indices tend to report the deviation from the mean
using different formulae, which sometimes end up in unrealistic interpretations. The
areas with very high annual rainfall may be considered as undergoing a severe drought
when the rainfall is reduced, even though the adverse impacts of a drought are not
significantly observed by the residents in that area. The complexity of the drought
phenomenon, the absence of a unique definition and data inaccessibility for a longer
duration may lead to variations in results given by different drought indices for the
same location and time period. Hence, it is a requirement to identify a suitable drought
index for a given area (Bayissa et al. 2018). Some indices such as Standardized Pre-
cipitation Index (SPI) and deciles give indications of several drought types, namely;
meteorological, hydrological and agricultural and this is due to the interdependency of
these droughts over the temporal scale (Zargar et al. 2011).
This study mainly focuses on two main aspects of drought evaluation. The first
aspect is identifying the applicability of drought studies in larger resolutions, especially
at sub-basin levels where the most of the empirical studies focus on an entire basin, a
large district or country and evaluating the usefulness of such a study in a country with
high ecological diversity. The other aspect is finding the most suitable drought eval-
uation methodologies (indices) to represent the drought patterns and behaviour in a
selected area. Most of the existing literature is based on larger extents, assuming that
the climatological patterns are approximately similar. However, the validity of such
methods for a country like Sri Lanka which has a higher variation in climatological
zones is questionable and the majority of the rivers originate from the central highlands
and reach the ocean at different destinations all around the country traversing over
several such zones.
One of the major challenges for a high spatial resolution drought study is the
availability of data. Precipitation and temperature are the most prominent parameters in
drought study. Temperature over a large extent of the area remains nearly constant
while rainfall patterns change rapidly. This difference cannot be easily tracked by
118 S. U. Hendawitharana et al.

traditional methodologies since the rainfall gauge stations are sparsely distributed.
Therefore, moving towards high resolution remotely sensed data sources is important.
The SPI is one of the most common drought indices used around the world,
including Sri Lanka. Other than that, many other drought indices have been considered
for research studies in different regions of the country.
Considering the drought studies conducted, it can be identified that most of the
studies on drought analysis have been conducted using gauge station data. Main issues
associated with gauge station data are the significant distances between the gauge
stations, high cost and the data discontinuity (Villarini et al. 2008). However, use of
remotely sensed data is becoming popular and the data sources selected by the
researchers include NOAA, TRMM, GMP, PERSIANN, MSWEP (Johnson et al.
1993; Thenkabail et al. 2004; Beck et al. 2017). The spatial and temporal resolution of
the available data is the main concern, since even though some regions of the world
have a complete set of satellite-based gridded data with high spatial accuracy, there are
limited data sources for Sri Lanka which have both the required temporal and spatial
resolutions. The data required for a drought study is high in comparison to other studies
and it requires a minimum of 30 years of continuous data in order to calculate the
available drought indices with sufficient accuracy (Juliani and Okawa 2017). Consid-
ering the inaccessibility of continuous data, drought-related studies using remotely
sensed data have received considerable attention and this study uses remotely sensed
gridded precipitation data with a spatial resolution of 0.1°.
The next challenge is determining the most suitable method for qualitative analysis
of drought impact and its severity. The SPI is one of the most common drought indices
used around the world including Sri Lanka. Other than that, many other drought indices
have been considered for research studies in different regions of the country.
Objectives of this research are,
• Evaluating the applicability and suitability of sub-basin scaled drought study over
basin-scaled drought study
• Evaluating different quantitative drought measures to identify the most suitable
measure
This paper critically examines the notion that drought-related studies can be more
accurate when they are conducted at the sub-basin scale using the Kirindi Oya as a case
study.
This study uses five selected drought indices with the intention of performing a
critical analysis of the applicability of them in the selected basin.

2 Study Area

Sri Lanka is a tropical island nation situated towards the North of equator in the Indian
Ocean. The island comprises of four main climatological regions; wet zone, dry zone,
intermediate zone and arid zone. Towards the southern central part of the island, a hilly
territory is observed while the coastal belt has a nearly flat terrain. These central hill
lands have a significant impact on the country’s climate. The island has an intricate
system of waterways which take water from the inland to the ocean.
 Comparative Study of Spatial and Temporal Variation of Drought 119

Considering the river basins of the island, they are either totally in one climate zone
or span over more than one climate zone. Hence, studying droughts in a river basin
should consider the fact that there can be an influence from the climate zone for
drought in that area. In order to study the spatial and temporal variation of meteoro-
logical drought, the authors have selected the Kirindi Oya basin, which spans in three
climate zones; intermediate zone, dry zone and arid zone while a major part lies in dry
and arid zones. According to the agro-ecological region map of Sri Lanka, the Kirindi
Oya spans over the zones; IM2, IU3, I(L1-L2), I(U2-U3), DL1 and DL5. Figure 1
indicates the location of the Kirindi Oya basin in Sri Lanka and the agro-ecological
zoning for Kirindi Oya basin is shown in Fig. 2.

Fig. 1. Kirindi Oya basin

The Kirindi Oya starts from Bandarawela and reaches the Indian Ocean at Bundala.
The basin is of a land extent of 1176 km2 and is situated in the Badulla, Moneragala
and Hambantota districts. High temperatures and low relative humidity leading to
higher evaporation and comparatively low mean annual rainfall are common features of
the area. A bimodal rainfall pattern can be observed in the area with a major proportion
of rainfall being contributed by the North-Eastern Monsoon (September–February)
which goes in line with the Maha season.
120 S. U. Hendawitharana et al.

Fig. 2. Agro-ecological regions - Kirindi Oya

3 Data and Methodology

Depending on the drought index being used, the past data requirement for drought
analysis varies. The SPI is the current drought index used in Sri Lanka which requires a
minimum of 30 years data (Juliani and Okawa 2017).
Obtaining continuous rainfall data for 30 years is challenging for some areas of the
country due to data discontinuity and other management issues. Remotely sensed pre-
cipitation data with a spatial resolution of 0.1° can be identified as an effective solution.
For this analysis, daily gridded precipitation data obtained from Multi-Source
Weighted-Ensemble Precipitation (MSWEP) was used for a temporal duration of 37
years (1979–2017).
Detection and real-time monitoring of droughts can be done by using drought
indices which enables us to identify the drought duration and drought severity. How-
ever, the definitions of drought indices differ from one another and the drought measure
is a qualitative one. Hence, a comparison of different drought indices is complex.
 Comparative Study of Spatial and Temporal Variation of Drought 121

For this study, five identified drought indices, namely Standardized Precipitation
Index (SPI), Deciles (DI), Percent of Normal (PN), Rainfall Anomaly Index (RAI) and
Z index, were calculated for each sub-basin in the Kirindi Oya basin.

3.1 Standardized Precipitation Index (SPI)

In calculating the SPI index, the rainfall received in a particular area is fitted to a
corresponding gamma distribution and the standard deviation from the long-term mean
is considered. This measure indicates the deviation from the historical mean rainfall of
a particular area and hence can be considered as a measure of drought. The SPI is
calculated for several durations; 1 month, 3 months, 6 months, 9 months, 12 months,
18 months, 24 months and 48 months. However, the SPI values corresponding to a
given month can differ from each other, when different durations are considered.
Hence, these different durations can be used when analysing diverse drought sce-
narios. The 1-month SPI would provide an indication of short-term soil moisture
conditions and crop moisture levels and is useful in decision- making during crop
growing seasons. The 3-month SPI reflects the seasonal rainfall estimates and mete-
orological drought. Indications of agricultural drought are given by 9-month SPI while
hydrological drought is reflected by 12-month SPI (Senay et al. 2015). The SPI is
calculated using Eqs. 1 and 2.

1 x
g(x) = xa1 e b ; x [ 0 ð1Þ
ba sðaÞ
P
lnð xÞ
A ¼ lnð xÞ  ð2Þ
n

Calculated SPI values for each month are categorized into regions as indicated in
Table 1 (Morid et al. 2006).

Table 1. SPI impact levels

SPI value Impact level
2.0 or more Extremely wet
1.5 to 1.99 Very wet
1.0 to 1.49 Moderately wet
0.99 to −0.99 Near wet
−1.0 to −1.49 Moderately dry
−1.5 to −1.99 Severely dry
−2.0 or less Extremely dry

The table is then converted to a common scale with the intention of analysing the
results. All the other drought indices are also converted to this common scale for the
ease of comparison.
122 S. U. Hendawitharana et al.

The drought period is indicated by the time period between the first month with
below −1.0 index value and first month after that month which has an index value
greater than or equal to 0.

3.2 Deciles (DI)

This method is one of the simplest methods used to calculate the impact of drought
where the precipitation data is arranged in decreasing order and the cumulative fre-
quency distribution is obtained. Then, it is split into ten parts using equal probability
distribution and each part is called a decile. According to the original work, precipi-
tations that are below the 5th decile is considered as drought (Sivakumar et al. 2010).
The original categorization is converted to the above-mentioned scale as in Table 2.

3.3 Percent of Normal (PN)

In this method, precipitation on a given month (Pi) is expressed as a percentage to the
long-term median/normal rainfall (Pn) of a particular area. Precipitations with PN value
below 80% is identified as droughts which are calculated as indicated in Eq. 3.

Pi
PN ¼  100% ð3Þ
Pn

However, since this method does not consider a time window, actual drought can
be identified by consecutive months with PN values in drought range. Table 2 indicates
the impact levels for the PN values.

Table 2. Impact levels

SPI RAI PN Z Index DI
No drought >0.99 >−0.3 >80 −0.25 to +0.25 5
Weak drought 0.99 to −0.99 −0.3 to −1.2 70–80 −0.52 to −0.25 4
Moderate drought −1.0 to −1.49 −1.2 to −2.1 55–70 −0.84 to −0.52 3
Severe drought −1.5 to −1.99 −2.1 to −3.0 40–55 −1.28 to −0.84 2
Extreme drought 2 or less <−3.0 <40 >−1.28 1

3.4 Rainfall Anomaly Index (RAI)

This index is used to calculate meteorological drought events and considered inde-
pendent of time and space. This is calculated as in Eq. 4, where Pn and m refer to mean
precipitation of a 30 years span and mean of 10 greatest rainfalls respectively. The
results are interpreted in Table 2.
 
P  Pn
RAI ¼  ð4Þ
mP
 Comparative Study of Spatial and Temporal Variation of Drought 123

3.5 Z Index
Z index is calculated using Eq. 5 where Pn and d refer to the mean precipitation and the
standard deviation of the region.

P  Pn
Z Score ¼ ð5Þ
d

Two approaches were taken in calculating the indices, where the first approach is to
consider the mean precipitation of each sub-basin in an index calculation. The other
approach was to consider the mean value for the basin as a whole. The results obtained
from the two approaches for five drought indices were compared and analysed.

4 Analysis of Results

In this study, spatial and temporal variations of drought within the Kirindi Oya basin is
analysed by using remotely sensed precipitation.

4.1 Rainfall Variation Within the Basin

First, the rainfall variation within the basin is analysed to ensure that our main
hypothesis – more accurate results can be obtained by sub-basin scaled analysis
remains valid for this basin. Since the precipitation distributions are significantly dif-
ferent among basins, it can be argued that drought patterns and severity may show
different distributions.

Fig. 3. Rainfall distribution histogram

124 S. U. Hendawitharana et al.

According to Fig. 3, Lower Kirindi Oya and Lunugamwehera sub-basins have

undergone more no-rainfall days compared to others. Moreover, they indicate a lower
frequency of high-rainfall days compared to the others. The highest proportion of
rainfall obtained for a given sub-basin differs from each other. The subbasins closer to
each other show a slight similarity while the subbasins away from each other show a
significant variation.

Fig. 4. Distribution of rainfall

According to Fig. 4, all the sub basins have a median of *50 mm rainfall.
However, outlier points (precipitation greater than 1.5IQR, extreme rainfall conditions)
can be observed in each sub-basin. Kuda Oya has the least number of outlier points and
they are concentrated around 150 mm whereas Upper Kirindi Oya and Maha Ara
which shows extreme rainfall events *250 mm.

4.2 Spatial Analysis

Spatial distribution of drought categories is indicated in Fig. 5. The frequency of
occurrence of each drought type identified by each drought index is shown. From the
total of 456 months considered, *200 months show a no drought condition according
to the majority of the indices. Z score has identified the lowest number of no drought
months.
 Comparative Study of Spatial and Temporal Variation of Drought 125

The drought indices, PN, RAI, Z score and SPI1 have identified a significant
variation in occurrence of extreme drought events within the basin and Lower
Lunugamwehera sub basin shows the highest frequency.
The indices, DI, PN, RAI, Z score and SPI 1 have been calculated for 1-month time
scale. However, the drought identification shows that SPI 1 has identified remarkably
low number of extreme, sever and moderate drought events compared to the other
indices. In contrast, SPI for all the time scales considered have identified higher number
of weak drought events.

Fig. 5. Drought variation in basins

It can be seen that a significant difference in the number of drought events of

drought classes among DI, PN, RAI, ZS which do not use time windows and SPI which
uses time windows respectively. Figures, 6, 7 and 8 represent temporal drought vari-
ation of three main basins.
PN has identified more drought events in all the three basins compared to the other
indices due to calculating drought in one-month intervals. This gives an inaccurate
result where ‘no precipitation’ events in between heavy rain events are also identified
as droughts. Therefore, it is suitable to use a windowing technique which can get an
average over windowing period and then quantify drought which results in an accurate
estimation. For this basin under consideration, SPI can be identified as the most suitable
one.
Confusion matrices in Figs. 9 and 10 explain the correlation of drought events
among sub-basins. Considering the drought events that occurred for the period under
consideration, for the spatial variation of the subbasins, the matrices provide a clear
picture of the drought occurrence.
126 S. U. Hendawitharana et al.

Fig. 6. Temporal drought variation in Kirindi Oya

Fig. 7. Temporal drought variation in Lunugamwehera

 Comparative Study of Spatial and Temporal Variation of Drought 127

Fig. 8. Temporal drought variation in Lower Kirindi Oya

Fig. 9. Correlation matrix - SPI 3

128 S. U. Hendawitharana et al.

Fig. 10. Correlation matrix - PN

Comparing SPI 3 and PN matrices, it can be observed that PN has identified a

higher variation of drought among the sub-basins while SPI 3 shows a comparatively
low variation.
Considering SPI 3, it can be observed that the Upper Kirindi Oya has a high
correlation to Kuda Oya and very low correlation to Lower Kirindi Oya. According to
the spatial distribution of basins, Kuda Oya is the nearest sub-basin to the Upper kirindi
Oya and is in the same agro-ecological region where Lower Kirindi Oya is the farthest
and in a different agro-ecological region.
Both Maha Ara and Lunugamwehera sub-basins have a high correlation to their
nearby sub-basins and a low correlation to the farthest sub-basins. Hence, it can be
concluded that nearby sub-basins may show similar drought patterns whereas the
farthest sub-basins show a different pattern depending on the agro-ecological and
climate region differences.
Same correlation distribution can be identified in PN drought index with slightly
low correlation for all subbasins compared to SPI-3.

5 Conclusion

The drought identification using the drought indices considered in this study shows that a
higher number of drought events including all the drought events identified by SPI have
been identified by PN, DI, RAI and Z score. However, they have also identified a
significant difference in rainfall in one month which falls amidst a drought event as the
 Comparative Study of Spatial and Temporal Variation of Drought 129

cease of drought which is impractical. Hence, drought predictions at the Kirindi Oya
basin can be effectively done using drought indices which use windowing technique;
hence, SPI.
Drought events in the Upper Kirindi Oya basin has significant differences during
some drought events while the drought variation in the nearby sub-basins being nearly
similar is the main observation of this study. Hence, the findings of this study help
understand that the effects of drought can vary even within a river basin and, that sub-
basin-scale analysis of drought gives a comparatively accurate interpretation of vari-
ation of drought within the basin. Furthermore, the present findings confirm that the
drought prediction at a sub-basin scale is more effective for river basins which lie in
several climatological regions.

Acknowledgement. The authors would like to acknowledge that the research project was
financially supported by the Senate Research Committee Grant (SRC/LT/2018/25), University of
Moratuwa.

References
Bayissa Y et al (2018) Comparison of performance of six drought indices in characterizing
historical drought for the Upper Blue Nile basin. Geosciences 8(3):81
Beck HE et al (2017) MSWEP: 3-hourly 0.25 global gridded precipitation (1979–2015) by
merging gauge, satellite and reanalysis data. Hydrol Earth Syst Sci 21:589–615
Hayes MJ (2017) Drought indices. The intermountain west climate summary
Johnson GE, Achutuni VR, Thiruvengadachari S, Kogan F (1993) The role of NOAA satellite
data in drought early warning and monitoring: selected case studies. Kluwer Academic
Publishers
Juliani BHT, Okawa CMP (2017) Application of a standardized precipitation index for
meteorological drought analysis of the semi-arid climate influence in Minas Gerais. Brazil.
Hydrology 4(2):26
Morid S, Smakhtin V, Moghaddash M (2006) Comparison of seven meteorological indices for
drought monitoring in Iran. Int J Climatol 26:971–985
Richard R, Heim J (2002) A review of twentieth-century drought indices used in the United
States. Am Meteorol Soc 83(8):1149–1166
Senay G et al (2015) Drought Monitoring and Assessment: Remote Sensing and Modelling
Approaches for the Famine Early Warning Systems Network. In: Drought Monitoring and
Assessment, pp 233–262. USGS Staff
Shahabfar A, Eitzinger J (2013) Spatio-temporal analysis of drought in semi-arid regions by
using meteorological drought indices. Atmosphere 4:94–112
Shah RD, Mishra V (2015) Development of and experimental near-real-time drought monitor for
India. J Hydrometeorol 16:327–345
Sivakumar MVK et al (2010) Drought risk and meteorological droughts
Svoboda M, Fuchs BA (2016) Handbook of drought indicators and indices. Integrated drought
management programme (IDMP), integrated drought management tools and guidelines series
2. World Meteorological Organization (WMO) and Global Water Partnership, Geneva
Thenkabail PS, Gamage MSD, Smakhtin VU (2004) The use of remote sensing data for drought
assessment and monitoring in Southwest Asia. International Water Management Institute,
Research Report 85
130 S. U. Hendawitharana et al.

Villarini G, Mandapaka PV, Krajewski WF, Moore RJ (2008) Rainfall and sampling
uncertainties: a rain gauge perspective. Adv Earth Space Sci 113(D11). https://agupubs.
onlinelibrary.wiley.com/doi/full/10.1029/2007JD009214
Wilhite DA (2000) Drought as a natural hazard: concepts and definitions. In: Drought: a global
assessment. Drought Mitigation Center Faculty Publications, pp 3–18
Wilhite DA, Glantz MH (1985) Understanding: the drought phenomenon: the role of definitions.
Water Int 10(3):111–120
Zargar A, Sadiq R, Naser B, Khan FI (2011) A review of drought indices. Environ Rev 19:333–
349
 Potential for Formation of Trihalomethane
in Diverted and Non-diverted Areas
of Mahaweli River in Sri Lanka

H. R. L. C. Bandara(&), W. B. M. L. I. Weerasekara,
and S. K. Weragoda

Water Safety Plan Advisory Unit, National Water Supply and Drainage Board,
Kandy, Sri Lanka
[email protected]

Abstract. Under the Mahaweli Development project the Mahaweli River was
diverted to Rajarata area at polgolla barrage in 1976. Along the river concentration
of the dissolved organic (DO), inorganic and suspended solids are gradually
increased. In drinking Water Treatment Plants (WTPs) can be removed mainly
suspended solid and some part of the dissolved component. As a result of
remaining DO in filtered water disinfection by products (DBPs) formed such as
Trihalomethane (THM) which is carcinogenic. This study was conducted to
investigate THM formation potential in Mahaweli Diverted Areas (DA) and Non-
Diverted Areas (NDA). Water samples were collected from following WTPs;
Ukuwela, Matale, Dambulla, Bakamuna, Minneriya, Medirigiriya and 2- Ela as a
DA and Meewathura, Greater Kandy, Pathadumbara, Mahiyanganaya, Dehiat-
takandiya, Gallella, Kantale and Mutur as a NDA. The raw water source of
Ukuwela and Matale is Sudu Ganga and others are Ibbankatuwa Lake, Yoda Ela,
Minneriya Lake, Kaudulla Lake and Parakrama Samudra respectively. Four THM
species, Trichloromethane, Bromodichloromethane, Dibromochloromethane and
Tribromomethane were measured using Gas Chromatography (GC) system and
finally Total THM (TTHM) were calculated. The highest TTHM concentration
was reported in 2- Ela (84.67 ± 35.17) µg/L and Mutur (83.85) µg/L which is
exceeded the United State Environmental Protection Agency (USEPA) maximum
contaminant level of 80 µg/L. Lowest concentration was reported from Matale
8.16 µg/L. According to the Kruskal-Wallis Test the median TTHM concentra-
tion of DA (22.73 µg/L) was not significantly different (p > 0.05) from the NDA
(25.36 µg/L). The mean Conductivity (lS/cm), pH, turbidity (NTU) and alka-
linity (mg/l) of the DA were 224.25 ± 65.95, 7.48 ± 0.36, 3.69 ± 1.91 and
87.5 ± 28.72 respectively and in NDA; 139.52 ± 62.39, 7.31 ± 0.37,
12.71 ± 12.13 and 56.25 ± 28.99 respectively. According to the statistical
analysis there was no significant difference (p > 0.05) between the measured
water quality parameters of DA and NDA.

Keywords: Diverted area
Mahaweli river
Non-Diverted Area

Trihalomethane
Water treatment plant

© Springer Nature Singapore Pte Ltd. 2020

R. Dissanayake and P. Mendis (Eds.): ICSBE 2018, LNCE 44, pp. 131–136, 2020.
https://doi.org/10.1007/978-981-13-9749-3_12
132 H. R. L. C. Bandara et al.

1 Introduction

The Mahaweli River is the longest river in Sri Lanka which starts in the Central
Mountains of the country. The mean annual runoff of the river is 7.2 million acre feet.
(Mahaweli Development Board 1970). The United Nations Development Programme
(UNDP) and the Food and Agriculture Organization (FAO) formulated the master plan
for the development of the Mahaweli river basin in 1960. (Hewavisenthi 2009). In
1976, the Mahaweli River was diverted to the Rajarata area at the Polgolla barrage.
Researchers have found that most patients having Chronic Kidney Diseases due to
unknown aetiology (CKDu) were reported in the Mahaweli Diverted area. (Bandara
et al. 2010). Along the river, the concentration of the dissolved organic (DO), inorganic
and suspended solids is gradually increasing. In drinking Water Treatment Plants
(WTPs) suspended solid can be removed mainly and some part of the dissolved
components.
Most of the water treatment plants in Sri Lanka have been using chlorine as a
disinfectant. Chlorination can help to reduce risk of water borne diseases such as
typhoid, cholera and hepatitis (Morris and Levine 1995). As a result of chlorination in
water treatment plants, disinfection by-products (DBPs) can be formed. (Weragoda
2005). The formation of these compounds is a function of precursor concentration,
contact time, chlorine dose, and pH (APHA 1995). Trihalomethanes (THMs) can be
considered as one of most important group of DBPs. Trihalomethanes (THMs) are
halogen-substituted single-carbon compounds with the general formula CHX3, where
X represents a halogen, which may be fluorine, chlorine, bromine, or iodine, or
combinations thereof. The THMs most commonly present in drinking-water are Tri-
chloromethane (CHCl3), Bromodichloromethane (CHCl2Br), Dibromochloromethane
(CHBr2Cl) and Tribromomethane (CHBr3). This study was conducted to investigate
THM formation potential in the Mahaweli Diverted Areas (DA) and Non-Diverted
Areas (NDA).

2 Methodology

2.1 Sampling Location

Water samples were collected from the following WTPs; Ukuwela, Matale, Dambulla,
Bakamuna, Minneriya, Medirigiriya and 2- Ela as DAs and Meewathura, Greater
Kandy, Pathadumbara, Mahiyanganaya, Dehiattakandiya, Gallella, Kantale and Mutur
as NDAs. The raw water source of Ukuwela and Matale is Sudu Ganga and others are
Ibbankatuwa Lake, Yoda Ela, Minneriya Lake, Kaudulla Lake and Parakrama Samudra
respectively.

2.2 Analytical Method

Water samples were collected to the pre washed clear glass bottles. Samples were pre-
served by using ascorbic acid. Recommended preservation dose of ascorbic acid was
25 mg for 40 ml of sample. Then, samples were transported to the laboratory by
 Potential for Formation of Trihalomethane in Diverted and NDA 133

maintaining temperatures below 4 °C. Four THM species, Trichloromethane (CHCl3),

Bromodichloromethane (CHCl2Br), Dibromochloromethane (CHBr2Cl) and Tribro-
momethane (CHBr3) were measured by using Thermo scientific trace 1300 computerized
gas chromatograph system with electron capture detector (GC-ECD) coupled with Tri
Plus RSH auto sampler, (Thermo Fisher scientific Inc., USA). GC-ECD capillary column
used to separate THMs (TRACE TR-5 GC) having 15 m  0.25 mm  0.25 lm film
thickness. Samples were injected through auto sampler (Sampler consist heating, incu-
bation agitation oven and standard X-length rail for headspace injection) in to split/split
less injector. GC-ECD system was configured according to headspace THMs determi-
nation method proposed by (Kuivinen and Johnsson 1999). Finally, Total THM (TTHM)
was calculated, which is a cumulative value of the four THM species. Conductivity, pH
and turbidity were measured by using portable water quality testing equipment. Alka-
linity was measured by a titrimetric method.

3 Results and Discussion

3.1 Diverted Area
Previous studies indicate that there was adverse water quality effect in Mahaweli River
water after the diversion, especially related to the Chronic Kidney Diseases due to
Unknown aetiology. (Bandara et al. 2010). One of the main targets of this study was to
identify the THM formation potential in Mahaweli DA (Table 1).

Table 1. Reported THM concentration (µg/L) in WTP in Mahaweli DA

Location CHCl3 ± SD CHCl2Br ± SD CHBr2Cl ± SD CHBr3 ± SD
Ukuwela 17.6 4.53 0.6 ND
Matale 5.445 1.93 0.785 ND
Dambulla 13.78 2.99 0.37 ND
Bakamuna 10.58 ± 0.21 5.78 ± 1.01 2.13 ± 0.61 ND
Minneriya 16.23 ± 2.43 7.97 ± 0.62 2.18 ± 0.07 ND
Medirigiriya 13.74 ± 3.33 4.83 ± 0.34 1.23 ± 0.36 ND
2-Ella 67.94 ± 29.45 14.59 ± 4.81 2.14 ± 0.90 ND

Lower concentration of THM values were reported in the Mahaweli DAs compared
to the NDAs. The following graph shows the variation of TTHM in Mahaweli DAs.
Highest TTHM values were reported in 2-Ella WTP which extracts source water from
the Parakrama Samudra. Water flow was divided into two directions from the Elahera
anicut. One flowing toward the Minneriya and Kaudulla tank along the Elehera Min-
neriya Yoda Ella and the other one flowing toward the Parakrama Samudra with the
combination of the Kalu Ganga. Parakrama samudra can be more polluted from
agricultural runoff coming from both the Kalu Ganga and the surrounding area (Fig. 1).
134 H. R. L. C. Bandara et al.

Fig. 1. Variation of TTHM concentration in Mahaweli Diverted area

3.2 Non Diverted Area

The Mahaweli river starts its journey in the wet zone of Sri Lanka and flows through
the dry zone. It carries most organic and inorganic pollutants from the urban and
agricultural areas. Organic level gradually increases when it comes downstream.
Therefore, THM formation potential also high in those areas. The following table
shows the recorded THM values in WTP of Mahaweli NDAs (Table 2).

Table 2. Reported THM concentration (µg/L) in WTP in Mahaweli NDA

Location CHCl3 ± SD CHCl2Br ± SD CHBr2Cl ± SD
Meewathura 7.58 ± 0.71 1.55 ± 0.10 0.70 ± 0.10
Greater Kandy 15.98 ± 8.18 4.47 ± 1.78 0.76 ± 0.58
Pathadumbara 27.64 ± 12.4 3.99 ± 1.29 0.40 ± 0.08
Mahiyanganaya 23.14 ± 1.29 6.61 ± 0.33 1.70 ± 0.12
Dehiaththakandiya 15.46 ± 8.46 5.58 ± 2.02 1.37 ± 0.45
Gallella 41.89 ± 18.1 9.72 ± 4.71 1.60 ± 0.91
Kantale 20.67 7.07 1.78
Mutur 63.87 16.76 3.22

The highest concentration of Trichloromethane, Bromodichloromethane and

Dibromochloromethane (CHBr2Cl) were recorded in the Mutur WTP and the lowest
values were recorded in Meewathura WTP. The Meewathura WTP is located in an
upstream area. Therefore, water contains less amounts of organic matter. Electrical
conductivity and turbidity value was 60.0 µs/cm and 3.23 NTU respectively.
 Potential for Formation of Trihalomethane in Diverted and NDA 135

Mutur is the final WTP which extracts source water from the Mahaweli River
before entering the sea. Source water flows through many of the agricultural areas and
contains considerable levels of organics. Electrical conductivity and turbidity value was
210.0 µs/cm and 17.7 NTU respectively in Mutur WTP source water.
Tribromomethane was not reported in any WTP along the Mahaweli River. Tri-
bromomethane forms when water contaminates with Bromide. Bromide reacts with
applied disinfectants and forms bromine, which oxidizes organic matter and finally
creates Tribromomethane. (Singer 1994; McTigue et al. 2014). Water can be polluted
from the bromide due to anthropogenic activity such as sewage or industrial effluent
discharges and landfill leachate. (Siddiqui et al. 1993; Boyer et al. 2013).
The following graph show the variation of TTHM (cumulative value of four THM
species) along the river. TTHM values gradually increased along the river. But there
were a few reductions at the Dehiaththakandiya and Kantale WTP. TTHM concen-
tration can be largely affected by treatment adequacy and the treatment technology of
the WTP. However, TTHM concentration of the all WTP were below the USEPA
maximum contamination level of 80 µg/L except at Mutur WTP (Fig. 2).

Fig. 2. Variation of TTHM concentration in Mahaweli Non-Diverted area

However, according to the Kruskal-Wallis Test, the median TTHM concentration

of DAs (22.73 µg/L) was not significantly different (p > 0.05) from the NDAs
(25.36 µg/L). The mean Conductivity (lS/cm), pH, turbidity (NTU) and alkalinity
(mg/l) of the DAs were 224.25 ± 65.95, 7.48 ± 0.36, 3.69 ± 1.91 and 87.5 ± 28.72
respectively and in the NDAs; 139.52 ± 62.39, 7.31 ± 0.37, 12.71 ± 12.13 and
136 H. R. L. C. Bandara et al.

56.25 ± 28.99 respectively. According to the statistical analysis, there was no sig-
nificant difference (p > 0.05) between the measured water quality parameters of DAs
and NDAs.

4 Conclusion

This study was conducted to investigate THM formation potential in the Mahaweli
Diverted areas and Mahaweli Non-Diverted areas. Some researchers have found that
there was an adverse effect on water quality after the diversion. Large number of CKDu
patient were reported in the Mahaweli Das, especially in the North Central province.
However, it was found that THM concentration in the Mahaweli DAs and NDAs were
below the USEPA maximum contamination level of 80 µg/L except at Mutur WTP.
According to the statistical analysis, it was found that there was no significant differ-
ence in THM concentration between Mahaweli DAs and NDAs.

References
APHA, AWWA, WEF (1995) Standard methods for the examination of water and wastewater,
19th edn, Washintion D.C
Bandara JMRS, Senevirathna DMAN, Dasanayake DMRSB, Herath V, Bandara JMRP,
Abeysekara T, Rajapaksha KH (2008) Chronic renal failure among farm families in cascade
irrigation systems in Sri Lanka associated with elevated dietary cadmium levels in rice and
freshwater Wsh (Tilapia). Environ Geochem Health 30:465–478
Boyer TH, Ged E, Motz L, Chadik P, Frank K, Hill CP (2013) Impact of sea-level rise on
saltwater intrusion and formation of brominated disinfection byproducts. In: Water quality
technology conference and exposition (WQTC) conference proceedings. American Water
Works Association, Denver
Hewavisenthi ACS (2009) Management of the Mahaweli, a river in Sri Lanka. Water Int 22
(2):98–107
Kuivinen J, Johnsson H (1999) Determination of trihalomethanes and some chlorinated solvents
in drinking water by headspace technique with capillary column gas-chromatography. Water
Res 33(5):1201–1208
Mahaweli Ganga Development Sri Lanka (1970) Mahaweli Development Board of Sri Lanka
McTigue NE, Cornwell DA, Graf K, Brown R (2014) Occurrence and consequences of increased
bromide in drinking water sources. J Am Water Works Assoc 106(11):E492–E508
Morris RD, Levine R (1995) Estimating the incidence of waterborne infectious disease related to
drinking water in the United States. Reichard
Siddiqui MS, Amy GL (1993) Factors affecting DBP formation during ozone–bromide reactions.
J Am Water Works Assoc 85(1):63–72
Singer PC (1994) Control of disinfection by-products in drinking water. J Environ Eng 120
(4):727–744. https://doi.org/10.1061/(ASCE)0733-9372(1994)120:4(727)
Weragoda SK (2005) THM formation modeling in treated water using the rapid organic
characterization technique. http://www.faculty.ait.ac.th/visu/Data/AIT-Thesis/MasterThesis
final/Weragoda05.pdf
 Studying About the Manganese Phase
Diagram When Manganese Is Reacting
with Calcium Hypochlorite

S. A. Abeysinghe1(&), S. K. Weragoda2,
and W. B. M. L. I. Weerasekara2
1
Department of Science and Technology, Faculty of Science and Technology,
Uva Wellassa University, Badulla, Sri Lanka
[email protected]
2
National Water Supply and Drainage Board, Kandy, Sri Lanka

Abstract. Oxidation Reduction behaviour is highly contributes to the condition

of the water. Oxidation reduction potential [ORP] gives an idea about the Redox
behaviour of water while the pH value give an indication about the behaviour of
the Hydrogen ion concentration. When Manganese is in water, it can exist either
as particulate matter or as dissolved matter and this can be showed in Man-
ganese phase diagrams. Due to the inconvenient consequences, Manganese
should be removed from the water if only it has a higher concentration. For this,
oxidation can be done by using Calcium hypochlorite and the obtained pre-
cipitate can be filtered. The behaviour of different Manganese species fluctuates
with different pH values and Eh values. The objective of this study was to
determine the effect of pH and Eh while changing the different concentration of
Calcium hypochlorite to the blended Manganese solution which is prepared by
using MnSO4. Final concentration of Manganese was reduced with lower
concentration of Calcium hypochlorite and ORP values were reduced with the
increasing pH. To conclude the result, Mn3O4, Hausmannite was obtain as the
Manganese species according to the Manganese phase diagram.

Keywords: Oxidation reduction potential
Redox behavior
Eh

Manganese phase diagram
Calcium hypochlorite
Hausmannite

1 Introduction
1.1 Effect of Manganese to the Water
Manganese naturally exists in earth’s crust and it aggregates to water due to the
dissolution of Manganese bearing minerals and rocks such as Manganosite, Manganite,
Bixbyte when there is weathering. Until Manganese is in groundwater, Manganese
exists as a stable form, but as soon as it is exposed to the outside air it oxidizes and
precipitates out naturally (Salem et al. 2012).
Manganese generally does not affected human health but it may cause undesirable
taste, appearance and straining (Rozainy et al. 2015). Presence of Manganese bearing

© Springer Nature Singapore Pte Ltd. 2020

R. Dissanayake and P. Mendis (Eds.): ICSBE 2018, LNCE 44, pp. 137–146, 2020.
https://doi.org/10.1007/978-981-13-9749-3_13
138 S. A. Abeysinghe et al.

bacteria in water distribution systems causes bacterial growth in pipes. From the
consumer’s view, washing machines and boilers can be damaged due to oxidization
and precipitation under conditions of excess Manganese (Barlokova and Ilavský 2009).

1.2 Manganese Phase Diagram

The Manganese phase diagram, (Fig. 1) shows the relationship of the redox potential
and hydrogen-ion concentration that are relevant with specific dissolved ions and solids
at chemical equilibrium in the presence of water. Phase boundary lines are called
Stability field boundaries and in their activities, each adjacent species is nearly equal.

Fig. 1. Manganese phase diagram

pe is accessible as tendency of solution to donate or accept protons while Eh is

known as the potential of the solution relative to the Standard Hydrogen Electrode
[SHE]. In here SHE is the reference point and when all conditions are constant no
reactions occur and the Eh value can be obtained from connecting the electrode con-
taining another solution.
When Eh is higher, the solution is desirable to accept electrons from another
species and when Eh is lower, the solution is desirable to donate electrons to the
species. Therefore solution is reducing and oxidizing respectively.

1.3 Redox Reactions

Redox reactions are typically happens in a solution by oxidizing or reducing an electron.
Hence, these reactions of oxidation or reduction do not exist in nature individually
because free electrons never exist in a solution. Releasing of any electron from one species
and its gain by another species happens simultaneously (Appelo and Pstma 2005).
 Studying About the Manganese Phase Diagram 139

Ox1 þ Red2 ! Red1 þ Ox2

Due to this, pe of the solution cannot be measured practically. Eh can be discovered

not only by measuring but also by calculations of the Nernst equation. Therefore, Eh is
the integration of both the calculated Eh and the measured Eh (Christensen et al. 2000).

Eh ¼ Emeasured þ Ereference ð1Þ

Both pe and Eh measure essentially the same and therefore the relationship between
them can be written as follows;

2:303 RT
Eh ¼ pe ð2Þ
F

1.4 Behavior of Manganese in Water

Manganese is the 25th element of the periodic table and the most stable oxidation state
is +2. When Manganese reacts with water, Mn(OH)2 is obtained but it can change with
the condition of the water.

Mn2 þ þ 2H2 O ! MnðOH Þ2 þ H2

1.5 Behavior of the Calcium Hypochlorite in Water

Calcium hypochlorite, which has the commercial name of bleaching powder, is an
inorganic compound. It is mostly used in water treatments because of the high effi-
ciency and low cost. It exists as a powder form and it should be dissolved with water to
acquire Calcium hypochlorite solution.

CaðOClÞ2 þ 2H2 O ! CaðOH Þ2 þ 2HClO

1.6 Reaction of Manganese and Calcium Hypochlorite

While Ca(OH)2 is a precipitate Mn(OH)2 is react with HClO.

MnðOH Þ2 þ 2HClO ! MnO2 þ CaCl2 þ 2H2 O

1.7 Objective of the Study

To investigate the behavior of pH and Eh of the solution while changing the different
concentration of Calcium hypochlorite, in certain concentrations of the Manganese rich
solution by studying the Manganese phase diagram.
140 S. A. Abeysinghe et al.

To discover the particulate form of the Manganese species which can be removed
by filtration.

2 Materials and Methodology

2.1 Analytical Techniques

2.1.1 Determination of pH Value
pH values of the water samples were directly measured by the pH meter. Before that,
the pH electrode was removed from the storage solution, rinse with distilled water and
dried with tissue paper. Then, the water sample was taken to the beaker and the pH
electrode was placed in it and the value on the digital screen was recorded. This pH
value is at fixed temperature.

2.1.2 Determination of Concentration of Manganese

UV spectrophotometer was used to measure the dissolved Manganese concentration of
the sample. A program was selected to measure the concentration of Manganese and
the wavelength was kept as 525 nm. Water sample was filled to the two 10 ml cells and
one is kept as reference. A Manganese citrate buffer powder pillow was added to the
other cell and shaken until it dissolved well. Then, Sodium periade reagent was added
and kept for two minutes to complete the reaction. After two minutes, the reference
sample was placed inside the hole of the UV spectrophotometer and calibrated to zero.
After it became zero, the working cell was placed and the reading on the digital screen
of UV spectrophotometer was recorded.

2.1.3 Determination of Oxidation Reduction Potential (ORP) Value

Same pH meter was used to measure the ORP value and it demonstrated the reference
Eh value. To get the ORP value, the program was changed by pressing the ‘Option’
button. ORP value is record as millivolt.

2.1.4 Determination of Residual Chlorine (RCl) Value

Residual chlorine value gives the remaining free chlorine amount of the sample. It can
be measured directly by using RCl meter. For that, the sample is filled into two cells
and one was kept as reference. DPD free chlorine reagent was added to the other
sample and shaken well until it fully dissolved. If Chlorine is present in the sample,
light pink color appears. The reference sample was placed in the hole inside the RCl
meter and calibrated to Zero and then the working sample was placed and the value
given in the digital screen was recorded.

2.2 Preparation of Manganese Solution

5 L of 0.9 ppm Mn2+ solution was freshly prepared by dissolving 0.0124 g of MnSO4
in to raw water sample collected from Maduru Oya. A raw water sample was used
instead of tap water or distilled water, in order to obtain the natural environmental
condition.
 Studying About the Manganese Phase Diagram 141

2.3 Preparation of Calcium Hypochlorite Solution Series

1L of 10 ppm Calcium hypochlorite solution was made by dissolving 0.01 g of
Bleaching powder in to distilled water. Bleaching solution series was selected with
respect to the Residual Chlorine value given from the stock sample.

2.4 Method
Initial concentration of the Mn2+ measured by using UV spectrophotometer and
Bleaching solution series was added to the 4 beakers which were filled with 500 ml of
Mn2+ sample. Each sample was mixed continuously using a shaker for almost 30 min.
Finally pH, RCl, ORP and final concentration of Manganese was measured.

3 Result and Discussion

3.1 Calculating the Standard Potential

When considering the redox reactions of Mn2+ and ClO−
Oxidation reaction of Mn2+

Mn2 þ þ 2H2 O ! MnO2 þ 4H þ þ 2e

Reduction reaction of ClO−

ClO þ 2e þ H2 O ! Cl þ 2OH 

Final reaction of Mn2+ and ClO−

Mn2 þ þ ClO þ H2 O ! MnO2 þ 2H þ þ Cl

According to Nernst equation

ðH þ Þ ðCl Þ
2
RT
E ¼ E0 þ lnð Þ
2F ðMn2 þ ÞðClO Þ

The standard potential at 1 atm and 25 °C for oxidation reaction is +1.23 V, while
for reduction reaction is +0.9 V.
To obtain E 0 for the final reaction, Standard potential of the reduction reaction
should be subtracted from standard potential of the oxidation reaction. It is expressed as
potential per electron.
Therefore standard potential of the final reaction is 1.23 V − 0.9 V = 0.33 V.

3.2 Removal Efficiency

Table 1 represents the removal efficiency of Manganese after adding the bleaching
powder solution. Initial concentration of Manganese is 0.9 ppm
142 S. A. Abeysinghe et al.

After 30 min, in each trial, Residual Chlorine value was zero. It indicated that at
that moment, all of the bleaching powder was almost finished.

Table 1. Final concentration of Mn with respect to different dosages of ClO− after 30 min of
reaction
Trial Concentration of Concentration of Removal Residual chlorine
ClO−/ppm Final Mn2+/ppm efficiency/% value/ppm
1 0.7 0.6 33.33 0
2 2 0.7 22.22 0
3 4 0.8 11.11 0
4 6 0.9 0 0

The variations of removal efficiency and the concentration of ClO−] are shown in
Fig. 2. It indicated that increasing concentration of ClO− reduces the Manganese
removal efficiency.

Fig. 2. Graph of removal efficiency vs concentration of ClO−

Since this is the result of synthetic Manganese contained in raw water, it did not
give the natural condition of raw water and it also influences this circumstance because
in nature, Manganese mostly coexists with Ferrous (Kasim et al. 2017).

3.3 Relationship Between pH and ORP

Figure 3 presented the relation between pH and ORP showed in Table 2. Since the
concentration of ClO− is zero, only an oxidation reaction is occurring. According to
that, the reaction produces H+ and due to that, pH reduces. Therefore, a higher con-
centration of final Manganese has a lower pH.
Figure 3 indicates that the ORP value is decreasing with increasing pH. For the
most part, ORP increased with increasing oxidant dosage (James et al. 2004). In this
experiment, ClO− is oxidant and from trial 1 to trial 4, concentration of the concen-
tration of ClO− is increased. Therefore, ORP value is increasing from trial 1 to trial 4.
 Studying About the Manganese Phase Diagram 143

Fig. 3. Graph of ORP value and pH

Table 2. pH and ORP value of the each sample

Trial Temperature/°c pH ORP [Emeasured]/mV
1 25.3 7.61 −28.5
2 25.2 7.27 −8.6
3 25.2 7.04 6.7
4 25.3 6.99 8.2

3.4 Calculating Ehreference

Ehreference value depends on the concentrations of Mn2+, ClO−, H+, and Cl−. Final
concentration of Cl− is zero, and under that condition Ehreference value of the each trial
is equal to the standard potential of the reaction as shown in Table 3.

Table 3. Ehreference values calculated from Nernst equation for each trial
Trial Concentration of Concentration of Concentration Concentration Temperature/K Ehreference/
Mn2+/ppm ClO−/ppm of H+/ppm of Cl−/ppm V
1 0.6 0.7 10.22  10−8 0 298.45 0.33
2 0.7 2.0 9.12  10−8 0 298.35 0.33
3 0.8 4.0 5.37  10−8 0 298.35 0.33
4 0.9 6.0 2.45  10−8 0 298.45 0.33

Since the Eh value is the integration of both ORP value and the calculated Eh value
from Nernst equation, Table 4 illustrates the final Eh value and pe value, which is
acquired from the relationship of Eh and pe (Appelo and Pstma 2005).
Since the higher pe value has higher tendency to accept protons, according to the
obtained pe, propensity to accepting protons is developed from trial 1 to trial 4
144 S. A. Abeysinghe et al.

respectively as shown in Table 5. Therefore, the solution of trial 1 is more oxidized

than the trial 2 solution.
Figure 4 shows that Eh is reduce with increasing pH of the solution.

Table 4. Indicate that Ereference values influence to the Eh more than Emeasured values.
Trial Emeasured/V Ereference/V Eh/V pe/V
4 −0.0285 0.33 0.3015 5.0906
3 −0.0086 0.33 0.3214 5.4285
2 0.0067 0.33 0.3367 5.6869
1 0.0082 0.33 0.3382 5.7103

In this experiment, pH of the samples exist between 6.8 and 7.8. According to the
Manganese phase diagram shown in Fig. 4, at that level, Manganese can exist as a
dissolved species (Mn2+) or as a precipitate such as Mn3O4, MnO2, Mn2O3, and MnO4
(Bruins et al. 2014).

Table 5. pH values and Eh values of each trial

Trial pH Eh/V
1 7.61 0.3015
2 7.27 0.3214
3 7.04 0.3367
4 6.99 0.3382

Fig. 4. Graph of Eh vs pH

Here, Eh value exists between 0.3 and 0.4. Type of the Manganese species to be
obtained varies with the pH and the Eh as shown in Fig. 5. From the above result of
Mn3O4, Hausmannite was obtained after adding bleaching powder solution between
0.7 ppm and 6 ppm to 0. 9 ppm Mn2+ solution.
 Studying About the Manganese Phase Diagram 145

Fig. 5. Phase diagram with analyzed data

After obtaining the particles, they may be removed by using a filtration process
(Isaeva and Castro 2011).
If water contains higher concentrations of Manganese as a dissolved form, before
removal it should be converted to its particular form by changing the pH or Eh values.
pH can be changed by adding diluted acid or base and Eh can be amended by changing
the concentration of the initial concentration of the Manganese solution and the
bleaching solution. Particulate form can be selected according to the Manganese phase
diagram considering the adequate pH and Eh values.

4 Conclusion

Mn3O4, Hausmannite occurs as a final product when 0.9 ppm synthetically prepared
Manganese solution reacts with different concentration of Calcium hypochlorite solu-
tions. Final product can be observed after some time as a brown colour precipitate. This
precipitate can be removed by using an adequate filtration method.

Acknowledgement. The authors would like to thank to staff of Water safety plan advisory unit
of National Water Supply and Drainage board in Katugastota for financial support and laboratory
support. Also give thanks to the Laboratory of the National Water Supply and Drainage board in
Kegalle for help of the analysis.
146 S. A. Abeysinghe et al.

References
Appelo CAJ, Postma D (2005) Redox processes. In: Geochemistry, groundwater and pollution,
2nd edn. A.A. Balkema Publishers, Leiden, pp 415–480. The Netherlands a member of
Taylor & Francis Group plc
Barloková D, Ilavský J (2009) Removal of iron and manganese from water using filtration by
natural materials. J Environ Stud 19:1118–1120
Bruins JH, Petrusevski B, Kruithof JC, Kennedy MD (2014) Manganese removal from
groundwater: characterization of filter media coating. Desalin Water Treat 1–13
Christensen TH, Bjerg P, Banwart SA, Jakobsen R, Heron G, Albrechtsen HJ (2000)
Characterization of redox conditions in groundwater contaminant plumes. J Contam Hydrol
45:165–241
Isaeva M, Castro NM (2011) Water treatment for the removal of iron and manganese. Bachelor
Degree Project in Mechanical Engineering – Development Assistance, 30 ECTS
James CN, Copeland RC, Lytle DA (2004) Relationship between oxidation reduction potential,
oxidant and pH in drinking water. In: American water works association WQTC conference
Kasim N, Mohammad AW (2017) Iron and manganese removal by nanofiltration and
ultrafiltration membranes: influence of pH adjustment. Malays J Anal Sci 21:149–158
Rozainy ZMR, Jamil R, Adlan MH (2015) A review of removal iron and manganese by using
cascade aeration systems. Sc Eng 74:69–76
Salem MG, El-Awady MH, Amin E (2012) Enhanced removal of dissolved iron and manganese
from nonconventional water resources in Delta District, Egypt. J Energy Proc 18:983–985
 Watershed Characterization of Demodara
Reservoir with Special Attention on Soluble
Manganese Concentration

W. G. Chathura Sanjeewa1(&), R. A. C. B. Ranaweera1,

Ruwan Liyanage1, W. B. M. M. W. Wijekoon1, and G. P. R. Ireshika2
1
National Water Supply and Drainage Board, Regional Support Center (Uva),
Badulla, Sri Lanka
[email protected]
2
Faculty of Science and Technology, Uva Wellassa University,
Badulla, Sri Lanka

Abstract. To ensure smooth operation and to manage water quality of the

Demodara Water Treatment Plant (DWTP), better knowledge of its watershed is
required. In line with this evolving response to water quality management; the
National Water Supply & Drainage Board is currently piloting the development
of a Water Safety Plan for the Demodara water supply scheme. For the purpose,
watershed assessment as a baseline study in a border context for the study area is
addressed in this paper. The DWTP is fed by the Demodara reservoir. Water
quality tests were conducted representing its catchment and none of its water
quality parameters except manganese (Mn) shows critical levels in the catch-
ment. The catchment was analysed to capture the Mn source. Results show that
the geological formation of the catchment consists of MnO2; especially in the
Charnockitic gneisses layers. With rainfall, Mn+4 converts to dissolved Mn+2. It
then comes to the Demodara reservoir from surface runoff and, it remains as
Mn+2 at the bottom levels of the reservoir due to anaerobic, reducing conditions.
Raw water is pumped to the DWTP from low levels of the reservoir, and
therefore Mn enters in to DWTP system. Manganese chemistry is very complex
and consists of multiple oxidative states. At the DWTP, Mn is converted to
insoluble manganese by oxidation using KMnO4 followed by the removal of
oxidized, insoluble manganese through filtration. However, manganese treat-
ment does not completely remove all manganese from the treated water. It will
allow some manganese to enter the distribution system, particularly in high rate
water treatment plants. Literature reports emphasize that Mn has no negative
health impacts at the levels commonly found in drinking water. However,
aesthetic problems caused by Mn are a continuing source of concern for cus-
tomers and DWTP operators. Therefore, treating Mn in raw water requires
continuous monitoring by plant operators. For the smooth functioning of the
DWTP, a number of recommendations are discussed in this paper.

Keywords: Manganese
Oxidization
Catchment
Distribution system

© Springer Nature Singapore Pte Ltd. 2020

R. Dissanayake and P. Mendis (Eds.): ICSBE 2018, LNCE 44, pp. 147–161, 2020.
https://doi.org/10.1007/978-981-13-9749-3_14
148 W. G. Chathura Sanjeewa et al.

1 Introduction

Badulla, Haliela and Ella integrated water supply scheme is mainly fed by the
Demodara Water Treatment Plant (DWTP) which is owned and operated by the
National Water Supply & Drainage Board (NWSDB), Sri Lanka. The DWTP was
designed and built by Tetra Tech, USA and commissioned towards the end of 2017. It
has a water treatment capacity of 15,000 m3 per day and is able to supply safe drinking
water to Badulla, Haliela and Ella demand centres and its suburbs. This integrated
water treatment system includes a new dam, new and upgraded water treatment plant,
storage tanks, pumping stations, and transmission and distribution pipelines. The water
system originates at the new Demodara reservoir, located on the Badulu Oya about
12 km upstream of the city of Badulla. It is a reservoir which has a capacity of
130,000 m3 to store raw water for the system.
To ensure smooth operation and to manage the quality of water of the DWTP,
better knowledge on its watershed is required. In line with this evolving response to
water quality management, the NWSDB is currently piloting the development of Water
Safety Plan for the Demodara water supply scheme. For the purpose watershed
assessment as a baseline study in a border context for the study area is addressed in this
paper.
Special attention was given to soluble manganese concentration since the DWTP
facility has experienced manganese (Mn+2) levels up to 1,000 µg per litre in certain
time periods. During the water treatment process and in the distribution system, the
manganese oxidized and caused brown staining of the plumbing fixtures leading to
public complaints. SLS 614 water quality standards has established a maximum con-
taminant level of total Mn as 100 µg per litre of treated water. However, the DWTP has
experienced Mn problems from effluent concentrations which is lower than its stipu-
lated standard limits.
Although Mn+2 oxidation has been practiced for years in water treatment, many
studies reported in literature have not fully explored oxidation of low initial Mn+2
concentrations and issues that may be important to water treatment plants. This study
also sought to address Mn related issues in the DWTP for its treatment.

2 Methodology

Location maps, catchment maps, elevation maps, land use patterns, population distri-
bution, river morphology maps and geological maps of the catchment were developed
using ArcGIS software. Rainfall data and river flow data were analysed using statistical
methods.
Spatial and Temporal distribution of water quality at the Demodara catchment
along the Badulu oya were analysed. Water samples were collected from 16 locations
along the Badulu oya covering the whole catchment area. Physical, chemical and
bacteriological parameters were checked for all samples and tests were carried out for
both dry and wet seasons. Then water quality parameter distributions were mapped to
get an average idea of the catchment. The Manganese variation of the catchment area is
presented in this paper.
 Watershed Characterization of Demodara Reservoir with Special Attention 149

Water samples were checked at the Demodara reservoir and an X-ray Powder
Diffraction (XRD) test was conducted for reservoir sediment samples. Phase identifi-
cation was done by using reference XRD graphs.
Since there are no mining sites, industrial discharge or leachates from landfills in
the Demodara catchment, the anticipated source for accumulated Mn in the Demodara
reservoir is eroded natural deposits from unplanned upper catchment. Therefore,
geological studies were done to identify Mn sources from the catchment. Fourteen rock
samples were analysed chemically representing all soil types of the Demodara catch-
ment area. Fertilizers used by farmers in the catchment were also studied.
To capture the correlation of raw water manganese level and upper catchment
rainfall, correlation coefficient analysis was conducted. The Mn level at raw water and
water at clear wells were analysed. The pH levels at raw water, tube settler surface,
adsorption clarifier surface, filter outlet, contact tank and clear well were taken and
analysed statistically.
The Demodara-Badulla main distribution pipe line (450 mm diameter HDPE) was
flushed. After continuous supply of 15 days, manganese at four wash-out locations
were checked in order to capture the ability of Mn accumulation in distribution lines.

2.1 Study Area

The Demodara dam catchment is entirely located in the Badulla district and is in the
South-West of the Namunakula mountain range in the hill country of Sri Lanka (be-
tween latitude 6° 51′ 00″ to 6° 55′ 30″ and longitude 80° 58′ 30″ to 81° 7′ 00″) (Fig. 1).
The Demodara reservoir is fed by the Badulu oya which is a major tributary of the
Mahaweli river, the longest river in Sri Lanka.

Fig. 1. Location of the catchment

150 W. G. Chathura Sanjeewa et al.

The Demodara dam location is just downstream of the confluence of the Nawela
oya and the Heel oya (Fig. 2). The Nawela Oya originates at an altitude of 1500 m
from Andagala and it joins the Heel Oya at Demodara. Heel Oya originates around
Bandarawela at an altitude of 1200 m.

Fig. 2. Demodara catchment

The total catchment area of the Demodara dam is 59.1 km2. Being located in an
upland area with narrow, steep and confined valleys, the catchment reacts quickly. The
terrain is generally mountainous with steep slopes where the elevation varies from
700 m to 1800 m. Soil cover in the basin is not deep and this will result in a quick
runoff response. In general, mountainous catchments have higher specific runoff than
lower regions due to high rainfall into the catchment. Furthermore, the storage capacity
of the surface zone and root zone is less in the mountainous areas.

Fig. 3. Elevation variation of the catchment

 Watershed Characterization of Demodara Reservoir with Special Attention 151

2.2 Drainage Density

Drainage density (DD) relates to the number of streams in a particular drainage basin
and can be measured by dividing total length of all streams in a basin (L) by its area
(A). As a rule, the higher the drainage density (DD) the more quickly water drains to a
river.
P
L
DD ¼ ð1Þ
A

2.3 The Correlation Coefficient (CC) Test

Correlation coefficient denoted by ‘r’ is a statistical analysis method representing the
degree of linear relationship between two variables (Taylor 1990). The correlation
coefficient takes on values ranging between +1 and −1.
Pn
i¼1 ððxi   xÞðyi  yÞÞ
r ¼ qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
Pn ð2Þ
2 Pn
i¼1 ðxi   xÞ i¼1 ðyi   yÞ 2

xi denotes the rainfall of the Bandarawela meteorological station at ith date. yi denotes
the manganese level of the incoming raw water to the DWTP at ith date. x and y denotes
average values of the corresponding group. The relationship between x and y is as
follows according to r value (Table 1).

Table 1. Correlation coefficient value interpretation

‘r’ value Interpretation
−1 Variables are completely anti-correlated
Between −0.7 to −1.0 Strong negative linear relationship
Between −0.3 to −0.7 Moderate negative linear relationship
Between 0 to −0.3 Weak negative linear relationship
0 Variables are completely uncorrelated
Between 0 to 0.3 Weak positive linear relationship
Between 0.3 to 0.7 Moderate positive linear relationship
Between 0.7 to 1.0 Strong positive linear relationship
1 Variables are absolutely correlated

3 Results

3.1 Land Use Pattern of the Catchment

Tea cultivation is the prominent land use pattern in the area, which is about 54% from
the total land (Table 2).
152 W. G. Chathura Sanjeewa et al.

Fig. 4. Land use patterns in the catchment

Table 2. Land use of the catchment

Land use type Percentage (%)
Tea 54
Build-up lands 21
Forests 9
Paddy 8
Rock 2
Other 8

3.2 Fertilizers Mainly Used for Cultivations

Most fertilizers used by farmers in the catchment consisted of calcium, nitrate, phos-
phate and sulphate. No fertilizers containing manganese were found during the study.

3.3 Drainage Density (DD) of the Catchment

In rainy season, the DD is nearly 1.5 and it means large channel frequency (tributaries).
Therefore, in the rainy seasons, higher sediment yields at the Demodara reservoir is
expected. However, all the eroded material does not enter the stream system as some of
the material is deposited at natural or manmade barriers within the watershed. Soil
erosion from the catchment also depends on rainfall intensity, soil type and geologic
formation, ground cover, land use, topography, runoff, sediment characteristics-grain
size, mineralogy, etc. and channel hydraulic characteristics which are interconnected.

3.4 General Geology of the Catchment

Rock types in the catchment area were Charnockitic gneisses, Garnet-sillimanite-biotite
gneiss, Garnetiferous quartzofeldspathic gneiss, Marble, Quartzites and Walawe
gneisses. Soluble Mn concentrations in each rock type are presented in Table 3.
 Watershed Characterization of Demodara Reservoir with Special Attention 153

Table 3. Rock types in the Demodara catchment and average soluble Mn in each rock type
Rock type Dominant minerals & chemical compositions Average soluble Mn
level (mg/l)
Charnockitic gneisses Hornblende [Ca2(Fe2,Mg)4(Al,Fe3)(Si7Al)O22(OH,F)2], 0.021
Graphite[C]
Garnet-sillimanite-bioite Sillimanite [Al2OSiO5], Biotite [KFe3,Mg3(AlSi3O10)(OH,F)2] 0.016
gneiss Graphite [C]
Garnetiferous Quartz [SiO2] 0
quartzofeldspathic gneiss
Marble Dolomite [CaMg(CO3)2], Calcite [CaCO3] 0.003
Quartzites Sillimanite [Al2OSiO5], Biotite [KFe3,Mg3(AlSi3O10)(OH, F)2] 0.007
Walawe gneisses Hornblende [Ca2(Fe2,Mg)4(Al,Fe3)(Si7Al)O22(OH,F)2], 0.011
Graphite [C], Quartz [SiO2],
Biotite [KFe3,Mg3(AlSi3O10)(OH,F)2]
Tourmaline [FeWO4],
Topaz [Al2SiO4(F,OH)2]

Results show that the major contributor to Mn in the Demodara catchment is

metamorphic Charnockitic gneisses rock type. Rock samples from Quartz and Marble
layers shows very low amounts of soluble Mn (Fig. 5).

Fig. 5. Geological map of the catchment

154 W. G. Chathura Sanjeewa et al.

3.5 Water Quality Parameters

Alkalinity, electrical conductivity, total hardness, phosphates, sulphates, nitrate, nitrite,
manganese, iron and other chemical parameters where checked and none of the water
quality parameters except that of manganese showed critical levels in the catchment.
Also, Heavy metals were not found in the Demodara raw water.

3.6 Correlation Coefficient (CC) Test Results

Results show a correlation coefficient value of 0.66 which indicates that there is a
moderate positive relationship with the DWTP raw water manganese levels and the
rainfall to the upper catchment. Therefore, manganese in raw water is critical in the raw
water of the DWTP.

3.7 Manganese Variation of the Catchment

According to the map, Mn concentration is comparatively high on the Bandarawela
side. This is due to the presence of a Charnokitic gneisses rock layer. Intermediately,
due to river flow, soluble Mn concentration is becoming low due to the self-aeration
capability of the stream. Near the Demodara Sri Shailatharama Viharaya, again, Mn
concentration is considerably high. Reviewing the characteristics of this area empha-
sized that soil erodibility near this area is high. The main reason for the determination is
that this area is surrounded by paddy cultivations (refer Fig. 4 for land use pattern).
During high rainfall events, sediment erosion is high in this area. Moreover, this area
has a Charnokitic gneisses stratum.
From the Demodara Sri Shailatharama Viharaya to the Demodara reservoir the
elevation variation is not high (refer Fig. 3 for the contour details). Therefore, self-
aeration from the Badulu oya is not expected and therefore Mn will transfer to the
Demodara reservoir (Fig. 6).

Fig. 6. Manganese variation in wet season

 Watershed Characterization of Demodara Reservoir with Special Attention 155

3.8 Presence of Manganese in Demodara Reservoir

Average pH, Mn and Fe levels of water samples collected from bottom of the reservoir
were presented in Table 4.

Table 4. Soluble Mn and Fe in reservoir sediment samples

Parameter Units Average concentration
pH At 25 °C ± 2 °C 7.10
Total Iron mg/l as Fe 3.3
Manganese mg/l as Mn 0.58

XRD result of the reservoir sediments is presented in Fig. 7.

Fig. 7. XRD analysis results for reservoir sediments

According to the XRD results the dominant minerals contained in sediments are
Gibbsite (Al(OH)3), Kaolinite (Al2Si2O5(OH)4) and Quartz (SiO2). Slight amount of
MnO2 was also present.

3.9 Scenario of Manganese Accumulation in Demodara Reservoir

and Issues Faced by DWTP Due to Manganese
According to these results, during the rainy season the DD is nearly 1.5 in the
Demodara catchment and it means large channel frequency (tributaries). Therefore,
during rainy seasons, higher sediment yields at Demodara reservoir are expected. In
addition to this, the pH in rainfall ranges from 4.46 to 6.89 in Sri Lanka (Silva and
Manuweera 2004). The most stable form of manganese in natural soil in Mn+4 and rain
156 W. G. Chathura Sanjeewa et al.

provided an acidic media which converted Mn+4 to Mn+2 which is a soluble form in
water. Therefore it flows with the surface runoff and accumulates, combined with the
high sediment yields in the Demodara reservoir.
Manganese (Mn) often occurs naturally in the environment and is widely dis-
tributed in the soil, water and air (Brandhuber 2013). Literature reports emphasized that
Mn has no negative health impacts at the levels commonly found in drinking water.
However, aesthetic problems caused by Mn are a continuing source of concern for
customers and DWTP operators. Problems associated with Mn includes water discol-
oration (usually black or dark red), clothes and fixture staining, turbid water, sediments
and at very high levels, metallic taste. If Mn remains soluble it will be undetectable to
the human eye. But if the customer washes clothes with bleach (NaClO), it will oxidize
Mn and precipitate directly on to the surface of the clothing, leaving small black dots
(Kohl and Medlar 2006). Manganese chemistry is very complex and consists of
multiple oxidative states. In fact, the most common process to treat manganese in
drinking water is to convert soluble manganese to insoluble manganese by oxidation
and to remove the insoluble manganese through filtration or sedimentation. However,
manganese treatment does not completely remove all manganese from the treated
water. It will allow some manganese to enter the distribution system, typically solid
MnO2 which can form colloidal bunches, do not settle and are not easily captured onto
a filter under moderate oxidation conditions (Brandhuber 2013). Increasing both the
ORP (oxidation reduction potential) and the pH is the most viable approach for con-
verting soluble manganese (Mn+2) to insoluble manganese (Mn+4) oxides for removal
(Tobiason et al. 2016).

Fig. 8. Manganese staining inside pipe fixtures

 Watershed Characterization of Demodara Reservoir with Special Attention 157

3.10 Methods to Minimize the Existence of Manganese in Water Supply

System
Methods to be adopted to minimize the existence of manganese in water supply system
are;
1. Selecting an alternative source or blending with other water sources
2. Management of manganese source
3. River catchment management
4. Sequestration
5. Water treatment process
The only feasible option at the DWTP is remove Mn by treatment since the source
of the Mn is in geological formation of the Demodara catchment and the Demodara
reservoir is the only source which is capable in suppling 15,000 m3/day to the area.
Sequestration is a temporary control measure to limit the aesthetic water quality
problems associated with oxidation of manganese (Kohl and Medlar 2006).
Removing manganese through water treatment process comprises two stages. The
first stage is to oxidize soluble forms of Mn to insoluble precipitates and the precipi-
tated material will be removed from the water through a solid separation process.
The DWTP uses Potassium permanganate (KMnO4) to oxidize manganese. The bal-
anced reaction between soluble Mn and Potassium permanganate is;
þ2 þ þ
2KMnO4ðaqÞ þ 2H2 OðlÞ þ 3MnðaqÞ ! 5MnO2ðsÞ þ 4HðaqÞ þ 2KðaqÞ

Potassium permanganate is more effective at oxidizing manganese than aeration or

chlorination (MWWOM 2009). Maintaining the pH from 7.2 to 7.4 during the water
treatment process has shown 85% removal of dissolved manganese from raw water.
Graph 1 shows pH variation throughout the water treatment process at DWTP under
similar ORP conditions and potassium permanganate dosage ratios (i.e.: amount of
permanganate dosing per mg/l of soluble Mn and Fe).

Graph 1. pH variation throughout the water treatment process at DWTP

158 W. G. Chathura Sanjeewa et al.

3.11 Controlling Manganese After Entering the Distribution System

It was noticed that even at Mn levels lower than 0.1 mg/l at Demodara clear well,
samples collected from the distribution pipe line contained a considerable amount of
Mn. Table 5 shows that the Mn level of samples and distance from the DWTP to
particularly wash out.

Table 5. Mn concentrations in distribution lines

Sample no. Distance from DWTP (m) Mn level (mg/l)
01 1550 (near Halpe Kovil) 0.08
02 1830 (near culvert 35/1) 0.21
03 2370 (near culvert 35/9) 0.16
04 3540 (near culvert 36/8) 0.02

To mitigate problems caused by Mn accumulation in the distribution systems,

periodic flushing of the distribution system is recommended according to a recognized
time period. However, minimizing the manganese concentrations in the lowest possible
limit before entering the distribution system is the most effective strategy.

4 Conclusions and Recommendations

The total area of the catchment is 59.1 km2 and terrain is generally mountainous with
steep slopes. The dominant land use pattern is tea plantations. During the rainy seasons,
due to high drainage density of the catchment, higher sediment yields at the Demodara
reservoir is expected.
Water quality analysis has shown physical and bacteriological contamination of the
reservoir to be considerably high due to poor catchment management practices as a
result of human activities in the catchment. But no considerably stressing chemical
contamination has occurred in the catchment except manganese which is a naturally
occurring and widely distributed chemical in the soil, water and air. Literature reports
emphasized that Mn has no negative health impacts at the levels commonly found in
drinking water. However, aesthetic problems caused by Mn are a continuing source of
concern for customers and for water utility.
The most common process to treat manganese in drinking water is to convert
soluble manganese to insoluble manganese by oxidation and to remove insoluble
manganese through filtration or sedimentation. Increasing both the ORP and pH is the
most viable approach for this conversion.
The DWTP receives Mn from its raw water by its catchment’s geological formation
and land use pattern. Soluble Mn is rich in the Charnockitic gneisses, Garnet-
sillimanite-biotite gneiss and Walawe gneisses geology layers in the catchment. Also,
as a major Mn source, paddy fields near Sri Shailatharama Viharaya, Demodara have
been highlighted both in the dry and wet seasons. Therefore, soil erosion at the
 Watershed Characterization of Demodara Reservoir with Special Attention 159

particular location has to be minimized through appropriate tree plantations, soil

retaining structures etc.
The correlation coefficient test result has shown that Mn level in the DWTP raw
water, has a moderately positive relationship with rainfall amounts in the upper
catchment. This happens due to insoluble Mn becoming soluble Mn with rainfall (pH
of the rainfall <7). It then comes to the Demodara reservoir from the surface runoff and,
it remains as Mn+2 at the bottom levels of the reservoir due to anaerobic conditions.
Raw water is pumped to the DWTP from the lower level of the reservoir, and therefore
Mn will enter in to the DWTP system.
However, manganese treatment does not completely remove all manganese from
the treated water. It will allow some manganese to enter the distribution system, par-
ticularly in high rate water treatment plants. According to the Federal-Provincial-
Territorial Committee on Drinking Water (2016); even manganese concentrations low
as 0.02 mg/l can cause complaints about discoloured water, plumbing fixture and
laundry staining, and general dissatisfaction with the water quality due to continued
accumulation of Mn in the distribution system from low but constant manganese
present in the source or treated water.
Therefore, treating Mn in raw water is very complex and required continues
monitoring by plant operators. For smooth functioning of the DWTP, following rec-
ommendations could be given for different capacities from this study.

4.1 Recommendations for Catchment Management

– Soil stabilization has to be done in paddy fields near the Sri Shailatharama
Viharaya, Demodara since both turbidity, colour and manganese levels of water
flowing through the Badulu oya highlighted that contamination from this location is
high in both the dry and wet seasons.
– From the Sri Shailatharama Viharaya, Demodara to the Demodara reservoir, the
self-aeration capability of the Badulu oya is less due to less elevation variation in
the river. Providing additional aeration facilities or pH adjustments in the river will
be able to provide added advantages for the DWTP operation. But it requires more
studies related to this and a holistic approach in conduct.

4.2 Recommendations for Plant Operators

– Continued monitoring of raw water quality is required to ensure proper dosing of

KMnO4. The dosage needs to be carefully controlled so that the KMnO4 residues
shall not remain in the water as Mn. KMnO4 feeding rate should be equal to
oxidization requirement of both Mn and Fe. Theoretical KMnO4 dosing is 1.92 mg
for 1.0 mg of Mn and 0.94 mg for 1.0 mg of Fe.
– Dosing KMnO4 and NaOH to raw water is best at the initial stage to ensure
sufficient contact time with KMnO4 and will be benefited by an improved plant
efficiency.
160 W. G. Chathura Sanjeewa et al.

– Maintaining pH levels tube settler, adsorption clarifier, filter and contact tank at
level 7.2 to 7.4 will ensure more than 85% of Mn removed from water with proper
the permanganate dosage.
– Periodic flushing of the distribution lines has to be conducted according to the Mn
levels at water even at the allowable levels.
– Mn intake to the WTP has been increased due to the fact the raw water line is fixed
at a lower level of the reservoir. If water can be extracted from an upper level of the
reservoir, especially during the wet season, the possibility of increased Mn con-
centration in raw water could be reduced.
– Air bubbling at a pre-sedimentation tank will be able to increase dissolved oxygen
in raw water and therefore will increase Mn removal efficiency at the filter.
– To reduce Mn accumulation of the reservoir, frequent flushing out of the sediments
trapped in the reservoir has to be done, especially during rainy seasons.

4.3 Recommendations for Packaged Water Treatment Plant Designers

and Manufacturers

– Since continued monitoring of manganese level and iron levels should be carried
out in order to ensure smooth operation of the plant including inline Mn and Fe
measuring meters and accordingly continued adjustment of the KMnO4 feeding
system will be an additional value added design.
– It is recommended to increase the hydraulic retention time of packaged treatment
units up to 15 min as per recommendations by researchers. It will provide additional
contact time with KMnO4.
– If the system can inject KMnO4 and NaOH prior to alum or PAC it will then
provide an additional benefit in pH support for Mn removal; since alum or PAC will
reduce pH which is a disadvantage in KMnO4 removal. However, PAC decreases
pH less than alum.
– Addition of oxide coated media in filters will also effectively remove Mn and Fe.

4.4 Recommendations for Further Studies

– According to SLS 614: 2013, which is the standard for drinking water in Sri Lanka,
stipulated Mn level of drinking water is 0.1 mg/l. This value is fairly high compared
to aesthetic requirements of drinking water. In addition, Mn can be accumulated in
distribution lines and therefore 0.1 mg/l level has to be modified according to the
high rate advanced water treatment plant requirements.
– Buffer zone establishment, particularly for drinking water reservoirs, has to be done
at a national level.
 Watershed Characterization of Demodara Reservoir with Special Attention 161

References
Taylor R (1990) Interpretation of the correlation coefficient: a basic review. J Diagn. Med.
Sonogr 6(1):35–39
Silva EIL, Manuweera L (2004) Surface and rainwater chemistry in Sri Lanka-a risk of
acidification. Asian J Water Environ Pollut 1(1 & 2):79–86
Kohl PM, Medlar SJ (2006) Occurrence of manganese in drinking water and manganese control.
American Water Works Association, Washington, D.C., pp 1–9
Brandhuber P (2013) Legacy of manganese accumulation in water systems. Water Research
Foundation
Tobiason JE, Bazilio A, Goodwill J, Mai X, Nguyen C (2016) Manganese removal from drinking
water sources. Curr Pollut Rep 2(3):168–177
Minnesota water works operations manual (2009) Fourth Publication, Chapter 14, pp 1–6
Federal-Provincial-Territorial Committee on Drinking Water (2016) Manganese in drinking
water. Health Canada
 Application of Safety Plans for Rainwater
Harvesting Systems in Sri Lanka

K. A. I. L. Rathnayake1, S. A. C. Upeka2(&), K. A. M. Kularathne2,

P. G. K. B. Ranathunga2, T. D. Jayasinghe3,
W. B. M. L. I. Weerasekara3, and S. K. Weragoda3
1
Faculty of Science and Technology, Uva Wellassa University,
Badulla, Sri Lanka
2
China Sri Lanka Research Grant Project,
Ministry of City Planning and Water Supply, Katugastota, Sri Lanka
[email protected]
3
Water Safety Plan Advisory Unit, National Water Supply and Drainage Board,
Katugastota, Sri Lanka

Abstract. Water is often considered the most basic need for sustaining life. As
the world population increases, the demand for quality drinking water increases.
Dug wells and tube wells are considered as the main water source among the
rural community in the dry zone. Rainwater harvesting is receiving increased
attention in the dry zone as an alternative source of drinking water. While
rooftop rainwater harvesting is being promoted, little attention has been paid to
the quality of collected rainwater until recently. It becomes nationally important
to look for low cost options which could be managed and afforded by the
communities themselves. Rain water systems are categorized as individual
systems and there are no proper regulations for maintaining the rain water
harvesting systems. Therefore, it is vital to develop a proper guideline for rain
water harvesting systems. Kebethigollewa was selected as the sample location
due to the high number of rain water harvesting systems available. A ques-
tionnaire survey and focus group discussions with the beneficiaries were con-
ducted to collect data on different rain water harvesting systems, rain water
usage pattern and health concerns due to consumption of rain water. Based on
the outcome of the data from observations, questionnaire survey and sample
analysis, cost effective and community manageable rain water harvesting safety
plan were introduced. Rain water was mainly used for the drinking and cooking
purposes and dug well water was used for other activities like bathing and
washing. Over hanging vegetation was identified above the roof catchment
which may attract animals. Fallen foliage which blocked the gutters were
identified in most of the guttering channels. According to the analysis, samples
shows slight acidity. Turbidity ranged from 0.22 to 2.80 and TDS varied from 5
to 120 mg/l. All the water samples have TDS, Nitrate, Sulphate and hardness
concentration below the prescribed limit.

Keywords: Guideline
Focus group discussions
Rainwater harvesting

Rural community

© Springer Nature Singapore Pte Ltd. 2020

R. Dissanayake and P. Mendis (Eds.): ICSBE 2018, LNCE 44, pp. 162–170, 2020.
https://doi.org/10.1007/978-981-13-9749-3_15
 Application of Safety Plans for Rainwater Harvesting Systems 163

1 Introduction

Water is often considered the most basic need for sustaining life. As the world pop-
ulation increases, the demand increases for quality drinking water. A large population
of the world do not have access to safe drinking water and is suffering from water
scarcity problems. Water shortage issues are a growing concern in Sri Lanka as well,
especially in the dry zone areas.
Dug wells and tube wells are the main water source among the rural community of
the dry zone areas. With the existing questionable hypothesis (excessive hardness,
fluoride and heavy metal contamination in groundwater) developed in the dry zone,
which has susceptible factors leading towards the major health issue of Chronic Kidney
Disease (CKD), the community is looking for alternative sources for drinking water.
Rainwater harvesting is receiving increased attention in the dry zone as an alternative
source of drinking water. Rainwater from individual household roof catchments is
becoming a common practice by the rural community in the dry zone today. While
rooftop rainwater harvesting is being promoted, little attention has been paid to the
quality of collected rainwater until recently. It becomes nationally important to look for
low cost options which could be managed and afforded by the communities themselves.

1.1 Problem Statement

Rooftop rain water harvesting systems can be contaminated at collection, conveyance,
and storage system. As rain water is obtained as a source of drinking water, it is
important to analyze the quality of the harvested rain water. Rain water systems are
categorized as individual systems and there are no proper regulations for maintaining
the rain water harvesting systems. Therefore, it is vital to develop a proper guideline for
rain water harvesting systems.

1.2 Objectives

• To investigate physio-chemical and bacteriological quality of the harvested rain

water.
• To study the water quality deterioration causes and pathways with the time period.
• To assess the health impact and social acceptability of rain water harvesting
systems.
• To introduce water safety plan for rain water harvesting systems.

2 Methodology

Kebethigollewa was selected as the sample location due to high number of rain water
harvesting systems available. A questionnaire survey and focus group discussions with
the beneficiaries were conducted to collect data on different rain water harvesting
systems, rain water usage patterns and health concerns due to consumption of rain
water. Water samples were collected from each rain water harvesting unit and sample
164 K. A. I. L. Rathnayake et al.

analysis was conducted for physicochemical parameters like turbidity, pH, electrical
conductivity, hardness, total dissolved solids, nitrate, phosphate, sulfate, alkalinity and
bacteriological parameters including total coliform and e-coli.
The water quality data was compared with the standards stipulated by SLS 614:
2013 for drinking water. Based on the outcomes of the data from observations,
questionnaire survey and sample analysis cost effective and community manageable
rain water harvesting safety plan was introduced.

3 Results and Discussion

The observed weather condition was dry during the inspection period and temperature
ranged between 25 °C–30 °C. Rain water was mainly used for drinking and cooking
purposes and dug well water was used for other activities like bathing and washing.
Over hanging vegetation was identified above the roof catchment which may attract
animals. Fallen foliage which blocked the gutters were identified in most of the gut-
tering channels. Inadequately sealed storage tanks of the rain water harvesting units
were also observed.

3.1 Water Quality Analysis

3.1.1 pH Value
The pH is a measure of the intensity of acidity or alkalinity and measures the con-
centration of hydrogen ions in water. It has no direct adverse effect on health. However,
a low value, below 4.0 will produce sour taste and higher value above 8.5 shows
alkaline taste. A pH range of 6.5–8.5 is normally acceptable as per guidelines suggested
by Sri Lanka Standards for potable water (SLS 614: 2013). The pH values of the
samples from Kebithigollewa area are in between 4.04–7.19. According to the study
samples collected, the rain water of Kebithigollewa shows slight acidity (Fig. 1).

pH
14

12

10
Concentration

0
0 5 10 15 20 25 30 35
Sample number

Fig. 1. pH variation
 Application of Safety Plans for Rainwater Harvesting Systems 165

3.1.2 Turbidity
Measurement of Turbidity reflects the transparency in water. It is caused by the sub-
stances present in water in suspension. In natural water, it is caused by clay, silt,
organic matter and other microscopic organisms. It ranged from 0.22 to 2.80 in the
Kebithigollewa area. However, the prescribed limit of Turbidity for drinking water is 2
NTU (SLS 614: 2013). There are 4 water samples with higher turbidity values in the
Kebithigollewa area (Fig. 2).

Turbidity
3.0

2.5

2.0
Turbidity [in NTU]

1.5

1.0

0.5

0.0
0 5 10 15 20 25 30 35
Sample number

Fig. 2. Turbidity variation

3.1.3 EC
EC refers to the electrical conductivity of the water solution. When the water is pure
without any dissolved solutes, it conducts electricity naturally. The amount of elec-
tricity that can pass through water is influenced by amount of dissolved ions in the
water. According to Sri Lanka Standards value of electrical conductivity of drinking
water should be lower than 750 mS/cm (Fig. 3).

EC
300

250
Conductivity [mS/cm]

200

150

100

50

0
0 5 10 15 20 25 30 35
Sample number

Fig. 3. EC variation
166 K. A. I. L. Rathnayake et al.

3.1.4 Total Dissolved Solids

Total Dissolved Solids may be considered as salinity indicator for water. The TDS in
water is due to the presence of calcium, magnesium, sodium, potassium, bicarbonate,
chloride and sulphate ions. In the study area TDS varied from 5 to 120 mg/l. As the
prescribed limit of TDS for drinking water is 500 mg/l, all the water samples have TDS
concentration below the prescribed limit (Fig. 4).

TDS

140

120
Total Dissolved solids(mg/L)

100

80

60

40

20

0
0 5 10 15 20 25 30 35
Sample number

Fig. 4. TDS variation

3.1.5 Hardness
Hardness of water is objectionable from the view point of water use for laundry and
domestic purposes since it consumes a large quantity of soap. Based on the present
investigation, hardness is low. However, the permissible limit of Hardness for drinking
water is 250 mg/l (SLS 614: 2013) (Fig. 5).

Hardness

140

120
Total Hardness [in ppm]

100

80

60

40

20
0 2 4 6 8 10 12 14
Sample number

Fig. 5. Hardness variation

 Application of Safety Plans for Rainwater Harvesting Systems 167

3.1.6 Nitrates
Nitrates are an essential source of nitrogen for plants. When nitrogen fertilizers are used
to enrich soils, nitrates may be carried by rains. Nitrate levels in drinking water can be
an indicator of overall water quality. According to Sri Lanka Standards for drinking
water, Nitrates concentration should be less than 50 ppm (Fig. 6).

Nitrate

5
concentration [in ppm]

0 5 10 15 20 25 30 35
Sample number

Fig. 6. Nitrate variation

3.1.7 Sulphate
Sulphate can be found in almost all natural water. The origin of most sulphate com-
pounds is the oxidation of sulphate ores, the presence of shales, or industrial wastes.
Sulphates is one of the major dissolved components of rain. According to Sri Lanka
Standards for drinking water, Sulphate concentration should be less than 250 ppm
(Fig. 7).

Sulphate
30

25

20
Concentration [in ppm]

15

10

-5

-10
0 5 10 15 20 25 30 35
Sample number

Fig. 7. Sulphate variation

168 K. A. I. L. Rathnayake et al.

3.2 Bacteriological Analysis

If coliform bacteria present in the water, it is not suitable for drinking purposes. (SLS
614: 2013). These samples are from Kebithigollewa from 5 through 11 are not suitable
as potable water.

Location of the Coliform bacteria present in 100 ml E coli in 100 ml of sample

sample table sample at 37 °C at 44 °C
Kebithigollewa 160 00
Rambewa 102 86
Rambewa 218 00
Rambewa 00 00
Kebithigollewa 00 00
Kebithigollewa 00 00
Kebithigollewa 00 00
Kebithigollewa 00 00
Rambewa 06 00
Rambewa 14 04
Rambewa 00 00

4 Recommendation

The results of the data indicates that all the water quality parameters are in acceptable
level for drinking purposes but the considerations on poor maintenance practices of the
rain water harvesting systems should be addressed.
The roof catchment and the guttering channels should be cleaned systematically
because contamination of the roof and guttering channels may be washed into the
storage tanks during the rainfall events and constitute a risk to water quality (When rain
falls after a long dry period, or, any rainwater collected may carry with it significant
amounts of contamination and debris which have accumulated on the roof and in the
gutters. It is therefore recommended that the water running off the roof after the first
storms of the season, and preferably for the first 5–10 min afterwards or until it runs
clear, should be discarded or used for purposes other than drinking).
The overhanging vegetation near the rain water harvesting systems should be
removed as the overhanging vegetation could attract animals that may contaminate the
roof catchment with faecal material.
The gutters should be kept clean as the fallen foliage could also block the gutters
and filters.
The filter box should be cleaned at least once in 6 months of time as clogged filters
may result in contaminating the storage tank due to overflow.
The storage tank should be sealed/covered adequately as lizards and geckos may
enter and produce elevated thermotolerant (faecal) coliform counts. (A fine mesh fitted
to all openings to the tank will prevent the entry of organic debris).
 Application of Safety Plans for Rainwater Harvesting Systems 169

Stagnant water of the water collection area should be cleaned as it may encourage
mosquito breeding and increase the likelihood of contamination entering the storage
tank.
An area of 10 m from the tank should be cleaned as the presence of waste material
(household, agricultural, animal faeces) may constitute a risk to water quality.
• Activities that may result in airborne contaminants such as spray drifts from agri-
cultural practices should be carried out at a safe distance from the water collection
area because it may contaminate the roof catchment.
• The storage tank should be drained, cleaned and disinfected at least annually.
Stored rainwater should be safely stored and handled, and ideally disinfected prior
to consumption (e.g. Chlorination, boiling or any other suitable household water
treatment option).
The quality of the collected rainwater can also be improved by proper maintenance
of the roof and gutters, and careful cleaning at the beginning of every wet season.
It is recommended to conduct an analysis determine the components of asbestos in
water.
Use a food grade paint for the roofing.
Guideline to develop WSP for the Rain Water Harvesting System
• Modification of specified rain water collection removable guttering system to
control contamination arising due to long dry seasons.
• Modification of specified rain water collection tanks with conical shape bottom for
better flushing out performances.
• Educate people to construct at least one part of tile roof Need to facilitate with
subsidy from the relevant authority, Establishment of new rules and regulations.
• The existing Asbestos roofs need to protected with food grade paint to minimise
possible fibre contaminations.
• Introduce household treatment for drinking rain water: Boiling, Bacteriological
filtration, Domestic Chlorination, Blending with existing ground water (10, 20%)
for minimising possible ion deficiencies

5 Conclusion

The most significant impact of the rainwater harvesting system in Sri Lanka is the
ensured supply of water in the homestead. Rain water was mainly used for drinking and
cooking purposes and dug well water was used for other activities like bathing and
washing. Over hanging vegetation was identified above the roof catchment which may
attract animals. Fallen foliage which blocked the gutters was identified in most of the
guttering channels. According to the analysis, samples shows slight acidity. Turbidity
ranged from 0.22 to 2.80 and TDS varied from 5 to 120 mg/l. All the water samples
have TDS, Nitrate, Sulphate and hardness concentration below the prescribed limit.
170 K. A. I. L. Rathnayake et al.

Acknowledgments. We express our sincere and honorable gratitude towards all the staff of
water safety plan advisory unit - Katugasthota for providing the necessary support and guidance
during the period of the research. Our sincere thanks go to the National Water Supply and
Drainage Board (Anuradhapura and Kegalle) for being a supportive hand to conduct the study
successfully. We would like to thank all the personalities including professionals and all other
supportive hands who directed us to succeed in this study.

References
Abu-Zeid MA (1998) Water and sustainable development: the vision for world water, life and the
environment1. Water Policy 1(1):9–19
Ariyananda T, Mawatha E (July 1999) Comparative review of drinking water quality from
different rainwater harvesting systems in Sri Lanka. In: Proceedings of 9th IRCSA conference
on Petrolina, Brazil
Coombes PJ, Kuczera G, Kalma JD, Dunstan HR (2000) Rainwater quality from roofs, tanks and
hot water systems at Figtree Place. In: Proceedings of hydro 2000: interactive hydrology,
p 1042
Despins C, Farahbakhsh K, Leidl C (2009) Assessment of rainwater quality from rainwater
harvesting systems in Ontario, Canada. J Water Supply: Res Technol-Aqua 58(2):117–134
Hapugoda D (1995) Action research study on rainwater harvesting. Community water supply and
sanitation project (CWSSP), Colombo, Sri Lanka
Morey A, Dhurve B, Haste V, Wasnik B (2016) Rainwater harvesting system. Int Res J Eng
Technol 3(4):2158–2162
Yaziz MI, Gunting H, Sapari N, Ghazali AW (1989) Variations in rainwater quality from roof
catchments. Water Res 23(6):761–765
 Hardness and Fluoride Removal Efficiency
of Groundwater: Use of Physiochemical
Methods in Water Purification
at Anuradhapura District Dry Zone
in Sri Lanka

K. M. S. Kumara1(&), K. A. M. Kularathne2,
P. G. K. B. Ranathunga2, S. A. C. Upeka2, D. C. Herath2,
and S. K. Weragoda2
1
Department of Civil Engineering,
General Sir John Kotelawala Defense University, Colombo, Sri Lanka
[email protected]
2
Water Safety Plan Advisory Unit, National Water Supply and Drainage Board,
Katugastota, Sri Lanka

Abstract. The limitation of drinking water critically affects the day-to-day life
of most of the community in Sri Lanka as well as in the world. The community
in the dry zone of Sri Lanka has especially been facing this difficulty for a long
period. It is identified that high concentration of Calcium, Magnesium and
Fluoride in groundwater is the main reason for this situation. Even though there
are various methods used for water purification, the problem remains the same
for rural communities and some urban communities. This study is focused on
installing the most suitable and effective water treatment pilot plant and the
collection of data to design an industrial water treatment plant. Anuradhapura
and Mihinthale were selected as the study area based on the data collected
during field visits. A pilot plant with a Reverse Osmosis (RO) membrane was
installed and operated for two months of a time period, at the army camp
Padulagama while comparing it with an industrial treatment plant at Rajarata
University. According to the physical and chemical parameter results it was
revealed that 32.43% of the raw water was successfully purified. Hardness and
Fluoride concentration could be reduced up to 42.07 mg/l and 0.06 mg/l
respectively. The common parameters which were reduced due to the use of RO
membrane were also analyzed and found acceptable. Operation lapses and
difficulties in treatment plants were identified and rectifications were carried out
successfully. Designing of a RO water treatment plant with a pre-treatment unit
and maintenance, regeneration and replacements were conducted. Recommen-
dations were made for appropriate discharging methods of the rejected water
from the RO membrane.

Keywords: Hardness removal
Fluoride removal
Dry zone

Water purification
RO membrane

© Springer Nature Singapore Pte Ltd. 2020

R. Dissanayake and P. Mendis (Eds.): ICSBE 2018, LNCE 44, pp. 171–182, 2020.
https://doi.org/10.1007/978-981-13-9749-3_16
172 K. M. S. Kumara et al.

1 Introduction

Generally, 70% of the earth’s surface area is covered with water, but the availability of
fresh water is less than 3% of the total global water. Most of it is in glaciers, ice caps,
the atmosphere and deep groundwater reservoirs. Only 13% of the total fresh water is
readily available to use for the human consumption. At the same time, water demand of
the world becomes double every 21 years due to industrial, agricultural and domestic
pollution of pure water. Water is becoming an increasingly precious resource which
should be managed judiciously. Therefore, drinking water purification has become an
essential requirement in the present world. Few decades ago in Sri Lanka, water at
acceptable quality was readily available in many parts of the island. Most of the pure
water sources have now been polluted and new solutions for accessibility of water in
drinking condition are essential. However, in certain parts of the island such as the
Northern and North central province (Dry Zone), drinking water under imposed
standards are not available with respect to hardness and the presence of fluoride in the
water.
The requirement of suitable drinking water for the community in the dry zone is
becoming a critical issue and it has already affected the survival of the community. By
now, it is a National problem of the country. The rural community of the dry zone are
highly affected due to the use of contaminated drinking water. As a solution to this,
there are various kind of drinking water treatment plants installed in several areas and
purified water is distributed among the community in the rural areas in the north east.
The final outcome of these plants is not guaranteed. Since it has become a fund-raising
event, the situation has become worse. None of the owners of these drinking water
treatment plants in the private sector and or government institutes has taken any pre-
caution for the discharge of those plants. The discharges might be affecting the envi-
ronment indirectly.
Therefore, correct identification of the drinking water treatment process very
important. It is high time to find possible solution for groundwater water purification
according to the ground water condition. This study is focused on the design a proper
drinking water purification system considering the raw water quality and the required
quantity while optimizing the available drinking water treatment systems.

2 Methodology

2.1 Overview
Primary data was obtained by testing the raw water samples collected in the initial field
visit at Anuradhapura. The lab scale water treatment plant including the sand filter unit,
granular activated carbon filter unit, ion exchange resin unit, RO membrane unit and
UV chamber was installed at the army camp at Pandulagama - Anuradhapura (GR-
8.331165, 80.362208) as shown in Fig. 1. For the operation of the lab scale water
treatment plant, raw ground water dug from well source in Pandulagama Army Camp
was used. Lab Scale reverse osmosis (RO) plant and industrial RO treatment plant were
used for the study as shown in Figs. 2 and 3. To identify the performance, practical
 Hardness and Fluoride Removal Efficiency of Groundwater 173

operating capability and maintenance of the lab scale RO treatment plant was compared
with the Industrial RO treatment plant which was installed at the University of Rajarata
in Mihinthale.

Fig. 1. Pilot plant site at Pandulagama Army camp (GR- 8.331165, 80.362208)

Fig. 2. Lab scale water treatment plant

Fig. 3. Lab scale water treatment plant

174 K. M. S. Kumara et al.

Sampling of Analysis Water Quality Parameters

Water samples were collected into clean 250-ml bottles for a period of two months
from raw water, sand filter, granular activated carbon filter unit, ion exchange resin
unit, RO membrane unit, UV chamber and RO rejected of both lab and industrial scale
RO water treatment plants. Physical water quality parameters such as pH, conductivity,
and turbidity were measured onsite and laboratory experiments were done to measure
the chemical water quality parameters such as Total Dissolved Solids (TDS), alkalinity,
total hardness, fluoride and Total Organic Carbon (TOC).

2.2 Sand Filter for Turbidity Removal

Results and Analysis
The Lab Scale Water Treatment Plant (RO Plant)
Influent flow rate of the lab scale water treatment plant is 37 l/h
Permeable flow rate of the RO plant is 12 l/h
Purification efficiency = 12  100/37 = 32.43%
The Industrial Water Treatment Plant
Influent Flow rate of the Industrial water treatment plant is 566 l/h
Permeable flow rate of the plant is 148 l/h
Efficiency of the Industrial water treatment plant = 26%
Turbidity values of the first 3 days at the lab scale treatment plant are not of the
expected manner. Turbidity of the treated water is higher than the raw water turbidity.
Increasing of back washing period and having proper supervision, efficiency of sand
filter can be increased. However, the raw turbidity and treated water turbidity are at the
permissible range (<1 NTU) (Fig. 4).

Lab Scale Plant

Industrial Plant
100
Turbidity Removal Efficiency (%)

50

-50

-100

-150

-200

0 2 4 6 8 10 12 14 16 18
Duration (Days)

Fig. 4. Comparison of turbidity removal efficiency with the time using the sand filters in lab
scale plant and industrial plant
 Hardness and Fluoride Removal Efficiency of Groundwater 175

2.3 Granular Activated Carbon (GAC) Filter for TOC Removal

Using the GAC filter, the main expectation is to reduce the TOC of the water. In this
study total organic carbon of raw water and treated water was tested for both plants as
follows (Fig. 5).

Fig. 5. Comparison of TOC removal in lab scale water treatment plant, industrial water
treatment plant

Raw water TOC concentration is higher at the lab Scale plant. TOC reduction
capacity is also high at the Lab Scale Plant. Where that is in the permissible range
(<2 mg/l).

2.4 Ion Exchange Resin for Hardness Removal

According to the data, the reduction of hardness using ion exchange resin is consid-
erably low. Improvement of that reduction is dependent on the maintenance of the ion
exchange unit. Therefore, the regeneration process should be observed properly. There
may be a practical problem of analyzing the data at the time the sample is taken, just
after the regeneration process so that adjustment can be done for the quantity of resin,
time of regeneration and frequency (Figs. 6 and 7).
176 K. M. S. Kumara et al.

Raw Water
1000 Treated by Resin

900

Total Hardness (mg/l)

800

700

600

500

400
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
Duration (days)

Fig. 6. Hardness of raw water and treated by ion exchange resin in lab scale water treatment plant

900 Raw water

Treated by Resin
800

700
Hardness (mg/l)

600

500

400

300

1 2 3 4 5 6
Duration (Days)

Fig. 7. Hardness of raw water and treated by ion exchange resin in industrial water treatment
plant

2.5 RO Membrane
Even though this study mainly focuses on reduction of hardness and fluoride, with the
use of an RO membrane, almost all other parameters are reduced considerably as well.
Generally, hardness removal efficiency of lab scale plants and industrial plants are
same. But hardness concentration of a lab scale plant is significantly higher than the
industrial plant. According to the data, it proves that hardness removal from both
treatment plants depends on the raw water condition. Anyway, the removal of hardness
using RO membrane is successful. Hardness of treated water is well below the per-
missible range (Fig. 8).
 Hardness and Fluoride Removal Efficiency of Groundwater 177

Fig. 8. Comparison of hardness removal capacity with lab scale water treatment plant industrial
water treatment plant

2.6 Fluoride Removal

It is observed that the fluoride concentration of raw water at the industrial water
treatment plant is already below the permissible range and the lab scale plant is mar-
ginal. Considering the fluoride removal efficiency of the treatment plants, the lab Scale
plant shows higher efficiency than others. Fluoride concentration of both plants are
successfully reduced and it is below of the permissible range (<1 mg/l) (Fig. 9).

Fig. 9. Fluoride removal capacity in lab scale water treatment plant, industrial water treatment
plant
178 K. M. S. Kumara et al.

2.7 Conductivity
According to the conductivity graph shown in Fig. 10, the maximum and minimum
efficiencies of the lab scale plant for electrical conductivity are 96.6% and 84.5%
respectively whereas the maximum and minimum electrical conductivity efficiency of
industrial water treatment plants are 96.08% and 93.89% respectively. As per the
figure, the efficiencies of conductivity reductions of both the plants varies between
96.2% to 83.9%. This shows the functions of the RO membranes of both plants which
are gradually reducing. This capacity may be reduced due to the clogging of mem-
branes. Furthermore, it shows that the efficiencies are varying within a short range. That
may depend on the daily deposits of substance on the membrane.

98

Lab Scale RO unit

Conductivity Reduction Efficiency(%)

96
Industrial RO unit
94

92

90

88

86

84

82
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 --

Duration (Days)

Fig. 10. Electric conductivity reduction efficiency of RO units in lab scale water treatment
plants and industrial water treatment plants

2.8 pH Value
Generally, the average pH value of treated water in both the plants are within 6.5 to 8.5
(Figs. 11 and 12).

Lab Scale Raw Water

Lab Scale Treated Water
12

11

10
pH Value

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Duration (Days)

Fig. 11. pH variation of between raw water and purified water in lab scale water treatment plant
 Hardness and Fluoride Removal Efficiency of Groundwater 179

10.0
Industrial Raw Water
9.5 Industrial Treated Water

9.0

8.5

pH value
8.0

7.5

7.0

6.5

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Duration (Days)

Fig. 12. pH variation of between raw water and purified water in industrial water treatment plant

2.9 Total Dissolve Solids

It is observed that the TDS concentration of raw water at the lab scale treatment plant is
fairly higher than the TDS concentration of raw water at the industrial water treatment
plant. The TDS concentration of raw water on the 10th day at the industrial treatment plant
was unexpectedly higher than others. It may be a sample error or a testing error (Fig. 13).

1000
TDS of Lab scale Raw Water
TDS of Lab Scale RO Treated Water
TDS of Industrial Raw Water
TDS of Industrial RO Treated Water
800
Total Dissolved solids (ppm)

600

400

200

0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
Duration (Days)

Fig. 13. Total dissolve solids of lab scale plant and industrial plant

2.10 Alkalinity
Alkalinity of the raw water at the Lab Scale plant is considerably higher than the raw
water of the Industrial plant. But the alkalinity of treated water at the lab scale plant and
the industrial plant has been reduced up to an average value of 6.38 mg/l and 7.43 mg/l
respectively. Reduced values of alkalinity are below the permissible range, which is
less than 200 mg/l. Alkalinity of raw water at the industrial plant is also less than the
permissible value (Fig. 14).
180 K. M. S. Kumara et al.

Lab Scale Raw Water

Lab Scale RO Treated Water
Industrial Raw Water
Industrial RO Treated Water

200

150

Alkalinity (mg/l)
100

50

1 2 3 4 5 6 7 8 9 10 11 12
Duration (Days)

Fig. 14. Variation of alkalinity in raw water and treated water of lab scale water treatment plant
and industrial water treatment plant

2.11 Ultra Violet Disinfection (UV)

An ultra violet lamp is used as a disinfection process at the lab scale plant and the
industrial plant. Using the UV lamp, bacteria content is reduced up to zero as shown in
Table 1.

Table 1. Removal of bacteria using uv chamber in lab scale treatment plant

Type of bacteria (per 100 ml) Raw water UV treated water
Coliformbacteria <1000 0
E coli bacteria 320 0

3 Conclusion

In the dry zone of Sri Lanka, existing traditional systems of water purification are not
capable enough to satisfy the requirement. Therefore, the latest technology should be
used for water purification. This study was done in a selected area at Anuradhapura and
Mihinthale to installing a pilot plant at Anuradhapura. The existing physicochemical
parameters of the water in the selected area were examined and those are hardness
844.33 mg/l, fluoride 1 mg/l, alkalinity 207.08 mg/l, conductivity 1186.61 µS/cm,
total dissolve solids 592.5 mg/l, total organic carbon 9.9 mg/l, pH 9, turbidity
0.11 mg/l.
Knowing the effects of the high concentration of fluoride and hardness in drinking
water, ground water in the selected area of dry zone in Sri Lanka was treated using a
RO membrane and pretreatment systems with combination of latest technology in order
to select the suitable and effective purification process. The selected purification pro-
cess of the RO membrane and the pre-treatment with sand filter, carbon filter, ion
exchange resin and ultra violet chamber was able to achieve the requirement of
reducing hardness and fluoride concentration of the ground water in the selected
 Hardness and Fluoride Removal Efficiency of Groundwater 181

general area in the dry zone of Sri Lanka. Purification efficiency of hardness and
fluoride are 95.17% and 94% respectively and physicochemical parameters of RO
treated water are hardness 42.07 mg/l, fluoride 0.06 mg/l, alkalinity 6.38 mg/l, con-
ductivity 108.61 µS/cm, total dissolve solids 32.31 mg/l, total organic carbon 0.7 mg/l,
pH 8.5, turbidity 0.057 mg/l.
Considering the above data, collected and analyzed, the purification process of the
RO membrane with pretreatment can be selected as the most effective and practical
system of water purification for drinking water.
The literature survey and the laboratory experiments have indicated that each of the
discussed techniques can remove fluoride and hardness under specified conditions. The
fluoride and hardness removal capacity varies according to many site-specific chemical,
geographical and economic conditions, so actual applications may vary from the
generalizations made. Any particular process, which is suitable at a particular region
may not meet the requirements at some other place. Therefore, any technology should
be tested using the actual water to be treated before implementation in the field.

Acknowledgments. I would like to express my deepest gratitude to Sir John Kotelawala

Defence University, Water safety Plan Advisory Unit, National Water Supply and Drainage
Board, Katugastota, National Institute of Fundamental Studies and Faculty of Engineering,
University of Peradeniya for giving the facilities to carry out experiments.

References
Kodikara1 DC, Kumara RDHA, Chaminda GT, Kawakami T (2015) Fluoride and hardness
removal from groundwater by electrolysis. In: 3rd international symposium on advances in
civil and environmental engineering
Degu G, Yigzaw T (2006) Research methodology. Ethiopia public health training initiative,
Ethiopia
Dharmagunawardhane HA, Dissanayake CB (1991) Hydrogeochemistry of deeptube wells in the
dry zone of Sri Lanka. In: The second international symposium on thegeology of Sri Lanka
Disanayaka C (2014) Water quality and dental health in the dry zone of Sri Lanka. Institute of
Fundamental Studies, Hantana Road, Kandy, Sri Lanka: s.n
Dissanayake CB (2007) The fluoride problem in the ground water of Sri. Int J Environ Stud 20
Emamjomeh MM (2006) Electrocoegulation technology as a process for defluoridation in water
treatment, s.l.: University of Wollongong thesis collection
Etiksson P (1988) Nanofiltration Extends the Range of membrane filtration. FilmTec
Corporation, 7200 Ohms Lane: s.n
Freeman B, Sagle A (nd) Fundamentals of Membranes for Water, s.l.: s.n
Haftenberger M, Viergutz G, Neumeister V, Hetzer G (2001) Total fluoride intake and urinary
excretion in German children. Fluoride
Herath KRPK, Ileperuma OA, Dharmagunawardhane HA (2005) Environmental health risk for
the chronic renal failure in Sri Lanka. Faculty of Science, University of Peradeniya,
Peradeniya, Sri Lanka: s.n
Hell F, Lahnsteiner J, Frischherzb H, Baumgartner G (1998) Experience with full-scale
electrodialysis. Elsevier, Burgenland
Jayalath CPG, Miguntanna NS, Perera HAKC (2016) Burnt clay bricks as an alternative filter
media for pebble matrix filters (PMF). Eng: J Inst Eng Sri Lanka 49:1–9
182 K. M. S. Kumara et al.

Jayasumana C (2013) Possible link of chronic arsenic toxicity with chronic kidney disease of
unknown etiology in Sri Lanka. Rajarata University of Sri Lanka, Faculty of Medicine and
Allied Sciences, s.n
Junco LI (1972) Renal failure and fluorosis. JAMA 222:783–785 Fluorine and dental health
Priyantha N, Keerthiratne S (1997) Removal of heavy metal ions from textile effluents. Ceylon J
Sci: Phys Sci 4(1):77–85
Kothari C (2006) Reseach methodology; methods and techniques. New Age International
Publishers, New Delhi
Kumar R (2014) Research Methodology. SAGA, New Delhi
Dapeng LI, Jiuhui QU (2009) The progress of catalytic technologies in water purification: a
review. J Environ Sci 21:713–719
Dharma-Wardana MWC, Amarasiri SL (2014) Chronic kidney disease of unknown aetiology
and. Environ Geochem Health
 Land-Use Planning Model: A Case Study
on Transit-Oriented Development (TOD)

G. A. M. A. S. B. Udapitiya1, R. M. T. Rathnayake1,
and K. K. K. Sylva2(&)
1
Department of Civil Engineering/Faculty of Engineering,
University of Peradeniya, Peradeniya, Sri Lanka
2
Department of Engineering Management/Faculty of Engineering,
University of Peradeniya, Peradeniya, Sri Lanka
[email protected]

Abstract. Many communities are shifting their land-use approaches by moving

towards cities for better access to transportation and other facilities resulting in
densification, particularly around main streets and urban centres. These trends
are driving the increased emphasis on better models of city planning for urban
centres, urbanization of suburbs of cities and crowding of either side of minor
and major highways of Sri Lanka. Transit Oriented Development (TOD), is
found to be a concept that has succeeded in reaching better sustainability for
both society and the environment closer to the main transit access. The concept
has been a success in the reduction of individual vehicle usage and the practice
in the usage of the main transit systems which is in lieu of the concept of
sustainability. In this milieu, this study was aimed at creating a TOD model for a
developing city as a case study to verify the pros and cons of the concept of
TOD in a context in Sri Lanka. A qualitative study was done with in-depth
interviews to verify the complementary factors of a selected village in Sri Lanka
to be converted to a TOD model as a case study. This paper presents the
conceptual model on TOD for Meewatura, Sri Lanka.

Keywords: Transit-Oriented development (TOD)
Land-Use planning

Model
Sustainability

1 Introduction

In recent decades, cities have expanded dramatically due to rapid urbanization pro-
cesses (Motieyan and Mesgar 2017). Consequently, several issues associated with the
management of urban built environments such as unplanned urban sprawl, unfair
distribution of land uses, and inappropriate utilization of infrastructures, have emerged.
As Motieyan and Mesgar (2017) points out a significant rise in the use of private cars
over public transit is one of the most conspicuous issues in many cities. According to
them, this issue can lead to both environmentally and socially harmful consequences,
such as traffic congestion, global warming, climate change, environmental pollution,
and socio-economic problems. On the other hand, in most developing countries, the
existing public transportation services are unsuccessful in attracting people because

© Springer Nature Singapore Pte Ltd. 2020

R. Dissanayake and P. Mendis (Eds.): ICSBE 2018, LNCE 44, pp. 183–192, 2020.
https://doi.org/10.1007/978-981-13-9749-3_17
184 G. A. M. A. S. B. Udapitiya et al.

land use characteristics are not considered when planning and designing public transit
for easy access of the public. According to Motieyan and Mesgar (2017) to facilitate
decision-making processes in urban areas, it is, therefore, necessary to integrate land
use and public transportation planning into a comprehensive index.
In order to achieve a favourable environment and a good image, there should be a
balanced built environment in a city. As Pathinayaka and Wijesundara (2015) state,
“cities are specially planned to enhance the functionality while balancing civic, com-
mercial and recreational activities”. With the failure of city planning models such as
sprawl development, models such as TOD and Business Improvement Districts
(BID) have been practiced in the world to achieve a better sustainability (Knowles
2012; Hesse 2009 and Cervero 2004; Hoyt 2004).
Sri Lanka is in the process of rapid development and population towards cities is
increasing making the cities unbearably crowded. This uncoordinated growth results in
an unplanned development of infrastructure and simultaneously increasing pollution.
Many recent incidents such as flooding in cities and problems with solid waste man-
agement are evidence for this uncoordinated and unplanned urbanization. When more
small villages add up to the same process of urbanization along major and minor
highways in Sri Lanka the problems are aggravated. It is, therefore, necessary to
introduce better land-use planning models to Sri Lankan villages on either side of the
major and minor highways. Introducing such models would enable ‘feed-in’ (Sylva
2018) solutions rather than concentrating on end-of-pipe solutions to the problems
created by improper planning such as solid waste management and dealing with traffic
related problems and pollution.
With the aim of reviewing better practices for community development and related
land use planning for small villages on either side of the minor and major highways of
Sri Lanka, this research study focuses on developing a model on Land-Use planning
based on the concept of TOD. It is expected to uplift the quality of life of the people
while maintaining sustainable practices by such a model. Meewathura village, with
similar characteristics depicted to the TOD concept, which is located in the Central
Province of Sri Lanka, is selected as the case study, for applying TOD concept
to research into the strengths and weaknesses of the application of the concept
in Sri Lanka.
A conceptual model for Meewathura, Sri Lanka, was developed through a com-
prehensive literature survey on the concept of TOD taking into consideration the
features of the concept. This paper presents the background to the study, the features of
the concept of TOD and the development of the conceptual model to Meewathura. The
paper aims to disseminate present practices in the sustainable built environment to the
research community specifically focused on city planning for residential communities.
Since there are no residential city models developed with the concept of TOD in Sri
Lanka this model will be an eye-opener for the planning sector or the public authorities
to create better models for villages developing into cities on either side of the major and
minor highways. Ultimately the society as a whole and the environment too will be
benefitted with the application of such sustainable city models in Sri Lanka.
 Land-Use Planning Model: A Case Study on Transit-Oriented Development (TOD) 185

2 City Development Models

Main city development models found in the literature are sprawl, TOD, and BID for
applications in different contexts (Knowles 2012; Hesse 2009 and Cervero 2004; Hoyt
2004). According to Knowles (2012), Hesse (2009) and Cervero (2004), sprawl
development, which encouraged human population away from urban centers into low-
density areas, was highly accepted in the past. Sprawl created monofunctional and
usually car-dependent communities, in a process called ‘suburbanization’ (Cervero and
Day 2008). This was the same reason for the failure of the sprawl concept; due to the
high dependency on private vehicles causing high expenses and high levels of envi-
ronmental pollution related to emissions.
With the identification of negative consequences of sprawl development models
such as TOD has been chosen in spite of other alternatives because of its broader
benefits. TOD is a type of community development that includes a mixture of housing,
office and other facilities integrated into a walkable neighbourhood and located within a
half-mile of quality public transportation (Cervero 1997). According to Cervero (1997),
TOD allows households to live in neighbourhoods where they can walk, bike, or take
transit to get around. With transportation, the second highest household expense after
housing, living near transit makes it easier for people to forego the high cost of auto
ownership. This not only benefits new and existing residents but also businesses, transit
agencies, local governments, merchants and developers are made to get positive effects
by mixing of these communities (Lu et al. 2018). BID, on the other hand, is a sus-
tainable concept introduced to improve business districts ranging from few blocks to
many miles of streetscape (Hoyt 2004). BID is mainly targeted on developing business
cities to fulfill the requirements of such a community with tax collection from the
community to run the city and a Public-Private Partnership to maintain the cities.

2.1 The Concept of TOD

TOD has emerged as one possible solution for sustainable urban transport, with sup-
porters claiming that it can help to reshape the quality and form of urban growth
towards enhanced accessibility and mobility, pedestrian friendliness, increased sus-
tainability and potentially a higher degree of human interaction (Curtis et al. 2009). By
definition (Knowles 2012; Hesse 2009 and Cervero 2004), a TOD community (or
neighbourhood) means creating a pedestrian-friendly built environment, in which the
centrally located rail or bus stations are surrounded by relatively high-density and mix-
used commercial and residential developments. According to Dittmar and Ohland
(2004) and Cervero and Day (2008), TOD communities are generally located within a
radius of about 600 m from a transit stop, as this is considered to be an appropriate
scale for pedestrians. A TOD city refers to the built urban forms that are composed of
numerous TOD communities built on the transit lines (Dittmar and Ohland 2004;
Cervero and Day 2008).
According to Cervero (1997), TOD is a type of community development that
includes a mixture of housing, office and other facilities integrated into a walkable
neighbourhood and located within a half-mile of quality public transportation. Research
by Hasibuana et al. (2013) found that the majority TOD is a concept of managing urban
186 G. A. M. A. S. B. Udapitiya et al.

growth in transit corridor which has characteristics of mixed land use, compact,
walking distance, and development focused around public transit area. Further research
(Shohani et al. 2016) introduce TOD as a process of reconnecting transportation and
land use and can be considered as a potential solution for revamping suburban
neighbourhoods by developing a new alternative transit system conforming to the
principles of TOD. They also suggest that TOD is a viable tool for achieving a sus-
tainable future. They too highlighted that transit can enable a city to use its market
forces to increase densities near stations where most services are situated, thereby
creating more efficient sub-centers and reducing urban sprawl, rediscovering ways in
which planning and design can better incorporate less automobile-dependent land use.
Thus, at its core, TOD is a process of reconnecting transportation and land use.
According to Shohani et al. (2016), on the basis of the extent of the area served,
TOD is identified as either Neighbourhood TODs or Regional TODs. Focusing on
Neighbourhood TODs the emphasis is on residential use with locally-oriented shop-
ping in facilities that are sized to serve the population living in close proximity.
Calthorpe (n.d.) has identified TOD as a mixed-use community within an average one-
fourth mile walking distance of a transit stops and core commercial area. Calthorpe (n.
d) further added more specifics to the definition as the design, configuration, and mix of
uses emphasize a pedestrian-oriented environment and reinforce the use of public
transportation; TOD is a mix residential, retail, office, open space, and public uses
within comfortable walking distance, making it convenient for residents and employees
to travel by transit, bicycle or foot, as well as by car.
Zeigler (2009) showed the practical implications of TOD as a metropolitan sus-
tainable development governing framework for growth management in the twenty-first
century is essential for a sustainable future. This includes Higher density urban centers
and a change in public attitude away from “not in my backyard” thinking.

2.2 Application of TOD

TOD allows efficient land use, energy and resources efficiency and helps to conserve
open space. Shohani et al. (2016) identified other benefits for the community of TOD
application as encouraging walking, a cleaner environment, and minimising traffic.
Referring to rail accessibility, Papa and Bertolini (n.d.) carried out a study to inves-
tigate how TOD structure affects accessibility in cities and found that the accessibility
increases in urban development clustering along rail corridors and stations. Lund et al.
(2004) studied the travel behaviour in California TODs. According to them, TOD
residents are more likely to use rail transit “if there is less of a time benefit for traveling
via highways (compared to transit), if there is good pedestrian connectivity at the
destination if they are allowed flexible work hours, and if they have limited vehicle
availability.” But they found that TOD residents are less likely to use transit if the trip
involved multiple stops or “trip chaining”, if there is good job accessibility via high-
ways, if they can park for free at their workplace, and if their employer helps to pay
vehicle expenses such as tolls, fuel, etc. Yap and Goh (2017) listed the consideration
factors to determine the potential requirements of TOD. Amenities related to jobs and
family for baby boomers, time-saving, location and house should be located in relation
to working place, school, etc.
 Land-Use Planning Model: A Case Study on Transit-Oriented Development (TOD) 187

Sung and Oh (2011) stated TOD development will depend upon the courage of the
political leaders and the effectiveness of the targeted policies. These will include
reshaping monotonous land use, comprehensive planning with a greater focus upon the
transport network, restricting the total freedom for citizens to own and drive their cars,
and focusing the power of real estate companies to meet wider sustainable transport
investments. According to them, these areas of policy will attract opposition and in
some cases go against the interest of filling the government coffers. According to Sung
and Oh (2011), it is expected that Chinese mayors with vision, courage and endurance
are likely to go this extra mile; in applying TOD in big Chinese cities it can only be
hoped that many of them indeed have these qualities.
According to Massachusetts Bay transportation authority & Massachusetts
department of transportation (2017), TOD means more than locating individual
buildings or projects near transit. As the department explains, it is a pattern of land use
that ties development to transit in a synergistic way at the station level, the corridor
level, and even the regional level. Every place is unique in a particular degree and the
department has identified four foundational principles for successful TOD’s.
• Density and a mix of uses: the first foundational principle of TOD is the concen-
tration of dense, mixed-use development within walking distance of stations.
Density places more homes, jobs, stores, offices, and recreational activities within
walking distance of transit which could relatively change with place and context.
• Equitable development: affordable and workforce housing is identified to be offsetting
gentrifying pressure on community residents and businesses since TOD can connect
housing with jobs, reducing a household’s combined cost of housing and commuting.
As employment becomes more decentralized, jobs located near transit remain
accessible to households that cannot afford to own one automobile per worker
• A great public realm: Every transit rider is a pedestrian, and successful TOD is
pedestrian-friendly. This means that station areas have safe, interconnected walking
environments with sidewalks, plazas, lighting, way-finding, and amenities.
• A TOD approach to parking: “right-size” overall parking supply in TOD settings
encourages shared parking, and ensure that parking is designed so as not to dom-
inate the streetscape or permanently block development on sites most convenient to
the station.
Hasibuana et al. (2013) have proposed ecological footprint, carbon emission, and
green open space carrying capacity as sustainable urban development indicators to
measure the success of TODs.

3 Development of TOD Model City for Meewathura

In the development of the conceptual model for Meewathura in relation to the concept
of TOD, first, a criteria comparison was done. The criteria comparison for Meewathura
through the research study is presented in Table 1. Then relevant areas to be developed
as residential and business, employment opportunities for residents and wildlife
habitats and parks were identified from the satellite map of Meewathura obtained from
Grama Niladari Boundary Map of Sri Lanka Survey Department. The Satellite Map of
Meewathura is presented in Fig. 1. The model was verified with a quantitative study.
188 G. A. M. A. S. B. Udapitiya et al.

Table 1. Criteria comparison for Meewathura TOD model

Features of the concept of TOD to be taken Features of Meewathura comparable to TOD
as indicators for development
Walkable neighbourhood and located within Meewathura is a small village of an area of
a half-mile of quality public transportation around 80.5 ha and a walkable distance of a
(Cervero 1997) quarter to half a kilometer
Pedestrian-friendly built environment, in The A5 road running through Meewathura
which the centrally located rail or bus and is currently being improved as a major
stations are surrounded by relatively high- connection between Colombo and
density and mix-used commercial and Nuwaraeliya. Main bus stop of Penideniya
residential developments (Knowles 2012; and railway station of Peradeniya are located
Hesse 2009 and Cervero 2004) in the visinity of the village at a walking
Communities are generally located within a distance of around a quarter to half a
radius of about 600 m from a transit stop kilometer
(Dittmar and Ohland 2004; Cervero and Day
2008)
Mixed-use community within an average The community in Meewatura has both
one-fourth mile walking distance of a transit residential and business mixed environment
stops and core commercial area which could be improved further with the
Calthorpe (n.d.) development of the A5 road
Sustainable with more efficient use of land, There are areas which are yet not subjected
energy and resources, helps conserve open to urbanisation which could be developed as
space, encourages walking, cleaner park-lands and wild-life habitats allowing
environment, Shohani et al. (2016) connectivity along the river flowing adjacent
to the village and along the railway track
where foot-paths could be located

Fig. 1. The satellite map of Meewathura

 Land-Use Planning Model: A Case Study on Transit-Oriented Development (TOD) 189

Connectivity for the wildlife habitats was identified for the sustainability of the city
either from ground level or other as appropriate owing to the indicators suggested by
Hasibuana et al. (2013) in order to reduce the ecological footprint and have more green
open space carrying capacity for sustainability. Some connectivity paths had to be
proposed above ground level or below ground level. Footpaths were identified along
the rail-track and major highways which could also be used as jogging tracks or cycling
lanes for residents. These footpaths were connected to railway stations and major bus
stops.
The principles identified by Massachusetts Bay transportation authority & Mas-
sachusetts department of transportation (2017) were used as the main criteria in
selecting spaces for proposed development which are presented in Table 2.

Table 2. Principles used in the development of TOD for Meewathura

Principle Description
Density and mis of • Mixed-use development
uses • Homes, jobs, stores, offices, and recreational activities within
walking distance of transit
Equitable • Affordable and workforce housing
development • Connect housing with jobs
• Decentralized employment opportunities
• Jobs near transit accessible to households that cannot afford to own
one automobile per worker
A greater public • Pedestrian-friendly
realm • Safe, interconnected walking environments with sidewalks, plazas,
lighting, way-finding, and amenities
A TOD approach to • “Right-size” overall parking supply
parking • Shared parking not to dominate the streetscape or permanently block
development on sites most convenient to the station

A proposal on a two-dimensional model was developed by identifying essential

features of TOD and trying to achieve these with the minimum disturbance to existing
features in Meewathura. A mixed housing, commercial and parking were proposed
with employment opportunities identified in the commercial and nearby university
premises for people who are unskilled and households that cannot afford to own one
automobile per worker. Transit access for other workers to move away from the city for
employment was identified within walking distance to main mass transit systems. Safe
walking paths preventing the illegal crossing of rail tracks were identified on either side
of the rail track with safe connectivity. The TOD model for Meewathura is presented in
Fig. 2.
190 G. A. M. A. S. B. Udapitiya et al.

Fig. 2. The TOD model for Meewathura

4 Conclusions

The city model developed for Meewathura using the concept of TOD is expected to be
a sustainable model in terms of economy, society and environment. The economic
factors of the society are expected to be developed by providing employment oppor-
tunities to the city dwellers in the vicinity and allowing transit access to employment
opportunities to others who are travelling out of the village. The society will benefit
 Land-Use Planning Model: A Case Study on Transit-Oriented Development (TOD) 191

from a cleaner and nature-friendly environment which allows public recreational areas
and walkways. The environment will benefit from the regeneration of natural habitats
and environmentally friendly living styles of people. The green connectivity proposed
will enhance the sustainability of biodiversity.
A mix-use development of residential and commercial activities will enable the
community to reduce transit requirements and this would reduce environmental
impacts related to settlements. In general, the sustainability of the city will be improved
in the application of a model built on the concept of TOD. The expected results are to
be verified after the application of the model.

References
Calthorpe P The next American metropolis: ecology, community, and the American dream.
Princeton Architectural Press, New York
Cervero R (1997) Transit-oriented development in the United States: experiences, challenges,
and prospects, vol 102. Transportation research board, Washington
Cervero R, Day J (2008) Suburbanization and transit-oriented development in China. Transp
Policy 15(5):315–323
Curtis C, Scheurer J (2017) Performance measures for public transport accessibility: learning
from international practice. J Trans Land Use 10(1):1–26
Dittmar H, Ohland G (2004) The new transit town. Best practices in transit-oriented
development. Island Press, London
Grama niladari boundary map of Sri Lanka. Survey department, Sri Lanka. https://www.survey.
gov.lk/gn_updating/. Accessed 1 May 2018
Hasibuana HS, Soemardia TP, Koestoerb R, Moersidik S (2013) The role of transit oriented
development in constructing urban environment sustainability, the case of Jabodetabek, no. 4.
Post Graduate of Environmental Science, University of Indonesia
Hesse M (2009) Transit oriented development: making it happen. J Trans Geogr. 17(6):510.
https://doi.org/10.1016/j.jtrangeo.2009.07.007
Hoyt LM (2005) Do business improvement district organizations make a difference? Crime in
and around commercial areas in Philadelphia
Knowles RD (2012) Transit oriented development in Copenhagen, Denmark: from the Finger
Plan to Ørestad. J Trans Geogr 22:251–261. https://doi.org/10.1016/j.jtrangeo.2012.01.009
Lu Y, Gou Z, Xiao Y, Sakar C, Zacharias J (2018) Do transit-oriented developments (TODs) and
established urban neighborhoods have similar walking levels in Hong Kon’. Int J Environ Res
Public Health 15:555
Lund HM, Cervero PR, Willson RW (2004) Travel characteristics of transit-oriented
development in California. https://www.bart.gov/sites/default/files/docs/Travel_of_TOD.pdf
Massachusetts Bay Transportation Authority & Massachusetts Department of Transportation
(2017) TOD Policies and Guidelines
Motieyan H, Mesgar MS (2017) Towards sustainable urban planning, through transit-oriented
development (a case study: Tehran). Int J Geo-Inf 6:402
Papa E, Bertolini L (n.d.) Accessibility impacts of TOD experiences in European metropolitan
areas. https://biblio.ugent.be/publication/5789325/file/5789360. Accessed 05 May 2018
Pathinayaka AD, Wijesundara J (2015) Imaging cities; social perception and physical
composition; case study of Pettah, Colombo. J Cities People Places 1(1):47–60
192 G. A. M. A. S. B. Udapitiya et al.

Sung H, Oh J (2011) Transit-oriented development in a high-density city: identifying its

association with transit ridership in Seoul. J. Science Direct Cities. 28:70–82
Shohani AV, Thomas M, Rao KVK (2016) World conference on transport research, masters
student. Civil Engineering Department, Bombay, Mumbai 400 076, India
Sylva KKK (2018) A circular economic model for a sustainable city in South Asia. In: Leal
Filho W, Rogers J, Iyer-Raniga U (eds) Sustainable development research in the Asia-Pacific
region. World Sustainability Series. Springer, Cham, pp 345–359
Yap JBH, Goh SV (2017) Determining the potential and requirements or transit oriented
development (TOD): the case of Malaysia. Property Manag 35(4):394–413. https://doi.org/10.
1108/PM-06-2016-0030
Zeigler EH (2009) The case for megapolitan growth management in the 21st century: regional
urban planning and sustainable development in the united states. Int J Law Built Environ 1(2)
 Effective Sizing of Natural Systems for Grey
Water Treatment Using Locally Derived
Parameters

S. Karunaratne(&)

Department of Civil Engineering, Faculty of Engineering,

Sri Lanka Institute of Information Technology, Malabe, Sri Lanka
[email protected]

Abstract. When properly managed, grey water can be a valuable resource,

which agricultural and horticultural practitioners as well as home gardeners can
benefit from using to provide an excellent nutrient sources for vegetation.
Natural treatment systems such as constructed wetlands emphasizing on sus-
tainability and wastewater reuse typically for agricultural irrigation/home gar-
dening can be considered as an ideal alternative to close the loop in terms of
nutrient capture and water conservation. This paper presents a part of an analysis
done to estimate the locally derived treatment wetland parameters based on
BOD5 removal mechanisms of a pilot scale subsurface horizontal flow con-
structed wetland (SSHF CW) treating greywater. A pilot scale SSHF CW with a
surface area of 8.64 m2 and a substrate depth (with 20 mm aggregates) of 0.7 m
was monitored. This is a part of a long term study carried out to test the viability
of SSHFCW for treating grey water for agricultural reuse. The wastewater flow
measurements were taken from the inlet and the outlet of the constructed wet-
land while monitoring the hydraulic (HRT) and organic loading rates. Water
samples from four sampling points were collected on a long term basis. Equa-
tions based on a kinetic model of a first-order piston flow reactor was used to
calculate the relationship among the design parameters appropriate for local
conditions. The HRT varied from 3.3 to 4.7 d for maximum and minimum
inflow loading rates observed. Using the data from the current study, a resizing
exercise for a subsurface horizontal flow constructed wetland was carried out.
Accordingly, a household with four people, discharging 720 L/d greywater with
an average influent BOD5 of 300 mg/L could be treated up a BOD5 concen-
tration of 25 mg/L using an area of 3.60 m2 of SSHF CW, implying only
0.975 m2 of per capita land usage.

Keywords: Hydraulic retention time
Inflow rate
Per capita land usage

Subsurface horizontal flow
Treatment efficiency

1 Introduction

According to the Sri Lanka Standard code of practise for the design and construction of
septic tanks and associated effluent disposal systems SLS 745-Part 2, generally black
water comprises of only 25% of combined (all) waste water. However, this small

© Springer Nature Singapore Pte Ltd. 2020

R. Dissanayake and P. Mendis (Eds.): ICSBE 2018, LNCE 44, pp. 193–201, 2020.
https://doi.org/10.1007/978-981-13-9749-3_18
194 S. Karunaratne

portion of black water contains 90% of Nitrogen, 60% of COD and 50% of P. On the
other hand, 75% of all wastewater contains relatively a small amount of pollutants.
Furthermore, the nitrogen found in grey water can be filtered out and used by plants
readily. However, grey water, if left untreated for a few days, will behave like sewage,
becoming malodorous becoming anaerobic. Also, grey water decomposes much faster
than black water.
When properly managed, grey water can be a valuable resource, which agricultural
and horticultural practitioners as well as home gardeners can benefit from providing an
excellent nutrient source for vegetation.
Grey water irrigation has long been practiced in areas where water is in short
supply. The engineering of these systems is still a relatively young technology; but it is
one making rapid progress. As these systems utilize the nutrient (potential pollutant)
content in the effluent, they constitute a real solution to the treatment/management of
grey water.
A largely undocumented practice which occurs in the country (Sri Lanka) in most
semi-urban and rural areas is the separation of black water from grey water. Usually the
toilet waste goes to a septic system while all grey water is discharged mostly untreated
directly into the environment, especially in rural areas or soakage pits in urban and
semi urban areas. This pre-existing culture of source separation has the potential for the
development of sustainable wastewater treatment systems based on the principles of
natural treatment or ecological sanitation. Natural treatment systems such as con-
structed wetlands emphasizing on sustainability and wastewater reuse typically for
agricultural irrigation/home gardening can be considered as an ideal alternative to close
the loop in terms of nutrient capture and water conservation.
The demand for treated water is reaching a level which might not be satisfied with the
available resources even in places where there is no physical lack of it. In such situations,
the necessity of effective and efficient wastewater treatment and promotion of water
reuse would be highly productive. Hence, the use of highly favoured treatment systems
such as constructed wetlands which are associated with natural removal mechanisms of
pollutant removal could be successful applications, especially in developing countries
since these systems combine low-cost, low-maintenance, simple and reliable operations,
and high removal efficiency (Economopoulou and Tsihrintzis 2003).
In general, Subsurface Flow (SF) constructed wetlands can treat a variety of
wastewaters including domestic (black water and greywater) and municipal wastewa-
ter, industrial wastewater, agricultural and urban storm water runoff, combined sewer
overflow and landfill leachate. Amongst most popular types of constructed wetlands,
the Subsurface Horizontal Flow Constructed Wetlands (SHFCW) are considerably
efficient in the removal of pollutants and eliminate some of the common problems such
as bad odour, growth of mosquitoes, etc. Relatively, less research has been done so far
on the use of constructed wetlands for greywater treatment in tropical areas and the
design parameters still remain as black boxes due to the complexity of treatment
mechanisms (Economopoulou and Tsihrintzis 2003). In Sri Lanka too, constructed
wetlands have not yet been identified as a very efficient and cost effective wastewater
treatment method especially due to the notion of the high land demand, which is
prohibitive in urban and peri-urban areas. The main reason for this fact could be the
 Effective Sizing of Natural Systems for Grey Water Treatment 195

insufficient research work on constructed wetlands, specifically under the local envi-
ronmental conditions of the country. The design information available in the temperate
region of the world cannot be translated to Sri Lanka for optimization of pollutant
removal mechanisms and design criterion of constructed wetlands due to the varying
climatic conditions. Therefore, it is an essential requirement to estimate both design and
operational parameters of the constructed wetland under the specific climatic and
environmental conditions in the country.
Few researches have been carried out focussing on parameter estimation and
optimization of SF constructed wetland systems in Europe and United States. Econ-
omopoulou and Tsihrintzis (2003) have presented a simple criterion including the
guidelines and mathematical models for the preliminary sizing of subsurface horizontal
flow constructed wetland systems by considering the conditions of definite seasonal
variations. But, this research work still has some limitations to apply the methodology
in a tropical climate with the values given for certain parameters. However, the
methodology could provide relevant guidelines for the estimation of parameters in a
tropical climate and determining removal efficiencies of existing systems.
Wynn and Liehr (2001a, b) developed a mechanistic compartmental simulation to
model and predict seasonal trends in the removal efficiencies of HFCWs. The model
consists of six sub-models representing the carbon cycle, the nitrogen cycle, an oxygen
balance, autotrophic and heterotrophic bacteria growth, and a water budget. Therefore,
this could make a close relationship with the project under concern.
Vymazal (2005) has provided a very important discussion on subsurface horizontal
flow and hybrid constructed wetlands including the pollutant removal mechanisms and
their efficiencies with experimental results. Since the experimental data and analyses
are based on the temperate continental climate conditions, it would not be possible to
make a direct relationship with the local conditions.
Werner and Kadlec (2000) and Chazarenc et al. (2003) modelled wetland residence
time distributions in temperate climes. Garcia et al. (2004) and Maloszewski et al.
(2006) used tracer experiments to determine hydraulic parameters with different filter
materials.
In Sri Lanka, the acceptance of constructed wetlands as an efficient and cost
effective wastewater treatment method is not yet established. This may be especially
due to the scarcity of the land due to high population density as well as the insufficient
research work on constructed wetlands, specifically under the local conditions.
Therefore, it is an essential requirement to estimate both design and operational
parameters of the constructed wetland under the specific climatic and environmental
conditions in the country (Vymazal 2011).
Treatment systems based on natural degradation processes, such as stabilization
ponds and constructed wetlands, are particularly suited for domestic wastewater
treatment and they require little or no energy, are relatively simple to operate, and show
reliable treatment performance.
This paper presents a part of an analysis done to estimate the locally derived
treatment wetland parameters based on BOD5 removal mechanisms of a pilot scale
subsurface horizontal flow constructed wetland (SSHF CW) treating greywater. This is
a part of a long term study carried out to test the viability of SSHFCW for treating grey
water.
196 S. Karunaratne

2 Methodology
2.1 Pilot Scale Subsurface Flow Constructed Wetland
The subsurface flow constructed wetland system consisted of two settling tanks with a
grease trap for the pre-treatment (Fig. 1).

S
S
S

Fig. 1. Configuration of pilot scale subsurface flow constructed wetland (SSFCW) indicating
the sampling points (SP 1-4).

The pre-treated water was introduced into the wetland cell through a 50 mm per-
forated pipe and both inlet and outlet partitions were filled with 40 mm aggregates. The
water passes through a 20 mm aggregate bed with baffles inside the wetland cell.
Wetland cell was partitioned width-wise to extend the flow path creating a long and
narrow treatment cell. Wetland cell maintained a bed slope of 1% with the cannas as
the vegetation.
Table 1 shows the characteristics of the pilot scale subsurface flow constructed
wetland (fed with grey water discharged from a canteen) used for the study.

Table 1. Characteristics of the pilot scale subsurface flow constructed wetland

Parameter Value
U - porosity 0.435
d - substrate depth 0.70 m
L - length 4.00 m
W - width 2.16 m
Ve - effective volume 2.643 m3
A - surface area 8.64 m2

2.2 Flow Measurement, Sample Collection and Analysis

The wastewater flow measurements were taken from the inlet and the outlet of the
constructed wetland while monitoring the loading rate. Water samples from four
sampling points (at the sink outlet, after primary settling, at inlet to the CW and outlet
 Effective Sizing of Natural Systems for Grey Water Treatment 197

of the CW) were collected on 1–2 week time intervals and analyzed for pH, Tem-
perature, Turbidity, Conductivity, Biological Oxygen Demand (BOD5), Chemical
Oxygen Demand (COD), Nitrate Nitrogen, Nitrite Nitrogen, Total Phosphorous, Total
Coliforms and Suspended Solids. However the present analysis concentrates mostly on
the BOD5 measurement data.
The efficiency of the concentration abatement of the pollutants were calculated
using Eq. 1.
Percentage abatement ð%Þ ¼ ðCo  CeÞ=Co ð1Þ

Similar to Andreo-Martínez et al. (2016), equations derived by Reed et al. (1995)

based on a kinetic model of a first-order piston flow reactor proposed was used to
calculate the relationship among the design parameters appropriate for local conditions
(Eq. 2). By using Eq. 2, the theoretical wetland surface areas were calculated for the
measured data.

As ¼ ½QðlnC0  lnCe Þ=Kt :d:n ð2Þ

Where, As-Surface wetland area (m2), Co-inlet BOD5 concentration, Ce- Outlet
BOD5 concentration, Q-inflow rate (m3/d), Kt temperature dependent reaction coeffi-
cient (as given in Eq. 3), d- treatment wetland depth in m, n- porosity of the substrate
media.

Kt ¼ ð1:104Þ : ð1:06ÞðTo 20Þ ð3Þ

Where, Kt rate is constant corresponding water temperature in wetland (d−1), To is

the water temperature in the treatment wetland, which was taken as 25 °C.
Hydraulic retention time (HRT) for corresponding observed data was calculated by
using Eq. 4.

HRT ¼ ðAs :d:nÞ=Q ð4Þ

Previous studies (Andreo-Martínez et al. 2016; Karunaratne et al. 2011;

Udukumburage et al. 2013) have shown that, BOD5 can be considered as the governing
parameter for the surface area optimisation of SSHF CW design when treated for reuse
in irrigation as P and N removal is not an objective. P and N could be retained as plant
nutrient.
Accordingly, using the observed BOD5 data, relationships between the optimum
wetland surface area, inflow rates and hydraulic retention time were calculated using
Eq. 2 and trend lines were established using power functions to make predictions.

3 Results and Discussion

The ratio of BOD5/COD is related to the biodegradability of the organic matter present
in the water. In the present study, the ratio BOD5/COD for influent wastewaters had an
average of 0.74 ± 0.08 in raw wastewater, a characteristic value for wastewater of high
198 S. Karunaratne

biodegradability. The average BOD5 of the influent was 892 ± 18 mg/L, a relatively
higher value than the usual domestic grey water in Sri Lanka. This higher value is
mainly due to the food varieties, mostly Sri Lankan style menus served in the staff
canteen. Average PO43-, NO3-and BOD of the influent greywater over the study
period was 1.79, 42.62 and 892.19 mg/L respectively.
Figure 2 shows the variation of treatment efficiencies with varying loading rates.
Inflow rate ranged from 0.56 m3/day to 1.55 m3/day.
At all loading rates, treatment efficiencies of BOD5 did not fall below 65%. At the
minimum loading rate, all parameters measured displayed efficiencies ranging from
62–88%, removal of P showing the minimum while BOD5 showing the highest. Same
trend was observed generally at all loading rates. Even at the highest loading rate of
1.55 m3/d, N removal efficiency of more than 50% was recorded.
efficiency
Treatment

Inflow rate (L/day)

Fig. 2. Variation of treatment efficiency of pollutants with varying inflow rate of the subsurface
horizontal flow constructed wetland

The optimum retention time that would yield the minimum required treatment level
was calculated using the measured data and the temperature dependent reaction rate
constant for BOD5 was calculated based on the Eqs. 2 and 3. The HRT varied from 3.3
to 4.7 d for maximum and minimum inflow loading rates observed. This is compara-
tively lesser than the HRT reported from temperate region but reasonable when con-
sidering the higher temperature observed in Sri Lanka (Andreo-Martínez et al. 2016).
Generally, sizing of wetlands treating greywater is performed based on the COD or
BOD5 removal efficiencies. Targeting higher removal efficiencies of N and P through
constructed wetlands demands higher land requirements, hence, making such targets
impractical. It was observed that very few models have been developed to simulate
pollutant removal processes in SFCWs and only a fewer number of models of these are
applicable for SSHFCWs (Langergraber 2008). In fact, many of those models are
 Effective Sizing of Natural Systems for Grey Water Treatment 199

applicable for a particular climate or a clime where the model has been developed.
Such models should be modified before applying for different conditions. Modification
of an existing model can be rather difficult and complex than developing a new model
in relation with the local conditions. Modelling of pollutant removal mechanisms of a
SSHFCW in a particular climate and sizing of the wetland systems would be a chal-
lenge because, the procedure should consist of the formulation of removal processes,
the estimation of parameters in the model and the sizing of the system.
It is known that efficiency of wastewater treatment depends on the influent water
quality, climate conditions as well as substrate type and plants used. The reaction rate
constants yielded in this study were calculated under high loading rates (i.e. three times
the original designed). Therefore, influent water quality had a higher variability over
the prolonged study period. Hence, it was important to calculate the reaction rate
constants under varied loading rates and compare the same for better accuracy. Kt
calculated at 25 °C was 1.5 d−1.
Figure 3 shows the relationship between the inflow BOD loading (in g) of grey-
water and the load reduction per unit surface area (kg m2).
Figure 4 shows the relationship between the wetland surface area and the hydraulic
retention time calculated for the SSHF CW in consideration. Equations 2 and 3 were
used to calculate the HRT and wetland surface areas based on the observed data. As
mentioned earlier, the estimations of hydraulic retention time for specific treatment
levels was based on the BOD5 treatment efficiency only. Figures 3 and 4 can be used
with reasonable accuracy to size treatment wetland cells to treat greywater discharged
from households. The relevant estimation of flowrates can be based on SLS 745 (Part 2)
standards.

600
BOD load reduction per m2 (kg m2)

500 y = 0.2701x + 1.3563

R² = 0.9135
400

300

200

100

0
0 250 500 750 1000 1250 1500 1750 2000
BOD load (kg)

Fig. 3. The relationship between the inflow BOD loading (in kgs) of wastewater and the load
reduction per unit surface area (kg m2).
200 S. Karunaratne

Hydraulic retention time (d)

5
4
3
2
y = 1.8721e0.2019x
1 R² = 0.4326

0
2.5 2.75 3 3.25 3.5 3.75 4 4.25 4.5
Wetland surface area (m2)

Fig. 4. Relation between the wetland surface area (As) and the HRT.

Using the outcomes from the current study, a resizing exercise of SSHFCW was
carried out as shown below;
A household with 4 people was considered. As per the SLS standards 745 (part 2),
greywater discharge of 180 L/p/d accounts for 720 L/d of grey water from the
household. An average influent BOD5 of 300 mg/L (according to literature it can vary
from 200–500) and effluent concentration of 25 mg/L was considered. Hence, BOD
load will be 216 g/d (720 * 300/1000). A more stringent discharge standard was
considered due to the undiluted discharge conditions prevailing at domestic environ-
ments for irrigation purposes.
Using Fig. 3, BOD5 load reduction per m2 corresponding to 216 g of BOD5 load
can be estimated to be 59.68 BOD/m2. As such, the surface area required is 3.6 m2
(216/59.68). Using Fig. 4, HRT corresponding to 3.6 m2 of wetland surface area is
estimated as 3.9 days. According to the present calculation, average strength grey
water, usually discharged from a typical Sri Lankan household can be treated using an
area of 0.975 m2/person surface area of a SSHF CW.
If a ratio L/W of 3:1 is assumed (Andreo-Martínez et al. 2016), L and W can be
calculated as, W = 1.10 m while L = 3.30 m, respectively. The plan dimensions can
even be readjusted to maintain the total surface area depending on the space availability
of the location. The depth of the substrate is 0.7 m. This can be considered as a size
which can be conveniently accommodated even in an urban setup and mimic a flower
trough if plants such as Cannas are used.

4 Conclusion

An analysis was done to estimate the locally derived treatment wetland parameters
based on BOD5 removal mechanisms of a pilot scale subsurface horizontal flow
constructed wetland (SSHF CW) treating greywater. The treatment wetland resizing
exercise for a subsurface horizontal flow constructed wetland revealed that for a
household with 4 people discharging 720 L/d greywater with an average influent BOD5
of 300 mg/L could be treated up a BOD5 concentration of 25 mg/L using an area of
3.6 m2, implying only 0.975 m2 of per capita land usage.
 Effective Sizing of Natural Systems for Grey Water Treatment 201

Acknowledgements. The author would like to acknowledge the contributions by Udukum-

burage RS, Ariyasena MADN, Rupasinghe GGMBMB, Wijesiri, B. S., Jayasooriya, V. M. for
the field work as well as laboratory analysis done in relation to this work.

References
Andreo-Martínez P, García-Martínez N, Almela L (2016) Domestic wastewater depuration using
a horizontal subsurface flow constructed wetland and theoretical surface optimization: a case
study under dry Mediterranean climate. Water 8:434. https://doi.org/10.3390/w8100434
Chazarenc F, Merlin G, Gonthier Y (2003) Hydrodynamics of horizontal subsurface flow
constructed wetlands. Ecol Eng 21:165–173
Economopoulo MA, Tsihrintzis VA (2003) Design methodology and area sensitivity analysis of
horizontal subsurface flow constructed wetlands. Water Resour Manage 17:147–162
García J, Chiva J, Aguirre P, Alvarez E, Sierra JP, Mujeriego R (2004) Hydraulic behaviour of
horizontal subsurface flow constructed wetlands with different aspect ratio and granular
medium size. Ecol Eng 23:177
Langergraber G (2008) Modeling of processes in subsurface flow constructed wetlands—a
review. Vadoze Zone J 7(2):830–842
Karunaratne S, Wijesiri BS, Jayasooriya VM (2011) Estimation of reaction rate constants of
pollution removal for subsurface flow constructed wetlands treating grey water. In:
Proceedings of IESL-SSMS joint international symposium, SSMS 2011, 14th – 16th
September 2011, Colombo, Sri Lanka, pp 289–293
Małoszewski P, Wachniew P, Czupryński P (2006) Study of hydraulic parameters in heterogeneous
gravel beds: constructed wetland in Nowa Słupia (Poland). J Hydrol 331:630–642
Reed SC, Crites RW, Middlebrooks EJ (1995) Natural systems for waste management and
treatment, 2nd edn. McGraw-Hill, New York, p 433
Sri Lanka standard code of practise for the design and construction of septic tanks and associated
effluent disposal systems SLS 745-Part 2
Udukumburage RS, Ariyasena MADN, Rupasinghe GGMBMB, Karunarathne S (2013) Effect of
hydraulic loading rate on treatment efficiency of a subsurface horizontal flow constructed
wetland. In: Proceedings of Sri Lanka water convention 2013, 20–21 March 2013, Colombo,
Sri Lanka
Vymazal J (2005) Horizontal sub-surface flow and hybrid constructed wetlands systems for
wastewater treatment. Ecol Eng 25(5):478–490. https://doi.org/10.1016/j.ecoleng.2005.07.010
Vymazal J (2011) Constructed wetlands for wastewater treatment. Environ Sci Technol 45:61–69
Werner TM, Kadlec RH (2000) Wetland residence time modeling. Ecol Eng 15:77–90
Wynn TM, Liehr SK (2001a) Development of a constructed subsurface flow wetland simulation
model. Ecol Eng 16:519–536
Wynn TM, Liehr SK (2001b) Development of a constructed sub-surface flow wetland simulation
model. Elsevier Science, Amsterdam, pp 1–18
 Sustainability via ERP and BIM Integration

F. S. T. Hewavitharana(&) and A. A. D. A. J. Perera

Department of Civil Engineering, University of Moratuwa, Katubedda, Sri Lanka

[email protected]

Abstract. Sustainability is a paramount factor in the present construction

industry which enhances the quality of life, thus allowing people to live in a
healthy environment and improves social, economic and environmental condi-
tions for present and future generations. Therefore, lots of construction com-
panies have attempted to use different kinds of software for the development of
sustainable structures. Two of those major software are Building Information
Modelling system (BIM) and Enterprise Resource Planning system (ERP). This
review seeks to identify how ERP and BIM addresses sustainability at the
current level, a way of integrating BIM and ERP and how the integration of
these two leads to sustainability of the projects according to the current litera-
ture. It is identified that there is a significant impact on sustainability by inte-
grating ERP and BIM together than implementing them separately.

Keywords: Sustainability
ERP
BIM
Integration

1 Introduction

The term sustainable development can be described as enhancing the quality of life,
allowing people to live in a healthy environment and improving social, economic and
environmental conditions for present and future generations [11]. More than ever, the
improvement of these indicators is drawing the attention of the construction industry,
which is a globally emerging sector and a highly active industry in both developed and
developing countries [6]. Because of this rapid development in the construction
industry, lots of sustainable issues are created. Some of them are problems related to
designs, over usage of energy and resources, problems related to the integration of
building systems, non-adherence to ethical standards, negative environmental impacts,
and excess utilization of money [1]. Thus, to overcome these negative impacts, sus-
tainability practices should be carried out during project design, construction, usage,
maintenance, and demolition periods respectively [20].
In the current construction practice, although BIM is developing a sustainable
design in the design phase of the project there is a lack of maintaining the sustainability
aspects throughout the operation. To continue that sustainable practice throughout the
project lifecycle, ERP will act as a bridge which connects sustainable design data with
the operation of activities.
In the design stage, BIM can assist in performing complex building performance
analyses to ensure an optimized sustainable building design due to its representation of
the building as an integrated database of coordinated information [12, 17]. Further,

© Springer Nature Singapore Pte Ltd. 2020

R. Dissanayake and P. Mendis (Eds.): ICSBE 2018, LNCE 44, pp. 202–210, 2020.
https://doi.org/10.1007/978-981-13-9749-3_19
 Sustainability via ERP and BIM Integration 203

BIM can manipulate energy performance data such as power consumption, tempera-
ture, CO2 emission and environmental performance data such as indoor air quality
[20]. But the issue is, BIM is not capable of conducting operations sustainably in the
construction phase.
However, in the construction stage, ERP can integrate and streamline the business
processes, reduce costs and improve the quality which leads to economic improve-
ments, social equity & justice and environmental protection [5, 13]. Further, as each
ERP module provides the standard business processes with an integrated set of func-
tions, it facilitates the organization to improve and optimize current business processes.
But the issue is, ERP is not capable of designing sustainable drawings in the design
stage.
Hence, it is clear that there is a gap in operationalization of business functions at the
design stage and the construction stage of the construction industry. The bridging of
this two software would be the best solution to fill this gap.
However, the integration of these two software is challengeable as they are not
compatible with each other with the functional structure. Although the integration of
different architectural features in BIM and ERP are difficult, there are significant
benefits which can be achieved through this integration [3, 4]. Therefore, this paper
aims to contribute to the body of knowledge by reviewing existing literature on this
specific topic. The objective of this review is multifold, first, to identify and summarize
the contribution of ERP and BIM for the sustainability aspects in the current con-
struction industry. Second, to identify the most effective way of integrating ERP and
BIM. Third, to identify how BIM and ERP integration can support the sustainable
development of the construction industry.

2 Methodology

Literature review can be named as a well-established method for accumulating existing

knowledge within a domain of interest [21]. As the methodology for this article, a
systematic review approach is expected to be applied. This approach is characterized by
adopting explicit procedures and conditions which minimize bias. To find relevant
articles for this review, various keywords searches have been conducted in Science
Direct. In Table 1 an overview of the included articles in this research is presented. The
period of publishing was set to the years from 2000–2017 in order to narrow down the
search. A combination of different keywords such as “ERP and Sustainability”, “BIM
and Sustainability”, “ERP and BIM integration”, “Sustainability with ERP and BIM
integration” were used to find relevant articles to choose the topic. However, few
articles were identified relating to the targeted topic in specific. Therefore, the topics
were broadened, and article’s abstracts were carefully read to check their relevance.
204 F. S. T. Hewavitharana and A. A. D. A. J. Perera

Table 1. Overview of the articles

Author(s) Title Year Journal
Hande Aladaga, Gökhan Building Information Modelling 2016 Procedia
Demirdögenb, Zeynep Isıkc (BIM) use in Turkish Engineering
construction industry
Ibrahim Motawa, Kate Carterb Sustainable BIM-based 2012 Procedia social
Evaluation of Buildings and behavioural
sciences
Donya Mehran Exploring the Adoption of BIM 2016 Procedia
in the UAE construction industry Engineering
for AEC firms
Eng.Parisa Esmaeili, Moakher Building Information Modelling 2012 IOSR Journal of
and Prof. Dr. S. S. Pimplikar (BIM) and Sustainability – Using Mechanical and
Design Technology in Energy Civil Engineering
Efficient Modelling
Saumyendu Ghosh, Sam Impact of Sustainability on 2008 Ph.D. Thesis
Negahban, Young Hoon Kwak, Integration and Interoperability
and Miroslaw J. Skibniewski between BIM and ERP–A
Governance Framework
Yen-Ray CHENab and H. Ping An Integrated Methodology for 2017 (ISARC 2017)
TSERNGb Construction BIM & ERP by
Using UML Tool
Mohamed Abdalla, Mohamed Sustainability of Enterprise 2015 Ph.D. Thesis
Badreldin Lotfy Resource Planning
(ERP) Benefits Post-
implementation: An Individual
User Perspective

3 Finding and Discussion

BIM and Sustainability

Previously, traditional computer-aided design (CAD) tools are used in modelling the
construction designs [2]. Further, energy simulation packages such as Energy plus,
Ecotect and IEs virtual environment were used to consider design features such as
Thermal insulation, Climatic response glazing, Shading, Solar gain, Solar penetration,
Airtightness, Natural ventilation, and Mechanical ventilation [17].
Today these multiple operations are replaced by one system called BIM which
improves the performance through the whole life cycle of building [1]. Based on these
understandings, BIM can be used to minimize energy consumption, reduce water
consumption, reduce carbon footprint, select sustainable materials, reduce waste,
conserve the natural environment and safeguard human health and wellbeing [8, 12].
For this type of sustainability assessment, it is essential to incorporate the green
building assessment criteria into BIM [11]. At the present condition, BIM is mostly
utilized in the design phase of the construction project, and it is a well-recognized tool
in sustainable building design [1]. However, at the current level, BIM is unable to help
 Sustainability via ERP and BIM Integration 205

with functionalities of the project at the operational stage. Therefore, there should be
separate software to carry out those operations. In that case, ERP would be the best
solution to integrate with the BIM.
When considering the evolution of BIM, it is useful to see the prediction which has
been made at the initial stage. Now, most of those predictions have become real.
According to the UK records, the government construction strategy published in May
2011 states that Government will require fully collaborative 3D BIM (with all project
and asset information, documentation and data being electronic) as a minimum by
2016”. Also, in the reports published by McGraw-Hill in 2010 more than one-third of
western European construction industry participants (36%) reported that they started to
use BIM technologies by the year 2010 [1]. Thus, it is clear the BIM has gained a
significant place in the construction industry.
Despite all the advantages, BIM implementation faces some difficulties. Some of
them are ignorant about the added value of the BIM in the industry, lack of
Standards/Legislations, non-coherence with organization culture, high initial invest-
ment, technological deficiencies of the stakes holders [1]. Mainly, these issues are due
to a lack of BIM standards, lack of BIM awareness and resistance to change [4]. Thus,
if these problems are sorted out by an organization, BIM will be an ideal solution to
address sustainable issues at the designing stage.
ERP and Sustainability
ERP integrates the functions of an organization and provides a set of flexible best
practices to improve and optimize current business processes through re-engineering
the current business processes [13]. It helps with cost reduction, integration of
resources, sharing of efficient information, flexibility, agility and improved perfor-
mance [5]. Thus, it impacts the individual, workgroup, management and the whole
organization in a positive manner. Further, when considering the reduction of using
hard copies and the documentation level due to the ERP implementation and the
efficiency and productivity obtained through ERP implementation, it is very clear that
ERP can handle all three pillars of sustainability. (People, Environment, Profit).
On the other hand, in the current construction industry, ERP is upgraded as S-ERP
(Sustainable ERP) with sustainable features [9]. The concept of a sustainable ERP system
can be named as a holistic and integrative information system that is driven by sustain-
ability concerns that cover all aspects [10]. However, introduction of S-ERP system is in
the primary stage as per the current situation [5]. The indicators embedded in S-ERP
system under the economic category are financial health, performance, quality, cost, and
flexibility. Under the social category, internal human resources and under environmental
category air resources, land resources, mineral and material resources, energy resources,
waste impact and environmental commitments are considered [5].
Therefore, it is clear that the ERP system is capable of addressing sustainability
parameters when conducting the project. But the problem is, it is not capable of
manipulating sustainable design for the project. Thus, BIM should also be incorporated
in the design phase with ERP.
However, despite the benefits, ERP implementation shows some issues and chal-
lenges. The implementation can be complex and expensive due to the re-engineering.
Further, it needs organizational change which needs considerable time and top
206 F. S. T. Hewavitharana and A. A. D. A. J. Perera

management commitment towards the implementation. Also, infrastructure may be a

barrier for the ERP implementation [5]. If these challenges are successfully achieved
then ERP would be the best tool to address sustainability concepts in the operation
phase of the construction projects.
Integration of BIM and ERP
BIM is a concept that uses digitalized and parametric form design to build an integrated
module [7]. On the other hand, ERP is a concept which integrates all operations in to
one place [7]. Thus, at the design stage, BIM can create sustainable design while in the
operational phase ERP is capable of carrying out the operation properly [7]. Therefore,
the combination of this BIM and ERP will provide the best practices for the con-
struction organizations.
The integration of BIM and ERP should happen in three ways. Namely, Knowledge
integration, Data integration and Component integration. Knowledge integration is the
task of identifying how new and prior knowledge interacts while incorporating new
information into a knowledge base. Data integration involves the combining of data
from these two separate systems and provides a unified view of the data. Component
integration means the identification of the way of integrating hardware and software
according to the requirement.
Subsequently, customization should be involved in the integration of BIM and
ERP. According to the literature described by Chen and Tserng, combining BIM into
ERP is much easier than combining ERP into BIM [7]. This is because customization
of ERP can be easily incorporated into the BIM design data rather than incorporating
ERP concepts into BIM. This ERP customization can include changing application
parameters, (Sustainability parameters), adjusting workflow models (Data flow from
BIM to ERP) and choosing which modules will integrate [7].
The integration between BIM and various components of the ERP systems such as
estimates, controls and schedules can result in flexibility and availability of real-time
information required to execute complex and huge projects while maintaining integrity
between operations and back office automation systems. Thus, in order to establish a
cohesive time-sensitive information flow throughout their operation, many construction
organizations that utilize both BIM and ERP would help the two systems to control
project management tasks [9].
Sustainability via BIM-ERP Integration
It can be assumed that ERP and BIM integration will be a novel experience for
construction companies. While BIM allows constructing a building virtually inside a
computer, ERP will help in understanding how much it will cost and how to account
for it [14]. A BIM model includes every element along with geometric and parametric
information, provides an accurate quantity of materials and takes the guesswork out of
cost estimation. It will generate a bill of quantities (BOQ) automatically and produce
accurate material quantities [20]. On the other hand, ERP takes it from there and
enables the management of inventory, invoices and financial reporting efficiently with
lower cost and lower resources [15, 16]. Benefits of BIM-ERP integration will be
further explained in the sections below of the study.
The environment issues related to construction practices can be evaluated through
leadership in Energy and Environmental Design (LEED) which was introduced by
 Sustainability via ERP and BIM Integration 207

United States Green Building Council (USGBC). To fulfill those requirements, doc-
umentation will be required in energy modeling water use, daylighting, and sun
shading. For that, proper documentation should prevail. Correct data flow will ensure
that through ERP with related to the information generated by BIM.
Sustainability of the design can be incorporated into BIM and ERP integration. The
sustainability issues related to actual construction like carbon emissions, heating/
cooling loads, ventilation/air flow, the data could be compared related to the project
easily by using the project BIM model [17].
The integration of building systems like rainwater collection, usage of renewable
energy, utilization of sustainable materials, and energy modeling can be related easily
with the projects based on the schedules. That comparison of schedules can be easily
handled by ERP. If there is a variation in the schedule, BIM allows flexible modifica-
tions [17]. Findings of the literature review can be summarized as below (Figs. 1 and 2).

Fig. 1. Current situation of BIM and ERP on sustainability

Fig. 2. Future situation of combined BIM and ERP on sustainability

The rapid development of the construction industry creates a lot of sustainable

issues, now more than ever. Some of the main issues are; overutilization of energy and
resources, negative environmental impacts, excess use of money, non-adherence to
standards [1]. Thus, new technologies are concentrated on solving these issues. BIM
and ERP are two such systems discussed throughout this paper.
The fundamental low of true sustainability is the integration of all building systems
within themselves as well as with the external economic and environmental realities of
the project. To fulfill that goal, understanding the global sustainability issues and
addressing them in BIM is not the final solution if it is not possible to integrate with
another system like ERP [22].
In the present construction industry, BIM is developing sustainable designs at the
design phase but cannot carry out operations in a sustainable manner. Thus, to continue
208 F. S. T. Hewavitharana and A. A. D. A. J. Perera

the sustainable aspects, ERP can act as a bridge which connects sustainable design data
throughout the operations. Hence, the integration of these two will create great benefits
for the construction industry.
In the design stage, BIM helps with performing complex building performance
analyses while in the construction stage, ERP helps with integrating and streamlining
the business processes, cost reduction, and quality improvements. Hence, this inte-
gration allows addressing sustainability aspects in both phases.
BIM and ERP integration will ideally suit the delivery of information needed for the
improved design and building performance. The two most significant benefits of BIM–
ERP integration would be integrated project delivery and design, operation optimiza-
tion [14].
The three bottom lines of sustainability will be assured by the BIM–ERP integration
as follows.
Environment Aspect:
The integration will eliminate the extra cost of design change during the period,
enhancing the delivery culture in the future. Design optimization will reduce the use of
material and energy significantly so that energy, water and material, wastage due to
inefficiencies or mistakes in the project delivery processes can also be reduced [18, 19].
Social Aspect:
Integrated project delivery will reduce the risk in project delivery with collaboration
and proper communication. It will also improve the safety by solving the anticipated
problems earlier at the planning stage. The quality of sustainable design will improve
the living environment of the people, enhancing living standards. Also, this integration
will create a number of new job opportunities with the ERP and BIM integration [14].
Economic Aspect:
The undesirable wastage throughout the planning stage and operation stage can be
eliminated through BIM–ERP integration. Further, most of the operations through BIM
and ERP reduce much of paperwork. Moreover, optimum design which comes out of this
will reduce the capital and lifetime cost through the improved material and resources [2].
However, it is not easy to integrate ERP and BIM as their design architecture is
different from one another [17]. Therefore, the integration of these two software should be
carefully monitored. On the other hand, if ERP and BIM have not been integrated, their
performance will limit into one stage of construction. So, the integration of BIM and ERP
is a very much beneficial to facilitate sustainability practices throughout the project.
Potential Benefits of the Research
In the current construction industry, ERP and BIM are being used for construction
activities separately [17]. Though ERP has a proper way of integrating all the func-
tionalities across the organization, there is no proper way of addressing the building
models for the analysis. Hence, BIM needs to be used to analyse the building model
[19]. On the other hand, BIM has no power of integrating generated building structure
with other functionalities of the system. Therefore, this BIM-ERP integration will
combine these two processes. Further, addressing of sustainability issues in construc-
tion by an integrated system will be more efficient rather than having two separate
systems. Moreover, these findings will motivate construction organizations to
 Sustainability via ERP and BIM Integration 209

implement more efficient and productive systems to their organizations. It will lead to a
more sustainable construction industry in the near future.
Future Work
According to the literature review, it is established that there are huge benefits from the
integration of the ERP and BIM. Hence, it will be interesting to carry out future
research work on this literature review. So further research analysis should be carried
out under following objectives.
1. Identifying a method to incorporate a sustainable BIM design into an ERP system.
2. Listing out the sustainable variables that should be included in ERP to combine the
BIM design and ERP.
3. Establishing the measurements to justify the impact on sustainability through BIM
and ERP integration.

References
1. Aladag H, Demirdögen G, Isık Z (2016) Building information modeling (BIM) use in
Turkish construction industry. Proc Eng 161:174–179
2. Motawa I, Carter K (2013) Sustainable BIM-based evaluation of buildings. Proc - Soc Behav
Sci 74:419–428
3. Migilinskas D, Popov V, Juocevicius V, Ustinovichius L (2013) The benefits, obstacles and
problems of practical BIM implementation. Proc Eng 57:767–774
4. Mehran D (2016) Exploring the adoption of BIM in the UAE construction industry for AEC
firms. Proc Eng 145:1110–1118
5. Lotfy MAMB (2015) Sustainability of enterprise resource planning (ERP) benefits post
implementation: an individual user perspective (Ph.D). College of Management and
Technology
6. Ortiz O, Castells F, Sonnemann G (2009) Sustainability in the construction industry: a
review of recent developments based on LCA. Constr Build Mater 23(1):28–39
7. Chen Y, Tserng H (2017) An integrated methodology for construction BIM & ERP by
using UML tool. In: International symposium on automation and robotics in construction,
Taiwan, p 6
8. Chong H, Lee C, Wang X (2017) A mixed review of the adoption of building information
modelling (BIM) for sustainability. J Clean Prod 142:4114–4126
9. Chofreh A, Goni F, Shaharoun A, Ismail S, Klemeš J (2014) Sustainable enterprise resource
planning: imperatives and research directions. J Clean Prod 71:139–147
10. Hasan M, Ebrahim Z, Mahmood W, Rahman M (2017) Sustainable-ERP system: a
preliminary study on sustainability indicators. J Adv Manuf Technol 11:61–74
11. Wong K, Fan Q (2013) Building information modeling (BIM) for sustainable building
design. Facilities 31(3/4):138–157
12. Love P, Matthews J, Simpson I, Hill A, Olatunji O (2014) A benefits realization
management building information modeling framework for asset owners. Autom Constr
37:1–10
13. Gholamzadeh Chofreh A, Goni F, Ismail S, Mohamed Shaharoun A, Klemeš J,
Zeinalnezhad M (2016) A master plan for the implementation of sustainable enterprise
resource planning systems (part I): concept and methodology. J Clean Prod 136:176–182
210 F. S. T. Hewavitharana and A. A. D. A. J. Perera

14. Lindblad H, Vass S (2015) BIM implementation and organizational change: a case study of a
large swedish public client. Proc Econ Finan 21:178–184
15. Babič N, Podbreznik P, Rebolj D (2010) Integrating resource production and construction
using BIM. Autom Constr 19(5):539–543
16. Hessari P, Kirkegaard P, Zandieh M, Kani I (2016) Adoption of BIM systems in the AEC
industry. PONTE Int Sci Res J 72(9):123–133
17. Ghosh S, Negahban S, Hoon Kwak Y, Skibniewski M (2011) Impact of Sustainability on
Integration and Interoperability between BIM and ERP – a governance framework
18. Schwegler B (2010) Green BIM: successful sustainable design with building information
modeling by Eddy Krygiel and Bradley Nies. J Ind Ecol 14(5):859–860
19. Chen C, Chen S, Li S, Chiu H (2017) Green BIM-based building energy performance
analysis. Comput-Aided Des Appl 14(5):650–660
20. Moakher E (2012) Building information modeling (BIM) and sustainability – using design
technology in energy efficient modeling. IOSR J Mech Civ Eng 1(2):10–21
21. Saunders M, Lewis, P, Thornhill A (n.d.) Research methods for business students
22. Bonenberg W, Wei X (2015) Green BIM in sustainable infrastructure. Proc Manuf 3:1654–
1659
 A Comparative Study of Speed and Flow
Measurements Methods as Applied to Four
Lane Dual Carriageway Roads

W. U. A. Lowe(&), H. S. A. Mendis, and I. M. S. Sathyaprasad

Department of Civil Engineering, Faculty of Engineering,

University of Peradeniya, Peradeniya, Sri Lanka
[email protected]

Abstract. Flow-density-speed relationships play an important role in traffic flow

modelling and provides basis to illustrate traffic flow behaviour under different
traffic conditions. Analyses of this traffic condition critically depends on the
collected traffic data. Traffic data collection under mixed traffic conditions is one of
the major problems faced by researchers as well as the traffic regulatory authori-
ties. Therefore a comparative study on traffic data measurement techniques has
been carried out in this study. The traffic data can be collected using different kinds
of techniques. In this study, traffic flow and speed measurements have been col-
lected using a pneumatic tube detector method, video capturing method, moving
observer method and manual classified vehicle counting method. By comparing
these methods, suitable methods are suggested for application depending on the
existing conditions. This study concentrated on traffic data collection along four
lane dual carriageway roads and they are William Gopallawa Mawatha (AB42)
and Gannoruwa road (B365). Along these roads, study areas have been selected
such that traffic flow is uninterrupted. Tube detectors have been installed and data
has been analysed using Metrocount software. For other methods data have been
analysed manually. The flow measurements and speed measurements graphs have
been obtained for each direction in both four lane roads. These graphs have been
drawn such that they represent both forced flow conditions and free flow condi-
tions. Comparison has been done suing graphs obtained.

Keywords: Speed
Flow
Four lane dual carriageway roads

1 Introduction

The fundamentals of traffic flow and their characteristics have become more important
and form the foundation for all theories, techniques and procedures that are being
applied in the design, operation and development of road transportation systems. The
three most important characteristics of traffic are flow, speed and density. These
characteristics can be observed and studied at microscopic level and macroscopic level.
Flow is the number of vehicles passing a point on a road during stated period of time.
The scientific study of traffic flow had begun in the early 80’s with the study of models
relating volume and speed. Measurement of speed requires observation over both time
and space. There are two methods of calculating average traffic stream speed: a time

© Springer Nature Singapore Pte Ltd. 2020

R. Dissanayake and P. Mendis (Eds.): ICSBE 2018, LNCE 44, pp. 211–223, 2020.
https://doi.org/10.1007/978-981-13-9749-3_20
212 W. U. A. Lowe et al.

mean speed and a space mean speed. The space-mean speed reflects the average speed
over a spatial section of roadway, while the time-mean speed reflects the average speed
of the traffic passing a specified point over a specified time. In this study both speeds
will be calculated. There are a lot of methods available to collect traffic data. Each of
these methods have their pros and cons. Identifying these pros and cons is very
important for transportation engineers because it reveals which method is more suitable
depending on the nature of the study. The considered methods in this study are
pneumatic tube detector method, video capturing method, moving observer method and
manual classified vehicle count (CVC). Each of these techniques is briefly described in
following section. This study it is limited to four-lane dual carriageway roads in Kandy.
This type of roads consists of four lanes (two lanes per each direction) and they are
divided by a central reservation. Even though relatively high speeds can be observed in
four-lane dual carriageway roads, they are much safer when comparing with two-lane
dual carriageway roads. Head-on collisions, slide swipe collisions (in same direction)
are completely eliminated less in four-lane dual carriageway roads because of the
central reservation. Even though study areas are the same, observation points are quite
different in each method. In the study areas road stretches have been selected such that
traffic flow is uninterrupted. In Gannoruwa road (B365) stretch of 200.0 m is selected
and in William Gopallawa Mawatha (AB42), road stretch is 268.0 m. At the end of the
study traffic flow and speed measured using several methods have been compared.

1.1 Pneumatic Tube Detector Method

With the aid of the Road Development Authority (RDA), tube detectors have been
installed in the middle of each road stretch. In the Gannoruwa road it has been installed
100.0 m from the Kandy side edge and in the William Gopallawa Mawatha it is
134.0 m from the Kandy side edge. To observe the speed measurements two tube
detectors are required. Tubes are fixed at 1.0 m apart. Since the roads selected are four
lane roads, to get the traffic flow by lane, four tubes have been installed in one
direction. So, there were four data loggers and 8 tubes collecting data along each road
stretch. Data has been collected for a Period of one week (Fig. 1).

Fig. 1. Pneumatic tube detectors

 A Comparative Study of Speed and Flow Measurements Methods 213

1.2 Moving Observer Method

In this method an observer witnessing vehicle travel with traffic flow collects data. Two
observers were in the vehicle since data collecting has to be done covering both
through traffic and against traffic. When the vehicle entered into the road stretch the
time is recorded by one observer and same done as the vehicle leaving the road stretch.
While the vehicle moved with traffic one observer recorded the number of vehicles in
the against traffic (vehicle in the opposite direction) while the other one recorded the
number of vehicles over taken by the observer vehicle and number of vehicles over
take the observer vehicle. Once this process is done in one direction, to complete one
cycle the same procedure has to be repeated in the opposite direction too. For a
complete data set 8–10 cycles have been done for one road stretch at a time.

1.3 Video Capturing Method

In this method videos have been captured covering the selected road stretches. All the
cameras have been mounted at the edges of each road stretch. Four cameras have been
used per one direction which had been synchronized earlier. This is because, the speed
measurement travel time of each vehicle is required and to do so each vehicle should be
identified when it is entering into the road stretch and when it is leaving. Two cameras
captured the inner lane flow while the other two captured the outer lane flow. For each
camera, a reference line has been marked 5 m away from the point the cameras are
mounted. When a vehicle crosses the reference line, the time, vehicle type and the
registration plate number are recorded. The same vehicle is identified from the other
camera which is mounted at the other end of the road stretch and time difference is
calculated.

1.4 Classified Vehicle Count

In Sri Lanka, the RDA uses 15 classes of vehicle for Classified Volume Count surveys,
it is shown in Table 1. CVC of the traffic flow is done manually with the aid data
loggers, which involves a group of people recording the number of vehicles passing on
a pre-determined location. These locations are installed 400 m away from the tube
installed location. It automatically gives CVC in 15 min intervals.

Table 1. Vehicle classification system used by RDA

Type of vehicle Class Description
Motor cycle MCL All type of motor cycles and scooters
Three wheelers TW All type of three wheelers
Car CAR All type of cars, Jeeps, cabs, dual cabs, micro van and van
with three rows of seats
Van VAN Van with seats more than three rows
Buses MBUS Buses with lower 26 seats
LBU Buses with more than 26 seats
(continued)
214 W. U. A. Lowe et al.

Table 1. (continued)
Type of vehicle Class Description
Light goods LGV Small lorries with 2 axels and 4 tyres
vehicle
Medium goods MG1 Small lorries with 2 axel and 6 tyres
vehicle MG2 Large lorries with 2 axel and 6 tyres
High goods HG3 Large lorries with 1 chacy with 3 axel
vehicles
Articulated AG3 3 axels with tailor
vehicle AG4 4 axels with tailor
AG5 5 axels with tailor
AG6 6 axels with tailor
Farm vehicle FVH Tractor with 4 tyres with tailor, Tractor with 2 tyres with tailor

2 Literature Review
2.1 Introduction
There are many researches available on traffic flow analyses. But only some of them
have been carried out for mixed traffic condition. Mixed traffic condition consists of
different kinds of vehicle categories. For an example car, vans, buses, trucks, three
wheelers, bikes and etc. Since in Sri Lanka also has a mixed traffic condition, the focus
of this section goes to the literature connected to mixed traffic conditions. Traffic flow
analysis can be used to simulate traffic. For example, to evaluate a new part of the
infrastructure, traffic flow models can be used. Also, traffic flow analysis can be used to
improve road safety.

2.2 Traffic Flow and Speed Measurement

There are a lot of methods available to collect traffic data. But when it comes to
heterogeneous traffic automatic collection methods seem to be more suitable. Most of
the time, video capturing method has been utilized. Sharma et al. (2011) used video
capturing method. In their study, they had considered the heterogeneous mix vehicles
along an undivided two-lane road facility. This study of traffic had not been influenced
by any nearby intersections or road geometry (Gradient). Therefore, this traffic is
considered as an uninterrupted flow of traffic. Also, verbal recordings of vehicle cat-
egories and license plate number were done. Certain key features were noted down so
that the identification of vehicles could be done. But some researchers have used
software to do that. Mallikarjuna and Rao (2011) have used this method to collect
traffic flow data under heterogeneous conditions. They have used video data to extract
speed, flow, density, lateral gap and several other parameters. There can be many other
lateral movements but, in this study, it is restricted to five movements. They have also
used a software named TRAZER to track vehicles and get their trajectory data over a
road length of 50 m. This software is capable of classifying vehicles capturing the
lateral movement of vehicles. Traffic speeds had been obtained using the trajectory
 A Comparative Study of Speed and Flow Measurements Methods 215

data. Also, Mallikarjuna et al. (2009) have used the same method to collect traffic data
under mixed traffic conditions. In their study the same software has been used. The
velocities and acceleration values were obtained by performing first and second order
differentiation on trajectory equations. In their study they have proved that TRAZER is
capable of detecting vehicles with an average accuracy of 95% and classifying with an
average accuracy of 85%. Using the software for traffic flow detection seems to be easy
and more useful in gathering other information too. But Sharma et al. (2011) have used
FMV6 (full motion video) player for video data extraction. This software is capable of
fragmenting the video into frames and uses that to obtain frame readings when traffic
vehicle just crosses the reference point. Ranthatiyawa et al. (2017) have also utilized
the video capturing method. But in that study, roads selected were residential roads.
They have used the video data for traffic flow counting and have manually counted the
traffic flow using the videos. For their study not using software was possible since they
only need the traffic flow.
Other than the video capturing method, the loop detector method has also been used
to collect traffic data. Stathopoulos and Karlaftis (2003) have used the loop detector
method. Speeds and travel times along selected routes have been obtained using flow
and occupancy data collected from the loop detectors. They had used these data to
develop ITS (intelligent transportation system). For this, six origin points had been
defined at major city entrances, from which a total of 17 possible routes had been
considered. Traffic data along these 17 routes had been collected directly from the
controllers installed at 144 locations of the down-town urban network. Also, Hueper
et al. (n. d) have used traffic flow and occupancy data from loop detectors. They have
used the collected data to calibrate their model. In their study, they had considered a
Californian freeway. They have also used PeMS database to achieve traffic measure-
ments like occupancy, flow, speed, delay, vehicle hours travelled and productivity loss
etc. The only direct measurements taken from the loop detections are the number of
cars which cross a detector station and the fraction of time a vehicle is present over the
loop. These researches used loop detector method because in their context it was more
convenient. And also, for a mixed traffic condition it is a very effective method.
Therefore, choosing video capturing method seemed to be a wise choice.

3 Methodology

3.1 Site Selection

Firstly, suitable sites have been selected. Few major aspects were paid attention to
when selecting the sites. First, was to avoid any junctions along the road section.
Suitable lengths for the road sections have also been selected to get accurate data. The
road stretches have been selected such that they do not consist of central openings,
pedestrian crossings and bus stops. The locations for tube detector installation were
selected such that it is in the middle of the road stretch. The places to video camera
mounting were selected such that they had the capability to capture the vehicle reg-
istration number clearly. On the Gannoruwa road, a stretch of 200 m is selected and on
the William Gopallawa Mawatha, the road stretch is 268 m (Figs. 2 and 3).
216 W. U. A. Lowe et al.

Fig. 2. Survey location at Gannoruwa Road

Fig. 3. Survey location at William Gopallawa Mawatha

3.2 Field Data Collection

In the video survey method, the distance between two camera locations and widths of
the road sections were measured using a road tracer.
In the tube detector method, tubes have been installed for a week with aid of the
RDA. In the same week, Moving observer method also was carried out. Manual CVC
was done on the same day. But the video capturing was done in a different day. In
manual CVC it was done from 6.00 am to 22.00 pm. In the video capturing method the
videos have been captured for two hours (7.00 am–9.00 am) so that it covered the peak
hours. Peak hours were firstly identified using tube detector data.

3.3 Data Extraction and Analysis

The data extraction was done only for the video capturing method because in all the
other methods recorded data can be directly retrieved in an Excel file format. In the
video capturing method, form each video data had to be fed into a computer database
with parameters such as vehicle registration plate number, time stamp and vehicle type
when the vehicle passed the reference point. In the tube detector method, recorded data
was converted into an Excel file format using Metrocount software. This software is
used for tube data extraction and analysis. In manual CVC also, recorded data was
downloaded from data loggers. It was directly collected as an Excel file format.
Analysis of the data collected is different in each method and it is briefly discussed in
following sections.
 A Comparative Study of Speed and Flow Measurements Methods 217

3.3.1 Pneumatic Tube Detector Method

In the tube detector method, data analysis was done using Metrocount software. From
each data file speed and flow is analysed along each road for each direction. To obtain
lane wise traffic flow, traffic flow obtained for inner lane was deducted from the traffic
flow obtained for both lanes. From the Metrocount software data files are converted in
to excel files as it directly gives the flow and speed data separately. Time interval which
data are required can be changed accordingly. Classification systems are also can be
changed. Classification systems in the Metrocount software are quite different and none
of them comply with RDA classification system. They are ARX +, Scheme F +,
Scheme F2 +, Scheme F99 +, AustRoads94 +, NAASRA +, Swedish +, TNZ1999 +,
Scheme axle +. Under each of these classification systems flow data has been plotted
and compared with the RDA classification system. Also, flow variation in the week has
been plotted along with lane distribution, flow variation in city-bound and out-bound
directions. Traffic flow and speed data have been calculated and analysed for both lanes
for each direction.

3.3.2 Video Capturing Method

In the video capturing method, the speed of each vehicle has been calculated using time
measurement extracted from the videos and length between the specified points.

L
V¼ ð1Þ
t1  t2

Where,
V: Speed of the vehicle
L: Distance between two survey points
t1 : Time of entering of the vehicle
t2 : Time of leaving of the vehicle
Since flow measurements are recorded according to the RDA classification, they are
directly taken from the extracted data. After obtaining the flow data, they were con-
verted into passenger car units by using the relevant passenger car equivalency factors.
Flow, speed calculation and analysis have been carried out only in William Gopallawa
Mawatha towards Kandy direction. Plots have been developed to represent the speed-
flow variation.

3.3.3 Moving Observer Method

In the moving observer method, the following equations have been used. When
observer is moving with the stream, let m0 be the vehicles overtake the observer and mp
be the vehicles overtaken by the observer. Let the difference be m. Then,

m ¼ m0  mp ð2Þ

m ¼ ðq  tÞ  ðk  v0  tÞ ð3Þ
218 W. U. A. Lowe et al.

Here m; t and v0 can be obtained from the survey. q can be obtained using below
equation.

mw þ ma
q¼ ð4Þ
tw þ ta
Where,
mw : Vehicle difference when observer vehicle is going with the traffic flow
ma : Vehicle difference when observer vehicle is going against the traffic flow
tw : Time taken to travel road stretch with the traffic flow
ta : Time taken to travel road stretch against the traffic flow
Mean stream speed vs can be obtained using below equation.

l
vs ¼ mw ð5Þ
tw  q

Where,
vs : mean stream speed
Flow and speed calculations have been carried out for both lanes for both direc-
tions. Plots have been developed to represent flow-speed variation.

4 Results and Discussion

Figure 4 shows the traffic flow variation along William Gopallawa Mawatha for the
city bound direction with different data collection methods. The methods considered
are pneumatic tube detector method, moving observer method and manual CVC.

Fig. 4. Traffic flow comparison using different methods

According to Fig. 4, during peak hours manual count shows an over estimation
compared to tube detector data while moving observer method shows an under
 A Comparative Study of Speed and Flow Measurements Methods 219

estimation. However, during low flow conditions it is quite different. Moving observer
method shows over estimation compared both other methods. Even though tube data
show an underestimation during low flow condition, it almost complies with the
manual count data. Here the reason for the over estimation could be traffic congestion.
Since the tube detector method is an automated method, when it detects vehicles during
congestion, errors could occur due to vehicles over lapping. Because vehicles move
closely to each other during congestion, tube detectors could detect them as one
vehicle. In the moving observer method, these flows are calculated using 8-10 cycles.
But the cycle time is less than one minute since the road stretch is small. The rec-
ommended time period to take a flow measurement mentioned in highway manual is
15 min. So, the results obtained using moving observer method could differ from the
manual count data.
Figure 5 shows the traffic speed variation along William Gopallawa Mawatha in the
city bound direction. The data represented here are collected using pneumatic tube
detector method and moving observer method.

Fig. 5. Traffic flow comparison using different method

According to Fig. 5, in the moving observer method speed variation is between 15–
35 kmph while it is in tube detector method 35–55 kmph. Even though speed variation
during the day doesn’t show any significant difference, the moving observer method
always shows an underestimation of speed. The reason could be the same as mentioned
above.
Figure 6 shows the shows the comparison of speed measurements using video
capturing method and tube detector method in the William Gopallawa Mawatha in the
city bound direction. As clearly seen in the figure tube detector method underestimates
the speed measurements. This is mainly because of the survey location. Since tube
detector was located in the middle of the road stretch, it gives the spot speed at the
middle of the road stretch. But in the video method speed measurements have been
calculated considering the whole road stretch and therefore it gives an average value for
the speed along the road stretch.
220 W. U. A. Lowe et al.

Fig. 6. Speed comparison between video and tube method

Fig. 7. Flow comparison between video and tube method

The Fig. 7 shows the flow measurement comparison using video method and tube
detector method in the William Gopallawa Mawatha in the outbound direction. So, the
figure shows zigzag variation pattern between the two methods. Since the video has
been taken from 7.00 a.m. to 9.00 a.m., it includes the peak hour. So, during the peak
condition, tube data gives an underestimation most of the time. The reason could be the
same as mentioned above.
Figure 8 plot shows the lane wise traffic flow variation along William Gopallawa
Mawatha. This data has been obtained using pneumatic tube detector method. Since
 A Comparative Study of Speed and Flow Measurements Methods 221

other methods cannot be used to collect lane distribution, here, explanation is done only
considering tube detector method.

Fig. 8. Lane distribution of traffic flow

A clear difference of lane usage can be obtained in the outbound. Whether it is low
traffic condition or high, it is always greater in the inner lane. The main reason can be
the commercial buildings located along outer lane side towards outer bound. Since the
vehicles that come to these buildings use the outer lane, through traffic allure to use the
inner lane. To the city bound it varies. During peak hours inner lane usage is higher
while during off peak hours it is the other way around.

Fig. 9. Flow variation at William Gopallawa Mawatha in city Bound direction

Since other methods hadn’t been carried out for a week, weekly flow variation is
described here only using pneumatic tube detector data. Graph in Fig. 9 has been
plotted showing the 24-h flow variation representing week days as well as weekends.
222 W. U. A. Lowe et al.

Following graph shows flow variation along the William Gopallawa Mawatha towards
the city (Table 2).

Table 2. Peak hour details using tube detector method

City-Bound Out-Bound
Day Peak hour Peak hour Total daily Peak hour Peak hour Total daily
volume (vph) volume volume (vph) volume
Thu 7–8 a.m. 1,330 17,526 17–18 p.m. 1,476 19,066
Fri 8–9 a.m. 1,401 18,368 17–18 p.m. 1,413 19,253
Sat 8–9 a.m. 1,434 17,687 15–16 p.m. 1,290 18,610
Sun 8–9 a.m. 1,124 15,539 17–18 p.m. 1,253 16,600
Mon 7–8 a.m. 1,390 18,187 17–18 p.m. 1,504 19,611
Tue 8–9 a.m. 1,352 17,679 - - -

Flow variation during weekdays does not show any significant difference. But flow
on Sunday is quite lower than other days. It doesn’t mean weekend flow is lower than
weekdays. Saturday flow’s variation is more similar to the weekdays flow. Peak hour,
during weekdays is from 8.00 a.m. to 9.00 a.m. except on Mondays, where it is from
7.00 a.m. to 8.00 a.m. During the weekend it is from 7.00 a.m. to 8.00 a.m. During the
low flow condition, flow variations of all days are very similar.
According to Fig. 10, in the outbound direction, flow variation is quite different
from it is for the city bound. In the city bound weekends flow is lower than weekdays
flow. The major change is in the peak hour. Even though the morning peak hour
remains the same for the weekdays, it varies on the weekend. The out bound peak hour
is in the evening and is from 5.00 p.m. to 6.00 p.m. except for Saturday. On Saturday it
is from 3.00 p.m. 4.00 p.m. This is mainly because the city consists of a lot of
commercial buildings and schools. Therefore, during the morning flow it is higher in
the city bound direction while it is higher in the afternoon for the outbound.

Fig. 10. Flow variation at William Gopallawa Mawatha - Outbound direction

 A Comparative Study of Speed and Flow Measurements Methods 223

So, in conclusion, the video capturing method and Manual counting method seem
more accurate than the tube detectors because its shows different results when it comes
congestion condition. Also, for a short road stretch, the moving observer method does
not seem suitable.

Acknowledgements. The authors acknowledge the continuing support from the Road Devel-
opment Authority and the assistance by the staff of the surveying laboratory.

References
Arasan T, Gnanavelu D (2009) Concentration of heterogeneous road traffic. In: Jayanthaku-
maran K (ed) Advance technologies. InTech, London, pp 503–530
Gerlough DL, Huber MJ (1975) Traffic flow theory. Transportation Research Board,
Washingtion
Mallikarjuna C, Rao KR (2011) Heterogeneous traffic flow modelling: complete methodology.
Transportmetrica 7:321–345
Bains MS, Ponnu B, Arkatkar SS (2012) Modeling of traffic flow on Indian expressway using
simulation technique, Changsha, s.n
Ranthatiyawa RDMDR, Perera BVLS, Sathayaprasad IMS, Nandalal HK (2017) Use of remote-
sensed spatial data in estimating traffic flows from small residential road catchments, Kandy,
s.n
Sharma N, Arkatkar SS, Sarkar AK (2011) Study on heterogeneous traffic flow characteristics of
a two-lane road. Transport 26(2):185–196
Stathopoulos A, Karlaftis MG (2003) A multivariate state space approach for urban traffic flow
modeling and prediction. Transp Res Part C 11:121–135
Transportation Reasearch Board (1995) Highway capacity manual. National Research Council,
Washington, D.C
Green Technology and Energy
 A Stochastic Differential Equation Based Wind
Speed Forecasting Model

H. M. D. P. Bandarathilake1(&) and G. W. R. M. R. Palamakumbura2

1
Sri Lanka Technological Campus, Ingiriya Road, Padukka, Sri Lanka
[email protected]
2
Department of Engineering Mathematics, Faculty of Engineering,
University of Peradeniya, Peradeniya, Sri Lanka

Abstract. Stochastic differential equation (SDE) based approaches have

demonstrated an improved capability in predicting temporal wind speed patterns
that have statistical properties which are very similar to those observed in reality.
However, no standard approach for deriving such models exist. One difficulty in
arriving at a unified framework is the presence of a seemingly wide variation in
the temporal statistical properties that one observes from location to location. In
this work, we propose an SDE model for short term wind speed pattern pre-
dictions for Kokkilai, an area that is located in the North-Eastern coastal region
of Sri Lanka. Here, we propose a systematic approach for modelling short term
temporal wind speed variations, using a 1-dimensional stationary stochastic
differential equation along with an approximated marginal distribution for the
long term empirical wind speed distribution and the corresponding autocorre-
lation function. The model parameters were estimated to fit the empirical wind
speed distribution of wind speed data that has been recorded at the wind mea-
surement center of Kokkilai, from February 2015 to February 2016. The normal
Gaussian distribution turned out to be the best fitted marginal distribution, in
terms of maximum likelihood, for long term wind speed data. Then, an
Ornstein-Uhlenbeck based SDE was obtained using the parameters of the fitted
normal Gaussian distribution and the empirical autocorrelation function. Sub-
sequently, considering special features such as cyclic behavior and certain trends
of the empirical autocorrelation function, a set of new parameters were intro-
duced to the drift and diffusion terms of the SDE. The obtained model was
capable of generating a short-term wind speed forecasting trajectories that fall
within an acceptable confidence level.

Keywords: Stochastic differential equation
Drift
Diffusion

Autocorrelation function

1 Introduction

During the last few years, the contribution of wind based electricity generation to the
global green energy industry has shown a remarkable increase. Wind is an energy
source that is inherently volatile and uncertain in nature. Therefore, an efficient wind
speed forecasting mechanism developed by means of mathematical/statistical mod-
elling, can play a pivotal role in the analysis of many aspects related to power systems

© Springer Nature Singapore Pte Ltd. 2020

R. Dissanayake and P. Mendis (Eds.): ICSBE 2018, LNCE 44, pp. 227–238, 2020.
https://doi.org/10.1007/978-981-13-9749-3_21
228 H. M. D. P. Bandarathilake and G. W. R. M. R. Palamakumbura

such as power generation economics, evaluation of the generation capacity and

ensuring reliability. Time series based methods, probability distribution based models,
four component composite models and Kalman filter approaches (Carta et al. 2009) are
some of the mathematical/statistical approaches that have been used to forecast wind
speeds in both metrology and the energy industry (Zhu and Genton 2012). However,
the development of an accurate and efficient forecasting modelling still remains a real
challenge.
In comparison with traditional modelling approaches of wind speed, stochastic
differential equations based methods have gained popularity in the field of wind power
forecasting during the last few years. The novel stochastic differential equations
(SDE) based approaches have indicated an increased capability of sufficiently and
accurately capturing the inherent stochastic nature of wind.
Due to the topographical dependence of wind flow patterns, the wind speed always
indicates high variations from one location to another location. Therefore, most studies
have only concentrated on studying the wind speed forecasting for a given location.
Two major types of wind speed forecasting exist, namely, probabilistic forecasting and
point-forecasting.
In the case of probabilistic forecasting, wind speed distributions are generated while
in point-forecasting wind speed predictions are explicitly provided for a certain time
period ahead. Point-forecasting models can be also split into models that generate long
term forecasting trajectories and short-term forecasting trajectories.
In (Zarate-Minano et al. 2016) (Zárate-Miñano and Milano 2016) (Calif 2012), a
one dimensional stationary stochastic differential equation has been employed to obtain
wind speed distributions. In these cases, the empirical wind speed distribution and the
corresponding autocorrelation function, of a moderately large wind speed data set, are
used to obtain the forecasting trajectories. According to the method employed in
(Zarate-Minano et al. 2016) (Zárate-Miñano and Milano 2016), the empirical wind
speed distribution has been approximated by a known standard probability distribution
and the exponentially decaying region of the empirical autocorrelation function, is used
to model the forecasting trajectories using the principles of Ito calculus as pointed out
in (Zárate-Miñano and Milano 2016). In (Calif 2012), the forecasted wind speed tra-
jectories have been presented as wind speed density functions. In (Zarate-Minano et al.
2016) where the marginal distribution was selected as a Weibull distribution, the model
provides probabilistic forecasting of long term wind speed variation together with
short-term autocorrelation forecasting. The work presented in (Calif 2012 and Zarate-
Minano et al. 2016), has indicated a situation where the empirical wind speed distri-
bution was approximated by a Gaussian mixture distribution. In addition to the above
cases, (Iversen et al. 2015) describes a model that refines a given numerical wind speed
prediction using a one dimensional SDE. In (Iversen et al. 2015), the developed SDE,
has generated accurate wind speed forecasting by modifying an SDE which generates
an Ornstein-Uhlenbeck process model by considering the cyclic variation of wind
speed patterns. Consequently, it provides a more accurate short-term wind speed
forecasting model in comparison with the numerical wind speed predictions.
In this paper, we present a short-term wind speed forecasting model which is
obtained by modifying an SDE model based on an approximated marginal distribution
and the corresponding autocorrelation function.
 A Stochastic Differential Equation Based Wind Speed Forecasting Model 229

In this work, we have considered the wind speed data that has been recorded at
Kokkilai, in the North-Eastern coast of Sri Lanka, from February, 2015 to February,
2016. The data set was recorded under a feasibility assessment of the “Quantum Leap
Wind Power Development” project funded by the Asian Development Bank (http://
www.energy.gov.lk/ta/node/380).

1.1 Methodology
A differential equation which consists of one or more stochastic terms is identified as a
stochastic differential equation. A one dimensional non-stationary Ito stochastic dif-
ferential equation has the following form,

d ðxðtÞÞ ¼ aðxðtÞ; tÞdt þ bðxðtÞ; tÞdWðtÞ ð1Þ

Where aðxðtÞ; tÞ and bðxðtÞ; tÞ terms are called drift and diffusion coefficients
respectively. Here, WðtÞ denotes the standard Wiener process (Øksendal 2003). The
integral form of Eq. (1) takes the form,

Zt Zt
xðtÞ  x0 ¼ aðxðuÞ; uÞdu þ bðxðuÞ; uÞdW ðuÞ ð2Þ
0 0

In the above Eq. (2), the first integral which is associated with the drift term can be
solved by means of ordinary Riemann-Stieltjes integration. The second integral, which
is termed the diffusion term, should be integrated using stochastic calculus (Øksendal
2003) (Higham and Higham 2001).
The wind speed data {vt : t ¼ 1; 2; 3; . . .} can be regarded as one realization of an
underlying stochastic process {Vt : t ¼ 1; 2; 3; . . .}. This says, for each time point t, vt
is treated as an observation of the random variable Vt . Such process can consist of a
deterministic part and a random component. Thus stochastic differential equations are a
natural choice for modeling wind speed data.
Provided that the stationarity conditions are met, statistics such as the autocorre-
lation function are not only useful as a descriptive tool of the process but can also can
be used to make inferences about the underlying process of which the observed series is
a sample (Brett and Tuller 1991). In this context, in addition to the selection of model
for forecasting, analytical methods based on the empirical autocorrelation function are
also employed to identify important periodicities in the underlying process.
In this work, we are interested in reproducing the wind speed values with statistical
properties similar to that of the empirical data, by means of a SDE of the form of
Eq. (1). Here, we have assumed that the considered wind speed variation preserves the
stationarity property. The stationary form of the Eq. (1) is given by

d ðxðtÞÞ ¼ aðxðtÞÞdt þ bðxðtÞÞdWðtÞ ð3Þ

and will be employed to obtain the forecasting trajectories.

In order to solve the stationary SDE in Eq. (3), the corresponding drift and diffu-
sion terms should be estimated. In the following, we present two approaches for doing
230 H. M. D. P. Bandarathilake and G. W. R. M. R. Palamakumbura

this. The first method is based on the estimation of the marginal probability distribution
function for the process while in the second method we explicitly account for diurnal
and other periodic effects on wind speed.
Method 1
Following the approach presented in (Zárate-Miñano and Milano 2016), the solutions
of the SDE are assumed to be distributed according to a known probability distribution
function. Then, by applying the Ito’s lemma to the stationary SDE, the corresponding
stationary Fokker-Planck equation associated with the considered SDE can be obtained
as follows,

1 @  2 
0 ¼ aðxðtÞÞpðxðtÞÞ þ b ðxðtÞÞpðxðtÞÞ ð4Þ
2 @xðtÞ

where pðxðtÞÞ is the probability distribution function of the stochastic process that is
generated by the SDE in Eq. (3). Since the wind speed distribution of the model
generated values should be the same as that of the empirical wind speed distribution,
the best fitted marginal distribution to the empirical wind speed distribution will be
selected for pðxðtÞÞ.
Then, the Fokker-Planck equation in Eq. (4), can be solved for aðxðtÞÞ and bðxðtÞÞ
in terms of this estimate as follows,

@bðxðtÞÞ 1 2 @
að xð t Þ Þ ¼ bð xð t Þ Þ þ b ð xð t Þ Þ ðlog pðxðtÞÞÞ ð5Þ
@xðtÞ 2 @xðtÞ

2 xðtÞZ
b2 ð x ð t Þ Þ ¼ aðzðtÞÞpðzðtÞÞdzðtÞ ð6Þ
pðxðtÞÞ 1

where pðxðtÞÞ 6¼ 0 and bðxðtÞÞ ¼ 0 if pðxðtÞÞ ¼ 0:

In order to require the desired drift term, we consider the regression theorem as in
(Zárate-Miñano and Milano 2016) Consequently, the drift term is obtained in the form:

aðxðtÞÞ ¼ aðxðtÞ  lÞ ð7Þ

where l is the mean of the considered probability distribution pðxðtÞÞ and in order to
obtain the value of a, the exponentially decaying part of the autocorrelation function of
the wind speed data is fitted with the function of the form, (here a and c are constants)

f ð xÞ ¼ ceax ð8Þ

It should be noted that the type SDE in Eq. (3), always generates a stochastic
process with exponentially decaying autocorrelation function. Therefore, the parameter
a plays a crucial role in the evaluation of the drift and the diffusion terms as shown in
(Zárate-Miñano and Milano 2016). Then using the obtained drift term, diffusion term is
 A Stochastic Differential Equation Based Wind Speed Forecasting Model 231

calculated using Eq. (6). In our case, the empirical distribution is best approximated by
a normal Gaussian distribution. So, taking

1 ðx  lÞ2
pðxðtÞÞ ¼ pffiffiffiffiffiffiffiffiffiffi expð Þ ð9Þ
2pr2 2r2

we have that the drift and diffusion terms must take the form,

aðxÞ ¼ aðx  lÞ ð10Þ

pffiffiffiffiffi
bð x Þ ¼ 2ar ð11Þ

Therefore, the above drift and diffusion terms generate an Ornstein-Uhlenbeck

process of the form,
pffiffiffiffiffi
dx ¼ aðx  lÞdt þ 2ardW ð12Þ

The parameters l; r and a are estimated so that Eq. (9) best fits the marginal
probability density of the empirical data and Eq. (8) fits the initial exponentially
decaying part of the empirical autocorrelation function. Using these parameters, the
solutions of Eq. (12) can be used to provide a point wise estimate of wind speeds.
Notice that the autocorrelation of the so modeled process is necessarily exponentially
decaying and cannot handle periodic variation as observed in reality.
Method 2
In (Iversen et al. 2015), they have incorporated cyclic variation of wind pattern by
modifying the drift term of an SDE model based on numerical weather prediction
which generates an Ornstein-Uhlenbeck process. The purpose of short-term model in
(Iversen et al. 2015), was to refine the numerical wind speed prediction which also
provides forecasts for short periods ahead. In this paper, we introduce a short-term
forecasting model which can be validated for several selected portions of data based on
the different features exhibiting in the autocorrelation function. After incorporating
these in the drift and the diffusion terms of Eq. (12), we propose the following model
for the selected portions of the series.

 p   
dX ðtÞ ¼ b sin tþu 1  eXðtÞ þ aðl  XðtÞÞ dt þ rX c ðtÞdWðtÞ ð13Þ
T
  
Here, bSin Tp t þ u term has been used to capture the sinusoidal behavior of
 
wind speed variation and 1  eX ðtÞ can avoid generating negative values. In addition
to that, value of T should be chosen to capture the cyclic variation of the considered
region. The term rX c replaced the constant value of the diffusion term of Eq. (12) in
order to capture the trends of the resultant process.
232 H. M. D. P. Bandarathilake and G. W. R. M. R. Palamakumbura

2 Results and Discussion

2.1 Estimation of the Parameters for the SDE Model Proposed in Method
1
Wind speed data which had been recorded as ten-minute average values, were con-
verted to hourly mean wind speed values and the obtained dataset of hourly average
wind speed consisted of 8419 observations. Resulting empirical wind speed distribu-
tion is shown in Fig. 1. In order to find the best fitted marginal distribution for the
empirical wind speed distribution, the maximum log likelihood values of each marginal
distribution was obtained as in Table 1. Figure 2, indicates, graphically, some
approximations of the marginal distributions for the empirical wind speed distributions.

Fig. 1. Empirical wind speed distribution of Kokkilai

Fig. 2. Approximation of empirical density function of hourly average wind speed of Kokkilai
 A Stochastic Differential Equation Based Wind Speed Forecasting Model 233

Table 1. Log likelihood values

Log-likelihood values Probability distribution
Logistic Extreme value t-Location scale Normal
Log-likelihood −19173.7 −19522.9 −18977.8 −18977.8

According to the Table 1 and Fig. 3, the normal Gaussian distribution was selected
as the most suitable marginal distribution with the parameter values, l ¼ 5:68398 and
r ¼ 2:30547.
The value of the parameter a was obtained by the empirical autocorrelation func-
tion. For this case, the exponentially decaying region of the empirical autocorrelation
function should be identified. The empirical autocorrelation function of Kokkilai which
was constructed by considering the entire period, is given in the Fig. 3. It indicates that
the autocorrelation coefficient decays exponentially only for a 12-hour time interval
from the beginning and then it indicates a complex cyclic behavior.

Fig. 3. Autocorrelation function of hourly wind speed of Kokkilai for entire time domain

Using the method of least squares, the parameter c and a in Eq. (8) were estimated
within a 95% confidence interval and estimated values were found to be c ¼ 0:6678
and a ¼ 0:2204 (Fig. 4).
234 H. M. D. P. Bandarathilake and G. W. R. M. R. Palamakumbura

Fig. 4. Curve fitting of autocorrelation function with 12 lags

Finally, the SDE model of Eq. (12) was simulated with obtained parameter values
by using the Euler-Maruyama numerical scheme with the initial value xð0Þ ¼ v0 ¼
5:428 and the wind speed forecast for the entire period was obtained. The wind speed
distribution of the model generated values for several simulations are compared with
the empirical wind speed distribution in Fig. 5.

Fig. 5. Comparison of original & initial SDE model generated wind speed distributions obtained
by kernel density estimation of Kokkilai

Figure 5 indicates that the model generated wind speed values and original wind
speed data have high deviation in the median velocity values. So, it is obvious that even
though the model has been capable of generating a wind speed distribution that has a
similar shape to that of the original data, it is incapable of generating a long term
prediction of wind speed with considerable accuracy.
 A Stochastic Differential Equation Based Wind Speed Forecasting Model 235

2.2 Estimation of the Parameters for the SDE Model Proposed

in Method 2
The autocorrelation function in Fig. 3, indicates a complex cyclic behavior accompa-
nied with a short period of exponential decay. The periods of the cyclic patterns
indicated in the autocorrelation vary with the time. Therefore, by observing these
patterns in the autocorrelation function, the wind speed data was grouped and labeled
as portion-11, portion-12, portion-21 and portion-22 accordingly. In this work, portions
11, 21 and 22 were chosen for forecasting since they exhibit sinusoidal behaviors with
almost constant periods. For the three cases the model (13) was to estimate the
parameters of for each respective portion (Fig. 6).

Fig. 6. Selected portions of the main dataset for incorporation of short-term models

In constructing the models for the three cases, the proposed model (13) was
implemented. The parameters were estimated by implementing the pseudo-maximum
likelihood estimation method. With estimated parameters, the SDE model was simu-
lated using the Euler-Maruyama numerical scheme with 100 simulations and mean
forecasting trajectory was obtained by taking the mean of all simulated trajectories for
all three cases. These are shown in Figs. 7, 8 and 9 respectively. The estimated
parameter values for the respective portions are in Tables 2, 3 and 4.
The corresponding parameter values of the SDE model, estimated for the portion-22
and portion-21 which have been indicated in Tables 3 and 4, display much similarity. It
shows us that the wind speed data of a total of 64 h including the data of portion-21,
consisted of almost constant periodic variation. Therefore, it has ensured a comparatively
longer period of time in generating a mean forecasting trajectory within the accepted
confidence levels. In addition, it should be observed that the corresponding parameter
values which are involved with the diffusion term have taken similar values. According to
the figures, the mean forecasting trajectory of portion 21, can be identified as the best.
236 H. M. D. P. Bandarathilake and G. W. R. M. R. Palamakumbura

Fig. 7. Mean forecasting trajectory of portion-11 (within a considered confidence interval)

Table 2. Parameter values of short-term model of portion-11

Parameter value Parameter
T a b c r u l
Value 19 0.2204 0.6500 0.4230 0.4210 1.0472 4.9500

Fig. 8. Mean forecasting trajectory of portion 22 (within a considered confidence interval)

 A Stochastic Differential Equation Based Wind Speed Forecasting Model 237

Fig. 9. Mean forecasting trajectory of portion 21 (within a considered confidence interval)

Table 3. Parameter values of short-term model of portion-22

Parameter value Parameter
T a b c r u l
Value 12 0.2204 0.4900 0.4230 0.4210 1.0472 3.5200

Table 4. Parameter values of short-term model of portion-21

Parameter value Parameter
T a b c r u l
Value 12 0.2204 0.3900 0.4230 0.4210 1.0472 3.5200

2.3 Conclusion
In this work, we present a short-term wind speed forecasting model which is capable of
generating a mean forecasting trajectory that agrees well with the empirical data for a
considerable period of time (32 h and 64 h) into the future. Considering the cyclic
variation and the diffusion process of the selected portions of data, a new set of
parameters were initially introduced to the SDE which generates a well-known
Ornstein-Uhlenbeck process. In addition to the new set of parameters, an exponential
expression was also added to the drift term of the SDE to avoid the generation of
negative values that can be observed as a drawback in previous SDE models on wind
speed forecasting. Then, using a pseudo-maximum likelihood estimation method the
parameter values were estimated for the considered portion of data. In all three portion
of data, the corresponding region of autocorrelation function has exhibited a sinusoidal
variation of almost constant period. Constant cyclic variation in autocorrelation
238 H. M. D. P. Bandarathilake and G. W. R. M. R. Palamakumbura

function confirms the cyclic variation in wind speed data in respective regions. Results
of the all three cases, demonstrate that the mean forecasting trajectory coincides very
well with the original wind speed values upto 10 h ahead. Of these three cases, the
mean forecasting trajectory obtained for portion 21, has been totally confined to the
confidence levels because it is the region where constant periodic behavior has been
greatly indicated.

References
Brett AC, Tuller SE (1991) The autocorrelation of hourly wind speed observations. J Appl
Meteorol 30:823–833
Calif R (2012) PDF models and synthetic model for the wind speed fluctuations based on the
resolution of Langevin equation. Appl Energy 99:173–182
Carta JA, Ramirez P, Velazquez S (2009) A review of wind speed probability distributions used
in wind energy analysis. Renew Sustain Energy Rev 13(5):933–955
Higham DJ, Higham DJ (2001) An algorithmic introduction to numerical simulation of stochastic
differential equations. SIAM Rev 43(3):525–546
Iversen EB et al (2015) Short-term probabilistic forecasting of wind speed using stochastic
differential equations. Int J Forecast 32(3):981–990. https://doi.org/10.1016/j.ijforecast.2015.
03.001
Øksendal B (2003) Stochastic differential equations: an introduction with applications, New
York. https://doi.org/10.1109/TAC.2006.882767
Zarate-Minano R, Mele FM, Milano, F (2016) SDE-based wind speed models with Weibull
distribution and exponential autocorrelation. In: IEEE power and energy society general
meeting. https://doi.org/10.1109/PESGM.2016.7741754
Zárate-Miñano R, Milano F (2016) Construction of SDE-based wind speed models with
exponentially decaying autocorrelation. Renew Energy 94:186–196. https://doi.org/10.1016/j.
renene.2016.03.026
Zhu X, Genton MG (2012) Short-term wind speed forecasting for power system operations. Int
Stat Rev 80:2–23
 Experimental Investigation to Achieve
Minimum Surface Roughness in Wire
EDM Process

K. P. M. Priyankara(&) and G. I. P. Perera

Department of Mechanical and Manufacturing Engineering,

Faculty of Engineering, University of Ruhuna, Matara, Sri Lanka
[email protected]

Abstract. This paper presents the behaviors and evaluation of Wire Electrical
Discharge Machining (Wire EDM) manufacturing process. Wire EDM is also
known as Wire-cut EDM, which uses a dielectric fluid submerged thin metallic
wire to cut the conductive work piece material. Taguchi orthogonal array was used
for planning the experiments to optimize the control variables for better surface
roughness while machining AISI 420 modified metallic material. Pulse-on/off
Time, Wire Feed, Spark Voltage and Peak Current were selected as the control
variables. AISI 420 metal, also called as Stavax®ESR has a higher metal strength
and is used for all types of moulds and moulding tools in the industry. The
experiments were carried out by changing the aforesaid control variable param-
eters and TR200 handheld roughness tester was used to measure the surface
roughness of metal cuts. Three surface roughness measurements were taken for
each machined section and averaged. Plotting method and Analysis of Variance
(ANOVA) were used for the analysis and interpretation of the results. Error
variances were estimated by using Pooling-up strategy. The influences of each
control parameter on surface roughness were estimated. Then, the results were
used to develop a mathematical model. The developed mathematical model can be
used to minimize surface roughness. Control variable parameter settings for the
optimized surface roughness can be obtained by using the mathematical model.

Keywords: Wire EDM
ANOVA
Surface roughness
Optimization

1 Introduction

Wire electric discharge machine (WEDM) is well known for its ability to machine
complex features with tough metals which normally cannot be easily machined in
transitional machines. Thus, EDM is an ideal process for producing cutting tools, dies,
and moulds for injection moulding, etc. (Chen and Cox 2017) since using the WEDM
for the above applications may give better surface finish for the work material. But to
achieve a better surface finish for the work a better selection of parameter control
should be included. Then, the selection of optimum machining parameters in WEDM is
an important step. Improperly selected parameters may result in serious problem on the
work surface damage which imposes certain limits on the product quality (Muthu
Kumar and Suresh Babu 2010). It involves the removal of metal by discharging an

© Springer Nature Singapore Pte Ltd. 2020

R. Dissanayake and P. Mendis (Eds.): ICSBE 2018, LNCE 44, pp. 239–251, 2020.
https://doi.org/10.1007/978-981-13-9749-3_22
240 K. P. M. Priyankara and G. I. P. Perera

electrical current from a pulsating Direct Current (DC) power supply across a thin inter-
electrode gap between the tool and the work piece. It is most commonly used for
machining hard and difficult to machine materials with very close tolerances. Gener-
ally, WEDM is perceived to be an extremely actuate process and there are various
reasons for this perception. Firstly, in WEDM, no direct contact takes place between
the cutting tool (electrode) and the work piece; as a result, the adverse effects such as
mechanical stresses, chatter, and vibrations normally present in traditional machining
are eliminated. Secondly, the wire used as a cutting tool has high mechanical properties
and small diameter 0.076 to 0.30 mm that produces very fine, precise, clean cuts.
Finally, in WEDM, the movements of the work piece during cutting are controlled by a
highly accurate Computer Numerical Control (CNC) system (Kumar et al. 2013).
A schematic diagram of WEDM is shown in Fig. 1. Here, the work material is
vaporized to perform a thin cut though the workpiece.

Fig. 1. Schematic diagram of WEDM (Goyal 2017)

2 Literature Review

To predict the Surface roughness and Kerf width while machining Aluminum and mild
steel, (Tilekar and Das 2014) developed an empirical model. The analysis of variance
resulted that spark on time has major influence on the surface roughness in the case of
aluminum. On the other hand, input current has major influence in case of mild steel. In
the case of Kerf width, wire feed rate and spark on time have significant effect on
aluminum and mild steel respectively. (Lodhi and Agarwal 2014) made an investiga-
tion to study the Pulse-on time, Pulse-off time, wire feed and Peak current on Surface
roughness. Lower amounts of surface roughness shows the high productivity of
Wire EDM. (Babu et al. 2017) studied the optimization of the trim cutting operation of
WEDM of AISI A3 tool steel for given machining conditions by the Taguchi technique
coupled with Grey relational analysis to optimize the input variables of WEDM on
Material Removal Rate (MRR), Surface Roughness (SR) and Gap Current (GC) of
ballistic grade aluminum alloy. Results confirm that Ton, peak current and spark
voltage are significant variables to the Grey relational grade.
 Experimental Investigation to Achieve Minimum Surface Roughness 241

(Prajapati et al. 2013) made an attempt to determine the important machining

parameters for performance measures like SR, MRR, GC and kerf separately in the
WEDM process. Response surface methodology experimental design method is used to
obtain optimum parameter combination for maximization of MRR, SR as well as
minimization of kerf while conditioning GC. The conclusion is made that cutting rate
and surface roughness, the pulse on and pulse off time is most significant. The spark
gape set voltage is significant for kerf. (Pasam et al. 2010) proposed, and investigated
the Optimizing Surface Finish in WEDM using the Taguchi Parameter Design Method.
In this work, an attempt is made to determine the optimal parameters for surface finish
under varying conditions using the Taguchi parameter design process in the WEDM of
Ti6Al4V alloy. This process was applied using a specific set of control parameters,
ignition pulse current, short pulse duration, time between two pulses, servo speed,
servo reference voltage, injection pressure, wire speed and wire tension with a response
variable of surface roughness. In this research it was observed that increase in spark
time meant discharge will remain for a longer time and leads to high cutting speed and
high discharge energy. This improves the MRR and worsens the surface finish. Hence,
small spark time gives better surface finish. To achieve better surface finish lower
values of delay time should be selected.
(Azhiri et al. 2014) proposed a new approach, for the application of Taguchi,
ANFIS and grey relational analysis for studying, modeling and optimization of wire
EDM process while using gaseous media. The machining parameters, namely; pulse on
time, pulse off time, wire feed rate, and wire tension and voltage are optimized with
consideration of performance characteristics SR and Cutting Velocity (CV) improving
the machining efficiency. Here, to correlate the relationship between process inputs and
responses, adaptive neuro-fuzzy inference system has been utilized. In the end, a grey
relational analysis has been used to maximize CV and minimize SR simultaneously.
Results indicated that oxygen gas and brass wire guarantee superior cutting velocity.
Also, according to Analysis of Variance (ANOVA), pulse on time and current were
found to have significant effect on CV and SR.
(Bobbili et al. 2013) used the Experimental results showing that both approaches
can be optimized in machining parameters effectively and input variables of WEDM on
MRR and SR of high strength armour steel have been studied. Design of Experiments
(DOEs) and regression techniques have been adopted to optimize the response char-
acteristics. The conclusions are as Pulse-on Time (Ton), Pulse-off Time (Toff), and
Servo Voltage (SV) are significant variables to the MRR and SR of the material. Also,
optimum process parameter combinations for better MRR and SR are achieved by
employing ANOVA. This technique is cost effective and time saving in evaluating the
machining parameters without conducting a large number of experiments. Kumar et al.
(2013) developed optimization of process parameters based on response surface
methodology for pure titanium using WEDM process. The experimental plan is based
on the Box–Behnken design. In this experiment, six set parameters, i.e., pulse on time,
pulse off time, peak current, spark gap voltage, wire feed and wire tension have been
varied to investigate their effect on output responses. Then concluded that SR and MRR
are highly depend on pulse on time. Also, the result says that the dimensional deviation
depends on moderate input current.
242 K. P. M. Priyankara and G. I. P. Perera

(Kumar and Ravikumar 2013) conducted an experiment on modeling and opti-

mization of the wire EDM process by optimizing the parameters of the (WEDM)
process. In the article, an attempt was made to determine the significant machining
parameters for performance measures surface roughness in the WEDM process. Factors
like speed, feed, Time on and Time off have been found to play a significant role for
MRR and surface roughness. Muthu Kumar and Suresh Babu (2010) proposed a new
approach, for optimization of the WEDM parameters on machining incoloy800 super
alloy with multiple quality characteristics. In this paper, an application of combined
Taguchi Method and Grey Relational Analysis made to improve the multi-response
characteristics of MRR, Surface roughness in the Wire-Cut EDM of Inconel 800 has
been reported. While applying the Grey-Taguchi method, The Material Removal Rate
shows an increased value of 0.05351 g/min to 0.05765 g/min, the Surface Roughness
shows a reduced value of 3.31 lm to 3.10 lm and the Kerf width shows a reduced
value of 0.324 to 0.296 mm respectively, which are positive indicators of efficiency in
the machining process.
(Maher et al. 2015) stated that the effect of process parameters on SR, Heat Affected
Zone (HAZ) and CS in WEDM varying pulse on time, pulse off time and wire tension
to reveal their impact on optimization parameters using Adaptive Neuro-Fuzzy Infer-
ence System (ANFIS). This study concludes that the peak current and pulse on time are
the most significant parameters affecting the cutting speed, surface roughness and heat
affected zone. The wire tension has minor effect on the cutting speed and heat affected
zone, but it has great effect on the surface roughness. (Goyal 2017) identified signif-
icance of input variables of WEDM on MRR and SR of Inconel 625 super alloy. DOEs
and regression techniques have been adopted to optimize the responses characteristics.
The main conclusions are that current and pulse on time are the most influential
parameters that affect the Material removal rate and surface roughness. Also, cryogenic
tool electrode provides better machining performance (maximum MRR and better
surface roughness) when compared to the normal tool electrode.
(Bobbili et al. 2015) performed modelling and analysis of material removal rate and
surface roughness in wire-cut EDM of armour materials which used Buckingham pi
theorem to model the input variables of WEDM on MRR and Ra of Al 7017 and
Rolled Homogeneous Armour (RHA) steel alloy. The article says that the dependence
of thermo-physical parameters on the mechanism of MRR and Ra has been interpreted,
and owing to the thermal properties of WEDM, MRR of machining is higher for
materials with a low melting temperature and specific heat, the rise in pulse-on time
from 0.85 ms to 1.25 ms causes improvement in MRR and deterioration of surface
finish and the wear rate of brass wire increases with increase in input energy, leading to
wire breakage.
(Kubade et al. 2015) Parametric Study and Optimization of WEDM Parameters for
Titanium diboride TiB2. The effect of pulse-on time, pulse-off time and wire feed are
investigated in machining of TiB2 by keeping other process parameters constant. The
experiment gave results that the pulse-on time and pulse-off time are most significant
whereas wire feed rate is insignificant for material removal rate. Also the research
proposes a suitable set of variable combinations or better output. (Babu et al. 2017)
conducted an experiment on multi response optimization of wire-EDM process
parameters of ballastic grade aluminum alloy based on Taguchi method for Inconel 625
 Experimental Investigation to Achieve Minimum Surface Roughness 243

super alloy. Results confirm that TON, IP and SV are significant variables. Mathe-
matical models were developed using response surface method for MRR, SR and IG to
determine the relation between machine variables and performance measures. Optimum
response characteristics such as MRR, SR and GC are improved with 6% error. To
predict the optimization of process parameters for surface roughness of Inconel 625 in
Wire EDM, using Taguchi and ANOVA method is noted.
(Babu and Soni 2017) developed the empirical model by taking different process
parameters like pulse on time, pulse off time and peak current for the surface rough-
ness. The results of the experiments say that surface roughness increases with increase
on pulse on time and decrease with increase in pulse off time. For surface roughness,
the most significant factor is pulse on time followed by pulse off time and peak current
is an insignificant factor for surface roughness. Taguchi-based optimization of surface
roughness and dimensional accuracy in wire EDM process with S7 heat treated steel is
new research which has been done by (Chen and Cox 2017). This research focuses on
the use of the Taguchi method to reduce the surface roughness and improve dimen-
sional accuracy by optimizing the pulse on time, pulse off time, servo voltage, voltage,
servo feed, wire tension, and wire speed. The researcher has given sets of parameters
for better optimization of output while proposing that this systematic Taguchi approach
can be further implemented to more complex manufacturing processes such as 3D
printing or injection moulding processes (Table 1).

Table 1. Conclusion of literature review

Reference Metal Used methods Expt. variables Control
variables
(Tilekar and Das Aluminum and Taguchi Ton, Toff, Peak SR, Kerf
2014) mild steel method Current (I), Wire Feed Width (KW)
(Fw)
(Lodhi and AISI D3 steel Taguchi Ton, Toff, I, Spark SR
Agarwal 2014) method Voltage (Fw)
(Babu et al. 2017) Ballistic grade Taguchi Ton, Toff, I, V SR, Gap
aluminum method Current (GC),
MRR
(Prajapati et al. AISI A3 tool
Response Ton, Toff, V, Fw, Sw SR, MRR,
2013) steel surface KW, GC
methodology
(Pasam et al. 2010) Titanium alloy Taguchi Ton, Toff, Tg, I, V, SR
method Fw, Pd
(Azhiri et al. 2014) Al/SI metal Taguchi Ton, Toff, V, Sw, Fw SR, CV
matrix method
(Bobbili et al. Nickel based Taguchi Ton, Toff, Fw, Pf, V, SR, MRR
2013) super alloy method Sw
(Kumar et al. 2013) Pure titanium Response Ton, Toff, I, V, Fw, SR, MRR,
surface Sw DD
methodology
(continued)
244 K. P. M. Priyankara and G. I. P. Perera

Table 1. (continued)
Reference Metal Used methods Expt. variables Control
variables
(Kumar and Al-SiC Taguchi Ton, Toff, Fw MRR, SR
Ravikumar 2013) method
(Muthu Kumar and Inconel 800 Taguchi Ton, Toff, Fw, V MRR, SR,
Suresh Babu 2010) super alloy method KW
(Maher et al. 2015) AISI 1050 ANFIS Ton, Toff, Sw SR, CS, HAZ
Carbone steel
(Goyal 2017) Inconel 625 Taguchi Ton, Toff, I, Fw, Sw SR, MRR
super alloy method,
ANOVA
(Bobbili et al. Al, RHA Buckingham pi Ton, I*V, Pf, S, H MRR, SR
2015) theorem
(Kubade et al. Titanium Taguchi Ton, Toff, Fw MRR, SR, OC
2015) Dibromide method
(Babu et al. 2017) Inconel 625 Taguchi Ton, Toff, I, V SR, GC, MRR
super alloy method
(Babu and Soni S7 heat treated Taguchi Ton, Toff, I SR
2017) steel method
(Chen and Cox Ballistic grade Taguchi Ton, Toff, Fw, Sw, V SR, DD
2017) aluminum method
alloy

3 Materials and Method

The machines, equipment and materials which are used for the experiment may be
discussed here in this chapter with the method.

3.1 Material List of the Experiment

• Wire EDM machine - BMXP wire cut EDM

• Surface roughness tester – Time TR200
• Workpiece material - Stavax ESR (AISI 420 modified stainless tool steel)
• Wire material – Copper

3.2 Methods, Levels and Experiment Arrangement

Method: Here, the experiment is arranged by considering Taguchi’s orthogonal array
method and the experiment outcomes are analysed by ANOVA (Analysis of variance)
method. Error variances were estimated by using Pooling-up strategy.
The five variables of the experiments are put into three levels to arrange the
experiments. Each parameter has three levels except voltage. Voltage has two levels in
which the levels are limited by machine capability (Table 2).
 Experimental Investigation to Achieve Minimum Surface Roughness 245

Table 2. Levels of variable parameters

No. Process parameter Unit Level 1 Level 2 Level 3
1 V V Low - High
2 Ton µs 48 55 62
3 Toff µs 2 5 8
4 Fw mm/min 1 2 3
5 I A 2 4 6

The assigned levels of parameters are used to plan the experiment setups by using
L18 Taguchi’s orthogonal array. Then, there are 18 experiments conducted according
to the Table 3 arrangement of variable parameters.

Table 3. Experimental arrangement

Expt. no Parameter
V (H/L) Ton (µs) Toff (µs) Fw (mm/min) I (A)
1 L 48 2 1 2
2 L 48 5 2 4
3 L 48 8 3 6
4 L 55 2 2 6
5 L 55 5 3 2
6 L 55 8 1 4
7 L 62 2 3 4
8 L 62 5 1 6
9 L 62 8 2 2
10 H 48 2 1 2
11 H 48 5 2 4
12 H 48 8 3 6
13 H 55 2 2 6
14 H 55 5 3 2
15 H 55 8 1 4
16 H 62 2 3 4
17 H 62 5 1 6
18 H 62 8 2 2

4 Results and Discussion

The results of the experiments and results of the analysis are presented in this section.
The analysed results are presented by using plotted graphs to get a better understanding
on variations of controlled parameters. A mathematical model is also developed for
surface roughness by using results. Then, the discussion of the experiment result is
further discussed.
246 K. P. M. Priyankara and G. I. P. Perera

4.1 Experiment Outcomes

Three roughness values are taken for each experiment set to obtain an accurate result.
Then, the mean value of measurements is taken as the roughness value which considers
an analysis of the experiment outcomes. We can clearly identify the variations of
outputs as follows in the Table 4.

Table 4. Experimental results

Sample no. Measurements (µm) Average (µm)
No. 01 No. 02 No. 03
1 5.953 5.927 6.086 5.989
2 5.803 5.403 5.865 5.690
3 6.474 6.249 6.192 6.305
4 5.824 5.894 6.018 6.088
5 5.173 5.453 5.301 5.309
6 6.952 7.070 6.850 6.957
7 5.811 5.744 5.800 5.785
8 6.287 6.293 7.118 6.566
9 6.328 6.289 6.240 6.286
10 5.082 5.357 5.762 5.400
11 6.937 6.999 6.839 6.880
12 7.225 6.467 6.838 6.843
13 6.525 6.181 6.428 6.378
14 5.321 5.647 5.261 5.410
15 7.217 7.465 6.975 7.219
16 5.348 5.566 5.440 5.451
17 7.068 6.381 7.419 6.956
18 6.211 5.779 6.205 6.065

The results are analysed, and the contribution factors of each variable is generated
(Table 5).

Table 5. Contribution factors of parameters to SR

Parameter Contribution factor (%)
V 2.8
Ton -
Toff 34.0
Fw 25.4
I 37.7
 Experimental Investigation to Achieve Minimum Surface Roughness 247

The roughness values of experiment outcomes are plotted with respect to levels of
each process parameter in identifying the variation of surface finish considering the
changes of process parameter values. Figure 2 shows that the surface roughness
increases when spark voltage is shift to the higher values. Figure 3 indicates that the
variation of surface roughness for three levels of pulse-on time is quite similar and have
a smooth peak at the middle level. There is dramatic increment of surface roughness
with respect to the increment of pulse-off time according to Fig. 4. Figure 5 shows a
strong negative relationship between surface roughness and the wire feed. The rela-
tionship between peak current and surface roughness is strong and positive according
to Fig. 6.

Spark votage (P1)

8
ROUGHNESS(µm)

0
0 1 2 3 4
LEVELS

Fig. 2. Levels of voltage Vs SR

Pulse-on time(P2)

6.23
ROUGHNESS(µm)

6.22
6.21
6.2
6.19
6.18
0 1 2 3 4
LEVELS

Fig. 3. Levels of pulse-on time Vs SR

248 K. P. M. Priyankara and G. I. P. Perera

Pulse-off time (P3)

ROUGHNESS(µm)
6.8
6.6
6.4
6.2
6
5.8
0 1 2 3 4
LEVELS

Fig. 4. Levels of pulse-off time Vs SR

Fig. 5. Levels of wire feed Vs SR

Peak current (P5)

6.6
ROUGHNESS(µm)

6.4
6.2
6
5.8
5.6
0 1 2 3 4
LEVELS

Fig. 6. Levels of peak current Vs SR

 Experimental Investigation to Achieve Minimum Surface Roughness 249

The following figure (Fig. 7) shows the change of roughness value with different
parameters while the levels of the parameters are in consideration.
The graph shows that the roughness values of the five process parameters are
almost close to 6 µm while the individual roughness value for different levels of the
same input parameter interpret some variations. The variation of surface roughness is
negligible when considering the variable parameter as Pulse-on time.

Fig. 7. Factor plot – surface roughness

4.2 Mathematical Model

SR=-61.1852+117.4868V-34.0843V^2+0.0214Toff+0.0106 Toff ^2-
ð1Þ
0.1373Fw-0.0487 Fw ^2+0.5897I-0.0494I^2

The above equation shows the mathematical model of surface roughness. However,
it needs to be verified. Thus, if any errors occur at the verification, the mathematical
model will have to be modified to minimize the error.

5 Conclusion

The Taguchi orthogonal array was used for planning the experiments to optimize the
control variables for better surface roughness while machining AISI 420 modified
metallic material. Pulse-on/off Time, Wire Feed, Spark Voltage and Peak Current were
selected as the control variables. AISI 420 metal, also called as Stavax®ESR has a
higher metal strength and is used for all types of moulds and moulding tools in the
industry. The experiments were carried out by changing the aforesaid control variable
parameters and TR200 handheld roughness tester was used to measure the surface
roughness of the metal cuts. Three surface roughness measurements were taken for
each machined section and averaged.
250 K. P. M. Priyankara and G. I. P. Perera

Plotting method and Analysis of Variance (ANOVA) were used for the analysis
and interpretation of the results. Error variances were estimated by using Pooling-up
strategy. The influences of each control parameter on surface roughness were esti-
mated. Then, the results were used to develop a mathematical model. The developed
mathematical model can be used to minimize surface roughness. Control variable
parameter settings for the optimized surface roughness can be obtained.

Acknowledgments. We would like to thank Department of Mechanical and Manufacturing

Engineering, Faculty of Engineering, University of Ruhuna to providing necessary facilities for
this research work. Finally, we would give our greater gratitude to the free education system of
Sri Lanka.

References
Azhiri RB, Teimouri R, Baboly MG, Leseman Z (2014) Application of Taguchi, ANFIS and
grey relational analysis for studying, modeling and optimization of wire EDM process while
using gaseous media. Int J Adv Manuf Technol 71:279–295
Babu TV, Kumar SA, Soni JS (2017) Multi response optimization of wire-EDM process
parameters of aluminum alloy 6063 using Taguchi method. Int J Curr Eng Technol 7
(3):1040–1045
Babu TV, Soni JS (2017) Optimization of process parameters for surface roughness of Inconel
625 in wire EDM by using Taguchi and ANOVA method. Int J Curr Eng Technol 7(3):1127–
1131
Bobbili R et al (2013) Effect of wire-EDM machining parameters on surface roughness and
material removal rate of high strength armor steel. Mater Manuf Process 28:364–368 6914
(August 2017)
Bobbili R, Madhu V, Gogia AK (2015) Modelling and analysis of material removal rate and
surface roughness in wire-cut EDM of armour materials. Eng Sci Technol Int J 18(4):664–
668. https://doi.org/10.1016/j.jestch.2015.03.014
Chen JC, Cox J (2017) Taguchi-based optimization of surface roughness and dimensional
accuracy in wire EDM process with S7 heat treated steel. World Acad Sci Eng Technol Int J
Chem Mol Nucl Mater Metall Eng 11(8):551–556
Goyal A (2017) Investigation of material removal rate and surface roughness during wire
electrical discharge machining (WEDM) of Inconel 625 super alloy by cryogenic treated tool
electrode. J King Saud Univ - Sci 0–7. http://dx.doi.org/10.1016/j.jksus.2017.06.005
Kubade PR et al (2015) Parametric study and optimization of WEDM parameters for titanium. Int
Res J Eng Technol 2:1657–1661
Kumar A, Kumar V, Kumar J (2013) Multi-response optimization of process parameters based
on response surface methodology for pure titanium using WEDM process. Int J Adv Manuf
Technol 68:2645–2646
Kumar K, Ravikumar R (2013) Modeling and optimization of wire EDM process. Int J Mod Eng
Res 3:1645–1648
Lodhi BK, Agarwal S (2014) Optimization of machining parameters in WEDM of AISI D3 steel
using Taguchi technique. Proc CIRP 14:194–199. https://doi.org/10.1016/j.procir.2014.03.080
Maher I et al (2015) ScienceDirect improve wire EDM performance at different machining
parameters – ANFIS modeling. IFAC-PapersOnLine 48(1):105–110. https://doi.org/10.1016/
j.ifacol.2015.05.109
 Experimental Investigation to Achieve Minimum Surface Roughness 251

Muthu Kumar V, Suresh Babu A, Venkatasamy R, Raajenthiren M (2010) Optimization of the

WEDM parameters on machining Incoloy800 super alloy with multiple quality character-
istics. Optimization 2(6):1538–1547
Pasam VK, Madar Valli P, Swapna M (2010) Optimizing surface finish in WEDM using the
Taguchi parameter design method. J Braz Soc Mech Sci Eng XXXII(2):107–113
Prajapati SB et al (2013) Effect of process parameters on performance measures of wire EDM for
AISI A2 tool steel. Int J Comput Eng Res 3:274–278
Tilekar S, Das SS, Patowari PK (2014) Process parameter optimization of Wire EDM on
aluminum and mild steel by using Taguchi method. Proc Mater Sci 5:2577–2584
 Eco-Friendly Porous Carbon Materials
for Wastewater Treatment

C. J. Thambiliyagodage1(&), V. Y. Cooray2, I. N. Perera2,

and R. D. Wijesekera2
1
Faculty of Humanities and Sciences, Sri Lanka Institute of Information
Technology, New Kandy Road, Malabe, Sri Lanka
[email protected]
2
Department of Chemistry, Faculty of Science, University of Colombo,
Colombo 03, Sri Lanka

Abstract. Different porous carbon materials were applied to remove organic

compounds and heavy metals from wastewater. Hierarchically, porous nitrogen
rich monolithic carbon (HPC) was synthesized through the sol-gel synthesis
method and carbonization process. Rice husk (RH), an agricultural waste
material followed by carbonization at two different temperatures (RH400 and
RH600) was used as the naturally generated porous carbon composite. Activated
carbon (AC) was used as the control. According to the FT-IR spectra, different
functional groups are present in all adsorbents. Scanning electron microscopic
(SEM) images show an irregular shaped, random disordered macropore struc-
ture in RH and a robust sponge-like disordered macropore structure with liga-
ments in HPC. Transmission electron microscopic (TEM) images of these
materials show a disordered mesopore network. Adsorption capacities of these
porous carbon materials were determined for organic dyes and metal ions.
Methylene blue and Methyl orange were used as the model organic compounds
and Pb2+, Cu2+, Ni2+, and Cd2+ ions were selected as the metal ions. The
experimental data demonstrate that the adsorption capacity of AC for Methyl
orange (5.88 mg/g) is much higher than that of the alternatives (RH400–
1.97 mg/g, RH600–0.69 mg/g, and HPC – 1.14 mg/g). Adsorption capacity of
all the adsorbents for Methylene blue is quite similar. Adsorption capacity of
RH400 for Pb2+ is much higher than that of RH600, AC and HPC. RH400 has a
greater adsorption for Cu2+ and Ni2+ than the other adsorbents. All adsorbents
show a low adsorption capacity for Cd2+. RH400 is a promising adsorbent for
wastewater purification.

Keywords: Rice husk
Hierarchically porous carbon
Organic compounds

Metal ions
Adsorbance

1 Introduction

Organic pollutants (Li et al. 2006) and heavy metals (Jaishankar et al. 2014) in
wastewater are toxic and non–biodegradable environmental hazards that can affect
human health. The volume of wastewater generated by the industrial parks of Sri Lanka
has been estimated as 30 million cubic meters per year but the extent of wastewater

© Springer Nature Singapore Pte Ltd. 2020

R. Dissanayake and P. Mendis (Eds.): ICSBE 2018, LNCE 44, pp. 252–260, 2020.
https://doi.org/10.1007/978-981-13-9749-3_23
 Eco-Friendly Porous Carbon Materials for Wastewater Treatment 253

used for agricultural activities has not been determined as yet (Senarath and Vis-
vanathan 2001). The chronic kidney disease of unknown etiology (CKDU) which is
prevalent in the North-central area of the country has become one of the major health
issues related to water pollution of agricultural wastewaters (Janakan et al. 2009).
Organic pollutants harm ecosystems by causing acute and chronic health effects, and by
destroying aquatic life etc. (Jayasumana et al. 2014). Hence, it is necessary to remove
these toxic materials from wastewater. Various methods have been used to remove
pollutants, such as precipitation (Kim et al. 2002), membrane filtration (Nasirabadi
et al. 2016), oxidation (Ebrahiem et al. 2017) and adsorption (Sophia and Lima 2018).
Adsorption technology is rapidly gaining prominence, and activated carbon (AC) is the
most popular adsorbent due to its porous structure, high specific area and high
adsorption capacity (Dias et al. 2007). However, the high cost of commercially
available AC (*US $90/kg) has limited its usage. The main objective of the present
study is to investigate potential alternatives to AC. Rice husk was selected as the
agricultural waste due to its abundance at low cost and the presence of silica in addition
to carbon after carbonization. In addition to making use of an agricultural waste,
synthesis of hierarchically porous monoliths was also of interest to investigate the effect
of organized interconnected porosity on adsorption.

2 Materials and Reagents

Rice husk was collected from a rice mill at Thalawathugoda, Sri Lanka. Methylene
Blue (Basic Blue 9) and Methyl Orange were purchased from Eurostar Ltd, UK,
Copper (II) nitrate trihydrate, A.R was purchased from Himedia Laboratories, India.
Lead acetate was purchased from British Drug Houses, England. Cadmium acetate was
purchased from Hopkin and Williams, UK. Ethyl alcohol (95%), formaldehyde (34.5
wt%), o-phenylenediamine (99.5%) and polyethylene glycol (PEG, Mol. Wt: 6000)
were purchased from Sigma-Aldrich. Resorcinol recryst. extra pure was purchased
from Merk Limited Mumbai.

3 Material Characterization

FT-IR spectra of the samples were collected by a Varian 600 FT-IR spectrometer.
For FT-IR studies, samples and KBr were mixed in the ratio of 1:10 and ground to
make a pellet. The morphology of the samples was analyzed by Hitachi SU6600
Analytical Variable Pressure FE-SEM and Oxford instruments EDX with AZtec
software scanning electron microscope operated at 10 kV. Transmission electron
microscope images of the samples were collected with a EOL - JEM – 2100 High
resolution transmission electron microscope. The acceleration voltage was 200 kV.
Samples were prepared on a lacey carbon film supported on a Cu grid. The
adsorption capacity of the adsorbents on metal ions was investigated using an Atomic
Absorption Spectrometer (GBS 932 plus AAS). The adsorption of organic dyes was
measured on a GENESYS 10S UV-VIS spectrophotometer.
254 C. J. Thambiliyagodage et al.

4 Methodology

Rice husk (RH) was washed with distilled water and ethanol followed by drying at 75 °
C. The sample was carbonized separately at 400 °C and 600 °C for 2 h in air (RH400
and RH600 respectively). Hierarchically porous carbon (HPC) was prepared by dis-
solving resorcinol (3.0 g) and polyethylene glycol 6000 (1.25 g) in a solvent mixture of
ethanol (11.4 mL) and distilled water (9 mL) with magnetic stirring at 50 °C. After-
ward, o-phenylenediamine (0.1 g) was added to the above solution and stirred for
30 min at 50 °C. Formalin containing formaldehyde (4.42 g, 35 wt%) was quickly
added to the solution. The reaction mixture was stirred at 50 °C for another 10 min.
The homogenous emulsion was sealed and transferred to an oven and heated at 90 °C
for 4 h. It was dried at 50 °C for 24 h followed by carbonization at 600 °C for 2 h in an
oxygen restricted environment. Both RH and HPC were separately crushed to a fine
powder and sieved through a 500 lm mesh.
Batch adsorption studies were carried out by stirring 20 mL of dye solution
(9.46 mg/L Methylene blue and 3.10 mg/L Methyl orange, separately) and 40 mL of
metal ion solution (6.05 mg/L Cu2+, 6.88 mg/L Pb2+, 1.20 mg/L Cd2+ and 8.23 mg/L
Ni2+, separately) with varying weights of adsorbent (0.025 g and 0.05 g) at 300 rpm
for 1 h. The adsorption was determined by UV-visible spectrophotometry (organic
dyes) and AAS (metal ions).

5 Results and Discussion

Rice husk is a naturally available agricultural waste material which is highly abundant
in rice cultivating countries. Rice husk contains mainly carbon based compounds such
as cellulose, hemicellulose and lignin, and silicon (Chuah et al. 2005). In addition,
compounds of Mg, K and Ca are also present (Chuah et al. 2005). Upon carbonization,
the above mentioned carbon based compounds convert to carbon which is rich with
functional groups. Silica is also produced while metal ions form their oxides. The
carbon and silica components function as adsorbents in this study. Hierarchically
porous monolithic carbon was synthesized by the sol-gel method where the solution
containing resorcinol and formaldehyde turned to a gel upon completion of the poly-
merization. Hydrophilic polymer PEG, acts as the porogen. O-phenylenediamine cat-
alyzes the polymerization of resorcinol and formaldehyde by providing the required
alkalinity for the reaction. The authors assume that it also acts as a crosslinker in the
polymerization through the Mannich reaction, as with diaminohexane as reported by
Hao et al. (2011). The reaction mixture was heated at 50 °C to provide the required
activation energy.
 Eco-Friendly Porous Carbon Materials for Wastewater Treatment 255

Table 1. FT-IR frequencies RH400, RH600, HPC and AC

Sample RH400 RH600 HPC AC
IR frequencies (cm−1) 3400–3800 3400–3800 3400–3800 3400–3800
1620 1620 1500–1600 1620
1095 1095 1500–1600
798.4 798.4
465 465

The chemical functionalities of the prepared porous carbon materials RH400,

RH600, HPC and AC were characterized by FT-IR spectroscopy as given in Table 1.
The bands in the region of 3400 cm−1–3800 cm−1 are ascribable to vibrations in non-
bonded O-H groups for alcohols, phenols and carboxylic acids (Allouche et al. 2015).
The band appearing at around 1620 cm−1 corresponds to the C = O vibration of lac-
tonic, carboxyl, or anhydride groups in AC and RH (Allouche et al. 2015). These
functional groups are absent in HPC. The bands in the region of 1500 cm−1−1600 cm−1
which are more apparent in AC and HPC are ascribable to aromatic ring C = C
stretching vibrations (Mopoung et al. 2015), and these are hidden in RH400 and RH600
due to the presence of a strong band around 1620 cm−1. In RH400 and RH600, the band
at 1095 cm−1 is assigned to Si-O-Si stretching, the band at 798.4 cm−1 to Si-O bending
and the band at 465 cm−1 to Si-O out of plane deformation (Shokri et al. 2009).

Fig. 1. TEM images of (a) RH (b) HPC.

TEM images of RH and HPC (Fig. 1 (a) and (b)) Polymers such as cellulose,
hemicellulose and lignin are subject to chain scissions or depolymerization, breaking of
C-C and C-O bonds within ring units and aromatization, and ultimately forming gra-
phite like structures. This process creates more mesopores and micropores (Galarneau
et al. 2016). The disordered mesopore structure of HPC is created mainly by the
thermal decomposition of PEG during the carbonization (Hou et al. 2004). PEG
molecules leave the carbon matrix leaving voids which remain in the structure as
mesopores.
256 C. J. Thambiliyagodage et al.

Fig. 2. SEM images of (a) RH  1200 (b) RH  5000 (c) HPC  2500 (d) HPC  5000.

The macropore structure of the prepared materials was studied by SEM. The
irregular shaped, disordered, naturally originated macropore structure could be seen in
RH 400 (Fig. 2(a)). Figure 2(b) shows the connectivity of the macropores providing
macro channels which facilitate the movement of electrolytes within the composite and
hence enhance the interaction between the electrolyte and the composite. SEM image
of HPC (Fig. 2(c)) shows the robust sponge-like disordered macropore structure with
ligaments. The structure of the ligaments is shown in Fig. 2(d). Ligaments are formed
during the phase separation of the synthesis. PEG acts as the porogen of this synthetic
procedure. The macropore structure is formed due to the polymerization-induced phase
separation (Mopoung et al. 2015). Resorcinol and formaldehyde monomers were
mixed with thermally unstable polyethyleneglycol polymer to produce a well dis-
tributed xerogel. Mesoporous structure are caused by the thermal decomposition of
PEG at high temperature (Hou et al. 2004). Resorcinol, PEG and water/ethanol solvent
is a homogeneous mixture where resorcinol molecules are surrounded by PEG mole-
cules. When formaldehyde is added to the mixture, base catalyzed resorcinol
formaldehyde condensation polymerization occurs. As the polymerization progresses,
hydrophilic groups of the resorcinol formaldehyde polymer interact with PEG through
H-bonding and form aggregates. This is the starting point of spinodal decomposition,
leading to a phase separation of the polymer rich phase and water rich phase. The phase
separation ends at the sol-gel transition. The macropore size will be determined by the
sol-gel transition kinetics, which in turn is controlled by the size of the polymer, the
ratio of PEG/polymer, the amount and the type of the polymer and the temperature
(Aghababazadeh et al. 2007). The SEM image (Fig. 2(c)), shows a heterogeneous
distribution of the macropore system. The amount of PEG and base, the amount and
type of polymer, and temperature are constant during the synthesis and hence, sol gel
 Eco-Friendly Porous Carbon Materials for Wastewater Treatment 257

transition kinetics is constant. During the synthesis, PEG 6000 was used. Lower
molecular weight PEG interacts more strongly with the oligomers of the resorcinol-
formaldehyde polymer via multiple interactions, adsorb on the surface of the oligomers
and cover the particle leading to faster condensation of the oligomers. This favors the
sol-gel transition resulting in smaller macropores. However, during the synthesis when
more and more oligomers are produced, the number of PEG molecules in the reaction
mixture is insufficient to interact with the oligomers and hence, incomplete conden-
sation results. Eventually, the sol gel transition decreases producing larger pores
(Galarneau et al. 2016).

Fig. 3. SEM image, element maps and EDX spectrum of (a) RH and (b) HPC.

Samples were analyzed by SEM-EDS in order to qualitatively determine the ele-

mental composition. Figure 3 shows the element maps and the energy dispersive X-ray
spectra of RH (Fig. 3(a)) and HPC (Fig. 3(b)). The element maps and EDX spectra
show that RH is rich in carbon and silica and that K, and Ca are also present. These
elements are probably present as their oxides (K2O and CaO respectively) as supported
by the presence of oxygen in the element maps and the spectrum. HPC on the other
hand consists mainly of carbon.

Fig. 4. Adsorption capacity of organic dyes, adsorbent dosage (a) 0.025 g (b) 0.050 g, and the
adsorption capacity of metal ions, adsorbent dosage (c) 0.025 g and (d) 0.050 g.
258 C. J. Thambiliyagodage et al.

The adsorption capacities of the adsorbents (adsorbent dosage as 0.025 g and

0.05 g) for organic dyes and metal ions are given in Fig. 4(a)–(d). The adsorption
capacity of AC for Methyl orange is much higher than that of the examined alterna-
tives, RH400, RH600 and HPC, irrespective of the adsorbent dosage used. However,
the adsorption capacity of all the adsorbents is almost the same for Methylene blue. The
mechanism of dye adsorption depends on many factors such as the physical and
chemical nature of the adsorbent and the mass-transfer process (Gong et al. 2013).
Methyl Orange and Methylene Blue are planar molecules and can be readily adsorbed
through p-p interactions between the aromatic rings of the dye molecules and the
hexagons of AC, RH and HPC (Gong et al. 2013). The higher adsorption capacity of
AC for Methyl Orange than the other adsorbents suggest that AC has more rings and
other functional groups than the other adsorbents. However, the adsorption of
Methylene Blue by these adsorbents is almost the same. Methyl Orange is an anionic
dye and Methylene Blue is a positively charged dye at the working pH (6.5).
Methylene Blue interacts electrostatically with negatively charged and electron rich
functional groups of the adsorbents. Negatively charged silica is found in RH but not in
AC and HPC. However, all of these adsorbents have electron rich functional groups.
The net result is that Methylene Blue would interact with all the adsorbents in a similar
manner. Interaction of Methylene Blue and Methyl Orange with the carbon structure is
shown in Fig. 5.

Fig. 5. Interactions of Methylene Blue and Methyl Orange with the carbon structure

The adsorption capacity of RH400 for all the metal ions selected for the study is
higher than for the other adsorbents, the highest being for Pb2+ (7.47 mg/g). In general,
RH600 showed a higher adsorption capacity than AC and HPC. All the examined
adsorbents adsorbed the other tested metals, Cu2+, Cd2+, and Ni2+ the lowest being for
Cd2+ (0.275 mg/g by RH400). Metal ions are positively charged and they can interact
only via electrostatic interactions. The negatively charged silica and electron rich
functional groups in RH400 can readily interact with metal ions. RH600 has silica as
 Eco-Friendly Porous Carbon Materials for Wastewater Treatment 259

well but the carbon content and concentration of functional groups is low due to the
higher temperature of pyrolysis. Functional groups leave as gaseous molecules with
increasing temperature of pyrolysis (Moldoveanu 1998). AC and HPC have only the
electron rich functional groups which can interact with metal ions resulting in low
adsorption capacities. Though HPC has a well-defined porous structure its adsorption
capacity is lower than that of the other adsorbents. The results obtained indicate that the
surface functionality of the adsorbent is more critical for adsorption of the studied
organic dyes and metal ions than the surface area and porosity. The adsorption capacity
of both dyes and metal ions studied decreased with increasing adsorbent mass.
Adsorption sites which adsorb the adsorbates remain saturated and at higher adsorbent
dosage aggregation of particles take place resulting in a lower surface area and hence
leads to lower adsorbent capacities.

6 Conclusions

Hierarchically porous nitrogen rich monolithic carbon could be successfully synthe-

sized by sol-gel synthesis. Rice husk based porous composites were also prepared as
the naturally available low cost adsorbent. HPC possesses a defined macropore and
mesopore structure and RH has a naturally originated porous structure.
Methyl Orange was removed more successfully by AC (77.5%) than the other
adsorbents. The adsorption of Methylene Blue was similar with all the adsorbents
(>90%). All the tested metal ions were removed most efficiently by RH400. The
adsorption capacities of the metal ions ranged from *0.28 mg/g for Cd2+ to 7.5 mg/g for
Pb2+ by RH400. RH400 is a promising adsorbent for removing Methylene Blue and Pb2+.

Acknowledgements. Authors acknowledge Sri Lanka Institute of Nanotechnology (SLINTEC)

for providing facilities for SEM and TEM analysis. This work was supported by a grant from Sri
Lanka Institute of Information Technology (SLIIT).

References
Sophia AC, Lima EC (2018) Removal of emerging contaminants from the environment by
adsorption. Ecotoxicol Environ Saf 150:1–17. https://doi.org/10.1016/j.ecoenv.2017.12.026
Aghababazadeh R et al (2007) Synthesis and characterization of silica nanoparticles. Solid State
Phenom 121–123:49–52. https://doi.org/10.4028/www.scientific.net/SSP.121-123.49
Allouche F-N, Yassaa N, Lounici H (2015) Sorption of methyl orange from aqueous solution on
Chitosan biomass. Proc Earth Planet Sci 15:596–601. https://doi.org/10.1016/j.proeps.2015.
08.109
Chuah TG et al (2005) Rice husk as a potentially low-cost biosorbent for heavy metal and
dye removal: an overview. Desalination 175(3):305–316. https://doi.org/10.1016/j.desal.
2004.10.014
Dias JM et al (2007) Waste materials for activated carbon preparation and its use in aqueous-
phase treatment: a review. J Environ Manag 85(4):833–846. https://doi.org/10.1016/j.
jenvman.2007.07.031
260 C. J. Thambiliyagodage et al.

Ebrahiem EE, Al-Maghrabi MN, Mobarki AR (2017) Removal of organic pollutants from
industrial wastewater by applying photo-Fenton oxidation technology. Arab J Chem 10:
S1674–S1679. https://doi.org/10.1016/j.arabjc.2013.06.012
Galarneau A et al (2016) Synthesis and textural characterization of mesoporous and meso-/
macroporous silica monoliths obtained by spinodal decomposition. Inorganics 4(2):9. https://
doi.org/10.3390/inorganics4020009
Gong R et al (2013) Adsorptive removal of Methyl orange and Methylene blue from aqueous
solution with finger-citron-residue-based activated carbon. Ind Eng Chem Res 52(39):14297–
14303. https://doi.org/10.1021/ie402138w
Hao G-P et al (2011) Structurally designed synthesis of mechanically stable poly (benzoxazine-
co-resol)-based porous carbon monoliths and their application as high-performance CO2
capture sorbents. J Am Chem Soc 133(29):11378–11388. https://doi.org/10.1021/ja203857g
Hou ZH et al (2004) Influence of polyethylene glycol on pore structure and electric double-layer
capacitance of carbon xerogel. J Cent South Univ Technol (Engl Ed) 11(3):255–260. https://
doi.org/10.1007/s11771-004-0052-z
Jaishankar M et al (2014) Toxicity, mechanism and health effects of some heavy metals. Interdisc
Toxicol 7(2):60–72. https://doi.org/10.2478/intox-2014-0009
Janakan N, Dissanayake LJ, Hewawasam S (2009) Chronic kidney disease of unknown aetiology
in Sri Lanka. Weekly Epidemiol Rep 36(November/December):1–4. http://www.epid.gov.lk
Jayasumana C, Gunatilake S, Senanayake P (2014) Glyphosate, hard water and nephrotoxic
metals: are they the culprits behind the epidemic of chronic kidney disease of unknown
etiology in Sri Lanka? Int J Environ Res Public Health 11(2):2125–2147. https://doi.org/10.
3390/ijerph110202125
Kim Y, Osako M, Lee D (2002) Removal of hydrophobic organic pollutants by coagulation-
precipitation process with dissolved humic matter. Waste Manag Res 20(4):341–349. https://
doi.org/10.1177/0734247X0202000405
Li QQ et al (2006) Persistent Organic Pollutants and Adverse Health Effects in Humans. J Toxicol
Environ Health Part A 69(21):1987–2005. https://doi.org/10.1080/15287390600751447
Moldoveanu SC (ed) (1998) The chemistry of the pyrolytic process. In: Analytical pyrolysis of
natural organic polymers. Elsevier (Techniques and Instrumentation in Analytical Chemistry),
pp 9–32. https://doi.org/10.1016/S0167-9244(98)80023-1
Mopoung S et al (2015) Characterization and properties of activated carbon prepared from
tamarind seeds by KOH activation for Fe (III) adsorption from aqueous solution. Sci World J
2015:1–9. https://doi.org/10.1155/2015/415961
Nasirabadi PS, Saljoughi E, Mousavi SM (2016) Membrane processes used for removal of
pharmaceuticals, hormones, endocrine disruptors and their metabolites from wastewaters: a
review. Desalin Water Treat 57(51):24146–24175. https://doi.org/10.1080/19443994.2016.
1140081
Senarath U, Visvanathan C (2001) Environmental issues in brackish water shrimp aquaculture in
Sri Lanka. Environ Manag 27(3):335–348. https://doi.org/10.1007/s002670010153
Shokri B, Firouzjah MA, Hosseini SI (2009) FTIR analysis of silicon dioxide thin film deposited
by metal organic-based PECVD. In: Proceedings of 19th international plasma chemistry
society, pp 1–4. www.ispc-conference.org
 Identification of Joint Efficiencies in 13 mm
Finger Jointed Timber Species Used
in Sri Lanka

C. K. Muthumala1(&), Sudhira De Silva2, K. K. I. U. Arunakumara3,

and P. L. A. G. Alwis4
1
Research, Development and Training Division, State Timber Corporation,
Battaramulla, Sri Lanka
[email protected]
2
Department of Civil and Environmental Engineering, Faculty of Engineering,
University of Ruhuna, Galle, Sri Lanka
3
Department of Crop Science, Faculty of Agriculture, University of Ruhuna,
Kamburupitiya, Sri Lanka
4
Department of Agricultural Engineering, Faculty of Agriculture,
University of Ruhuna, Kamburupitiya, Sri Lanka

Abstract. When using timber in construction and the furniture manufacturing

industry, waste timber materials and short length sections of timbers which are
dumped by sawmills are considered to be matter of concern. ‘Finger joint’ is a
recognized technique connecting two small pieces of waste timber together to
ensure their sustainable utilization. Currently, the technique is used in Sri Lanka
for non-structural purposes such as making timber boards and furniture. How-
ever, issues related with the strength of the joints have not fully investigated in
Sri Lanka. This study was undertaken to determine the tensile strength perfor-
mance of seven timber species in both unjointed (clear) and finger-jointed
methods with 13 mm finger lengths and 4 mm finger pitch. An SWR adhesive
(PVA) type was used as bonding material at normal exposure conditions. Well-
seasoned Grandis, Jack, Kumbuk, Mahogany, Pine, Satin and Teak timber
materials were used for the study.
BS 373: 1957 and BS EN 15497:2014 were used as standards for tests. The
tests for tensile properties were performed using the Universal Testing Machine
(UTM 100 PC) with a loading plate moving speed of 01 mm/min. Maximum
load was read on graph to calculate the ultimate tensile strength.
The highest ultimate average tensile strength values were recorded in control
specimens and different joint efficiencies were recorded in different finger
jointed species. The highest mean finger joint efficiency percentage was
obtained from Mahogany timber species and the least mean finger joint effi-
ciency percentage was recorded in Kumbuk timber species.

Keywords: Finger joint
Tensile strength
SWR glue type
Joint efficiencies

© Springer Nature Singapore Pte Ltd. 2020

R. Dissanayake and P. Mendis (Eds.): ICSBE 2018, LNCE 44, pp. 261–267, 2020.
https://doi.org/10.1007/978-981-13-9749-3_24
262 C. K. Muthumala et al.

1 Introduction

The history of wood architecture of Sri Lanka has revealed that architects and crafts-
man with the natural creative skills have existed since the beginning of the country’s
long history and civilization. As a result of low quality and low strength of the local
wood used by some of these ancestral architects and craftsmen, most wood creations
have not lasted for long and have perished with time (Ruwanpathirana and Muthumala
2010).
The forests of Sri Lanka have nearly 450 types of hardwood timber species which
are used for structural and non-structural purposes. The wood is excellent material for
roofing and other construction works, furniture, interior decorations, doors and window
frames, panelling, partition borders, floorings, wood carvings, musical instruments etc.
Physical and mechanical properties are very important in structural and non-
structural purposes of timber. The properties not only vary with the species but also
with factors such as the moisture content and defects of wood.
Finger joint is a sustainable, eco-friendly and economically viable concept for the
furniture industry. It ensures the sustainable utilization of small wood cut pieces which
are removed as waste (Sandika 2017). Finger joint timber manufacturing is considered
to be a viable solution for minimizing the waste production in furniture manufacturing
activities. Types of glue applied for the joining process of finger joint is one of the key
factors which determines the strength of the product. The highest tensile strength of
finger jointed timber solids was recorded in the SWR glue type in Sri Lanka
(Muthumala et al. 2018a, b).
However, issues related with the strength of the joints have still not been fully
investigated in Sri Lanka. This study was undertaken to determine the tensile strength
performance of both unjointed (solid) and finger-jointed in seven species of timber with
13 mm finger lengths, which is the commonly used finger type in Sri Lankan furniture
manufacturing.

2 Materials and Methods

Unjointed and jointed finger joint samples were cut from seasoned, defect-free wood
timber to calculate tensile strength properties. Finger joint specimens were made at the
finger joint factory at Boossa in the State Timber Corporation (STC) using 13 mm
finger joint length cutters and an assembling pressure of 6 MPa was used in this study.
Clear timber specimens were taken as control specimens and finger jointed timber
specimens with the same dimensions were made with constant finger geometry such as
13 mm Finger Length, 1 mm Tip width and 4 mm finger pitch. Finger joint geometry
is shown in Fig. 1. Polyvinyl acetate (PVA) SWR adhesive was used as a bonding
agent (Glue type) for finger jointed wood.
Table 1 shows that timber species mainly used in furniture industries in Sri Lanka.
In this study those timber species were used to calculate tensile strength properties.
The specimens were prepared from defect – free, sawn woods and dimensions of
the sample prepared for above test is shown in Fig. 2.
 Identification of Joint Efficiencies in 13 mm Finger Jointed Timber Species 263

Fig. 1. Finger joint geometry

Table 1. Timber species mainly used in the furniture manufacturing industry in Sri Lanka.
Timber species Botanical name Timber class (STC)
Grandis Eucalyptus grandis Class-II
Jack Artocarpas heteroplyllus Luxury
Kumbuk Terminalia arjuna Special
Mahogany Swietenia macrophylla Luxury
Pine Pinus caribaea Class-III
Satin Chloroxylon swietenia Luxury
Teak Tectona grandis Supper Luxury

Fig. 2. The dimensions of the samples prepared for tension test.

The density values were calculated for seven timber species using following
equation (Eq. 1). Dry weight of the timber samples was taken by placing them in the
electric oven for a period of 48 h at 105 °C (BS EN 373:1957).

Weight of oven dry wood ðkgÞ

Density ¼ ð1Þ
Volume of wood ðm3 Þ

Determination of basic density was done based on the green volume and oven-dry
weight using a water displacement method. Moisture contents were measured using a
moisture meter.
264 C. K. Muthumala et al.

Ten replicates were made for each timber species. Samples which were placed in
normal room temperature conditions showed good structural performance compared to
hot and wet conditioned ones (Vievek et al. 2016).
BS 373: 1957 and BS EN 15497:2014 were used as standards for tests. The test for
tensile strength were performed by the Universal Testing Machine (UTM 100 PC)
(Figs. 3 and 4).

Fig. 3. a. The loading setup b. Finger joint specimen

Fig. 4. a. Moisture meter b. Elec. Oven

Determination of the Tensile Strength

Maximum load act in timber section was taken into calculation. The following equation
was used to calculate the Tensile strength (Eq. 2).

Maximum Load
Tensile strength of timber ¼ N=mm2 ð2Þ
Average cross section area of specimen
 Identification of Joint Efficiencies in 13 mm Finger Jointed Timber Species 265

3 Results and Discussion

Dry density values and moisture content of the timber species are shown in Table 2
(Fig. 5).

Table 2. Dry density values and moisture content of timber species.

Timber species Density (kg/m3) Moisture content %
Grandis 570 ± 5 11 ± 2
Jack 645 ± 3 12 ± 2
Kumbuk 750 ± 4 11 ± 2
Mahogany 570 ± 3 12 ± 1
Pine 460 ± 2 12 ± 2
Satin 975 ± 5 12 ± 1
Teak 720 ± 5 12 ± 2

Fig. 5. Tensile strengths of clear and finger jointed specimens

According to Table 3, clear specimens and finger jointed specimens have statisti-
cally significant effect. P < 0.0005

Table 3. Statistical data from SPSS for tensile strength in both c lear and finger-jointed
specimens
ANOVA
Sum of squares df Mean square F Sig.
Clear specimens Between groups 33890.571 6 5648.429 7.372 0.000
Within groups 48272.864 63 766.236
Total 82163.435 69
FJ specimens Between groups 9385.919 6 1564.320 14.253 0.000
Within groups 6914.491 63 109.754
Total 16300.410 69
266 C. K. Muthumala et al.

According to the overall results of the tests, control (unjointed) samples gave
significantly (p < 0.005) higher tensile strength than 13 mm finger jointed specimens.
The highest tensile strength was recorded in clear specimens of Teak and the least
tensile strength was showed by finger jointed Kumbuk.
Clear specimen of Teak showed the highest tensile strength followed by Grandis,
Kumbuk, Satin, Pine, Jack, while the lowest recorded was Mahogany. Finger joint
specimen of Grandis showed the highest tensile strength followed by Teak, Jack,
Mahogany, Satin, Pine, while the lowest was Kumbuk.
According to Table 4, the highest mean efficiency percentage (75.08%) for finger
joint was obtained from Mahogany and the least mean efficiency percentage for finger
joint (19.60%) was recorded in Kumbuk. Mean finger joint efficiency percentage varied
as, Mahogany > Jack > Grandis > Teak > Satin > Pine > Kumbuk.

Table 4. Finger joint efficiencies of seven timber species

Timber Mean tensile Mean tensile strength Strength Joint
species strength of clear of finger-joint difference efficiency
specimens specimens (N/mm2) (Clear - joint) %
(N/mm2) (N/mm2)
Grandis 103.16 50.24 52.92 48.70
Jack 73.88 36.06 37.82 48.80
Kumbuk 86.13 16.88 69.25 19.60
Mahogany 47.64 35.77 11.87 75.08
Pine 83.23 17.04 66.19 20.47
Satin 83.57 21.13 62.44 25.28
Teak 124.25 36.47 87.78 29.35

Yeh et al. (2011) reported that no significant difference in the tensile strength of
finger-jointed lumber was found between horizontal and vertical finger formation.

4 Summary and Conclusion

The following conclusions can be drawn from this study.

Control (unjointed) samples gave higher tensile strength than 13 mm finger jointed
specimens.
(1) The highest tensile strength was recorded in clear specimens of Teak and the least
tensile strength was showed by finger jointed Kumbuk.
(2) The highest mean efficiency percentage (75.08%) in finger joint was obtained
from Mahogany and the least mean efficiency percentage (19.60%) in finger joint
was recorded in Kumbuk.
(3) The study illustrates the joint efficiency of seven finger jointed timber species.
Finger joint manufacturers can use these results in increasing their production
efficiency.
 Identification of Joint Efficiencies in 13 mm Finger Jointed Timber Species 267

References
British Standards Institution (1957) BS 373: 1957, Methods of testing small clear specimens of
timber. British Standards Institution, London
British Standards Institution (2014) BS EN 15497:2014, Structural finger jointed solid timber-
performance requirements and minimum production requirements. European Committee for
Standardization, B 1000 Brussels
Yeh M-C, Lin Y-L, Huang Y-C (2011) Evaluation of the tensile strength of structural finger-
jointed lumber. Taiwan J Sci 26(1):59–70. www.airtilibrary.com. Accessed 3 Aug 2018
Muthumala CK, De Siva S, Alwis PLAG, Arunakumara KKIU (2018a) Factors affecting the glue
strength of finger-joints in commonly used timber species in Sri Lanka. In: International
symposium on agriculture and environment. Faculty of Agriculture, University of Ruhuna, Sri
Lanka, pp 126–128
Muthumala CK, De Siva S, Alwis PLAG, Arunakumara KKIU (2018b) Investigate the most
suitable glue type for finger-joints production in Sri Lanka. Res J Agric For Sci 6(11):6–9
Ruwanpathirana ND, Muthumala CK (2010) Wooden wonders of Sri Lanka. State Timber
Corporation, Sri Lanka
Sandika AL, Pathirana GDPS, Muthumala CK (2017) Finger joint timber products for effective
utilization of natural resources: an analysis of physical properties, Economic factors and
Consumers’ perception. In: International symposium on agriculture and environment,
University of Ruhuna, Sri Lanka, pp 109–111
Vievek S, De Silva S, De Silva S, Muthumala CK (2016) Finger joint and their structural
performance in different exposure conditions. In: 7th international conference on sustainable
build environment, Kandy, Sri Lanka, pp 207–210
 Experimental Investigation of Top Hinged Flap
Type Wave Energy Device with a Bottom
Gap – Preliminary Experimentations

L. U. Bakmeedeniya(&) and S. D. G. S. P. Gunawardane

Department of Mechanical Engineering, University of Peradeniya,

Peradeniya, Sri Lanka
[email protected]

Abstract. The “Pendulor” type wave energy converting system essentially

consists of a caisson, a water chamber facing the offshore open, and a flap that is
hung down in that. The flap is driven by the oscillatory motion of the standing
waves generated inside the caisson. The flap then drives a generator through a
hydraulic power transmission system. The main objective of the proposed study
is to study the hydrodynamics of the device with a gap between the bottom of
the flap and the caisson. As an initiative, this paper describes the test facility
arrangements with wave tank modification, calibrations and preliminary model
test results. Accordingly, the experimental facility, the wave tank at the Faculty
of Engineering, University of Peradeniya was modified to comply with a scale
models of (1/40) of the proposed device configuration for typical wave climatic
conditions in which the wave length is 100 m, maximum wave height is 6 m
and the power density is 20 kW/m.

Keywords: Wave energy
Power capture
Flap-type converter

Hydrodynamic characteristics
Caisson
Bottom gap

1 Introduction

1.1 Significance of the Research Work

The energy from ocean waves is the most conspicuous form of ocean energy, possibly
because of often spectacular but destructive effects. The waves are produced by wind
action and are therefore an indirect form of solar energy. The potential for extracting
and using the energy in ocean waves has been recognized for more than 200 years with
the first patent for a wave energy converter being submitted by a father and a son
named Girad in 1799 in Germany [1]. More than thousand patents had been registered
by 1980 and the number has increased markedly since then but no viable technology
has gained merit as a fully functional commercial product to harness wave energy.
After the first oil crisis in 1974, significant investments were done in wave energy
research. As a result, during 1973–1984 a number of large scale (2 GW) wave energy
concepts were developed. By the time of the oil crisis many investors were inspired by
the possibility of converting wave energy into usable energy, but in over three decades
there are not many commercial uses compared to other renewables. Up until the year

© Springer Nature Singapore Pte Ltd. 2020

R. Dissanayake and P. Mendis (Eds.): ICSBE 2018, LNCE 44, pp. 268–281, 2020.
https://doi.org/10.1007/978-981-13-9749-3_25
 Experimental Investigation of Top Hinged Flap Type Wave Energy Device 269

2000, there was a development of small scale wave energy projects. By the year 2000
and moving to the 21st century world awakened with the key challenges of the growing
concerns regarding global warming, the need for a sustainable energy form and the
utilisation of renewable form. Wave energy is a non-polluting and renewable source of
energy, created by natural conversion of part of the wind energy above the oceans. The
wave power density is higher than the wind and solar power since it is a concentrated
form of solar energy through wind energy. Just below the ocean water surface the wave
energy flow, is typically five times denser than the wind energy flow 20 m above the
sea surface, and 10 to 30 times denser than the solar energy flow. As a non-polluting
renewable source of energy it can give consistent power for low operating costs. Also it
will provide the least visual impact in the point of view of structure since most of the
systems will be fully submerged. More importantly these systems can be used effi-
ciently and as shoreline protection systems as well.
Having considered the fact that there is significant potential for extracting and using
the energy in ocean waves the world’s attention now has focused on the possibilities of
extracting power from wave energy. The “pendulor” device discussed in this paper
invented in Japan is one of the pioneers among wave energy converters (WEC) which
is the attention of this research. This was first invented in Japan at Muroran Institute of
Technology as a result of work done in the early 1980s and has proven 40–50% of
wave energy conversion efficiency at the sea of Muroran.

1.2 Background
The situation on wave energy in Europe was changed from 1990s and in northern
America by 2000. Since wave energy absorption is a hydrodynamic process of con-
siderable theoretical difficulty, much research needed to be done to get maximum
benefit. Due to the problems that lie in the variability of the energy flux absorbed from
the waves, in several time scales; sea states and seasonable variations and the sur-
vivability in extreme conditions are major issues. In general development, from con-
cept to commercial stage in wave energy utilization has found to be difficult, slow and
expensive process. This research is primarily focused on a top hinged flap type wave
energy device which is known as Pendulor (Fig. 1(a)) invented in Japan with a bottom
gap. Experimental sea trails at the Muroran sea, has shown that overall conversion
efficiency is about 40–50%, while the theoretical efficiency of this type device is 100%
(regular wave assumption without gaps) under the normal running conditions [2, 4].
However related experimental studies are still limited on testing the energy capture of
Pendulor wave energy devices through hydrodynamic investigations. There is not
much research done to explore the avenues in this area in Sri Lanka although a
substantial progress has been achieved in the theoretical and numerical modelling of
WECs [2]. Model testing in wave basins and evaluation of optimum conditions is still
essential and finally, testing in real sea conditions [3].

1.2.1 Operation of Pendulor Wave Energy Conversion Device

In Fig. 1(a) basic components of a “Pendulor” type energy device are shown. A flap
hung inside the caisson which faces to the ocean waves is driven by the horizontal
water particle motion of the waves (Fig. 1(b)). The device operates on standing waves
270 L. U. Bakmeedeniya and S. D. G. S. P. Gunawardane

Fig. 1. (a) and (b) Basic “Pendulor type energy conversion device configuration and movement
of water particle in the wave

which are created by superposition of the incident waves and reflected waves from the
back wall. The excitation moment of the flap depends on the caisson configuration as
well as the wave frequency and the wave height. The pendulum is placed at the nodal
point of the standing waves Fig. 1(b) where the water particle motion is almost hori-
zontal and to extract the maximum power output (where the kinetic energy is maxi-
mum). The nodal point is at a distance ‘d’ where d = L/4 for a wave inside a caisson,
(L = wave length) for regular waves.
Inside the caisson, the pendulum is placed. The natural frequency (xn) of the
pendulum and the water chamber (water quantity inside the caisson) is set to the
incident wave frequency, to obtain the maximum efficiency of the device. The pen-
dulum drives a rotary vane pump and the wave energy is converted into pressure
energy of a fluid by vane pump. This pressurized fluid is then used to fulfil any energy
requirement.

2 Objectives

In the ocean, the wave’s frequency is varied according to local wave climate as well as
global wave climate. To absorb the maximum wave energy, most of the device should
resonate with the wave frequency. In the case of Pendulor device, various researches
have been done to develop mathematical models and experiments to improve the
performances [3, 5, 6]. Based on those models, analysis of motion response and cor-
responding power conversion capability of the device were conducted and discussed
for regular and irregular wave climate in Sri Lanka [2]. The related theory and most of
the research reveals that the performance of the device highly depends on the caisson
configurations. These configurations involve the placement of the flap inside the
caisson and it is highly necessary to understand the hydrodynamics to optimize the
 Experimental Investigation of Top Hinged Flap Type Wave Energy Device 271

configuration [14]. No research has been done to look into the hydrodynamic char-
acteristics of caisson with gaps around the flap. Therefore in this research a caisson
with a flap hung on from the top with a bottom gap was tested for maximum power
capture under different wave frequencies.

3 Research Design and Methodology

3.1 Development of the Scale Model of 1/40
A scale model of 1/40 was designed and constructed to test with some modifications to
the existing towing tank. The experimental setup is shown in the Fig. 2 below.

Fig. 2. The profile of the final test model in the wave tank

Fig. 3. (a) The platform model, (b) The caisson model

272 L. U. Bakmeedeniya and S. D. G. S. P. Gunawardane

1. Flap Arms
2. Potentiometer
3. Main Pulley
4. Main shaft
5. Cable
6. Tension pulley
7. Main bearings
8. Flexible coupling
9. Magnetic particle brake
10. Rigid coupling
11. Torque sensor

Fig. 4. Components of power harnessing unit

The test model scale (1/40) has been selected mainly to match the width of the
wave channel to achieve a 3D effect around the caisson (the model is nearly 1/3 rd of
the width of the channel). To match the required water depth, for the model testing the
tank needed to be modified using a platform as shown in the Fig. 2. The slope of the
bottom was kept at 1/20 at seaward and 1/10 in the leeward direction. The dimensions
of the platform and the basic caisson configuration with dimensions are shown sepa-
rately in Fig. 3(a) and (b).
The caisson is placed at the top of the platform and the energy absorbed by the flap
is transferred to the power harnessing unit. The main features of the power harnessing
unit are shown in Fig. 4.

3.2 Calibration of the Modified Experimental Setup

The existing towing tank was modified as indicated previously and the wave charac-
teristics were investigated for model study. The generated waves were recorded by the
three wave gauge method and analyzed to obtain the required wave characteristics for
the proposed study using a least squares method (Fig. 5).
Simultaneous measurements of the waves at three positions in flume which are in
reasonable proximity to each other and are in a line parallel to the direction of wave
propagation were done to separate the incident and reflected a spectra of generated
irregular sea state (for different frequencies) for the experiment. Such measured waves
are shown in Fig. 6. Then, to compare the generated waves with the anticipated
sinusoidal waves, in other words, the fluid motion was done using a 3-point method
which uses a least square analysis for decomposing the measured spectra into incident
and reflected spectra with greater accuracy and range. Thereby, the wave channel was
calibrated with the analyzed results to measure the wave heights required for proposed
WEC model study (Table 1).
Experimental Investigation of Top Hinged Flap Type Wave Energy Device 273

Fig. 5. Wave gauge assembly

Table 1. Incident wave height results for wave generation

Stroke/cm , Wave Height/mm

PT
19 16 15 14 13 11 9 8 6
0.9
1 55
5
1.1 70 60 0
5 4
1.2 65 55 0 0
5
1.3 70 60 0
5
1.4 65 58 0
5
1.5 70 62 57 2
1.6 75 60 57 55 50
1.7 70 60 55 50
1.8 65 55
1.86 55
274 L. U. Bakmeedeniya and S. D. G. S. P. Gunawardane

Fig. 6. Wave profiles for the three wave probes at periodic time 1.1 s

3.3 Power Takeoff System (PTO) to Measure the Power Capture

The arrangement shown in Fig. 4 is the PTO measuring system developed for the
model. A torque sensor model TQM 202-0.7 which has a sensitivity of 2.4435 mV/V
has the range from 0–0.7 Nm is used to measure the excitation torque by the waves.
A reaction torque, magnetic particle BRAKE B35 model which has a torque range of
0.07–4 Nm is used to apply the load to the system and a high precision potentiometer
was used to measure the angle of rotation of the flap (Model 157 from Vishay Spectral
can rotate 360º). Installation and measurements in action are shown in Fig. 7(a) and (b).
 Experimental Investigation of Top Hinged Flap Type Wave Energy Device 275

Fig. 7. (a). Assembling and assembled unit for preliminary model testing (b). Assembling and
assembled unit for preliminary model testing

3.4 Measurements and Calculations

Measurements were obtained through a LABVIEW program as real time data for
Torque on the shaft due to wave power, flap rotation angle and the incident wave
height. A video camera was used as shown in the Fig. 2 to capture the motion of the
flap and the water inside the caisson. The two signals from the power harnessing unit
(torque and angle) will lead to interpret the wave energy absorbed by the flap which is
generated through the torque sensor and the potentiometer. A magnetic particle brake
was used to adjust the damping torque of the power harnessing unit (Fig. 8).
276 L. U. Bakmeedeniya and S. D. G. S. P. Gunawardane

Fig. 8. Graphical representation for incident wave, angle of rotation and the torque

4 Results

4.1 Mathematical Model for the Flap Dynamics

The equation of motion of the flap for the constant damping (Coulomb damping Nm)
brake torque device is given as the following single degree of freedom Coulomb
damped forced oscillation system.
 Experimental Investigation of Top Hinged Flap Type Wave Energy Device 277

X X
I €h þ ½Nh ðxÞ:h_ þ Nm þ K:h ¼ M0 sin xt ð1Þ
P P
Where I ¼ Ih ðxÞ þ Im and K ¼ Kh ðxÞ þ Km
Im : moment of inertia of the flap, Ih(x): moment of inertia of the added water mass,
Nm: Coulomb damping torque, Nh: Hydrodynamic damping torque, Km: Coefficient of
restoring moment by the flap gravitation, Kh(x): Coefficient of restoring moment by the
water mass elevation, M0: Wave excitation moment. For a straight caisson, the rela-
tionship between the above parameters and the wave parameters were discussed in
detail in literature [2].
And the Power capture is given by
 
Pmax ðxÞ ¼ Nm h_ ð2Þ

Where HðxÞ: Complex amplitude of flap rotation. The normalized power capture is
the capture factor Cf defined by the ratio of the power capture divided by the wave
power incident on the width of the flap.

~ max ðxÞ ¼ Cf ðxÞ ¼ Pmax ðxÞ=½Pi ðxÞ

P ð3Þ
P 
t2 Torque  Angle difference
PTO ¼ KðxÞ ¼ t1 Time is the power take off (power capture)
from the system. And the wave power is given by
   
1 H 2 x 2k0 h
Pw ¼ b qg 1þ ð4Þ
4 2 k0 sinhð2k0 hÞ

where b: flap width, H: incident wave height, h: water height, x: wave frequency, k0 is
the wave number which can be calculated from the dispersion relationship given by
x2 =g ¼ k0 tan(k0 h).

4.2 Experimental Analysis of the Model

Initially A 1/40th scale model caisson configuration for a water depth of 25 cm and a
wave length of 2 m, was considered as a preliminary test flap arrangement without a
bottom gap for wave periods 1.1(x = 5.712), 1.4(x = 4.488), 1.6(x = 3.927) and 1.86
(x = 3.378). Figure 9 shows the variation of average power capture and conversion
efficiency with PTO damping for different wave periods. It indicates that the peak for
efficiency will be beyond the experimented range and some modifications (large
damping) to the PTO damping is required.
278 L. U. Bakmeedeniya and S. D. G. S. P. Gunawardane

Fig. 9. Variation of average power and efficiency with average torque for different frequencies

Then, a 1/40th scale model caisson configuration for the same water depth of 25 cm
and a wave length of 2 m, was considered as a preliminary test flap arrangement with a
bottom gap of 25 mm. Two positions were considered in the experiments as the flap
placed at distances of one quarter and one-eighth of a wave length from the back wall
of the caisson.
 Experimental Investigation of Top Hinged Flap Type Wave Energy Device 279

Experiments were conducted for the periodic times 1.1(x = 5.712), 1.2(x = 5.236),
1.3(x = 4.833).

Fig. 10. Variation of average power with average torque for periodic times 1.1, 1.2 and 1.3
(Flap placed at quarter of a wave length with a bottom gap)
280 L. U. Bakmeedeniya and S. D. G. S. P. Gunawardane

5 Results and Discussion

The conventional model without a gap that is the sealed caisson chamber in the first set
of tests indicates a peak as can be seen in Fig. 9 in the power capture with a value of
0.4 W for the wave frequency of 5.7 rad/s (PT 1.1). That is, it produces an average
power of 2.25 W when there is a damping torque of 0.65 Nm applied, resulting with an
efficiency of 10.4%. A peak can be anticipated from the trend lines for the periodic
times of 1.4, 1.6 and 1.86 as can be seen in the Fig. 9. It is observed that there is a trend
to increase the capture efficiency with the increase of torque. Nevertheless, in this stage
of the experimentation, the applied torque had to be limited due to the limitations of the
PTO system. In this case, the peak in wave torque occurs as a standing wave is
generated in front of the flap and then the wave torque reduces as diffraction begins to
reduce the wave torque.
Then, the variation of the average power and the damping torque for the settings of
the flap with a bottom gap of 25 mm in the scale model which was placed at a distance
of a quarter of a wave length from the back wall is shown in Fig. 10. In both cases of
the caisson without a gap and caisson with a bottom gap the chamber length was set at
50 cm (that is, the flap was placed at a distance of 50 cm from the back wall-a quarter
of a wave length) and the chamber width was 30 cm in 1/40 scale. As can be seen in
the figures there are peaks in the capture factor for the vented caisson too. Also,
repeated results for the flap at the distance of approximately one-fourth and one-eighth
of a wave length indicates a maximum power capture as expected. These maxima occur
as the sloshing mode is excited by the incident wave resulting in the formation of a
standing wave within the chamber. Although the tank tests carried out so far have not
been extensive, the analysis shows promising results. However, extensive tank
experiments are needed to validate and come to a better conclusion with the numerical
model. Further work is required of wave tank experiments with utilizing PIV and video
analyzing to validate the numerical models.

6 Conclusion

The objective of the experimental work is to investigate the hydrodynamics of a top

hinged wave energy converter with a bottom gap and to compare with the numerical
model. The experimental setup was established in the existing towing tank by modi-
fying the tank to match with the testing requirements. Preliminary test results showed
that the experimental setup needs further modifications to capture the essential char-
acteristics of the device (peak efficiency etc.) for broader spectrum of cases.
 Experimental Investigation of Top Hinged Flap Type Wave Energy Device 281

References
1. de Antonio FO (2010) Wave energy utilization: a review of the technologies. Renew Sustain
Energy Rev 14:899–918
2. Watabe T, Yokouchi H, Gunawardana SDGSP, Obeysekara BRK, Dissanayake UI (2001)
Preliminary study on wave energy utilization in Sri Lanka. In: Proceedings of ISOPE 2001,
Kithakyushu, Japan
3. Gunawardane SP, Kankanamge CJ (2016) Study on the performance of the “Pendulor” wave
energy converter in an array configuration. Energies 9(4):282. https://doi.org/10.3390/
en9040282
4. Watabe T, Kondo H, Matsuda T, Yano K, Dote Y, Takagi M (1984) Method and apparatus
for generating electric power by waves, U.S. Patent 4,490,621, 25 December 1984
5. Watabe T, Kondo H, Yano K (1985) Characteristic of a pendulor wave energy converter.
Ocean Eng 12:475–598 1st symposium of wave energy utilization
6. López I, Andreu J, Ceballos S, de Alegría IM, Kortabarria I (2013) Review of wave energy
technologies and the necessary power equipment. Renew Sustain Energy Rev 27:413–434
7. Wan Z, Zheng H, Sun K, Zhou K (2017) A model and experimental study of an improved
pendulor wave energy converter. Energy Procedia 105:283–288 8th international conference
on applied energy 2016
8. Whittaker T, Folley M, Osterried M The oscillating wave surge converter. https://www.
researchgate.net/publications/266404049
9. Whittaker T, Folley M Nearshore oscillating wave surge converters and the development of
Oyster, Philos Trans Roy Soc. http://rsta.royalsocietypublishing.org. Accessed 31 July 2017
10. Wilkinson L, Whittaker TJT, Thies PR, Day S, Ingram D (2017) The power-capture of a
nearshore, modular, flap-type wave energy converter in regular waves. Ocean Eng 137:394–
403
11. Porter R, Biggs RT (2012) Wave energy absorption by a flap-type oscillating wave surge
converter. https://people.maths.bris.ac.lk/*marp/abstracts/single-flap-110912.pdf. Accessed
27 July 2017
12. Renzi E, Dias F (2013) Hydrodynamics of the oscillating wave surge converter in the open
ocean. Eur J Mech B/Fluids 41:1–10
13. Sarkar D, Renzi E, Dias F (2015) Effect of straight coast on the hydrodynamics and
performance of the oscillating wave surge converter. Ocean Eng 105:25–32
14. Gunawardane SDGSP, Folley M, Kankanamge CJ (2019) Analysis of the hydrodynamics of
four different oscillating wave surge converter concepts. Renew Energy. https://doi.org/10.
1016/j.renene.2018.06.115
15. Dean RG, Dalrymple RA (1991) Water wave mechanics for engineers and scientists. World
Scientific Publishing Co. Pte. Ltd., Singapore ISBN 9810204205
 Increasing the Efficiency of a Dye-Sensitized
Solid-State Solar Cell by Iodine Elimination
Process in Hole Conductor Material

A. W. M. V. Ekanayake1(&), G. R. A. Kumara2, R. M. G. Rajapaksa3,

and A. Pallegedara4
1
Postgraduate Institute of Science, University of Peradeniya,
Peradeniya, Sri Lanka
[email protected]
2
Institute of Fundamental Studies, Kandy, Sri Lanka
3
Department of Chemistry, University of Peradeniya, Peradeniya, Sri Lanka
4
Department of Chemical and Process Engineering, University of Peradeniya,
Peradeniya, Sri Lanka

Abstract. The use of silicon in photovoltaics are popular these days. Besides
the use of silicon, the use of titanium dioxide as an n-type semiconductor
material is trending in the emerging dye-sensitized solar cells due to its low cost,
nontoxicity and ease in preparation when compared to silicon. The objective of
this research is to find avenues to enhance the efficiency of a dye-sensitized solid
state solar cell, fabricated using D-149 and N-719 dyes as sensitizers and excess
iodine-eliminated cuprous iodide hole conductor. P-25 Degussa titanium dioxide
in a titanium dioxide colloidal suspension was used to deposit titanium dioxide
thin films with a thickness less than 20 µm on fluorine-doped tin oxide glasses.
Titanium dioxide films were then immersed in the dye solution for 12 h to
enable dye adsorption. Acetonitrile solution containing refined copper powder
treated cuprous iodide was used to deposit excess iodine-eliminated cuprous
iodide on to the working electrode. Optimum efficiency of 2.11% was obtained
for the dye-sensitized solid-state solar cell fabricated using D-149 dye and 1.6%
was obtained for N-719 dye. The highest solar parameters obtained were shown
for the dye-sensitized solid-state solar cell fabricated for D-149 dye which were
an open circuit photo-voltage of 475 mV and short circuit current density of
12.7 A cm−2.

Keywords: Iodine
Titanium dioxide
Dye
Solar cells

1 Introduction

Without any controversy, it is an acceptable fact that world relies on energy [1]. As the
population of the world increases, the supply of energy should be equal or more,
corresponding to the demand. Most abundant energy sources come either directly or
indirectly from the sun’s energy [2]. As an alternative to the dwindling energy sources,
solar energy can be used to heat buildings, generate electricity, heat water and a variety
of other domestic and industrial uses. Solar energy provides a tremendous resource for

© Springer Nature Singapore Pte Ltd. 2020

R. Dissanayake and P. Mendis (Eds.): ICSBE 2018, LNCE 44, pp. 282–287, 2020.
https://doi.org/10.1007/978-981-13-9749-3_26
 Increasing the Efficiency of a Dye-Sensitized Solid-State Solar Cell 283

generating clean and sustainable electricity with no air and water pollution [3], no
global warming, no risk of electricity price hikes and no threats to public health. The
principle behind every solar cell today is the photovoltaic effect [4] which produces an
electric current and a voltage in a material upon exposure to solar radiation. Belonging
to the third generation [5] of solar cells, dye-sensitized solid-state solar cells (DSSCs)
mimics the action of photosynthesis [6]. DSSC configuration sandwiches a monolayer
of dye in between two inorganic wide band gap semiconductor materials [7], one with a
p-type hole conducting mechanism and the other one exhibiting n-type conduction
mechanism. Here, the purpose is to enhance the efficiency of the DSSC by changing the
properties of the p-type CuI semiconductor material.

2 Materials and Methods

The following procedure was used to prepare TiO2 colloidal suspension solution. First,
5.0 ml of Titanium Tetraisopropoxide (Sigma Aldrich) was mixed with 5.5 ml of
glacial acetic acid (Sigma Aldrich) and 20.0 ml of iso-propyl alcohol (Sigma Aldrich)
and ground in a ceramic motor. Then, 5.0 ml of distilled water was added to the
solution and a gelatinous mixture was obtained. Afterwards 0.650 g of P25 Degussa
TiO2 powder was added to the gelatinous solution and ground well. Afterwards,
fluorine doped tin oxide (FTO) glasses were immersed in a detergent solution and
ultrasonicated for 15 min. TiO2 films were prepared on FTO glass substrates by using
the drop coating method. Cleaned FTO glass substrates were kept on a hot plate at
150 °C with the conducting side facing upwards. TiO2 colloidal suspension paste was
deposited by using a dropper. Due to the heat, the liquid component of the thick paste
may evaporate leaving a brownish residue on the glass surface. This brownish residue
was blown off using a jetstream of air. After repeating this process for several times, the
plates were sintered at 500 °C for 2 min and then allowed to cool. After following this
process for a few rounds, a uniform layer of white TiO2 is formed. Plates were sintered
in a furnace at 500 °C for 10 min.
Indoline (D-149) & N-719 dyes were used as the sensitizer dyes in this research.
100 ml of solvent was prepared by mixing acetonitrile and tert-butyl alcohol in the ratio
of 1:1. Then 0.3 mM of D-149 & N-719 dyes were dissolved in separate solvents to
prepare separate dye solutions.
TiO2 coated electrodes were preheated to 80 °C and dipped in a newly prepared
dye solution for 12 h and allowed to deposit a monolayer of dye on the TiO2 surface.
The electrodes were taken out of the dye solution and washed with acetonitrile to
remove the excess dye molecules on the TiO2 surface. Then, the dye-sensitized TiO2
plates were dried using a hot air-gun.
Initially, copper oxide powder was washed with dilute HNO3 (0.1M) in a Buchner
funnel. Next, it was washed with iso-propyl alcohol. Finally, copper powder was dried
at 150 °C for 30 min. Purified copper powder was stored in a desiccator.
0.60 g of CuI powder was dissolved in a *10 ml of moisture free acetonitrile
solution and stirred for 10 min to obtain a saturated solution of CuI. Then, it was
transferred into a flat bottom flask and 3 g of oxide-free refined copper powder was
284 A. W. M. V. Ekanayake et al.

added to the saturated CuI solution and stirred overnight in the dark. Finally, clear
colorless supernatant was separated using filter paper.
Dye-coated TiO2 electrodes were placed on a hot plate at 80 °C. Excess iodine free
CuI solution was spread over the exposed area of 0.25 cm2 while rest of the area was
covered with a mask. This procedure was repeated until the CuI layer was filled just
above the TiO2 surface while measuring the I-V characteristics for the best result. After
the deposition of CuI on the dye-coated TiO2 film, CuI layer on the working electrode
was sandwiched between the dyed TiO2 layer and the platinum counter electrode. Then
current-voltage (I-V) characteristics of DSSC at 1.5 AM (100 mW cm−2) simulated
sunlight irradiation were recorded using a calibrated solar-cell evaluation system.

3 Results

It indicates that, the prepared solid-state solar cells sensitized with D-149 show an
overall efficiency (η) of 2.11% for Iodine free CuI over CuI with Iodine. Short circuit
current density (Jsc) was 12.76 mA cm−2 for Iodide free CuI which was higher than
the short circuit current density of (Jsc) for CuI (Table 1 and Fig. 1).

Table 1. Comparison of parameters of DSSC constructed using Iodine free CuI and CuI for
D-149 dye
Hole conducting material Jsc/mA cm−2 Voc/V FF η/%
Iodine free CuI 12.7 0.475 0.35 2.11
CuI with Iodine 11.3 0.348 0.31 1.22
Jsc - Short circuit current density
Voc - Open circuit voltage
FF - Fill factor
η - Efficiency

▪ Iodine free CuI ▪ CuI with Iodine

Fig. 1. I-V characteristics of DSSC sensitized with D-149 at highest conversion efficiency
 Increasing the Efficiency of a Dye-Sensitized Solid-State Solar Cell 285

According Table 2 and Fig. 2, prepared solid-state solar cells sensitized with N-719
show an open circuit voltage (Voc) of 425 mV, Short circuit current density (Jsc) of
11.2 mA cm−2 and fill factor (FF) of 0.34 with an overall efficiency of 1.6% for Iodide
free CuI over pure CuI.

Table 2. Comparison of the variation of the solar cell parameters with the DSSC produced
using N-719 dye.
Hole conducting material Jsc/mA cm−2 Voc/V FF η/%
Iodine free CuI 11.20 0.425 0.34 1.60
CuI 6.23 0.416 0.43 1.11

▪ Iodine free CuI ▪ CuI with Iodine

Fig. 2. I-V characteristics of DSSC sensitized with N-719 at highest conversion efficiency

This shows that the removal of excess Iodine in CuI leads to an increase in current.
In other terms, the charge mobility of the Iodine free CuI is higher than CuI with
Iodine.
The SEM image in Fig. 3(b) shows Iodine free CuI crystal structure with a highly
porous structure since the size of the crystals have decreased as shown in the Fig. 3(b)
compared to Fig. 3(a). Since excess iodine blended with CuI crystals are now detached,
pure CuI crystal surfaces are free to attract holes and transport the charges efficiently
via the crystal chains as there is no insulating impurity to block the pathways of the
charges (Fig. 4).
Both the XRD patterns corresponding to CuI and Iodine free CuI give comparable
peak positions. Thus, it confirms that the removal of excess iodine in the system
preserved the crystalline structure of pure CuI.
286 A. W. M. V. Ekanayake et al.

(a) CuI crystals with Iodine

(b) Iodine free CuI crystals

Fig. 3. Scanning electron microscope images of CuI

 Increasing the Efficiency of a Dye-Sensitized Solid-State Solar Cell 287

Fig. 4. XRD spectrum of Iodine free CuI

4 Conclusion

Optimum efficiency (η) of 2.11% and the short circuit current density (Jsc) of
12.7 mA cm−2 were obtained for the DSSC built for Indoline dye (D-149) with excess
iodine eliminated CuI saturated acetonitrile solution.

References
1. Poizot P, Dolhem F (2011) Energy Environ Sci 4:2003–2019
2. Manahan SE (2008) Fundamentals of environmental chemistry, vol 3, pp 632–951
3. Suttie JM, McDowell RW (2008) Environmental impact of pasture-based farming, vol 45,
pp 1–3
4. Dincer I (2000) Renew Sustain Energy Rev 4:157–175
5. Carlson DE, Wronski CR (1976) Appl Phys Lett 28:671–673
6. Green MA (2004) Sol Energy 76:3–8
7. Nozik AJ (1978) Annu Rev Phys Chem 29:189–222
 Monitoring Volatile Fatty Acids
in an Anaerobic Process with Microbial
Fuel Cell

H. M. K. S. B. Thennakoon(&), S. M. W. T. P. K. T. Ariyarathne,
and M. Danthurebandara

Department of Chemical and Process Engineering, University of Peradeniya,

Peradeniya, Sri Lanka
[email protected]

Abstract. Volatile fatty acids (VFAs) concentration is a good indicator to

control and optimize the anaerobic digestion process. Though VFA concentra-
tion is a high sensitive parameter, most of the methods of measuring it are
complex and require time. Therefore this study is to find out a reliable, quick
method to monitor VFA concentration in the process. In this study, a microbial
fuel cell (MFC) was developed using synthetic samples of VFA. As the first part
of the study, the correlation between voltage and the acetic acid concentration in
the anaerobic digestion process was developed using developed MFC. Microbes
to digest the acetic acid were grown in the coir brush. That coir brush was used
as a bio film in the MFC. The salt bridge, which is made using 1 M KCl and
agar was used as a proton exchange membrane in the MFC. The relationship
was observed between voltage and acetic acid concentration with acetic acid
level from 0.09 M to 0.01 M.

Keywords: Microbial fuel cell
Volatile fatty acid
Salt bridge

1 Introduction

The growth of population with the growing level of modernization presents a signifi-
cant challenge on energy consumption. Eighty six percent of the global energy
requirement is covered using fossil fuels (Parkash 2015). But, it is a declining natural
resource and the burning of fossil fuels accounts for a lot of negative environmental
issues. Then, there is a high demand for the search for new materials and technology to
enhance power generation. Out of these, the energy resources of Bio fuels, solar,
nuclear, hydraulic, wind, geo thermal, tidal power, wave and ocean currents play the
major roles.
Currently, by going through various research directions, biogas has been identified
as a renewable and sustainable energy source since there are a lot of bio degradable
waste in Nature. But bio gas production is challenged by the instability of the
Anaerobic Digestion (AD) process. Therefore, it is not easy to maintain an optimum
biogas production and the overall anaerobic digestion process should be monitored to
reduce the instability of anaerobic digestion process. The following parameters: pH

© Springer Nature Singapore Pte Ltd. 2020

R. Dissanayake and P. Mendis (Eds.): ICSBE 2018, LNCE 44, pp. 288–294, 2020.
https://doi.org/10.1007/978-981-13-9749-3_27
 Monitoring Volatile Fatty Acids in an Anaerobic Process 289

level, Alkalinity, volume of biogas generation, concentration of Methane and CO2 will
determine the state of the process. However, several investigators have pointed out
disadvantages of these parameters as process indicators due to low sensitivity and
reliability. But researchers have found Volatile fatty acids (VFA) as intermediate
products of the fermentation stage. These are substrate of the methanogenesis stage and
more sensitive reliable parameters to monitor the AD process. Most researchers have
tried to use VFA as an indicator in anaerobic digestion systems. They have tried to
correlate current to VFA concentration. Also, traditional off line methods of VFA
quantitative measurement have been tried in last few years. Most of the methods of
measuring VFA concentration are complex and require more time. For example, Gas
chromatographic (GC), high pressure liquid chromatographic (HPLC), UV absorption
are advanced methods but depend on expensive instruments, skilled operators and so
on. Therefore, researchers have been moved to find easy and reliable methods to
measure the VFA concentration in the AD process. The literature so found explains a
way of using Microbial Fuel Cell (MFC) to monitor the VFA. Accordingly, it has been
found out that correlation can be developed between electrical signal that generates
from MFC and VFA concentration in AD process (Jin et al. 2016).
Literature reveals that there are researches on this area and they have experimented
to build a correlation between voltages generated from MFC with total volatile fatty
acid concentrations (Lee et al. 2015). But it has been identified that it is difficult to find
the exact cause for the instability of anaerobic digestion systems only with those
correlations. The instability depends on individual VFAs and monitoring of overall
VFAs generation would not help much. Correlation of individual fatty acid with
voltage may be more advantageous in finding the instability in the process. Therefore,
this study is to find out voltage variation with selected single acid concentration.

2 Materials and Method

Fabricating Microbial Fuel Cell

MFC was fabricated according to the Method given by Parkash in 2015 (Parkash
2015).
Two commonly available (500 ml) plastic containers were selected as the cathode
and anode chambers of the MFC. Two holes of diameter 6 mm and 1.5 mm were made
on each of the lids to insert the salt bridge and electrodes. The anode chamber was filled
with 300 ml of anaerobic digestion solution. The cathode chamber was filled with
300 ml 0.05 mM NaCl solution. Only the anode chamber was closed tightly and sealed
with tape. Anode and cathode electrodes were connected using copper wire through a
millimeter.
Electrodes
Graphite rods of 10 cm length and 0.5 mm diameter was used as electrodes. Initially, a
carbon rod which is inside the HB pencil was used as the graphite rod in the cell. The
selected graphite rods were heated in 100% Ethanol for 30 min to remove the impu-
rities (Parkash 2015). Then, the electrodes were washed using 1 M hydrochloric acid
290 H. M. K. S. B. Thennakoon et al.

followed by 1 M Sodium hydroxide, each for 1 h to neutralize. Electrodes were stored

in distilled water until use.
Anode Solution
In order to obtain a relationship among VFA concentration and generated voltage a
solution of acetic acid with 0.1 M was used.
Bio Film Preparation
Firstly, discarded vegetables and fruits were collected from a stall. It was blended
together to form a slurry with a domestic blender. Following the method carried out by
Prakash in 2016, a sugar solution was added to the slurry to make it nutrient rich
(Parkash 2015). Then, anaerobic bacteria sample collected from wastewater treatment
plant at Kandy Citadel hotel was added to the slurry. Attached growth anaerobic
bacteria were cultivated by dipping coir brush to the mixture. The coir brush was kept
in the slurry under anaerobic condition for enrichment of bacteria for 2 months. The
enriched coir brush (bio film) was set aside to starve for about 1 day (Jin et al. 2016).
Salt Bridge Preparation
Salt bridge was made with 5 mm diameter level tube as shown in Fig. 1. The salt
bridge was prepared using 1 M Potassium chloride with 5% Agar. Initially 7.43 g of
KCl was added to the 100 ml of distilled water to prepare a 1 M KCl solution. Then
0.5 g of agar powder was added to the KCl solution and mixture was boiled for 10 min.
Then, the mixture was sucked into the level tube and allowed to solidify. After
solidifying this salt, the bridge was connected to the corresponding MFC and sealed
with tape as shown in Fig. 1.

Fig. 1. Salt bridge connection experiments

Experiment 1
The bio film (Coir brush) and 0.1 M acetic acid 300 ml was added to the anode
chamber and it was sealed properly using tape (Parkash 2015). Then, the anode and
cathode electrodes were connected using copper wire through a millimeter. After that,
the generated voltage was recorded with 1 h time intervals and the corresponding acetic
acid concentrations in the anode chamber were measured using a titration method
(Titrations.info 2018).
 Monitoring Volatile Fatty Acids in an Anaerobic Process 291

Then, measured acetic acid concentration was plotted against the generated voltage
and the relation between them was obtained.
Experiment 2
Experiment 1 was repeated with duplicates. In this experiment, generated voltage and
the acetic acid concentration was recorded in 30 min intervals as in the first trial.

3 Results and Discussion

Experiment 1
Table 1 shows the gathered result of 1st experiment. In experiment 1, the MFC was run
only for 4 h. During this time period only 3 readings were able to be taken. With these
observation, the correlations of generated voltage and acetic acid concentration were
developed and the relevant graph is shown in Fig. 2. There were not enough obser-
vations that could be collected to come to a conclusion and therefore experiment 2 was
designed by minimizing the errors of experiment 1.

Table 1. Results of Experiment 1

Hours pH Voltage/mV Volume of 0.01 M NaOH Concentration/M
sample/ml volume/ml
1 4.15 0.3 9.8 28 0.0286
3 4.05 0.1 9.4 18 0.0191
4 3.7 0 9.4 13 0.0138

Fig. 2. Acetic acid concentration vs voltage in Experiment 1

292 H. M. K. S. B. Thennakoon et al.

Table 2 shows the errors encounter in experiment 1 and the remedies that were
followed to develop the experiment 2.

Table 2. Errors encountered in Experiment 1 and remedies taken to develop

Errors encountered Remedies
No voltage after 4 h. Therefore couldn’t take more Planned to take readings in
readings by taking reading in 1 h time intervals 30 min time interval
No replicates to compare the results Planned to duplicate the
experiment

Experiment 2
Table 3 shows the result of experiment 2.

Table 3. Results of Experiment 2

Hours pH Voltage/mV Volume of 0.01 M NaOH Concentration/M
sample/ml volume/ml
Sample 1
0.5 3.65 0.1 10 52.5 0.0525
1 3.75 0 10 50.5 0.0505
1.5 3.91 0 10 50.3 0.0503
Sample 2
0.5 3.44 0.1 10 52 0.0520
1 3.65 0 10 50.3 0.0503
1.5 3.7 0 10 50.2 0.0502

In this trial, the generated voltage has been reduced within 1 h. Concentration also
has not been changed after 1 h. Relevant graphs obtained from experiment 2 are shown
Fig. 3.

Causes to Obtain Different Operating Times of MFC’s

Sudden pH Drop
In the first trial pH of bio film was 4.88 and the pH of the three samples were 4.18, 4.05
and 3.70. There was no significant pH difference that could be observed with the
samples and the bio film generated. But, in the second trial, the pH of bio film was 4.72
and the pH of the three samples was the same around 3.7. There were significant pH
differences between these two and it may suppress the microbial activity in the unit.
 Monitoring Volatile Fatty Acids in an Anaerobic Process 293

Fig. 3. Acetic acid concentration vs voltage in Experiment 2

Reduction of Microbes in Coir Brush and Microbes Decay with Time

Figure 4 explains the microbe growth pattern with time. According to that, it follows
four phases as lag, log, stationary and death. Time between 1st and 2nd experiment is
around 2 weeks. It can be assumed that coir brush which used for 1st experiment has
taken out from the slurry (Mentioned in bio film preparation) when it was in log or
stationary phase. At that time microbes were very strong and active. It could have
potential to resist the sudden pH drop. After two-week time microbes in slurry could be
in death phase. Therefore, are not rich enough to resist sudden pH drop. Therefore,
microbes could be under stress when it exposes to the environment which have low pH
value.

Fig. 4. Cell growth vs time

294 H. M. K. S. B. Thennakoon et al.

4 Further Developments

This research is a preliminary study to verify the method that has been developed using
previous researches. Because of this, only acetic acid was used to develop the corre-
lation. But there were many experimental difficulties identified with the study. In this
present study, two trials were completed and the problems encountered identified.
Future studies were planned to repeat the experiments as follows with remedies taken
to overcome the problems.
Observe the voltage variation vs different initial acetic acid concentrations at
10 min intervals.
At that next step, the three main VFA with the following concentration will be fed
into the anode chamber.
Acetic acid - 5431 mg/L
Butyric acid - 6014 mg/L
Propionic acid - 2248 mg/L (Lee et al. 2015).
Take ammeter readings at 30 min intervals and corresponding individual VFA
concentrations will be taken using gas chromatography.

References
Titrations.info (2018) Determination of acetic acid in vinegar by titration. http://www.titrations.
info/acid-base-titration-acetic-acid-in-vinegar. Accessed 9 Feb 2018
Jin X, Angelidaki I, Zhang Y (2016) Microbial electrochemical monitoring of volatile fatty acids
during anaerobic digestion. Environ Sci Technol 50(8):4422–4429. https://doi.org/10.1021/
acs.est.5b05267
Lee DJ et al (2015) Effect of volatile fatty acid concentration on anaerobic degradation rate from
field anaerobic digestion facilities treating food waste leachate in South Korea. J Chem.
https://doi.org/10.1155/2015/640717
Parkash A (2015) Design and fabrication of a double chamber microbial fuel cell for voltage
generation from biowaste. J Bioprocess Biotech. https://doi.org/10.4172/2155-9821.1000246
 Embodied Carbon Reduction Strategies
for Buildings

L. M. T. Kumari, U. Kulatunga(&), N. Madusanka, and N. Jayasena

Department of Building Economics, University of Moratuwa,

Moratuwa, Sri Lanka
[email protected]

Abstract. Global warming has become a major environmental issue faced by

the international community. Greenhouse gases (GHG) have been suspected to
be the source of global warming and carbon emission has identified as the main
contributor for global warming. CO2 could be released due to activities carried
out by human such as burning fossil fuels, electricity consumption and trans-
portation. Hence, the carbon footprint is the total amount of greenhouse gases
produced directly or indirectly as a result of an activity. Up to 40% of all energy
have been consumed by the building sector and it has contributed up to 30% of
global annual GHG emissions. Therefore operational and embodied carbon were
identified as two main general groups of carbon emissions related to buildings.
Numerous researchers have paid their attention on operational carbon reduction
and there are limited study on embodied carbon reduction. Therefore the
researcher aims to identify embodied carbon reduction strategies for buildings.
After a comprehensive literature review, selection of low carbon materials at the
design stage of the building, reuse and recycling of carbon intensive materials,
material minimization, optimum building design, local sourcing of materials,
transportation minimization, efficient construction processes, policies and reg-
ulations by government, adaptive reuse of buildings and carbon labelling
schemes were identified as the global embodied carbon reduction strategies for
buildings.

Keywords: Embodied carbon
High-rise buildings
Strategies

1 Introduction

Global warming has become a significant environmental issue and, it is considered as

one of the biggest challenges faced by the community (Lai et al. 2012). Methane,
carbon dioxide, ozone, water vapour, nitrous oxide, and halogenated hydroxide have
been identified as greenhouse gases (GHG) that are suspected to be the sources of
global warming (Gajewski et al. 2013). According to Ozkaymak et al. (2017), CO2 has
been recognized as the main contributor for global warming and Gajewski et al. (2013)
have revealed that CO2 accounted for 9–26% of greenhouse effect.
Material extraction, material processing and component fabrication, construction
and assembly, operation and service phase and end-of-life phase have been identified as

© Springer Nature Singapore Pte Ltd. 2020

R. Dissanayake and P. Mendis (Eds.): ICSBE 2018, LNCE 44, pp. 295–308, 2020.
https://doi.org/10.1007/978-981-13-9749-3_28
296 L. M. T. Kumari et al.

the different life cycle stages of a building and energy use and carbon emissions can
occur in each stage (Akbarnezhad and Xiao 2017).
According to RICS information by Lockie and Berebecki (2012), operational
carbon (OC) and embodied carbon (EC) have been identified as the two main general
groups of carbon emissions. Kang et al. (2015) recognized that the CO2 emitted from
the buildings from the use of heating, cooling and lighting as the operational carbon of
a particular building. Kang et al. (2015) have further revealed that, CO2 emitted during
manufacturing, transportation and construction of building materials as the embodied
carbon.
According to Crawford (2011), the lack of consideration and effort have been
directed towards embodied carbon emissions occurring from construction materials
while significant effort has been placed on measuring and reducing the impact of
emissions occurring in the building operational stage. Further, Crawford (2011) has
recognized that major impact is created by embodied emissions of the building and that
it accounts for more than 50% of whole the life impacts of buildings towards global
GHG emissions.
According to Wheating (2017), the building sector has been recognized as the
major contributor for GHG emissions and the building sector has been identified as the
sector with major opportunities for emission reduction strategies and mitigation of
GHG emissions that contribution to climate change.
Most building owners tend to reduce operational carbon of the building in the
operational stage with the aim of reducing the operational cost by using different
strategies to save energy. However, there are limited previous studies for the reduction
embodied carbon in buildings (Pomponi and Moncaster 2016). This situation has given
rise to the need to conduct an analysis to identify suitable embodied carbon reduction
strategies for buildings.
Therefore, the aim of this research paper is to review embodied carbon reduction
strategies for buildings.

2 Research Methodology

A comprehensive literature survey was carried out to identify the concept of carbon
footprint, carbon emission in building life cycles, embodied carbon emission and global
embodied carbon reduction strategies.

3 Carbon Emissions in Building Life Cycle

Typically, a building life cycle could be divided into stages such as material manu-
facturing, transportation of building materials, construction of the building, operation
and maintenance and demolition (Gustavsson et al. 2010). Further, according to the
views of Monahan and Powell (2011), energy and carbon emissions related to these
stages could be considered.
 Embodied Carbon Reduction Strategies for Buildings 297

According to the review of the building life cycle assessment by Ibn-Mohammed

et al. (2013), it has been proven that embodied carbon can account for anywhere
between 2% and 80% of whole-life carbon emissions.

4 Embodied Carbon Emission

Embodied carbon emission is identified as the energy consumption associated with the
mining, manufacturing, transportation, assembly, replacement and deconstruction of
construction materials or products and generally is expressed in kilograms (Kg) of
CO2e per kilogram of product or material (Lockie and Berebecki 2012). Further authors
stated that, all the activities relevant to the construction and demolition of a building
will result in embodied emission. According to the Embodied Carbon Assessment,
carbon footprint of a material is defined as embodied carbon (Anon n.d.). Embodied
carbon can be measured from cradle-to-gate, cradle-to-site, and cradle-to-end of con-
struction, cradle-to-grave, or cradle-to-cradle (Lockie and Berebecki 2012). Further,
Lockie and Berbecki (2012) stated that, cradle-to-gate is the typical embodied carbon
data sets.
Initial embodied emission and recurring embodied emission are identified as the
two main categories of embodied emissions (Ramesh et al. 2010; Chen et al. 2001).
Ding (2004) defined the initial embodied emission as the emissions associated with the
acquisition of raw materials, their processing stage, transportation to the site and
construction. Further, Ding (2004) stated that the two major categories of initial
embodied emissions could be identified as direct emissions and indirect emissions.
Accordingly, the emissions associated with the transportation of building materials or
products to the construction site and the emissions related to the construction of the
building were revealed as direct initial embodied emissions. According to Chen et al.
(2001), the emissions incurred with the acquiring, processing and manufacturing of
building materials including transportation linked to these activities were defined as the
indirect initial embodied emission of a building.
Different types of materials are being used in building construction and various
materials have a low life span compared to the building life span. Therefore, new
materials are required to be replaced at the operation stage of the building (Ramesh
et al. 2010). Accordingly, regular annual maintenance should be carried out in the
building for its proper operation. Therefore, the emissions incurred with building
maintenance, repair, restore, refurbishments or replacement of a system, components or
a system during the effective life of the building were defined as the recurring
embodied emissions (Ramesh et al. 2010; Chen et al. 2001; Ding 2004). Ibn-
Mohammed et al. (2013), stated of recurring embodied emissions as the total emissions
associated with the materials used in the rehabilitation and maintenance.
298 L. M. T. Kumari et al.

5 Importance of Embodied Carbon Reduction

According to Dutil et al. (2011) the importance of embodied carbon emission has
increased and further, (Anon n.d.) mentioned that the construction sector has con-
tributed for a large portion of embodied emissions. Due to this, embodied carbon has
gained higher attention from both the industry and the government. Reducing
embodied energy, energy consumption of buildings and switching to low carbon fuels
were identified as the measures to reduce GHG emissions in buildings (IPCC 2007).
According to Monahan and Powell (2011), most buildings have used less energy in
their operational phase and have a lower GHG emission over the life span of the
building. Therefore, a considerable amount of embodied carbon emissions have been
included in the whole life carbon emission of the building and the importance of the
GHG emissions associated with the building materials has increased.
In addition to operational emission reductions, the green construction board low
carbon route map for the built environment defined that, embodied carbon reduction by
39% in 2050 is essential compared to the 2010 baseline (Green Construction Board
2013). Therefore, nine strategies were identified through a comprehensive literature
review.

6 Global Strategies Used for Embodied Carbon Reduction

6.1 Use of Low Energy and Low Carbon Materials

According to the study carried out by Monahan and Powell (2011), cement, sand,
gravel and concrete products were identified as the types of materials, which includes
minerals. Further, minerals were identified as the most important material types, which
was accounted for 45% of material related embodied carbon. High or low embodied
carbon footprint of a building is decided according to the amount of carbon that exists
in the selected building material (Wheating 2017). According to Monahan and Powell
(2011) up to 82% of the total amount of embodied carbon is found in construction
materials. Further, Strain, Simonen, Yang, and Webster (as cited in Wheating 2017)
stated that building materials with highest embodied carbon have high energy con-
sumption during their production. Depending on the type and amount of energy used
for the building materials’ manufacturing processes, the embodied carbon of a par-
ticular building material would increase (Gonzalez and Navarro 2006). Kwok et al.
(2016) revealed that embodied carbon of building materials is controllable at its pro-
duction sources.
Globally 60% of raw materials are being used for civil works and building con-
struction. Out of this volume, the construction sector has consumed 40% and it con-
tributes 40%–50% GHG emissions (Huang et al. 2017; Bribian et al. 2011). Gonzalez
and Navarro (2006) found that a varied selection of materials with low environmental
impact could reduced up to a maximum of 30% of carbon emissions. After evaluating
embodied carbon of several office buildings constructed with different building mate-
rials, the study revealed that the highest portion of embodied energy in these buildings
relates to structural materials such as concrete and reinforced steel which have
 Embodied Carbon Reduction Strategies for Buildings 299

accounted for 50%–66% of total embodied energy of a building (Akbarnezhad and

Xiao 2017).
Therefore, according to the literature, Concrete could be identified as the major and
widely used construction material for buildings and other infrastructures (Chen and Ng
2015). Further, Imbabi et al. (2012) revealed that 9 kg of CO2 will be released as a
byproduct of 1 kg of ordinary Portland cement. According to Imbabi et al. (2012), the
construction industry has investigated alternative materials, which do not produce CO2
in the process, to replace the raw materials.
Supplementary cementation materials (SCM) which are the byproducts of other
industries could be used for restructuring the concrete mixes with the aim of reducing
the carbon footprint of concrete (Lee et al. 2011). According to Imbabi et al. (2012), fly
ash (a byproduct of coal combustion), blast furnace slag (a byproduct from iron
smelting) and silica fumes (a byproduct of silicon production) are the best common
identified SCM substitutes for traditional mixes.
Application of alternative fuels for the processing stage of cement could be iden-
tified as another strategy used to reduce the carbon footprint (Imbabi et al. 2012;
Higuchi et al. 2014). Biomass is identified as an alternative fuel for cement clinker kilns
where kilns have been traditionally fired using coal, petroleum coke, gas or fuel oil.
Moreover, the author stated that emission from a biomass could reduce 20%–25% of
the emissions from coal combustion (Imbabi et al. 2012).
There is a new strategy to convert production emissions into an aggregate and form
artificial limestone aggregates which could be used as SCMs to reduce carbon emis-
sions in a carbon emitting facility. This has been identified by (Blue Planet Company n.
d.). CO2 could be captured through the extension attached to the exhaust of the power
plant and the cement manufacturing facility and finally, it is converted into calcium
carbonate mineral. Moreover, they have stated that this strategy could tackle the two
largest causes of embodied emissions in concrete, which could reduce the emissions
from cement production and emissions from the burning of fuel.
According to Zhong and Wu (2015), concrete and steel have been identified as the
most common building materials which frequently account for the highest proportion
of materials in the building. Further, Chen et al. (2001) mentioned that steel and
aluminium have accounted for more than three quarters of total embodied emission of
the building. Compared to reinforced concrete and composite materials, structural steel
is identified as a more carbon intensive material (Gan et al. 2017).
When compared to the masonry cavity structure, a timber frame has provided
lighter weight (Monahan and Powell 2011). Further to the authors, materials for the sub
structural supports could be reduced due to the lighter structure and accordingly, the
foundation materials and high embodied building materials such as concrete and steel
reinforcing could be reduced. There is low carbon emission in bamboo structures
compared to brick concrete buildings (Yu et al. 2011).
According to studies carried out by Reddy (2009), stabilized mud blocks
(SMB) could be an alternative for burnt clay bricks and it is a product of compacting a
mixture of sand, soil, water and a stabilizer such as cement or limestone. Moreover,
Reddy (2009) has mentioned that, these blocks could be used for wall construction and
these are eco-friendly and energy efficient. Due to the fact that construction waste such
as stone quarry dust and fly ash could be utilized in manufacturing process of SMB,
300 L. M. T. Kumari et al.

there is no need to burn them in a clinker kilns, and it is possible to save 60%–70% of
energy compared to burnt clay brick and it incurs less transportation emission because,
these SMD could be easily produced at the construction site itself.
Ortiz et al. (2009) investigated that timber, glass, aluminum, concrete, slate, plas-
terboard, ceramic tiles, damp course and mortar are in a single family house and
defined that concrete, timber and ceramic tiles have embodied emission of 61%, 14%,
and 15% respectively. Further, Rai et al. (2011) mentioned that, using timber instead of
steel cladding in distribution warehouse could reduce CO2 emissions by 18%.

6.2 Reduce, Reuse and Recycling of Carbon Intensive Construction

Materials
Metal, masonry, carpet, concrete, acoustic tiles, ceramic tiles and insulations are
identified as the commonly used materials which have recycled content and these
materials have provided an opportunity to reduce carbon and energy emissions in the
initial phase of a building (Yu et al. 2011; Saghafi and Teshnizi 2011). Recycling of
construction materials could be identified as the highest embodied emission saving
potential, which accounts for 53% and reusing of construction materials has 6.2%
potentials of embodied emission reduction (Ng and Chau 2015).
Therefore, recycling and closing the materials loop are being identified as the
efficient and effective strategies for reduce GHG emissions of a building (Saghafi and
Teshnizi 2011). Further, authors have mentioned that converting already made products
consumes less energy than obtaining raw materials from scratch and producing a totally
new product. According to Ng and Chau (2015), 2%–3% of GHG reduction could be
gained by increasing the recycling rate of concrete from 27%–50% from deconstructed
buildings. According to the studies carried out by Thormark (2002), it was found that
35%–40% of embodied energy could be save through recycling building waste.
Concrete, steel, wood and sand are the main construction and demolition waste that
could be found in building constructions (Thongkamsuk et al. 2017).
There is less volume of waste generated in precast and prefabricated structures due
to designs that are detailed and manufacturing process has generated a less volume of
waste (Thongkamsuk et al. 2017; Martos et al. 2018). Moreover, Martos et al. (2018)
have explained that the amount of construction and demolition waste generated in
reinforced or concrete structures vary compared to the prefabricated structures.
Chau et al. (2012) mentioned that reduction in quantity of material used and
construction waste could be achieved through maintaining the existing building ele-
ments. This could be reduced CO2 emission of the building. Further Chau et al. (2012)
have revealed that, maintaining existing walls, roofs and floors could maximize the
benefits of material reduction and construction waste reduction by reusing the whole
building. Moreover, the authors have revealed that, properly maintained and stored
construction materials and components from previously demolished from sites could be
recovered for the current project site. Therefore, the embodied emissions could be
reduced due to less construction materials.
Basbagill et al. (2013) defined that optimizing key parameters anywhere in the
building such as thickness of piles and footings, external and internal walls accounts for
63% to 75% reduction in total embodied impact of the building.
 Embodied Carbon Reduction Strategies for Buildings 301

Therefore, reusing and recycling building materials was considered during the
primary evaluation for installation on a building (Ng and Chau 2015). Further to
authors, at the end of life span of an initial recycled value and recyclability product
could result in considerable reductions of embodied carbon value throughout the total
lifecycle of a building.

6.3 Material Minimization

The usage of more materials results in the highest carbon footprint of a building
because, embodied carbon footprint of a building is directly proportionate to the
quantity of materials used in that building (Akbarnezhad and Xiao 2017). Therefore,
according to Akbarnezhad and Xiao (2017), the reduction in the quantity of materials
used would be quick method for embodied carbon reduction in a building. Further,
authors have mentioned that it is necessary to measure the total amount of materials at
the design stage without comparing materials with their embodied carbon value sep-
arately. According to Nadoushani and Akbarnezhad (2015), the type of construction
materials used, height of the structure and selected structural system for the building
were identified as the main factors which could affect the total amount of materials used
in a building structure.
Yeo and Gabbai (2011) revealed that avoiding overdesign and use of optimum
design results in a significant amount of reduction in the total amount of materials used
and therefore, the total embodied carbon of the building could be reduced. Further,
authors have mentioned that, minimization of material usage should be accomplished
while retaining the capability of the structure to encounter all other performance and
technical requirements.
According to studies carried out by Rashid and Yusoff (2015), they have identified
that precast concrete floors are capable of having an extended life between beams.
Therefore, the total amount of concrete used in precast concrete buildings could be
reduced by using less amount of columns and footing needed for the building.
According to studies carried out by Lopez-Mesa et al. (2009), it was identified that a
considerable amount of embodied emission could be reduced by using hollow core
concrete for load bearing constructions than using solid reinforced.

6.4 Optimizing Building Design

According to Iwaro and Mwasha (2013), sustainable development considerations have
to be considered in the early design stage because the highest impact to the final impact
of a building takes place at the early design stage. Therefore, construction of sus-
tainable buildings could be appropriately integrated with this concept (Hakkinen et al.
2015). Further, Hakkinen et al. (2015) have mentioned that clear goals could be rec-
ognized during the design stage for environmental impact restrictions and building
performance, which could be guided in the material selection process. Dixit et al.
(2010) revealed that the selection of low energy and carbon intensive materials and
products in the early stage of a building design has resulted in the reduction of
embodied emission of a building. According to Shi and Yang (2013), the decisions
302 L. M. T. Kumari et al.

taken in the conceptual design phase has the greatest impact on the total building
performance.

6.5 Local Sourcing of Materials and Transport Minimization

Transportation activities have accounted for 16.5% of the total embodied carbon
emissions in the construction of buildings (Serrano and Alvarez 2016). The transport
requirements of a construction site is determined according to the selection of the
location (Akbarnezhad and Xiao 2017). Moreover to the authors, according to the
location, the transportation distance and transportation mode would be changed.
Numerous researchers have stated that increasing local material usage would reduce
the embodied carbon emission through the reduction of transportation impacts
(Gustavsson et al. 2010; Asdrubali et al. 2015; Chou and Yeh 2015). According to
Gonzalez and Navarro (2006), the mode of transport, number of trips, country of origin
and distance to be traveled have been identified as the main factors affecting transport
emission of materials. It is not enough to reduce transport distance to achieve low
carbon emission, the mode of transportation has a greater impact on the amount of
embodied carbon in construction material (Weight et al. n.d.).
According to Hammad et al. (2016), site layout planning problem (SLPP) and
container loading problem (CLP) have been identified as the two main problems related
to carbon reduction in transportation. Under SLPP, optimization of tower crane loca-
tions on construction sites would account for low carbon emissions due to minimiza-
tion of their operations. Further to Hammad et al. (2016), all the off-site transportation
activities related to moving construction materials from manufacturers to suppliers and
from suppliers to construction sites have considered under CLP and an optimized
framework has been proposed in order to reduce embodied carbon emissions in
constructions.

6.6 Policies and Regulations by Government

Policies and regulations implemented or reviewed by the government could be iden-
tified as another frequently cited strategy for the reduction of embodied carbon emis-
sions (Dakwale et al. 2011; Foraboschi et al. 2014). Giesekam et al. (2015) mentioned
that, this strategy has supported the other reduction strategies. For example, Giesekam
et al. (2015) revealed policies related to the broader use of low embodied carbon
materials. According to the studies carried out for China and Japan by Dhakal (2010),
due to the influence of policies and regulations on design and construction practices
50% of CO2 could be reduced.

6.7 Efficient Construction Processes and Techniques

Increase of efficiency in the construction sector was identified as a vital opportunity for
embodied carbon reduction (Monahan and Powell 2011; Upton et al. 2008). Further,
the efficiency in building materials manufacturing, use of new construction processes,
reducing wastages during the construction stage or the combination of all these things
have been identified as strategies that would enhance the efficiency of the construction
 Embodied Carbon Reduction Strategies for Buildings 303

process. According to Sandanayake et al. (2016), carbon emissions from materials,

equipment usage and transportation were recognized as sources of emissions and
finally concluded that emissions from equipment usage and transportation would
contribute to the highest portion of embodied emission in construction of foundations.
Therefore, concentrating on equipment usage and transportation could reduce the
emissions in the foundation of constructions.
According to Junnila and Horvath (2003), the best practices in operation of con-
struction equipment and use of temporary construction materials such as formwork,
temporary structures which are not attached to the building permanently were identified
as the opportunities to reduce embodied carbon emission of constructions. Selection of
optimum equipment for the construction process, reduction in idle time of construction
equipment to enhance the construction process, enhancement of the operation of
equipment and reduction of both horizontal and vertical transportation within the site
were identified as approaches to be used in order to reduce embodied carbon emission
in the construction phase (Lewis et al. 2009; Avetisyan et al. 2012). According to the
studies carried out by Carmichael et al. (2012), optimum earthmoving operations has
resulted in the reduction of embodied carbon emission related to the construction
process.
Nadoushani et al. (2017) revealed that the idle time of vehicles (trucks) and pumps
could be reduced through optimizing the concreting operation on construction sites in
order to reduce carbon emission occurred in the construction. Further, Nadoushani
et al. (2017) have mentioned that a significant amount of emission could be reduced
with the optimum use of cranes in the construction site. Operation time of the crane
could be minimized by enhancing the places of supply and demand points in the
construction site to reduce carbon emission.

6.8 Adaptive Reuse of Buildings

Moving to reusing and adapting of buildings has become a growing style in last decade
(Wilkinson et al. 2009; Bueren and Jong 2007; Bradley and Kohler 2007). According
to the Department of the Environmental and Heritage (2004), adaptive reuse was
defined as a process that converts unproductive items into a new product which could
be used for an altered purpose. Buildings could contribute to climate change and
sustainability via the reduction of carbon emission by adapting this adaptive reuse
process (Bullen 2007). Moreover to previous author, reduction of material usage,
transportation and energy consumption could be achieved through extending the life
span of a current building. It has contributed to the enhancement of the sustainability
and carbon reduction of a building.

6.9 Carbon Labelling Schemes

Carbon labelling schemes are designed for construction materials in order to reduce the
impact to the global climate (Wu et al. 2014). Further, according to Wu et al. (2014),
appraising the inputs of raw materials, energy demand and emission to air as an output
of production of a product or a service, consists in the total labelling process and
environmental impact of a product or a material is represented in a single score.
304 L. M. T. Kumari et al.

Wu et al. (2015) revealed that this carbon labelling scheme is only focused on global
climate change. According to Boardman (2008), to provide information on low carbon
products to consumers to take correct decisions on their selections, to inspire sellers to
sell non carbon intensive products, and to provide a method for organizations to be
obligated in the reduction in embodied carbon emissions of a product are the three main
aims of this carbon labelling schemes (Fig. 1).

Fig. 1. Carbon reduction strategies for buildings

 Embodied Carbon Reduction Strategies for Buildings 305

7 Conclusion

Carbon dioxide, methane, hydro fluorocarbon, per fluorocarbon, nitrous oxide and
sulphur hexafluoride are the six main GHG exists in the atmosphere. These GHGs have
increased due to activities carried out by humans. Therefore, the carbon footprint is
defined as the total amount of GHGs produced directly or indirectly as a result of an
activity. The researcher focused on a building’s carbon emission and there are two
general groups of carbon emission related to buildings; operational emission and
embodied emission. According to the research gap, priority was given to embodied
carbon in buildings and initial and recurring embodied carbon was identified as the two
major categories of embodied carbon.
The use of low energy and low carbon materials, reuse, recycle and reduce of
carbon intensive construction materials, materials minimization, optimizing building
design, local sourcing of materials and transport minimization, policies and regulations
by government, efficient construction process and techniques, adaptive reuse of
buildings and carbon labelling schemes are identified as the main global embodied
carbon reduction strategies.

Acknowledgement. Authors express the gratitude to each and every individual for their
encouragement, values and ideas, assistance and specially their commitment towards this
research to make it a success.
Further, authors would like to acknowledge the technical assistance given by Miss
L A Siriwardena, Technical Officer (Grade 1), Department of Building Economics, University of
Moratuwa.

References
Akbarnezhad A, Xiao J (2017) Estimation and minimization of embodied carbon of buildings: a
review. Buildings 7(4):5
Anon (n.d.) Embodied carbon assessment. Circular Ecology. http://www.circularecology.com/
embodied-carbon.html. Accessed 07 Feb 2018
Asdrubali F, Baldinelli G, D’Alessandro F, Scrucca F (2015) Life cycle assessment of electricity
production from renewable energies: review and results harmonization. Renew Sustain
Energy Rev 42:1113–1122
Avetisyan HG, Miller-Hooks E, Melanta S (2012) Decision models to support greenhouse gas
emissions reduction from transportation construction projects. J Constr Eng Manag 138
(5):631–641
Basbagill J, Flager F, Lepech M, Fischer M (2013) Application of life-cycle assessment to early
stage building design for reduced embodied environmental impacts. Build Environ 60:81–92
Blue Planet Company (n.d.) Economically sustainable carbon capture. http://www.blueplanet-ltd.
com/?page_id=54
Boardman B (2008) Carbon labelling: too complex or will it transform our buying? Significance
5(4):168–171
Bradley PE, Kohler N (2007) Methodology for the survival analysis of urban building stocks.
Build Res Inf 35(5):529–542
306 L. M. T. Kumari et al.

Bribian IZ, Capilla AV, Uson AA (2011) Life cycle assessment of building materials:
Comparative analysis of energy and environmental impacts and evaluation of the eco-
efficiency improvement potential. Build Environ 46(5):1133–1140
Bullen PA (2007) Adaptive reuse and sustainability of commercial buildings. Facilities 25
(1/2):20–31
Bueren EV, Jong JD (2007) Establishing sustainability: policy successes and failures. Build Res
Inf 35(5):543–556
Carmichael DG, Williams EH, Kaboli AS (2012) Minimum operational emissions in
earthmoving. In: Construction research congress
Chau C, Hui W, Ng W, Powell G (2012) Assessment of CO2 emissions reduction in high-rise
concrete office buildings using different material use options. Resour Conserv Recycl
61:22–34
Chen T, Burnett J, Chau C (2001) Analysis of embodied energy use in the residential building of
Hong Kong. Energy 26(4):323–340
Chen Y, Ng ST (2015) Integrate an embodied GHG emissions assessment model into building
environmental assessment tools. Proc Eng 118:318–325
Chou J-S, Yeh K-C (2015) Life cycle carbon dioxide emissions simulation and environmental
cost analysis for building construction. J Clean Prod 101:137–147
Crawford RH (2011) Life cycle assessment in the built environment. Spon Press, London
Dakwale VA, Ralegaonkar RV, Mandavgane S (2011) Improving environmental performance of
building through increased energy efficiency: a review. Sustain Cities Soc 1(4):211–218
Department of the Environmental and Heritage (2004) Adaptive reuse: preserving our past,
building our future, s.l.: s.n
Dhakal S (2010) GHG emissions from urbanization and opportunities for urban carbon
mitigation. Curr Opin Environ Sustain 2(4):277–283
Ding GK (2004) The development of a multi-criteria approach for the measurement of
sustainable performance for built projects and facilities. University of Technology, Sydney
Dixit MK, Fernandez-Solis JL, Lavy S, Culp CH (2010) Identification of parameters for
embodied energy measurement: a literature review. Energy Build 42(8):1238–1247
Dutil Y, Rousse D, Quesada G (2011) Sustainable buildings: an ever evolving target.
Sustainability 3(2):443–464
Foraboschi P, Mercanzin M, Trabucco D (2014) Sustainable structural design of tall buildings
based on embodied energy. Energy Build 68:254–269
Gajewski A, Siergiejuk J, Szulborski K (2013) Carbon dioxide emission while heating in selected
European countries. Energy Build 65:197–204
Gan VJL et al (2017) Sustainability analyses of embodied carbon and construction cost in high-
rise buildings using different materials and structural forms. HKIE Trans 24(4):216–227
Giesekam J, Barrett JR, Taylor P (2015) Construction sector views on low carbon building
materials. Build Res Inf 44(4):423–444
Gonzalez MJ, Navarro JG (2006) Assessment of the decrease of CO2 emissions in the
construction field through the selection of materials: practical case study of three houses of
low environmental impact. Build Environ 41(7):902–909
Green Construction Board (2013) Low carbon routemap for the UK built environment, s.l.: s.n
Gustavsson L, Joelsson A, Sathre R (2010) Life cycle primary energy use and carbon emission of
an eight-storey wood-framed apartment building. Energy Build 42(2):230–242
Hakkinen T, Kuittinen M, Ruuska A, Jung N (2015) Reducing embodied carbon during the
design process of buildings. J Build Eng 4:1–13
Hammad AWA, Akbarnezhad A, Rey D, Waller ST (2016) A computational method for
estimating travel frequencies in site layout planning. J Constr Eng Manag 142(5):04015102
 Embodied Carbon Reduction Strategies for Buildings 307

Higuchi T, Morioka M, Yoshioka I, Yokozeki K (2014) Development of a new ecological

concrete with CO2 emissions below zero. Constr Build Mater 67:338–343
Huang W et al (2017) Carbon footprint and carbon emission reduction of urban buildings: a case
in Xiamen City, China. Proc Eng 198:1007–1017
Ibn-Mohammed T et al (2013) Operational vs. embodied emissions in buildings – a review of
current trends. Energy Build 66:232–245
Imbabi MS, Carrigan C, McKenna S (2012) Trends and developments in green cement and
concrete technology. Int J Sustain Built Environ 1(2):194–216
IPCC (2007) Climate change 2007: mitigation of climate change Working Group III contribution
to the fourth Assessment Report of the International Panel on climate Change. Cambridge
University Press, Cambridge, United Kingdom
Iwaro J, Mwasha A (2013) The impact of sustainable building envelope design on building
sustainability using Integrated Performance Model. Int J Sustain Built Environ 2(2):153–171
Junnila S, Horvath A (2003) Life-cycle environmental effects of an office building. J Infrastruct
Syst 9:157–166
Kang G et al (2015) Statistical analysis of embodied carbon emission for building construction.
Energy Build 105:326–333
Kwok KYG, Kim J, Chong WKO, Ariaratnam ST (2016) Structuring a comprehensive carbon-
emission framework for the whole lifecycle of building, operation, and construction.
J Architect Eng 22(3):04016006
Lai JH, Yik FW, Man C (2012) Carbon audit: a literature review and an empirical study on a
hotel. Facilities 30(9/10):417–431
Lee B, Trcka M, Hensen JL (2011) Embodied energy of building materials and green building
rating systems – a case study for industrial halls. Sustain Cities Soc 1(2):67–71
Lewis P et al (2009) Requirements and incentives for reducing construction vehicle emissions
and comparison of nonroad diesel engine emissions data sources. J Constr Eng Manag
135:341–351
Lockie S, Berebecki P (2012) Methodology to calculate embodied carbon of materials (RICS).
Royal Institute of Chartered Surveyors, United Kingdom
Lopez-Mesa B, Pitarch A, Tomas A, Gallego T (2009) Comparison of environmental impacts of
building structures with in situ cast floors and with precast concrete floors. Build Environ 44
(4):699–712
Martos JL, Styles D, Schoenberger H, Zeschmar-Lahl B (2018) Construction and demolition
waste best management practice in Europe. Resour Conserv Recycl 136:166–178
Monahan J, Powell J (2011) An embodied carbon and energy analysis of modern methods of
construction in housing: a case study using a lifecycle assessment framework. Energy Build
43(1):179–188
Nadoushani ZM, Akbarnezhad A (2015) Effects of structural system on the life cycle carbon
footprint of buildings. Energy Build 102:337–346
Nadoushani ZSM, Hammad AWA, Akbarnezhad A (2017) Location optimization of tower crane
and allocation of material supply points in a construction site considering operating and rental
costs. J Constr Eng Manag 143(1):04016089
Ng W, Chau C (2015) New life of the building materials - recycle, reuse and recovery. Energy
Proc 75:2884–2891
Ortiz O, Castells F, Sonnemann G (2009) Sustainability in the construction industry: a review of
recent developments based on LCA. Constr Build Mater 23(1):28–39
Ozkaymak M et al (2017) CO2 emission during the combustion of Orhaneli lignite coal. World J
Eng 14(1):27–34
Pomponi F, Moncaster A (2016) Embodied carbon mitigation and reduction in the built
environment – what does the evidence say? J Environ Manag 181:687–700
308 L. M. T. Kumari et al.

Rai D, Sodagar B, Fieldson R, Hu X (2011) Assessment of CO2 emissions reduction in a

distribution warehouse. Energy 36(4):2271–2277
Ramesh T, Prakash R, Shukla K (2010) Life cycle energy analysis of buildings: an overview.
Energy Build 42(10):1592–1600
Rashid AF, Yusoff S (2015) A review of life cycle assessment method for building industry.
Renew Sustain Energy Rev 45:244–248
Reddy BV (2009) Sustainable materials for low carbon buildings. Int J Low-Carbon Technol 4
(3):175–181
Saghafi MD, Teshnizi ZSH (2011) Recycling value of building materials in building assessment
systems. Energy Build 43(11):3181–3188
Sandanayake M, Zhang G, Setunge S (2016) Environmental emissions at foundation construction
stage of buildings – two case studies. Build Environ 95:189–198
Serrano AR, Alvarez SP (2016) Life cycle assessment in building: a case study on the energy and
emissions impact related to the choice of housing typologies and construction process in
Spain. Sustainability 8(3):287
Shi X, Yang W (2013) Performance-driven architectural design and optimization technique from
a perspective of architects. Autom Constr 32:125–135
Thongkamsuk P, Sudasna K, Tondee T (2017) Waste generated in high-rise buildings
construction: a current situation in Thailand. Energy Proc 138:411–416
Thormark C (2002) A low energy building in a life cycle – its embodied energy, energy need for
operation and recycling potential. Build Environ 37(4):429–435
Upton B, Miner R, Spinney M, Heath LS (2008) The greenhouse gas and energy impacts of
using wood instead of alternatives in residential construction in the United States. Biomass
Bioenergy 32(1):1–10
Weight D, Connaughton J, Jones C (n.d.) Information sheet for construction clients and
designers: cutting embodied carbon in construction projects. http://www.wrap.org.uk/sites/
files/wrap/FINAL%20PRO095-009%20Embodied%20Carbon%20Annex.pdf. Accessed 29
Apr 2018
Wheating NC (2017) Embodied carbon: a framework for prioritizing and reducing emissions in
the building industry. s.l.: s.n
Wilkinson SJ, James K, Reed R (2009) Using building adaptation to deliver sustainability in
Australia. Struct Surv 27(1):46–61
Wu P, Feng Y, Pienaar J, Xia B (2015) A review of benchmarking in carbon labelling schemes
for building materials. J Clean Prod 109:108–117
Wu P, Xia B, Pienaar J, Zhao X (2014) The past, present and future of carbon labelling for
construction materials – a review. Build Environ 77:160–168
Yeo D, Gabbai RD (2011) Sustainable design of reinforced concrete structures through embodied
energy optimization. Energy Build 43(8):2028–2033
Yu D, Tan H, Ruan Y (2011) A future bamboo-structure residential building prototype in China:
life cycle assessment of energy use and carbon emission. Energy Build 43(10):2638–2646
Zhong Y, Wu P (2015) Economic sustainability, environmental sustainability and constructabil-
ity indicators related to concrete- and steel-projects. J Clean Prod 108:748–756
Sustainable Building and Infrastructure
 Lessons to Learn and Challenges to Address
in Resilient Historic Towns

M. De Fino(&), E. Cantatore, A. Scioti, and F. Fatiguso

Polytechnic of Bari, Via Orabona 4, 70125 Bari, Italy

[email protected]

Abstract. The concept of resilient cities, including the environmental, social

and economic dimensions of the urban sustainability against both slow and fast
moving changes over long time horizons, is very distinctive when applied to
historic towns. For built heritage, beyond the conventional goals, namely
wellbeing and safety of citizens, functionality and reliability of services and
infrastructures, effectiveness of management and development strategies, as well
as availability of economic resources under disruptions or threads, a further
aspect should be considered. It concerns the preservation and promotion of the
original value and identity of the places by the capability to prevent, adjust and
overcome the obsolescence at different levels – physical, technological, nor-
mative and functional. In this regard, two approaches are required: on the one
hand, the critical understanding of the inherent qualities that made the historic
towns resilient to past transformations – lessons to learn; on the other hand, the
identification of strategies that could enhance such a resilient behaviour in the
future against an exceptionally precipitous evolution of the outer alteration
factors – challenges to address. In the light of the above-mentioned issues, the
paper is going to develop some insights on the aspects that most influence the
resilience of the historic towns, with specific attention toward the Mediterranean
climate of South Italy. Firstly, some performance-based categories are identified,
connected to the inherent resilience of the built heritage – e.g. relationships with
the surroundings, employment of resources, natural regulation of microclimate,
aggregation schemes and uses – as well as to the requirements of contemporary
resilience – e.g. safety, accessibility, fruition, and efficiency. Thus, some qual-
itative and quantitative indicators are introduced and discussed, based on the
validation by representative case studies, as a support toward the development of
a resilience metrics for historic towns.

Keywords: Urban resilience
Historic towns
Built heritage conservation

1 Introduction

The concept of resilient cities, including the environmental, social and economic
dimensions of urban sustainability over long time horizons, is very distinctive when
applied to historic towns. In reality, the urban resilience of the built heritage certainly
involves the requirements set by several international organizations, such as GFDRR,
UNISDR, UN-HABITAT, in terms of its capability to guarantee wellbeing and safety
of citizens, functionality and reliability of services and infrastructures, effectiveness of

© Springer Nature Singapore Pte Ltd. 2020

R. Dissanayake and P. Mendis (Eds.): ICSBE 2018, LNCE 44, pp. 311–322, 2020.
https://doi.org/10.1007/978-981-13-9749-3_29
312 M. De Fino et al.

management and development strategies, as well as availability of economic resources

under disruptions or threads. Nevertheless, it also implies the opportunity to safeguard
the integrity and identity of the places themselves, as tangible and intangible evidence
of culture, tradition and history of the communities. In this regard, with reference to the
resilience of the historic towns, various studies and researches refer to ancient sites as
heuristic models of the centenary ability to resist, adapt, mitigate and recover against
both slow and fast moving changes, as discussed by (Correia et al. 2014), (Hassler and
Kohler 2014), Langenbach (2016), Garcia and Vale (2017). However, the same sci-
entific and technical literature highlights how such an inherent resilience is nowadays
facing an unprecedented acceleration of landscape and climate changes, as well as
cultural, economic and relational modifications of the societies.
Consequently, a challenging declination arises about the conservation-
transformation relationship that typically features the building refurbishment and
requalification processes. On the one hand, the identification and preservation of the
original qualities that made the historic towns capable of showing resilient behavior in
the past - lessons to learn – should be pursued. On the other hand, the development of
strategies that might extend such a behavior in the future against an exceptionally
precipitous evolution of the outer alteration factors – challenges to address – is highly
desirable. In the light of the above-mentioned issues, the recognition of the aspects that
most influence the resilience of the historic towns is paramount, in order to point out
present values and envisaged goals. Yet, such a recognition should preliminarily
identify the investigation boundaries, so that the multi-disciplines and multi-scales
character of the urban resilience might not “collapse into meaninglessness”, as
underlined by Vale (2014).
In this regard, the present study is going to focus mainly on the physical spaces and
structures of the historic towns (streets, squares, buildings, infrastructures …), thus
neglecting the issues related to management policies, involvement of citizens and
availability of funds. Moreover, it will concern the slow risk coming from obsoles-
cence, instead of sudden threads, such as flooding, landslide and external attacks. More
specifically, the paper is going to classify some performance-based categories of
resilience, in relation to both the original identity of the built environment toward its
conservation and the required contemporary quality in view of transforming, adapting
and improving the existing heritage. The categories, based on the critical review of the
state-of-the-art, are herein referred to the Mediterranean are of South Italy. Further-
more, some of them are discussed and validated by assessment of a representative case
study, where qualitative and quantitative indicators are proposed, as a support toward
the development of a resilience metrics for historic towns.

2 General Framework

As outlined in the introduction, the first research phase is concerned with the identi-
fication of a general framework of performance-based categories addressing the con-
cept of resilience of Mediterranean historic towns. In detail, based on the state-of-the-
art-review, as well as on the authors’ experience on the topic, as in De Fino et al.
(2017). Fatiguso et al. (2017), it seems suitable to refer to such an identification of two
 Lessons to Learn and Challenges to Address in Resilient Historic Towns 313

technical-scientific areas. The former relates to the “inner” resilience, namely the
characteristics and values developed by traditional settlements and architectures in
order to interact with local environment, resources and communities. The latter con-
cerns the “outer” resilience, which comprises, in the authors’ view, all the issues and
aspects deriving from external obsolescence factors, in terms of physical, normative,
technological and functional decrease of building performances.
According to the above-described classification, a set of requirements are proposed
below. They are linked to the three dimensions – technical-environmental (TE), socio-
cultural (SC) and socio-economic (SE) - that are commonly used for urban resilience. It
is worth mentioning that the requirements are not meant to provide with a closed
taxonomy. On the contrary, they should pave the way for an open discussion on facets
and implications, even conflicting, toward the development of robust decision-making
support tools for the preservation and revitalization of resilient historic towns.
Particularly, the “inner” resilience, as partially addressed by Sharma (2017), Currà
and D’Amico (2014), Choudhary (2016), Ray and Shaw (2018), Qian and Li (2017), is
herein classified in:
RI.1. Relationship between the urban settlement and the surroundings

RI.1.1. Selection and adaptation to the geo-morphological and hydro-geographic

characteristics of the site - TE
RI.1.2. Arrangement and exposure to climatic agents (wind, sun …) - TE

RI.2. Bioclimatic response in summertime

RI.2.1. Use of passive solutions for natural ventilation - TE

RI.2.2. Use of passive solutions for thermal regulation - TE
RI.2.3. Use of passive solutions for shading - TE

RI.3. Management of resources

RI.3.1. Use of local materials and construction techniques - SC

RI.3.2. Water management - TE
RI.3.3. Storage of resources - SE

RI.4. Use models

RI.4.1. Shared aggregation spaces - SC

RI.4.2. Common services and facilities - SE
RI.4.3. Typological and functional layout - SE
In parallel, the “outer” resilience related to the physical, normative, technological
and functional obsolescence, as referred to some extent by Martínez-Molina et al.
(2016), Webb (2017), Marchi et al. (2018), Maio et al. (2018), Srinurak et al. (2016),
Micelli and Pellegrini (2018) and Zagroba (2016), is herein classified in:
314 M. De Fino et al.

RE.1. State of conservation (physical obsolescence)

RE.1.1. Environmental quality of outdoor spaces - SC

RE.1.2. Structural stability - SC
RE.1.3. Protection from dampness - SC
RE.1.4. Surface integrity - SC

RE.2. Compliance with codes (normative obsolescence)

RE.2.1. Seismic vulnerability - TE

RE.2.2. Energy efficiency - TE
RE.2.3. Accessibility - SC
RE.2.4. Habitability of indoor and outdoor spaces - TE
RE.2.5. Privacy - SC

RE.3. Energy infrastructures (technological obsolescence)

RE.3.1. Availability of energy grids - TA

RE.3.2. Employment of RES - TA

RE.4. Fruition (functional obsolescence)

RE.4.1. Connection with the modern city - SE

RE.4.2. Typological and functional flexibility - SE
RE.4.3 Distribution of services and infrastructures - SE
In order to validate the above-described framework, the following sections intro-
duce, as an illustrative case study, the historic town of Bitonto in South Italy. For that,
some “inner” resilience requirements are assessed by quantitative and qualitative
indicators. Specifically, a multi-level approach is applied, where the assessment is
developed at the building component scale – herein the roof due to its relevance – and
thus reported at the building scale and extended at the district scale.

3 Case Study

The historic centre of Bitonto shows some typological and technical solutions of the so-
called genius loci – namely the distinctive vernacular identity of a place – making the
town highly representative of the coastal and sub-coastal cities of the Apulia Region. In
general, it is a medieval settlement, developed under the Norman, Swabian and
Angevin dynasties, with additions and modifications during the Spanish domination in
the XVI century. It is characterized by compact urban arrangement with very high
housing density and use of local calcarenitic limestone as construction material. The
main building typologies are the “tower-house” featured by a simple plan, with only
one room per floor, two, three or four floors and inner staircase and the “palace-house”,
 Lessons to Learn and Challenges to Address in Resilient Historic Towns 315

typically coming from the aggregation of two/three tower-houses and, therefore, fea-
tured by several rooms per floors on several floors.
In detail, with reference to some of the above-described resilience requirements, the
settlement is naturally shaped following the close Balice watercourse (RI.1.1), it is
arranged through narrow streets providing solar shading on the building facades and
refreshing wind flows in summer (RI.1.2.), it is featured by inner and outer courtyards
that could enhance social interaction and sharing (RI.4.1). Nonetheless, the historic
town has been undergoing, for the last century, a progressive process of transformation,
in order to overcome physical, normative, technological and functional obsolescence,
by structural strengthening (RE.1.2), provision of energy grids (RE.3.1), functional
changes from residential to commercial and entertainment use for improving the overall
touristic attractiveness (RE.4.2). Among the occurred and ongoing transformations, the
most relevant seem to be those related to the roofs. In fact, on the one hand, the roofs
have undergone several renovation either/or replacement works throughout the years
due to static failures and water infiltrations, thus potentially losing their “inner” resi-
lience features (RI.2.2–RI.3.1). Furthermore, they are crucial in relation with “outer”
resilience performances, such as structural, environmental and energy quality (RE.2.1–
RE.2.2), due to their exposure to external aggressive agents (wind, rain, pollution,
direct solar radiation …). Consequently, for such building components, the research
has developed the assessment of some illustrative “inner” resilience requirements
(RI.2.2–RI.3.1) and, thus, it has drawn in the conclusions some insights on the rela-
tionship among these requirements and some “outer” resilience ones (RE.2.1–RE.2.2).

3.1 Assessment at the Building Scale

As far as the resilience requirement RI.3.1 is concerned - use of local materials and
construction techniques – the assessment is qualitative: it corresponds to the degree of
compatibility of the current building component with the original roof solution, in
terms of both structural system and external finishing layer. On the contrary, as far as
the resilience requirement RI.2.2 is concerned - use of passive solutions for thermal
regulation – the assessment is quantitative: it corresponds to the indoor surface tem-
perature of the roof in summer (Tin), as calculated by theoretical dynamic simulation,
given two main parameters of the current building system:
• Periodic thermal transmittance (YI,E) calculated according to normative standard
EN ISO 13786:2017 “Thermal performance of building components - Dynamic
thermal characteristics - Calculation methods”. It is an indicator of the thermal
inertia, deriving from the attenuation factor (fa) and the steady-state thermal
transmittance U;
• Albedo (a) assessed according to common values for categories of materials. It is
the measure of the diffuse reflection of solar radiation out of the total solar radiation
received by the external finishing layer.
Both the assessment criteria – degree of compatibility with the original solution for
RI.3.1 and indoor surface temperature for RI.2.2 - have been applied to the case study,
after the detailed identification of the current roof components, based on documentary
316 M. De Fino et al.

research and onsite inspection. Specifically, n.6 scenarios have been surveyed, corre-
sponding to several combinations of structural system and external finishing layer:
Comb1.a and Comb 1.b. They both correspond to horizontal slabs made out of steel
reinforced concrete beams and hollow brick blocks. They were typically used in the
60’s to replace the original timber slabs, because of their poor conditions and due to
missing local government control, or to cover a new built upper floor. They are
generally covered by waterproofing bituminous membranes, when inaccessible
(Comb1.a). Otherwise, they were finished by traditional stone tiles (Comb1.b);
Comb2.a e Comb2.b. They both correspond to horizontal slabs made out of steel
beams and hollow brick blocks. They generally result from replacement work of the
early XX century and can be found in those areas with medium-low state of con-
servation, where limited intervention have been carried out in the last decades.
Consequently, they might be considered a historic transformation according to the
local construction techniques, consistent with the tradition featuring that period at
regional levels. Similar to the previous combinations, they might be covered by
waterproofing bituminous membranes, when inaccessible (Comb2.a). Otherwise,
they are finished with local clay tiles (Comb2.b);
Comb3.a e Comb3.b. They both correspond to masonry vaults made out of regional
calcareous stone structure and upper filling of mortar mixture and stone fragments.
They typically feature in the noble palaces, which still have residential functions.
Again, they might be covered by waterproofing bituminous membranes, when
inaccessible (Comb3.a). Otherwise, they are finished with stone tiles (Comb3.b);
As previously explained, following the above-described identification, the
requirements R1.3.1 and R2.2 have been assessed. Particularly, the former has been
further classified into R1.3.1a and R1.3.1b, corresponding respectively to the use of
local techniques for the structural component and the use of local materials for the
finishing layer. For the assessment, each requirement has been described by a score
scale, ranging from 0 to 1, according to the roof configuration.
In detail, for the requirement RI.3.1a, the following scores are proposed:
• “0” for horizontal slabs made out of steel reinforced concrete beams and hollow
bricks (Comb1.a-b) due to their inconsistency with the construction and technical
characteristics of the original structures, namely timber slabs and stone vaults;
• “0.5” for horizontal slabs made out of steel beams and hollow brick blocks (Comb2.
a-b), which still witness the indigenous construction tradition, although resulting
from a subsequent alteration of the original features;
• “1” for stone masonry vaults (Comb3.a-b) as representative of the very primary
identity of the local built environment places.
Similarly, for the requirement RI.3.1b, the following scores are proposed:
• “0” for waterproofing bituminous membranes;
• “0.5” for clay tiles as representative of a subsequent historic transformation;
• “1” for original stone tiles.
Finally, the same combinations have been assessed against the requirement R.I.2.2.
For the purpose, the thermal properties of the structural layers (Table 1) and the optical
 Lessons to Learn and Challenges to Address in Resilient Historic Towns 317

properties of the finishing layers (Table 2) have been identified, based on common
values reported in literature.

Table 1. Thermal properties

U(W/(m2K) fa Yi,e u(h)
Comb1a-b 1,63 0.29 0.471 9.25
Comb2a-b 1,91 0.35 0.662 8
Comb3a-b 2,07 0.136 0.282 10.6

Table 2. Optical properties

a e
Bituminous membrane 0.1 0.95
Clay tiles 0.4 0.9
Stone stiles 0.6 0.95

Therefore, the above-reported parameters have been implemented into software-

based dynamic simulations of the energy and environmental behavior of typical
building unit, here three-storey tower-houses, covered by each roof combinations. The
simulations, run by Design Builder©, enabled the calculation of the indoor surface
temperature in summertime, based on the climatic data of the meteorological station at
the near international airport of Bari Palese.
Figure 1 reports some representative results of the roof intrados and extrados mean
temperatures, as calculated for a representative hot day – 4th August – with 24.5 °C and
34.5 °C average and maximum daily temperature. Thus, based on the internal mean
surface temperatures (Table 3), the requirement R.I.2.2 has been also described by a
score scale, ranging from 0 to 1, according the temperature ranges.
In detail, for the requirement RI.2.2, the following scores are proposed:
• “0” corresponding to the range 30°–32 °C. The score is associated with all the
combinations featured by bituminous waterproofing membranes as the finishing
layer (Comb1.a, Comb2.a and Comb3.a). The results show particularly how the
worst scenario corresponds to the combination of horizontal slabs made out of steel
reinforced concrete beams and hollow brick blocks (Comb1.a), where both poor
thermal properties of the structural component and poor optical properties of the
finishing layer occur.
• “0.5” corresponding to the range 28°–30 °C. The score is associated with the
horizontal slabs covered by clay tiles (Comb1.b e Comb2.b), where the poor
thermal inertia of the structural components are somehow compensated by the
medium solar reflectance of the finishing layer.
“1” corresponding to the range 26°–28 °C. The score is associated with stone vaults
and stone tiles (Comb3.b), where the density of the massive structural component
provides with good thermal inertia and the light color of the finishing reduces the
surface temperature peaks over the daytime.
318 M. De Fino et al.

Fig. 1. Roof intrados and extrados mean temperatures on 4th August

Table 3. Internal mean surface temperatures

Comb 1.a 1.b 2.a 2.b 3.a 3.b
Tmean [°C] 31,30 28,51 31,70 28,53 31,92 26,41
T env 24,45

3.2 Assessment at the District Scale

Based on the assessment at the building scale, the entire historic town has been mapped
(Fig. 2) to calculate the incidence of the surface of each roof solution out of the total
and, thus, derive for each requirement a synthetic score at the district scale as weighted
average of the scores achieved by the six combinations (Table 4).
 Lessons to Learn and Challenges to Address in Resilient Historic Towns 319

Fig. 2. Distribution of the combinations

Table 4. Assessment scores for the historic town

% extent RI.3.1a RI.3.1b RI.2.2
Comb1a 61% 0 0 0
Comb1b 17,9% 0 0,5 0,5
Comb2a 9,8% 0,5 0 0
Comb2b 5,3% 0,5 0,5 0,5
Comb3a 2,6% 1 0 0
Comb3b 3,3% 1 1 1
Total 0,135 0,15 0,15

3.3 Discussion of Results

At the construction component scale, as related to the local building typologies and
climatic conditions, it should be noted that all the solutions preserving original roof
materials and techniques - herein the stone masonry vaults covered by stone tiles
(Comb.3.b) – show the most satisfying indoor comfort. Nevertheless, some further
systems, including horizontal slabs made out of steel reinforced concrete beams and
320 M. De Fino et al.

hollow bricks and horizontal slabs made out of steel reinforced concrete beams and
hollow brick blocks, perform relatively better if combined with “semi-traditional” clay
tiles, instead of modern waterproofing bituminous membranes (Comb.1.b and Comb.2.
b). However, the worst combination from the conservative point of view, namely the
horizontal slabs made out of steel reinforced concrete beams and hollow brick blocks
and covered by external bituminous membrane (Comb.1.a), is also the poorest in terms
of environmental performances. Moreover, at the district scale, the high incidence of
this solution affects the overall assessment of the historic town, resulting in a quite low
score for all the three indicators under study. Consequently, in view of future requal-
ification and retrofitting policies, the strict correlation between recovery of original
materials and techniques and improvement of quality levels and the opportunity to
apply the assessment routine as comparative guide and control tool for selecting
suitable alternatives should be taken into account. For instance, if all the external
finishing layers of that era currently made out of bituminous membranes would be
covered by clay tiles, great improvements could be achieved, in terms of compatibility
(RI.3.b = 0.53), as well as attenuation of thermal peaks in summertime and improve-
ment of indoor comfort (RI.2.2 = 0.53). Similarly, the replacement of all the existing
horizontal slabs made out of concrete beams and hollow brick blocks with structural
timber beams and borders and finishing stone tiles would imply a further increase of the
district overall score (RI.3.a = 0.92, RI.3.b and R2.2 = 0.85).

4 Conclusions

The paper set out to identify in Mediterranean historic towns – some categories of
“inner” resilience – characters and values that the traditional settlements and archi-
tectures have developed in relation to the local environment, resources and commu-
nities – and some categories of “outer” resilience – issues and aspects of the historic
built environments coming from external obsolescence factors. Thus, a score-based
assessment of some relevant requirements has been proposed, with specific focus on
the “use of passive solutions for thermal regulation” and the “use of local materials and
construction techniques” in the roof components. The outcome shows how the con-
struction culture and practice are very well compatible with the desirable capability of
protection from atmospheric agents, particularly the direct solar radiation in summer-
time, and mitigation of indoor discomfort. Such a compatibility has been lost in some
modern solutions, where the employment of industrial building products and processes,
beyond the missing opportunity to transmit a traditional knowledge to the future
generations, has caused a performance drop, in terms of micro-environmental and
micro-climatic quality. Nonetheless, although not investigated by the present paper, it
should be outlined for further studies that the recovery of traditional solutions seems to
be very promising even for some requirements of “outer” resilience, related to seismic
protection and energy efficiency among others. In fact, the steel reinforced concrete
solutions with outer finishing membrane might negatively affect the vulnerability to
earthquakes due to their weight and rigidity, as well as the heat conductivity on the
inside due to the high solar absorption and thermal gradient of the outside. The
development of a resilience metrics, to which the present paper gives a limited
 Lessons to Learn and Challenges to Address in Resilient Historic Towns 321

contribution, is highly desirable in order to understand and manage the multiple

implication connected with the institutional actions and the technical measures. Finally,
it should be highlighted that the safeguard and transmission of the “lessons” about
inherent resilience of these historic towns is paramount and vital. For this purpose, the
“challenges” against the obsolescence according to modern quality level, although
equally relevant, might be addressed by the compatible integration of traditional
solutions with cutting edge products and systems, as an alternative to the replacement
or alteration of the existing structures.

References
Choudhary P (2016) Vernacular built environments in India: an indigenous approach for
resilience. In: Urban disasters and resilience in Asia. Elsevier, pp 269–286. https://doi.org/10.
1016/b978-0-12-802169-9.00017-3
Correia M, Dipasquale L, Mecca S (2014) VERSUS: heritage for tomorrow. Correia M,
Dipasquale L, Mecca S (eds) Firenze University Press. ISBN 9788866557418 (print)
Currà E, D’Amico A (2014) The role of urban built heritage in qualify and quantify resilience.
Specific issues in Mediterranean City. Proc Econ Finance 18:181–189. https://doi.org/10.
1016/S2212-5671(14)00929-0
Fatiguso F et al (2017) Resilience of historic built environments: inherent qualities and potential
strategies. Proc. Eng. https://doi.org/10.1016/j.proeng.2017.04.262
De Fino M et al (2017) Methodological framework for assessment of energy behavior of historic
towns in Mediterranean climate. Energy Build 144:87–103. https://doi.org/10.1016/j.enbuild.
2017.03.029
Garcia EJ, Vale B (2017) Unravelling Sustainability and Resilience in the Built Environment.
Routledge, Abingdon ISBN 9781138644021 (hardcover)
Hassler U, Kohler N (2014) Resilience in the built environment. Build Res Inf 42(2):119–129.
https://doi.org/10.1080/09613218.2014.873593
Langenbach R (2016) What we learn from vernacular construction. In: Nonconventional and
Vernacular Construction Materials. Elsevier, pp 3–36. https://doi.org/10.1016/b978-0-08-
100038-0.00001-9
Maio R, Ferreira TM, Vicente R (2018) A critical discussion on the earthquake risk mitigation of
urban cultural heritage assets. Int J Disaster Risk Reduction 27:239–247. https://doi.org/10.
1016/j.ijdrr.2017.10.010
Marchi M et al (2018) Environmental policies for GHG emissions reduction and energy
transition in the medieval historic centre of Siena (Italy): the role of solar energy. J Clean Prod
185:829–840. https://doi.org/10.1016/j.jclepro.2018.03.068
Martínez-Molina A et al (2016) Energy efficiency and thermal comfort in historic buildings: a
review. Renew Sustain Energy Rev 61:70–85. https://doi.org/10.1016/j.rser.2016.03.018
Micelli E, Pellegrini P (2018) Wasting heritage. The slow abandonment of the Italian Historic
Centers. J Cult Heritage 31:180–188. https://doi.org/10.1016/j.culher.2017.11.011
Qian Z, Li H (2017) Urban morphology and local citizens in China’s historic neighborhoods: a
case study of the Stele Forest Neighborhood in Xi’an. Cities 71:97–109. https://doi.org/10.
1016/j.cities.2017.07.018
Ray B, Shaw R (2018) Changing built form and implications on urban resilience: loss of climate
responsive and socially interactive spaces. Proc Eng 212:117–124. https://doi.org/10.1016/j.
proeng.2018.01.016
322 M. De Fino et al.

Sharma AK (2017) Historic city–a case of resilient built environment. Proc Eng 180:1103–1109.
https://doi.org/10.1016/j.proeng.2017.04.270
Srinurak N et al (2016) Analysis of urban morphology and accessibility character to provide
evacuation route in historic area. Proc-Soc Behav Sci 216:460–469. https://doi.org/10.1016/j.
sbspro.2015.12.061
Vale LJ (2014) The politics of resilient cities: whose resilience and whose city? Build Res Inf 42
(2):191–201. https://doi.org/10.1080/09613218.2014.850602
Webb AL (2017) Energy retrofits in historic and traditional buildings: a review of problems and
methods. Renew Sustain Energy Rev 77:748–759. https://doi.org/10.1016/j.rser.2017.01.145
Zagroba M (2016) Issues of the revitalization of historic centres in small towns in Warmia. Proc
Eng 161:221–225. https://doi.org/10.1016/j.proeng.2016.08.537
 Pedestrian Wind Comfort Study Using
Computational Fluid Dynamic
(CFD) Simulation

Sarah Fernando1(&), Shiromal Fernando1, and Priyan Mendis2

1
Civil & Structural Engineering Consultants (Pvt) Ltd., No. 01, 33rd Lane,
Bagathale Road, Colombo 03, Sri Lanka
[email protected]
2
University of Melbourne, Melbourne, Australia

Abstract. Many high-rise buildings are presently being designed and con-
structed in the central business district of Colombo. These buildings will change
the wind patterns and directions which may create an uncomfortable environ-
ment at the pedestrian level. Urban authorities in many countries grant a
building permit for a high-rise building, only if the wind comfort study proves
that the wind environment for the pedestrian remains comfortable and safe.
Pedestrian wind comfort and safety studies have been conducted in the past by
wind tunnel tests. However, due to the high cost involved, pedestrian wind
comfort studies are not generally carried out. With a significant improvement in
computer resources in the recent past, Computational Fluid Dynamic
(CFD) simulations can be used as an effective alternative method to simulate the
pedestrian wind comfort around high rise buildings. This paper presents a val-
idation of pedestrian wind comfort results obtained from a wind tunnel test,
using CFD simulation.

Keywords: High-rise buildings
Wind comfort and safety

Computational Fluid Dynamics (CFD)
Wind environment

1 Introduction

Outdoor pedestrian wind environment plays a major role in providing comfort and
safety to the pedestrians around high rise buildings. High-rise buildings in urban areas
can often introduce uncomfortable or even dangerous wind environments at the
pedestrian level due to the form of complex wind patterns. The death of two old ladies
has been reported due to a disastrous fall which was caused by high wind speed at the
base of a tall building (Lawson and Penwarden 1975). Hence, it is necessary to evaluate
the condition of the pedestrian wind environment in terms of comfortability for high
rise buildings.
High-rise buildings are particularly influential to complex wind patterns due to the
building’s complex geometrical features (Ex: sharp/leading edges, trailing edges,
complex corner configurations, slender columns etc.) that could result in the behaviour
of the wind around the structure. Today, urban authorities in many countries grant a
building permit for a high-rise building only if the wind comfort study proves that the
© Springer Nature Singapore Pte Ltd. 2020
R. Dissanayake and P. Mendis (Eds.): ICSBE 2018, LNCE 44, pp. 323–339, 2020.
https://doi.org/10.1007/978-981-13-9749-3_30
324 S. Fernando et al.

wind environment for the pedestrian remains comfortable and safe (Blocken and
Carmeliet 2008). Pedestrian wind comfort studies can be either be performed using
wind tunnel tests or using Computational Fluid Dynamic (CFD) simulations. In the
past, wind tunnel tests have been conducted intensively as a tool for evaluating
pedestrian wind comfort conditions around buildings. Wind tunnel tests are expensive
in simulation process and in maintenance. Moreover, only point measurements can be
obtained by wind tunnel experiments (Kim et al. 2013). However, with a significant
improvement in computer resources in the recent past, Computational Fluid Dynamic
(CFD) simulations can be used as an effective alternative method to simulate the
pedestrian wind comfort and safety around high rise buildings. CFD studies are most
often used due to its ability of full-scale modeling and it facilitates the obtaining of a
variety of results from single demonstration. In fact, CFD simulations can be used to
evaluate physics that cannot be achieved practically in the wind tunnel test.
This paper presents the CFD evaluation of the pedestrian wind comfort and safety
in a high-rise apartment building. The validation process was carried out with the
experimental results and the numerical simulation results were obtained using the
commercially available software Solidworks Flow Simulation. Wind mitigation mea-
sures and improvements are evaluated for the proposed development premises
according to the results.

2 Wind Criteria

2.1 Criteria for Comfort and Safety

Many researchers have proposed a variety of criteria to assess the pedestrian level wind
climate condition. According to William et al. (1990), either a fixed wind speed is
selected and evaluated where the percentage of time that this wind speed can occur
varies depending on the activities planned for the area or a fixed frequency of occur-
rence is selected with acceptable wind speed varying depending upon the anticipated
pedestrian activity. Situations where winds are very strong and it directly affects the
people’s balance, Gust speeds are more important. However, mean wind speeds might
be just as important an indicator of comfort in areas such as outdoor pedestrian usage
areas. Therefore, it is desirable for the criteria to capture the effects of both mean speeds
and gust speeds (ASCE 2003).
Durgin (1995) and Lawson (1978) proposed a way of achieving this, which is to
express the criteria in terms of mean speeds, but then to compare not just with the mean
speed, but also compare with a parameter called “Gust Equivalent Mean Speed”
(UGEM).

b
U
U GEM ¼ ð1Þ
G

b is the peak gust speed and G is a representative fixed gust factor.

Where U
 Pedestrian Wind Comfort Study Using Computational Fluid Dynamic 325

For the purpose of feasibility, the pedestrian level wind micro-climate criteria that is
described in Table 1 was employed in this study. Wind conditions acceptable for sitting
(having lowest wind speeds) are appropriate for outdoor amenity areas and seating
areas in open air cafes/parks. Furthermore, a small increase of wind speeds will reduce
the humidity of amenity spaces which will be more beneficial. Hence, a mix of sitting
and standing conditions for such kind of places would be more appropriate. Standing
conditions are preferred at main entrances, drop off areas and bus stops, where
pedestrians are about to linger. Strolling conditions are appropriate if pedestrians are
expected to move at a more leisurely place (Eg: shop windows, park, Children’s play
area or landscape area). Moreover, walking conditions are appropriate for side walks
and pathways, when pedestrians are not expected to move for a purpose. It should be
noted that wind speeds consistently above this level are likely to be a nuisance to
pedestrians.

Table 1. Wind comfort criteria used for the study (RWDI 2017)
Comfort UGEM/(m/s) Description
category
Sitting  2.38 Calm or Light breezes desired for outdoor restaurants and
seating areas where one can read a paper without having it
blown away
Standing  3.88 Gentle breeze acceptable for main building entrances and
bus stops
Strolling  4.72 Moderate winds that would be appropriate for window
shopping and strolling along a downtown street, plaza or
park
Walking  5.56 Relatively high speeds that can be tolerated if one’s
objective is to walk, run or cycle without lingering
Uncomfortable  5.56 Strong winds of this magnitude are considered a nuisance
for most activities, and wind mitigation is typically
recommended

3 Analysis of Wind Data

Wind statistics recorded at the Katunayake International Airport between 1998 and
2017 were observed between May and October, which includes the southwest monsoon
and the November to April Period, which includes the northeast monsoon. Figure 1
illustrates the distribution of wind frequency and directionality for those two periods of
the year.
During the period of May to October, prevailing winds are from west through
south-west directions. During the period of November to April, prevailing winds are
from north-east direction.
326 S. Fernando et al.

Fig. 1. Seasonal wind roses for Colombo (data from Katunayake International Airport
(1998–2017)

4 Wind Behavior Around Structures

The effect of the pedestrian wind microclimate may depend on the building shape, size,
orientation and interaction with neighbouring building obstacles. Tall buildings tend to
intercept stronger winds at higher elevations and redirect them to the ground level.
Figure 2 shows the down washing effect of wind, which is deflected down to the street
level, creating swirling turbulence at the pedestrian level. This effect is often the main
cause for wind accelerations around large buildings at the pedestrian level.
When oblique winds are deflected down by a building, an increment of the wind
activity can be expected around the building corners at the pedestrian level as shown in
Fig. 3. (RWDI 2017). This is known as the corner acceleration effect of wind due to
building geometry and the behaviour of the wind.

Fig. 2. Down washing effect

 Pedestrian Wind Comfort Study Using Computational Fluid Dynamic 327

Fig. 3. Corner acceleration effect

5 Case Study

5.1 Model
The Proposed Development is a 36-storey residential development comprising of a
tower, a top, and a 4-level podium. The development will be 136 m high above the
ground level after completion. Podium level at level 5 provides common facilities and
above the podium level, all the other floors are apartments.

5.2 CFD Simulation

In order to evaluate the wind condition related to the pedestrian wind comfort, a CFD
simulation was conducted by using the commercially available software Solidworks
Flow Simulation. A full-scale solid model of the proposed development was modelled
(see Fig. 4) for the purpose of conduction a CFD simulations.

Fig. 4. Solid model of the building

328 S. Fernando et al.

Hence, no scaling was involved in length and time scales of the results. Model
setup has been selected for optimizing the computational time whilst keeping the
numerical errors to be at a minimum.
Wind profile within the atmosphere extends from the ground level to few hundred
meters. The shape, speed and the direction of the wind profile changes with various
factors such as terrain type, topography features, direction etc. However, for this
analysis, the wind profile which was calculated using AZ/NZS1107.2:2002 (as shown
in Fig. 5) was used.
Similar to the wind profile, the turbulence intensity profile (shown in Fig. 6) was
used as mentioned in the AS/NZS 1170.2:2002 assuming that the terrain category of
the region as 2 and the region as B.

500

450

400

350

300
Height/(m)

250

200

150

100

50

0
8.00 9.00 10.00 11.00 12.00 13.00
Wind Velocity/(m/s)

Fig. 5. Relationship of height vs wind velocity

Wall conditions are assumed to be adiabatic condition for this analysis. The size of
the computational domain depends on the region that shall be represented by the
simulation (see Fig. 7). It should be large enough to avoid reflecting of fluid streams,
which may cause abnormal pressure fields around the model.
However, it is also true that unnecessarily large fluid domain demand higher
computational cost. Therefore, arbitrary domain sizes were used for this analysis,
 Pedestrian Wind Comfort Study Using Computational Fluid Dynamic 329

which are large enough to give a satisfactory level of accurate results with reasonable
computational cost.

500

450

400

350
Height/ (m)

300

250

200

150

100

50

0
5 7.5 10 12.5 15 17.5 20
Turbulence Intensity / (%)

Fig. 6. Relationship of height vs turbulence intensity

Fig. 7. Computational domain of the study

The mesh has been done via SOLIDWORKS and the model setup consists of two
mesh types; one for the global domain and the other for near building surface. This
model setup was selected for optimizing the computational time whilst keeping the
numerical errors to a minimum.
330 S. Fernando et al.

5.3 Definition of Wind Direction

0° wind direction has been chosen to coincide with the geographical North as shown in
Fig. 8. Major wind directions were selected as North-East and South-West for the
pedestrian wind comfort analysis.

Fig. 8. Geographical north of the building (plan view)

6 Results

Results were obtained using CFD Simulation at three levels (Ground Level, Podium
Level and Roof Terrace Level as shown in Fig. 9) to assess the condition where wind
directly affects the comfortability of people.
Maximum wind velocity values (from North-East and South-West Direction) at
each point obtained from the wind tunnel test (which were shown in Figs. 10, 11 and
12) were compared with the CFD simulation results.

Fig. 9. Considered floor levels of the apartment building

 Pedestrian Wind Comfort Study Using Computational Fluid Dynamic 331

Fig. 10. Ground floor level-locations

Fig. 11. Podium level-locations

Fig. 12. Roof terrace level-locations

Figures 13, 14 and 15 shows the velocity contours at the ground level, podium
level and roof terrace level respectively for the wind direction of North-East.
332 S. Fernando et al.

Figures 16, 17 and 18 shows the velocity contours at the ground level, podium
level and roof terrace level respectively for the wind direction of South-West.

Fig. 13. Contours for velocity in a horizontal level at a height of 1.5 m above ground level for
wind direction-North East

Fig. 14. Contours for velocity in a horizontal level at a height of 1.5 m above podium level for
wind direction-North East
 Pedestrian Wind Comfort Study Using Computational Fluid Dynamic 333

Fig. 15. Contours for velocity in a horizontal level at a height of 1.5 m above roof terrace level
for wind direction-North East

Fig. 16. Contours for velocity in a horizontal level at a height of 1.5 m above ground level for
wind direction-South West
334 S. Fernando et al.

Fig. 17. Contours for velocity in a horizontal level at a height of 1.5 m above podium level for
wind direction-South West

Fig. 18. Contours for velocity in a horizontal level at a height of 1.5 m above roof terrace level
for wind direction-South West

Table 2 shows the results of wind condition of each locations, considering the
results obtained from both wind tunnel and CFD simulations for the two main direc-
tions (North-East and South-West).
 Pedestrian Wind Comfort Study Using Computational Fluid Dynamic 335

Table 2. Summary of the results of pedestrian wind comfort conditions

Location Wind CFD Condition Remarks
tunnel/ Simulation/
(m/s) (m/s)
Ground 1 7.8 9.6 Uncomfortable Discrepancies of the results may due
level 2 5.8 6.9 Uncomfortable to absence of the surrounding of the
3 6.9 7.3 Uncomfortable CFD model, since it may directly
4 6.1 6.5 Uncomfortable affect to the wind behaviour around
the structure
5 4.4 7.5 Strolling-
Uncomfortable
6 3.3 5.4 Standing-
Walking
7 4.2 3.5 Strolling-
Standing
8 3.8 3.0 Standing
9 5.0 5.6 Walking
10 3.8 4.3 Strolling
11 5.3 5.5 Walking
12 6.1 7.6 Uncomfortable
13 8.3 10.3 Uncomfortable
14 5.3 6.8 Walking-
Uncomfortable
15 5.3 5.0 Walking
16 4.7 4.9 Strolling-
Walking
17 4.2 3.9 Strolling
18 3.8 2.0 Standing-Sitting
19 3.6 2.1 Standing-Sitting
20 3.8 2.0 Standing/Sitting
Podium 21 2.5 1.6 Sitting Discrepancies of the results may due
level 22 5.0 2.0 Walking - Sitting to absence of the surrounding of the
23 3.3 3.6 Standing CFD model, since it may directly
24 5.8 6.3 Uncomfortable affect to the wind behaviour around
the structure
25 6.4 7.1 Uncomfortable
26 5.0 5.5 Walking
27 4.7 5.0 Walking
28 5.0 10.5 Walking-
Uncomfortable
29 5.0 5.6 Walking
30 4.7 4.9 Strolling/Walking
31 3.8 1.5 Standing-Sitting
32 3.8 3.0 Standing
33 4.2 4.7 Strolling
34 3.8 3.7 Standing
(continued)
336 S. Fernando et al.

Table 2. (continued)
Location Wind CFD Condition Remarks
tunnel/ Simulation/
(m/s) (m/s)
Roof 6.8 7.3 Uncomfortable
terrace 36 7.8 8.8 Uncomfortable
37 5.3 5.5 Walking
38 4.7 3.9 Strolling
39 4.4 2.0 Strolling-Sitting

7 Wind Comfort Assessment

The discussion of the wind comfort at the three levels is based on results obtained using
wind tunnel data and CFD simulations.

7.1 Ground Level

Locations represented by 1 to 27 are expected to have wind conditions ranging from
suitable to sitting to uncomfortable according to the results obtained in Sect. 6.
Uncomfortable wind conditions are formed in the corners of the building due to corner
acceleration effect of the wind flow. Mitigation measures are required to achieve
desired calmer conditions at the ground level.
Locations represented by 18, 19 and 24 (Main Entrance) shows a calmer wind
condition throughout the year. However, lowest wind condition may lead to create
discomfort to the pedestrians (due to formation stagnation condition). Such areas need
to be considered to avoid that kind of discomfort where pedestrians are likely to stay.

7.2 Podium Level

The wind condition at the podium level ranges from standing to uncomfortable con-
ditions during both monsoons. The amenity area along the podium terrace includes the
swimming pool, Children’s pool and built-in seating areas. Wind condition of the
swimming pool areas (28 and 31) are likely to have walking to uncomfortable con-
ditions according to the results, due to the podium which is exposed to the wind. Wind
flow may lead to accelerate through confined spaces as well as corner areas of the
building. Open sky seating areas (36 and 37) are predicted to have higher wind con-
ditions which would require mitigation measures to maintain desired (sitting) wind
conditions.

7.3 Roof Terrace Level

Roof level terrace areas represented by 38 to 42 and wind condition along the terrace
ranges from sitting to uncomfortable according to the simulation. Mitigations measures
 Pedestrian Wind Comfort Study Using Computational Fluid Dynamic 337

are required to achieve the desired wind conditions if any seating areas are proposed
along this terrace.

8 Mitigation Strategies

In areas where wind condition is unacceptable for the intended pedestrian use or unsafe
and will be accessible to pedestrians, it is advisable to develop wind control strategies.

8.1 Mitigation at Ground Level

Wind control measures are recommended along ground areas to eliminate the
uncomfortable walking conditions that were described in Sect. 7.1. The addition of
hard or soft landscaping elements are advisable as shown in Fig. 9. However, ever-
green trees having 2 to 4 m height with low level landscaping elements will diffuse the
energy of high wind.
In order to diffuse high wind speeds, porous wind screens can also be used in
ground areas. To achieve sitting wind localised sheltering and 1.5 m–2 m planting or
screening elements can be used.
It is recommended, that the areas that indicating stagnant conditions, landscaping
elements and planting trees should be avoided or minimized in order to unblock the air
flow entering such places.

8.2 Mitigation at Podium Level

Seating areas and swimming pool along the podium level, need to be considered when
adding high wind speed mitigation measures. The addition of 1.5 m to 2 m height
columnar plantations along the balustrade would be significant in breaking up any
oncoming flow and provide a desired wind condition in such places. In addition to that,
low-level landscaping elements along the periphery of the pool area could be more
helpful in achieving calm conditions. Furthermore, low level landscaping such as
shrubs in planters, porous screens or sculptures can be used within the podium level to
confuse the wind and break high wind speeds.

8.3 Mitigation at Roof Terrace Level

Mitigations that are mentioned Sect. 8.2 are applicable to the roof terrace level. In
addition to that, increasing the height of the balustrade from 1 m to 2 m will be
beneficial in providing shelters for the areas which are nearby (immediately behind) the
balustrade.
338 S. Fernando et al.

9 Concluding Remarks

CFD has been applied to evaluate the pedestrian wind comfort in specific locations in a
High-rise Apartment Building. The CFD Simulations were conducted based on the two
seasonal monsoons that occur in Sri Lanka (North-East monsoon and South-West
Monsoon). Results obtained from CFD simulation have been compared with the results
obtained from wind tunnel test for verification. Pedestrian comfort was assessed for the
ground level, podium level and the roof terrace level of the development. The wind
microclimate in these levels is predicted to range from sitting to uncomfortable during
the windiest season.
Overall, the pedestrian comfort conditions around the selected levels is acceptable
for its intended use and any windier than desired locations are proposed to be mitigated
by the suggested mitigation measures.

References
Lawson TV, Penwarden AD (1975) The effect s of wind on people in the vicinity of buildings.
In: 4th international conference on wind effects on buildings and structures, Heathrow
Janssen W, Blocken B, Van Hoof T (2013) Use of CFD simulations to improve the pedestrian
wind comfort around a high-rise building in a complex urban area. In: Proceeding of BS2013:
13th conference of international building performance simulation association, pp 1918–1925
Williams CJ, Soligo MJ, Cote J (1992) A Discussion of the components for a comprehensive
pedestrian level comfort criterion. J Wind Eng Ind Aerodyn 44–41:2389–2390
Fadl MS, Karadelis J (2013) CFD simulation for wind comfort and safety in urban area: a case
study of Coventry University central campus. Int J Archit Eng Constr 2(2):1–13
Blocken B, Carmeliet J (2008) Pedestrian wind conditions at outdoor platforms in a high-rise
apartment building: generic sub configuration validation wind comfort assessment and
uncertainty issues. Wind Struct 11(1):51–70
ASCE Aerodynamics Committee (2003) Outdoor human comfort and its assessment. State of the
art report, Task Committee on Outdoor Human Comfort, American Society of Civil
Engineers, Boston, VA, USA
Stathopoulos T (2006) Pedestrian level winds and outdoor human comfort. J Wind Eng Ind
Aerodyn 94:769–780
William CJ, Hunter MA, Waechter WF (1990) Criteria for assessing the pedestrian wind
environment. J Wind Eng Ind Aerodyn 36:811–815
Montazeri H, Blocken B, Janssen WD, Hooff TV (2012) CFD analysis of wind comfort on high-
rise building balconies: validation and application. In: The seventh international colloquium
on bluff body aerodynamics and applications, Shanghai, China, pp 1674–1681
Kim H, Kim T, Leigh S (2013) Assessment of pedestrian wind environment of high-rise complex
using CFD Simulation. In: Sustainable procurement in urban regeneration and renovation
Northern Europe and North-West Russia, pp 1–8
AS/NZS-1170.2 (2002) Australian/New Zealand Standard, Structural Design Action-Part 2:
Wind Actions, Standards Australia & Standards New Zealand
RWDI (2017) Pedestrian level wind micro-climate assessment, Apartment complex, Colombo,
Sri Lanka, 170 1325
 Pedestrian Wind Comfort Study Using Computational Fluid Dynamic 339

Mohotti D, Danushka K, Mendis P (2015) Wind design of slender tall building: CFD approach.
In: 6th international conference on structural engineering and construction management,
Kandy, Sri Lanka, pp 194–203
Urban Design Terms of Reference (2014) Pedestrian comfort and safety studies. Planning and
Building Department, Development and Design Division, City of Mississauga
Mendis P, Ngo T, Hariots N, Hira A, Samali B, Cheung J (2007) Wind loading on tall building.
EJSE special issue: loading on structures, pp 41–54
 Performance of Paving Materials in Outdoor
Landscaping

G. H. Galabada(&) and R. U. Halwatura

University of Moratuwa, Moratuwa, Sri Lanka

[email protected]

Abstract. Many developments today, have made the human lifestyle more
comfortable but seem to be concerned only about the technical and economic
viability rather than the environment or the user comfort. Hence, these devel-
opments are causing severe environmental issues such as global warming, heat
island effect, losing the natural appearance etc. The selection of construction
materials considering their properties is vital with respect to sustainable outdoor
development. Materials such as clay bricks, asphalts, concrete etc.; are used and
since recent times, concrete block pavements have become an attractive engi-
neering and economical material in the construction of walkways, worldwide.
However, there are many drawbacks with respect to user comfort via temper-
ature effect in the use of these materials. Therefore, the study of temperature
effects of materials, which are used for outdoor developments is important.
Accordingly, this research was done to evaluate and compare the thermal per-
formance of existing paving materials used for outdoor development in Sri
Lanka. Separate sections were paved with different paving materials and same
colour was applied on the surface to avoid the colour effect. The results obtained
were evaluated and the thermal performances were compared. The results
indicated that the material and its surface colour of surface influenced thermal
performance.

Keywords: Outdoor development
Paving materials
Thermal comfort

Surface colour

1 Introduction

Continually increasing urbanization and careless development of infrastructure and

open spaces with hard surfaces creates a major impact on the urban climate (Doulos
et al. 2004; Babić et al. 2012). Most of the outdoor open spaces adversely affects the
behavior of users and the aesthetic aspects of built environment and causes the dete-
rioration of the visual image of most urban areas (Mendoza et al. 2012). One of the
most dominant surface elements contributing to the city is artificial pavements exposed
to solar radiation because there is no evaporative cooling effect over the surface. They
generate long wave radiation than a vegetated surface. They create not only high
surface temperatures but also influence the near surface air temperature (Chudnovsky
et al. 2004; Kleerekoper et al. 2012; Spagnolo and de Dear 2003). (Rehan 2013)
recommended to use locally sourced paving materials manufactured with sustainable

© Springer Nature Singapore Pte Ltd. 2020

R. Dissanayake and P. Mendis (Eds.): ICSBE 2018, LNCE 44, pp. 340–352, 2020.
https://doi.org/10.1007/978-981-13-9749-3_31
 Performance of Paving Materials in Outdoor Landscaping 341

practices to reduce life-cycle environmental impacts (Rehan 2013). Therefore, sus-

tainability in the outdoor development is one of the most important strategies for urban
design.
On the other hand, well planned and well connected infrastructures will lead to
increased walking activities while limiting the demand for motorized transport. The
environmental benefit of limiting motorized transportation are the reduction of fossil
energy consumption and greenhouse gas (GHG) emissions (van der Zee 1990).
Creating a pedestrian-friendly environment could promote walking as a mode of sus-
tainable transportation and safe recreation, thus encouraging citizens to be physically
active (Mendoza et al. 2012), Chiesura 2004, (Kelly et al. 2011). Therefore, pedestrian
pavements and recreational areas are designed to be technically, economically and
aesthetically suitable for users and developers should involve a pedestrian need hier-
archy in outdoor development. Therefore, the design process must apply comprehen-
sive life cycle environmental data to identify suitable construction solutions and urban
management strategies that contribute to minimizing environmental impacts (Grim-
mond et al. 2010; Rahaman and Ohmori n.d.). Construction materials play a very
important role in the thermal balance in the urban environment (Santamouris et al.
2011; Scudo and Dessì 2006). The materials’ thermal balance is determined mainly by
their reflectivity of solar radiation and their emissivity of long wave radiation during
daytime (Babić et al. 2012). (Doulos et al. 2004) has determined that the materials
thermal balance by their reflectivity of solar radiation and their emissivity of long wave
radiation during day time and has reported that the mean daily surface temperature
differences are mainly caused by the different albedo value. Further they have men-
tioned that the physical characteristics of the material that affected their albedo are the
colour, surface texture and the construction material. The rough and dark coloured
surfaces tend to absorb more solar radiation than smooth light coloured and flat sur-
faces and therefore the rough and dark coloured surfaces are warmer than the light
coloured (Doulos et al. 2004). They have defined the materials light in colour and
having smooth surface as ‘cold’ materials and those having rough and dark coloured
materials as ‘warm’ materials.
The use of proper materials to reduce the temperature effect and improve the
thermal characteristics of the outdoor spaces has had extended interest during the resent
years. Researchers are trying to examine the impact of materials’ thermal performance
on the urban temperature (Doulos et al. 2004; Erell et al. 2014).
The urban heat island phenomenon (UHI) is one of the most important occurrences
affecting the climate and environment quality and negative influence on the standard of
living and health of city residents as well as increasing energy consumption (Babić
et al. 2012; Chudnovsky et al. 2004; Kleerekoper et al. 2012). Most of the urban areas
where large proportions consist of dry impervious pavements or buildings increase
surface temperature than many rural areas, where plant cover and evaporation of soil
moisture may be moderate (Santamouris et al. 2012; Kleerekoper et al. 2012).
The urban heat island phenomenon is mainly related to the high density of
buildings and urban structures that absorb solar radiation, the use of highly absorbent
materials, the lack of green spaces and the absence of vegetation (Chudnovsky et al.
2004; Santamouris et al. 2012). Since, it is very important to identify the thermal
behavior of the urban surface, researchers recommend to use cool pavements wherever
342 G. H. Galabada and R. U. Halwatura

possible to reduce the urban heat island effect, particularly in the urban environment of
cities with hot climate. However, warm materials are used in urban areas instead of
cold material due to economic and aesthetic reasons. As a result, the temperature in the
urban environment is raised and the demand of cooling energy in the building is
increased (Doulos et al. 2004; Rehan 2013; Santamouris et al. 2012).
Heat island phenomenon increases temperatures in urban areas, increases the
energy consumption for cooling purposes and affects the global environmental quality
of cities. Therefore, the hidden approach of the lowering top layer air temperature in a
city gives the benefit of reduced air conditioning loads in buildings; and improved
thermal comfort for pedestrians in outdoor urban spaces. To mitigate the heating of
such surfaces that result in elevated air temperature and contribute to the urban heat
island phenomenon, numerous researchers have suggested the use of paving materials
which have a high albedo as much as possible. These studies have directed recom-
mendations and guidelines for urban planners and architects on practical implemen-
tation of cool paving which are presenting a high solar reflectivity and emissivity as an
effective mitigation technique when applied to buildings and open spaces (Santamouris
et al. 2012; Dareeju et al. 2013).
Although many experimental data are available to evaluate the performance of cool
materials in isolated buildings (Jaffal et al. 2012; Halwatura 2013) very limited
information is available concerning the mitigation potential of the reflective materials
when applied in open urban areas (Santamouris et al. 2011).
Therefore, this paper represents the results of field experiment done to compare the
temperature behavior on commonly used paving materials in Sri Lanka for outdoor
landscaping.

2 Objective of the Research

To compare the actual performance of existing paving materials used for outdoor
development in Sri Lanka.

3 Materials and Method

The research was conducted in University of Moratuwa premises, one of the main
universities in Sri Lanka. The climate of the area is characterized as hot and dry during
the observations, which is similar to its climate for most parts of the year as well as
most city areas in Sri Lanka. The climatic data of the area shows that February and
March were comparatively the hottest months of the year, with the daily average
temperature ranging from 27–28 °C and the daily maximum temperature ranging from
30 °C to 33 °C. The daily average relative humidity is around 70–80%, and the
average wind speed is 5–6 km/hr during the period of measurement [10]. Figure 1
shows the locations of the four observation areas, which were paved with three different
paving materials and the fourth left natural ground. The selected paving materials were
cement concrete blocks, burnt clay bricks and asphalts. An areas of 2 m  2 m was
paved with selected paving materials while keeping the same sub-base. Temperature
 Performance of Paving Materials in Outdoor Landscaping 343

measurements were taken as two sets of observations. The materials were tested with
their native colour and then it was tested after painting them with the same colour to
avoid the Albedo effect. The colour applied was beige. Figure 2 shows the four
observation areas with three different paving materials after painting while Fig. 3
shows the change in colour on the paving areas.

Natural Ground Cement concrete blocks

Asphalt Burnt Brick

Fig. 1. Paving areas with their native colour.

For all observation areas, the temperatures were recorded on the top surface and
point which is 300 mm above from the top surface simultaneously at 10 min intervals.
The temperatures were collected using a data logger. The ground surface and the
ambient temperatures were also recorded at the same time intervals. Observations were
recorded for four days and 24 h for each day.
344 G. H. Galabada and R. U. Halwatura

Asphalt Burnt Brick

Cement concrete Block Data Logger

Fig. 2. Coloured paving areas and data logger

Asphalt
k

Burnt brick

Cement
concrete

Fig. 3. Change in colour on paving areas.

 Performance of Paving Materials in Outdoor Landscaping 345

4 Results and Discussion

The average surface temperature of the studied objects, the ambient temperature and
the natural ground temperature were evaluated to compare temperature variation of
manmade hard surface with natural surface. Results obtained from temperature mea-
surements were analyzed to compare the temperature variation for each paving
material.
The temperature variations on the paved area for 24 h in the 1st case of this study
were analyzed first. (i.e. the areas paved with selected paving materials with their native
colour). The results observed are shown in Fig. 4. From 00.00 h to 7.00 h all the
surfaces show nearly same temperature with variation less than 0.8 °C. After 7.00 h the
temperatures increased rapidly for all three paved areas. The maximum average
ambient temperature of these day was 32.71 °C at 12.00 h. The maximum temperature
of the asphalt top surface was 52.78 °C at 13.00 h, which is 20.07 °C higher than the
ambient temperature at the time. At the same time, the temperature at a height of
300 mm above the surface reached 33.6 °C, showing a difference between the ambient
temperature and the temperature at a height of 300 mm above of 0.89 °C. The tem-
perature difference between top surface and the 300 mm above was 19.18 °C.
The maximum temperature on the top surface of burnt brick paved area was
47.56 °C at 14.00 h. At this time the ambient temperature was 31.33 °C. The tem-
perature difference between top surface and the ambient temperature was 16.23 °C. At
the same time, the temperature at a height of 300 mm above the surface reached
32.74 °C. This shows temperature difference of 1.4 °C between ambient temperature
and the temperature at 300 mm above and 14.82 °C between the top surface and
300 mm above.
In the case of cement block paved area the maximum temperature on the top surface
of was 49.79 °C, which is 18.46 °C higher than the ambient temperature at 14.00 h. At
the same time, the temperature at a height of 300 mm above the top surface reached
31.99 °C, showing a difference of 0.66 °C between ambient temperature and the
temperature at a height of 300. The temperature difference between top surface and the
300 mm above was 17.80 °C.
All most all the cases the top surface temperature remains above the ambient
throughout the day, while this effect was become null after 20.00 h at a height of
300 mm above the paved surface.
The maximum temperature difference between surface and the ambient temperature
was shown on dark asphalt, while the second maximum was given by cement concrete.
The temperature difference between 300 mm above and the ambient temperature shows
a significant difference in the day time for dark asphalt and burnt brick. However, the
same difference was not that significant for cement paving. Hence, it’s clear that the
darker paving materials give more contribution to the urban heat island than the light
colour surfaces.
346 G. H. Galabada and R. U. Halwatura

Fig. 4. The temperature variations on paved areas with their native colour
 Performance of Paving Materials in Outdoor Landscaping 347

Then the same testing was repeated for the painted paved areas. (i.e. the areas paved
with same selected paving materials by painting the surface). Figure 5 shows the
temperature variation for painted areas. The temperature differences from 01.00 h to
7.00 h were marginal and increased slowly, however there was slight temperature
differences between different materials. After 7.00 h the temperatures increased rapidly
for all three paved areas. The maximum average ambient temperature of this day was
34.43 °C at 12.00 h. The maximum temperature on the asphalt top surface was
recorded as 51.73 °C showing a difference of 17.3 °C with the ambient temperature.
The temperature difference between the ambient and surface temperature was 3.27 °C
less than the temperature difference between ambient temperature and surface tem-
perature on their native colour. At the same time, the temperature at a height of
300 mm above the top surface reached 38.88 °C, showing a difference between the
ambient temperature and the temperature at a height of 300 mm above the top surface
of 4.45 °C. The temperature difference between top surface and the 300 mm above was
12.85 °C. This shows 6.33 °C lesser value than that of first case.
The maximum temperature on the top surface of burnt brick paved area was
43.99 °C at 13.00 h. showing a 10.11 °C difference between the ambient temperatures
at that time. This was 6.12 °C less than the temperature difference between ambient
temperature and surface temperature on their native colour. At the same time, the
temperature at a height of 300 mm above the top surface reached 38.03 °C, showing a
difference between the ambient temperature and the temperature at a height of 300 mm
above the top surface of 4.15 °C. The temperature difference between top surface and
the 300 mm above was 5.96 °C. This shows 8.87 °C lesser value than that of first case.
In the case of the cement block paved area, the maximum temperature on the top
surface was 42.97 °C at 13.00 h showing a difference between the ambient temperature
of 9.1 °C. At the same time, the temperature at a height of 300 mm above the top
surface reached 33.43 °C, showing very little difference between the ambient tem-
perature and the temperature at a height of 300 mm above of 0.45 °C. The temperature
difference between top surface and the 300 mm above was 9.54 °C again after 20.00 h,
the temperatures of the top surface and at a height of 300 mm above the top surface
cooled down nearer to ambient temperature.
After applying the same colour on the surfaces, the difference in temperature with
ambient temperature decreased by a considerable amount. The difference in tempera-
tures between surface and 300 mm above were also decreased by a significant amount.
This shows that the colour is an important factor, which influences the thermal behavior
of paving surface and the urban heat island.
Figure 6 shows the comparison of temperature variation on the top surface of
selected paving areas with their own colour, natural ground and ambient temperature
while Fig. 7 shows the temperature variation for paving after painting to the same
colour. In the first case almost all the surfaces showed a minor temperature variations
during morning (i.e. 00.00 h to 7.00 h). Asphalt, burnt brick and cement concrete block
paving surfaces heated up quicker than ground. During 13.00–14.00 h asphalt surface
showed the highest temperature with a difference of 10.45 °C with natural ground.
Cement blocks and brick surfaces took the second and third places with temperature
differences of 6.88 °C and 4.75 °C respectively.
348 G. H. Galabada and R. U. Halwatura

Fig. 5. The temperature variations on paved areas after applying colour

 Performance of Paving Materials in Outdoor Landscaping 349

Fig. 6. Comparison of temperature variation on the top surface with native colour

Fig. 7. Comparison of temperature variation on the top surface with coloured surface

In the second case, temperature variation on the top surfaces were less during early
morning (i.e. 00 h to 07 h). At peak temperature during day time (at 13–14 h), the
asphalt surface showed the highest temperature difference of 5.87 °C with natural
ground. However, it was less when compared to the temperature difference between
natural ground and the asphalt top surface with their native colour, which was 10.45 °C.
Figures 8 and 9 show the comparison of temperature variation on a 300 mm above
the surface of all types of paving areas and the ambient temperature. For both cases,
350 G. H. Galabada and R. U. Halwatura

Fig. 8. Comparison of temperature variation of 300 mm above top surface with native coloured
surface

Fig. 9. Comparison of temperature variation of 300 mm above top surface with coloured
surface
 Performance of Paving Materials in Outdoor Landscaping 351

though almost all the temperatures were a similar pattern during 00.00 h to 7.00 h and
20.00 h to 00.00 h, the temperature between 7.00 h to 20.00 h significantly differed
than that of the ambient temperature. In the first case, during the peak time at 13.00–
14.00 h, the asphalt surface showed the highest temperature with a difference of
27.41 °C than the ambient temperature and that of the coloured asphalt was 14.21 °C.

5 Summary and Conclusions

Nowadays, much development is seen with many models and techniques for designing
pedestrian ways or sidewalks around the world. However, in Sri Lanka, the transport
authorities still think of managing only the motorized vehicles. Therefore, the most
vulnerable road users, the pedestrians, are neglected with regard to their safety as well
as convenience. This research work is an attempt to raise awareness in transportation
planners and researchers of the existing paving materials and their thermal properties,
with regard to their comfort. From the present study, it can be concluded that manmade
hard surfaces heat more than natural ground. Further, paved areas exposed to direct
solar radiation elevated the temperature of the atmosphere above them throughout the
day. This indicates that such impervious artificial surfaces can be regarded as “heat
sources” to the urban environment and the usage of impervious pavements should be
reduced. The knowledge of thermal behavior of various urban surfaces and landscape
components is an important tool for planners and designers. If utilized properly, it can
lead to climatic rehabilitation in urban areas and an improvement of the outdoor
thermal environment.

References
Babić S, Deluka-Tibljaš A, Cuculić M, Šurdonja S (2012) Analysis of pavement surface heating
in urban areas. Građevinar 64:125–132
Chiesura A (2004) The role of urban parks for the sustainable city. Landsc Urban Plan 68
(1):129–138
Chudnovsky A, Ben-Dor E, Saaroni H (2004) Diurnal thermal behavior of selected urban objects
using remote sensing measurements. Energy Build 36:1063–1074. https://doi.org/10.1016/j.
enbuild.2004.01.052
Dareeju B, Meegahage JN, Halwatura RU (2013) Influence of grass cover on flat reinforced
concrete slabs in a tropical climate
Doulos L, Santamouris M, Livada I (2004) Passive cooling of outdoor urban spaces. The role of
materials. Sol Energy 77:231–249. https://doi.org/10.1016/j.solener.2004.04.005
Erell E, Pearlmutter D, Boneh D, Kutiel PB (2014) Effect of high-albedo materials on pedestrian
heat stress in urban street canyons. In: Urban climate, ICUC8: the 8th international conference
on urban climate and the 10th symposium on the urban environment, vol 10, pp 367–386.
https://doi.org/10.1016/j.uclim.2013.10.005
352 G. H. Galabada and R. U. Halwatura

Grimmond CSB, Roth M, Oke TR, Au YC, Best M, Betts R, Carmichael G, Cleugh H,
Dabberdt W, Emmanuel R, Freitas E, Fortuniak K, Hanna S, Klein P, Kalkstein LS, Liu CH,
Nickson A, Pearlmutter D, Sailor D, Voogt J (2010) Climate and more sustainable cities:
climate information for improved planning and management of cities (producers/capabilities
perspective). In: Procedia environmental sciences, world climate conference, vol 3, no 1,
pp 247–274. https://doi.org/10.1016/j.proenv.2010.09.016
Halwatura RU (2013) Effect of turf roof slabs on indoor thermal performance in tropical climates:
a life cycle cost approach [WWW Document]. J Constr Eng. https://doi.org/10.1155/2013/
845158
Jaffal I, Ouldboukhitine S-E, Belarbi R (2012) A comprehensive study of the impact of green
roofs on building energy performance. Renew Energy 43:157–164. https://doi.org/10.1016/j.
renene.2011.12.004
Kelly CE, Tight MR, Hodgson FC, Page MW (2011) A comparison of three methods for
assessing the walkability of the pedestrian environment. J Transp Geogr 19:1500–1508.
https://doi.org/10.1016/j.jtrangeo.2010.08.001 Special section on Alternative Travel futures
Kleerekoper L, van Esch M, Salcedo TB (2012) How to make a city climate-proof, addressing
the urban heat island effect. Resour Conserv Recycl Clim Proofing Cities 64:30–38. https://
doi.org/10.1016/j.resconrec.2011.06.004
Mendoza J-MF, Oliver-Solà J, Gabarrell X, Rieradevall J, Josa A (2012) Planning strategies for
promoting environmentally suitable pedestrian pavements in cities. Transp Res Part D: Transp
Environ 17:442–450. https://doi.org/10.1016/j.trd.2012.05.008
Rahaman KR, Ohmori N, Harata N Evaluation of the roadside walkway environment of Dhaka
city. Proc East Asia Soc Transp Stud 5:16
Rehan RM (2013) Sustainable streetscape as an effective tool in sustainable urban design.
HBRC J 9:173–186. https://doi.org/10.1016/j.hbrcj.2013.03.001
Santamouris M, Gaitani N, Spanou A, Saliari M, Giannopoulou K, Vasilakopoulou K,
Kardomateas T (2012) Using cool paving materials to improve microclimate of urban areas –
design realization and results of the Flisvos project. Build Environ 53:128–136. https://doi.
org/10.1016/j.buildenv.2012.01.022
Santamouris M, Synnefa A, Karlessi T (2011) Using advanced cool materials in the urban built
environment to mitigate heat islands and improve thermal comfort conditions. Sol Energy
Prog Sol Energy 2(85):3085–3102. https://doi.org/10.1016/j.solener.2010.12.023
Scudo G, Dessì V (2006) Thermal comfort in urban space renewal. In: Proceeding 23rd PLEA
Spagnolo J, de Dear R (2003) A field study of thermal comfort in outdoor and semi-outdoor
environments in subtropical Sydney Australia. Build Environ 38:721–738. https://doi.org/10.
1016/S0360-1323(02)00209-3
van der Zee D (1990) The complex relationship between landscape and recreation. Landscape
Ecol 4:225–236
 Determining Criticality of Performance
Indicators for a Construction Company

K. K. Tripathi1(&), K. N. Jha2, and A. K. Jain2

1
B. S. Anangpuria Institute of Technology and Management,
Faridabad 121004, India
[email protected]
2
Indian Institute of Technology Delhi, Hauz Khas, New Delhi 110016, India

Abstract. Because of stiff competition in the construction business, it is nec-

essary for construction companies to measure their performance on a regular
interval for long-term survival. Therefore, it is imperative for construction
companies to determine the parameters critical to the measurement of their
performance. This study identifies the performance indicators, financial as well
as non-financial, for construction companies to measure their performance. In a
previous study, the authors identified a total of 20 performance attributes and
applied principal component analysis which extracted six components called
performance indicators. These performance indicators are profitability and asset
management; key stakeholders’ satisfaction; time and cost predictability; envi-
ronment, health, and safety; quality consciousness; and employee turnover. The
present study utilized structural equation modelling (SEM) to determine the
criticality of these performance indicators for which a questionnaire survey and
structured interview approach were adopted. A total of 106 responses were
collected from 90 different construction companies operating in various parts of
India. The findings indicate that ‘quality consciousness’ is the most critical
performance indicator whereas the ‘employee turnover’ is the least critical
performance indicator in measuring the performance of a construction company.
The performance indicators obtained from this study may help construction
companies to measure and compare their performances with their competitors
and setting out a strategy to remain competitive in the market.

Keywords: Performance attributes
Performance indicators

Questionnaire survey
Principal component analysis
Construction company

1 Introduction

Performance measurement has been considered fundamental to management planning

and control by both academicians and practitioners and hence gained more attention
from many researchers (Tsolas 2011). According to Kagioglu et al. (2001), it is defined
as the process of evaluating how successfully the companies or individuals have been
achieving their strategies and objectives. Performance measurement helps the con-
struction companies to concentrate on their long-term strategic view, objectives, and
optimization of their operations in the companies.

© Springer Nature Singapore Pte Ltd. 2020

R. Dissanayake and P. Mendis (Eds.): ICSBE 2018, LNCE 44, pp. 353–364, 2020.
https://doi.org/10.1007/978-981-13-9749-3_32
354 K. K. Tripathi et al.

To measure the performance of the construction companies, the companies must

first identify the appropriate key performance indicators (KPIs). According to Cox
et al. (2003), “performance indicators can be defined either by quantitative measures
($/unit) or qualitative measures such as worker behaviour on the job”. It helps top-level
management to watch the performance of the corporate or department in one place. The
identification of KPIs is the commencement in developing a correct framework for
measurement of a construction company’s performance (Lin et al. 2011).
Traditionally, performances of the companies were measured in financial terms
such as turnover, profit, sales per employee, return on investment, etc. which has been
criticized (Love and Holt 2000). It has also been found that financial parameters
measure only the short-term performance of the company and ignores the issues related
to the long-term sustainability of the company (Kaplan and Norton 1996). The tradi-
tional financial measures of performance do not reflect the need of stakeholders, fail to
provide information about what customers are getting against what they really wanted
and do not identify the performance of the competitors (Isik 2009). Hence, construction
companies should use the performance indicators, both financial as well as non-
financial to evaluate and compare their performances with others which will enhance
the effectiveness and efficiency of the company.
Performance measurement of any construction company in terms of their failure or
success is very complex because of the participation of various stakeholders like the
clients, project management consultants, contractors, and public etc. For a given
company, the objectives of all the stakeholders need not be the same. For one stake-
holder, the parameter for measuring the performance of a company may be high profit
while for the other stakeholder, it can be client satisfaction. Hence defining the per-
formance in terms of failure or success without defining the objectives of the various
stakeholders involved has no meaning (Tripathi and Jha 2018a).
In the present study, the factors based on which the performance of a construction
company can be measured, have been described. Researchers have identified various
performance indicators for measuring the performance of a construction company. But
their research works are basically project specific and in the context of developed
countries which may not be applicable in the performance measurement of construction
companies in other countries. Very few of them have focused on the performance
indicators of construction companies working in developing countries like India
(Tripathi and Jha 2018a). Moreover, the performance of projects and the construction
companies were measured mainly based on cost, time, and quality achieved on those
projects whereas, the performance relative to these three parameters is influenced by
some other parameters like environment, health, and safety (EHS), productivity of the
employees, and key stakeholders’ satisfaction etc. This study attempts to identify all
such parameters which measure the performance of a construction company.

2 Literature Review

The purpose of the current study is to determine those indicators that are critical for the
measurement of construction company’s performance. Very limited studies are avail-
able in the literature in this area. These are briefly discussed below.
 Determining Criticality of Performance Indicators for a Construction Company 355

Mbugua et al. (1999) developed a framework which will develop the information
and hence give improved construction performance grounded on financial (liquidity,
profit, potency etc.) and non-financial (leadership, customer’s satisfaction, impact on
society, learning and growth, quality etc.) measures. Cox et al. (2003) established a
correlation between quantitative and qualitative performance measurement parameters
to identify the most extensively used parameter. They found that quality control, timely
completion, cost, safety, etc. are the extremely vital indicators for the construction
profession. But they neither incorporated company level indicators like the company’s
financial standing, market condition, relationship with external agencies etc. nor
evaluated its efficiency.
Bassioni et al. (2004, 2005) proposed a theoretical framework employing a bal-
anced scorecard and business excellence model to evaluate the business performance of
construction companies. They divided the framework into performance driving factors
(resource management, leadership, strategic management, capital management, risk
management, work culture, etc.) and performance results factors (people, partnership
and supplier results, project results, customer and society results, organizational
business results, etc.). Elyamany et al. (2007) developed a performance measurement
model based on the financial ratios along with economic and trade factors and con-
cluded that the company with higher financial performance index had better perfor-
mance. Balatbat et al. (2010) measured the performance of Australian publicly listed
company with other listed companies in Australia using profitability ratio, market
performance, and equity analysis and performance ratio.
Luu et al. (2008) performed the strengths, weaknesses, opportunities, and threat
(SWOT) analysis to judge the strategic performance of huge contractors in Vietnam.
The KPIs known were from four perspectives: money, customer, internal method, and
learning and growth. Chan (2009) developed a scientific performance measuring
framework for the Malaysian industry to watch their progress towards achieving the
goals commenced within the industry master plan 2006–2015. The performance
measures known were identical to the ones known by Luu et al. (2008). Tsolas (2011)
integrated the data envelopment analysis (DEA) and the ratio analysis while developing
a framework to gauge the performance of the construction companies with regards to
profit and effectiveness of the construction companies listed on the Athens Exchange.
Ali et al. (2013) used 47 indicators for performance evaluation of construction
companies in Saudi Arabia via relative importance index (RII). Profitability, cash flow,
service and work quality, financial stability, market share, growth, customer satisfac-
tion, safety, business efficiency, and effectiveness of planning were the highest 10
KPIs. Hassaan et al. (2013) applied the analytical hierarchy process (AHP) along with a
fuzzy set theory for performance evaluation of contractors on 34 quantitative and
qualitative criteria during an award of tender in Egypt. According to the respondents,
the financial stability and past performance were the foremost significant measures for
contractor selection.
The literature review reveals that various frameworks/models were developed by
researchers to gauge the performance of construction companies. Most of the
researchers administered the research focused on a specific country, however, none of
them have administered the research for India. In most of the studies, only contractors
were considered by the researchers, but this study has thought about the opinions of
356 K. K. Tripathi et al.

other stakeholders such as the clients and project management consultants along with
the contractors. This study also attempts to include all possible performance evaluation
parameters pertinent to a construction company. Hence, authors have attempted to fill
the gaps in the previous studies with the help of the current study.

3 Research Method

The research method contains a total of four steps. The first two steps were carried out
by the authors in their previous study and the remaining two steps are pertaining to the
present study. These are described in detail in the subsequent steps.
Step 1: Identification of Performance Attributes and Data Collection
Twenty attributes were extracted from the literature to measure the performance of a
construction company. Table 1 shows the list of these attributes along with their
sources. Based on the attributes identified, a questionnaire was designed. A five-point
Likert scale was used to measure the degree of importance of these attributes in
measuring the performance of a construction company. In the Likert scale, the degree
of importance of performance attributes ranges from 1 (very low importance) to 5 (very
high importance).
A total of 106 completed questionnaires were received from 90 different con-
struction companies. Out of 106 questionnaires, 29 questionnaires were collected via
email and 77 questionnaires via personal meeting. The average experience of the
respondents was 20 years whereas the average experience of the companies partici-
pated in the survey was 21 years.

Table 1. List of performance attributes and their sources

Sl. Performance attributes with id Sources
no.
1 Size of the organization (PA-1) Mbugua et al. (1999), Chan (2009)
2 The productivity of employees (PA-2) Mbugua et al. (1999), Cox et al. (2003),
Chan (2009)
3 Good track record of timely completion Luu et al. (2008), Skibniewski and Ghosh
of the projects (PA-3) (2009)
4 Health and safety consciousness (PA-4) Cox et al. (2003), Chan (2009), Luu et al.
(2008)
5 Customer satisfaction in terms of Rimbalova and Vilcekova (2013),
product and services (PA-5) Menches and Hanna (2006)
6 Client satisfaction in terms of product Chan (2009), Skibniewski and Ghosh
and services (PA-6) (2009), Nemcova-Zunana (2009)
7 Cost performance of projects (PA-7) Menches and Hanna (2006), Bassioni et al.
(2004)
(continued)
 Determining Criticality of Performance Indicators for a Construction Company 357

Table 1. (continued)
Sl. Performance attributes with id Sources
no.
8 Impact on society (PA-8) Mbugua et al. (1999), Nemcova-Zunana
(2009)
9 Impact on the environment (PA-9) Rimbalova and Vilcekova (2013),
Nemcova-Zunana (2009)
10 Optimum liquidity ratio (PA-10) Elyamany et al. (2007), Balatbat et al.
(2010)
11 Higher profitability ratio (PA-11) Skibniewski and Ghosh (2009), Horta
et al. (2010)
12 The higher annual growth rate of the Chan (2009), Luu et al. (2008), Horta et al.
organization (PA-12) (2010), Yu et al. (2007)
13 Predictability of cost in construction Rimbalova and Vilcekova (2013), Chan
(PA-13) (2009)
14 Predictability of time in construction Nemcova-Zunana (2009), Bassioni et al.
(PA-14) (2004)
15 Rework or defect rectification (PA-15) Menches and Hanna (2006), Kagioglou
et al. (2001)
16 Adopting learning and growth culture Mbugua et al. (1999), Chan (2009), Luu
in the organization (PA-16) et al. (2008)
17 Higher wages of the employees Rimbalova and Vilcekova (2013),
(PA-17) Nemcova-Zunana (2009)
18 Employee turnover (PA-18) Chan (2009), Rimbalova and Vilcekova
(2013)
19 A good relationship with the client Mbugua et al. (1999), Menches and Hanna
(PA-19) (2006)
20 Annual construction demand/market Chan (2009), Luu et al. (2008), Yu et al.
share (PA-20) (2007)

Step 2: Identification of Performance Indicators

Performance attributes identified in Step 1 were grouped into six components, called
performance indicators when subjected to principal component analysis. The output of
the principal component analysis indicated that the Kaiser Meyer Olkin (KMO) value
was 0.793 (>0.05) indicating sample adequacy for principal component analysis. The
result of the principal component analysis is shown in Table 2. 1. These indicators are
profitability and asset management (PI-1), key stakeholders’ satisfaction (PI-2), time
and cost predictability (PI-3), environment, health, and safety (PI-4), quality con-
sciousness (PI-5), and employee turnover (PI-6) as shown in Table 2 (Adapted from
Tripathi and Jha 2018a). The total variance explained by all the indicators accounts for
67.448%.
358 K. K. Tripathi et al.

Table 2. Results of the principal component analysis

Sl. Performance indicators with variance Performance attributes with factor loading
no. (%)
1 PI-1 Profitability and asset PA-12 The higher annual growth rate
management (13.893%) (0.801)
PA-11 Higher profitability ratio (0.742)
PA-10 Optimum liquidity ratio (0.637)
PA-2 The productivity of employees
(0.609)
2 PI-2 Key stakeholders’ satisfaction PA-5 Customer satisfaction (0.830)
(12.688%) PA-6 Client satisfaction (0.816)
3 PI-3 Time and cost predictability PA-13 Predictability of cost in
(11.698%) construction (0.876)
PA-14 Predictability of time in
construction (0.862)
4 PI-4 Environment, health, and safety PA-9 Impact on the environment (0.806)
(11.161%) PA-4 Health and safety consciousness
(0.734)
5 PI-5 Quality consciousness (10.581%) PA-15 Rework or defect rectification
(0.769)
PA-16 Adopting learning and growth
culture (0.719)
PA-20 Annual construction
demand/market share (0.525)
6 PI-6 Employee turnover (7.427%) PA-18 Employee turnover (0.673)

Reliability Test
To ensure the internal consistency within the attributes grouped under components and
reliability of the data, Cronbach’s alpha (Ca) reliability test was performed using
statistical package for social science (SPSS) version 21. The Ca value ranges between
0 and 1. A higher Ca value points towards the greater internal consistency and vice
versa. As per Tripathi and Jha 2018b, a Ca value greater than 0.7 is acceptable. In this
study, the Ca value is 0.844, which shows a good overall internal consistency among
the attributes. Hence, the attributes grouped under the individual performance indica-
tors were considered reliable for further analysis (Chan et al. 2014).
Step 3: Development of a Hypothesized Model
After grouping of the performance attributes, a hypothesized SEM model was proposed
to test the relationship between performance indicators and performance of a company
as shown in Fig. 1.
 Determining Criticality of Performance Indicators for a Construction Company 359

Fig. 1. Hypothesized and final model

Based on the model proposed, the hypothesis that performance indicators measure
the construction company’s performance was tested as follows:
1. Null hypothesis (H0): Path coefficient between performance indicators and the
construction company’s performance is not significantly different from zero.
2. Alternate hypothesis (Ha): Performance indicators measure the performance of a
construction company.

Step 4: Validation of the Hypothesized Model

The adequacy of the SEM model is established by assessing the results of the
covariance structural analysis, which is shown by the various goodness-of-fit
(GOF) indices suggested by different researchers. If the model is not adequate to
interpret, it needs to be revised. Different GOF indices evaluate the adequacy of a
model from different aspects. From the many GOF indices available within the SEM
literature, following GOF indices were proposed in this study for validating the
hypothesized model (Tripathi and Jha 2018b).
360 K. K. Tripathi et al.

(1) The ratio of chi-square (v2) to the degree of freedom (df), (2) Incremental fit
index (IFI), (3) Tucker-Lewis index (TLI), (4) Comparative fit index (CFI), and (5) The
root-mean-square error of approximation (RMSEA).
The recommended level of these indices is given in Table 3 (Tripathi and Jha
2018b). The hypothesized model was analysed using an analysis of moment structure
(AMOS) version 21.

Table 3. GOF measures (Adapted from Tripathi and Jha 2018b)

Sl. The goodness of fit Recommended level Values of GOF indices
no. (GOF) indices of GOF indices obtained in the final model
1 Chi-square/degree of freedom 1 to 2 1.171
(v2/df)
2 Incremental fit index (IFI) 0 (no fit) −1 (perfect 0.969
fit)
3 Tucker-Lewis index (TLI) 0 (no fit) −1 (perfect 0.952
fit)
4 Comparative fit index (CFI) 0 (no fit) −1 (perfect 0.967
fit)
5 Root mean square error of <0.05 (very good) - 0.041
approximation (RMSEA) 0.1 (threshold)

Table 3 shows the GOF indices of the hypothesized SEM model. The values of
GOF indices indicate that the hypothesized SEM model was adequate to explain the
interrelationships between performance indicators and the construction company’s
performance. Hence, the hypothesized SEM model was not revised and was acceptable
for interpretation.
All the standardized path coefficients shown on the arrow lines in the Fig. 1 were
positive and statistically significant which indicate that there is a relationship between
the construction company’s performance and performance indicators. The larger path
coefficient indicates that the attributes or indicators are more critical for measuring the
construction company’s performance. Accordingly, ‘quality consciousness’ emerged as
the most critical performance indicator whereas the ‘employee turnover’ emerged as
the least important performance indicator. The hypothesis H1, which assumes that the
performance indicators measure the performance of the construction company is found
to be supported because the path coefficients were significant at the 0.05 significance
level.

4 Result and Discussion

The results of the study reveal that the quality consciousness (PI-5) is the most critical
performance indicator with a path coefficient of 0.78. Quality consciousness is an
important area for good quality construction and its integration in the company.
Nowadays, clients are more worried about the quality rather than the cost. Also,
 Determining Criticality of Performance Indicators for a Construction Company 361

correcting a work of poor quality involves an investment of time and money and
creates an obstruction in the construction. Hence, quality consciousness is considered
as a significant indicator for the measurement of the performance of a construction
company. Construction companies can increase their market by enhancing client sat-
isfaction with a superior product.
Environment, health, and safety (PI-4) is the second most critical performance
indicator with a path coefficient of 0.71. It must be an environmental responsibility for
both an individual as well as the company. Much of the earth’s resources are controlled
by the construction companies which diminish environmental nuisance. In addition, the
cost of construction can also be reduced by environment-friendly construction using
energy-efficient construction materials and techniques. Inefficient health and safety
management on construction sites results in the loss of life and wastage of construction
materials and hence increases the associated costs. Therefore, construction companies
must be held responsible for the society as well as the environment because they
interfere in various areas of our life.
Profitability and asset management (PI-1) received a path coefficient of 0.70 and
was placed in the third position. Asset management can be defined as the technique of
managing assets for the companies through cash, bonds, and stocks. A company’s
performance is gauged by the profitability and assets of the company. Profitability is
one of the most significant parameters for measuring the performance of a construction
company (Yu et al. 2007).
Time and cost predictability (PI-3) received a path coefficient of 0.67 and was
placed in the fourth position. Predictability is very important for any construction
company because it psychologically influences client. If clients are confident that their
project will be handed over by the company within the cost and within the scheduled
time, then it is most likely that they can deal with the company. The lack of pre-
dictability in the company negatively influences the clients and reduces the chances of
doing business with the company. Therefore, all companies must possess predictability
otherwise its number of clients will decrease affecting its overall profits.
Key stakeholders’ satisfaction (PI-2) received a path coefficient of 0.54 and was
placed in the fifth position. Key stakeholders’ satisfaction is a significant indicator for
measuring the construction company performance. Client satisfaction is found to be
fundamental to the success of the business. Construction companies always attempt to
maximize the Key stakeholders’ satisfaction through good quality of products and
services rendered by them. They manage their resources in such a way that their
client’s objectives are achieved to have repeat business.
Employees turnover (PI-6) was placed at the last position with a very low path
coefficient of 0.41. It is expressed as the percentage of the staff leaving the company for
one year. For any company, the objective should be to have employee turnover as low
as possible to maintain a consistent workforce and to grow more skilled staff. High
employee turnover results in higher cost to the company because additional time and
resources are required to fill the vacant position and to train a new employee. Low
employee turnover permits the company to concentrate on its business rather than
preparing the new employee, adapting them to the company’s needs. Therefore, low
employee turnover is an indicator of the company performing well in the industry.
362 K. K. Tripathi et al.

The above results show that the respondents have given more weight to the per-
formance indicators ‘quality consciousness’ and ‘environment, health, and safety’
when compared to the performance indicator ‘profitability and asset management’. It
indicates that the construction companies apprehend that the traditional financial
parameters are no longer a comprehensive measure of company performance. The
insufficiency of the traditional financial performance indicators has drawn the attention
of construction companies in non-financial performance indicators. Several construc-
tion companies have a long-term strategy that focuses on the satisfaction of the
stakeholders, good quality of the work, time and cost predictability etc. Hence, the
company performance from the perspective of their key stakeholders has become a
priority for construction companies.
The results of this study recommend that these indicators could be considered as a
yardstick for the Indian construction companies while measuring their performance
with their competitors. Most of these indicators are the prerequisite of the award of the
tender. Clients prefer the above criteria to be satisfied by the construction companies to
decrease the risk of budget and time overruns, low-quality work and workmanship and
so forth. Construction companies can enhance their performance by meeting these
criteria and increase the chances of getting increased business despite the competition
in the market.

5 Conclusion

This study tries to find performance indicators that can measure the performance of the
construction companies executing real estate projects. This study utilised a question-
naire survey and an SEM approach to analyse the causal relationship between per-
formance indicators and company’s performance. The results enable a construction
company to understand and formulate a strategy to perform well with their competitors.
The SEM supports the hypothesized positive relationships between performance
indicators and the performance of a construction company with very strong path
coefficients at the significance level of 0.05.
The SEM model reveals that the quality consciousness (PI-5) is the most critical
indicator for the measurement of construction company’s performance followed by
environment, health, and safety (PI-4), profitability and asset management (PI-1), time
and cost predictability (PI-3), key stakeholders’ satisfaction (PI-2), and employees
turnover (PI-6). Due to rapid changes and challenges in the construction business, the
traditional parameters of performance measurement may not be enough in developing
adequate results for the various stakeholders. It has been found that high financial
performance is essential but not enough for satisfactory performance of a company. All
these indicators contribute to enhancing the performance of a construction company
leading to sustainable growth. The probability of getting the increased business of a
construction company will increase if proper attention is drawn towards improvement
in these areas. The clients select a construction company which fulfils the above criteria
to ensure enhanced value for money and overall satisfactory results in construction.
The scope of this study was limited to the construction companies associated only
with the real estate business operating in India. Hence, the result of this study may or
 Determining Criticality of Performance Indicators for a Construction Company 363

may not be applicable to the construction companies in other countries. However,

despite various limitations, this study may be beneficial for neighboring countries like
Bangladesh, Nepal, Sri Lanka, Pakistan, Afghanistan etc. and other developing
countries because of resemblance in working conditions.

References
Ali HAEM, Al-Sulaihi IA, Al-Gahtani KS (2013) Indicators for measuring performance of
building construction companies in Kingdom of Saudi Arabia. J King Saud Univ - Eng Sci
25:125–134
Balatbat MCA, Lin C, Carmichael DG (2010) Comparative performance of publicly listed
construction companies: Australian evidence. Constr Manag Econ 28(9):919–932
Bassioni HA, Price ADF, Hassan TM (2004) Performance measurement in construction. J Manag
Eng ASCE 20(2):42–50
Bassioni HA, Price ADF, Hassan TM (2005) Building a conceptual framework for measuring
business performance in construction: an empirical evaluation. Constr Manag Econ 23
(5):495–507
Chan TK (2009) Measuring performance of Malaysian construction industry. Constr Manag
Econ 27(12):1231–1244
Chan IYS, Liu AMM, Richard Fellows R (2014) Role of leadership in fostering an innovation.
Constr Manag Econ 30(6):1–7
Cox RF, Issa RRA, Ahrens D (2003) Management’s perception of key performance indicators
for construction. J Constr Eng Manag ASCE 129(2):142–151
Elyamany A, Basha I, Zayed T (2007) Performance evaluating model for construction
companies: egyptian case study. J Constr Eng Manag ASCE 133(8):574–581
Hassaan HS, Fors MN, Shehata MS (2013) Fuzzy decision model for construction contractor’s
selection in Egypt: tender Phase. In: Proceedings of the international conference on industrial
engineering and engineering management, institute of electrical and electronics engineers
(IEEE), Bangkok, Thailand
Horta IM, Camanho AS, Da Costa JM (2010) Performance assessment of construction companies
integrating key performance indicators and data envelopment analysis. J Constr Eng
Manag ASCE 136(5):581–594
Isik Z (2009) A conceptual performance measurement framework for construction industry. Ph.
D. thesis, The Graduate School of Natural and Applied Science of Middle East Technical
University
Kagioglou M, Cooper R, Aouad G (2001) Performance management in construction: a
conceptual framework. Constr Manag Econ 19(1):85–95
Kaplan RS, Norton DP (1996) The balanced scorecard. Harvard Business School Press, Boston
Lin G, Shen GQ, Sun M, Kelly J (2011) Identification of key performance indicators for
measuring the performance of value management studies in construction. J Constr Eng
Manag ASCE 137(9):698–706
Love PED, Holt GD (2000) Construction business performance measurement: the stakeholder
perspective measurement alternative. Bus Process Manag J 6(5):408–416
Luu T-V, Kim S-Y, Cao H-L, Park Y-M (2008) Performance measurement of construction firms
in developing countries. Constr Manag Econ 26(4):373–386
364 K. K. Tripathi et al.

Mbugua LM, Harris P, Holt GD, Olomolaiye PO (1999) Framework for determining critical
success factors influencing construction business performance. In: Hughes W (ed) 15th
Annual ARCOM conference of association for research in construction management held in
Liverpool, John Moores University, 15–17 September 1999, vol 1, pp 255–264
Menches CL, Hanna AS (2006) Quantitative measurement of successful performance from the
project manager’s perspective. J Constr Eng Manag ASCE 132(12):1284–1293
Nemcova-Zuzana (2009) Application of key performance indicators in the construction industry:
a comparison of the United Kingdom with the Czech Republic. The European Council of
Construction Economists, 6 p
Rimbalova J, Vilcekova S (2013) The proposal of key performance indicators in facility
management and determination the Weights of Significance. J Civil Eng 8(2):73–84
Skibniewski MJ, Ghosh S (2009) Determination of key performance indicators with enterprise
resource planning systems in engineering construction firms. J Constr Eng Manag ASCE 135
(10):965–978
Tripathi KK, Jha KN (2018a) An empirical study on performance measurement factors for
construction organizations. KSCE J Civil Eng 22(4):1152–1166
Tripathi KK, Jha KN (2018b) Determining success factors for a construction organization: a
structural equation modeling approach. J Manag Eng 34(1):1–15
Tsolas IE (2011) Modelling profitability and effectiveness of Greek-listed construction firms: an
integrated DEA and ratio analysis. Constr Manag Econ 29(8):795–807
Yu I, Kim K, Jung Y, Chin S (2007) Comparable performance measurement system for
construction companies. J Manag Eng ASCE 23(3):131–139
 Seepage in Iranamadu Dam and Its Influence
on the Stability

T. Venuja and L. C. Kurukulasuriya(&)

Faculty of Engineering, University of Peradeniya, Peradeniya, Sri Lanka

[email protected]

Abstract. Earth dams have played an important role in agriculture since

ancient times. Failure of side slopes caused due to seepage or drawdown is one
of the major considerations in the design of earth dams as the failure of a slope is
catastrophic. The Iranamadu dam which was raised at different times showed a
significant increase of seepage when the reservoir level increased beyond a
certain elevation. In this research, the seepage problem of the Iranamadu dam
was analysed using Geo-Studio (SEEP/W) software to estimate the seepage rate
with the determination of the phreatic surface and to investigate the influence of
the anisotropy of hydraulic conductivity of core material on seepage. For these
analyses, volumetric water content function to be used in the unsaturated zone
above the phreatic surface was selected from the van Genuchten model. Both
steady and transient state seepage were carried out to compare the seepage rates
with field measurements and the effect of anisotropy of hydraulic conductivity
using the observed reservoir level history spanning over 170 days. Furthermore,
the stability of the dam was assessed under seepage conditions using the
SLOPE/W software. SEEP/W results show a good agreement in seepage rates
with field measurements under both seepage condition when the reservoir level
is above a certain elevation. Within a small range of reservoir level, some
deviation of the predicted seepage rate is observed. Additionally, when the
phreatic surface passes through the core section, a drop-in the seepage rate was
observed due to the low permeability of the clay core. The seepage rate
increased with ky/kx ratio and the Iranamadu dam seepage field measurements
showed a good agreement when ky/kx ratio is 1. Therefore, the behaviour of the
Iranamadu dam with respect to the seepage flow, shows a near isotropic
hydraulic behaviour. The SLOPE/W results conclude that when the reservoir
water level increases, the stability of the downstream slope reduced beyond the
recommended value but the stability of the upstream slope was found to be
satisfying the safety criteria under both steady and transient seepage conditions.

Keywords: Earth dam
Seepage
Steady state
Transient state

Slope stability

1 Introduction

Dams are constructed to meet various functional requirements such as flood control,
hydroelectric power generation, navigation, recreation, irrigation and as a source of
water. Most of the large embankment dams in the world were built during the middle

© Springer Nature Singapore Pte Ltd. 2020

R. Dissanayake and P. Mendis (Eds.): ICSBE 2018, LNCE 44, pp. 365–377, 2020.
https://doi.org/10.1007/978-981-13-9749-3_33
366 T. Venuja and L. C. Kurukulasuriya

decades of the twentieth century and can be classified into two main categories; earth-
fill and rock-fill dams. These dams represent about 85% of all dams built in the world
(Elshemy et al. 2002). These embankment dams are more common than other type of
dams because of using ordinary construction methods utilizing cheap raw soil materials
and subsurface materials (Abhilasha and Antony Balan 2015). Furthermore, the self-
weight of these earth dams can be effectively used to prevent sliding and overturning.
Failure of an earth dam may occur due to different reasons, such as failure of
upstream or downstream slope under earthquake conditions, seepage through the dam,
piping through dam body or under the foundation, and drawdown (Athani et al. 2015;
Kirra et al. 2015a). One of the main causes of failure is seepage in their body and
foundation. Failure due to seepage has occurred in more than 35% of all earth dam and
failures and structural failures account to about 25% of dam failures (Elshemy et al.
2002). Hence, seepage analysis is an important aspect in the design of an earth dam.
Normally, seepage flow occurs along the path that connects through the pores within
soil particles. This flow most likely is laminar that complies with Darcy’s law. Darcy’s
law originally applied for the saturated domain, but can also be used in the unsaturated
domain (Andreea 2016).
This seepage can give rise to stability problems under high pressure head that
causes reduced shear strength. All earth dams will have some seepage and it is unre-
alistic not to expect this (Andreea 2016). Therefore, seepage must be controlled to
prevent leeching out of embankment or foundation material. The ground conditions and
the geological features of the dam site greatly influence the amount of seepage and its
relevant effects.

1.1 Research Area

The Iranamadu dam is one of the largest earth dams in Sri Lanka. Figure 1 presents the
cross section of the Iranamadu dam comprising of a central puddle clay core, original
shoulder fill, shoulder fills placed in 1951 and 1975 and the embankment foundation.

Fig. 1. Iranamadu dam cross section

In this research, seepage problem of the Iranamadu dam was analysed using Geo-
Studio (SEEP/W) software to estimate the seepage rate under both steady and transient
state conditions and, the influence of anisotropy of hydraulic conductivity of core
 Seepage in Iranamadu Dam and Its Influence on the Stability 367

material on seepage was also examined. In addition, stability of upstream and down-
stream slopes of the Iranamadu embankment was also investigated under both steady
and transient state seepage conditions using SLOPE/W limit equilibrium software.

1.2 Research Objectives

• Establish historical data on the seepage and reservoir water level measurements
over a continuous period of time.
• Compare field seepage data and results from numerical analysis under steady and
transient states using SEEP/W software.
• Study the effect of soil anisotropy on seepage through the dam by numerical
modelling for transient state.
• Investigate the stability of the dam incorporating the effect of seepage.

2 Literature Review

Seepage analysis can be divided into two stages; steady state analysis and transient
analysis. Steady state flow analysis is where the boundary conditions inside and outside
of the ground do not change with time. Therefore, the inflow is always equal to the
outflow within the analytical range. Transient analysis is used when the boundary
conditions inside or outside of the ground changes with time. Transient analysis can
display different inflow and outflow with time, even when the same boundary condi-
tions as those for the steady state analysis are applied. The main differences between
transient and steady state analyses are that the boundary conditions change as time
passes and the fact that the transient analysis requires the definition of the volumetric
water content function (Andreea 2016). In the slope stability analysis, several limit
equilibrium methods have been developed to determine the critical slip surface and the
corresponding minimum factor of safety.
Recently, the Finite Element Method (FEM) and the Limit Equilibrium Method
(LEM) are used for the seepage and stability analyses of embankment dams respec-
tively. Several authors, namely Uromeihy and Barzegari (2007), Kamanbedast and
Delvari (2012), Hasani et al. (2013) and Kirra et al. (2015a, b) had performed seepage
and/or stability analysis of an embankment dam. Abo (2001) used FEM to determine
the free surface position, quantity of seepage, pore water pressure distribution, total
head distribution and the effect of anisotropy of the material in AL-Adheem zoned
earth dam, Iraq.
GeoStudio has been used for seepage and slope stability analyses of earth dams by
many researchers. Kamanbedast and Delvari (2012) used the GeoStudio (SEEP/W)
software to investigate seepage condition and flow of water through the soil in the
Maroon dam, Iran. Hasani et al. (2013) also studied the seepage analysis using
SEEP/W software in the Ilam earth dam, Iran for four mesh sizes in order to assess the
effect of meshing on accuracy of results and slope stability analysis have been carried
out under different conditions and methods using SLOPE/W software. Kirra et al.
368 T. Venuja and L. C. Kurukulasuriya

(2015b) also used GeoStudio software to analyse the seepage and stability of the
Mandali earth fill dam, Iraq.

3 Materials and Methods

Based on the geotechnical investigation details of the Iranamadu dam, soil parameters
were identified for each layer and which are tabulated in Table 1.

Table 1. Soil parameters used in the software analysis (Report of design development of
remedial works Iranamadu embankment 2013)
Layer name Soil type c′ (kPa) /′ (°) Unit weight Permeability
(kN/m3) (m/s)
Weathered rock Silty clayey sand 0 45 21 1  10−7
Residual soil Silty sand 0 35 21 1  10−5
Puddle clay core Sandy clay with 5 20 16 1  10−7
gravels
Original Clayey gravel 0 30 20 1  10−4
shoulder fill
1951 Shoulder Clayey sand 0 30 20 1  10−7
fill
1975 Shoulder Silty sand 0 30 20 1  10−5
fill

3.1 Field Measurements

Based on the Iranamadu dam field records collected from the Department of Irrigation,
Kilinochchi, a seepage location has been identified downstream in the side of the dam
embankment at Ch 0+1544 m and seepage discharges were measured from 26th
December 2014 to 13th June 2015 totalling 170 days under different reservoir levels at a
location which is 41 m below the dam crest. Also, a temporary filter of 6.0 m  3.2 m
has been built at that measuring location and the seepage discharge was collected into a
measuring cylinder through a pipeline from where the rate of seepage was calculated.
Then the upstream reservoir water levels were recorded with the time period. Based on
the field measurements, upstream reservoir water levels were plotted with time and
seepage rates were plotted with different reservoir water levels as shown in Figs. 2 and
3, respectively.
In Fig. 2, upstream reservoir water level has reduced from 30.33 m to 27.71 m
during the period of 170 days.
As expected, seepage rate has decreased with the reservoir water level as shown in
Fig. 3. There is a uniform reduction of the seepage rate in zone I from 30.33 m to
29.0 m of the reservoir water level. After that, a sudden reduction can be seen in zone II
between 29.0 m to 28.75 m of the reservoir water level. In zone III, the seepage rate is
seen reducing uniformly with the reduction of the reservoir water level.
 Seepage in Iranamadu Dam and Its Influence on the Stability 369

Fig. 2. Variation of upstream reservoir water level with time

Fig. 3. Variation of Seepage rate with upstream reservoir water level

3.2 Seepage Analysis of Iranamadu Dam

Seepage analysis was carried out using SEEP/W software assuming isotropic hydraulic
characteristics. The saturated/unsaturated hydraulic model was selected to accommo-
date seepage flow above the phreatic surface under unsaturated conditions. Therefore,
the hydraulic conductivity function and the volumetric water content function were
defined for both analyses. To estimate the hydraulic conductivity function, van Gen-
uchten model was used as the estimation method and sample function option was used
to estimate the volume water content function.
In the steady state analysis, the seepage rate was calculated by considering different
upstream reservoir levels. In transient analysis, a steady state analysis at maximum
370 T. Venuja and L. C. Kurukulasuriya

reservoir water level of 30.33 m was taken as the parent analysis. In this analysis, the
variation of reservoir level with time (Fig. 2) was used to define the input hydraulic
head function. Furthermore, the influence of anisotropy of hydraulic conductivity of
core material on seepage was investigated by varying soil anisotropy ratios (ky/kx) from
1 to 3 and by keeping kx constant for both steady and transient state analyses, where kx
and ky are the coefficients of permeability measured in the horizontal and vertical
direction respectively.

3.3 Stability Analysis of Iranamadu Dam

The stability analysis of the dam under steady and transient seepage conditions was
carried out by assigning shear strength parameters (Table 1) for the selected Mohr-
Coulomb material model and using Bishop’s simplified method in SLOPE/W. The
stability was analysed for both upstream and downstream slopes of the Iranamadu dam
embankment. Then, the results of the analysis were compared with the safety regulation
criteria set by the British Dam Society (BDS 1994), according to which the factor of
safety should be more than 1.2 under rapid drawdown and 1.3 under steady state
seepage conditions.

4 Results and Discussion

Seepage rates through the Iranamadu dam and the determination of the position of the
phreatic surface were computed using SEEP/W under steady and transient states.

4.1 Steady and Transient State Seepage Analysis

In the steady state seepage analysis, the variation in reservoir upstream water level was
assumed as constant with time in each drop of 0.15 m from 30.33 m to 27.71 m.
Figure 4 shows the position of the phreatic surface and flow fields when the reservoir
water level was at 30.33 m.

Fig. 4. Phreatic line and flow fields for the reservoir water level at 30.33 m

As indicated in Fig. 4, the seepage rate at the filter location is computed to be

1.612  10−5 m3/s/m (1.393 m3/day/m) for the reservoir water level at 30.33 m. The
 Seepage in Iranamadu Dam and Its Influence on the Stability 371

seepage rates thus computed for different reservoir water levels are plotted in Fig. 5
along with the field measurements.

Fig. 5. Variation of seepage rates with different upstream reservoir water levels

Based on Fig. 5, it can be seen that SEEP/W software results and field measure-
ments closely match for reservoir water levels varying from 28.8 m to 30.33 m. The
significant drop in seepage rate for reservoir water levels below 29.0 m for both field
and SEEP/W results is due to the change of the phreatic surface position from passing
through the original shoulder fill to the clay core of the dam. Figures 6 and 7 show the
change of phreatic surface for reservoir water level at 29.0 m and 28.8 m, respectively.

Fig. 6. Phreatic line for reservoir water level at 29.0 m

When the reservoir water level further decreases below 28.8 m, the software results
showed a greater difference with the seepage rate calculated from field measurements.
It may be due to the presence of non-homogeneous soil in clay core section. However,
in the analysis, the clay core was taken as a low permeability homogenous section.
Therefore, when the phreatic surface reaches the clay core section, there is no signif-
icant change in the seepage rate when reservoir water level reduces other than that due
to changes in the water head.
372 T. Venuja and L. C. Kurukulasuriya

Fig. 7. Phreatic line for reservoir water level at 28.8 m

Seepage rate corresponding with the time history of reservoir level variation was
calculated using transient analysis in order to capture the real time behaviour of
seepage flow. The variation of seepage rates obtained from the above analysis and field
measurements along with the variation of reservoir water level with time are shown in
Fig. 8.

Fig. 8. Variation of seepage rate and reservoir water level with time

Based on Fig. 8, the seepage rates and reservoir water level reduce with time. The
variation of seepage rates in field and SEEP/W software results are almost same up to
115 days when the reservoir water level reached 29.1 m. As the water level further
decreased, the software results showed a significant variation from field measurements
due to the presence of non-homogeneous clay core. After 115 days, the phreatic surface
passes through the puddle clay core. Hence, there is a significant difference in field and
SEEP/W results. Figure 9 shows a comparison of phreatic surfaces in 115 and 144
days. Further, it was noticed that the upstream water levels are at 29.1 m and 28.8 m
after 115 and 144 days, respectively. Therefore, both steady state and transient seepage
analysis show almost similar variation due to the change in phreatic surface.
 Seepage in Iranamadu Dam and Its Influence on the Stability 373

Fig. 9. Comparison of phreatic surface at reservoir level of 29.1 and 28.8 m after 115 and 144
days respectively

Figure 10 shows a comparison of observed seepage rate under steady and transient
state analyses with field measurements at different reservoir water levels.

Fig. 10. Comparison of observed seepage rate with results of steady and transient state analyses

In Fig. 10, the transient and steady state analyses match with field data for the
phases from 1 to 2 and 2 to 3. For the phase between 3 and 4, a deviation of the seepage
rate from each line is observed. The deviation of the transient analysis is greater than
that in the steady state analysis, especially at lower reservoir elevations, indicating the
over prediction of seepage rate in the transient analysis due to the effect of drawdown
phase of reservoir water level. When reservoir water level draws down suddenly, the
pore water pressure will reduce and more pores will drain out. Therefore, a greater
seepage rate is observed in transient analysis.

4.2 Effect of Anisotropy of Hydraulic Conductivity on Seepage

The effect of anisotropy of hydraulic conductivity on seepage was examined under
steady and transient states using SEEP/W software. Hydraulic conductivity ratio, ky/
kxwas changed by keeping kx constant and seepage rates were calculated with time
374 T. Venuja and L. C. Kurukulasuriya

under ky/kx ratios of 1, 2 and 3. Figure 11 shows the variation of seepage rate with time
for different ky/kx ratio.

Fig. 11. Variation of seepage rate in transient state with time for different ky/kx ratio

As shown in Fig. 11, the results indicate that the seepage rate is increased with the
increase of ky/kx ratio. Except for time period between 115 to 150 days, the field
seepage measurements show good agreement when ky/kx ratio is 1. Therefore, the
behaviour of the Iranamadu dam with respect to the seepage flow rate shows a near
isotropic behaviour.
In the steady state seepage analysis, four different water levels were selected and
seepage rates were calculated for different ky/kx ratio. The calculated seepage rates are
given in Table 2.

Table 2. Seepage rates under steady state for different reservoir water levels
Reservoir Seepage rate  10−5 m3/s/m
water level (m) ky/kx = 1 ky/kx = 2 ky/kx = 3 ky/kx = 5
30.33 1.610 2.204 2.583 2.931
29.10 1.230 1.564 2.067 2.436
28.75 0.164 0.214 0.260 0.325
27.70 0.135 0.194 0.237 0.282

Based on the results of the SEEP/W analysis, the rate of seepage increases when
increasing the upstream water level and ky/kx ratio. This agrees with Seepage mod-
elling with SEEP/W (2012), which stated that the seepage rate increases with aniso-
tropy ratio (ky/kx).
 Seepage in Iranamadu Dam and Its Influence on the Stability 375

4.3 Slope Stability Analysis Using SLOPE/W Software

The slope stability analysis was carried out corresponding to the transient seepage
analysis. The variation of minimum factor of safety corresponding with the reservoir
water level with time for downstream (D/S) and upstream (U/S) slopes are shown in
Figs. 12 and 13, respectively.

Fig. 12. Variation of downstream minimum factor of safety and reservoir water level with time
in transient state analysis

Fig. 13. Variation of upstream minimum factor of safety and reservoir water level with time in
transient state analysis
376 T. Venuja and L. C. Kurukulasuriya

Based on Fig. 12, the Iranamadu dam downstream slope does not satisfy the
minimum FOS recommended up to 144 days. Thus, it infers that when the reservoir
water level increases above 28.85 m, there is a possibility of a slip in the downstream
slope. This is due to the passage of the phreatic surface through the critical slip surface.
As per the Fig. 13, the minimum factor of safety values for upstream slope varies
from 1.46 to 1.32 when the reservoir water level is reduced, due to the removal of the
balancing hydraulic forces acting on the upstream side embankment and the near
undrained condition prevailing within the clay core. Also, after 140 days, when the
reservoir water level reduces significantly, the upstream minimum FOS increases due
to the dissipation of pore water pressure from the saturated zone within the embank-
ment. Souliyavong et al. (2012) also concluded the same results. As a result, the
Iranamdu dam upstream slope satisfies the minimum FOS of 1.3 under transient state
seepage condition as mentioned in the safety criteria of British Dam Safety (1994).
Furthermore, the slope stability analysis under steady state condition also gives
similar variation of results to that of transient state analysis. The results of stability
analyses compare well with the observations of the Iranamadu dam slopes, where there
was no upstream slope failure recorded in the Iranamadu dam while a downstream
slope failure had occurred in 1983.

5 Conclusions

Based on the results obtained from the seepage and stability analyses, the following
conclusions can be made.
• In both steady and transient state seepage condition, SEEP/W software results show
a good agreement in seepage rates with those obtained from field measurements.
• When the phreatic surface is intercepted by the puddle core section, the sudden
drop-in seepage rate is observed due to the low permeability of puddle core section.
• The seepage rate increases with increase of ky/kx ratio by keeping kxconstant and
the Iranamadu dam seepage field measurements show a good agreement when ky/kx
ratio is 1. Therefore, the behaviour of the Iranamadu dam shows a near isotropic
hydraulic behaviour.
• When the reservoir water level increases, the minimum FOS of the downstream
slope reduces which could lead to failure of the Iranamadu dam and the possibility
of upstream slope failure of the Iranamadu dam for different reservoir water levels is
highly unlikely.

Acknowledgements. The Engineers and Technical Officers of the Department of Irrigation,

Kilinochchi are gratefully acknowledged for the permission granted to use the data related to the
Iranamadu dam.
 Seepage in Iranamadu Dam and Its Influence on the Stability 377

References
Abhilasha PS, Antony Balan TG (2015) Numerical analysis of seepage in Embankment dams.
J Mech Civil Eng 13–23. https://pdfs.semanticscholar.org/2ef9/5e1b61aa0133de1f5b6c05d7f
d767fbc2694.pdf
Abo AA (2001) Calculating seepage through earth dams by finite element method. College of
Engineering, University of Baghdad, Iraq
Andreea C (2016) Unsaturated slope stability and seepage analysis of a dam. Energy Procedia
85:93–98
Athani SS, Shivamath, Solanki CH, Dodagoudar GR (2015) Seepage and stability analysis of
earth dam using finite element method. Aquat Procedia 4:876–883
BDS (1994) The British dam society at the institution of civil engineers, Great George Street,
London
Elshemy M, Nasr RI, Bahloul MM, Rashwan IM (2002) The effect of blockages through earth
dams on the seepage characteristics. Faculty of Engineering, Tanta University, Egypt
Hasani H, Mamizadeh J, Karimi H (2013) Stability of slope and seepage analysis in earth fills
dams using numerical models (case study: Ilam dam - Iran). World Appl. Sci. J. 21(9):1398–
1402
Kamanbedast A, Delvari A (2012) Analysis of earth dam, seepage and stability using ANSYS
and GeoStudio software. World Appl. Sci. J. 17(9):1087–1094
Kirra MS, Shahien M, Elshelmy M, Zeidan BA (2015a) Seepage and slope stability analysis of
Mandali earth dam, Iraq: case study. In: International conference on advance in structural and
geotechnical engineering
Kirra MS, Zeidan BA, Shahien M, Elshelmy M (2015b) Seepage analysis of Walter F. Geoge
Dam, USA: a case study. In: International conference on advance in structural and
geotechnical engineering, pp 1–13
Report of design development of remedial works Iranamadu tank embankment (2013)
Seepage modelling with SEEP/W (2012). GEO-SLOPE International Ltd
Souliyavong T, Gallage C, Egodawatta P, Maher B (2012) Factors affecting the stability analysis
of earth dam slopes subjected to reservoir drawdown. In: Second international conference on
geotechnique, construction materials and environment
Uromeihy A, Barzegari G (2007) Evaluation and treatment of seepage problems at Capar-Abad
Dam, Iran. Eng Geol 91:219–228
 Application of Principle of Network
Connectivity in Creation of Sustainable Urban
Form: Case of Negombo Water Fronts

D. S. N. Jayakody(&)

Urban Development Authority, Gampaha District Office,

Minuwangoda, Sri Lanka
[email protected]

Abstract. City is a vital organism that entitled with the diverse planning
attempts. Blue green infrastructure planning is the one of attempt which has
been practised recently. Mainly the Blue Infrastructure Planning involves with
the four major core principles of Green-Grey Integration, Network connectivity,
multifunctionality and social inclusion. The principle of network connectivity
has been considered in the research and it involves with the both structural and
functional connections between blue green spaces in order to create added
values from an interlinked system. Numerous quantative approaches have been
developed to measure and map the connectivity. Measures of structural con-
nectivity are based on the concepts of presence or absence or the size, form and
shape of corridors. Initially, connectivity was calculated using the various
indices of distance, frequency and density. The research finding has been
adopted in Negombo Development Plan 2018–2030 which is formulated by
Urban Development Authority. It adheres with the strategic planning approach
with the use of diverse methods of multy criteria analysis techniques, field
surveys, observation and key formal interviews to integrate the blue green
spaces of the city with the purpose of enhancing the sectoral tourism, recreation
and self-sufficient affluent city economy connecting the all blue spaces of urban
waterfronts. The research attempts to find how the network connectivity influ-
ence to reinforce a sustainable urban form in the city and activate the networks
of green and blue spaces with full spectrum of benefits to denizens and the
journeyrs at all spatial scales that enable to open a positive platform to merge
with the emerging urban challenges of the city and ultimately it influences to
show the creation of sustainable urban form in Negombo with the integration of
blue-green networks and considerable positive outcomes.

Keywords: Blue-green infrastructure
Network connectivity

Self sufficient city
Sustainable urban form

1 Introduction

Blue green infrastructure planning associates with the major core principles of green
grey integration, network connectivity, multifunctionality and social inclusion that
contribute to create a green space networks and multiple function and services. The

© Springer Nature Singapore Pte Ltd. 2020

R. Dissanayake and P. Mendis (Eds.): ICSBE 2018, LNCE 44, pp. 378–393, 2020.
https://doi.org/10.1007/978-981-13-9749-3_34
 Application of Principle of Network Connectivity 379

principle of network connectivity is an innovative approach in the creation of sus-

tainable city form to face the major urban challenges and mitigate the risks of the cities.
more than half the world’s population live in cities are rising rapidly (UNEP 2011).
The major factors of Increase of population density, development pressure lead to
densification of cities. Many cities are situated in the coastal or delta areas where
disaster risk increases. This is firstly a result of climate change, including rising sea
levels, increased river discharge and extreme precipitation. Secondly, this results from
land use changes such as the increase of sealed surfaces and buildings and from
subsidence.
Frequent flooding and the consequent damage are a threat to a healthy and vital
urban environment. Furthermore, heat waves become more frequent because of climate
change. This may lead to health problems in cities becoming urban heat islands.
Well planned and designed cities adhere with the green infrastructure, including
water, greenery and soil. This can contribute to climate change adaptation and promote
and support healthy urban living. Effective application requires design principles that
address the direct relationship between green infrastructure and impacts of climate
change is import but also the quality of the ecosystem (Kemperman and Timmermans
2014).
Three categories of health aspects of water and ecosystem services support human
health as directing exposure to water contributing to medical health, encouraging healthy
living by creating possibilities to exercise and enhancing the aesthetical aspects of water
contributing to mental health. Further the principle affects to generate the economic
benefits for the city that can be derived from the aquatic ecosystem services (Fig. 1).

Fig. 1. Design principle of blue green infrastructure on city level

The European Commission has adopted a Green Infrastructure Strategy (COM

2013) and states that green infrastructure (GI) can make a significant contribution to
380 D. S. N. Jayakody

effective implementation of all policies where some or all of the desired objectives can
be achieved in whole or in part through nature-based solutions (Merk 2012).
There is usually a high return on GI investments and overall reviews of restoration
projects typically show cost-benefit ratios in the range of 3 to 75. Well planned and
designed green infrastructure, including water and soil, can contribute to climate
change adaptation and at the same time promote and support healthy urban living
(Worpole 2007).

1.1 Objectives of the Research

The key objectives of this research open to assess how the functional use of urban blue
and green infrastructure both at the street and city level can contribute to a liveable,
healthy and sustainable city. Liveability can be improved with a variety of ecosystems
services with the creation of connectivity and integration of blue green infrastructure
and develop innovative combinations of existing solutions with added value, using
state of the art expertise, tools and models to climate change mitigation, climate change
adaptation and resource efficiency and giving rise to recognition of the central role that
green space networks.
Successful green-blue infrastructure planning naturally requires an integration of
different scales of planning and application. Accordingly, green-blue infrastructure
planning will likely be for a whole urban area or catchment, but will need to consider
possible interventions at all scales of local, neighbourhood and city scales their
cumulative impact (Fig. 2).

Fig. 2. Interlinked green blue infrastructure systems across scale

1.2 Study Area

City of Negombo was selected as the study area. It is situated in Negombo Municipal
Council Area in Gampaha District, Western Province and the west coast of Sri Lanka.
It is located 35 km away from Colombo, the commercial capital of Sri Lanka and 7 km
away from the Katunayake Bandaranayake International Airport and free trade zone. It
is bounded, North by Maha Oya, South by Diya honda Ela, East from Roma Ela and
the west from lagoon and Indian Ocean.
 Application of Principle of Network Connectivity 381

The total extent of Negombo Municipal Council area is 2800 ha and it is within the
boundary of Negombo Divisional Secretariat Division with 39 Grama Niladari Divi-
sions. The total population of Negombo is 175000. Negombo features a tropical
rainforest climate under the Koppen climate classification. Mainly the city receives
rainfall from the South-western monsoons from May to August and October to January.
During the remaining months there is a little precipitation due to Convective rains. The
average annual precipitation is about 2400 mm. The average temperature varies 24 to
30 °C, and there are high humidity levels from February to April (Fig. 3).

Fig. 3. Location of study area

Negombo currently act as the one of major commercial, transport center, an edu-
cational Facilitator and health service provider for the Region that is function with both
rail and road network together in Gampaha District.
Negombo has an identifiable blue green network that blend with the Lagoon, sea,
Hamilton Canal, Ma oya river, 24 km long golden sandy beach and 18 number of
mangrove species. This contribute to strengthen the self-sufficient economy of the city
that is created through three majors of tourism, fishery and commerce. Simply 45% of
economy is based on fishery, 35% of economy is based on tourism and 25% of
economy is based on commerce and other service based on the blue green network.
This affect to Attract 350000 tourists annually and they come to Negombo for
leisure activities and this lead to become the major tourist destination. Negombo has an
active fishery industry which has a plenty of fish production. It gets the third place in
Sri Lanka and Contributes 16% for the national economy by fish production. Which is
80% of marine fish and 20% of lagoon fish. Shrimp and prawn’s production highly
contribute for its economy (Fig. 4).
382 D. S. N. Jayakody

Ma Oya

Hamilton Canal
Sea

Mangroves

Lagoon

Fig. 4. Blue green network of Negombo

2 Methods of the Research

The research involves with the Multi criteria analysis of development pressure, envi-
ronmental, sensitivity, liveability, potential and density and GIS mapping, field survey
and key interviews. 39 number of Grama Niladari Divisions of Negombo MC area
were selected and surveyed for the research.

3 Analysis and Synthesis

The value of green spaces in Negombo Municipality area were rediscovered with the
use of multi criteria analysis of development pressure, sight scene hotspot, sensitivity,
liveability, potential and density analysis. The development pressure analysis proved
that the development is spread towards the coastal belt and sensitive blue green spaces
with urban trend phenomena (Figs. 5 and 6).
 Application of Principle of Network Connectivity 383

Fig. 5. Criteria of development pressure analysis

The sensitivity analysis enhanced that the existing blue green spaces of Negombo
lagoon, Ma oya, Hamilton Canal and mangroves species are the most sensitive ele-
ments in the city and the it is the most important blue green infrastructure for the entire
eco system of Negombo and the inhabitants who are dependent on this infrastructure
(Fig. 7).
Thee green corridors were provided to access and integrate the beach parks as a
single promenade. The proposed public outdoor recreation space plan of Negombo has
created a positive linkage among the Ma oya, lagoon and beach strips with Liner parks,
mangrove parks, and active pedestrian friendly cycling and walking paths.
The network connectivity integrated with the proposed Water transport along the
Hamilton Canal to create an active blue space up to Ma oya environment. The proposed
strategic action opened a platform for the conservation of blue green spaces and
activation of the spaces in to the healthy and economically viable context of Negombo
with promoting the tourism and fishery activities (Auttred 2015) (Fig. 8).
384 D. S. N. Jayakody

Fig. 6. Development pressure analysis

Fig. 7. Criteria of environmental sensitivity analysis

 Application of Principle of Network Connectivity 385

Ma oya

Hamilton
Canal

Mangroves

Lagoon

Fig. 8. Environmental sensitivity analysis

Further the field survey and key interview elaborated that the main functions,
economy and other services of Negombo are oriented with the existing green spaces
which provide the aesthetic value and recreational facilities. There is a growing
awareness that green infrastructure can be more beneficial to society than merely
serving aesthetics and recreation. For example, the municipality of Negombo is
interested in making better use of green infrastructure in the search of measures that can
help to achieve a healthy city.
Town planners want to take the capacity of green elements for regulating tem-
perature, air quality, water storage and drainage and noise reduction into account. In
order to be able to do so, there is a need for translation of scientific knowledge on
functionality of green infrastructure into practical design principles. It also requires
practical guidelines for how to integrate these principles into the design of multi-
functional green infrastructure with the principle of blue green network connectivity.
The principle was adapted with the use of sight scene hotspot analysis of the south
part of beach promenade of Negombo. The five major hotspots were identified and the
hotspots were converted on to central, mini and minor beach parks (Fig. 9).
386 D. S. N. Jayakody

Fig. 9. Criteria of sight scene hotspots analysis

In order to optimize the south part of the coastal belt and create an active beach
promenade with creating the hierarchy of networks among the beach parks. The pro-
vision of the typical amenities of Open Spaces, Hotels, playing areas, strolling, cycling,
inline-skating, beer gardens and restaurants, open-air-cinema in summer, sunbathing,
and occasional public events, markets, exhibitions, concerts, a meeting place, Chatting,
restrooms, picnic shelters, Seating on benches under plane trees, clean and well-
maintained, safe, no vehicles in sight were activated to create an active open space for
the users and journers of Negombo (Fig. 10).

Fig. 10. Sight scene hotspots analysis

 Application of Principle of Network Connectivity 387

Thee green corridors were provided to access and integrate the beach parks as a
single promenade. The proposed public outdoor recreation space plan of Negombo has
created a positive linkage among the Ma oya, lagoon and beach strips with Liner parks,
mangrove parks, and active pedestrian friendly cycling and walking paths.
The network connectivity integrated with the proposed Water transport along the
Hamilton Canal to create an active blue space up to Ma oya environment. The proposed
strategic action opened a platform for the conservation of blue green spaces and
activation of the spaces in to the healthy and economically viable context of Negombo
with promoting the tourism and fishery activities (Fig. 11).

Fig. 11. Proposed PORS plan

4 Findings and Results

The research emphasises the key findings. This attempts to optimize the blue green
infrastructure. The identified gap of blue green infrastructure and city scape were
integrated with the principle of network connectivity and it affects to interact between
two major components of Blue green infrastructure and the city fabric.
Further the application of principle of network connectivity influences to create a
vibrant active, accessible, comfort sociability and a reputed image for the city (Fig. 12).
388 D. S. N. Jayakody

Fig. 12. Active beach promenade

The principle of network connectivity further influenced to upgrade the liveability

and potential of Negombo and the direct and indirect impacts for city economy were
interlinked and the tourist arrivals will be increased up to 300000 with the tourist
prediction scenarios (Baudry and Merriam 1998) (Figs. 13 and 14).

Fig. 13. Criteria of liveability analysis

 Application of Principle of Network Connectivity 389

Fig. 14. Livability analysis of Negombo

The entire area is emerged as more potential in terms of hotels, tourism, fishery,
residential, commercial and other uses and many projects were initiated to accom-
modate the area (Figs. 15 and 16).

Fig. 15. Criteria of potential analysis

390 D. S. N. Jayakody

Fig. 16. Potential analysis of Negombo

Land value of the area was risen with the integration of blue green infrastructure
and it became the major influence on city economy and the coastal areas will be
converted as the potential tourism areas and active open spaces and the significant hotel
development, demand and competition for the space will be originated on the ground of
the application of blue green network connectivity (Worpole 2007) (Fig. 17).
Further the application of principle of Connectivity encourages the structural and
functional upgrading between green spaces to create an added value from inter linked
systems and a sustainable, healthy form of the city with self-sufficient economy (Taylor
2006).
 Application of Principle of Network Connectivity 391

Fig. 17. Land value of Negombo

5 Conclusion and Recommendations

This research interprets the how far the application of the design principle of blue green
network connectivity is worth for the sustainable city form with addressing the urban
challenges of climate resilience, healthy and liveable urban environment with sus-
tainable self-sufficient city economy. A healthy and liveable urban environment con-
tributes to the strengthening of the socio-economic climate in cities and improve a
variety of ecosystems services.
Research opens a platform for the future researchers to apply the modern design
principals for the cities.
The urban environment in which people live, work and play affect for their health
with the characteristics of a healthy a clean and safe environment, sufficient green
space, nature and water, healthy and sustainable homes, attractive and varied public
spaces, wide range of public services.
The strategic approach followed in the research interlinks to harness the potential of
blue green spaces to make a major contribution to sustainable development by
enhancing social cohesion, supporting the economy, and adapting to a changing cli-
mate, and highlights the importance of green infrastructure solutions in cities. The
design principle can adhere to the urban challenges of protecting bio diversity, pro-
moting green economy, increasing the social cohesion and minimising the disparities
(Fig. 18).
392 D. S. N. Jayakody

Fig. 18. Application of principle of blue green network connectivity in Negombo

Meanwhile the physical arrangement of the city is converted with the use of blue
green infrastructure and the colours, building designs and blue green oriented regula-
tions, conservation buffers, and green blue corridors will be adapted in the city and
green walkable street areas, user friendly open spaces are activated with the multiple
benefits for the users and journeyrs of the city.

Acknowledgements. Author wishes to thank the Plnr. Janak Ranaweera, Director of Western
Province of Urban Development Authority, Plnr J. Somasiri, Deputy Director of the Gampaha
District Office and the staff and the Plnr Samantha Kumara, Deputy Director of Matale District
Office and A.M.C. Samanthilaka, U. D. Hettiarachchi, Town planners of Kelaniya and Biyagama
Development Plans in Gampaha District office guidance, supervision and sponsorship to become
the success of the research study. Sumithra Wijesinghe, City Planner of Negombo MC sup-
porting for the field surveys and the process of data collection and Plnr, Nileesha Senavirathna
encouraging to launch the research and those who support to carry out the research.

References
UNEP (2011) Green Jobs towards a green economy, pathways to sustain able development and
poverty eradication. Kenya
UNEP (2013) Measuring progress towards an inclusive green economy. Kenya
Merk O (2012) Financing green urban infrastructure. Paris
Dunn AD (2010) Sitting Green intra legal and policy solution to alleviate urban poverty and
promote a healthy communities. Boston College
Kemperman A, Timmermans H (2014) Green spaces in the living environment and social
contacts of the aging population, Landscape and Urban Planning
Worpole K (2007) The social value of public spaces. Joseph Rowntree Foundation, York
 Application of Principle of Network Connectivity 393

Taylor PD (2006) Landscape connectivity. Cambridge

Baudry J, Merriam G (1998) Connectivity and connectedness, functional versus structural
patterns in landscapes connectivity
Auttred AG (2015) The spatial and temporal components of functional connectivity in
fragmented landscapes
Department of Environment, Land, Water and Planning (2017) Planning a Green-Blue City,
Australia
Gehrels H, van der Meulen S, Schasfoort F, Bosch P, Brolsma R, van Dinther D, Geerling G,
Goossens M, Jacobs C, Kok S, Massop HTL (2016) Designing green and blue infrastructure
to support healthy urban living. TO federate
 Development of a Post-tentioned Masonry
Retaining Wall System

W. W. P. K. Perera(&)

Department of Civil Engineering, National Engineering Research

and Development Centre of Sri Lanka,
No: 2P/17B, Industrial Estate, Ekala, Ja-Ela, Sri Lanka
[email protected]

Abstract. A Retaining Wall is a commonly used structural element in the field

of construction. A retaining wall is a structure constructed to hold back any
material (mainly soil) and prevent it from failure. A retaining wall is capable of
resisting the material pressure of the material that it is holding back. Random
Rubble Masonry (RRM), Reinforced Cement Concrete (RCC) and Gabion Wall
(GW) are the most common types of construction materials used in the con-
struction of retaining walls. RRM & GW are gravity type retaining walls and
they require significant width from the land. RCC retaining walls require large
amount of tor steel reinforcement, concrete and vertical form work, which are
costly. Masonry walls such as brick walls and block walls cannot be used as
retaining walls as they cannot withstand the tensile stresses developed due to
lateral earth pressure. This paper discusses the development of a masonry block
wall system which can be used as a retaining wall by post-tensioning eccen-
trically. The compressive stresses developed due to post-tensioning is used to
neutralize the tensile stresses developed due to lateral earth pressure. This
system eliminates concrete and vertical formwork and minimizes steel usage. It
has advantages such as cost effectiveness, pleasing aesthetic appearance and
simplicity.

Keywords: Retaining walls
Post-tensioned masonry

Eccentrically pre-stressing
Embankment
Lateral pressure of soil

1 Introduction

A retaining wall is a structure designed and constructed to resist the lateral soil pressure
and hydrostatic pressure if any of a slope that exceeds the angle of repose of the soil.
Angle of repose is the maximum slope a soil type can withstand without any external
support. Movement of unstable soil masses can cause lot of inconveniences resulting in
loss of property and life. The soil movements are known as slips or landslides
depending on their magnitude (Fig. 1).

© Springer Nature Singapore Pte Ltd. 2020

R. Dissanayake and P. Mendis (Eds.): ICSBE 2018, LNCE 44, pp. 394–410, 2020.
https://doi.org/10.1007/978-981-13-9749-3_35
 Development of a Post-tentioned Masonry Retaining Wall System 395

Fig. 1. Failure of retaining wall

Traditional retaining walls can be classified as gravity walls, piling walls, cantilever
walls and anchored walls based on their behaviour. Random Rubble Masonry (RRM),
Reinforced Concrete (RC) and Gabion walls (GW) are the most common material types
used to construct retaining walls. RC retaining walls are costly as it requires high
quality vertical formwork. Since reinforced concrete retaining wall are in-situ con-
struction, lack of supervision and lack of skilled supervisors results in construction of a
different structure than what is designed. RRM and GW require more space from the
land. This is critical in smaller plots of land with higher land value.

2 Literature Review on Post-tensioned Masonry

Masonry is the first man made construction material in the world invented long before
other construction materials such as steel and concrete. Masonry walls have advantages
such as availability, simplicity, durability, strength and ability to incorporate archi-
tectural elements in to its design. But the application of masonry is limited only to
compression members as masonry is weak in resisting tensile stresses. However, it has
sufficiently large compressive strength. Due to this fact, traditional masonry cannot be
used to construct retaining walls as it cannot withstand the tensile stresses developed
due to lateral earth pressure. Thus a traditional brick work structure subjected to
bending will fail in tension long before its compressive strength is exploited [1]. If
these tensile stresses can be neutralized by applying a pre-compression to the masonry
elements, there is a possibility to use masonry elements in retaining walls. Post-
tensioning technique can be applied to brick masonry to achieve this target and it can
also be used to get the maximum benefits from this versatile construction material. It
396 W. W. P. K. Perera

offers new application for masonry with simple construction technique and competitive
cost. Post tensioning will increase cracking resistance, and ductility. Post tensioning
will also provide effective connection between wall and binding material to avoid the
failure of the joints [2].
Advantages of post-tensioning
• Enhances the flexural tensile resistance of the brickwork.
• Highly cost effective and durable
• Simple technique.
• Aesthetically pleasing.
• Convert the brickwork from a brittle material to a ductile material. (It will bend
without cracking/ deflect and recover when the load is removed)
• Increases the robustness, resistance to impact and high improvement in accidental
damage resistance.
Disadvantages
• Post-tensioning of high-shrinkage units shall be avoided due to possibility of
excessive losses of pre-stress.
• Difficulty in finding anchoring methods.
• High accuracy has to be maintained in positioning and tensioning pre-stressing
wires.

3 Research Methodology

The methodology described below is adopted in the development of Post-Tensioned

Masonry (PTM) retaining wall system.
• Observe the failure modes of retaining walls and loads on a typical retaining wall
section.
• Analysis and design of post-tensioned retaining wall system by first principles
considering retained height, effective prestressed force and eccentricity as variables.
• Manufacturing of cement: quarry dust blocks using the hydraulically operated
cement: soil block making machine
• Construction of pilot project to validate the theoretical results.

3.1 Failure Modes of Retaining Wall and Loads on a Typical Retaining

Wall
3.1.1 Modes of Failure of a Retaining Wall
Failure modes of retaining wall are illustrated in Fig. 2.

3.1.2 Loads on a Typical Retaining Wall

Loads on a typical retaining wall is illustrated in Fig. 3 and mainly consist of active soil
pressure, hydrostatic pressure, passive soil pressure, surcharge, weight of concrete
sections, weights of soil mass on the footing and upward soil pressure.
Development of a Post-tentioned Masonry Retaining Wall System 397

Fig. 2. Modes of failure of a retaining wall

Surcharge

Ww
Ws

Wb

Passive Hydrostatic Active

soil pressure soil
pressure pressure

Bearing pressure

Fig. 3. Loads on a typical retaining wall

398 W. W. P. K. Perera

3.2 Analysis of Post-tensioned Retaining Wall Section

For the analysis of the retaining wall system with first principles, the section of the
retaining wall shown in Fig. 4 is selected.

150

G.L.
150x150 RC Tie beam

Continuous duct

2150 2 Nos. of 6mm H.T.S.Wires

300x150x100mm 1:5 cement

quarry dust masonry block

Anchor cone & pin

250 200

G.L.
100

T12-200 c/c
200mm wide R.C.C. wall
T10-200 c/c

T10 -200 c/c

900 T10 -250 c/c
T10 -250 c/c
T10 -200 c/c
Screed concrete

200

50
200
1050

Fig. 4. Section of the retaining wall used to analysis

A retaining wall needs to be checked for sliding, overturning, bearing and bending.
The following parameters were selected as variables for the design of Post-Tensioned
Masonry (PTM) retaining wall.

Density of insitu concrete: 24 kN/m3

Grade of insitu concrete: 25 N/mm2
Density of soil: 18 kN/m3
Angle of internal friction: 30
Allowable bearing pressure: 125 kN/m2
 Development of a Post-tentioned Masonry Retaining Wall System 399

Active earth pressure coefficient: 0.33

Passive earth pressure coefficient: 3.0
Surcharge: Nill
HTS wire diameter: 5 mm
Jacking force: 24 kN
Wire spacing: 150 mm
Total losses of prestress: 30%
Design bending moment: 23.83 kN/m

3.3 Construction of Pilot Project

3.3.1 Selection of Site for the Pilot Project
A site at Dalupitiya Clinical Building premises belonging to the Mahara Pradeshiya
Sabha was selected to construct the pilot project of the research (Fig. 5). The necessary
approval was obtained from the Mahara Pradeshiya Sabha. The height of the soil
embankment to be retained at the site was 2400 mm. The expected length of the
retaining wall is 6.7 m. The section of the PTM retaining wall illustrated in Fig. 4 is
selected.

Fig. 5. Site selected for the pilot project

3.3.2 Selection of Masonry Block Type for Casting

The available bricks and blocks are first considered for construction of PTM retaining
wall. However it was found that the characteristic direct compressive strength of these
bricks and blocks are in the range of 1–2 N/mm2 and are not suitable for PTM retaining
wall. A potential axial compressive strength could be obtained from the cement: quarry
dust mixed interlocking blocks manufactured using the NERDC developed hydraulic
block making machine. A series of sample tests were carried out to find the axial
compressive strength (Fig. 6) of masonry blocks produced. The results indicate an axial
compressive strength in the range of 7–8 N/mm2 is possible with a mix proportion of
1:5 cement: quarry dust.
400 W. W. P. K. Perera

Fig. 6. Direct compression test

3.3.3 Testing of Flexural Tensile Strength of Masonry Blocks

According to the literature the characteristic flexural tensile strength of masonry is
assumed as zero. Since a masonry block with high compressive strength is used in this
research, a test is carried out to find out the actual characteristic flexural tensile strength of
masonry blocks produced. The test results show an average characteristic flexural tensile
strength of 2.73 N/mm2 and this value is used in analysis under this research (Fig. 7).

Fig. 7. Flexural tensile strength test

The following values base on the test results are also used for the calculation.
Density of block: 17 kN/m3
Compressive strength of blocks, fk: 7.9 N/mm2
 Development of a Post-tentioned Masonry Retaining Wall System 401

Characteristic flexural compressive strength is equal to 1.5 times the characteristic

direct compressive strength [2].
Material safety factor: 1.2
Design flexural com. strength, ff: 9.87 N/mm2
Flexural tensile strength, fk: 2.73 N/mm2
Self weight of the block work: 0.03 N/mm2

3.3.4 Results of Analysis of Stresses on the Post-tensioned Masonry

Retaining Wall
Results of Analysis of different type of tresses acting on the masonry retaining wall is
illustrated in Fig. 8.

50 25
5
5

0.03N/mm2
self weight

0.74 N/mm2

Pe/A
-0.59 N/mm2
0.59 N/mm2
Pe.e
Z

-1.81 N/mm2

1.81 N/mm2

M/Z

-0.522 N/mm2

2.06 N/mm2

Resultant stress

Fig. 8. Stresses on the post-tensioned masonry retaining wall

402 W. W. P. K. Perera

3.3.5 Casting of Masonry Blocks

Blocks are manufactured using the hydraulically operated block making machine with
the cement: quarry dust ratio of 1:5 (Fig. 9). The machine is originally designed and
manufactured for cement: soil block making. Since the cement: quarry dust mix is less
compressible compared to cement: soil mix, the volume of mix used to manufacture the
blocks has to be carefully selected.

Fig. 9. Casting of masonry blocks

3.3.6 Curing of Masonry Blocks

Blocks are kept under a polythene cover and spray water frequently for 14 days for
curing. Blocks are stored another 14 day to allow for shrinkage (Fig. 10).

Fig. 10. Curing of masonry blocks

 Development of a Post-tentioned Masonry Retaining Wall System 403

3.3.7 Preparation of Site for Construction

The existing embankment of approximate width of 1.0 m to 1.2 m is excavated up the
boundary using an excavator. Then the foundation trench was excavated (Fig. 11) as
per the structural detail drawing.

Fig. 11. Excavation of embankment and foundation trench

3.3.8 Construction of Foundation

The insitu concrete foundation base and wall up to the ground level was constructed
with necessary reinforcement arrangements placed. Vertical and horizontal ducts which
are required to insert High Tensile Steel wires are kept in the exact required positions.
PVC pipes and electrical conduit pipes are used to keep ducts in insitu concrete
(Fig. 12). Completed RCC foundation and RCC wall is shown in Fig. 13.

Fig. 12. Positioning of vertical & horizontal ducts

404 W. W. P. K. Perera

Fig. 13. Completed RCC foundation and wall up to ground level

3.3.9 Construction of Masonry Wall

Interlocking masonry block wall is constructed using normal construction procedure as
shown in Fig. 14. Cement slurry is used to paste the interlocking blocks. A reinforced
concrete tie beam is constructed over the masonry block wall (Fig. 15). Block wall is
allowed a period of curing of 14 days.

Fig. 14. Construction of masonry wall

 Development of a Post-tentioned Masonry Retaining Wall System 405

Fig. 15. Casting of tie beam

3.3.10 Inserting the H.T.S. Wires

High Tensile Steel (HTS) wires cut to the exact measurement are inserted through the
duct from the top and fixed against the RCC wall using the anchor cone and the pin
fixed permanently at the bottom.

3.3.11 Grouting the Ducts

The ducts into which the H.T.S. wires are inserted are filled with cement grout to avoid
corrosion. Grouting is done before the anchor cone and pin is placed to the jacking end
as shown in Fig. 16.

Fig. 16. Grouting the ducts

406 W. W. P. K. Perera

3.3.12 Post-tensioning of the Wall

H.T.S. wires of 5 mm diameter with characteristic tensile strength 1670 N/mm2 is used
in post-tensioning the masonry wall. Anchor cones and pins which can accommodate 2
H.T.S. wires are used to hold the wires in position in the fixed end as well as the
jacking end. 10 mm thick mild steel bearing plates of size 75 mm  75 mm are placed
in between the cone and the concrete to facilitate smooth transfer of stresses. The wires
are then are stressed to 24 kN each using the hydraulic jack system and the jacking
force is transferred to the permanent anchor cone and pin as shown in Fig. 17.

Fig. 17. Stressing the H.T.S. wires

The basic stress-strain relationship

Fj=A
E ¼
e=l

is used to calculate the elongation relevant to the jacking force with following values.
E = Young’s modulus (205 kN/mm2)
Fj = Jacking Force (24 kN)
A = Cross sectional area of the tendon (p  2.52)
l = Initial length (2.35 m)
e = Elongation (12 mm)
After the H.T.S. wires are stressed, the anchor cone and the pin are fixed perma-
nently. The steel bearing plates, anchor cones and pins at both the jacking end and the
fixed end are covered with the 1: 2 cement sand mortar to prevent corrosion as shown
in Fig. 18.
 Development of a Post-tentioned Masonry Retaining Wall System 407

Fig. 18. Covering of anchor cones and pins

3.3.13 Back Filling of Foundation and the Embankment

The excavated trench was backfilled with good quality excavated soil and compacted in
300 mm thick layers with a soil compactor as shown in Fig. 19.

Fig. 19. Backfilling the foundation

Dial gauges were positioned as shown in Fig. 20 and the gap between the retaining
wall and the embankment was filled with selected excavated soil. The short term
deflection of the P.S.C. wall panels were measured in 300 mm intervals.
408 W. W. P. K. Perera

Fig. 20. Monitoring the deflection

4 Results

4.1 Short Term Deflection of the Wall

Readings of two dial gauges were recorded and the average of these readings is
tabulated in Table 1.

Table 1. Deflection against fill height

Fill height (mm) Defection (mm)
0 0
300 0.1
600 0.5
600 1.0
1200 1.3
1500 1.7
1800 2.0
2100 2.2

The verticality of the wall is measured with a plumb bob and recorded to check the
long term deflection or rotation. The distances to the adjacent building from the RCC
part of the retaining wall is measured with a steel tape and recorded to check the long
term sliding.

4.2 Results of Cost Compression

The cost of RRM, RC and Pre-Stressed Concrete (PSC) retaining walls of two different
heights were calculated and compared with the PTM retaining wall with the similar
 Development of a Post-tentioned Masonry Retaining Wall System 409

height and tabulated in Table 2. The structural detail for RRM retaining wall is
obtained from typical RRM sections designed by the Road Development Authority of
Sri Lanka and the structural design for RC is done by computer software “Prokon”. The
cost of a PSC retaining wall is obtained from the previous study of the author on
“Development of Pre-stressed Concrete Retaining Wall System [3]. The parameters
described in Sect. 3.2 above is considered for design of all Retaining walls.

Table 2. Cost comparison for 3 m wall length

Clear wall height (m) Cost (Rs.) and wall width (mm)
RRM RC PSC PTM
1.8 80,000 115,000 95,000 80,000
1050 200 100 150
3.0 140,000 185,000 160,000 135,000
1350 300 175 150

5 Conclusion

PTM retaining wall system can be used as a competitive alternative to the traditional
methods of retaining wall construction.
Though the cost of a PTM retaining wall is similar to that of an RRM retaining
wall, it needs only 15% of land width when compared to RRM. Hence it is suitable to
replace an RRM retaining wall with a PTM retaining wall in places where the land
value is critical.
The cost of a PTM retaining wall is 25%–30% less than that of an RC retaining wall
and the width of land requirement is less than that of an RC.
As the masonry blocks used to construct the retaining wall is manufactured using
the hydraulic cement: soil block making machine, manufacturing of masonry blocks
can be done without additional cost for purchasing machines.
The aesthetic appearance of PSC retaining wall is high compared to traditional
retaining walls.
Since manufacturing of masonry blocks and construction of foundation can be done
in parallel, the construction time can be reduced as well.

6 Future Research

The height of the post-tensioned masonry retaining wall can be increased with a
specially designed masonry block with reduced cross sectional area, increased sectional
modulus (Z) and the possibility of increasing the eccentricity (cellular block).

Acknowledgement. The author wishes to acknowledge the support given by the Technical
Officer, Mr. A.D.A.R. Premawansa in manufacturing of masonry blocks and construction of pilot
project in clinical building premises at Dalupitiya.
410 W. W. P. K. Perera

References
1. Curtin WG, Shaw G, Beck JK, Howard J. Design of post-tensioned brick work: the brick
development association, January 1989
2. Ganz HR (2002) Post-tensioned masonry around the world. In: Post-tensioning institute
engineering conference, San Antonio, TX, 5–7 May 2002
3. Modulus. Society of Structural Engineers Sri Lanka 27(01), 8–13 (2013)
 Roads for Sustainable Cities – What Is Wrong
with Us?

K. S. Weerasekera(&)

Department of Civil Engineering, The Open University of Sri Lanka,

Colombo, Sri Lanka
[email protected]

Abstract. When ‘roads for sustainable cities’ are to be considered, firstly, the
sustainability of roads should be brought into discussion. As an initiative, a road
user behaviour study was conducted in a busy suburban area outside South-
Eastern city limits of Colombo. The study area consisted of heterogeneous
vehicle composition; typically combined with high proportion of motorcycles
and three-wheelers which are poor in adhering to road rules that can commonly
be seen in most South Asian developing countries. Although heavy investments
are utilized in road improvement projects in these countries, the outcomes
gained often seems to be unsatisfactory. The intent of this study was to find out
the reasons for this unsatisfactory outcomes of the investments spent on road
improvement projects and suggest measures to overcome these unsatisfactory
outcomes for sustainability. The study was conducted in two parts; (i) road user
behaviour study, and (ii) field consultation study. Through the road user
behaviour study, the undesirable practices of all road users such as passenger
vehicles, motorcycles, three-wheelers, buses, trucks, pedestrians and excessive
noise were directly identified. It was seen from the consultation study results that
the varied expectations of different categories of road stake holders were dif-
ferent from category to category and biased towards their advantage. Finally,
through this study it was observed that the maximum benefits expected from
investing on road improvement projects are not fully realized when poor road
user practices are adopted by the stake holders’.

Keywords: Poor road practices
Colombo urban roads
Road indiscipline

1 Introduction

This study was conducted on roads just outside Colombo city limits to find out the
negative impacts on urban roads due to poor road user practices. It was suggested that
due to non-homogeneous vehicle composition; especially with high proportion of
motor-cycles and three-wheelers combined with other motorized vehicles may result in
the dual problem of poor traffic flows and high rate of road risk on urban roads
(Howorth, 1995). This was substantiated from the current study conducted around the
relatively busy suburban areas of Nawala and Nugegoda (i.e. outside South-Eastern
city limits of Colombo, Sri Lanka). It was observed that the inefficient use of the road
systems and poor road user behaviour is a major contributor to traffic problems, even
where there is no shortfall in road capacity. Hence, this study justifies the views of
© Springer Nature Singapore Pte Ltd. 2020
R. Dissanayake and P. Mendis (Eds.): ICSBE 2018, LNCE 44, pp. 411–418, 2020.
https://doi.org/10.1007/978-981-13-9749-3_36
412 K. S. Weerasekera

some recent studies conducted by (Edirisinghe 2011; Jayatilake et al. 2017 and
Dharmakeerthi and Sathyaprasad 2017) which indicate that heavy investments spent on
road improvements are not fully harnessed unless road user behaviour practices are
improved.
The area indicated in Fig. 1 was selected as the area of study which is the South-
Eastern city limits of Colombo. Figure 2 indicates some typical sections of roads in the
study area. During this study, a number of road users who have been accustomed to the
area for the last 25 years, consisting of drivers, residents, businessmen, daily office
travellers and pedestrians were consulted. Consultations consisted of question and
answer sessions posing different types of questions regarding road user behaviour
patterns of multitude of road users who are familiar with the study area.

2 Methodology

Currently, road safety is a complicated issue due to the increase of road users over time
and the improvement of vehicle engine efficiencies. The problem of accidents has
become more severe in highway engineering. Hence, the responsibility of traffic
engineers to carry out systematic studies of traffic accidents and to find the main causes
for accidents and finally come-up with preventive or remedial measures is important.

Fig. 1. Road network of the study area (South-Eastern city limits of Colombo)
 Roads for Sustainable Cities – What Is Wrong with Us? 413

Fig. 2. Typical road sections of the study area

Adhering to road rules and improved road discipline helps greatly in improving
road safety standards. The level of safety of a road system is influenced by the
behavioural practices of all road users, not only driver behaviour alone. The other types
of road users in addition to standard passenger cars and SUV drivers include motor-
cycle riders, three-wheel operators, bus and truck drivers and pedestrians. The traffic
composition of Colombo is different from most developed countries due to its extre-
mely heterogeneous vehicle fleet and very high proportion of motor-cycles and three-
wheelers. Hence, all road users and their behavioural practices had to be studied
carefully to improve road safety and road efficiency under the existing extremely
complicated situation. This has observed by Roberts and Simmonds (1995) and
Howorth (1995) in their work overseas. The current study tends to observe the local
misbehaviour patterns of the local stake holders.

2.1 Road User Behaviour

An observational study was performed to observe and determine road user behaviour
operations. Twelve observation locations were selected as indicated in Fig. 1 which is
distributed along the road network under study. The observations carried-out were
initially focused on studying the behaviours reflecting violations of road rules and other
dangerous road behaviour practices. The road user characteristics which were carefully
observed were; lane discipline, roadside parking and friction, turning movements,
failure to giveaway, irregular lane swapping, bus driver arrogant behaviour and not
properly stopping at bus-stops, road rage by large vehicles and the pedestrian move-
ments. Each study location (see Fig. 1) was under observation for the duration one-
hour during busy daytime.
414 K. S. Weerasekera

2.2 Field Consultations

In addition to the road user behaviour study, a field consultation study was conducted
for a total study sample of 500 consultations, which consisted of 100 drivers, 100 local
residents, 100 local businessmen, 100 office travellers and 100 pedestrians. Consul-
tations were conducted at random locations in the study area. A questionnaire survey
was executed in the field consultation study.
The questionnaire was focused on:
• Sufficiency of surrounding road network
• Adequacy of pedestrian facility
• Adequacy of vehicle parking facilities
• Safety of roads
• Road user behaviour
• Necessity of road widening

3 Results

3.1 Road User Behaviour Study Results

Through field observation studies conducted at 12 randomly selected locations in the
study area, different types of adverse road-user behaviour practices of different cate-
gories of road-users were identified and are categorized as indicated below.
All Vehicles
• Reluctance to giveaway at un-signalised junctions and small roundabouts
• Reluctance to giveaway at prioritized junctions
• Unauthorized roadside parking
• Unnecessary lane changes and overtaking
• Obstruction to overtaking

Motor-cycles
• Pillion passengers not wearing helmets
• Squeezing through other piled vehicles
• Keeping too close to other vehicles
• Riding on sidewalks
• Excessive speeding every so often
• Cutting across lanes dangerously
• Riding while talking on mobile phones
• Overtaking from the left
• Excessive noise
• Not using indicators during lane changing or turning
 Roads for Sustainable Cities – What Is Wrong with Us? 415

Three-wheelers
• Not sticking to lanes
• Sudden lane changes
• Excessive noise
• Travelling on the wrong side
• Cutting across lanes dangerously
• Overtaking from the left
• Keeping too close to other vehicles
• Riding on sidewalks
• Not using indicators during lane changing or turning

Buses
• Blocking other vehicles
• Not always sticking to assigned bus stops
• Overstaying at bus stops
• High speeds
• Travelling on two lanes
• Dangerously sudden stops
• Not using indicators during lane changing or turning
• Excessive use of loud horns

Trucks
• Blocking other vehicles
• Aggressive driving
• High speeds
• Travelling on two lanes
• Not using indicators during lane changing or turning

Pedestrians
• Often not using footpaths
• Jaywalking
• Not using zebra crossings
• Not waiting for green phase for crossing at signalized junctions

Excessive Use of Horns

Excessive use of horns (sometimes very loud) very frequently by Colombo drivers is
frustrating. When closely studied, it was found that the reasons are as follows:
• To alert slower vehicles to make way
• To alert pedestrians on the road of approaching vehicles
• To express annoyance
• To facilitate overtaking
• To alert motor-cycles and three-wheelers
• By three-wheel operators to mark their presence
416 K. S. Weerasekera

4 Discussion

It is clear from Fig. 3A that drivers are most unhappy about the inadequacy of sur-
rounding road network and closely followed by office travellers. On contrary it is
observed that pedestrians are happy about the surrounding road network. It also
observe that residents and local businessmen too are generally satisfied with the ade-
quacy of the surrounding road network.
From Fig. 3B it is seen that pedestrians are generally worried about the inadequate
pedestrian facilities available. Generally, residents too are not very satisfied about the
adequacy of pedestrian facilities that are available currently. On the contrary, drivers
are least worried about the availability of pedestrian facilities.
Figure 3C clearly indicates the unhappiness of local businessmen and drivers about
the inadequacy of parking facilities. It’s also clear that even the local residents expect a
higher level of parking facility in the area. It also could be noted that pedestrians and
office travellers are not bothered about parking facilities available in the surrounding
areas.
From Fig. 3D it is indicated that except pedestrians, all the other categories are
generally satisfied about the safety level of these urban roads. Majority of the pedes-
trians do conclude that the roads are not safe.
Figure 3E indicates that it is mainly drivers and residents who complain about the
unsatisfactory level of road user behaviour, but local businessmen and office travellers
are not very concerned about road user behaviour. The pedestrians have an opposite
view; that they think the road user behaviour is satisfactory from their point of view.
From Fig. 3F it is interesting to note that greater majority of residents and local
businessmen are not in favour of widening the roads fearing that their land will get
affected. But on contrary majority of all the other categories do agree in widening the
roads.
It could be seen from the consultation study results that the expectations of different
categories of stake holders’ were different from category to category probably biased
towards their advantage. For instance, many drivers and travellers preferred road
widening to be carried-out, but residents and road side business community were
against road widening since they were going to be affected. It also could be seen that
pedestrians were unaware of the road capacity situation, but they were worried about
the lack of pedestrian facilities in the area. When it comes to adequacy of parking in the
area, mainly the business community and drivers were not happy about the inadequacy
in parking availability although travellers and pedestrians were least bothered about
parking facility in the area. It can be seen that generally all the categories of road users
were concerned about road safety, especially pedestrians and drivers who had a high
priority concern for improving road safety. Most of the drivers were not satisfied with
the behaviour of the road users in general. In this regard, pedestrians were the least
bothered.
 Roads for Sustainable Cities – What Is Wrong with Us? 417

5 Conclusions

Road safety depends when road user behaviour is combined with vehicle and road
characteristics. Hence, road user behaviour of vehicle drivers is due to; lack of
understanding of road signs and some carriageway markings, lack of knowledge of
priority rules, right of way concept not being properly understood and practiced,
reversing to main roads, obstructing the main flow, roadside parking causing side
friction, and road raging by bus and heavy vehicle drivers. It could be clearly seen that
efficiency of a road depends very much on smooth traffic flow along the roads with;
(i) least side friction along mid-blocks, and (ii) orderly maximum throughputs at
intersections. The main reasons behind the increase in side-friction and flow in mid-
blocks is due to unnecessary lane changing on roads, roadside parking and slowing
down due to business activities along the sides of main roads, vehicles cutting across
lanes and parking along opposite side of the road and buses overtaking and competing
with each other.
The road regulations and conditions that the road users have to adhere are; con-
dition of vehicle (road worthiness), speed limits in urban areas, give-way and priority
regulations, adopting correct overtaking practices, buses sticking to assigned bus stops,
proper post-accident practices, correct parking practices, sticking to lane discipline and
strictly adhering to obligations at pedestrian road crossings.
Addressing all those issues scientifically is vital for the sustainability of roads for
sustainable cities. Finally, through this study, it was observed that the maximum
benefits expected from investing on road improvement projects are not fully realized
when poor road user practices are adopted by the stake holders.

Acknowledgements. Author wishes to acknowledge all who patiently took part in the con-
sultations which consisted of drivers, residents, businessmen, daily office travellers and pedes-
trians who were familiar with the study area that supported this study with their valuable views
and suggestions.

References
Bandara JMSJ (2011) Need for an integrated highway management system: Developments in Sri
Lankan context. In: Proceedings of the Joint International Symposium on Social Management
Systems (SMSS-2011), pp. 700–703 Colombo, Sri Lanka
Dharmakeerthi UGS, Sathyaprasad IMS (2017) Development of trip attraction rates and parking
standards for supermarkets in Kandy area. In: Proceedings of the 5th International
Symposium on Advances of Civil and Environmental Engineering Practices for Sustainable
Development (ACEPS-2017) March 2017, pp. 198–205. University of Ruhuna, Galle, Sri
Lanka
Edirisinghe JE (2011) Driver behaviour at pedestrian crossings. In: Proceedings of the Joint
International Symposium on Social Management Systems (SMSS-2011), pp. 696–700
Colombo, Sri Lanka
Howorth N (1995) Road user behaviour and safety in a developing country. In: Proceedings of
the 7th World Conference on Transport Research (WCTR), July 1995, pp. 302–306, Sydney,
Australia
418 K. S. Weerasekera

Jayatilake PASR, Premachandra BMNM, Wickramasinghe WMVSK (2017) Variation of the

start-up time of pedestrians with their familiarity to the signal phase. In: Proceedings of the
5th International Symposium on Advances of Civil and Environmental Engineering Practices
for Sustainable Development (ACEPS-2017) March 2017, pp. 206–212. University of
Ruhuna, Galle, Sri Lanka
Roberts M, Simmonds DC (1995) A strategic modelling approach for urban transport policy
development. In: Proceedings of the 7th World Conference on Transport Research (WCTR),
July 1995, pp. 285–289, Sydney, Australia
 A Sustainable Green Façade Retrofit
for An Iconic Built Structure:
A Case Study at SAINTGITS

Jacob Thottathil Varghese1,2(&), Sajan Thomas1, and Joselin Herbert3

1
SAINTGITS College of Engineering, Kottayam, Kerala, India
[email protected]
2
VIT Vellore, Vellore, Tamil Nadu, India
3
NICHE, Kumaracoil, Tamil Nadu, India

Abstract. The depleting fossil fuel reserves and the constantly surging levels of
atmospheric pollution have made the architects and building owners join hands to
foster sustainable initiatives. A modest experimentation in a sustainable domain
had been inculcated at SAINTGITS, a much sought academic institution at Kot-
tayam, in the state of Kerala, India. The campus houses multiple high rise buildings
and a plenty of other small structures. The iconic structure, the Central Lecture
Complex (CLC) exhibits its uniqueness with an arc shaped plinth which houses 3
stories to accommodate several conventional lecture rooms, faculty cabins,
sophisticated seminar halls and other essential amenity spaces. It provides natural
shade to the SAINTGITS Amphi theatre, structured and maintained at the eastern
side. The proposed green façade retrofit on to the East and the West facing walls of
CLC has improved the indoor operative temperature and hence, an approximate
15% reduction has been estimated in the building cooling load. An additional 2 m
high temporary parapet wall has been recommended on to the existing CLC to
provide extended shading hours to the SAINTGITS Amphi, right from 14:15 h
onwards. The comprehensive acoustic modelling of the latter has given promising
results for the Reverberation Time (RT60), which is on par with the international
standards for performance spaces - ISO 3382-1 and ASTM E2235. The analyses
inferred that the state-of-the art design of the existing CLC and the proposed green
façade retrofits will reduce the entire energy demand by 20% and it certainly will
provide guidelines to similar structures at tropical metropolises.

Keywords: Green façade
Sustainable retrofits
SAINTGITS Amphi

Operative temperature
Built structure
Energy analyses

1 Introduction

The industrial revolution has turned the metropolises denser and made way for the
evolution of new ones. Massively Built structures, unavailability of free ground space,
energy crunch, traffic density, rising city heat and the alarming levels of atmospheric
pollution are some of the major concerns that seek urgent intervention (Jacob et al.
2015). High dependence on fossil fuel has made the scientific community to deliberate
on alternate sustainable initiatives (Takakura et al. 2000; Jacob et al. 2018). One such

© Springer Nature Singapore Pte Ltd. 2020

R. Dissanayake and P. Mendis (Eds.): ICSBE 2018, LNCE 44, pp. 419–432, 2020.
https://doi.org/10.1007/978-981-13-9749-3_37
420 J. T. Varghese et al.

alternative is the implementation of sustainable retrofits on to the existing high rise

built structures. These retrofits are trellises of green façades, either draped directly on to
the outer fabric of built structures to absorb the direct radiation or erected alongside
busy pavements to facilitate the absorption of the atmospheric pollutants (Yang et al.
2008; Jacob et al. 2010, 2015) (Fig. 1).

Fig. 1. A fully grown Green façade trellis (almost 100 m length) alongside the Main Central
Road, Kerala state, India.

The former initiative helps to convert 60% of the absorbed radiation into latent heat
(Campbell et al. 1998). Hence, the thermal insulation and noise attenuation of the
building can have improved (Pal et al. 2000; Wolverton et al. 1989; Jacob et al. 2015).
For every 1 °C dip in the operative temperature, the energy load of the respective built
structure reduces by 12–15% (Stec et al. 2005; Jacob et al. 2018). Green façades having
larger Leaf Area Index (LAI) cools the building more efficiently (Fig. 1) and at least
35% deduction can be expected in the annual cooling load (Wong et al. 2010;
Alexandri and Jones 2008)
A modest initiative has been proposed at SAINTGITS, a renowned educational
institution in India, to improve the indoor operative temperature. The SAINTGITS
group of Institutions, founded in the year 2002, has its lush and verdant campus located
at Central Travancore (9.5090 N and 76.5510 E), Kerala state, India (Fig. 2).

Fig. 2. Areal view of the SAINTGITS group of professional institutions.

 A Sustainable Green Façade Retrofit for an Iconic Built Structure 421

2 Methodology

Among the numerous multi-storied structures at SAINTGITS campus, the Centre

Lecture Complex (CLC) and the SAINTGITS Amphi are the state-of-the art built
structures. Whilst both the structures have arc shaped plinths, the latter is naturally
carved as multiple stairs on the hilly terrain (Fig. 3).

Fig. 3. The arc shaped Centre Lecture Complex (CLC) and the SAINTGITS Amphi

At the initial phase, we were successful in preparing and analysing an architectural

model which replicated the entire CLC. The simulation results were at par with the
previous similar studies (Jacob et al. 2015) and were later validated by conducting real
time energy audits.
The next stage of the work estimated the benefits that could be attained, if a suitable
green façade was retrofitted on to the East and West facing walls of the CLC building.
The third and the fourth phases contain detailed daylight and shadow analyses,
performed exclusively for the entire CLC. Whilst the conclusions from the latter
rendered an insight in to the real hours of occupant comfort at the SAINTGITS Amphi,
the former had explicitly revealed the ability of CLC’s first and the second floors to
harness the maximum daylight.
The final phase incorporated the acoustic analyses at SAINTGITS Amphi. The
reverberation time was estimated by recreating an architectural model of the entire
acoustic space.

3 Results and Discussion

The exclusive analyses, the first of its kind at SAINTGITS, were done for the entire CLC
and SAINTGITS Amphi. For the ease of discussing in detail, the entire results are
included into five - Subsect. 3.1 whole building’s energy load calculation, Subsect. 3.2
attainable dip in operative temperature of a fully air-conditioned & Green façade retro-
fitted CLC block, Subsect. 3.3 harnessing the natural light source - daylight analyses of
the CLC block, Subsect. 3.4 extended occupant comfort at the SAINTGITS Amphi and
Subsect. 3.5 estimating the reverberation time at SAINTGITS Amphi.
422 J. T. Varghese et al.

3.1 Whole Building’s Energy Load Calculation

The entire CLC is a full-fledged 3 storied structure with a temporary third floor (Fig. 4).

Fig. 4. Spatial arrangement of various zones at the individual levels of the Centre Lecture
Complex (CLC).

The ground floor has 11 zones. The first and the second floors share equal number
of zones (10 zones per floor) (Fig. 4). All zones in the ground and the topmost floor are
earmarked as class rooms (capacity 72 seats in each zone) to accommodate the

Fig. 5. A stark comparison (floor wise) of the various energy consuming sources and the related
monetary outflow in USD – monthly data for the entire CLC block.
 A Sustainable Green Façade Retrofit for an Iconic Built Structure 423

sophomores, juniors and senior students. The first floor serves as exclusive zones for
two major engineering departments–the Mechanical Engineering and the Electrical
Engineering.
Real time energy audits of all the zones were conducted. Monthly analysis of the
audit summary revealed that the artificial lighting consumed the least energy units
(1071 units) (Fig. 5).
Whilst the energy expended by the ordinary air-coolers and fans was 3348 units per
month, a proposed fully air-conditioned CLC’s monthly cooling load was approxi-
mately 51544 units; 15 times more than the energy expended to operate all ordinary
cooling equipment (Fig. 5). Since the first and the second floors share almost the same
physical geometry, the monthly power consumption (cooling load alone) of both the
floors together is nearly 4400 USD.
Since the combined energy expended by the cooling devices (ordinary air-coolers,
fans and air-conditioners) is significantly higher (nearly 10 times) than the artificial
lighting load, it is advisable to inculcate reliable and sustainable initiatives to assuage
the cooling load.

3.2 Attainable Dip in Operative Temperature of a Fully Air-Conditioned

and Green Façade Retrofitted CLC Block
Year after year, as the tropics get warmer, a fully air-conditioned CLC block will rather
become essential than luxury. Though it increases the operative cost, to a certain extent,
it can be negated by a custom designed green façade retrofits. Earlier experimental and
simulation studies on two plant species (Vernonia elaeagnifolia and Passiflora caer-
ulea) at Vellore Institute of Technology (VIT, Vellore, Tamil Nadu, India) proved that
these green drapes can bring down the indoor operative temperature of built structures
(Jacob et al. 2011, 2015). Particularly, in summer, the ambient air temperature near the
vegetated façade dropped by 1 °C to 3 °C (Hoyano 1988).
The entire Centre Lecture Complex (CLC block) with pitched roof was modelled
using the most modern and versatile architectural software, the DesignBuilder (Fig. 6).

Fig. 6. A comprehensive architectural model of the Centre Lecture Complex (CLC block),
modelled in the DesignBuilder.
424 J. T. Varghese et al.

The energy calculations were done according to the ASHRAE standards by directly
plugging in the pre-written energy analyses codes from EnergyPlus, a building energy
simulation program, developed by the National Renewable Energy Laboratory
(NREL), Colorado, USA.
The East and the West facing walls of the CLC are fully retrofitted by trellises of
green façades. The plant species used is a mix of Vernonia elaeagnifolia and Passiflora
caerulea – whilst the former can be draped on high rise structures, the latter is efficient
in removing the atmospheric pollutants, hovering at lower heights – roughly up to 4 m
(Jacob et al. 2015).
A stark comparison of the average leaf area of the two different plant species shows
that a 5 lobbed leaf structure covers nearly twice the area than a single lobbed structure
(Fig. 7).

Fig. 7. Comparing the average leaf area of a five lobbed leaf structure (average leaf
area = 16 cm2) and single lobbed Vernonia elaeagnifolia (average leaf area = 8 cm2).

The input parameters of the model are the façade dimensions, building orientation,
the weather data file and the U-values. For both the plant species, the U value is taken
as 0.192 W/m2K, which is much closer to the value of straw bale (Jacob 2017; Behzad
et al. 2011). The weather data for Trivandrum was used and it was a reasonable
approximation for Kottayam.
The operative temperature, which is a direct indication of the indoor occupant
comfort, is governed by the air temperature, mean radiant temperature and the air speed
- see Eq. (1).
  pffiffiffiffiffiffiffiffi
tr þ ta 10v
Operative tempeature ¼  pffiffiffiffiffiffiffiffi ð1Þ
1 þ 10v

Where
tr = air temperature (degree C)
ta = mean radiant temperature (degree C)
v = air speed (m/s)
 A Sustainable Green Façade Retrofit for an Iconic Built Structure 425

Hence, if the air speed (v) is 0.1 m/s, the analytical expression for the operative
temperature attains a further simplified form – see Eq. (2).

½tr þ ta 
Operative tempeature ¼ ð2Þ
2

Fig. 8. Indoor operative temperature for a Bare wall vs Green façade retrofitted wall – general
trend during the various quarters of the year.
426 J. T. Varghese et al.

The DesignBuilder model was run for all the days of the year (four quarters) with
constant air speed 0.1 m/s. Whilst the operative temperature exhibits mutually
opposing trends for first and the last quarters respectively, the trend almost plateaus for
both the 2nd and the 3rd quarters (Fig. 8). It is validated by the general perception of the
quarterly temperature variation that occurs every year - the first and the last quarters are
cooler than the remaining.
It was observed that, throughout the year, with a green façade retrofit on the outer
fabric, the average indoor operative temperature dipped by approximately 1 °C than the
respective bare wall (Fig. 9).

Fig. 9. Dip in the operative temperature experienced due to the green façade retrofit.

To analyse the trend of dip in operative temperature for different values of air speed
(v), the data has been plotted for the 15th and the 30th day of every month (Fig. 10).

Fig. 10. Pattern of the dip in the operative temperature for different values of air speed.

It is inferred that the increase in air speed improves the air change/hour (ac/h) inside
the respective zones and thus, more dip in the operative temperature is experienced. If
the air speed is 1 m/s, a maximum dip of 1.1 °C is attainable with a fully grown green
façade retrofit.
 A Sustainable Green Façade Retrofit for an Iconic Built Structure 427

Based on the earlier similar studies, if the operative temperature drops by more than
1 °C, a modest energy saving of 15% is achievable (Stec et al. 2005; Jacob et al. 2018).
Hence, a monthly saving to the tune of 660 USD is reachable with a suitable green
façade retrofitted CLC block.

3.3 Harnessing the Natural Light Source – Daylight Analyses of the CLC
Block
Though the energy expended for providing artificial lighting is the least among all the
other expended categories, it cannot be neglected. The daylight analyses show that the
interiors of every zone receive 50% less illuminance than the exteriors (Fig. 11).
Hence, throughout the working hours, the existing design of all the zones of the CLC
demands artificial lighting. The zones in the ground floor are deprived of natural
daylight than the first and the second floors (Fig. 11). The close proximity of the
ground floor to a confined veranda is one of the major reasons.

Fig. 11. Daylight analyses of the various zones – arranged floor wise.

By extrapolating the lighting demand, if the artificial lights are assumed to be

functioning throughout the day, the monthly lighting load for the entire CLC sums to
an approximate amount of 450 USD (31,000 INR). The analyses revealed that by
providing sufficient windows at the East facing wall, the artificial lighting load can be
reduced by 1/3rd from the existing.

3.4 Extended Occupant Comfort at the SAINTGITS Amphi

The first floor’s balcony of the Centre Lecture Complex (CLC) provides extended
seating for SAINTGITS Amphi (Fig. 3). During sunny days, the occupant comfort at
the latter is challenging – especially during the forenoon hours. Until the afternoon of
all the days, the entire amphitheatre receives direct sunlight. The annual sun path
diagram explicitly reiterates the fact (Fig. 12).
428 J. T. Varghese et al.

Fig. 12. The annual sun path diagram over the Centre Lecture Complex (CLC Block) and the
SAINTGITS Amphi.

Hence, the complete utilization of the amphitheatre could not be claimed during the
entire active hours of the day (Fig. 13).

Fig. 13. The extent of shadows cast by the CLC block on SAINTGITS Amphi at 3:10 pm and
5:15 pm.
 A Sustainable Green Façade Retrofit for an Iconic Built Structure 429

The shadow details collected for different quarters reveal that the topmost seating at
the amphitheatre receives the foremost comfort shadow of CLC at 15:00 h (3.00 pm) -
during certain quarters, it even extends to 3:30 pm. The front most row of the
amphitheatre receives the shadow, not earlier than 5:15 pm on any of the sunny days
(Fig. 13).
The data in Table 1 indicates the respective timings at which the occupants at the
amphitheatre receive the benefits of CLC’s shadow. The attainable comfort duration
also is appended (Table 1).

Table 1. Occupant comfort – the extended duration of attainable benefits at SAINTGITS

Amphitheatre.
Month & Time of receiving Time of having complete Time and attainable
quarter the first shadow at occupant comfort at all the extended benefits/gains
of the the topmost row rows of SAINTGITS Amphi with 2 m high parapet wall
year (pm) (pm) above CLC
Shadow from Gain
(pm)
Jan (Q1) 3:30 5:15 2:45 45 min
April 3:30 5:30 3:00 30 min
(Q2)
July (Q3) 3:30 5:45 3:15 15 min
Oct (Q4) 3:00 5:15 2:15 45 min

If the height of the existing CLC block is increased temporarily by 2 m, either by a

parapet wall or vertically arranged temporary corrugated roofing sheets, the duration of
the live hours of all the afternoons can be extended by at least 45 min.
While conducting several major events at SAINTGITS, the existing practice is to
raise temporary roofs above the amphitheatre. It can comfortably accommodate at least
2500 seats and thus the live and comfortable hours are made available throughout the
day.

3.5 Estimating the Reverberation Time at SAINTGITS Amphi

The two major events at SAINTGITS, the cultural festival (NAKSHATRA) and the
technical festival (PRATITYA) attract student participants from all parts of the country.
These annual events are conducted at SAINTGITS Amphi, right from the morning
hours. Every time, makeshift arrangements are done at the amphitheatre to accom-
modate the crowd.
An architectural model was created in Autodesk’s Ecotect to find the reverberation
effects at the amphitheatre (Fig. 14). For evaluating the reverberation effects, the
acoustic zone assumed the entire amphitheatre and CLC as a single zone of volume
100187 m3.
430 J. T. Varghese et al.

Fig. 14. An Ecotect model of the acoustic zone of SAINTGITS Amphi – the wire frame and
rendered view.

The input parameters have been fixed slightly higher than the nominal values. The
entire acoustic space is designed to accommodate 2500 seats (persons). Two tower
speaker systems having 1000 W each was given as the sound input. The reverberation
time (RT60 – the time elapsed to shed 60 decibels, measured from the originating time)
was calculated by two independent algorithms – (i) the Norris – Eyring (NE) and
(ii) the Millington –Settle (MS) (Fig. 15).

Fig. 15. The Reverberation Time (RT60) measured independently by Norris-Eyring and
Millington-Settle algorithms for the SAINTGITS amphi.

The values of RT60 were plotted for varying levels of occupancy, independently by
each of algorithms (Fig. 15). The Millington-Settle algorithm is widely accepted for its
acceptable range of RT60 values.
Whilst for the Norris-Eyring algorithm, the plot yielded a steady gradient (RT60
between 2.8 s to 2.5 s), the trend observed for Millington-Settle is a sharp gradient
followed by a plateau (RT60 between 0.82 s to 0.79 s), for 0 to 4000 occupancy
(Fig. 15).
 A Sustainable Green Façade Retrofit for an Iconic Built Structure 431

Based on the International Standard for performance spaces, the ISO 3382-1 and
ASTM E2235, the occupied state is defined as 80%–100% occupancy. Hence, the
range of analyses had been limited to the values ranging between 2000 to 2500
occupants respectively for 80% to 100% occupancy. For a performing space to have
richer musical sound, the suggested RT60 values lie in between 2.5 s–3.5 s. The RT60
based on the Norris-Eyring algorithm validates the same (Fig. 15).

4 Conclusion

The five stage scientific analyses of the Centre Lecture Complex (CLC) and the
SAINTGITS Amphi are the first of its kind at SAINTGITS campus. The study has
revealed the capability of sustainable retrofits on existing built structure. It has
explicitly measured the hours of occupant comfort at the amphitheatre and suggested a
viable measure to extend the comfort hours. Whilst similar analysis stands promising
for several upcoming multi-storied projects at SAINTGITS campus, it revealed the
significance of energy analyses for built structures at tropical climes.

References
Alexandri E, Jones P (2008) Temperature decreases in an urban canyon due to green walls and
green roofs in diverse climates. Build Environ 43:480–493
Behzad S, Deepak R, Barbara J, Jabuk W, Rosi F (2011) The carbon reduction potential of straw
bale housing. Build Res Inf 39(1):1–56
Campbell D, Gaylon S, Norman JM (1998) Introduction to Environmental Biophysics, 2nd edn.
Springer, New York
Hoyano A (1988) Climatological uses of plants for solar control and the effects on the thermal
environment of a building. Energy Build 11:181–199
Jacob Thottathil V, Balamuralikrishna C, Ghosh S (2010) Use of green facades in sustainable
building environments: Quantifying the uptake rates of air pollutants by facades draped with
tropical creepers. Proc Int Conf Sustain Built Environ ICSBE, Kandy, Colombo, Sri Lanka
1:363–373
Jacob Thottathil V, Balamuralikrishna C, Ghosh S (2011) Role of green facades in creating
sustainable environments: Comparing chemical analyses and SEM images to quantify the
atmospheric cleansing by green creepers. In: Proceedings of the International Conference on
Ecosystem Conservation and Sustainable Development, ICECSD, Paris, France, vol 77,
pp 343–346
Jacob Thottathil V, Ghosh S, Swentank P, Rudranshu S (2015) Evaluating the cleansing
efficiency of an extended living façade draped with Vernonia elaeagnifolia. J Green Build 10
(2):157–177
Jacob Thottathil Varghese (2017) Mass and Energy Transfer Process In High Rise Green façade
Structures. LAP Lambert Academic Publishing, Beau Bassin. ISBN 978-3-659-51994-9
Jacob Thottathil V, Ghosh S, Anshul SG, Manchanda R, Afsan A, Prathyusha M (2018) The
projected thermal attributes and energy saving potential of Green Façade draped multi-storied
cuboid buildings at major Indian cities: an ecotect analysis. World J Modell Simul 14(1):12–20
Pal AK, Kumar V, Saxena NC (2000) Noise attenuation by green belts. J Sound Vib 234
(1):149–165
432 J. T. Varghese et al.

Stec WJ, Van Paassen AHC, Maziarz A (2005) Modelling the double skin facade with plants.
Energy Build 37(5):419–427
Takakura T, Kitade S, Goto E (2000) Cooling effect of greenery cover over a building. Energy
Build 31:1–6
Wolverton BC, Johnson A, Bounds K (1989) Interior landscape plants for indoor air pollution
abatement. Interiorscape 11:37–63
Wong N, Tan A, Tan P, Chiang K, Wong N (2010) Acoustics evaluation of vertical greenery
systems for building walls. Build Environ 45:411–420
Yang J, Yu Q, Gong P (2008) Quantifying air pollution removal by green roofs in Chicago.
Atmos Environ 42:7266–7273
 Embankment Construction on Power Blender
Cement Mixed Soft Soil Treated Ground
for CKE Bypass

N. Vasanthan1(&), H. Ohashi1, F. Obara1, S. D. L. Tilak2,

M. C. W. Dissanayake2, and A. Karunawardena3
1
Oriental Consultant and Katahira Engineering International JV,
Colombo, Sri Lanka
[email protected]
2
Road Development Authority, Ministry of Highway & Road Development,
Colombo, Sri Lanka
3
National Building Research Organization, Colombo, Sri Lanka

Abstract. Power Blender cement mix soft soil improvement method was
adopted for the construction of embankment of the 4-lane bypass of Colombo-
Katunayake Expressway for the New Bridge Construction over the Kelani River
in Sri Lanka. This treatment method is a rapid construction method, first applied
in Sri Lanka. The embankment height varies from 2 m to 5 m and peat/organic
clay exists for a depth of about 10 m. Soil cement columns are spaced at
1 m
2 m rectangular shape and the improvement depth is about 10 m. After
completion of the ground treatment for the entire area and confirming the quality
of the improvement, the embankment was constructed with 2 layers of geo-
textiles. During the embankment construction, settlement and lateral movement
were measured. The expected settlements and lateral deformations were esti-
mated and compared with field measurements. This paper describes the geo-
logical condition, embankment construction sequence, settlement estimate and
stability analysis, measurement results and the comparison with the respective
estimates.

Keywords: Power blender method  Embankment construction 

Peat/organic clay  Settlements  Geotextiles  Stability analysis

1 Introduction

The Project consists of a bypass of about 480 m length with 4-lane traffic at the CKE
(Colombo - Katunayake Expressway) end section as shown in Fig. 1.

© Springer Nature Singapore Pte Ltd. 2020

R. Dissanayake and P. Mendis (Eds.): ICSBE 2018, LNCE 44, pp. 433–446, 2020.
https://doi.org/10.1007/978-981-13-9749-3_38
434 N. Vasanthan et al.

Katunayake

Colombo

Bypass Road

Fig. 1. Bypass section at the CKE end section

The height of the embankment varies from 2 m to 5 m over the soft ground. The
existing CKE is running parallel apart to the proposed bypass at about 20 m to 30 m
distance. The CKE was constructed on a similar soft ground condition of the proposed
bypass using sand compaction piles (SCP) at 1.3 m centre to centre spacing.
Thick fibrous peat and organic clay/silt deposits are distributed over the area with
localised swamps. The larger portion of the deposits predominantly consist of a thick
layer of very soft to firm fibrous peat and organic clays, and below this layer stiff to
hard gravelly clay layers and stiff to very stiff organic silt/clay exist at a deeper level.
The profile indicates some variation of engineering properties of the stratification
(Fig. 2). The upper stratum of the bypass area consists of a very soft to firm clay layer
with undrain shear strength varying from 10 kPa to 25 kPa. Thick, very soft to soft
peat/organic clay/organic silt layers of undrain shear strength of 5 kPa is underlying the
above. This layer, if untreated, exhibits significant settlement characteristics of the
bypass section. Then a relatively hard clayey layer exists underneath. Soft ground
treatment is designed using data from recent geotechnical investigation (12 Nos. of
exploratory boreholes) and limited laboratory test results with the experience gained in
similar projects in Sri Lanka.
 Embankment Construction on Power Blender Cement 435

Fig. 2. Subsoil profile with SPT-N value

2 Proposed Ground Treatment

Column Supported Embankment (CSE) using Power Blender Cement Mixing tech-
nique is considered (Ohashi et al. 2018) as a technically feasible option for foundation
improvement of the bypass due to time constraint in completing the project. Based on
subsoil condition and layer thickness recorded from soil investigation, the ground
improvement area was divided into three cases; namely Case 1, Case 2 and Case 3, for
different mixes and area improvement ratios of 40%, 50%, 33% respectively as shown
in Fig. 3.
The load from the embankment is designed to be transferred to Power Blender
Cement Mixing columns with two layers of geotextiles of 600 kN/m which is con-
sidered in a manner such as to minimize the differential settlements and laid in both
directions.
Spacing of the columns is designed to transfer the load effectively to the columns.
Soil arching theory helps to load transfer mechanism of column supported embank-
ment. The stresses developed within Power Blender Cement Mixing columns increase
with time due to the development of arches above the columns.
Part of the load of the soil wedge below the soil arch is also transferred to the
columns by the tension developed within the geotextile layers laid on top of the treated
soil. This is called the membrane action of the geotextile layer.
436 N. Vasanthan et al.

Case III, ap=33% Case II, ap=50% Case I, ap=40%

Fig. 3. Improvement sections with different mixes

The load transfer mechanism is a combination of arching within the embankment

fill over the Power Blender Cement Mixing columns and membrane action of the
geotextile layer.
Figure 4 shows the basic arrangement of a column supported embankment.
 Embankment Construction on Power Blender Cement 437

Geotextile layers

Embankment fill

Improved soil Unimproved soil

Fig. 4. Basic arrangement of a column supported embankment

3 Methodology

The following procedures have been adopted in the design of CSE with Power Blender
Cement Mixing piles for the 480 m long bypass section based on the Federal Highway
Administration Design (2013).
• Estimate preliminary column spacing.
• Determine required column load capacity.
• Select preliminary column type based on required column load capacity and site
geotechnical requirements.
• Determine capacity of column to satisfy the serviceability design requirements.
• Determine extent of columns required across the embankment width.
• Determine the geotextile requirements based on estimated column spacing and
lateral spreading.
• Check the global stability.

4 Settlement Analysis

Road embankments usually experience settlements when they are built on soft com-
pressible foundation soils. Since the area is cement-treated, theoretically, the settlement
is relatively very small. However, clayey soil layer exists underneath the treated soil,
and may lead to some settlement. Settlements that occur very slowly, over decades or
more, may be tolerable by pavements during the service within the scheduled main-
tenance period. If these settlements occur within a period of a few months or years, they
would result in severe pavement damage.
It is, therefore, important to be able to measure the settlement to reasonable
accuracy over a period of time to confirm whether the settlement is tolerable by the
pavement and contributes to good riding quality by road users.
438 N. Vasanthan et al.

4.1 Calculation of Stress Increase Due to the Proposed Embankment

Construction
Stress increments (r) due to the proposed embankment construction with varying
embankment height (He) and traffic load (Q) extracted from Japanese manual are given
in Table 1.

Table 1. Stress increment with height of embankment

Case no. Unit weight, c He Q Dp
(kN/m3) (m) (kN/m2) (kN/m2)
1 19 3.2 20 80.8
2 19 2.5 20 67.5
3-1 19 3.4 20 84.6
3-2 19 5.0 10 105.0

4.2 Consolidation Settlement Without Ground Improvement

Consolidation settlement, S0, was calculated using the coefficient of volume com-
pressibility, mv0, stress increase, r and soft soil thickness, H, as shown in Eq. (1).

S0 ¼ mm0
H
r ð1Þ

The consolidation settlement, S0 estimated without ground improvement is sum-

marized in a Table 2.

Table 2. Consolidation settlement without ground improvement

Case no. He (m) r (kN/m2) mv0 S0
(m2/kN) (m)
1 3.2 80.8 0.006 1.53
2 2.5 67.5 0.008 1.36
3-1 3.4 84.6 0.003 0.81
3-2 5.0 105.0 0.002 0.94

4.3 Design Standard Strength of Soft Soil for the Ground Improvement
From the stress increase, r, area improvement ratio, ap and the safety factor, Fc, the
design strength of the improved ground, quck is obtained by the Eq. (2).

quck ¼ r=ap
Fc ð2Þ

For each case, the improvement ratio and the design standard strength are given in
Table 3.
 Embankment Construction on Power Blender Cement 439

Table 3. Improvement ratio and design strength

Case no. He Dp ap Fc quck
(m) (kN/m2) (%) (kN/m2)
1 3.2 80.8 40 1.2 242 250
2 2.5 67.5 50 1.2 162 200
3-1 3.4 84.6 33 1.2 308 350
3-2 5.0 105.0 33 1.2 382 400

4.4 Residual Settlement After Ground Improvement

For the calculation of residual settlement after improvement, the following two
methods were adopted:
(1) Based on the stress sharing ratio
(2) Considering the arch effect of the embankment fill.

4.4.1 Method Based on the Stress Sharing Ratio

The total residual settlement based on the stress sharing ratio method (S1) is calculated
by the Eqs. (3), (4), (5) and (6).

S1 ¼ rs =r
S0 ð3Þ


rs =r ¼ 1= ðn  1Þ
ap þ 1 ð4Þ

n ¼ rc =rs ð5Þ

S0 ¼ mv0
E
r ð6Þ

Where S0 is the settlement of the untreated ground, r is the average stress (or stress
increase), rs is the stress on the surrounding soil, rc is the stress on the column, ap is
the area improvement ratio and n is the stress sharing ratio (or stress concentration
ratio).
A schematic diagram of the method based on the stress sharing ratio is shown in
Fig. 5. The “n” value was adopted from Manual for Soft Ground Improvement in
Japan.
The “n” value adopted for this project was between 15 to 20.
440 N. Vasanthan et al.

Fig. 5. Schematic diagram of the method based on the stress sharing ratio

For each case, the residual settlement was obtained as given in Table 4. For Case 1
and Case 3, the grounding type (G) was adopted and for Case 2, floating type (F) was
implemented owing to a soft ground layer.

Table 4. Total settlement after ground improvement

Case no. S0 (m) ap n b′ S′ S″ S1 (m)
(%) (m) (m)
Case 1 1.53 40 15 0.152 0.23 – 0.23
G 20 0.116 0.18 – 0.18
Case 2 1.36 50 15 0.125 0.16 0.07 0.23
F 20 0.095 0.12 0.07 0.19
Case 3-1 0.81 33 15 0.178 0.14 – 0.14
G 20 0.138 0.11 – 0.11
Case 3-2 0.94 33 15 0.178 0.17 – 0.17
G 20 0.138 0.13 – 0.13
S′- Settlement where column height is equal to soft soil
thickness
S″- Settlement where column height is less than soft soil
thickness

4.4.2 Method Considering the Arch Effect of the Embankment Material

The total residual settlement after ground improvement is calculated from the defor-
mation coefficient of the composite ground. An unequal settlement between the
improved soil and original ground (poor ground) were calculated with consideration of
arch action of embankment between each improvement column. The embankment load
is transferring between the improved bodies as shown in Fig. 6 due to the arching
effect. Soil weight below the arch acts as a load on the unimproved part between the
improved bodies.
 Embankment Construction on Power Blender Cement 441

Fig. 6. A schematic diagram of Arching action

The total residual settlement based on the arching effect (S2) is calculated using
Eqs. (7) and (8) by the composite ground which was formed by improved and original
soil.


S2 ¼ r=Eeq
H ð7Þ


Eeq ¼ ap
Ec þ 1  ap
Es ð8Þ

Where Eeq is the modulus of elasticity of composite ground, Ec is the modulus of

elasticity of improved column and Es is the modulus of elasticity of unimproved soil.
In order to obtain the amount of settlement of the original ground between each
improved column, the consolidation settlement of original ground, S0 is obtained as
described in Sect. 4.2 and the amount of settlement of original ground between
improved column, Ss is calculated by the ratio of the stress, rs acting on the unimproved
portion and the average stress, r acting on the entire improved area using Eq. (9).

Ss ¼ Ss
ðrs =rÞ
H ð9Þ

The settlement amount of the improved column, Sc was calculated by the ratio of
the stress, rc acting on the improved portion and the modulus of elasticity of improved
column, Ec by Eq. (10).

Sc ¼ ðrc =Ec Þ
H ð10Þ

The differential settlement amount, ds between the improved column and original
ground was calculated by the Eq. (11).

dc ¼ Ss  Sc ð11Þ
442 N. Vasanthan et al.

A summary of total residual settlement and differential settlement are given in

Table 5.

Table 5. Differential settlement after ground improvement

Case no. S0 ap S2 Differential
(m) (%) (m) settlement
amount
Ss Sc ds
(m) (m) (m)
Case 1 1.53 40 0.08 0.14 0.06 0.08
G
Case 2 1.36 50 0.14 0.28 0.04 0.24
F
Case 3-1 0.81 33 0.06 0.09 0.05 0.04
G
Case 3-2 0.94 33 0.07 0.08 0.06 0.02
G

5 Stability of the Embankment Slopes

In these designs the analysis was done with Geostudio-Slope/W software.

The software considers a large number of trial failure surfaces and selects the one
with minimum factor of safety. It is taken as the factor of safety of the slope.
In this analysis, the Spencer method was used which covers both circular and non-
circular failure surface. It satisfies both force and moment equilibrium conditions.

5.1 Shear Strength Parameters Adopted for the Stability Analysis

of the Embankment Slopes
Based on the result of the soil test and experience earned on this area, the shear strength
parameters of the original ground were adopted as given in Table 6.

Table 6. The shear strength parameters of the original ground

Soil type c Drained Undrained
(kN/m3) C′ /′ Cu /u
(kN/m2) (Deg) (kN/m2) (Deg)
Clay 16.0 8.0 22 20 0.0
Peat 12.0 5.0 20 5 0.0
Sand 17.0 0.0 30 – –
Embankment 19.0 10.0 30 – –
 Embankment Construction on Power Blender Cement 443

Shear strength parameters for the improved area were obtained from the design
strength of the improved ground, quck and the area improvement ratio, ap by the
Eq. (12) as it is a composite of improved and original ground and presented in Table 7.

C ¼ quck
0:5
ap ð12Þ

C is the cohesion for the improved column and the internal friction angle, / for
improved column is assumed as 40°.

Table 7. The shear strength parameters of the improved ground

Case no. Soil parameter for improved ground
ap (%) quck (kN/m2) C (kN/m2) / (deg)
Case 1 40 250 50.00 40
Case 2 50 200 50.00 40
Case 3-1 33 350 57.75 40
Case 3-2 33 400 66.00 40

5.2 Stability Analysis of the Embankment Slopes

Stability analysis of the embankment slope for case 3-2 area is presented in Fig. 7 and
the results of all cases are summarized in Table 8. The calculation result was suffi-
ciently secured against the required factor of safety of 1.2.

2.14

35

30
Elevation (m)

25

20

15

10

0
0 10 20 30 40 50 60

Distance (m)

Fig. 7. Stability analysis for Case 3–2 area

444 N. Vasanthan et al.

Table 8. The results of the stability analysis

Case no. He Q Factor of safety
(m) (kN/m2) achieved
Case 1 3.2 20 2.27
Case 2 2.5 20 2.39
Case 3-1 3.4 20 2.51
Case 3-2 5.0 10 2.14

6 Monitoring of Ground Improvement

Settlement plates and horizontal displacement stakes were installed at every 50 m

interval to monitor the settlement and horizontal movements respectively as shown in
Fig. 8.

L1
L5

R1 L2 L3 L4
R5

R2 R4
R3

Settlement plate

Displacement stake Displacement stake

Embankment fill

Fig. 8. Plan layout of settlement plates and horizontal displacement stakes and sectional view of
A-A

Some of the monitoring results for vertical settlement are given in Table 9, which
indicate horizontal displacements to be minimal.
While Point Nos. 2, 6 and 9 are located between improved areas (unimproved
portion), the remaining are on improved areas.
 Embankment Construction on Power Blender Cement 445

Table 9. The results of the vertical settlement

Case Point no. Actual settlement until 31st October 2018/(mm)
Case 3 1 31
2 34
3 28
12 11
Case 2 4 33
5 36
6 42
Case 1 7 41
8 59
9 63
10 16
11 23

7 Current Progress of CKE Bypass

Current progress of CKE bypass is shown in Fig. 9 as of October 2018.

Fig. 9. CKE bypass paving work as of October 2018

446 N. Vasanthan et al.

8 Conclusion
(1) It does not show any consolidation settlement behaviour. Typical total settlement
limits ranging between 11 mm to 63 mm.
(2) The settlement behaviour of both the columns and the soils show that there was a
significant soil arching in the embankment, and the load acting on the columns are
much higher than that acting on the soils between the columns. The measured
differential settlements between the columns and the soils are small.
(3) The measurements have proven the effectiveness of this technique in reducing
total and differential settlements. The bypass will be open to traffic in early
November 2018. The measurement is on-going at the time of writing this paper.
(4) Accordingly, the power Blender method is an efficient method for improving soil
in a short period as the required strength can be achieved within curing period, 28-
day generally.

Acknowledgement. Authors wish to express their thanks to PMU for their guidance and
valuable advice. Also, our thanks extend to Mr. Sato, Mr. Suzuki and Mr. Teruya of SMCC-
SANKEN JV who performed the planning, testing and construction.

References
Federal Highway Administration Design: Deep Mixing for Embankment and Foundation
Support, Publication No. FHWA-HRT13-046, October 2013
Ohashi H, Jayasekare DA S, Amarasena DGSR, Karunawardena A (2018) Power Blender
Cement Mixed Soft Soil Treatment for Colombo Katunayake Expressway (CKE) Bypass,
ICSBE 2018. University of Peradeniya, Kandy
 Reinforced Concrete Bridge Deck Repair
with Titanium NSM

S. L. Platt1(&), K. A. Harries2, and M. J. McCabe3

1
Architecture and Civil Engineering, University of Bath, Bath BA2 7AY, UK
[email protected]
2
Civil and Environmental Engineering, University of Pittsburgh,
Pittsburgh 15260, USA
3
United States Department of Agriculture Forest Service, Montana, USA

Abstract. The use of titanium bars as a near surface mounted (NSM) repair
method for highway bridge deck slabs is investigated in this pilot study. Four
full-scale slabs had half of their internal reinforcing steel cut resulting in an
approximately 40% loss of slab capacity; these represented ‘damaged’ slabs.
The objective of this study was to evaluate the performance of replacing these
two cut bars in order to restore the original capacity of the slab. The four slabs
were repaired using NSM titanium bars designed to restore either their capacity
or stiffness of the cut bars. One slab from each group was repaired using a full-
length straight bar while the other implemented a ‘staple’ type repair. Both
approaches effectively restored the capacity of the slabs although each had
significant impacts on the available ductility of the repaired slab. Repairs based
on restoring lost bar strength resulted in preferable behavior. Restoring lost
stiffness required four times as much titanium and resulted in over-reinforced
sections showing little ductility.

Keywords: Concrete
Corrosion
Retrofit
Bond
Ti6al-4V
Serviceability

1 Introduction

Titanium reinforcing bars for concrete rehabilitation have been proposed (Adkins and
George 2017) and demonstrated in laboratory tests (Platt and Harries 2018a and
2018b). The primary advantage and motivation for using titanium is its corrosion
resistance. An anticipated application for titanium reinforcement is near surface
mounted (NSM) reinforcement for strengthening reinforced concrete members in cases
where the advantages of titanium may be realized such as in highly corrosive envi-
ronments, non-magnetic applications, and others in which ductility or high service
temperatures may be required.
A single demonstration application is known (Higgins et al. 2015 and 2017) in
which NSM titanium ‘staples’ – straight bars having a 90° bend at either end to affect
anchorage – were used in a bridge application.
The fundamental material behaviours of titanium are similar in form to those of
steel. Titanium, like steel, exhibits an elastic behaviour to a proportional limit, a
definable yield value followed, typically, by some degree of strain hardening, and

© Springer Nature Singapore Pte Ltd. 2020

R. Dissanayake and P. Mendis (Eds.): ICSBE 2018, LNCE 44, pp. 447–457, 2020.
https://doi.org/10.1007/978-981-13-9749-3_39
448 S. L. Platt et al.

displays a great deal of ductility in most cases. This pilot study is part of a large study
investigating specific aspects of the use of 6Al-4V titanium (UNS designation R56400)
bars having ribbed deformations as a concrete reinforcing material (Platt 2018). When
compared to conventional ASTM A615 reinforcing steel, Ti6Al-4V exhibits nearly
double the yield strength and half the extensional modulus. Titanium reinforcing bars
used in this study were nominally #5 bars, having a nominal diameter and area of
15.9 mm and 200 mm2, respectively. Results of tension tests compliant with ASTM
A370-14 (including Annex A9) in all ways, other than being titanium, are shown in
Fig. 1. Nominal (for design) and measured material properties of the bars used in this
study are shown in Table 1.

Fig. 1. Representative experimentally determined stress-strain curves.

6Al-4V titanium bars exhibit yield strength approximately twice that of ASTM
A615 reinforcing steel and an extensional modulus about 55% of steel. As such, the
yield strain of Ti6Al-4V is on the order of 0.008, approximately four times greater than
A615 steel. The softer response affects the assumed concrete behaviour and can be a
significant disadvantage when considering serviceability; primarily in the form of
increased crack widths.
The behaviour of reinforcing steel, particularly under service loads, is a function of
axial stiffness, EA; thus to directly replace steel with titanium, twice as much titanium is
required. Typically, this will not be practical. Alternatively, the design paradigm used
for glass fibre-reinforced polymer (GFRP) reinforcing bars [ACI 440.1R], which pre-
sents an elastic behaviour to failure with modulus ranging from 20% to 50% that of
steel, may be used for titanium-reinforced concrete or for using titanium as a retrofit
reinforcement.
 Reinforced Concrete Bridge Deck Repair with Titanium NSM 449

Table 1. Geometric and mechanical properties of reinforcing bars reported in this study
#5 6Al 4V Titanium bars # 5 ASTM A615 Steel bars
Nominal Measured Nominal Measured
properties properties properties properties
(Platt and Harries (McCabe et al.
2018a) 2014)
2
Bar area 200 mm 236 mm2 200 mm 2

Bar diameter 15.9 mm 17.4 mm 15.9 mm

Density 4430 kg/m3 4407 kg/m3 7900 kg/m3
Modulus ETi = 114 GPa Es = 200 GPa
Yield strength fyTi = 965 MPa fyTi = 999 MPa fys = 414 MPa fyTi = 467 MPa
Tensile fuTi = 1100 MPa fuTi = 1054 MPa fus > 586 MPa fuTi = 744 MPa
strength
Elongation at eTiu > 0.100 eTiu > 0.084 esu > 0.150
rupture

2 NSM Titanium Slab Tests

Four steel-reinforced bridge deck slabs, cast in 2013, were available to develop a pilot
study to demonstrate the efficacy of titanium NSM retrofit. These slabs were designed
based on the AASHTO (2017) prescriptive design method and had four #5 primary
reinforcing bars spaced at 140 mm, top and bottom, across the 559 mm slab width. The
‘control’ specimen used for this study is Slab A, tested in 2013 (McCabe 2013;
McCabe et al. 2014). Details of the slab geometry are shown in Fig. 2.

Fig. 2. Details of laboratory control specimen (McCabe 2013) (1 in. = 25 mm)

450 S. L. Platt et al.

Each of the four slabs to be retrofit was ‘damaged’ by cutting two of the four
existing #5 steel bars at increments along their length sufficient to ensure that the bars
no longer contributed to the flexural capacity of the slabs. The slabs were then repaired
with one of two types of NSM repairs: a straight NSM bar or an NSM ‘staple’. The
staples, as shown in Fig. 3, were short lengths of NSM bar having 90° anchorages at
either end. Such a staple may be practical for repairing local damage (as was the case in
this study) or in locations were straight bar development is not possible.
Two scenarios are investigated based on repairs that restore either the capacity
(Asfys = ATifyTi) or equivalent stiffness (AsEs = ATiETi) of the cut bars. In either case, #5
titanium deformed bars were embedded within channels cut along the length of the
tension face of the slabs. Sikadur 31 (a commercially available two-part structural
adhesive) was used to embed the bars. The installations, shown in Fig. 3, were, other
than the use of titanium, compliant with the design approach of ACI 440.2R-17.
After allowing the NSM installations to cure for a minimum of 72 h, they were
placed in a test frame. Loading and support conditions, shown in Fig. 4a, were identical
to those used to test control Slab A (McCabe 2013). The slabs were tested in mid-point
flexure over a simple span length (L) of 2135 mm. Primary instrumentation consisted
of multiple DEMEC gauges, having a resolution of 8 microstrain and gauge length of 8
in. (203.2 mm), arranged vertically at mid-span (Fig. 4).
Tests of the four NSM titanium retrofit slabs were carried out in a manner (to the
extent possible) identical to the control specimen: Slab A reported previously by
McCabe et al. (2014).
Load was applied using a 267 kN capacity hydraulic cylinder; load was measured
with a precision of 320 N. Load was applied at intervals of approximately 4.45 kN
while mid-span displacement was recorded manually at each load interval with a
precision of 0.8 mm.
Since there were no test cylinders remaining from the original casting of the slabs
and due to undocumented environmental conditions during storage, core samples were
taken from the slabs to confirm present in situ compressive and tensile strengths. Cores
having a diameter of 57 mm (2.25 in.) were removed from the [essentially undamaged]
support region of the slabs following testing. The compression and tension strengths
obtained from the 57 mm cores (including all corrections) are shown in Table 2. An
expected increase in compressive strength is observed. The 1666 day tested strengths
reported in Table 2 were used for all subsequent analyses of the slabs. The 132-day
strength was used for Slab A. Internal reinforcing steel and NSM titanium material
properties are reported in Table 1.
 Reinforced Concrete Bridge Deck Repair with Titanium NSM 451

TiNSM-1 and TiNSM-2 elevation with NSM titanium straight bars (units = mm)

NSM channels on [inverted] slab soffit; transverse cuts through internal reinforcing
bars can also be seen.

TiNSM-2 section: 4 bars at 83 mm

TiNSM-1 section : 1 bar centered in slab soffit
across soffit

TiNSM-3 and TiNSM-4 elevation with NSM titanium staples (units = mm)

TiNSM-3 section – 1 staple centered in slab TiNSM-4 section – 4 staples at 83 mm

soffit across soffit

TiNSM4 staples before epoxy embed-

Staple, channel and installation tools ment

Fig. 3. Repaired cross sections and elevations of slabs based on strength (TiNSM-1 and 3) and
Service (TiNSM-2 and 4) criteria (all dimensions in mm).
452 S. L. Platt et al.

267 kN hydraulic ram

ball joint
spreader beam
50 mm bearing pad
5 - 203 mm DEMEC gauges
(both sides)

midspan deflection
2135 mm simple span (Laboratory specimens)
test Set-up (Slab A shown)

DEMEC and vertical deflection instrumentation (Slab A shown)

Fig. 4. Test set-up and instrumentation

Table 2. Summary of concrete properties

Age ASTM C39 ASTM C496 ASTM C78
Compression test Split cylinder test Modulus of rupture
Days n fc0 COV n fsp COV n fr COV
pffiffiffiffi pffiffiffiffi
28a 3 44.8 MPa 0.034 3 3.12 MPa = 0:47 fc0 0.131 3 5.45 MPa = 0:81 fc0 0.056
132a 3 45.8 MPa 0.084 – – – – – –
pffiffiffiffi
1666b 8 50.7 MPa 0.081 5 2.98 MPa = 0:42 fc0 0.186 – – –
a b
McCabe (2013) Platt (2018)

3 Test Results

The key results for NSM-repaired slabs and the control Slab A are presented in
Table 3. Plots of applied moment versus curvature at mid-span are shown in Fig. 5.
The reported moments are calculated as: M = PL/4, were P includes the cross-head
contribution but neglects the weight of the slab. The curvature is calculated by dividing
the difference in strain between the uppermost DEMEC gauge and the gauge located at
the existing A615 steel tension reinforcement by the vertical distance separating these
gauges. This is consistent with the procedure used for the control Slab A.
The behavior of all slabs was analyzed using the program RESPONSE (Bentz
2000). All material properties used in the RESPONSE models are the measured values
 Reinforced Concrete Bridge Deck Repair with Titanium NSM 453

Table 3. Summary of tested slabs

Slab A TiNSM-1 TiNSM-2 TiNSM-3 TiNSM-4
Depth of slab mm 191 216 191 191 191
Design condition Strength: Stiffness: Strength: Stiffness:
Asfys+ ATifyTi  AsEs+ ATiETi  Asfys+ ATifyTi  AsEs+ ATiETi
Slab A Slab A Slab A  Slab A
NSM bars None 1–#5 straight 4–#5 straight 1–#5 4–#5
bar bars staple staples
Load at first crack kN 22.6 23.1 28.2 28.2 32.6
Moment at first crack kN-m 12.1 22.1 15.0 15.0 17.4
Load at steel bar yield kN 81.0 81.8 104 72.5 113
Moment at steel bar kN-m 43.1 43.6 55.3 38.7 60.1
yield
Ratio yield capacity to – – 1.01 1.28 0.90 1.39
Slab A 0.89a
Deflection at steel bar mm 9.65 5.59 6.35 5.59 11.2
yield
Curvature at steel bar rad/km 25.6 16.6 21.2 18.4 14.4
yield
Ultimate load kN 125 179 192 117 130
ultimate kN-m 66.6 95.7 103 62.5 69.6
moment
Ratio ultimate capacity – – 1.44 1.55 0.94 1.05
to Slab A 1.12a
Deflection at ultimate mm – 32.5 15.0 25.4 31.0
load
Failure mode Flexural Flexural Shear Flexural Flexural at end
of staples
a
value normalized to 191 mm slab depth; i.e.: Slab TiNSM-1 ratio multiplied by (191/216)2

Fig. 5. Summary of Moment-Curvature results from slab flexure tests

454 S. L. Platt et al.

given in Tables 1 and 2. Figure 6 shows the predicted moment-curvature responses

superimposed with the as tested results.
The predicted moment-curvature responses are for mid-span performance; effects of
shear-moment interaction along the relatively short shear spans will result in a marginal
reduction of the predicted “pure” moment capacity. In the more heavily reinforced
Slabs TiNSM2 and 4, the retrofit flexural capacity exceeds the slab shear capacity and
the latter controls the ultimate behavior. RESPONSE-predicted capacities are shown in
Table 4.

Table 4. Summary of predicted capacities of NSM-reinforced slabs

Slab Slab A with TiNSM-1 TiNSM-3 TiNSM-2 TiNSM-4
A two bars cut
Predicted moment 57.3 34.3 84.1 65.6 134
capacity (kNm)
Predicted peak applied 94.8 52.4 137 100 154
load (kN)
moment capacity 50.5 27.9 72.8 53.5 82.3
accounting for shear
(kNm)
Observed moment 66.6 – 84.6a 62.5 103 69.6
capacity (kNm)
Capacity normalized to 1.0 – 1.27 0.94 1.55 1.05
Slab A
a
value normalized to 191 mm slab depth; i.e.: Slab TiNSM-1 ratio multiplied by (191/216)2

Fig. 6. Summary of moment-curvature results with RESPONSE predicted results

 Reinforced Concrete Bridge Deck Repair with Titanium NSM 455

The proposed staple repair was investigated on two slabs having two of the four #5
ASTM A615 reinforcing bars cut only at mid-span (Fig. 4). This effectively dropped
the capacity of the slab approximately 40% (Table 4). The development length for a
#5 bar is ld = 340 mm. Thus not until 340 mm to either side of mid-span, is the
capacity of the cut #5 bars theoretically redeveloped. The staples used were only
406 mm long and thus did not span the entire region of reduced capacity (680 mm).
Figure 7 describes the resulting moment capacity envelopes (normalized to the pre-
dicted capacity of Slab A). The applied moments at failure are shown by the dashed
lines. Clearly Slab A is critical at mid-span, as is TiNSM-3, although there is a rela-
tively small margin at the end of staple. The applied moment of TiNSM-4 is seen to
intersect its capacity envelope at the end of the staples, rather than at the increased mid-
span capacity. Thus, the staples in this case were too short to develop the predicted
capacity of the repaired slab, as reflected in Table 3.
To mitigate the potential weak section at the staple anchorage, the anchorage
should be located beyond the point where the cut bars have been fully redeveloped.
That is, the staple length should exceed 2ld (of the cut/damaged bars).

Fig. 7. Generalized capacity envelope of the damaged slab with and without staple repair

4 Conclusion

The application of titanium NSM reinforcement was investigated through a pilot study
using four intentionally damaged steel-reinforced bridge deck slabs. Both straight bar
repairs and the staple repair based on equivalent stiffness (4 staples) restored the
original slab capacity. The repair having the single titanium staple, theoretically
456 S. L. Platt et al.

sufficient to restore the capacity lost by cutting the internal reinforcement, restored only
94% of the original slab capacity. It is hypothesized that if the staples had been
designed to extend beyond the development length of the cut internal steel reinforcing
bars, the resulting performance of the NSM repairs would have been improved.
The slabs having only a single titanium bar, developed ductility similar to the
undamaged control Slab A although exhibited significant slip at ultimate capacity. The
slabs having four bars were both effectively over-reinforced leading to increased
capacity but decreased ductility. The slab having four full-length titanium NSM bars,
failed upon reaching its shear capacity whereas that having four staples failed ‘pre-
maturely’ in flexure at the section at the end of the staples. The staples were not long
enough to develop the full theoretical capacity of this slab.
Cost of Titanium Bars
Currently the associated costs of implementing titanium as a reinforcement material
must be overcome. On a unit weight basis, 6Al-4V titanium bars are approximately 15
times more expensive than A615 steel bars and 5 times more expensive than solid
stainless steel bars (Platt 2018; Triantafillou 2012). However, the final in-place cost of
titanium bars will be affected by the design paradigm. Platt and Harries (2019) report a
preliminary study of the cost of using #5 6Al-4V titanium bars for bridge deck rein-
forcement (based on an earlier study by Triantafillou (2012) that considered several
other bar materials). Depending on the design paradigm used, 6Al-4V reinforcing bars
were estimated to cost between $430/m2 [of bridge deck constructed] and $1700/m2.
A615 bars were estimated to cost $117/m2 and solid stainless steel bars were $294/m2.
The typical bridge construction unit cost for medium span bridges using conventional
A615 reinforcing bars (in 2010) was $989/m2 (Triantafillou 2012). Therefore, the use
of titanium bars represents a significant cost premium although this may be offset by
long term savings in maintenance for a non-corrosive deck.

Acknowledgement. The research presented was sponsored by the Perryman Company. The
findings and conclusions reported in this paper reflect the opinion of the authors and are not
necessarily those of the Perryman Co. Slab construction and testing Slab A was sponsored by the
Pennsylvania Department of Transportation. All testing was conducted in the Watkins Haggart
Structural Engineering Laboratory at the University of Pittsburgh. The second author acknowl-
edges the support of the Leverhulme Trust (UK).

References
AASHTO (2014) LRFD bridge design specifications (7th edn), American Association of State
Highway and Transportation Officials
ACI 440.2R-17 Design and Construction of Externally Bonded FRP Systems, American
Concrete Institute
ACI 440.1R-15 (2015) Guide for the Design and Construction of Structural Concrete Reinforced
with FRP Bars, American Concrete Institute
Adkins J, George W (2017) Titanium finds a home in civil engineering. In: Concrete
International, December 2017, pp 51–55
 Reinforced Concrete Bridge Deck Repair with Titanium NSM 457

ASTM A370-14 (2014) Test Methods and Definitions for Mechanical Testing of Steel Products.
ASTM International
ASTM A615-16 (2016) Standard Specification for Deformed and Plain Carbon-Steel Bars for
Concrete Reinforcement. ASTM International
ASTM C39-16 (2016) Standard Test Method for Compressive Strength of Cylindrical Concrete
Specimens, ASTM International
ASTM C496-11 (2011) Standard Test Method for Splitting Tensile Strength of Cylindrical
Concrete Specimens, ASTM International
Bentz E (2000) Sectional Analysis of Reinforced Concrete Members, PhD Thesis, Department of
Civil Engineering, University of Toronto 2000, pp 310
Higgins C, Knudtsen J, Amneus D, Barker L (2017) Shear and flexural strengthening of
reinforced concrete beams with titanium alloy bar. In: Proceedings of the 2nd World Congress
on Civil, Structural, and Environmental Engineering (CSEE 2017), Barcelona, Spain, April
2017
Higgins C, Amneus D, Barker L (2015) Methods for Strengthening Reinforced Concrete Bridge
Girders Containing Poorly Detailed Flexural Steel Using Near-Surface Mounted Metallics,
Report No. FHWA-OR-RD-16-02, Oregon Department of Transportation, Salem, OR, pp 138
McCabe M, Harries KA, Sweriduk M (2014) Evaluation of concrete bridge deck rehabilitation by
the method of hydrodemolition and latex-modified overlay. In: Proceedings of 15th
International Conference on Structural Faults and Repair, London
McCabe M (2013) Structural Evaluation of Slab Rehabilitation by Method of Hydrodemolition
and Latex Modified Overlay. (Master’s thesis), University of Pittsburgh, PA
Platt S, Harries KA (2019, in revision) Proposed flexural design methodology for titanium
reinforcing bars in concrete. Engineering Structures
Platt S (2018a) Development of Titanium Reinforcing Bars for Concrete and Masonry, Doctoral
dissertation, University of Pittsburgh
Platt S, Harries KA (2018a) Geometry, Material Properties and Bond Performance of Titanium
Reinforcing Bars. Constr Build Mater 187:1253–1266
Platt S, Harries KA (2018b) Study of Galvanic Corrosion Potential of NSM Titanium
Reinforcing Bars. Case Stud Constr Mater 9:e00175
Triantafillou L (2012) Cost Comparison of Corrosion Resistant Reinforcing Steel Deployment
Considerations. FHWA Office of Infrastructure Research and Development as updated by
Wong W (2014)
 Physical Behavior of CFRP Retrofitted
Reinforced Concrete Slab-Column
Connections

R. H. M. Dissanayaka, M. A. L. Silva(&), L. P. G. Magallagoda,

and J. C. P. H. Gamage

Department of Civil Engineering, University of Moratuwa, Moratuwa, Sri Lanka

[email protected]

Abstract. Carbon Fiber Reinforced Polymer (CFRP) retrofitted application has

become widespread in the last decade due to its unique advantages. This paper
presents the punching shear strength enhancement of flat slabs which can be
achieved with CFRP by embedding it through the slab section and attaching it
externally on the tension surface. A common anchorage system that would
optimise the use of CFRP has been introduced in CFRP shear reinforcement.
The maximum capacity increment in punching shear was 22% and further
improvements to achieve even higher load capacities have been discussed.
Theoretical expressions have also been compared with the experimental results
with recommendations.

Keywords: CFRP
Flat slabs
Punching shear
Reinforced concrete

Internal strengthening
External strengthening

1 Introduction

Strengthening of existing flat slabs might become crucial due to the aging of the
structures, changes in the use of the building, exposure to heavy loads, poor detailing at
the design stage, early loading of the structure, deterioration and corrosion of rebar as
flat slabs are all reasons for punching shear failure.
The key factors which affect the punching shear strength of flat slabs are com-
pressive strength of concrete, ratio of longitudinal reinforcements, dimension of the
column and thickness of the slab. Therefore, some of the conventional strengthening
methods like concrete jacketing of columns to increase the loading area on the slab,
inserting longitudinal steel reinforcements by casting a new concrete layer or intro-
ducing Fiber Reinforced Polymer (FRP) and insertion of steel shear reinforcement [3]
or FRP [2] have become popular solutions for enhancing punching shear capacity.
Carbon Fiber Reinforced Polymer (CFRP) has become one of the most used ret-
rofitting materials in the construction industry due to the unique combination of
characteristics like high tensile strength and low self-weight - thus, a higher strength to
weight ratio and ease of installation. However, limited literature is available on the
effective use of CFRP to enhance the punching shear capacity.

© Springer Nature Singapore Pte Ltd. 2020

R. Dissanayake and P. Mendis (Eds.): ICSBE 2018, LNCE 44, pp. 458–469, 2020.
https://doi.org/10.1007/978-981-13-9749-3_40
 Physical Behavior of CFRP Retrofitted Reinforced Concrete 459

In recent past, as more attention was given for the use of CFRP materials to
enhance the structural performance of strength deficient members [6, 7], different
methods of retrofitting using CFRP were found to gain strength in flat slabs. Since
punching shear behavior is a type of shear failure which is critical due to its tendency
for sudden failure, it has been found that the application of shear reinforcements has the
most effective resistance capacity against the shear action. The behavior of flat slab
against punching shear was studied in literature using CFRP fabric dowels as shear
reinforcement embedded through the slab section near the column where punching
shear is critical [1]. It was found out that when the number of shear reinforcement rows
around the column increases, the punching shear capacity increases and the mid span
deflection at failure decreases. Further, strengthening of flat slabs using FRP rods, bolts
and screws as shear reinforcements have been performed [2, 12]. Debonding of FRP
rods was observed during the study indicating the insufficiency of area contact to the
FRP rod for a proper stress transfer from FRP rod to concrete. Further, the steel screws
with proper anchorage provided using tightened nuts gave a higher punching capacity
increment than that of given by FRP rods because, FRP rods were not provided with
end anchorage at both ends, while steel screws were mechanically anchored with nuts,
indicating the importance of proper end anchorage of shear reinforcement.
In this study, taking into account, all the aforementioned drawbacks in previous
studies, CFRP fabrics were used in the form of strips to retrofit two alternative
arrangements of strengthening by considering the effect of the location of vertical shear
reinforcements in terms of the punching shear capacity, mid span deflection and crack
patterns. Further, this study proposes a common anchorage system for two separate
vertical CFRP shear reinforcements in order to optimize the use of CFRP fabrics.

2 Experimental Details

2.1 Specimen Preparation and Material Properties

Eight slab-column connections were cast with the slab dimensions of 1200 mm 
1200 mm  100 mm and the stub column dimensions of 100 mm  100 mm 
150 mm. Reinforcement detailing of each aforementioned specimen is shown in Fig. 1.
Concrete and steel material properties are listed in Table 1.

2.2 Strengthening Schemes

Two specimens were kept as control specimens without retrofitting using CFRP and
other six were strengthened with CFRP fabric strip widths of 25 mm by embedding
them through the slab section and attaching them on to the tension surface of slab as
shown in Fig. 2. The strengthening schemes consisted of three different alternative
arrangements including two internal embedding of CFRP and one external attachment
of CFRP.
In Figs. 2(a) and (b), the two arrangements of CFRP embedment are shown and in
each of them, two rows of shear reinforcement were inserted. As shown in the sectional
view, two consecutive shear reinforcements of a considering radial section were made
460 R. H. M. Dissanayaka et al.

Fig. 1. Steel reinforcement details (a) plan view (b) cross sectional view

Table 1. Material properties of concrete and steel

Properties of concrete Value
Average compressive strength (N/mm2) 28.3
water/cement ratio 0.59
Maximum aggregate size (mm) 20
Properties of steel Value
Yield strength (N/mm2) 500

of single CFRP strip of 25 mm with separate anchorage provided at the compression

face and a common anchorage provided at the tension face. The compression face
anchorage length was 200 mm and the tension face common anchorage lengths were
different depending on the spacing between two rows of shear reinforcement. The first
row of shear embedding was placed at an effective depth (d) away from the column
face, which was 70 mm. The second row of shear reinforcements in the type RF-180
and type RF-250 was 110 mm and 180 mm away from the first row of shear rein-
forcements, respectively. This means approximately, 1.5 d and 2.5 d away from first
row of shear reinforcement.
As specified in ACI 318 and Eurocode 2 standards, the critical punching shear
perimeter formation of flats slabs is a rectangular perimeter 2, 1.5 and 0.5 times the
effective depth away from the column face respectively. Therefore, in both arrange-
ments, CFRP embedment was in the shear critical zone as specified in ACI 318, BS
8110 and Eurocode 2. The arrangement in Fig. 2(c) shows the combination of
orthogonal and skewed attachment of CFRP on the tension face of flat slab to identify
the contribution of enhancing punching shear capacity.
The surface areas where CFRP sheets needed to be bonded were ground until the
low strength cement grout layer got removed, because it is important to maintain a
proper bond between CFRP sheet and the concrete substrate.
 Physical Behavior of CFRP Retrofitted Reinforced Concrete 461

(i)

(ii)

(a) Specimen Type RF-180

(i)

(ii)

(b) Specimen Type RF-250

(i)

(ii)

(c) Specimen Type RF-EX

Fig. 2. CFRP applied strengthening scheme (i) plan view (ii) sectional view

The primer was applied on the ground and cleaned concrete surface in order to
provide a proper bond between the adhesive and the dry concrete interface. The
application of epoxy saturated CFRP was done followed by the application of epoxy
adhesive on primer applied concrete surface. The two ends of each CFRP strip were
inserted through the drilled holes and attached on the compression surface of the slab.
Afterwards, the drilled holes were refilled using a concrete repairing grout ensuring the
force transfer between the shear reinforcement and the concrete slab. The properties of
CFRP fabric and adhesive are mentioned in Tables 2 and 3.
462 R. H. M. Dissanayaka et al.

Table 2. Material properties of concrete and steel

Properties of CFRP fabrics [5] Value [4] Properties of CFRP fabric Adhesive Value
Thickness (mm) 0.166 Tensile strength (MPa) 45
Modulus of elasticity (GPa) 240 Flexural strength (MPa) 60
Yield strength (MPa) 4000 Compressive modulus (GPa) 1.67
Elongation at break 2%

Table 3. Test Results

Specimen Failure Percentage increment Deflection at Percentage increment of
type load (KN) of load (%) failure (mm) deflection (%)
C1 103 – 13.8 –
C2 98.1 – 11.2 –
RF-180-1 112.815 12.2 17.5 40
RF-180-2 112.815 12.2 15.42 23.36
RF-250-1 98.2 −2.34 18.9 51.2
RF-250-2 107.91 7.2 18.16 45.28
RF-EX-1 117.72 17.08 12.05 −3.6
RF-EX-2 127.53 26.83 10.71 −14.32

2.3 Testing of Specimens

Specimens were tested after the curing period, providing simply supported conditions
at four edges of each specimen using steel I sections. A point load was applied on the
middle stub column using the hydraulic jack at a rate of 10 kN/min. A steel cap with
rubber was used to support the column stub to distribute the load monolithically to the
column from the hydraulic jack ensuring the column would not fail prior to the slab.
Test setup and instrumentation done, are shown in Fig. 3. The mid span deflection
corresponding to the applied load was measured using three dial gauges fixed 75 mm
away from the column face while the point load was applied. Crack propagation at the
tension and compression surfaces of the specimens were examined throughout the
loading procedure and the failure load was recorded.

Fig. 3. Instrumentation and test set-up

 Physical Behavior of CFRP Retrofitted Reinforced Concrete 463

3 Test Results and Discussion

3.1 Failure Mechanisms
Six flat slab specimens were failed in sudden punching shear failure mode while two
failed in flexural failure mode (Fig. 4). The punching shear crack distribution was a
conical shape which initiated from the tension surface and ended at compression
surface near the column face. In all specimens flexural cracks started appearing before
punching shear failure happened.

(a) (b)

(c) (d)

Fig. 4. Crack patterns on tension surface of tested specimens (a) control specimens (b) type RF-
180 (c) type RF-250 (d) type Ex-RF

Figure 4(a) shows the crack pattern on the tension surface of the control sepeci-
mens. Flexural cracks started appearing around the load application of 60 kN and the
specimen failed due to punching shear failure, forming a punching shear crack of one to
four times the effective depth away from the column face.
Punching shear cracks were between two to five times the effective depth away
from the column face in specimen type RF-180 (Fig. 4(b)). In some places, the
punching shear crack on tension surface passed through the outer row of CFRP shear
reinforcement. The width of the flexural cracks was around 0.3 mm.
Punching shear failure was observed after the appearance of a set of flexural cracks
on the tension surface of the type of specimens RF-250 as shown in Fig. 4(c). The
punching shear crack formation was in between three to four times the effective depth
away from the column perimeter and therefore, the crack passed through the outer row
of CFRP shear reinforcement. The number of flexural cracks compared to the type
464 R. H. M. Dissanayaka et al.

RF-180 was higher in type RF-250 and large crack widths were observed under the
column. The type RF-EX only showed flexural cracks as shown below in Fig. 4(d).
Figure 5 shows the compression surface of RF-250-1 slab, after the collapse. It
shows that the crack has passed through the holes. The area has become weak due to
the damage occurred during drilling. Truncated cone at the punching shear failure was
occurred by pushing the concrete section from this crack.

Fig. 5. Crack patterns on compression surface of tested specimens

3.2 Load Versus Deflection

The load increment of the type RF-180 was 12.2% and the average deflection incre-
ment at the failure was 31.6% compared to the control panels. Due to the strengthening
of slabs with the CFRP sheets, the load carrying capacity and the deflection of the slab
has increased. The deflection increment rate of the type RF-180 seemed to be increased
than that of control panels as shown in Fig. 6 indicating a reduction in stiffness. This is
contradictory to the literature available related to CFRP because, with the application of
CFRP, the structure stiffness should increase. The reason behind this is either the
damage done in the slab to install CFRP shear reinforcement might have created micro-
cracks resulting in stiffness reduction or the refilling material might not have transferred
the stresses from slab to CFRP reinforcement properly.

Fig. 6. Load versus mid span deflection of specimens

 Physical Behavior of CFRP Retrofitted Reinforced Concrete 465

The type RF-250 has a punching shear capacity increment of 2.5%. The deflection
at failure has been increased 48%, compared to the control panels. The deflection
increment rates were increased as observed in type RF-180 with respect to the control
panels. Due to the retrofitting of the slab, the load carrying capacity and the stiffness of
both RF-180 and RF-250 types has been reduced. The reason might be the inadequacy
of refilling material of drilled holes to transfer the forces from slab to CFRP shear
reinforcement or the damage done to the concrete slab while drilling must have become
dominant.
RF-EX panels have achieved the maximum average load bearing capacity of 22%
and the deflection at failure was reduced compared to that of control specimens. This
shows, the attaching of CFRP at the tension surface has improved punching shear
capacity as well as the stiffness of the structure without any damage to the structure.

4 Theoretical Analysis and Comparison

Theoretical analysis was conducted to evaluate the applicability of available analytical

models in predicting the strength gain of flat slabs retrofitted with CFRP.

4.1 ACI 318

The punching shear strength of none shear strengthened normal weight concrete can be
calculated theoretically using the equations of ACI 318 by ACI Committee [8] given
below. The minimum of the following three Qup values denotes the punching shear
capacity.
qffiffiffiffi
1 0
Qup ¼  f c b0 d ð1Þ
3
qffiffiffiffi
  fc
0
1
Qup ¼ 2þ  b0 d ð2Þ
bc 12
qffiffiffiffi
  fc
0
as d
Qup ¼ 2þ  b0 d ð3Þ
b0 12

Where, Qup is the punching shear capacity; b0 is the critical section perimeter 0.5 times
the effective depth away from the face of loading area; d is the effective depth of the
slab, as is a constant of 40 for interior columns, 30 for edge columns and 20 for corner
0
columns; bc is the ratio of long side to the short side of the loading area; f c is the
concrete cylinder compressive strength. Here, all there equations have been derived not
considering the effect from flexural reinforcement ratio.
For slabs strengthened against punching shear, the above equations get modified as
below by Inacio [11].
466 R. H. M. Dissanayaka et al.

qffiffiffiffi
1 0 Asw f yt d
Qup ¼ 0:5   f c b0 d þ ð4Þ
3 sr

Here, Asw is the cross sectional area of shear reinforcement f yt is the tensile strength
of shear reinforcement and sr is the distance between two consecutive rows of shear
reinforcement.

4.2 Eurocode2-2004
The punching shear capacity of none shear strengthened flat slabs has been calculated
using Eurocode 2 by British Standard Institution [9] considering the effect of flexural
reinforcement and not considering the aspect ratio of loading area as follows.
qffiffiffiffiffiffiffiffiffiffiffiffiffiffi
0
Qup ¼ 0:18K ð5Þ
3
100qf c b0 d

Where, K = 1 + (200⁄d)1/2 and should not be taken more than 2; q is the flexural
reinforcement ratio and it is limited to 0.02; b0 is the perimeter of the critical section
with round corners at a distance of 2 times the effective depth away from the surface of
loading area.
The punching shear strength of retrofitted slabs can be obtained by the equation that
has been modified using the simplified critical shear crack theory (CSCT), proposed by
Faria [10] considering the concept of equivalent effective depth, dv .

ds Asl þ df Aft Ef =Es

dv ¼ ð6Þ
Asl þ Aft Ef =Es

And also by using equivalent reinforcement ratio qeql that can be obtain by using,

qeql ¼ qsl þ Ef =Es qfl ð7Þ

The modified punching shear evaluation equation for retrofitted flat slat slabs is as
follows by Inacio et al., (11).
qffiffiffiffiffiffiffiffiffiffiffiffiffiffi
0 Asw f yt d
¼ 0:75  0:18K 100qf c b0 d þ 1:5  ð8Þ
3
Qup
sr

As shown in Table 4, the non-shear strengthened specimen punching shear capacity

prediction by Eurocode 2 is more accurate than ACI 318 prediction. The reason is that
the amount of flexural reinforcement has an influence on enhancing the punching shear
capacity and the ACI code predicts an underestimated capacity for punching shear.
When it comes to the comparison of test results with theoretical results in CFRP
shear strengthened specimens, ACI 318 code predicts more accurate results than
Eurocode 2. The contribution of concrete to resist punching shear was estimated as half
the contribution when no shear reinforcement exists and the contribution from the shear
reinforcement has been taken 100% to enhance the punching shear capacity.
 Physical Behavior of CFRP Retrofitted Reinforced Concrete 467

Table 4. Theoretical versus Experimental Results

Experimental ACI 318-14 Eurocode
punching shear Code 2-2004
load (kN) Code predicted Pexp/PACI Code predicted Pexp/PEC2
[Pexp] load (kN) load (kN)
[PACI] [PEC2]
Type C 100.6 83.96 0.83 95.3 0.95
Type RF-180 112.8 126.5 1.12 147.5 1.31
Type RF-250 103.1 93.6 0.91 117.95 1.14
Type RF-EX 122.63 85.16 1.44 98.3 1.25

Eurocode 2 has overestimated the punching shear capacity. Therefore, it is rec-

ommended to reduce the safety factors in the equations.
However, CFRP flexural reinforced specimen experimental results are more com-
patible with Eurocode 2 prediction values which have been modified using Muttoni
equivalent reinforcement ratio and effective depth. The reason is ACI 318 code does
not provide sufficient provisions for the amount of flexural reinforcement ratio and
therefore, it could not predict accurate results considering the flexural reinforcement
ratio increment change.

5 Conclusions

CFRP strengthened flat slab behavior is yet a study area which needs to be explored by
performing research studies and analysing them using modified theoretical expressions.
Therefore, in this study, the behavior of post installation of CFRP as shear rein-
forcements was monitored. By observing the ultimate punching shear capacities, the
deflection at failure, crack patterns and failure modes, the following conclusions can be
made:
• Punching shear crack patterns of non-strengthened flats slabs indicates the punching
shear critical perimeter should be approximately at two times the effective depth
away from the column face because, the punching shear crack was two to four times
the effective depth away from the column face. This observation comples with the
Eurocode 2 predictions.
• The maximum punching shear capacity observed in post-installed shear reinforce-
ment strengthening scheme was 12.2% with respect to the non-strengthened panels.
• The type RF-180 obtained a higher punching shear capacity because; the
arrangement had two rows of shear reinforcements at the shear critical area while
type RF-250 had only one. Therefore, the number of rows of shear reinforcement at
the shear critical area is important.
• The punching crack pattern on RF-180 denotes the critical punching shear perimeter
location should be closer to 2.5 the effective depth away from column face. This
indicates, the application of shear reinforcements in the shear critical area would
distribute the shear stress further away from column face.
468 R. H. M. Dissanayaka et al.

• Load versus mid span deflection curves indicates stiffness reduction in CFRP shear
strengthened slabs. The reasons must be the damage done while drilling the slab to
install CFRP shear reinforcements has become server.
• The load carrying capacity and the stiffness has been increased in the specimens
where, CFRP was applied externally. The load carrying capacity increment was
22% in average.
• Theoretical predictions calculated using Eurocode 2 and ACI 318 predicted com-
patible results with the experimental results.
• The punching shear capacity prediction of ACI 318 is more accurate for CFRP shear
strengthened specimens because, the proportions provided by concrete and shear
reinforcements to resist punching shear are accurate than that of in Eurocode 2.
• Eurocode 2 accounts for the amount of flexural reinforcement ratio in the punching
shear capacity. Therefore, it predicts more accurate results when the amount of
flexural reinforcement ratio changes than that of ACI 318, using Muttoni modified
equivalent effective depth and equivalent flexural reinforcement ratio.

Acknowledgments. Authors would like to extend their gratitude for the technical officers, Mr.
D.M.N.L.Dissanayake and Mr. Roshan at Structural Testing Laboratory and Building Material
Laboratory; and staff at Structural Testing Laboratory, Department of Civil Engineering,
University of Moratuwa.

References
1. Meisami MH, Mostofinejad D, Nakamura H (2014) Strengthening of flat slabs with FRP fan
for punching shear. Compos Struct. doi.org/10.1016/j.compstruct.2014.08.041
2. Meisami MH, Mostofinejad D, Nakamura H (2013) Punching shear strengthening of two-
way flat slabs using CFRP rods. Compos Struct 99:112–122. https://doi.org/10.1016/j.
compstruct.2012.11.028
3. Askar HS (2015) Usage of prestressed vertical bolts for retrofitting. Alexandria Eng J
4. X-wrap lamination adhesive (2017). http://www.x-calibur.us/files/X-Wrap%20Lamination%
20Adhesive.pdf. Accessed 21 Jan 2018
5. X-wrap c300 (2017). http://www.x-calibur.us/files/X-Wrap%20C300.pdf. Accessed 21 Jan
2018
6. Wijerathna DMN, Gamage JCPH (2013) Investigation on Effective Use of CFRP laminates
for Flexural Performance, In: 4th International Conference on Structural Engineering and
Construction Management (ICSCM2013), Kandy, Sri Lanka
7. Wijerathna DMN, Gamage JCPH (2014) A review on Alternative Bonding Techniques to
Delay End debonding of CFRP/Concrete composites. In: 5th International Conference on
Sustainable Built Environment (ICSBE2014), Kandy, Sri Lanka, pp 144–150
8. ACI Committee 318. (1999). Building code requirements for structural concrete: (ACI 318);
and commentary (ACI 318R). Farmington Hills, Mich, American Concrete Institute
9. British Standards Institution (2004) Eurocode 2: design of concrete structures: British
standard. London, BSi
 Physical Behavior of CFRP Retrofitted Reinforced Concrete 469

10. Faria DMV, Einpaul J, Ramos AP, Fernández Ruiz M, Muttoni A (2014) On the efficiency
of flat slabs strengthening against punching using externally bonded fibre reinforced
polymers. Construction and Building Materials, Netherlands, vol 73, pp 366–377. https://
doi.org/10.1016/j.conbuildmat.2014.09.084
11. Inácio M, Almeida A, Faria D, Lúcio V, Ramos A (2015) Punching of high strength concrete
flat slabs without shear reinforcement Elsevier. https://www.sciencedirect.com/science/
article/pii/S0141029615005738. Accessed 8 Nov 2018
12. Bartolac M, Damjanović D, Duvnjak I (2015) Punching Strength of Flat Slabs with and
without Shear Reinforcement. GRAĐEVINAR 67(8):771–786. https://doi.org/10.14256/
JCE.1361.2015
 Review on Fire Performance of Cellular
Lightweight Concrete

I. R. Upasiri1(&), K. M. C. Konthesingha2, K. Poologanathan2,

S. M. A. Nanayakkara3, and B. Nagaratnam2
1
University of Sri Jayewardenepura,
41 Lumbini Avenue, Dehiwala-Mount Lavinia, Sri Lanka
[email protected]
2
Northumbria University Newcastle, Ellison Building ELB117 Northumbria
University, Newcastle upon Tyne NE18ST, UK
3
University of Moratuwa, Katubedda, Moratuwa, Sri Lanka

Abstract. Structural fire damage can be identified as a common accidental

disaster throughout the world which cause thousands of deaths, injuries and
millions in property damage each year. Fire represents one of the most severe
conditions to which structures may be subjected. Generally, structural elements
will be exposed to very high temperature (1200 °C) during a fire propagation.
Fire safety of a structure is measured in terms of fire resistance, which is the
duration that a structural member can exhibit resistance with respect to structural
integrity, stability and heat transmission. Concrete generally provides better fire
resisting characteristics compared to the other construction materials due to its
low thermal conductivity, high heat capacity and slower strength degradation
with temperature. Cellular lightweight concrete (CLC) is one of the novel type
of concrete which can be identified as a better construction material than con-
ventional concrete due to its numerous advantages. However, limited research
work has been carried out to determine the fire performance of CLC. Fire
response of structural members depends on the thermal, mechanical and
deformation properties of the structural material at elevated temperatures. Even
though properties at elevated temperatures for normal weight concrete is
available in literature, properties of CLC at elevated temperatures (ambient to
1200 °C) is not thoroughly investigated. Further, CLC fire rating under
natural/parametric fire situations and under hydrocarbon fire situations needs to
be studied. EN 1992.1.2 provides minimum thickness requirements under
standard fire situations for non-loadbearing and load bearing normal weight
concrete walls, but for CLC, these values are not available, hence required to be
included. Also, parameters and material property limitations related to spalling
effect of CLC during fire exposure has not being investigated. Moreover,
residual characteristics of CLC walls after fire situations and ability to withstand
a second fire situation needs to be assessed.

Keywords: Cellular lightweight concrete
Fire performance

Design guidelines
Insulation
Integrity

© Springer Nature Singapore Pte Ltd. 2020

R. Dissanayake and P. Mendis (Eds.): ICSBE 2018, LNCE 44, pp. 470–478, 2020.
https://doi.org/10.1007/978-981-13-9749-3_41
 Review on Fire Performance of Cellular Lightweight Concrete 471

1 Introduction

Fire represents one of the most severe condition that structures may be subjected to
(Kodur 2014). Structural fire damage causes thousands of deaths, injuries and millions
in property damage throughout the world each year (Brushlinsky et al. 2007).
According to the Centre of fire statistics data (CTIF 2019), there are nearly 510,000
structural fire situations reported each year, which means once in every 62 s, a
structural fire is occurring somewhere in the world. World trade centre fire in 2001,
Shanghai fire in 2010 and Grenfell Tower Fire in 2017 are recent massive structural
fires reported in the world (En.wikipedia.org 2018). Therefore, appropriate fire safety
construction is an important aspect in building construction.
Concrete generally provides the best fire resistant properties compared to other
construction materials (Kodur and Raut 2010). This excellent fire resistant character-
istics are due to its low thermal conductivity, high heat capacity and slower strength
degradation with temperature (Kodur 2014). Concrete is available in various forms and
it is often grouped under different categories based on weight (as normal weight and
light weight concrete), strength (as normal strength, high strength, and ultrahigh
strength concrete), presence of fibres (as plain and fibre-reinforced concrete), and
performance (as conventional and high performance concrete) (Kodur 2014).
Lightweight concrete has many advantages over conventional concrete, excellent in
acoustic performance, earthquake resistant, good insulation, workability, long-life span
due to termite and fire resistance, weather proof and, material savings (Singh 2016) and
therefore can be utilised as a better construction material in low rise and multi storey
buildings.
Lightweight concrete can be defined as a type of concrete which has low density
compared to normal weight concrete. Lightweight concrete has dry density of
300 kg/m3–1800 kg/m3, while conventional concrete has a dry density of 2400 kg/m3.
Lightweight concrete can be categorised as lightweight aggregate concrete and cellular
lightweight concrete (CLC). Porous lightweight aggregates with low specific gravity
such as pumice, scoria and volcanic aggregates are used to produce lightweight
aggregate concrete. Due to low specific gravity of aggregates, resultant concrete also
has less density. Cellular lightweight concrete does not contain coarse aggregate, thus
can be regarded as mortar. It is produced by introducing air or other gas into a cement
slurry and fine sand. Cellular concrete can be categorised as aerated concrete and foam
concrete. Fine powder of Aluminium is used as air entraining agent to produce aerated
concrete. Aluminium powder mixed with the cement slurry, reacts with the calcium
hydroxide to produce hydrogen gas, which produces a cellular structure and thus makes
the concrete lighter than the conventional concrete. Foam concrete is produced by
introducing foam into cement mortar. The foam is created using a foaming agent,
mixed with water and air from a foam generator. When the sand in cement mortar is
replaced with fly ash, better performance can be obtained in cellular concrete. Since it
consumes fly ash which is a waste material from coal fired thermal power plants,
cellular lightweight concrete can be identified as a green building material. With the
reduction of density, the strength and stiffness of cellular concrete is decreased (Mydin
and Wang 2011). CLC is a porous material; since at high temperatures, heat transfer
472 I. R. Upasiri et al.

through porous material is influenced by radiation, CLC with its lower thermal con-
ductivity and diffusivity results in better fire resistance. Furthermore, lower density of
foam concrete has exhibited better fire performance characteristics (Ramamurthy et al.
2009). However, cellular concrete cannot be used as reinforced concrete as it has a
cellular structure and therefore corrosion of steel reinforcement is quite proactive
(Singh 2016). In this paper, existing research studies on CLC with respect to fire
performance will be assessed and research gaps will be highlighted.

2 Fire Performance of Cellular Lightweight Concrete

Fire resistance of structural members are generally evaluated through standard fire tests
(ASTM E119-08b 2008). However, due to high cost and time consumption for fire
tests, numerical approaches and finite element modelling has been identified as an
effective method of evaluating fire performance of structures (Kodur 2014). These
numerical models are developed considering the material properties at elevated
temperatures.

2.1 Fire Performance of Cellular Lightweight Concrete with Fire Tests

The simplest way of representing fire is a time temperature curve. ASTM E 119 and
ISO 834 are most widely used standard fire curves (Denoël 2007) (see Fig. 1). To
determine the fire performance of a structural member, fire tests are carried out. There
are full scale fire tests and furnace tests. Full scale fire tests will give exact fire
behaviour of the entire structure. However, it is very expensive and time-consuming.
Furnace tests are used to evaluate the fire behaviour of structural elements. Furnace
temperature needs to be controlled according to the designated fire curve (standard or
natural fire curve). Usually, furnaces are equipped with devices to measure tempera-
ture, and deformations, and to load test specimens.

Fig. 1. Time temperature variation of Standard fire curves.

 Review on Fire Performance of Cellular Lightweight Concrete 473

Fire safety of a component is assessed relevant to the standard fire situation.

ASTM E 119 considers the inhabit passage of flame or hot gases through the com-
ponent as a failure criterion while ISO 834 measures it by the ignition of a cotton pad
held close to an opening for 10 s. ASTM E 119 has limited the average temperature
rise of an unexposed surface to 139 °C while ISO 834 allows the maximum average
temperature rise of 140 °C at unexposed surface. Further, ASTM E 119 considers the
temperature rise of reinforcement. It is limited to 593 °C. ISO 834 considers the
deformation of the member. Downward deformation of flexural member is limited to
L/30 where L is the span of the member.
Standard fires are suitable for comparison of fire resistance of materials, but in
actual fire situations, ventilation effects, environmental conditions, material composi-
tion, boundary conditions and fuel availability of the compartment will affect the
intensity of fire. Considering the above effects, more realistic fire curves (natural or
parametric fire curves) have been developed. Lie’s parametric curve, Swedish curve,
Law’s maximum temperature equations, Mehaffey’s Japanees parametric curves, Ma
and Makelainen’s parametric curves, Barnett’s BFD curves and Euro code parametric
curves are such parametric curves developed by various researchers to interpret natural
fire situations (Ariyanayagam and Mahendran 2014).
Laurent (2014) has investigated the fire performance of foamed concrete under
insulation criteria up to 900 °C temperature and concluded that foamed concrete has
better fire performance characteristics compared to normal weight concrete.
For cellular lightweight concrete with different mix proportions, standard fire
performance has not been investigated, hence, needs to be determined. Further, with
different parametric fire situations and hydrocarbon fire situations research studies on
CLC behavior in fire is not available.
Research studies on residual properties of CLC after fire scenarios is not available
in the literature where these properties are important to investigate the fire performance
of CLC. During fire exposure, structural element will undergo high temperature vari-
ation which will induce internal stresses and variation of bonding in CLC. Therefore,
though the material is capable of handling fire, after the fire situation, CLC properties
might be different from the initial state. Hence, post fire load bearing capacities and
capability of handling second fire, needs to be determined.
Breaking up of pieces of concrete when it is exposed to high and rapidly rising
temperatures is defined as spalling (Kodur 2014). Spalling can be identified as a special
characteristic in concrete (Kodur 2014). Fire-induced spalling is one of the major
components to be considered in evaluating the response of concrete members exposed
to fire. It occurs due to the build-up pore pressure during heating (Dwaikat and Kodur
2010). Spalling depends on moisture content, permeability, type of fire exposure and
tensile strength of concrete (Kodur 2014). Research studies were not done on inves-
tigating of spalling of CLC. Spalling might cause a significant effect on fire perfor-
mance of CLC due to its cellular structure.

2.2 Fire Performance of CLC with Properties at Elevated Temperature

Fire performance of material is influenced by the characteristics of the material, thermal
properties, mechanical properties, deformation properties and special characteristics of
474 I. R. Upasiri et al.

material in fire. Thermal properties influence the heat transfer in the structural element,
whereas mechanical properties influence the strength and stiffness variation. Defor-
mation properties together with mechanical properties influence the extent deformation
and strains in the structural member (Dwaikat and Kodur 2010). All these properties
vary as a function of temperature and depends on the composition of the material.
Hence to predict fire performance, elevated temperature properties need to be
determined.
Thermal properties that influence temperature rise and heat distribution in structural
members are thermal conductivity, specific heat and mass loss (Kodur 2014). The
thermal conductivity of concrete is the rate of heat transferred through a unit thickness
of the material per unit temperature difference. Thermal conductivity can be measured
using several techniques such as pulse method, hot-guarded plate method (HGP), two-
linear parallel probe method, transient hot wire method and laser flash method. Specific
heat is the amount of heat per unit mass required to change the temperature of material
by one degree. The thermal diffusivity of concrete, often denoted as the ratio between
its thermal conductivity and thermal capacity. It is a measure of the rate of heat
transported from the exposed surface of concrete to the inside. Flash method is more
convenient in measuring thermal conductivity, specific heat and thermal diffusivity of
CLC in elevated temperatures. Density of concrete depends primarily upon the type of
aggregate. With the increment of temperature density also changes (Kodur 2014). This
effect is more predominant in cellular lightweight concrete than conventional concrete
since CLC consists more voids. Thermo gravimetric analysis is more suitable to
determine the density variation of CLC at elevated temperatures.
Mechanical properties that influence the fire performance of concrete are com-
pressive and tensile strength, modulus of elasticity, and stress strain response at ele-
vated temperatures (Kodur 2014). Thermo-mechanical analysis needs to be done to
determine the mechanical properties at elevated temperatures. Both “steady state” and
“transient” test methods are available for determining the mechanical properties at
elevated temperatures (Kodur 2014). Unlike at ambient temperature, mechanical
property tests at high temperature are usually carried out on a wide range of specimen
sizes due to a lack of standardized test specifications for undertaking high temperature
mechanical property tests (RILEMTC129-MHT 2000).
Deformation properties that affects the fire performance of concrete are thermal
expansion and creep (Kodur 2014). Creep, can be defined as the time-dependent plastic
deformation of the material. At normal stresses and ambient temperatures, deforma-
tions due to creep are not significant. At higher stress levels and at elevated temper-
atures, however, the rate of deformation caused by creep can be substantial (Kodur
2014). Dialatometric analysis can be done to determine the deformation characteristics
at elevated temperatures.
Othuman and Wang (2011) have studied thermal properties of foam concrete at
elevated temperatures. They have measured thermal conductivity of foam concrete with
various densities up to 600 °C using the hot guarded plate test. Further, they have
proposed analytical models to predict thermal conductivity and specific heat of foam
concrete at elevated temperatures. In the analytical method, lightweight foamed con-
crete (LFC) was considered as a mixture of dried LFC and water up to 170 °C. Above
170 °C, LFC was considered as a mix of dried LFC and air pores. Average pore
 Review on Fire Performance of Cellular Lightweight Concrete 475

diameter found from the microscopic images used was in the analytical model. Heat
transfer of solid material through conduction and heat transfer of air voids through
convection and radiation were considered. They have concluded that the models give
accurate results predicting the thermal conductivity of foamed concrete at elevated
temperatures. Further, they have mentioned that specific heat of LFC can be approx-
imated as a sum of base value of dried LFC with addition of heat required to evaporate
free water.
Mydin and Wang (2011) have measured mechanical properties of foamed concrete
with various densities in elevated temperatures. They have measured compressive
strength, modulus of elasticity, compressive stress-strain relationship and flexural
bending strength of foamed concrete with densities 650 kg/m3 and 1000 kg/m3. All the
tests were conducted increasing the load while keeping the specimen at a constant
elevated temperature (100 °C, 200 °C, 300 °C, up to 600 °C) and it was found that the
compressive strength of foamed concrete decreases with temperature. However, up to
200 °C, foamed concrete is in a position to maintain 94% of original unheated strength.
But at 400 °C. 75% and at 600 °C, only 40% of original strength was retained.
Moreover, modulus of elasticity at 200 °C, 400 °C and 600 °C was respectively 75%,
40% and 25% of the original value.
A number of concrete mechanical property predictive models have been proposed
for normal-strength concrete and these models have been checked to determine their
applicability to foamed concrete. Sayadi et al. (2016) have concluded that the following
models as given in Eqs. (1–6) give accurate results for foamed concrete properties in
elevated temperatures.
Compressive Strength

fcT ¼ fc for T  100  C ð1Þ

fcT ¼ fc ð1:067  0:00067 T Þ

ð2Þ
for 100  C  T  400  C

fcT ¼ fc ð1:44  0:0016 T Þ for T  400  C ð3Þ

fcT – Compressive strength at elevated temperature

fc – Compressive strength at ambient temperature
T – Temperature
Elastic Modulus

EcT ¼ Ec for T  60  C ð4Þ

800  T
EcT ¼ Ec for 60  C  T  800  C ð5Þ
740
476 I. R. Upasiri et al.

EcT – Elastic modulus at elevated temperature

Ec – Elastic modulus at ambient temperature
T – Temperature
Stress-Strain Relationship

0 3ecT fcT
fcT ¼ ð6Þ
cT 3
e0T ½2 þ ðee0T Þ

0
fcT – Compressive stress at elevated temperature
ecT – Strain at elevated temperature
e0T – Strain at maximum stress
Vinith Kumar et al. (2018) have compared thermal behaviour of foamed concrete with
clay bricks and concluded that thermal conductivity of foamed concrete varies from
0.021–0.035 W/mK while brickwork has a thermal conductivity of 0.6 W/mK to
1.0 W/mK. Sayadi et al. (2016) has studied the effect of expanded polystyrene
(EPS) partials in foamed concrete. Increase of EPS cause reduction in thermal con-
ductivity and higher fire endurance can be obtained with 28% EPS volume in foamed
concrete. Kashani et al. (2017) studied the influence of recycled tyre crumbs in cellular
lightweight concrete as an insulating constituent. They have concluded that excellent
sound and thermal insulation with very low water absorption and total porosity have
been achieved by the inclusion of recycled tyre crumb in lightweight cellular concrete.
She et al. (2018) proposed a numerical method of predicting thermal behaviour of
cellular concrete having different pore shapes.
Even though experimental results for thermal properties of CLC are available up to
600 °C, variation of thermal conductivity, specific heat, thermal expansion with tem-
perature need to be measured up to 1200 °C to model an actual fire situation. Further,
density, porosity, strength and elasticity variation with ambient temperature to 1200 °C
need to be determined for better prediction in fire performance of CLC. Normally,
mechanical properties at elevated temperature are measured keeping the specimen in a
constant elevated temperature and increasing the load. However, it will be more
realistic if a test method could be developed to increase the temperature while keeping
the load constant when determining the elevated temperature mechanical properties.

3 Design Guidelines for Cellular Lightweight Concrete

in Fire Situation

EN 1992-1-2 (2004) Design of concrete structures -part 1-2 General rules- Structural
fire design describes the principles, requirements and rules for the structural design of
buildings exposed to fire. It includes safety requirements, design procedures and design
aids for concrete structures under fire situations. The methods described in EN 1992-1-
2 are applicable to normal weight concrete up to strength class C90/105 and for
lightweight concrete up to strength class LC55/60.
 Review on Fire Performance of Cellular Lightweight Concrete 477

EN 1992-1-2 has tabulated minimum wall thickness requirement for 30, 60, 90, and
120, 180 and 240 min for fire resistant walls for normal weight concrete with siliceous
aggregates. It also states that thickness requirement is reduced by 10% when calcareous
aggregates are used in normal weight concrete. However, minimum thickness
requirement for CLC panels which can be used as non-load bearing walls is not
available in the design guidelines.
Further, EN 1992-1-2 states that both load bearing and non-load bearing walls
should be limited to wall clear height to thickness ratio of 40 for normal weight
concrete. However, applicability of this ratio for CLC walls should be investigated.
Moreover, explosive spalling of concrete can be observed in concrete structures
during fire exposure. EN 1992-1-2 has given limitations to the moisture content of
normal weight concrete to prevent explosive spalling in the member. However,
moisture content limitation for CLC to prevent explosive spalling is not available in
design codes of practice and needs to be investigated.

4 Concluding Remarks

Cellular lightweight concrete can be identified as a novel type of concrete which can be
used as a construction material for load bearing and non-load bearing walls in low-rise
and multi storey buildings. It has numerous advantages against other construction
materials. A comprehensive literature survey was done on the fire performance of CLC
and the research gaps were identified.
CLC elevated temperature property variations are available up to 600 °C. However,
to study the behaviour of CLC in fire situation, CLC elevated temperature properties
(ambient to 1200 °C) need to be measured. Properties such as, thermal conductivity,
specific heat, thermal expansion, density, porosity, strength and elongation need to be
measured in elevated temperatures. Standard fire tests need to be conducted for non-
load bearing and load bearing CLC walls to determine the fire rating and spalling of
CLC. Moreover, CLC behaviour in hydrocarbon fire situations and fire behaviour
under different parametric fire situations needs to be assessed. Post fire residual
properties of CLC needs to be measured to predict the adequacy of CLC to withstand in
a second fire situation. Properties of CLC at elevated temperatures and fire tests results
can be used to update the design guidelines for CLC walls in fire situations. Minimum
thickness requirement for non-load bearing and load bearing CLC walls, maximum
wall height to thickness ratio and material property limitations to prevent explosive
spalling need to be included in relevant design guidelines.

References
Ariyanayagam A, Mahendran M (2014) Development of realistic design fire time-temperature
curves for the testing of cold-formed steel wall systems. Front Struct Civil Eng 8(4):427–447
Brushlinsky N, Ahrens M, Sokolov S, Wagner P (2007) World Fire Statistics. Center of fire
Statistics
478 I. R. Upasiri et al.

Ctif.org (2019) World Fire Statistics Issue no 22 - 2017–CTIF - International Association of Fire
Services for Safer Citizens through Skilled Firefighters. https://www.ctif.org/news/world-fire-
statistics-issue-no-22-2017. Accessed 6 Oct 2018
En.wikipedia.org (2018) List of building or structure fires. https://en.wikipedia.org/wiki/List_of_
building_or_structure_fires. Accessed 15 Nov 2018
Kashani A, Ngo T, Mendis P, Black J, Hajimohammadi A (2017) A sustainable application of
recycled tyre crumbs as insulator in lightweight cellular concrete. J Cleaner Prod 149:925–
935
Kodur V (2014) Properties of concrete at elevated temperatures. ISRN Civ Eng 2014:1–15
Laurent C (2014) Investigating the fire resistance of ultra light weight foam concrete. Revista
Técnica de la Facultad de Ingeniería Universidad del Zulia 37(1):11–18
Dwaikatand MB, Kodur VKR (2010) Fire induced spalling in high strength concrete beams. Fire
Technol 46(1):251–274
Mydin M, Wang Y (2011) Mechanical properties of foamed concrete exposed to high
temperatures. Constr Build Mater 26(1):638–654
Othuman M, Wang Y (2011) Elevated-temperature thermal properties of lightweight foamed
concrete. Constr Build Mater 25(2):705–716
Ramamurthy K, Kunhanandan Nambiar E, Indu Siva Ranjani G (2009) A classification of studies
on properties of foam concrete. Cem Concr Compos 31(6):388–396
RILEMTC129-MHT (2000) Test methods for mechanical properties of concrete at high
temperatures, Part 4—tensile strength for service and accident conditions. Mater Struct
33:219–223
Sayadi A, Tapia J, Neitzert T, Clifton G (2016) Effects of expanded polystyrene (EPS) particles
on fire resistance, thermal conductivity and compressive strength of foamed concrete. Constr
Build Mater 112:716–724
She W, Zhao G, Cai D, Jiang J, Cao X (2018) Numerical study on the effect of pore shapes on the
thermal behaviors of cellular concrete. Constr Build Mater 163:113–121
Singh N (2016) Effective uses of light weight concrete. J Civil Eng Environ Technol 3(3):208–
211
Standard test methods for fire tests of building construction and materials (2008) ASTM E119-
08b, ASTM International, West Conshohocken, PA, USA
The European Union (2004) Eurocode 2: Design of concrete structures - Part 1-2: General rules -
Structural fire design
Kodur VR, Raut N (2010) Performance of concrete structures under fire hazard: emerging trends.
Indian Concr J 84(2):23–31
Vinith Kumar N, Arunkumar C, Srinivasa Senthil S (2018) Experimental study on mechanical
and thermal behavior of foamed concrete. Mater Today: Proc 5(2):8753–8760
 Numerical Modelling of Re-cracking
Behaviour in Retrofitted Crack Stop Holes
in Steel Structures

S. Abeygunasekera1(&), J. C. P. H. Gamage1, and S. Fawzia2

1
University of Moratuwa, Katubedda, Moratuwa, Sri Lanka
[email protected]
2
Queensland University of Technology, Brisbane, Australia

Abstract. This paper focuses on the delaying of crack initiation process in a

crack stop hole (CSH) in steel structures. The cyclic J-Integral (DJ) numerical
model was adopted to develop this study. Also, the Power law was applied to
estimate the required number of cycles for crack initiation at the crack stop hole
with the effects of carbon fiber reinforced polymer (CFRP) strengthening CSH
improvement techniques. CFRP strengthened CSH indicated a significant
improvement in the CSH. This retrofitting technique reported an optimal per-
formance; in the range between 17.7% and 163.6% delay in required cycles for
the crack initiation when compared to an un-strengthened CSH.

1 Introduction

Crack stop hole (CSH) technique is a widespread method in the steel industry as a
crack repairing method. In this method, a drilled hole is fabricated at the crack tips to
crack transfer into a notch. The main purpose of the crack stop hole is the re-
distribution of the stresses which reduces stress concentrations at the crack tip. The
result is delay in crack propagation and changes in the direction of crack propagation.
The crack stop hole technique has been verified as enhancing the fatigue service life of
a structure [1, 2]. This technique is a simple and non-expensive method [3]. As a result,
this crack stop hole technique is widely applied in the maintenance of metallic struc-
tures [4]. This technique is especially used in the aerospace industry.
Due to the removal of material to prepare CSH, the load is redistributed throughout
the structure. Hence, the stress is maximum at the edge of the crack stop hole.
Therefore, the maximum stress concentration occurs around the CSH. Stress concen-
tration in ductile materials is higher when compared with the yield strength of material.
Such behaviour causes the redistribution and deformation of the local plastic stresses.
However, stress intensity factor (SIF) could considered as a good indicator representing
the stress at the vicinity of the crack tip as it could be used to predict stress and fracture
behaviour of the CSH. The simulation of the crack is important for predicting the
fatigue life with different techniques which could be used for this purpose. Thus, the J-
Integral technique could be considered as one of the best alternatives, as an extension
of SIF.

© Springer Nature Singapore Pte Ltd. 2020

R. Dissanayake and P. Mendis (Eds.): ICSBE 2018, LNCE 44, pp. 479–488, 2020.
https://doi.org/10.1007/978-981-13-9749-3_42
480 S. Abeygunasekera et al.

The potential of fatigue damage occurring at the crack stop hole is maximum with
respect to the remaining areas, since such a crack stop hole acts as a structural dis-
continuity. When the CSH is continuously subjected to cyclic loads, it results in the
stress concentration increase near the CSH. Due to the increase in the stress concen-
tration, crack initiation restarts from the CSH and this action is called a crack re-
cracking on CSH. In fact, only a few numerical investigations are available in the
literature to estimate delaying crack incubation and growth at the CSH. According to
Tanaka and Mura (1981), the reason for crack initiation originates at the slip bands,
without accounting for the effect of cyclic hardening of the material. Also, Bozek et al.
(2008) suggested a probabilistic simulation technique to understand the re-cracking
behaviour in members. Similarly, Fine and Bhat (2007); Mura and Nakasone (1990);
Bobylev et al. (2010) and Xie et al. (2016) have emphasized the ability of delay in
crack initiation due to energy barrier techniques.

2 Background

Computerized finite element analysis techniques are more popular due to the practical
difficulties of laboratory testing. This study also utilized the J-integral technique in
ABAQUS for analysis. The Paris law was used to estimate the rate of crack propa-
gation at the CSH and its regression curve was fitted to obtain the power law
approximation. In fact, the rate of crack growth could be estimated accurately by using
the results of the power law approximation. When comparing the results of Paris law,
the proposed power law approximation showed negligible differences (less than 3%).
The Paris law behaviour was based on characteristic of material such as homogeneity,
isotropy, and elasticity.
The main problem in CSH is the initiation of re-cracking at the tip of the hole.
There are some techniques already applied in this field to understand the delay in this
crack propagation. The main objective of this investigation is the analysis of the re-
cracking behavior of CSH. In addition, the behaviour of retrofitted CSH was also
explored in this context.
Carbon Fiber Reinforce (CFRP) materials show excellent fatigue resistance in
strengthened steel structures. Furthermore, CFRP materials exhibit exceptional char-
acteristics; few of them being lightweight, high strength, corrosion resistance, and ease
of installation. Field applications also confirmed the usefulness of the CFRP related
strengthened steel structures. Therefore, this investigation focuses on the numerical
modelling of CFRP strengthened CSH as an alternative solution to delaying re-cracking
in a CSH.

3 Methodology

In this method, a single layer of CFRP was attached to the CSH at the bottom surface of
the member which is shown in Fig. 1. A tensile load was applied on the bottom surface
of the member. Thus, the tendency of crack initiation is high at the bottom surface
rather than the upper side of the specimen. Crack stop hole strengthened with CFRP
 Numerical Modelling of Re-cracking Behaviour in Retrofitted CSH 481

assist in reducing the stress at the edge of the hole. Thus, the dimensions of the CFRP
layer were selected as 100 mm length and 40 mm width. Also, uni-directional CFRP
material was selected which is commercially denoted as CF 130. Thickness of the
CFRP layer was taken as 0.167 mm according to the data sheet provided by the
manufacturer. Young’s modulus and Poisons ratio of the material was taken as 240
GPa and 0.28 respectively. Interaction mode in the ABAQUS was utilized for mod-
elling of the CFRP layer under vicinity of the crack stop hole in the structural member.
Steel plates were considered as the master surface and the CFRP layer was considered
as the slave surface.

Fig. 1. Schematic view of the CFRP patched with CSH.

CFRP strengthened CSH specimens were kept on two cylindrical shaped, sup-
portive rollers with each having a 25 mm diameter and the span length was fixed at
240 mm between centers of the rollers as shown in Fig. 2. The specimen module was
loaded with a low cycle fatigue (LCF) mode using the direct cyclic option in ABA-
QUS. A Cyclic load was applied on the specimen at the mid-plane using the loading
nose and 8 mm radius loading nose was used as shown in Fig. 2. The load was applied
at the mid-plane of the specimen. The maximum deflection in the load direction (Y
direction) at mid-plane was limited to 5 mm. The magnitude of the frequency was
selected according to the ASTM D7774 standards [5]. A constant amplitude load was
applied to the model while during the analysis, 5 Hz frequency and 0.05 stress ratio
were maintained. Parametric study was conducted by varying the diameter of CSH and
alternative bonding techniques of CFRP. The procedure followed is listed below:
1. The geometry of the model using the part modulus of ABAQUS was created.
2. A pre-defined crack profile using a partition cell extrude/sweep edges option of the
part modulus was assigned.
482 S. Abeygunasekera et al.

3. The material properties, boundary conditions, interaction properties and loads for
both the static general step and the cyclic step were utilized in this module. (Static
general step for the applying step 1 load, direct cyclic step 2 for fatigue load)
4. Mesh of the model using CPE4R elements was utilized.
5. A special enrichment function called “seams” using contour integral, under inter-
action modulus were performed to this procedure.
6. The analysis was created and the results were taken from the data file of visual-
ization modulus.

Fig. 2. 3D object with 240 mm span length under the 3 points cyclic flexure load

In this numerical analysis, the diameter of CSH varied from 8 mm to 24 mm.

According to the model of the numerical analysis, the minimum and maximum J-
integral values were obtained while the cyclic J-integral value was calculated according
to Eq. 1. The required number of cycles for crack initiation was calculated by using the
power law approximation.

4 Theory
4.1 Theory of Fracture Mechanics
At present, two main fracture mechanics approaches are used for predicting the fatigue
life. They are the S-N curve approach method and fracture mechanics approach
method. The fracture mechanics approach is utilized in this study where the concept of
fracture mechanics was introduced by Leonardo da Vinci several centuries earlier [6]
Theory of fracture mechanics is based on the initiation and propagation of the cracks.
Since, fracture mechanics are based on the mechanical behaviour of steel structures, the
ultimate effect of the fracture is the catastrophic (Unrecoverable) failure of the struc-
ture. Therefore, understanding of fracture and its control is important in the real-world
application to assure safety. The best way to estimate fatigue life is a prediction of
crack initiation. Studies of Griffith (1921), Irwin(1957), and Rice (1968) have greatly
contributed to develop the theory on fracture mechanics [7]. Similarly, Irwin in 1948
 Numerical Modelling of Re-cracking Behaviour in Retrofitted CSH 483

has extended the Griffith approach to metals [8]. This extension was utilized as energy
dissipates due to local plastic flow. During the same period, Orowan [10] indepen-
dently proposed similar modifications as the Griffith theory [9, 10]. In 1948 Mott
extended the Griffith theory considering the rapid propagation of the crack [11]. In
1956, Irwin developed the concept of strain energy release rate [12]. This concept
showed a direct relation with the Griffith’s theory.

4.2 J-integral Technique

Dr. James Rice in 1968 introduced a path-independent contour integral analysis
technique for crack analysis [13]. Rice explained the J-integral method with elastic-
plastic fracture mechanics (EPFM). The cyclic J-integral (DJ-integral) could be con-
sidered as a crack tip parameter. The cyclic J-integral (DJ) value can be obtained by
using finite element analysis. The result of crack driving force is propagation of a short
fatigue crack. In fact, the crack driving force can be alternatively named as a crack tip
stress or as the strain energy release rate during the crack growth under elastic-plastic
condition. Therefore, the J-integral technique could be viewed as an energy related
parameter as well as a stress intensity parameter for non-linear materials. Hutchinson,
Rice, and Rosengren have confirmed that the J-integral technique is unique for non-
linear materials. Rice has confirmed the basis of EPFM methodology well beyond the
validity limits of LEFM. The estimation of cyclic J-integral values from using mini-
mum and maximum values of the J-integral could be explained as follows.

DJ ¼ Jmin þ Jmax  ðJmin þ Jmax Þ1=2 ð1Þ

Where DJ is the cyclic J-integral value, Jmin is the minimum value of the J-integral
taken from FEA and Jmax is the maximum J-integral value available from the same
analysis.

4.3 The Power Law

Prediction of crack growth is important for structural elements during their service life,
as such predictions helps the estimation of the remaining service life of any structural
element. In addition, such a prediction supports to estimate the inspection intervals of
structures for maintenance. Therefore, accurate and reliable approximation is required
to predict the behaviour of crack propagation in structures. On the other hand, the
method of prediction should not be complicated. However, to satisfy all these condi-
tions at the same time is not an easy task as, numerous variables significantly influence
the process of crack initiation. The cyclic J-integral technique is based on the theory of
fracture mechanics. Dowling and Begley [8] have proposed and implemented the J-
integral technique considering the relations of crack growth rate (da/dN), and cyclic J-
integral (DJ) [14]. The power law approximation is similar to the Paris law. The Power
law and fatigue crack growth rate of structures exposed to cyclic loading can be
calculated as follows:
484 S. Abeygunasekera et al.

da=dN ¼ AðDJÞn ð2Þ

In the Power law application, it is assumed that the cyclic J-integral (DJ) includes the
sum of elastic and fully plastic solutions. The rate of fatigue crack growth in the Power
law depends on constants A and n. These two parameters are called the Ramberg-
Osgood coefficient and strain hardening index, respectively. These constants spe-
cially represent the tensile data of materials. However, these two parameters do not
depend on the geometry and it is based-on material properties. In this analysis, theses
power law constants, A and n were considered as 0.95 and 0.47 respectively, using the
Paris law results approximation with the Power law. The cyclic J-integral values were
obtained from numerical analysis.

5 Model Results

In this study, CFRP strengthened and un-strengthened CSH were modelled to compare
the results. All other parameters which critically affected crack behaviour was kept as a
constant during the process of numerical analysis. The boundary conditions, properties
of materials, loading rates, frequency, stress ratio and aspect ratio were kept as con-
stants throughout the analysis (Fig. 3).

Fig. 3. Visualization of the 3D view under 3 - points flexural cyclic load with an adhesive CFRP
strengthen CSH.

The stress concentration at CFRP strengthened CSH contours is shown in Fig. 4.

 Numerical Modelling of Re-cracking Behaviour in Retrofitted CSH 485

Fig. 4. Visualization of the bottom view under 3 - points flexural cyclic load with CFRP
strengthened CSH.

When the three points cyclic flexural loads are applied on a CFRP strengthened
crack stop hole, the change of stress distribution pattern is shown in Fig. 5. In this case,
the stress variation around the CSH is not clearly visible as the elastic plastic zone is
controlled by the CFRP layer. This is the main reason for the delay in crack initiation at
the CSH compared to unstrengthened CSH.

Fig. 5. Stress distributions at CFRP strengthen CSH.

6 Results and Discussion

The analytical results of the CFRP strengthened CSH and un-strengthened CSH were
compared as shown in Table 1. It summarizes the maximum and minimum J-integral
values together with the variation of different CSH diameters from the range of 8 mm
to 24 mm. Interestingly, the number of cycles for crack initiation varied with the hole
size. The number of cycles for crack initiation compared with un-strengthened and
CFRP layer attached CSH is shown in Fig. 6.

Table 1. J-integral values and cyclic J-integral of CFRP strengthened CSH.

Diameter of JMin JMax Cyclic J integral (DJ)
the CSH (mm)
8 1.44E−04 8.54E−04 0.0006473207
12 1.24E−04 1.40E−04 0.0001322426
16 1.71E−04 2.25E−04 0.0001998495
20 4.54E−03 8.90E−03 0.0070834286
486 S. Abeygunasekera et al.

The variation of the number of cycles for crack initiation with respect to the
diameter of the CSH is shown in Fig. 6. The comparison of the un-strengthened CSH
and the CFRP strengthened CSH are shown in the same graph. The common behavior
of both graphs is sinusoidal wave patterns, which can be observed. According to the
investigations conducted by Kormay and Disnig [15], Ack et al. (1988) it was con-
firmed that theoretical stress concentration factor for a plate with a hole under bending
loads were a response to a sinusoidal stress pattern [8, 15]. When the diameter is
changed from 8 mm to 24 mm it showed significant delaying of the crack initiation
when compared to un-strengthened CSH.

Fig. 6. Comparison of number of cycles for crack initiation with un-strengthened and CFRP
strengthened crack stop hole with a hole diameter.

According to this study, the CFRP material exhibited a significant crack initiation
delay and its range in variation were recorded as 17.7% to 163.6% with respect to an
untreated crack stop hole. This technique is performed with 8 mm to 24 mm range of
the diameter of the CSH. However, up to medium size of the CSH diameter, significant
improvements were exhibited with respect to the untreated crack stop hole (Table 2).

Table 2. Delaying of crack initiation due to CFRP strengthened

Diameter of Percentage of delaying
CSH (mm) crack initiation compare
to un-strengthened CSH (%)
8 162.5
12 90.6
16 163.6
20 17.7
24 50.3
 Numerical Modelling of Re-cracking Behaviour in Retrofitted CSH 487

6.1 Discussion
CSH method shows a high potential in successfully delaying crack initiation at the
crack stop hole. When the CFRP layer was attached to the bottom surface of the
specimen, the effects of the tension load on bottom surface was controlled by the CFR,
since the tension plunge shows a high tendency of crack initiation rather than the
compression surface (upper side) of the specimen. The CFRP laminate also helps to
reduce the stresses at the edge of the CSH. Thus, a number of fiber layers, the module
of CFRP material, lay-up angles and bond length could be considered as the main
variables which affect the process of crack initiation.
Stress concentration at the CSH is the main reason for crack initiation and it is
governed by the residual stress and notch stress. When the CSH diameter is small, the
required number of cycles is less. The residual stress is high due to the mechanical
stress on the process of manufacturing such as cold forming and rolling. When the hole
diameter gradually increases in the notch, the stress decreases and the residual stress,
increases due to the process of drilling. According to the net effects of both stresses, the
required number of cycles for crack initiation increases as, a comparatively less stress
concentration is shown at the CSH. When the hole size was further increased, the
required number of cycles for crack initiation decreased again, because residual stress
increases due to the drilling process. Net effects of both actions contribute to create
high stress concentration at the CSH. When the hole diameter becomes larger, the
required number of cycles for crack initiation had drastically increased. This is due to
the reduction of notch stress when compared to the residual stress at the CSH. The
bottom surface is retrofitted with CFRP and the required number of cycles for crack
initiation was increased because the CFRP layer contributed to reduce the stress at the
edge of the holes.

7 Conclusions

In this study, CSH was strengthened with externally bonded CFRP sheets which were
numerically studied with the required number of cycles and crack initiation for CSH
were estimated using J-integral technique. The following conclusions were obtained:
1. Stress concentration at the crack stop hole edges effectively decreased with the
CFRP material.
2. The plastic zone of the vicinity of crack stop hole edges can be reduced by the
externally bonded CFRP sheets.
3. This retrofitting technique reported the performance; which is in the range of a
17.7% to 163.6% delay in crack initiation compared to the un-strengthened CSH
with the diameters changed from 8 mm to 16 mm.

Acknowledgments. The authors gratefully acknowledge the support of the staff in computer
laboratory, the Department of Civil Engineering at the University of Moratuwa for their valuable
support.
488 S. Abeygunasekera et al.

References
1. Wu H, Imad A, Benseddiq N, Castro JTPD, Meggiolaro MA (2010) On the prediction of the
residual fatigue life of cracked structures repaired by the stop-hole method. Int J Fatigue 32
(4):670–677
2. Crain JS, Smmons GG, Bennett CR, Barrettgonzalez R, Matamoros AB, Rolfe ST (2010)
Development of a technique to improve fatigue lives of crack-stop holes in steel bridges.
Transp Res Rec: J Transp Res Board 2200:69–77
3. Wang Y, Pan Q, Wei L, Li B, Wang Y (2014) Fracture toughness and fatigue crack growth
analysis of 7050-T7451 alloy thick plate with different thicknesses. J Central South Univ 21
(8):2977–2983
4. Makabe C, Kaito N, Ferdous MS (2014) Method of arresting crack growth for application at
a narrow working space. Mech Eng J 1(6):1–12
5. Lin F, Sun JG, Nakamura H, Maeda K (2012) Fatigue crack repair using drilled holes and
externally bonded CFRP strips. In: Bridge maintenance, safety, management, resilience and
sustainability. Taylor & Francis Group, London
6. ASTM D7774-12 (2013) Standard test method for flexural fatigue properties of plastics,
D20.10.24, Ed. West Conshohocken, PA: ASTM International
7. Anderson TL (1991) Fracture mechanics: fundamentals and applications. CRC Press Inc.,
Florida
8. Rice R (1968) A path independent integral and the approximate analysis of strain
concentration by notches and cracks. J Appl Mech 35:379–386
9. Irwin GR (1948) Analysis of stresses and strains near the end of a crack traversing a plate.
J Appl Mech 24:361–364
10. Orowan E (1948) Fracture and strength of solids. Rep Progress Phys XII:185
11. Griffith AA (1920) The phenomena of rupture and flow in solids. Philos Trans Ser A
221:163–198
12. Mott NF (1948) Fracture of metals: theoretical considerations. Engineering 165:16–18
13. Irwin GR (1956) Onset of fast crack propagation in high strength steel and aluminum alloys.
In: Sagamore research conference proceedings, vol 2, pp 289–305
14. Begley JA, Landes JD (1972) The J-integral as a fracture criterion. ASTM STP 514.
American Society for Testing and Materials, Philadelphia, pp 1–20
15. Kormay K, Disnig W (1991) pp 66–69
 Comparison of Properties of Masonry Walls
Constructed with Commonly Used Bricks
and Non-standard Large Bricks

V. L. W. Seneviratne(&) and D. Nanayakkara

Department of Civil Engineering, University of Moratuwa, Moratuwa, Sri Lanka

[email protected]

Abstract. Brick masonry is a very popular construction material especially

among the house builders, because of its better heat insulation properties,
compressive strength, high durability, availability and the low cost. In Sri
Lanka, brick masonry construction using commonly available burnt clay bricks
(normal bricks), in different bond patterns can be seen in practice. However, the
English bond is still considered to be the ideal bond pattern with normal bricks.
Recently it has been noticed that a burnt clay brick with larger dimensions (non-
standard brick) is gaining popularity in the local construction industry due to
some advantages such as fast construction and reduction in the required mortar
volume due to the reduced number of mortar joints. These larger bricks are used
with a header bond pattern to build typical one brick thick walls. Although there
are several benefits in using these larger bricks, it is important to determine the
strength properties as well, since the strength of brick masonry highly depends
on the strengths of brick units, mortar and the bond pattern. If these walls are
used as load bearing walls or in-fill walls, the determination of strength prop-
erties is essential. The effect on strength properties due to the difference in bond
pattern was addressed in this study. Experimental investigation was carried out
to compare the strength properties of normal and non-standard brick units and
corresponding brickwork in terms of compressive, flexural and shear strengths.
Locally available normal brick and non-standard brick samples have been
selected and tested according to the relevant standards to assess the strengths of
bricks and brickwork. It was observed that the brickwork constructed with non-
standard bricks has slightly higher compressive strength compared to brickwork
with normal bricks. However flexural strength of masonry constructed with non-
standard bricks with a header bond pattern is lower than the flexural strength of
normal brick masonry which is constructed with the English bond pattern.
Therefore, brickwork constructed with non-standard bricks is not efficient
compared to normal brickwork, especially when the walls are subjected to lat-
eral loads. This may be due to the weak bond in the brickwork constructed with
non-standard bricks as the header bond pattern was used in them.

Keywords: Compressive strength
Flexural strength
Non-standard brick

Standard bricks

© Springer Nature Singapore Pte Ltd. 2020

R. Dissanayake and P. Mendis (Eds.): ICSBE 2018, LNCE 44, pp. 489–498, 2020.
https://doi.org/10.1007/978-981-13-9749-3_43
490 V. L. W. Seneviratne and D. Nanayakkara

1 Introduction

Considering masonry construction in Sri Lanka, burnt clay bricks are the most com-
monly used masonry units in the construction industry. Generally, the dimensions of
commonly available bricks in the local market vary slightly from the specified dimen-
sions of standard bricks. However, recently it was found that a larger clay brick having
dimensions of 225 mm  150 mm  75 mm is becoming popular in the construction
industry. The properties of brick walls depends on the properties of brick units, mortar
and bond patterns. Hence, the investigation on the influence of size difference on the
strength of brick units, the effects on strength properties of brickwork due to the dif-
ference in bond pattern and the unit size were the main objectives of this study.
Few benefits can be expected when using the large brick units in masonry con-
struction. One benefit is the fast construction due to large unit size. The number of
mortar joints would also be reduced due to the larger units. In the construction of one
brick thick walls with non-standard large bricks, a better finish could be seen since the
header bond pattern is used. Generally, with the normal clay bricks, the English bond is
used and due to the uneven sizes of bricks, mortar fillings will be required to maintain
the constant wall thickness. Since there are benefits in using non-standard large bricks
as explained, it is necessary to investigate the strength characteristics for the effective
use of non-standard large size bricks in masonry construction.

2 Objectives

The main objectives of this research study are;

1. Investigation on the strength characteristics of the normal and non-standard large
brick units and the corresponding masonry panels.
2. Comparison of the properties of masonry panels constructed with normal and non-
standard bricks to ascertain the suitability of the non-standard bricks in masonry
construction.

3 Methodology
3.1 Materials Used
3.1.1 Brick Units
In this study, two locally available bricks of different sizes were used. One was a
commonly available brick having dimensions of 220 mm  105 mm  65 mm while
a larger brick having 225 mm  150 mm  75 mm was used as the non-standard
brick. Recently, this non-standard brick has been widely used in the construction
industry. Since the objective of this study is to compare the properties of masonry
panels constructed with these two brick types with different sizes, all bricks used for
investigation were bought from the same manufacturer ensuring that they are manu-
factured from the same clay and the same procedure.
Compressive strength of brick units was determined according to the SLS 39: 1978
(Sri Lanka Standards 1978). Figure 1 shows the loading arrangement used.
 Comparison of Properties of Masonry Walls Constructed 491

Fig. 1. Testing for compressive strength of bricks.

A sample of ten bricks was used to determine the compressive strength of bricks.
Prior to the compressive load application, bricks were immersed in water at room
temperature for 72 h. Compressive load was applied on the bricks using 200T
Universal Testing Machine.

3.1.2 Mortar Mix

Mortar mix of 1: 5 (cement: sand) was used for the masonry panels. Ordinary Portland
cement and local river sand were used for the mortar. Mortar cubes of 70.7 mm
70.7 mm  70.7 mm were tested for each batch of mortar as per BS 5628-1: 1992
(British Standard 1992). The mortar strengths were also tested on the same day when
masonry panels were tested.

3.2 Compressive Strength of Masonry Panels

Practically, normal size bricks are used for construction of both one brick thick and half
brick thick walls. Generally the English bond is used for one brick thick (225 mm) or
in other words 9″ thick walls and the stretcher bond is used for half brick or 4½″ thick
walls. Using the selected non-standard bricks, 9″ thick walls can be constructed with
the header bond and using stretcher bond, 150 mm thick walls can be constructed.
For the comparison of compressive strength of masonry constructed with different
sizes of bricks, one brick thick (225 mm) walls were considered. Panel sizes were
decided according to BS EN 1052-1: 1999 (British Standard 1999a). Specimen sizes
are shown in Table 1. All panels were covered with polythene sheets for three days
after construction and then left uncovered until tested. The panels were tested 28 days
after construction.

3.2.1 Brickwork Panels Constructed with Normal Bricks

One brick thick masonry panels were constructed with normal bricks using the English
bond and the header bond patterns. For the English bond, ten courses high wall panels
that were three bricks long, were constructed. For the header bond, wall panels of ten
courses high, six headers in a course were constructed. Three masonry panels per each
492 V. L. W. Seneviratne and D. Nanayakkara

case were constructed. Figure 2 shows these two types of wall panels and their
dimensions are given in Table 1.

Table 1. Specimen sizes for compressive strength test

Brick type & bond pattern Thickness (mm) Length (mm) Height (mm)
Normal (English bond) 220 680 740
Normal (Header bond) 220 680 740
Non-standard (Header bond) 225 630 755

Fig. 2. Masonry panels constructed with normal bricks using different bond patterns.

3.2.2 Non-standard Brickwork Panels

One-brick thick masonry wall panels were constructed with non-standard bricks using
header bond pattern. The wall panels were nine courses high with a length of four
headers. Three specimens were constructed for testing. Figure 3 shows the non-
standard brickwork panels and the dimensions are given in Table 1.

Fig. 3. Non-standard brickwork panels

 Comparison of Properties of Masonry Walls Constructed 493

According to the BS EN 1052-1: 1999 (British Standard 1999a) wall panels were
tested to determine the compressive strength of brick walls constructed with normal
and non-standard bricks.

3.3 Flexural Strength of Brick Masonry

Investigation on flexural strength of masonry was carried out in accordance with
BS EN 1052-2: 1999 (British Standard 1999b). According to this standard, the flexural
strength of masonry is derived from half-brick thick wall specimens tested under four-
point loading. Figure 4 shows the loading arrangement used in the flexural strength
test. A further study on one brick thick wall specimens was also considered in the
flexural strength test to compare the influence of bond patterns on flexural strength of
brickwork. The flexural strength was determined for two directions,
1. Flexural strength when the plane of failure is parallel to the bed joints,
2. Flexural strength when the plane of failure is perpendicular to the bed joints.

Fig. 4. Layout diagram of flexural strength tests

3.3.1 Flexural Strength of Masonry When Plane of Failure Is Parallel

to the Bed Joint
For flexural strength test for failure parallel to the bed joints, normal brick and non-
standard brick panels were tested for both half brick and one brick wall thicknesses.
According to the standard, the specimen size was two bricks long by ten courses high.
Five specimens were tested for each case. Immediately after the construction of wall
specimens, they were covered with polythene and pre-compressed by applying a
uniformly distributed vertical stress of 3  10−3 N/mm2. All panels were kept
undisturbed until testing and panels were tested at an age of 28 days (Fig. 5 and
Table 2).
494 V. L. W. Seneviratne and D. Nanayakkara

Table 2. Panel sizes for flexural strength test (parallel to bed joint)
Type of brick Thickness (mm) Height (mm) Length (mm)
Normal 105 740 450
Non-standard 150 840 460
Normal 220 740 450
Non-standard 225 840 470

Fig. 5. Loading arrangement for flexural strength test (failure parallel to bed joint)

3.3.2 Failure Perpendicular to the Bed Joint

For flexural strength test (when failure perpendicular to the bed joints), four bricks long
by four courses high specimens were tested for both standard and non-standard bricks
(Table 3).

Table 3. Specimen sizes for flexural strength test (perpendicular to bed joint)
Type Thickness (mm) Height (mm) Length (mm)
Normal 105 290 910
Non-standard 150 330 930
Normal 220 290 910
Non-standard 225 330 950

Five specimens were tested for each case. Specimens were pre-compressed by
applying a uniformly distributed load to give a vertical stress of 3  10−3 N/mm2 and
kept undisturbed until testing. Panels were tested at an age of 28 days (Fig. 6).
 Comparison of Properties of Masonry Walls Constructed 495

Fig. 6. Flexural strength test (failure perpendicular to bed joint)

3.4 Initial Shear Strength

Initial shear strength was determined according to BS EN 1052-3: 2002 (British
Standards Institute 2002). Six brick masonry triplets were tested for each case, 28 days
after casting. Specimens were tested under four point loading at zero pre compression.

4 Results
4.1 Average Unit Compressive Strength
Table 4 gives the compressive strength results obtained for normal and non-standard
large brick units.

Table 4. Average unit compressive strengths.

Brick type Dimensions Average Unit compressive
(mm) mass (kg) strength (N/mm2)
L W H
Normal 212 105 61 2.6 1.67
Non-standard 216 146 78 4.6 1.74

4.2 Average Compressive Strengths of Wall Panels

Compressive strength of brickwork was determined by calculating the mean com-
pressive strength of three masonry specimens. Compressive strength results are given
in Table 5.
496 V. L. W. Seneviratne and D. Nanayakkara

Table 5. Average compressive strength results of brickwork panels

Brick type Bond type Wall Average compressive
thickness (mm) strength (N/mm2)
Normal English 225 1.1
Normal Header 220 1.3
Non-standard Header 225 1.5

4.3 Shear Strength of Brick Masonry

Average shear strength results obtained for triplets made using normal bricks and non-
standard bricks are given in Table 6.

Table 6. Shear strength of brickwork

Brick type Shear strength (N/mm2)
Normal 1.16
Non-standard 1.15

4.4 Average Flexural Strength of Wall Panels

Flexural strength of masonry was determined in accordance with BS EN 1052-2: 1999
and the results are presented in Table 7.

Table 7. Flexural strength of masonry panels

Brick type Wall thickness (mm) Average panel Average flexural
size (mm) strength (N/mm2)
Length Height
Plane of failure parallel to the bed joints
Normal (Stretcher) 105 450 750 0.20
Non-standard (Stretcher) 150 460 920 0.18
Normal (English) 220 450 750 0.33
Non-standard (Header) 220 460 920 0.25
Plane of failure perpendicular to the bed joints
Normal (Stretcher) 105 890 240 0.72
Non-standard (Stretcher) 150 925 340 0.50
Normal (English) 220 905 300 0.46
Non-standard (Header) 220 940 355 0.35
 Comparison of Properties of Masonry Walls Constructed 497

By analyzing the results obtained from this study, significant differences in results
could not be seen for normal and non-standard size bricks. When considering the brick
unit strengths, 4% increase can be seen in non-standard brick units compared to normal
bricks. It was also found that the compressive strength of non-standard brick masonry
is higher than that of normal brickwork. Apart from the higher unit strength of non-
standard large bricks, the reduction of the number of vertical mortar joints may be the
other reason for this increase in compressive strength. Even though the brickwork with
non Standard bricks showed higher compressive strength, flexural and shear strengths
of normal brickwork were higher. Figures 7 and 8 clearly show the flexural strength
results for one-brick thick and half-brick thick walls constructed with normal bricks
and non-standard large bricks.

One brick thick wall Flexural

Strengths
0.5
N/mm2

0.4
0.3
0.2
0.1
0
Parallel to bed joint Perpendicular to
bed joint
Standard Non Standard

Fig. 7. Flexural strength of one brick thick wall panels

Half brick thick wall Flexural

Strengths
0.8
0.6
0.4
0.2
0
Parallel to bed Perpendicular
joint to bed joint

Standard Non Standard

Fig. 8. Flexural strength of half brick thick wall panels

498 V. L. W. Seneviratne and D. Nanayakkara

5 Conclusion

According to the test results, when considering the compressive strength of brickwork,
the non-standard large bricks can be selected as the better option. However, flexural
strength of normal brickwork is higher than that of brickwork made with large bricks.
Therefore, non-standard bricks would be ideal for vertical load bearing walls. When
walls are subjected to lateral loads, the better option would be normal bricks.
In this study, the main focus was on construction of one-brick thick masonry walls,
as the bond patterns were different in walls constructed with two brick sizes. From the
results (Table 5) it can be observed that the difference in bond pattern has not caused a
significant influence on compressive strength of brickwork. However, a considerable
influence of bond pattern on flexural strength can be seen in normal and non-standard
brickwork walls.
The English bond is used to construct one-brick thick walls with normal bricks, for
which the bricks are laid in two different directions alternatively. However, when
header bond pattern is used in non-standard bricks, all bricks are laid in one direction.
When bricks are laid only in one direction, resistance to flexure is reduced due to the
weak bond pattern. Therefore, the flexural strength of brickwork made with non-
standard bricks is low.
Even though brickwork constructed with non-standard bricks does not provide
higher flexural and shear strength, it will enhance the speed of construction and reduce
the required mortar volume due to the reduced number of mortar joints. Therefore,
these non-standard bricks can be used for masonry construction where the walls are not
subject to very high lateral loads.

Acknowledgement. The authors wish to acknowledge the financial support provided by the
Department of Civil Engineering, University of Moratuwa. Support from all the laboratory
technical staff members is also acknowledged.

References
British Standards Institute (1992) Code of practice for use of masonry—Part 1: Structural use of
unreinforced masonry. BS 5628-1
British Standards Institute (1999a) Methods of test for masonry—Part 1: Determination of
compressive strength. BS EN 1052-1
British Standards Institute (1999b) Methods of test for masonry—Part 2: Determination of
flexural strength. BS EN 1052-2
British Standards Institute (2002) Methods of test for masonry—Part 3: Determination of initial
shear strength. BS EN 1052-3
Jayasinghe C (2007) Comparative performance of burnt clay bricks and compressed stabilized
earth bricks and blocks. Eng J Inst Eng Sri Lanka 40(2):33–40
Sri Lanka Standards Institution (1978) Specifications for common burnt clay building bricks.
SLS 39
 Sustainable Maintenance of Steel Truss
Bridges: A Conceptual Framework

Sudath C. Siriwardane(&)

Department of Mechanical and Structural Engineering and Materials Science,

University of Stavanger, 4036 Stavanger, Norway
[email protected]

Abstract. This paper proposes a conceptual framework for the essential main-
tenance of truss bridges based on an accurate remaining fatigue life estimation
technique. The framework mainly consists of structural appraisal based critical
members and connections identification, remaining fatigue lives estimation of
critical connections and time dependent member replacement or strengthening
scheme. In the stage of critical member identification, detailed structural
appraisal has to be conducted giving priority to visual inspection for condition,
FE analysis, and material testing, experimental static and dynamic load testing.
Accurate fatigue life estimation techniques of both members and connections
consist of measured stress histories and recently developed new fatigue models.
A verification of the proposed framework is conducted by comparing the pre-
dicted replacement scheme with a sequence of experimental failure and fatigue
life of selected test specimens. Finally, the proposed framework was applied to a
case study railway bridge to obtain time dependent member replacement scheme
and obtained results were compared with previous estimations. Hence, validity
and merits of the proposed framework were confirmed.

Keywords: Essential maintenance
Fatigue life
Truss bridges

Member replacement

1 Introduction

Today, many of the structures in the world are getting old and a very large existing
stock of civil infrastructures is in need of maintenance, rehabilitation or replacement.
Considering the management of structures in terms of maintenance, member replace-
ment had a wide acceptance during this period (Kong and Frangopol 2004). However,
there are few studies regarding the member replacement based maintenance guidelines
for railway bridges. Most of them are especially based on rating factors against failure
(rating of load carrying capacity). Since railway bridges are most liable to fail due to
the effect of fatigue, remaining fatigue life based maintenance frameworks illustrates
more assurance than rating factor based policies (Network Rail 2001). However, very
few studies have only been concentrated to remaining fatigue life based member
replacement frameworks. There the remaining fatigue life estimation technique is
generally based on a combination of specification load given stress histories, Miner’s
rule and design fatigue curve. But the experiences from engineering practices have

© Springer Nature Singapore Pte Ltd. 2020

R. Dissanayake and P. Mendis (Eds.): ICSBE 2018, LNCE 44, pp. 499–509, 2020.
https://doi.org/10.1007/978-981-13-9749-3_44
500 S. C. Siriwardane

indicated that fatigue analysis based on specification loads and distribution factors
usually underestimates the remaining fatigue life of existing bridges by overestimating
the live load stress ranges (Spyrakos et al. 2004). Further, Miner’s rule does not
properly take account of the loading sequence effect (Suresh 1998; Siriwardane et al.
2008a, b). Therefore, there is a doubt in utilizing the remaining fatigue life based
available frameworks/approaches for bridges.
To overcome this problem, this paper proposes a conceptual framework for essential
maintenance of truss bridges based on an accurate remaining fatigue life estimation
technique. Initially, paper describes the proposed framework in detail. Then, the veri-
fication of the proposed framework is confirmed. Finally, the proposed framework is
applied to a case study railway bridge to obtain the time dependent member replacement
scheme. The obtained results are compared with previous estimations.

2 Proposed Framework

A proposed maintenance framework for truss bridges is discussed in this section. This
is a member replacement based essential maintenance framework. It consists of two
major parts such as identification of critical members/connections and member
replacement/strengthening scheme. The identification procedure of the critical mem-
bers is dependent on the remaining fatigue life of each member since railway bridges
are most liable to fail due to the effect of fatigue. The proposed remaining fatigue life
estimation technique is especially dependant on structural appraisal-based stress
evaluation and recently developed damage model. Flow chart of the proposed frame-
work is shown in Fig. 1.

Essential Maintenance Strategy

Structural appraisal and stress evaluation

Remaining fatigue life estimation of members

Identification of critical members/ connections

Remaining fatigue life estimation of critical

connections

Member replacement/ strengthening scheme

Member replacement

Fig. 1. Flow chart of proposed maintenance framework

 Sustainable Maintenance of Steel Truss Bridges: A Conceptual Framework 501

2.1 Structural Appraisal and Stress Evaluation

In remaining fatigue life estimation, it is essential to determine the stress ranges gen-
erated by the passage of trains over the bridge. Therefore, it is a requirement to know
the stress cycles of all the critical members for trains which are included in present and
past timetables.
Initially, a condition survey has to be carried out to assess the present geometric
condition and damages. Generally, it consists of a detailed visual examination, in-situ
measurements of each component of the bridge and non-destructive field examinations.
Then, laboratory tests are carried out to determine the current mechanical properties
and chemical composition of the bridge materials. Static and dynamic load testing is
recommended as the next major step to study the real behavior of the bridge under
various load combinations. The obtained results are used to develop a proper analytical
model and further assists in evaluating actual dynamic factors of each structural
component. Finally, the bridge is subjected to finite element (FE) analysis under test
and actual loadings to determine stresses and deflections, as well as variations of
stresses under moving loads. Material properties which are obtained through laboratory
tests and current geometric properties obtained from condition assessment are applied
to the FE model for more realistic outputs. The validation of the FE model has to be
done by comparing the results from analysis with those from field-tests. The FE model
which gives better comparison to load test results is nominated as “validated analytical
model”. Hence, the validated analytical model is used to obtain past and present static
stress histories due to passage of trains specified by the owner.
Due to the dynamic effect of moving trains, the actual working stresses should be
higher than the analytical static stress. Therefore, dynamic factors are used to multiply
the static stress to get the service stresses of each member. Finally, the stress histories
have to be converted into stress ranges using the rainflow counting method (Downing
and Socie 1982). The described stress evaluation method is briefly summarized as
shown in Fig. 2.

2.2 Remaining Fatigue Life Estimation of Members

The recently developed sequential law (Mesmacque et al. 2005; Siriwardane et al.
2007, 2008a, b) is recommended to use for remaining fatigue life estimation, instead of
Miner’s rule, which was used as the fatigue theory in previous methods. A detailed
description of the damage stress model and the definition of damage indicator, Di is
described in the corresponding paper (Siriwardane et al. 2007, 2008a, b). Here, only the
concept of new damage indicator is motioned for comprehension. Suppose a compo-
nent is subjected to certain stress amplitude or stress range ri for ni number of cycles at
load level i and Ni is the fatigue life (failure number of cycles) corresponding to ri .
Here, the residual life at load level i can be obtained as (Ni − ni). The stress rðiÞeq which
corresponds to the failure life (Ni − ni) is named as ith level damage stress amplitude or
stress range (otherwise can be introduced as stress amplitude or stress range relevant to
the residual life). Hence, the new damage indicator, Di is stated as,
502 S. C. Siriwardane

rðiÞeq  ri
Di ¼ ð1Þ
ru  ri

where ru is the intercept of the Wöhler curve with the ordinate at one-quarter of the
first fatigue cycle. Furthermore, it can be stated that, ru is the ultimate tensile strength
amplitude or range for rotating bending test-based S-N curves and it is the ultimate
shear strength amplitude or range for torsional fatigue test-based S-N curves.

Truss Bridge

Condition Material Load

Survey Testing Testing

Present geo- Current me- Present

metric details, chanical proper- stresses, deflec-
damages and ties and chemi- tions, accelera-
corrosions cal composition tions and dy-
of materials namic factors

FE Analysis

Validated Analytical Model

Past /Future Stress Cycles

Rainflow Counting
Method

Stress Ranges

Fig. 2. Flow chart for the past/future stress evaluation method

At the first cycle, the damage stress amplitude or range rðiÞeq is equal to applied
stress ri and corresponding damage indicator becomes Di = 0. According to the pro-
posed methodology, current damage has to be then transformed to the next load level.
Therefore, at the last cycle, the damage indicator becomes Di = 1 when rðiÞeq is equal
to ru Therefore, the damage indicator is normalized to one (Di = 1) at the fatigue
failure of the material and the same procedure is followed until Di = 1. Here, the
defined fatigue failure is the time taken for the occurrence of the first through-thickness
 Sustainable Maintenance of Steel Truss Bridges: A Conceptual Framework 503

crack at the location of maximum stress of the structural component. In the case of
railway bridge components, it can probably be taken as the time taken for initiation of
crack near a connection (rivet or bolt).
To capture the fatigue damage due to secondary stresses near the riveted connection
or discontinuities, detail class (BS 5400 1980) of riveted connection based Wöhler
curves are considered for life estimation. But, chosen fatigue curve only describes
stress ranges, which are corresponding to more than ten thousands of failure cycles
(usually called the partially known Wöhler curve). In the case of the sequential law it is
essential to use the Wöhler curve for the full range of the number of cycles. Therefore,
the chosen partially known Wöhler curve has to be transferred to fully known Wöhler
curve by using Kohout and Vechet Wöhler curve modeling technique (Kohout and
Vechet 2001).

2.3 Identification of Critical Members/Connections

The members which have the lowest remaining fatigue life of each member set (set of
members which has the same load capacity) is called as the “critical members” in this
study. Generally, these members are to be subjected more attention in the member
replacement based maintenance. From the previously obtained remaining fatigue lives
(Sect. 2.2), one can easily identify these critical members. The connections which are
joined to the previously obtained critical members (Sect. 2.2) are termed as “critical
connections”.

2.4 Remaining Fatigue Life Estimation of Critical Connections

The stress concentration effect in connections between the primary members of bridges
was found to be one of main reasons for fatigue damage (Imam et al. 2005; Fisher et al.
1980). Most such connections are subjected to multiaxial fatigue. To capture this effect
at connections or discontinuities, detail class (BS 5400 1980, Eurocode 3) of con-
nection based Wöhler curves are considered in the previous life estimation. However,
the variation of real rotational fixity, clamping force and geometry at the connection
causes changes in the real stress distribution at the connections. Such changes may
result in giving over or under predictions to estimated fatigue life (in Sect. 2.2) of the
corresponding member (which joined to the connection). As a result, replacement of
members based on previously determined remaining lives (Sect. 2.2) may not be an
appropriate maintenance procedure. Therefore, replacement of members based on
fatigue lives of critical connections is found to be a more appropriate strategy. This
section describes the methodology to estimate remaining fatigue life of such
connections.
Initially, all the critical connections should be investigated non-destructively to
determine their current condition. This is conducted using different types of tests such
as X-ray, ultrasonic, magnetic particle, radiographic examinations except visual
examination. The connections which do not illustrate significant change from the initial
state or condition are not subjected any unexpected stress concentration. Other con-
nections which have been subjected to significant change may have to be assessed for
remaining fatigue life sense to determine their degree of criticality.
504 S. C. Siriwardane

The remaining fatigue lives of the connections, which are not subjected to signif-
icant deviation from the initial condition, is finalized as same as the lowest remaining
fatigue life of the member that is joined to the particular connection. The remaining
fatigue lives of other connections, where the conditions have been significantly
changed, are to be evaluated based on the current geometric condition, secondary stress
distribution etc. The latter part of this section describes multiaxial stress-based fatigue
life estimation procedure of structural connections.
This procedure is especially based on the newly proposed multiaxial fatigue model.
The considered failure mechanism (damage process) is in the mesoscopic scale
and detailed descriptions are shown in the authors’ previous work (Siriwardane et al.
2008a, b). Initially, the accumulated plastic strain per each stabilized cycle,

4 2k   kmax  kmin
s ¼ pffiffiffi
epc ð2Þ
3 c

has to be obtained from finite element employed secondary stress analysis of the
connection or part of the member (sub model). Where c ¼ b þ 2g. The b and g are the
mesoscopic linear hardening modulus and the shear modulus respectively. The k* is the
radius of the smallest hypersphere which contains the entire history of the macroscopic
deviatoric stress amplitude of the stabilized cycle. The kmax and kmin are the maximum
and the minimum values of mesoscopic yield stresses that can be reached during the
loading cycle. Then, the fatigue life is calculated from the new damage indicator,

ðepc
s ÞðiÞeq  ðes Þi
pc
Di ¼ ð3Þ
ðepc pc
s Þu  ðes Þi

where ðepc
s Þi is accumulated plastic meso-strain per stabilized cycle of ni number of
cycles at load level i. The Ni is the fatigue life (number of cycles to crack nucleation)
corresponding to ðepc
s Þi can be estimated from,

n
N ¼ A ðepc
s Þ ð4Þ

where n and A are material parameters to be determined from fatigue tests. The
s Þiq , which corresponds to the failure life (Ni − ni)
accumulated plastic meso-strain ðepc
is named as i level damage accumulated plastic meso-strain. The ðepc
th
s Þu is the accu-
mulated plastic meso-strain which corresponds to one-quarter of first fatigue cycle.
According to the proposed methodology, current damage has to be then transformed to
the next load level. Similar to the previous sequential law in Sect. 2.2, here too, the
damage indicator is normalized to one (Di = 1) at the fatigue failure of the material and
the same procedure is followed until Di = 1.

2.5 Member Replacement/Strengthening Scheme

The lowest remaining fatigue life of the connections describes the remaining fatigue
life of the bridge. Once the age of the bridge reaches this value (when bridge life
becomes zero), it is advisable to replace the corresponding critical member with a new
 Sustainable Maintenance of Steel Truss Bridges: A Conceptual Framework 505

member with longer fatigue life. At the same time, the associated connection is also
recommended to be strengthened. After this essential repair, it is recommended to
obtain a new sequence for future member replacement by following the same procedure
from the beginning. This type of maintenance frameworks extends the service life of
the (fatigue capacity) bridge in the safest manner.

3 Verification of Proposed Framework

In this section, the experimental (real) failure life-based member replacement sequence
is compared with the proposed framework predicted scheme of five specimen members.
The experimental fatigue lives were obtained in the high cycle fatigue regime under
variable amplitude combined bending and torsion loading (Marciniak et al. 2008). The
comparison between the predicted member replacement scheme and the experimental
results-based scheme are shown in Table 1. In addition, it has been compared with two
previous frameworks-based member replacement schemes as shown in Table 1. The
first scheme is based on load rating factors (AASHTO 1990; Bridge Diagnostics 2002).
The second approach is based on remaining fatigue life estimation of members
(Caglayan et al. 2009). Here, Miner’s rule is considered as the fatigue law (damage
model). The Table 1 shows that there is better agreement between the proposed
framework-based sequence than the previous frameworks-based sequences. This ver-
ification reveals the validity of the proposed maintenance framework in predicting the
member replacement scheme in variable amplitude proportional loading conditions.

Table 1. Comparisons of member replacement schemes

Designation of Experimental Previous approach Previous approach Proposed
test specimen 1 (AASHTO 1990) 2 (Caglayan et al. framework
2009)
Service Sequence Service Sequence Service Sequence Service Sequence
life (s) life (s) life (s) life (s)
NWL 4 42795 1 27265 1 58243 1 49628 1
KWL 4 90705 2 44325 2 93497 2 78239 2
KWL 2 98948 3 68573 4 202229 5 116975 4
KWL 8 123275 4 158623 5 162519 4 109076 3
NWL 1 130630 5 48329 3 147265 3 123568 5

4 Case Study: Determination of Member Replacement

Scheme for a Rail-Way Truss Bridge

The selected bridge is one of the major railway bridges in Sri Lanka spanning 160 m
(Fig. 3). It is a six span-riveted bridge with double lane rail tracks having warren type
semi through trusses, supported on cylindrical piers. The bridge deck is made of
wrought iron and the piers are made of cast iron casings with infilled concrete. The
bridge was constructed in 1885. The proposed framework is used to determine the
suitable member replacement scheme for the bridge.
506 S. C. Siriwardane

Fig. 3. General views and identified corroded locations of the bridge

The condition survey reveals that no visual cracks were observed in any component
of the super structure. The obtained values from material testing for elastic modulus,
yield strength, ultimate strength in tension, fatigue strength and density are 195 GPa,
240 MPa, 383 MPa, 155 MPa and 7600 kg/m3 respectively. Static and dynamic load
testing were performed to study the real behavior of the bridge under various load
combinations. The Bridge deck was analyzed using the finite element (FE) method
employed general-purpose package SAP 2000. The FE model validations are per-
formed by comparing analytical results of the FE model and the measurement of the
actual bridge. Then, the stress histories were obtained and converted into stress ranges
using rainflow counting method. The proposed method (Sect. 2.2) was used to cal-
culate the remaining fatigue life of members. The riveted connections were classified as
class Wrought-iron (WI), which is proposed by the UK railway assessment code
(Network Rail 2001). The identified critical members which exhibits the lowest
remaining life of each member set are shown in Table 2. The corresponding critical
connections are also tabulated in Table 2. Designations of the members and connec-
tions of the bridge are schematically shown in Fig. 4. The fatigue damage of the critical
joints is evaluated based on the state of stress due to release of clamping force while all
the rivets of these two connections have no clamping force. The obtained remaining
fatigue lives for the all the critical connections are shown in Table 3.

Table 2. Remaining fatigue lives for critical members and corresponding critical connections
Designation of members Corresponding connections Remaining fatigue life (years)
TMB3 TB4 323
TMB5 TB6 165
TMB6 TB7 169
CG4, 5 D15-45, D16-46 12
ST 16, 26, 36, 46 D17-47 13
TDT1 TT1, TB2 191
TDT2 TT2, TB3 171
TDT3 TT3, TB4 138
TDT4 TT4, TB5 162
 Sustainable Maintenance of Steel Truss Bridges: A Conceptual Framework 507

Fig. 4. Designations of (a) Members: i. Main truss girder, ii. Horizontal bridge deck;
(b) Connections: i. Main truss girder, ii. Horizontal bridge deck

Table 3. Remaining fatigue lives for critical connections

Bridge component Designation of Corresponding Remaining fatigue
critical connection critical member life (years)
Main girder bottom TMB3 TB4 323
chord
Main girder bottom TMB5 TB6 165
chord
Main girder bottom TMB6 TB7 169
chord
Cross girders CG4, 5 D15-45, D16-46 12
Stringers ST 16, 26, 36, 46 D17-47 13
Truss diagonal TDT1 TT1, TB2 45
(tension member)
Truss diagonal TDT2 TT2, TB3 171
(tension member)
Truss diagonal TDT3 TT3, TB4 24
(tension member)
Truss diagonal TDT4 TT4, TB5 162
(tension member)

The member replacement sequence and the time of replacement are then deter-
mined from the remaining fatigue life of each critical connection. These are shown in
Table 4. The determined scheme was also compared with two previous approaches-
based member replacement schemes (similar schemes considered in Sect. 3 in Table 1)
as shown in Table 4. The comparisons revealed that the proposed framework-based
scheme deviates from previous approaches-based schemes.
508 S. C. Siriwardane

Table 4. Comparisons of member replacement schemes

Member should be Previous approach 1 (AASHTO Previous approach 2 (Caglayan et al. Proposed framework
replaced (connection) 1990) 2009)
Rating Replacement Sequence Service Replacement Sequence Service Replacement Sequence
factor years life years life years
TMB3 3.69 110 4 305 300 5 323 320 6
TMB5 2.70 50 3 156 150 3 165 160 4
TMB6 2.48 35 1 157 150 3 169 160 4
CG4-5 5.63 190 6 20 20 1 12 12 1
ST16-46 4.84 145 5 24 20 1 13 12 1
TDT1 3.39 110 4 179 170 4 45 45 3
TDT2 2.72 50 3 168 170 4 171 170 5
TDT3 2.61 40 2 131 130 2 24 20 2
TDT4 2.55 40 2 132 130 2 162 160 4

5 Conclusions

A remaining service life-based conceptual framework was proposed for essential

maintenance of truss bridges. The verification of the proposed conceptual framework
was conducted by comparing the predicted replacement scheme with a sequence of
experimental failure and fatigue life of selected test specimens. The proposed frame-
work was further utilized to determine the member replacement scheme of a railway
truss bridge. It was shown that the proposed framework gives a much safer and sus-
tainable member replacement scheme for railway truss bridges where detailed stress
histories are known. Further verifications of the proposed framework are currently
under way.

Acknowledgements. The author wishes to express their sincere gratitude to Senior Professor
Ranjith Dissanayake, Professor Mitao Ohga, Emeritus Professor M.P Ranaweera and the team of
experts who worked in the Sri Lankan Railway Bridge project for their great advice which laid
the foundation for this research. The kind support given by the Sri Lanka Railways (SLR) is also
appreciated.

References
AASHTO (1990) Guide specification for fatigue evaluation of existing steel bridges,
Washington. D.C
Bridge Diagnostics (2002) Load testing, evaluation and rating for railroad flatcar bridge spans
over Trinty river, Redding, California, Bureau of Reclamation, Colorado, USA
BS 5400, Part 10 (1980) Code of practice for fatigue, Steel Concrete and Composite Bridges.
British Standard Institutions, UK
Caglayan BO, Ozakgul K, Tezer O (2009) Fatigue life evaluation of a through-girder steel
railway bridge. Eng Fail Anal 16:765–774
Downing SD, Socie DF (1982) Simple rainflow counting algorithms. Int J Fatigue 4(1):31–40
Fisher JW, Yen BT, Wang D (1980) NCHRP Report No 302: Fatigue and fracture evaluation for
rating riveted bridges. Transportation Research Board, National Research Council, Wash-
ington, DC
 Sustainable Maintenance of Steel Truss Bridges: A Conceptual Framework 509

Imam B, Righiniotis TD, Chryssanthopoulos MK (2005) Fatigue assessment of riveted railway

bridges. Int J Steel Struct 5(5):485–494
Kong JS, Frangopol DM (2004) Prediction of reliability and cost profiles of deteriorating bridges
under time and performance-controlled maintenance. J Struct Eng 130(12):1865–1874
Kohout J, Vechet S (2001) A new function for fatigue curves characterization and its multiple
merits. Int J Fatigue 23(2):175–183
Marciniak Z, Rozumek D, Macha E (2008) Fatigue lives of 18G2A and 10HNAP steels under
variable amplitude and random non-proportional bending with torsion loading. Int J Fatigue
30(5):800–813
Mesmacque G, Garcia S, Amrouche A, Rubio-Gonzalez C (2005) Sequential law in multiaxial
fatigue, a new damage indicator. Int J Fatigue 27(4):461–467
Network rail RT/CE/C/025 (2001) Rail track line code of practice: the structural assessment of
under bridges, Railtrack
Spyrakos CC, Raftoyiannis IG, Ermopoulos JCh (2004) Condition assessment and retrofit of
historic steel-truss railway bridge. J Const Steel Res 60:1213–1225
Siriwardane SC, Ohga M, Dissanayake R, Taniwaki K (2007) Different approaches for remaining
fatigue life estimation of critical members in railway bridges. Int J Steel Struct 7(4):263–276
Siriwardane SC, Ohga M, Dissanayake R, Taniwaki K (2008a) Application of new damage
indicator-based sequential law for remaining fatigue life estimation of railway bridges.
J Const Steel Res 64(2):228–237
Siriwardane SC, Ohga M, Kaita T, Dissanayake R (2008b) Mesoscopic plastic strain based new
fatigue model and its application to a riveted connection. In: Proceeding of international
conference on continuum mechanics, nanoscience and nanotechnology, Sri Lanka, pp 77–85
Suresh S (1998) Fatigue of materials. Cambridge University Press, Cambridge
Sustainable Manufacturing
 Properties of Portland Blast Furnace Slag
Cements Affecting Their Autogenous
Shrinkage Behaviors

S. Suntharalingam(&) and Y. Takahashi

The University of Tokyo, Tokyo, Japan

[email protected]

Abstract. In Japan, Portland blast furnace slag cements that contain 30–60% of
blast furnace slag (BFS) are classified in the slag cement class B in JIS R 5211.
Even if slag cements are classified in a category, different autogenous shrinkage
behaviors are observed among them. In this study, class B slag cements are
collected from eight different plants in Japan. Physical and chemical properties
of slag cements are investigated to find the cause of different shrinkage behavior.
Physical properties of slag cement are expressed in terms of surface area, mean
particle size and particle size distribution. The experimental results showed that
the collected slag cements can be differentiated into two different groups based
on physical properties. However, different behaviors of autogenous shrinkage
cannot be explained, because slag cements, which contain finer particles, have
not displayed significant differences in the shrinkage of mortar specimen.
Therefore, chemical properties of slag cement could be the main reasons behind
the different shrinkage behavior among slag cements. To compare the chemical
property of slag cements, the phases of mineral (C3S, C2S, etc.) and content of
cement and slag mineral (%) in slag cements are analyzed by thermal gravi-
metric analysis (TGA) and X-ray diffraction (XRD) Rietveld analysis. The
results showed that in each plant, different types of BFS are used for producing
the slag blended cements, and different quantity levels of slag minerals are found
in BFS used in each cement plant. In slag cements classified as Class B in JIS R
5211, different physical and chemical properties are observed. However, the
chemical properties of BFS in slag cement mainly influences the shrinkage
behavior of slag cement. If the user of BFS cement is concerned about the
shrinkage behavior of the concrete, the chemical composition of BFS should be
checked to ensure that the concrete has sufficient resistivity against shrinkage.

1 Introduction

Blast furnace slag (BFS) is a by-product of the steel industry that is considered a
sustainable alternative to Ordinary Portland cement (OPC). Concrete with slag cement
is known for good durability properties and long-term strength development due to its
dense microstructure. Japan is one of the major users of blended cements containing
BFS and the common slag blended cement is specified as class B in JIS 5211 and it
contains blast furnace slag in the range of 30% to 60%.

© Springer Nature Singapore Pte Ltd. 2020

R. Dissanayake and P. Mendis (Eds.): ICSBE 2018, LNCE 44, pp. 513–525, 2020.
https://doi.org/10.1007/978-981-13-9749-3_45
514 S. Suntharalingam and Y. Takahashi

However, even if supplied class B in JIS 5211 Slag cement concrete is satisfying
the demand, different behaviors such as setting time, heat of hydration, shrinkage, etc.
in concrete are observed, and some previous research (Sagawa et al. 2010; Nito et al.
2005; Ekaputri et al. 2010) has also reported that commercial cements, classified in slag
cement class B with satisfactory existing allowable range of cement properties, are
exhibiting different behavior in autogenous shrinkage.
Therefore, slag cements, which are commercially available in Japan, should be
differentiated through the physical and chemical properties, to determine the cause of
different shrinkage behavior among slag cements; additionally, an investigation should
be conducted into what properties of slag cement mainly influence the behavior of
shrinkage of concrete.
In this study, the physical properties of slag cements are analyzed through a laser
diffraction test; to compare the chemical properties of slag cements, thermal gravimetric
analysis (TGA) and X-ray diffraction (XRD) Rietveld analysis are conducted.

2 Properties, Behavior of Compressive Strength

and Autogenous Shrinkage of OPC and Slag Cements
2.1 Properties of OPC and Slag Cements
In this experimental study, eight different commercially use slag cements that are
available in Japan, are collected and they are coded as sca, scb, scc, scd, sce, scf, scg,
sch and opc which is used as a check system. Table 1 shows that, the percentages of
BFS and physical properties of OPC and slag cements.

Table 1. Properties of OPC and slag cements.

No Symbol Cement type BFS [%] Density [kg/m3] Surface area [cm2/g]
1 opc Ordinary Portland 0 3.15 3430
cement, JIS R 5210
2 sca Slag cement class B, 40–45 3.02 3620
3 scb JIS R 5211 40–45 3.04 3750
4 scc 40–45 3.04 3780
5 scd 40–45 3.04 3820
6 sce 40–45 3.04 4140
7 scf 40–45 3.04 4170
8 scg 40–45 3.05 3790
9 scg 40–45 3.04 3860

2.2 Compressive Strength of OPC and Slag Cements

The compression test results are from each data sheet of cement plant, and the
experimental procedure was followed according to the JIS B 5201 for all samples.
Cube mortar specimens, with 40
40
40 mm dimensions, were casted, to test the
 Properties of Portland Blast Furnace Slag Cements 515

compressive strength of OPC, and slag cements and fresh mortar were prepared by 1: 3:
0.5 mass ratio between cement, sand and water, respectively. All casted specimens
were cured under a constant temperature at 20 °C and a relative humidity of 50%. All
specimens were demolded after 24 h, and then immersed into water at a constant
temperature of 20 °C. Compressive strength tests were conducted at 3, 7 and 28 days of
age.
Figure 1 shows the compressive strength at 3, 7 and 28 days of collected OPC and
slag cements. Different compressive strength developments are observed among the
OPC and slag cements at 3 and 7 days. However, it is interesting to note that at 28 days,
the OPC and Slag cements show an almost equal compressive strength development.

Fig. 1. The compressive strength of OPC and slag cements.

2.3 Autogenous Shrinkage of OPC and Slag Cements

Cylindrical mortar specimens, of ø 5
10 cm, were used to measure autogenous
shrinkage and fresh mortar (Takahashi and Suntharalingam 2018). These were prepared
with 35% of water to binder ratio and sand where the maximum aggregate size and
density were 2 mm and 2.63 g/cm3, respectively, as fine aggregate for 40% volume of
fresh mortar. To improve the workability of the fresh mortar, a water-reducing agent of
0.55% cement weight was used (BASF 6500 XD2). Strain gauges of the embedded
type (KM-30-120-H2-11) are used to measure the autogenous shrinkages of different
OPC and slag cement mortar specimens, under sealed conditions at a constant tem-
perature 20 °C (Ekaputri et al. 2010). The results are processed to eliminate the effects
of the thermal expansions, through the proposed model of early age heat expansion
coefficients by Holt (2001) and temperature histories at the early ages.
The results showed that slag cements had larger autogenous shrinkage than ordi-
nary Portland cement. Slag cements, which are classified in a particular category, show
differences in level of autogenous shrinkage behavior, with the results being presented
in Fig. 2.
516 S. Suntharalingam and Y. Takahashi

Fig. 2. Autogenous shrinkage of OPC and slag cements.

It is clearly shown that, even if collected slag cements are classified in a category in
JIS R 5211, different autogenous shrinkage behaviors among slag cements can be
observed, while almost equal compressive strength among OPC and slag cements at 56
days of age. Therefore, physical and chemical properties of slag cements were inves-
tigated to find the cause of different shrinkage behavior.

3 Physical Properties of OPC and Slag Cements

The particle size and distributions of OPC and slag cements were measured using a
laser diffraction technique, through the SALD-3000 J particle Size Analyzer in the
range of particle size from 0.1 µm to 2000 µm. The mean particle size and the rela-
tionship between surface area and mean particle size of OPC and Slag cements are
shown in Figs. 3 and 4, respectively.

Fig. 3. Mean particle size of OPC and slag cements.

 Properties of Portland Blast Furnace Slag Cements 517

Fig. 4. Relationship between surface area and mean particle size of OPC and slag cements.

The results show that the mean particle size and surface area of the collected slag
cement particles are different; when the surface area of slag cement particles with their
mean particle size is compared, sce and scf are significantly deviate from other slag
cement, as they have a smaller particle size and a greater surface area among collected
slag cement. However, there are no significant differences in the autogenous shrinkage
behavior of sce and scf from other slag cements.
It is evident that the physical properties of Class B slag cement in JIS 5211 are not a
major influence factor for showing the significant difference in autogenous shrinkage
among slag cements. Therefore, the chemical properties of slag cement could be the
main factor. Chemical properties of slag cement are investigated by the thermal
gravimetric analysis (TGA) and the X-ray diffraction (XRD) Rietveld analysis for
differentiation of collected slag cements.

4 XRD Rietveld and TGA Analysis

The Shimadzu XRD 6100 machine was used to determine the diffraction profile pattern
for OPC and slag cements sample in XRD test, and the measurement of XRD was
conducted using an internal standard substance with an average particle diameter of
3 lm. Corundum (a1 Al2O3) was added by 10% of the sample. In the XRD mea-
surement conditions were as follows: target Cu-Ka, tube voltage 45 kV, tube current
40 mA; range of scanning from 5 to 70°; 2h step width 0.02. Rietveld analysis is
distributed by Siroquant version 3 software. Micro absorption was corrected with the
particle diameter of quantitative minerals being 10 lm.

4.1 Quantitative Value of OPC Minerals in Slag Cement

In the quantitative value of cement mineral in slag cement, the following minerals were
subjected for Rietveld analysis, C3S (Belov), C3S (mono.), C2S, beta, C2S, alpha, C3A,
518 S. Suntharalingam and Y. Takahashi

(cubic), C3A (ortho.), C4AF, Periclase, Mayenite, Gypsum, Calcite 2, Portlandite,

Bassanite of cement mineral and as internal standard substance corundum (a1 Al2O3).
Obtained quantitative values for cement mineral in Rietveld analysis were corrected
by the internal standard substance amount and ignition loss according to Eq. (1).

WCX ¼ ðWX =WC Þ * 10 * ð1 þ Ig=100Þ ð1Þ

Where
WCX = Corrected quantitative value (%)
Wx = Quantitative value of mineral before correction (%)
WC = Quantitative value of corundum (%)
Ig = Ignition loss between 30 °C and before starting the crystallization effect of slag
materials (%)
In determining ignition loss by the TGA test, mass losses are measured between 30
and 1000 °C and they can be used for Ordinary Portland cement that does not contain
BFS. However, when heating the slag cement in the range between 30 and 1000 °C,
further losses or increases in weight of slag cement sample nearly above 760 °C
temperature were observed; this is because of the slag materials being in the form of
amorphous material nearly below 760 °C temperature. However, when continuously
heating slag cement above 760 °C, slag materials start to shift into their crystal forms.
Figure 5 shows that slag materials in scd start to transform from amorphous phase
to crystal form, above 760 °C. Therefore, the quantitative value of each mineral should
be corrected with ignition loss, to compare the percentage of each minerals in slag
cement before the start crystallization effect, for an accurate analysis. Here, to correct
quantitative values of each mineral, ignition loss of slag cements is calculated at a
weight loss temperature between 30 °C, and before starting the crystallization effect of
slag materials.

TGA
[%]

Temperature [oC]

Fig. 5. TGA result of scd.

 Properties of Portland Blast Furnace Slag Cements 519

Figure 6 shows that the corrected quantitative value of main cement mineral (C3S
and C2S) in OPC, which was used to blend with blast furnace slag to produce slag
cements in each plant (here, the percentage of each cement minerals were calculated
without considering the blast furnace slag portion in slag cement). The result shows
that significant differences were observed in percentages of C3S and C2S of used OPC
for producing slag cement. However, the pozzolanic reaction in slag cement is what
mainly decides the long-term behavior of autogenous shrinkage of mortar specimens.
Therefore, the content of slag material should be analyzed to understand behavior of
autogenous shrinkage.

Fig. 6. Corrected quantitative value of main mineral (C3S and C2S).

4.2 Quantitative Value of Crystallized BFS Minerals in Slag Cements

BFS fine powder is in amorphous phase in slag cement. Therefore, BFS should be
transformed into a crystal, through heat treatment for the XRD test. In a TGA test, it is
clearly observed that if heated nearly above 760 °C, the blast furnace slag starts to
obtain a crystal structure. Sagawa and Nawa (2006) proposed that, for determining the
content of BFS in slag cement, the sample should be heat-treated at 900 °C for 30 min
to crystallize slag material; as for the above crystallization conditions, a complete
crystallization of the slag, and a minimization the decomposition of the cement mineral,
can be achieved.
Due to the aforementioned heating process, there was C2S (a′L), which was
thought to have been generated by decomposition of C3S in cement mineral and blast
furnace slag fine powder changed into Gehlenite (2CaOAl2O3SiO2), Akermanite
(2CaOMgO2SiO2), Merwinite (3CaOMgO2SiO2), which are represented by
abbreviations of C2AS, C2MS2 and C3MS2, respectively. The crystal structure of the
above minerals was based on data preset in Rietveld analysis software (Siroquant
version 3).
520 S. Suntharalingam and Y. Takahashi

The following cement and crystalized slag minerals were the subject of Rietveld
analysis: C3S (Belov), C3S (mono.), C2S, beta, C2S, alpha, C3A, (cubic), C4AF,
Anhydrite, Wollastonite, Lime of cement mineral, Akermanite, Gehlenite, Merwinite,
Mayenite of crystalized slag mineral and as internal standard substance corundum (a 1
Al 203).
Obtained quantitative values of cement and crystalized slag mineral in Rietveld
analysis are corrected by the internal standard substance amount and ignition loss
according to Eq. (2).

WCX ¼ ðWX =WC Þ * 10 * ð1  Ig=100Þ ð2Þ

Where
WCX = Corrected quantitative value (%)
Wx = Quantitative value of mineral before correction (%)
WC = Quantitative value of corundum (%)
Ig = Ignition loss between starting point of the crystallization of slag materials and
end of heat-treatment (%)
The quantitative value of each slag and cement mineral in heat-treated samples (at
900 °C for 30 min) are listed with a degree of crystallization effect in slag material
between the starting temperature of crystallization (at nearly 760 °C) and the end of
heat treatment. Therefore, to accurately analyze, the quantitative value should be
corrected with ignition loss when comparing the percentage of each mineral in slag
cement, at the starting point of the crystallization effect. In this heat-treated sample
analysis, Ignition losses are calculated for correcting the quantitative value of each
mineral weight loss between the starting point of the crystallization of slag materials
and end of heat treatment.
Figure 7 shows the corrected quantitative value of crystallized slag minerals
(C2AS, C2MS2 and C3MS2) at the starting point of slag mineral crystallization (at
nearly 760 °C) through XRD Rietveld method in blended slag cements. Among slag
cements, several percentage differences are observed in each crystallized slag minerals:
sca have the higher content of Merwinite and Gehlenite, compared with other slag
cements and Akermanite in sce. Figure 8 shows the sum of corrected quantitative value
of crystallized slag minerals (C2AS, C2MS2 and C3MS2) at the starting point of
crystallization of slag mineral in slag cements: among these, several percentage
changes were observed in total crystallized slag minerals. There are two possibilities
for this difference in content of total crystallized slag mineral in slag cements, which are
(a) Mixing percentage of blast furnace slag and
(b) Chemical composition of blast furnace slag (CaO, SiO2, etc.)
 Properties of Portland Blast Furnace Slag Cements 521

Fig. 7. Content of crystalized slag minerals in slag cements.

Fig. 8. Content of crystalized slag minerals in slag cements.

The autogenous shrinkages of slag cements at 3, 7, 28 and 56 days are compared

with the sum of crystallized slag minerals (Akermanite + Gehlenite + Merwinite),
which is shown in Fig. 9. The higher the content of crystallized slag in cement, the
greater the autogenous shrinkage. The strong relationship between content of crystal-
lized slag and autogenous shrinkage can be observed at the 56 days period, because of
the pozzolanic reaction by slag mineral which will be influenced at a later stage.
Therefore, behavior differences in autogenous shrinkage among slag cements are
affected by to the level of crystallized slag mineral in heat treated slag cements.
522 S. Suntharalingam and Y. Takahashi

Fig. 9. Relationship between autogenous shrinkage and crystalized slag minerals of slag
cements.

Additionally, the types of BFS used for producing the blending slag cements in
each manufacturing plant, were analyzed in terms of crystallized slag minerals com-
position. Compositions of crystallized slag minerals in blended slag cements is shown
in Fig. 10, and they are calculated through the percentage of each crystallized slag
mineral in total crystallized slag mineral by Eq. (3).

PS ¼ ½WS =RWS Þ  100 ð3Þ

Where
PS = The percentage of mineral in blast furnace slag (%)
WS = Quantitative value of each slag mineral in slag cement (%)
RWS = Sum of Quantitative value of slag mineral in slag cement (%)
From this analysis, different compositions of crystallized slag minerals are observed
in slag cements from each plant. sca has the higher percentage of Merwinite among
crystallized slag minerals, compared to other slag cements. Similar types of slags
were used for blending with OPC to produce scb and scd, and, in sce and scf, nearly
the same type of slag without Gehlenite mineral were observed in heat treated samples.
Figure 11 shows the relationship between crystallized slag minerals of heat-treated
slag cement and composition of slag material (CaO, SiO2, MgO, etc.) (Sagawa and
Nawa 2012). The composition of crystallized slag minerals in heat treated samples is
dependent on the composition of slag in cement. It is evident that different types of BFS
are used in each plant for production of slag cement.
 Properties of Portland Blast Furnace Slag Cements 523

Fig. 10. Types of blast furnace slag used in slag cements.

Merwinite
Gehlenite
[%]

[%]

Al2O3 [%] (CaO+MgO)/SiO2 [%]

(a) (b)

Fig. 11. (a) & (b): The relationship between crystallized slag minerals and composition of slag
material (Sagawa and Nawa 2012).

From studies about slag cement chemical properties, different quantities and
chemical compositions of slag are observed in collected slag cements (Class B in JIS
5211). These different blast furnace chemical compositions change the Ca/Si ratio and
the density of C-S-H product among slag cements, due to the pozzolanic reaction
(Sagawa and Nawa 2010), which is shown in Fig. 12. The Ca/Si ratio and the density
of C-S-H product mainly influence the autogenous shrinkage behavior of cement, and
low densities of C-S-H produces larger shrinkage than high densities of C-S-H product,
a fact reported by Maruyama and Kurihara (2016).
524 S. Suntharalingam and Y. Takahashi

Density of C-S-H [g/cm3]

Ca/Si ratio

Hydration of BFS [%]

Fig. 12. The relationship between Ca/Si ratio and the density of C-S-H and hydration of BFS
(Sagawa et al. 2010).

5 Conclusion

In this study, slag cement (Class B in JIS 5211), which is commercially available in
Japan, is differentiated in two different groups based on their physical properties.
However, the physical properties of slag cement does not constitute a major influencing
factor for the different shrinkage behavior, because slag cements contain finer particles
that do not significantly affect autogenous shrinkage behavior. In chemical property
analysis, the different quantities and types of BFS that were used to produce slag
cements, were studied by TGA and XRD Rietveld analysis; autogenous shrinkage is
correlated more with the content of crystallized slag minerals in slag cement at 56 days
of hydration. If the user of BFS cement is concerned about the shrinkage behavior of
the concrete, the chemical composition of BFS should be checked, to ensure that the
concrete has sufficient resistivity against shrinkage.

Acknowledgments. This study was financially supported by JSPS KAKENHI Grant

No. 17H01284 and Research scholarship by Asian Development Bank-JSP.

References
Ekaputri JJ, Ishida T, Maekawa K (2010) Autogenous shrinkage of mortars made with different
types of slag cement. Proc JCI 32(1):353–358
Holt EE (2001) Early age autogenous shrinkage of concrete. Technical Research Centre of
Finland, Espoo, VTT Publications, p 446
Hoshino S, Yamada K, Hirao H (2006) XRD/Rietveld analysis of the hydration and strength
development of slag and limestone blended cement. J Adv Concr Technol 14(3):357–367
Maruyama I, Kurihara R (2016) Impact of two types of calcium silicate hydrate on drying
shrinkage of Portland cement paste. In: The 70th cement technology convention, vol 1109
 Properties of Portland Blast Furnace Slag Cements 525

Nito N, Hiroshima A, Ohtomo T, Miyazawa S (2005) Property of concrete using portland blast-
furnace slag cement b class with different slag grain and a chemical composition. Cement Sci
Concr Technol 59:231–238
Sagawa T, Nawa T (2006) Quantitative hydration analysis of blast furnace slag cement by
rietveld method. Concr Res Technol 17(3):1–11
Sagawa T, Nawa T (2010) Hydration-microstructural evolution and drying shrinkage of portland
cement-blast furnace slag system. J Struct Constr Eng 75(652):1029–1037
Sagawa T, Ishida T, Luan Y, Nawa T (2010) Hydrate composition analysis and micro structure
characteristics of Portland cement-blast furnace slag system. Cement Sci Concr Technol 66
(3):311–324
Sagawa T, Nawa T (2012) Quantitative determination of blast-furnace slag content in slag-
blended cement by combination method of heat-treatment and XRD/Rietveld analysis.
Cement Sci Concr Technol 68
Takahashi Y, Suntharalingam S (2018) Experimental study on the autogenous shrinkage
behaviors of different Portland blast furnace slag cements. In: Proceedings of the 8th
international conference the Asian Concrete Federation, pp 409–418
 Fire Performance of CFRP Strengthened
Steel I Beams Cured at Elevated Temperature

E. R. K. Chandrathilaka(&) and J. C. P. H. Gamage

Department of Civil Engineering, University of Moratuwa, Moratuwa, Sri Lanka

[email protected]

Abstract. Elevated temperature curing increases the glass transition tempera-

ture of the polymeric bond between CFRP and steel. Fire performance of an
elevated temperature cured and ambient temperature cured CFRP strengthened
steel I beam was evaluated using a numerical model. The effects of a range of
thicknesses and conductivities of insulation layers were also analyzed. Model
results are in good agreement with the theoretical results. Insulation is a com-
pulsory requirement for CFRP strengthened steel I sections. On average, a 35%
increase in fire performance was noted at elevated temperature cured CFRP
strengthened I beams compared to the ambient cured I beam.

Keywords: Glass transition temperature (Tg)
CFRP/steel bond
Fire

Steel I beams
Insulation

1 Introduction

The use of CFRP (Carbon Fiber Reinforced Polymer) in strengthening of steel struc-
tures is rapidly spreading due to its superior properties such as high strength to weight
ratio, easy installation process and corrosion resistance. However, one of the main
drawbacks of the use of CFRP in strengthening of steel structures is strength degra-
dation of CFRP/steel bond at elevated temperatures. The main cause of the strength
degradation is low glass transition temperature in epoxy adhesive (ARELDITE 420
A/B 2018).
The elevated temperature curing has been used to increase the bond performance of
CFRP/epoxy/steel bond (Nguyen et al. 2011; Bai et al. 2014; Nguyen et al. 2012).
However, the use of the methods mentioned in the literature (Nguyen et al. 2011; Bai
et al. 2014, Nguyen et al. 2012) were not practically feasible. Therefore, a better
feasible elevated temperature curing method is required. Chandrathilaka et al. (2019)
introduced the use of halogen floodlights in elevated temperature curing of
CFRP/epoxy/steel bond. The new method was compared with the traditional laboratory
experiment procedure with standard oven. It was concluded that the new method was
more effective than the standard method.
Gamage et al. (2006) have performed experimental thermo-mechanical analysis to
determine the effective thickness of insulation for the concrete/epoxy/CFRP bond.
Other than the insulation thickness, thermal conductivity of insulation and type of fire
was considered in the analysis. Imran et al. (2018) has predicted the temperature
variance of CFRP strengthened steel tubular columns and checked for the effective
© Springer Nature Singapore Pte Ltd. 2020
R. Dissanayake and P. Mendis (Eds.): ICSBE 2018, LNCE 44, pp. 526–537, 2020.
https://doi.org/10.1007/978-981-13-9749-3_46
 Fire Performance of CFRP Strengthened Steel I Beams Cured 527

thickness for the insulation. In both cases experimental thermal analysis was carried out
to validate the results of the numerical thermal analysis. However, performing exper-
imental thermal analysis is much more complex, expensive and time consuming pro-
cess. Therefore, using of theoretical equations to validate the model will be a good
alternative.
Zang and Usmani (2015) have investigated the thermal behaviour of steel I sections
in an event of fire numerically and using theoretical formulas. The performance of both
analytical methods was significantly similar at lower temperatures. When the temper-
ature exceeds 200 °C, the difference between the results of numerical thermal analysis
and theoretical thermal analysis tends to increase.
However, at temperatures less than 100 °C, the results of theoretical and numerical
thermal analysis had the same values. Zang et al. (2012) have also used different
equations to perform a sensitivity study on their accuracy and predicting applicability
to the situation.
In this study, a numerical model was developed to predict the fire performance of
CFRP strengthened steel I beam cured at elevated temperature.

2 Numerical Analysis

Transient heat analysis was performed using transient thermal tools in the software
(ANSYS mechanical user’s guide 2013). The thermal behavior of the CFRP
strengthened steel I beams were predicted for a range of environmental temperature
variations with the time. A 3-D model was developed to predict the thermal behavior of
CFRP strengthen I beam with different insulation layers with varying thermal con-
ductivities and insulation layer thicknesses.

2.1 FE Mesh and Boundary Conditions

Steel I section UC 254  254  167 was used in the FE analysis to evaluate the fire
performance. It was assumed that the CFRP strengthening was done by wrapping the
CFRP around a steel column section. A 1 m long steel I section was modelled to
evaluate the thermal behaviour of CFRP strengthened steel I section. SOLID70, 8-node
brick elements were used in each material. A programme controlled mesh was used in
the model with user input maximum mesh size between 25 mm to 50 mm with a
minimum mesh size of 1 mm. The aspect ratio was selected to be 1:1 (Fig. 1). The two
ends of steel I beam, CFRP layer and insulation were modelled as perfectly insulated as
shown in Fig. 2. Heat was introduced on the surface of insulation as a convection
(Fig. 2).
528 E. R. K. Chandrathilaka and J. C. P. H. Gamage

Fig. 1. FE mesh

Fig. 2. Boundary conditions

2.2 Material Properties

Thermal properties of the used materials in the FE analysis are listed in Table 1.

Table 1. Material properties (X-Wrap C300, ARELDITE 420 A/B, CAFCO 300, CAFCO 300
AC, CAFCO® BLAZE-SHIELD® II)
Material Thermal conductivity Specific heat Density
(W/m.°C) (J/kg.°C) (kg/m3)
Steel 60.5 511 7800
CFRP (X-Wrap C300) 5 1000 1000
Epoxy adhesive (ARELDITE 1 1000 1000
420 A/B)
Insulation 1 (CAFCO 300) 0.0839 700 240
Insulation 2 (CAFCO 300 AC) 0.078 700 240
Insulation 3 (CAFCO® BLAZE- 0.043 700 240
SHIELD® II)

2.3 Analysis Procedure

The solving was done using a computer with 8 GB RAM and i7 processing unit. The
transient thermal solver in ANSYS was used. The solving steps were applied as 20 s
 Fire Performance of CFRP Strengthened Steel I Beams Cured 529

time steps. Approximately 1 h was taken to complete the solving process. The con-
vection heat was applied as a time dependent temperature profile on the insulation
surface of CFRP strengthen I beam. The temperature of convection surface was
inserted according to ISO 834-11:2014 as shown in Eq. (1) and Fig. 3.

Tf ¼ 345:log10 ð0:133t þ 1Þ þ T0 ð1Þ

Tf = Fire temperature (°C)

t = time (s)
T0 = Initial temperature (°C)

Fig. 3. Standard fire curve (ISO 834-11:2014)

Insulation layer parameters were varied as shown in Table 2.

Table 2. Insulation layer parameters

Series Thermal conductivity Thickness of Fire rating (mins)
of insulation (W/m.°C) insulation (mm) Ambient Elevated
Cured temperature cured
Series 1 0.0839 50 18 26
0.078 19 28
0.043 31 44
Series 2 0.043 25 17 23
40 25 36
50 31 44
55 33 49
60 37 53
75 46 66
530 E. R. K. Chandrathilaka and J. C. P. H. Gamage

3 Theoretical Analysis

Milke (2000) has suggested an equation to predict the fire performance of an insulated
structural steel section.
" #
ki Tf  Ts
DTs ¼
 Dt ð2Þ
D þ 0:5ci qi h
h cs W

Ts = Steel temperature (°C)

Tf = Fire temperature (°C)
ki = Thermal conductivity of insulation material (W/m.°C)
h = Insulation thickness (m)
cs = Steel specific heat (J/kg.°C)
W = Steel weight per unit length (kg/m)
D = Heat perimeter (m)
ci = Specific heat of insulation material (J/kg.°C)
qi = Density of insulation material (kg/m3)
Dt = Time interval (s)
DTs = Steel temperature change (°C)
Tf = Standard fire temperature (°C)
However, the main vulnerability of CFRP strengthened steel sections are the epoxy
adhesive layer which connects the CFRP to steel. The glass transition temperature of
the bond has a value between 50 °C to 61 °C, varying with the curing type and
condition (Chandrathilaka et al. 2019). Therefore, the validity of Eq. (2) depends on
the thermal properties of CFRP and adhesive. Zang and Usmani (2015) has shown the
similarity of heat conductivity and Ohm’s law in electric-circuit theory. The thermal
resistance of the insulators was defined as,

hi
Ri ¼ ð3Þ
ki

Ri = Thermal resistance
hi = Insulation thickness (m)
ki = Thermal conductivity of insulation material (W/m.°C)
However, in cases of low insulation thicknesses and high thermal conductivities the
thermal resistance becomes very low, resulting in low temperature variance. The
thermal conductivity of insulation material (0.043 W/m.°C–0.0839 W/m.°C) is much
lower compared to thermal conductivity of CFRP (5 W/m.°C) and adhesive layers
(1 W/m.°C). The insulation thickness varies between 25 mm to 75 mm, while the
CFRP and adhesive layer thicknesses were 0.166 mm and 0.5 mm, respectively.
Therefore, the accuracy of Eq. (2) should be the main focus in this case, ignoring the
CFRP and adhesive layers and considering the steel temperature as bond temperature.
 Fire Performance of CFRP Strengthened Steel I Beams Cured 531

The same steel I beam was used in the theoretical analysis to evaluate the fire
performance. It was assumed that the CFRP strengthening was done by wrapping the
CFRP around a steel column section as mentioned in above.

4 Results and Validation

Temperature variation in epoxy layer of non-insulated steel I section and standard fire
curve is shown in Fig. 4. At the start of the fire, the temperature difference between
environment and CFRP strengthen steel beam was about 45%. However, after 100 s,
the difference was almost negligible. The temperature of epoxy layer has reached the
glass transition temperature of the epoxy layer (50 °C) within 10 s. The fire rating of
non-insulated CFRP strengthen steel I section was less than 10 s, which was very weak
in fire performance. Therefore, insulation is a compulsory requirement in CFRP
strengthened steel I sections in civil engineering applications to maintain required fire
rating of buildings.

Fig. 4. Temperature variation in epoxy layer of non-insulated CFRP strengthen steel I beam and
standard fire with time

Theoretical thermal analysis and numerical thermal analysis results for varying
insulation layer thickness and thermal conductivity is shown in Figs. 5 and 6. For the
different thermal conductivities of insulation material, the results had an average less
than 1% difference between numerical and theoretical analysis. However, for the dif-
ferent insulation layer thicknesses, the variance between theoretical and numerical
analysis results were an average of 2%. For the insulation thicknesses of 40 mm, 50 mm
and 55 mm, the difference is far more negligible. However, at 25 mm insulation
532 E. R. K. Chandrathilaka and J. C. P. H. Gamage

thickness, the difference between numerical and theoretical analysis were as high as 5%
with theoretical thermal analysis, showing higher temperature value than the numerical
analysis at the same time. When the insulation thicknesses are 60 mm and 75 mm the
numerical analysis has over-predicted the temperature on average difference of 3% than
the theoretically predicted temperature. However, the overall slight variation between
theoretical and numerical results provides evidence for the accuracy of the model.
Variation of temperature across an arbitrary cross section of the insulated CFRP
strengthened steel I section (with thermal conductivity of 0.043 W/m.°C and insulation
layer thickness of 50 mm) is shown in Fig. 7.

Fig. 5. Validation of FE model with theoretical analysis for different thermal conductivities of
insulation materials

Fig. 6. Validation of FE model with theoretical analysis for different insulation layer thicknesses
 Fire Performance of CFRP Strengthened Steel I Beams Cured 533

Fig. 7. Variation of temperature across an arbitrary cross section of insulated CFRP strengthened
steel I section

5 Fire Rating of CFRP Strengthened Steel I Section

Fire rating of the CFRP strengthened steel I sections were calculated by referring the Tg of
CFRP/epoxy/steel bond for elevated temperature curing (61 °C) and ambient curing
condition (50 °C) (Chandrathilaka et al. 2019). On average, 35% increase of fire rating
was noted in elevated temperature cured CFRP strengthened I sections compared to
ambient cured CFRP strengthened I sections as shown in Figs. 8 and 9. However, to
achieve 1-hour fire rating there is a requirement of thermal conductivity of 0.043 W/m.°C
for the insulation material with 75 mm insulation layer thickness. At 75 mm insulation
layer thickness, 20 min increase in fire rating was noted in elevated temperature cured
CFRP strengthened steel I sections compared to ambient cured ones. However, use of
0.839 W/m.°C and 0.78 W/m.°C thermal conductivities for the insulation, indicated the
same fire rating values as they were having a comparatively similar thermal conductivity
while 0.043 W/m.°C thermal conductivity insulation performs 63% better than those two
types. This shows that decreasing the thermal conductivity by 45%, fire rating can be
increased by more than 65%. However, when the insulation thickness increases by 300%,
the increase in fire rating was only 190%. Therefore, the use of low thermal conductive
insulation with lesser thickness may increase the fire rating than using high thermal
conductivity insulation with high insulation layer thickness. Thermal profiles within the
epoxy bond line of insulated beam with 0.043 W/m.°C conductivity and thickness of
50 mm are shown in Fig. 10.
534 E. R. K. Chandrathilaka and J. C. P. H. Gamage

Fig. 8. Variation of fire ratings with thermal conductivities of insulation materials

Fig. 9. Variation of fire ratings with insulation thickness

 Fire Performance of CFRP Strengthened Steel I Beams Cured 535

Fig. 10. Thermal profile within the epoxy layer for insulated I beam with conductivity
0.043 W/m.°C and thickness of 50 mm at (a) 10 min (b) 20 min (c) 30 min (d) 40 min
(e) 50 min (f) 55 min
536 E. R. K. Chandrathilaka and J. C. P. H. Gamage

6 Conclusions

A three dimensional heat transfer model was developed to predict the thermal per-
formance of CFRP strengthened steel I beam. Theoretical and numerical analysis were
compared under standard fire conditions. The following conclusions were made,
1. Fire rating of non-insulated CFRP strengthened steel I section was less than 10 s.
Insulation is a compulsory requirement in CFRP strengthened steel I sections.
2. A good agreement was noted between numerical and theoretical fire analysis results
for the steel I beam. The difference between theoretical and numerical results was
less than 5%.
3. To achieve at least 1-hour fire rating, 75 mm thick insulation layer is required with
0.043 W/m.°C thermal conductivity for the insulation material. On average, 35%
increase in fire rating was noted for the same insulation thickness and the thermal
conductivity when elevated temperature curing was done after installation of CFRP.

Acknowledgments. A special thank goes to the staff in the computer laboratory of Department
of Civil Engineering, University of Moratuwa for their valuable assistance. Senate Research
Council of the university of Moratuwa is greatly acknowledged for arranging the necessary
financial support (Grant No: SRC/LT/2016/19).

References
ARELDITE 420 A/B (Technical data sheet) Two component epoxy adhesive system, Huntsman
advanced materials. https://www.huntsman.com. Accessed 12 Aug 2018
Nguyen T, Bai Y, Zhao X, Al-Mahaidi R (2011) Mechanical characterization of steel/CFRP
double strap joints at elevated temperatures. Compos Struct 93:1604–1612. https://doi.org/10.
1016/j.compstruct.2011.01.010
Bai Y, Nguyen TC, Zhao XL, Al-Mahaidi R (2014) Enhancement of steel/CFRP adhesively-
bonded joints at elevated temperatures through curing. Aust J Struct Eng 15(4):367–376
Nguyen TC, Bai Y, Zhao XL, Al-Mahaidi R (2012) Effects of ultraviolet radiation and associated
elevated temperature on mechanical performance of steel/CFRP double strap joints. Compos
Struct 94(12):3563–3573
Chandrathilaka ERK, Gamage JCPH, Fawzia S (2019) Mechanical characterization of
CFRP/steel bond cured and tested at elevated temperature. Compos Struct 207:471–477.
https://doi.org/10.1016/j.compstruct.2018.09.048
Gamage JCPH, Wong MB (2006) Bond characteristics of CFRP plated concrete members under
elevated temperatures. Compos Struct 2006(75):199–205. https://doi.org/10.1016/j.
compstruct.2006.04.068
Imran M, Mahendran M, Keerthan P (2018) Heat transfer modelling of CFRP strengthened and
insulated steel tubular columns. Constr Build Mater 184:278–294. https://doi.org/10.1016/j.
conbuildmat.2018.06.205
Zhang C, Usmani A (2015) Heat transfer principles in thermal calculation of structures in fire.
Fire Saf J 78:85–95. https://doi.org/10.1016/j.firesaf.2015.08.006
Zhang C, Li G, Wang Y (2012) Sensitivity study on using different formulae for calculating the
temperature of insulated steel members in natural. Fire Technol 2012:343–366. https://doi.
org/10.1007/s10694-011-0225-x
 Fire Performance of CFRP Strengthened Steel I Beams Cured 537

ANSYS mechanical user’s guide (2013) ANSYS, Inc. http://www.ansys.com

X-Wrap C300 (Technical data sheet) High strength carbon fiber fabric for structural
strengthening, X-CALIBUR structural systems. https://www.x-calibur.us. Accessed 12 Aug
2018
CAFCO 300 (Technical data sheet) Wet mix spray-applied fire resistive materia. https://isolatek.
com. Accessed 12 Oct 2018
CAFCO 300 AC (Technical data sheet) Wet mix spray-applied fire resistive materia. https://
isolatek.com. Accessed 12 Oct 2018
CAFCO® BLAZE-SHIELD® II (Technical data sheet) Wet mix spray-applied fire resistive
materia. https://isolatek.com. Accessed 12 Oct 2018
ISO 834-11:2014 (2014) ISO. https://www.iso.org/standard/57595.html. Accessed 10 Aug 2018
Milke JA (2000) Analytical methods for determining fire resistance of steel members n.d.: 209–
238. Section four Chapter 9
 Treatment of Wastewater with High Nitrogen
Loading by Moving Bed Biofilm Reactor

H. K. D. T. Karunarathne(&), L. K. U. De Silva,
and K. B. S. N. Jinadasa

Faculty of Engineering, University of Peradeniya, Peradeniya, Sri Lanka

[email protected]

Abstract. Wastewater of municipal toilets contains higher nitrogen to carbon

ratio than other types of wastewater. The removal of nitrogen from municipal
toilet wastewater has become a major concern due to problems caused by dis-
charge of insufficiently treated wastewater, such as eutrophication and toxicity.
Moving Bed Biofilm Reactor (MBBR) has been developed in order to facilitate
nitrogen removal. Treatment of nitrogen is commonly achieved in separate
aerobic and anoxic tanks in typical MBBR systems. However the treatment
process using a single MBBR tank as an advantageous method because of the
low cost and low complexity, has not been thoroughly studied. This research is
focused on the performance of MBBR in nitrogen removal based on simulta-
neous nitrification and de-nitrification in single aerobic MBBR with continuous
operation. A moving-bed biofilm reactor system with an anoxic–aerobic
arrangement was set up and the wastewater extracted from the wastewater
treatment plant at temple of Tooth Relic, Kandy, was kept in the prepared
laboratory set up for the accumulation of microorganisms. The operation was
carried out in 3 steps. In the first step, the DO concentration in the reactor was
maintained at 5 mg/l and the synthetic wastewater was introduced initially to the
reactor at low nitrogen concentration. Then, the nitrogen concentration was
increased varying the BOD/N. In the second step BOD/N ratio was maintained
at 4.2, which is the BOD/N value of public toilet wastewater and the DO
concentration was varied. Volatile Suspended Solids (VSS) of the reactor was
kept at around 2200 mg/l. In the third step the DO concentration was maintained
at 1.5 mg/l since it was the optimum DO concentration for highest de-
nitrification in the 2nd step. Then the VSS of the reactor was maintained at
2200 mg/l and the BOD/N ratio was varied. Under the operating conditions,
67% of BOD and COD removal and 60% of TN removal could be achieved
from the prepared laboratory set up. Further it was observed that the removal of
nitrogen increased when the BOD/N ratio was increased.

Keywords: MBBR
Public toilet wastewater
High nitrogen loading

DO concentration

© Springer Nature Singapore Pte Ltd. 2020

R. Dissanayake and P. Mendis (Eds.): ICSBE 2018, LNCE 44, pp. 538–549, 2020.
https://doi.org/10.1007/978-981-13-9749-3_47
Treatment of Wastewater with High Nitrogen Loading by Moving Bed Biofilm Reactor 539

1 Introduction

Many processes are being used for treatment of sewage in Sri Lanka. These conven-
tional methods are not favorable for the treatment of BOD in wastewater and they are
not successful when considering the removal of nutrients such as nitrogen and
phosphorous.
The moving bed biofilm reactor was developed for the first time when nitrogen
removal from wastewater became a major problem. It was first developed in Norway in
the late 1980s and in early 1990s. Now, the MBBR has been successfully used for
treatment of industrial and municipal wastewater.
The MBBR has several advantages compared to Conventional Activated Sludge
(CAS) processes because of simple and reliable operation. There is no clogging. It does
not need periodical backwashing because in the MBBR, the biomass grows on bio film
carrier which moves freely with wastewater in the MBBR reactor. The process in the
MBBR can be aerobic, anaerobic or anoxic. In the aerobic process, the biofilm carrier
movement is done through the agitation of the air supplied into the reactor, while in the
anoxic and anaerobic processes a mixer inside the reactor is used for agitation (Øde-
gaard 2006).
In conventional methods of wastewater treatment, nitrogen removal is achieved by
nitrification and de-nitrification using two separate reactors with different environ-
mental conditions. However, this research is specified on evaluating the performance of
new MBBR setup with simultaneous nitrification and de-nitrification in a single tank.
This method is advantageous because of the low cost and the less complexity.

2 Background Information on Biological Wastewater

Treatment Using MBBR

The original and mostly used biofilm carrier (K1) is made of high density polyethylene
(Density 0.95 g/cm3) and shaped in a small cylinder with a cross in the inner side of the
cylinder and “fins” to the outer side, as illustrated by the Fig. 1. Biomass attached to
these carries is used for treatment in MBBR.

Fig. 1. Bio carriers (Ødegaard 2006)

540 H. K. D. T. Karunarathne et al.

The objectives of biological treatment of wastewater are to oxidize dissolved and

particulate biodegradable constituents into acceptable end products, to capture and
incorporate suspended and non-settleable colloidal solids into biological floc or bio-
film, to transform or remove specific nutrients such as nitrogen and phosphorus. The
removal of carbonaceous BOD, the coagulation of non-settleable colloidal solids, and
the stabilization of organic matter are accomplished biologically using a variety of
microorganisms, principally bacteria. The colloidal and dissolved carbonaceous
organic matter is converted into various gases and into cell tissues by microorganisms.
These products are then removed by gravity settling (Metcalf et al. 2003).
Organic matter and carbon dioxide are very common sources of cell carbon for
microorganisms. Organisms that use organic carbon for the formation of cell tissue are
called heterotrophs. Organisms that derive cell carbon from carbon dioxide are called
autotrophs. The energy needed for cell growth may be supplied by light or by a
chemical oxidation reaction. Nutrients, rather than carbon or energy source, may be at
times the limiting material for microbial cell growth. The principal inorganic nutrients
needed by microorganisms are N, S, P, K, Mg, Ca, Fe, Na, and Cl.
Nitrogen removal of wastewater is achieved in a treatment process by nitrification
and de-nitrification. Nitrification is the term used to describe the two step biological
process in which ammonia (NH4-N) is oxidized into nitrite (NO2-N) and nitrite is
oxidized into nitrate (NO3-N) under aerobic conditions. The biological reduction of
nitrate into nitric oxide, nitrous oxide and nitrogen gas under anoxic condition is
termed as de-nitrification (Metcalf et al. 2003).

3 Literature Review

Sidek et al. (2015) stated that MBBR is effective in COD, nitrogen and Total Sus-
pended Solids (TSS) removal than the CAS process, while the CAS process is better in
BOD removal than the MBBR. In terms of sludge production, MBBR produces fewer
amount compared to that of CAS. For overall performance, MBBR has better con-
stituent removal efficiency compared to CAS process in the experimental basis study.
Most of the literature regarding nitrogen removal from MBBR is based on nitri-
fication and de-nitrification done in two separate reactors (Zinatizadeh and Ghaytooli
2015). Separate reactors or sequences in intermittent aeration systems are required to
provide the two different environmental conditions. However, recent studies have
revealed that these two important steps can occur in the same reactor (Pochana and
Keller 1999).
According to Metcalf et al. (2003) under low DO concentration conditions, de-
nitrification can occur in the biomass floc interior, while nitrification is occurring at the
floc exterior. Under these conditions, nitrogen removal that occurs in a single aeration
tank is referred to as simultaneous nitrification and de-nitrification (SND).
Pochana and Keller (1999) stated that the effectiveness of the de-nitrification
process decreases when oxygen concentrations are higher than 0.2 mg O2/L. DO
concentration for nitrification should be higher than 2 mg O2/L otherwise DO can be
the limiting factor. At the same time, 0.2 mg O2/L is considered as a critical value, at
which nitrification no longer occurs (Bliss et al. 1986). However DO concentration
Treatment of Wastewater with High Nitrogen Loading by Moving Bed Biofilm Reactor 541

around 0.5 mg/L is suitable to achieve a nitrification rate equal to the de-nitrification
rate which can lead to Simultaneous Nitrification and De-nitrification (SND).
Wang et al. (2006) stated that SND could be successfully established at DO of
about 2 mg/l. About 89.9% of total nitrogen (TN) removal efficiency through SND had
been achieved in the system of moving bed biofilm system which was used to treat
municipal wastewater.
To accomplish de-nitrification in any process, organic carbon is found to be one of
the most essential factors since it is needed as a carbon and energy source for the
bacteria (Pochana and Keller 1999). Isaacs et al. (1994) proposed that COD/N ratio for
de-nitrification is in the range of 3.5 to 4.5 g COD/g N.
Sumino et al. (2006) examined the effects of C/N ratio and total organic carbon
(TOC) loading on nitrogen removal through simultaneous nitrate reduction and anaer-
obic ammonium oxidation in a single reactor and proved that nitrogen removal ratio was
affected by C/N ratio and TOC loading, not by the amount of granular sludge provided.
In the case of a moving bed system, the thickness of the biofilm should be con-
sidered. Several Experiments have been performed to determine the effect of the floc
size on SND in order to test the hypothesis that Simultaneous Nitrification and De-
nitrification (SND) is a physical phenomenon. The underlying physical explanation is
that a substantial anoxic mass fraction exists in the center of the biomass flocs resulting
from an oxygen diffusion limitation into the flocs. With respect to activated sludge
processes, large floc sizes create an anoxic zone inside the flocs leading to de-
nitrification in that area (Rosenwinkel and Cornelius 2005).

4 Methodology

4.1 Reactor Configuration

A moving-bed biofilm reactor system with an anoxic–aerobic arrangement, as shown in
Fig. 2, was used to treat the synthetic wastewater. The reactor working volume was
4.0 L. And a square shaped Perspex tank of total volume of 8.0 L was used. The biofilm
carriers filled 50% by volume of reactor. Influent was fed using an overhead tank and the
treated effluent was let to be overflowed through a pipe just above the head level of the
reactor into a secondary clarifier. An aeration unit was used to aerate the reactor.

Fig. 2. Schematic diagram of the reactor

542 H. K. D. T. Karunarathne et al.

4.2 Wastewater Characteristics

Public toilet wastewater was collected from three different locations in the Kandy town
area and those were tested to find out the real characteristics of public toilet wastewater.
According to the laboratory experiment results, the average BOD/N ratio of public
toilet wastewater was 4.2. Synthetic wastewater was prepared and fed into the system
according to the results obtained.

4.3 Operation and Procedure

After setting up the model, the operational phase was started. A batch experiment was
carried out to find the HRT value for the setup. Finally the HRT of 6 h was maintained
in the reactor. Then the wastewater from the wastewater treatment plant at temple of
Tooth Relic was kept for accumulation of microorganisms for two weeks.
The testing procedure adopted comprised three major steps. In the first step the
synthetic wastewater was introduced initially to the reactor at low Nitrogen concen-
tration and then the Nitrogen concentration was increased varying the BOD/N ratio
(Table 1). Four samples were tested for different BOD/N ratios. The BOD concen-
tration was maintained as a constant parameter and Nitrogen concentration was
increased using NH4Cl and NaHCO3 to obtain those four BOD/N ratios. Each sample
was treated for 2 weeks. Samples from influent and the effluent were collected and
tested to find out the treatment efficiency. The inflow rate, HRT, DO concentration and
pH of the reactor during the first step were 11.1 ml/min, 6 h, 5 mg/l and 7 respectively.

Table 1. Sample sets used for step 1

Sample BOD/N
1 9.5
2 7.5
3 5.5
4 3.5

In the second step of the procedure the BOD/N ratio was maintained at 4.2, which
was the BOD/N value for the real public toilet wastewater and the DO concentration
was varied. Since the de-nitrification was not efficient in the first step of the procedure,
the DO concentration was varied as shown in the Table 2. VSS of the reactor was kept
at around 2200 mg/l. Once the influent was treated for 6 h, samples from influent and
effluent were taken and were tested for the same parameters as in the first step.
Treatment of Wastewater with High Nitrogen Loading by Moving Bed Biofilm Reactor 543

Table 2. DO concentrations used for step 2

Sample no. DO concentration (mg/l)
1 5
2 4
3 2
4 1.5
5 1
6 0.6
7 0.3

In the third step, the DO concentration was maintained at 1.5 mg/l since it was the
optimum DO concentration for the highest de-nitrification in the second step. Then the
VSS of the reactor was maintained at 2200 mg/l and the BOD/N ratio was varied as
shown in the Table 3.

Table 3. Sample sets used for step 3

Sample no. BOD/N ratio
1 3
2 4.2
3 5
4 7

4.4 Data Collection

Samples from influent, effluent and from the reactor were collected for each test.
Samples of biomass attached carriers also were collected and biomass was separated by
vigorously stirring in water. The parameters measured are shown in the Table 4.

Table 4. Parameters measured for three steps

Parameter Influent Effluent Reactor
p p
BOD
p p
COD
p p
NH4+-N
p p
NO3−N
p p
NO2−N
p p
VSS
544 H. K. D. T. Karunarathne et al.

5 Results and Discussion

5.1 Observation for Step 1
During the first step the DO concentration was maintained at 5 mg/l. The initial value
of VSS of the reactor was 2200 mg/l. The initial BOD/N ratio was 9.5. Then the
BOD/N ratio was reduced by increasing nitrogen concentration in influent. The BOD
removal of the system is shown in the Fig. 3. These results have proven the fact that,
this system can be used to remove organic matter efficiently. From the initial stage
onwards, the efficiency of the system for BOD removal was around 90%.

Fig. 3. BOD removal of system in step 1

The COD removal of the system is shown in the Fig. 4. According to the results it
has been proved that the COD removal of the system also was efficient. From the initial
stage onwards, the efficiency of the system for COD removal was higher than 90%.

Fig. 4. COD removal of system in step 1

The source of Nitrogen used when preparing the synthetic wastewater was NH4Cl.
And the nitrogen concentration was varied in the 4 samples used. Figure 5 shows the
variation of the NH3–N in the system. According to the results, it shows a removal of
NH3–N in all four samples. However the reduction of NH3–N increased the NO3-N
concentration of the effluent.
Treatment of Wastewater with High Nitrogen Loading by Moving Bed Biofilm Reactor 545

Fig. 5. NH3–N variation of system in step 1

The results showed that the NO3−-N in the effluent was increased after treatment.
That was due to the fact that conditions of the reactor were not suitable for de-
nitrification. The reactor was in high aerobic condition in the first step of the operation.
The DO concentration was 5 mg/l in this step. Figure 6 shows the variation of TN in
the system.

Fig. 6. TN removal of system in step 1

5.2 Observation for Step 2

During the second step of this study, the BOD/N ratio was maintained at the BOD/N
ratio value of the public toilet wastewater (BOD/N = 4.2). The VSS of the reactor was
also maintained at 2200 mg/l. Then the DO concentration was varied from 5 mg/l to
0.3 mg/l and the treatment efficiency of the system was tested.
The BOD removal of the system is shown in the Fig. 7. BOD removal varied from
60%–80% when the DO concentration was higher than 1 mg/l. When the DO con-
centration was further reduced, the BOD removal efficiency decreased. That was due to
the anaerobic conditions in the reactor. Therefore it shows that the BOD removal of the
system was not affected much by varying the DO concentration.
546 H. K. D. T. Karunarathne et al.

Fig. 7. BOD removal of system in step 2

The COD removal of the system is shown in the Fig. 8. COD removal had not been
affected by the decreased DO concentration until it reduced to 1 mg/l. The COD
removal efficiency was around 60% to 70% when the DO concentration was higher
than 1 mg/l.

Fig. 8. COD removal of system in step 2

NH3-N variation of the system during the step two is presented in Fig. 9. It shows
that the nitrification was efficient when the DO concentration was higher than 1.5 mg/l.
When the DO concentration was lower than 1.5 mg/l the anaerobic conditions occurred
in the reactor. Therefore the nitrification had been reduced. The NO3-N concentration
in effluent was the highest when the DO concentration was 5 mg/l. When the DO
concentration was decreased, the removal of nitrogen was increased. The optimum de-
nitrification could be achieved at 1.5 mg/l DO concentration. When the DO concen-
tration was further reduced, the de-nitrification decreased again due to low nitrification
rate. The optimum DO concentration for Simultaneous Nitrification and De-
nitrification for this system was obtained to be 1.5 mg/l. The Fig. 10 shows the
NO3-N variation of the system.
Treatment of Wastewater with High Nitrogen Loading by Moving Bed Biofilm Reactor 547

Fig. 9. NH3-N variation of the system in step 2

Fig. 10. NO3-N variation of system in step 2

5.3 Observation for Step 3

In the third step, The DO concentration was kept constant at the optimum value that
was obtained in the previous step. The VSS of the reactor was maintained at 2200 mg/l.
Then the BOD/N ratio was changed from 3 to 7 step by step and the removal effi-
ciencies were obtained from tests. This study was done to obtain the effect of BOD/N
ratio for the SND.
The DO concentration was maintained at 1.5 mg/l. The BOD removal and COD
removal increased when the BOD/N ratio was increased. Figure 11 shows the BOD
removal of the system and Fig. 12 shows the COD removal of the system.
Figure 13 shows that the NH3-N removal had not been affected by the increased
BOD/N ratio. The nitrification efficiency was nearly a constant value for all 4 BOD/N
ratios. The de-nitrification efficiency was increased when the BOD/N ratio increased.
When the BOD/N ratio was increased, the available additional carbon source for de-
nitrification increased. Adding an external carbon source is therefore favorable to
maximize the nitrogen removal. Figure 14 shows the NO3-N variation of system.
548 H. K. D. T. Karunarathne et al.

Fig. 11. BOD removal of system in step 3

Fig. 12. COD removal of system in step 3

Fig. 13. NH3-N variation of the system in step 3

Fig. 14. NO3-N variation of system in step 3

Treatment of Wastewater with High Nitrogen Loading by Moving Bed Biofilm Reactor 549

6 Conclusion

Treatment of public toilet wastewater with BOD/N ratio of 4.2 is viable when main-
taining the DO concentration at 1.5 mg/l and VSS at 2200 mg/l. It is clear that under
these operating conditions, 67% of BOD and COD removal and 60% of TN removal
can be achieved. Furthermore the removal of nitrogen can be increased when the
BOD/N ratio is increased.

Acknowledgements. The authors wish to extend their gratitude to the staff members of
wastewater treatment plant at The Temple of the tooth relic, Kandy and the Engineers and staff of
Kandy municipal council for their support, and the evaluation panel for their guidance and
comments to complete this research successfully.

References
Rusten B, Eikebrokk B, Ulgenes Y, Lygren E (2004). Design and operation of Kaldnes moving
bed biofilm reactor
Pochana K, Keller J (1999) Study of factors affecting simultaneous nitrification and de-
nitrification (SND). Water Sci Technol 39:61–68
Ødegaard H (2006) Innovations in wastewater treatment: the moving bed biofilm process. Water
Sci Technol 53:17–33
Zinatizadeh AAL, Ghaytooli E (2015) Simultaneous nitrogen and carbon removal from
wastewater at different operating conditions in a moving bed biofilm reactor (MBBR) process
modeling and optimization. J Taiwan Inst Chem Eng 53:98–111
Rosenwinkel KH, Cornelius A (2005) Deammonification in the moving bed process for the
treatment of wastewater with high ammonium content. Chem Eng Technol: Ind Chem-Plant
Equipment-Process Eng Biotechnol 28:49–52
Wang XJ, Xia SQ, Chen L, Zhao JF, Renault NJ, Chovelon JM (2006) Nutrients removal from
municipal wastewater by chemical precipitation in a moving bed biofilm reactor. Process
Biochem 41:824–828
Sumino T, Isaka K, Ikuta H, Saiki Y, Yokota T (2006) Nitrogen removal from wastewater using
simultaneous nitrate reduction and anaerobic ammonium oxidation in single reactor. J Biosci
Bioeng 102:346–351
Liebert MA (2015) Effective biological nitrogen removal treatment process for domestic
wastewater
Metcalf E (2003) Wastewater engineering Treatment and reuse, 4th edn. McGraw-Hill, New
York
Sidek LM, Mohiyaden HA, Basri H, Salih GHA, Birima AH, Ali Z, Sabri AFM, Noh MNN
(2015) Experimental comparison between moving bed biofilm reactor (MBBR) and
conventional activated sludge (CAS) for river purification treatment plant. Adv Mater Res
11:806–811
 Experimental and Analytical Study of Residual
Yield and Tensile Strengths in Partially
Corroded Structural Steel

P. A. K. Karunananda(&), K. A. D. D. T. Kariyapperuma,
R. P. Weeraratne, and H. D. M. Sanjeewani

Department of Civil Engineering,

The Open University of Sri Lanka, Nawala, Nugegoda, Sri Lanka
[email protected]

Abstract. Corrosion has been one of the main structural degradation processes
and has been the root cause of many engineering structural failures around the
world. Therefore, proper understanding of corrosion process and especially
residual strength measurement is an important consideration in the engineering
discipline. This study was performed to find the residual strengths of partially
corroded steel using a cost effective reliable method which can be adopted
easily. For that, Average Weight Ratio (AWR) was introduced instead of the
Minimum Thickness Ratio (MTR), proposed by previous researchers. In this
context, partially corroded steel specimens were subjected to accelerated cor-
rosion tests in a corrosion incubator. There, correlations were built between
hardness verses AWR, ultimate tensile strength verses AWR and ultimate tensile
strength ratio verses AWR of partially corroded steel.

Keywords: Corrosion
Average Weight Ratio
Hardness
Residual strength

1 Introduction

The use of steel structures all over the world increased gradually due to its many
advantages such as its versatility, high strength, uniformity, ease of use, toughness,
ductility and many other properties. Structural steel has very high yield strength for
compression and tension. Therefore, it is convenient to use structural steel as a building
construction material than other construction materials such as concrete, timber, etc.
But one of the major disadvantages of using steel as a construction material is the
fast deterioration in steel due to corrosion when exposed to aggressive environmental
conditions and poor maintenance. Because of this, reduction in their strength capacities,
corrosion wastage, fatigue cracking, or mechanical damage may occur during their
service life. Therefore significant issues occur on safety, health, environment and
financial costs. The catastrophic collapse of the silver bridge on Ohio River in 1967,
USA is a classic example for structural failure due to corrosion.
Therefore, it is essential to carefully evaluate the remaining strength of steel
structures to understand the feasibility of the steel structure for future usage. Several
experimental studies and detailed investigations have been done by some researchers

© Springer Nature Singapore Pte Ltd. 2020

R. Dissanayake and P. Mendis (Eds.): ICSBE 2018, LNCE 44, pp. 550–558, 2020.
https://doi.org/10.1007/978-981-13-9749-3_48
 Experimental and Analytical Study of Residual Yield and Tensile Strengths 551

during the past few decades, to introduce methods of estimating the remaining strength
of corroded structural steel members (Appuhamy et al. 2011).
This research study is mainly based on developing a simple and convenient method
to estimate residual ultimate tensile strength and yield strength of the corroded struc-
tural steel specimens. Experimental results were used to develop and determine more
reliable results. Those developed results can be used for the existing corroded structures
by measuring the remaining thickness and calculating the average thickness ratio with
respect to initial thickness and thereby, the remaining strength can be easily calculated.

2 Theoretical Background

It has been mentioned that regular coarse surface measurement could be sufficient for
minor corrosion members as surface roughness does not affect minor corrosion.
Therefore, for our experiments, average thickness ratio (ATR) was considered instead
of minimum thickness ratio (MTR).
Considering that the density of steel remained unchanged and only thickness is
reduced rather than length and breadth, it can be proven that average thickness ratio is
equal to average weight ratio as below.

MassðmÞ
DensityðqÞ ¼
VolumeðvÞ

Initial Density ¼ Final Density

initial mass ðwoÞ final mass ðwf Þ

Therefore; ¼
initial volume ðVoÞ final volume ðVf Þ

Wo Wf
 B  to ¼  B  tf
L L

As considering unchanged of length (L) and breadth (B)

tf Wf
¼
to Wo

Hence,

Average thickness ratio ¼ Average Weight ratio

2.1 Corrosion Acceleration

Corrosion accelerates due to factors such as the over potential, the aggressive medium,
the temperature and the concentration. However, the materials react differently to
different types of electrolytes. The accelerated corrosion test is done to obtain results in
a shorter period of time than from natural exposure. In general, accelerated corrosion
552 P. A. K. Karunananda et al.

tests can be performed increasing temperature and potency with time to get different
corrosion levels.

2.2 Techniques for Inducing Accelerated Corrosion of Steel

Corrosion of steel is a slow process in the natural environment under normal condi-
tions. It takes reasonably long time to achieve a significant degree of corrosion in steel.
Different types of techniques were introduced to accelerate corrosion in past research
studies. One of it is “Impressed current technique” which is a method to induce
significant corrosion in a short period of time to steel embedded in the concrete.
Further, another alternative technique for inducing accelerated reinforcement cor-
rosion is to build up artificial climate environment in the room which controlled by a
computer system. In that artificial climate room, environmental condition, temperature,
relative humidity and repeated wetting and drying cycle consisted of salt water (5%
NaCl solution) spraying for 1 h and infrared light shining for 7 h were controlled. This
test set up is shown in Fig. 1 (Ahmad 2009)

Fig. 1. Set-up for accelerating reinforcement corrosion in large-size reinforced concrete beam
specimen. (Source: Ahmad 2009)

2.3 Correlations Among Mechanical and Strength Properties

In addition to yield strength and ultimate tensile strength of steel, hardness is an
important mechanical property which can be measured using hardness testing
machines. The above strengths can be found according to formulae among hardness,
yield strength and tensile strengths, which have been developed by past researches. Of
them, the following formulae are very convenient to study;
 Experimental and Analytical Study of Residual Yield and Tensile Strengths 553

H  n n
TS ¼ ð1Þ
2:9 0:217
H
YS ¼  ð0:1Þn ð2Þ
3

(Pavlina and Van Tyne 2008)

where, TS = Tensile strength
YS = Yield strength
H = Diamond pyramid hardness (Vickers hardness)
n = Strain-hardening exponent (n = 0.21)
Further Residual Ultimate Tensile Strength of corroded steel could be determined
using the following formula by substituting Brinell hardness number.

TS ¼ 3:45  HB ð3Þ

(Pavlina and Van Tyne 2008)

where,
HB = Brinell hardness
In addition to the above developed formulae, many conversion tables among
hardness and tensile strengths of steel have been developed by recognized
organizations.

2.4 Yield to Tensile Strength Ratio (YTR) of Steel

It is a very important parameter to determine Yield Strength of steel when the Ultimate
Tensile Strength of particular steel is known. Normally YTR of range from 0.5 to 0.9 is
recommended for steel (Gasko and Rosenberg 2011).

3 Experimental Set up

Two experiments were performed and the following gives the details of these
experiments.

3.1 Experiment 1
Accelerated corrosion test was performed in an incubator where temperature and
potency with time was controlled to check the feasibility of selected corrosion accel-
erated factors to be used for experiment 2 which was done to observe different cor-
rosivity levels of specimens by changing one control factor with respect to others.
1. Two numbers of 50  20  3 mm steel specimens were prepared out of Grade S
275 structural steel (one of them was selected as a standard specimen while the
other was subjected to corrosion).
554 P. A. K. Karunananda et al.

2. A glass tank (300  200  150 mm), (12 V–28 V) capacity DC transformer,
thermostat (30 °C–50 °C), 0.125 M Sodium Chloride solution (as a catalyst) and
flexible wire pieces were used to prepare an incubator.
3. Two numbers of 50  20  3 mm structural steel specimens were submerged in
the sodium chloride solution and connected to a 12 V power supply to create
corrosion on surfaces of steel specimens (in 30 °C temperature).
4. The specimen connected to the cathode end acts as a standard specimen while the
other specimen attached to the anode end corroded with time.
5. The above procedure was repeated twice for increasing temperature (35 °C) and
increasing voltage (28 V) respectively (Fig. 2).

Fig. 2. (a) 50  20  3 mm steel specimens. (b) Arrangement of apparatus

3.2 Experiment 2 - Phase I

Experiment 2 (Phase I) was performed to get different corrosivity levels of steel.
Corrosivity levels were evaluated as a measure of Average Weight Ratio instead of
average thickness ratio according to the above proof.
Initial weight of each specimen (W0) was measured using a digital scale.
The glass tank was filled with 4 L of 0.125 M NaCl solution. The electric heater
coupled to the thermostat was inserted to the glass tank and temperature of the ther-
mostat was adjusted to 30 °C.
The test specimen to corrode was connected to the anode end of the 12 V DC
power supply while another specimen was connected to the cathode end of the power
supply and both specimens were inserted to the solution and the power supply switched
on.
After an hour, the corroded steel specimen was removed from the solution and the
rust of the specimen was removed thoroughly.
• Then, the final weight of the specimen was measured.
• The above steps were repeated by increasing the time period and the solution was
removed and refilled for each repetition.
• The same procedures were adopted for the other experiments by increasing the
voltage and temperature as (12 V, 35 °C), (28 V, 30 °C), (28 V, 35 °C) and (28 V,
40 °C) respectively (Fig. 3).
 Experimental and Analytical Study of Residual Yield and Tensile Strengths 555

Fig. 3. Arrangement of apparatus by supplying of apparatus

4 Results

A considerable number of experiments were performed and the relationships with

AWR obtained are shown in Figs. 4, 5, 6, 7 and 8.

Fig. 4. The graph of Rockwell hardness (HRB) vs average weight ratio (Wf/W0)

Fig. 5. The graph of Brinell hardness (HB) vs average weight ratio (Wf/W0)
556 P. A. K. Karunananda et al.

Fig. 6. The graph of Vickers hardness (HV) vs average weight ratio (Wf/W0)

Fig. 7. The graph of Ultimate Tensile Strength (UTS) vs average weight ratio (Wf/W0)

Fig. 8. The graph of UTS ratio vs average weight ratio (Wf/W0)

 Experimental and Analytical Study of Residual Yield and Tensile Strengths 557

4.1 Identify Correlations to Find Out Residual Strengths

From the graph of Rockwell hardness verses average weight ratio, when average
weight ratio decreases, Rockwell hardness value decreases accordingly producing a
linear correlation between those two parameters up to a certain extent of moderate
corrosivity as shown in Fig. 4.
By scrutinizing the rest of the graphs, similar linear correlation could be identified
between Brinell hardness verses average weight ratio, Vickers hardness verses average
weight ratio, ultimate tensile strength verses average weight ratio and ultimate tensile
strength ratio verses average weight ratio as shown in Figs. 5, 6, 7 and 8.
Residual ultimate tensile strength of partially corroded structural steel could be
determined using the graph of ultimate tensile strength ratio verses average weight ratio
if the initial ultimate tensile strength was known. That obtained value can justify by
using Eq. (3),

TS ¼ 3:45  HB

Brinell hardness value is obtained from the graph of Brinell hardness verses
average weight ratio.
Residual yield strength of partially corroded structural steel could be calculated by
substituting Vickers hardness value obtained from the graph of Vickers hardness verses
average weight ratio to Eq. 2. That obtained value is to be multiplied by 9.81 as it is
given kgf/mm2 and the unit of residual yield strength is N/mm2.

H
YS ¼  ð0:1Þn
3

If yield to tensile strength ratio (YTR) is known, residual yield strength value could
be obtained by substituting residual ultimate tensile strength value and the above
residual yield strength value could be verified.

5 Conclusion

The residual strength measurements were conducted for partially corroded structural
steel and the following conclusions can be made from this study. The corrosion causes
strength reduction of structural steel members and average weight ratio (AWR) can be
used as the measure of the level of corrosion. The experimental and analytical results
can be used to predict the residual strength of actual corroded members accurately.

Acknowledgment. Authors would like to thank to Dr. C.S. Bandara of Civil Engineering
Department in University of Peradeniya for the support given in this research.
558 P. A. K. Karunananda et al.

References
Ahmad S (2009) Techniques for inducing accelerated corrosion of steel in concrete. Arab J Sci
Eng 34(2C):95 Department of Civil Engineering King Fahd University of Petroleum &
Minerals Dhahran, Saudi Arabia
Appuhamy JMRS et al (2011) Analytical study on significance of corroded surface measurement
on residual strength prediction. Proc Eng 14:2260–2268
Gasko M, Rosenberg G (2011) Correlation between hardness and tensile properties in ultra-high
strength dual phase steels – short communication. Mater Eng 18:155–159
Pavlina EJ, Van Tyne CJ (2008) Correlation of yield strength and tensile strength with hardness
for steels. J Mater Eng Perform 17(6):888–893
 Mathematical Modeling of the Hydrolysis
Phase in Anaerobic Digestion of Solid Waste

D. I. U. Edirisinghe(&), S. M. W. T. P. K. Ariyarathna,
and C. S. Kalpage

Department of Chemical and Process Engineering, Faculty of Engineering,

University of Peradeniya, Peradeniya, Sri Lanka
[email protected]

Abstract. Anaerobic digestion has been in use for many decades, and its range
of uses keep on expanding although it has a reputation of being unreliable and
unpredictable. Anaerobic digestion could provide a sustainable solution to waste
disposal as well as power generation, if the process could be controlled as
desired. This study was done in order to observe the behavior and to optimize the
hydrolysis phase in anaerobic digestion of biodegradable solids, by developing a
mathematical model for the process. Considering all aspects, a surface related
model was used to describe the process since hydrolysis of complex solid sub-
strates is dominated by surface phenomena. In order to narrow down the scope of
the research, initial experimentation was focused on pure carbohydrates only.
Boiled rice was selected as the substrate since it is readily available and because it
is a pure form of carbohydrate. Through experimental data and the aid of the
software Matlab2013, the model parameters were determined to be KS = −1 and
rmax = 1.004 mmol/dm3.day under operating conditions of 1 atm pressure and
temperature of 28 °C. Further developments and improvements to this model are
essential to be able to predict hydrolysis of complex substrates at any given
operating condition. The improved model may someday result in a capable and
efficient waste management system which is sustainable, ergonomic and efficient.

Keywords: Anaerobic digestion (AD)
Mathematical modelling
Hydrolysis

Surface related two phase kinetic model
Solid waste

1 Introduction

Anaerobic digestion has been in use for over ninety years (Gene and William 1986) to
treat agricultural, municipal and other waste sludge having a high organic material
content. Anaerobic digestion is becoming more and more popular as a waste treatment
method because of its low power requirement and ability to produce methane gas. Despite
the attractiveness of the process, it has a reputation of being unstable and therefore
unreliable. One of the reasons for this is the lack of fundamental knowledge on anaerobic
digestion process principles. The other reasonable explanation is that originally, anaer-
obic digestion was used to stabilize domestic sludge, but now it is being used to treat
chemically treated sludge, industrial waste and complex solid waste and therefore the
treated sludge quality has changed over the period (Gene and William 1986).

© Springer Nature Singapore Pte Ltd. 2020

R. Dissanayake and P. Mendis (Eds.): ICSBE 2018, LNCE 44, pp. 559–566, 2020.
https://doi.org/10.1007/978-981-13-9749-3_49
560 D. I. U. Edirisinghe et al.

Anaerobic digestion takes place in four steps as hydrolysis, acidogenesis, aceto-

genesis and methanogenesis. Each of the stages involve different bacterial groups
which result in a very complex reaction mechanism overall. Also, substrate digestibility
and bio gas production are affected by process parameters such as pH, temperature and
concentration of Volatile Fatty Acids (VFAs) as well. Due to this complexity, it is
difficult to predict the output even though the input conditions are known. Therefore it
is difficult to scale up the operation and optimize it.
Therefore, a mathematical model of the process can be very helpful in predicting
the stability, understanding the behaviour under different process conditions and
optimizing it. A mathematical model provides considerable insight to the process
mechanism and is less time consuming and less expensive compared to a physical
model. It helps in quantifying output and improving design (Andrews and Graef 1971).
Considering the facts, a mathematical model based on surface related phenomena
was developed for the hydrolysis phase of carbohydrates within this study. Boiled rice
was selected as the carbohydrate rich substrate and the model parameters were deter-
mined with the selected model in literature.

2 Reaction Mechanism of Carbohydrates

The first stage of anaerobic digestion is hydrolysis. Particulate carbohydrates, proteins,

lipids and inert material present in a complex organic substrate are disintegrated into
monosaccharides, amino acids and long chain fatty acids, solubilized and finally acted
upon by hydrolytic enzymes in the hydrolysis stage. Next, these hydrolysed products
are degraded into VFAs: acetate, butyrate, propionate etc. by acidogenic bacteria in the
acidogenesis stage. The produced intermediate VFAs are further degraded into acetic
acid by the action of acetogenic bacteria. Finally methane forming bacteria acts on the
acetic acid to form methane and carbon dioxide in the methanogenesis stage.

3 Reaction Kinetics of the Hydrolysis Phase

Since the study focuses on solid waste, hydrolysis plays an important role over the
other phases. Unless the complex solid substrate is hydrolysed, the acidogenic, ace-
togenic and methanogenic bacteria cannot digest the substrate to produce methane
directly.
It is known that hydrolysis occurs in two pathways explained below;
• Micro-organisms are secreted in to the bulk of the liquid where they are adsorbed on
to a particle or react with a soluble substrate (Jain et al. 1992).
• Micro-organisms attach to a particle, produce enzymes in its vicinity and benefit
from soluble products released by the enzymatic reaction (Vavlin et al. 1996).
In complex substrates, hydrolysis occurs in both the mechanisms simultaneously.
A two phase surface related model coupled with Mechaelis-Menten kinetics and the
Contois model are widely used in describing the hydrolysis kinetics. The surface
related model is a modification of the Mechaelis-Menten kinetics coupled with 1st order
 Mathematical Modeling of the Hydrolysis Phase in Anaerobic Digestion 561

rate kinetics and bacterial growth kinetics to suit realistic conditions. The Mechaelis-
Menten model starts with the preliminary rate equation, and is developed gradually to
suit realistic conditions.
However the parallel reaction is dominated by the surface related mechanism
mentioned secondly, and the first order kinetics for the reaction can be written as;

dCS
¼ kCS ð1Þ
dt
dCP
¼ akCS ð2Þ
dt

Where CS is the substrate concentration, CP is the product concentration, a is the

conversion coefficient, k is the first order rate coefficient and t is the time.
Equations (1) and (2) considers the whole substrate to be degradable. However, in a
complex substrate, only part of the substrate is degradable. Also, it has been found out
that a hydrolysis reaction does not follow first order reaction kinetics all the time and
the biomass concentration affects the hydrolysis rate as well (Liang et al. 2013).
Therefore, the non-degradable fraction of the substrate and effect of biomass concen-
tration have been incorporated in Eq. (3) (Liang et al. 2013);

dCS
¼ kX n ðCS  bCS0 Þ ð3Þ
dt

Where X is the biomass concentration, n is the order of the reaction and b is the
non-degradable fraction of the substrate.
The Mechaelis-Menten kinetics too can be used to describe the same first order
reaction as;

dCS rmax CS
¼ ð4Þ
dt KS ðCS  CS0 Þ þ CS

Where CS0 is the initial substrate concentration, KS is the model coefficient and rmax
is the maximum hydrolysis rate.
All the models explained above by Eq. (1) through (4) only consider the surface
related hydrolysis mechanism. It was found out that Vavlin et al. (1996) had developed
the two phase surface related model which takes the colonization of waste particles by
bacteria into account as well. The model equation related to this is given in Eq. (5).

dCS bX CS
¼ rmax ð5Þ
dt 1 þ bX KS þ CS

Where CS is the substrate concentration, b is the ratio between adsorption and

desorption rate constants, KS is the half saturation coefficient, rmax is the maximum
hydrolysis rate and t is the time.
562 D. I. U. Edirisinghe et al.

4 Methodology
4.1 Mathematical Development of the Model
This research is based on the model expressed by Eq. (4). Since only the initial sub-
strate concentration and variation of product concentration with time is obtained
through experimentation, the rest of the parameters required for the modelling of the
hydrolysis phase should be determined with the aid of mathematical equations gov-
erning hydrolysis.
Relationship shown in Eq. (6) was obtained by integrating the basic rate equations
applicable to hydrolysis and using mass balance for the conversion of substrate, shown
in Eqs. (1) and (2).


CP ¼ CP0 þ aCS0 1  ekt ð6Þ

Once the conversion coefficient for the conversion of volatile solids to products (a)
and the first order rate coefficient for the hydrolysis of carbohydrates (k), is determined,
Eq. (6) and Eq. (2) can be used to obtain a relationship between the substrate and
product concentration. Both these values can be determined through a simultaneous
simulation.
After obtaining a relationship between substrate and product concentration, the
variation of substrate concentration with time can be monitored with the aid of Eq. (2).
Once the substrate consumption rate is known, the model parameters rmax and KS
can be determined through a simultaneous simulation.

4.2 Experimental Procedure

The substrate was prepared by blending 680 g of boiled rice, 4.8 L of water and 30 g
of sugar together. Then the blended rice slurry was transferred to a container and mixed
with inoculum so that the inoculum: substrate ratio was nearly equal to 1, which is the
optimum ratio according to the literature (Yoon 2014). Nine identical 500 mL digesters
were filled with the prepared sample and left under anaerobic conditions. One digester
was opened at a time and sampled periodically to measure the amount of acetate
formed. Assuming that the rate of acidogenesis is much higher than the hydrolysis rate
and acetic acid concentration is ten times higher than the rest of the VFAs, it is
reasonable and accurate enough to use the acetate concentration to represent the
hydrolysed product concentration (Vavlin et al. 2008). Therefore, in this study acetic
acid concentration was measured with time.

5 Results and Discussion

The variation of the product concentration was determined experimentally and plotted
as depicted in Fig. 1. Through a simulation, it is possible to model the variation of
product concentration with time according to Eq. (6) in order to determine the con-
version coefficient a and first order rate coefficient k.
 Mathematical Modeling of the Hydrolysis Phase in Anaerobic Digestion 563

Fig. 1. Variation of product concentration with time

The model for the variation of product concentration with time is given in Eq. (7).
The R2 value of the model is 86.63%. According to Eq. (7), first order rate coefficient
k = 0.2321 per day and conversion coefficient a = 0.0012.

CP ¼ 0:000542e0:2321t þ 0:0004861 ð7Þ

Figure 2 shows the rate of product formation, and is obtained by differentiating the
above fitted model equation. Figure 3 shows the variation of substrate concentration
with time which is calculated with the aid of Eq. (2).

Fig. 2. Variation of the product formation rate with time

564 D. I. U. Edirisinghe et al.

Fig. 3. Variation of the substrate concentration with time

The variation of the substrate concentration with time too can be modelled into a
mathematical equation through a simulation. Using the model, the variation of the rate
of substrate utilization with time and the variation of the rate of substrate utilization
with substrate concentration can be obtained as shown in Figs. 4 and 5 respectively.
The model for the variation of substrate concentration is given in Eq. (8). The R2
value of this fit is 100%, which means that it is a perfect fit.

CS ¼ 0:4517e0:2321t ð8Þ

Fig. 4. Variation of substrate utilization rate with time

 Mathematical Modeling of the Hydrolysis Phase in Anaerobic Digestion 565

Fig. 5. Variation of substrate utilization rate with substrate concentration

After obtaining the plot in Fig. 5, the model parameters of Eq. (4) can be deter-
mined easily by linearizing the equation according to the Langmuir isotherm.
The Langmuir plot obtained by linearizing the model Eq. (4) is shown in Fig. 6.
Using the plot, the model parameters were determined to be KS = −1 and
rmax = 1.004 mmol/dm3.day.

Fig. 6. Langmuir plot for the model

6 Conclusion

Even though the anaerobic digestion process is an unreliable and unpredictable method
of waste treatment, given the present environmental status, sustainable and economic
implementations are essential for the common benefit of the community. Therefore,
during this study, effort was put in to developing a mathematical model that could be
used to predict the behaviour of the hydrolysis phase in anaerobic digestion. The final
566 D. I. U. Edirisinghe et al.

model developed during this study for the hydrolysis of solid carbohydrates in
anaerobic digestion is given below as Eq. (9).

dCS 1:004CS
¼ ð9Þ
dt ðCS0  CS Þ þ CS

Although the results seem to have no contradiction to theory, more improvements

should be made to this model before we could entirely rely on the model as a scientific
community. The behaviour of the model should be studied under various operating
conditions as well as initial substrate concentrations. The approach taken during this
study should be adopted to other biodegradable waste components such as proteins,
lipids and ultimately to complex solid waste which would be a combination of these
components.
These improvements and developments to the present model may someday result in
a capable and efficient waste management system which is sustainable, ergonomic
and efficient.

References
Andrews JF, Graef SP (1971) Dynamic modeling and simulation of the anaerobic digestion
process. Environmental Systems Engineering Department, Clemson University, Clemson, S.
C., pp 126–162
Angelidaki II, Ellegaard L, Ahring BK (1999) A comprehensive model of anaerobic
bioconversion of complex substrates to biogas. Biotechnol Bioeng 63:363–372
Batstone DJ, Keller J, Angelidaki I, Kalyuzhnyi SV, Pavlostathis SG (2002) The IWA anaerobic
digestion model no 1 (ADM1). Water Sci Technol 45:65–73
Gene FP, William FW (1986) Fundamentals of anaerobic digestion of waste water. J Environ
Eng 112(5):867–920
Jain S, Lala AK, Bhatia SK, Kudchadker AP (1992) Modeling of hydrolysis controlled anaerobic
digestion. J Chem Technol Biotechnol 53:337–344
Liang Y, Wensel PC, Jingwei M, Shulin C (2013) Mathematical modeling in anaerobic digestion.
J Bioremediat Biodegradation S4
Pavlostathis SG, Giraldo-Gomez E (1991) Kinetics of anaerobic treatment. CRC Crit Rev
Environ Control 21:411–490
Speece RE (1983) Anaerobic biotechnology for industrial wastewater treatment. Environ Sci
Technol 17:416A–427A
Vavilin VA, Rytov SV, Lokshina LY (1996) A description of hydrolysis kinetics in anaerobic
degradation of particulate organic matter. J Bioresour Technol 56:229–237
Vavlin VA, Fernandez B, Palatsi J, Flotats X (2008) Hydrolysis kinetics in anaerobic degradation
of particulate organic material: an overview. J Waste Manag 28(6):939–951
Yoon YM (2014) Effects of substrate to inoculum ratio on the biochemical methane potential of
piggery slaughterhouse wastes. Asian-Aust J Anim Sci 27(4):600–607
 Design of a Robotic Firefighting Machine

R. Munzara1, S. M. Ganyani1, and T. Mushiri1,2(&)

1
Department of Mechanical Engineering, University of Zimbabwe,
Mt Pleasant, P.O. Box MP167, Harare, Zimbabwe
[email protected]
2
Department of Quality and Operations Management,
University of Johannesburg, Bunting Road Campus,
P.O. Box APB524, Johannesburg, South Africa

Abstract. The destruction of property and the environment by fires has neces-
sitated the need for the design of an autonomous firefighting robot. Extreme cases
of uncontrolled fires result in injury or even death. This paper presents a design
for a firefighting robot with two degrees of freedom. The robotic arm will be able
to direct a jet of water within a radius of 15 m. The ability of the machine to
manoeuvre on its own, through the use of proximity sensors and flame finders
reduce the risk of exposure to fire and smoke of the operator. The device is
powered by two DC 40 Ah Lithium Ion batteries. At USD$3,455 initial value,
the machine is estimated to depreciate to a value of USD$2,277.13 after five
years, at a depreciation rate of 8%. The cost benefit analysis shows that the
implementation of such a project is justifiable since having such a machine
reduces the risk of losing properties and other valuables to fire. Despite the
undisputed importance of fire, it can also be dangerous and costly if not handled
well. Unwanted fires can result in heavy losses including properties and
important documents. They can also cause injuries or even death. The use of an
automatic machine (robot) to fight fire ensures less risk to firefighters at the same
time increasing efficiency in putting out the fire. The use of robotics and control
enhances accuracy and ease of operation of the machine. Careful design of each
component guarantees fulfilment of the aim, which is putting out fire, with high
efficiency and effectiveness. The cost of the machine is easily overshadowed by
the privileges assumed due to its manufacture. The production of this machine
thus ensures reduced financial losses and increased human safety.

Keywords: Design
Robotic
Fire
Styling
Firefighting

1 Introduction

Fire-fighting is a very imperative but hazardous occupation. A fire-fighter must be in

position to access the fire safely and quickly extinguish the fire, averting further
damage and lessen the possibility of fatalities. Technological advancements have
finally bridged the gap that exist between firefighting and machines allowing for a more
efficient and effective method to employ when extinguishing a fire. Robots are designed
to locate fire, before it rages out of control and prevent it from spreading reducing the
risk of injury to victims.

© Springer Nature Singapore Pte Ltd. 2020

R. Dissanayake and P. Mendis (Eds.): ICSBE 2018, LNCE 44, pp. 567–580, 2020.
https://doi.org/10.1007/978-981-13-9749-3_50
568 R. Munzara et al.

1.1 Background
According to (Fire Service Info, 2016) the first pump designed for firefighting was
created by Ctesibius of Alexandria, Egypt, around the second century BC. Thomas
Lote, an engineer, built the first fire engine made in America in 1743. John Ericsson is
saluted with building the first steam powered fire engine. The same self-propelled
steam engine or pumper was built in New York in 1841. The self-propelled (motorized)
fire apparatus did not gain acceptance until after 1900. The list of great firefighting
equipment inventions through the ages can go on even up to the modern technologies
like the fire sprinkler system.
Despite all the great technologies already in place to combat fire, there is need for
more alternatives to put out fire as different fire scenarios need very different methods
to be used. An apartment fire is so different to a veld fire. Other factors to consider are
cost and ease of use of the available technologies. Not everyone can afford a fire truck.
Portable fire extinguishers may never be effective in putting out fires at a burning farm.
There is need for more efficient, affordable and easy to use means of putting out
different fires such as an automatic fire fighting machine.

2 Literature Review
2.1 Mechanics Behind a Firefighting Device
Storage system: The reservoir tank is the agent storage facility. The water tank is the
simplest as no special requirements other than less corrosive tank material are needed.
For gaseous agents like CO2, high pressure tanks are need [1].
Power Supply: According to [2] a number of power sources can be used depending on
design of the machine. These include 36 V or 24 V 65 Ah gel lead acid batteries, DC
power via umbilical cable from 24 V electrical supply (from vehicle or socket) or
diesel/petrol engines. If an engine is used then a fuel tank should be installed.
Pumping System: Motor; The motor’s basic function is to convert the electrical energy
from the power supply to mechanical energy (rotating shaft) so as to drive the pump.
Pump: Coupled to the motor shaft, the pump’s function is to pump out the firefighting
agent during the firefighting.
Agent Discharge System: Piping.
Various types and sizes of pipes are used to transfer the agent from its reservoir to the
nozzle for discharge. Standardized hose pipes are used when the reservoir is separate
from the machine.
Nozzle: This is the end effector of the robot [1]. The sizes vary from 5/8 to 2 in.
according to the same source. Foam nozzles come in three types; high, medium and
low expansion nozzles.
 Design of a Robotic Firefighting Machine 569

2.2 Robotics
Robotics is an interdisciplinary subject involving information, electronics, mechanics,
automation, and control theory [3]. A robot is an electromechanical device with multiple
degrees of freedom that can be programmed to execute a variety of tasks [4]. Sensors or
the sensory system acquires data on the internal and external environment which involves
design of sensors, data processing, signal conditioning and information retrieval [3].

2.3 General Parts of a Robot

Controller: According to [5] the controller coordinates all the movements of the
mechanical system. It receives inputs from the immediate environment through sensors
and also from the remote control/manipulation by an operator. According to [5] there
are three levels of hierarchical control by the controller. Control signals are sent from
lower levels, while feedback results are sent to the higher levels. Each level depends on
the next level (above it) for instructions.
Level 1 - Actuator Control. This is the elementary level. Separate motions of the
robot along various planes (x, y and z) are controlled here.
Level 2 - Path Control. This level coordinates the various movements along the
planes determined in the first level into the desired path.
Level 3 - Main Control. This is the highest control level. Its function is to interpret the
coded instructions from a human programmer regarding the tasks and works required
to be done. Together with various environmental signals, the instructions are trans-
lated by the controller into more elementary instructions that Level 2 can understand.

Manipulator: The manipulator is an electromechanical device capable of interacting

with its environment [4]. The manipulator has segments that are joined together to
allow the robot to do work. It is the arm of the robot (like the human arm), providing
useful work by moving materials, parts, tools, or special devices through various
motions [5]. Tachometers are used in a robotic system to monitor the acceleration and
deceleration movement of the manipulator [5].
End Effector: An end effector is a tool, gripper, or any other device that is mounted at
the end of a manipulator for accomplishing the task. The area within reach of the robot
end effector is called its work envelope [5]. In the case of the firefighting robot, the end
effector will be the nozzle.
Actuators: Actuators are a motor or mechanism that converts supplied power into
motion. An actuator can also be described as a motor that creates motion in robot
articulations [6]. The Actuator can also be called the stimulus action of a machine or
muscle of a machine. There are basically three types of actuators, each is associated
with a different form of energy, be it electrical, hydraulic or pneumatic [6].
Pneumatic robots are frequently limited to tasks such as pick and place where on or off
control is possible. Pneumatic energy is also implemented in gripper controls, eradicating
the added motor weight [7]. Electric actuators are not as efficient as hydraulic actuators,
570 R. Munzara et al.

especially in large sizes. Hydraulic actuators have the largest capability to lift and hold
their final position without expending much additional power [7].
Sensors: A robot requires sensing to be an active member in the environment. Each
sensor is founded on a transduction principle which is the conversion of energy from
one form to another [5]. Sensors can be classified as proprioceptive and exteroceptive.
Proprioceptive means sense of self or internal state in which the sensor measures values
internally to the system robot for example battery level, wheel position, joint angle.
Exteroceptive means the external state [8].

3 Materials and Methods

The following softwares and methods were used to design the robot
(1) Achieve number of segments of the robot.
(2) Show the mechanical anatomy of the machine.
(3) Determine the end effector and its properties (in this case the nozzle).
(4) Highlight the various softwares used in the design. The writer is to use:-
(a) RoboAnalyzer
(b) SolidWorks
(c) Arduino
(5) Determine the type of controller to be used.
(6) Determine the kind of actuator to be used (AC or DC).
(7) Highlight the programming language to be run on the robot.
(8) Perform the dynamic stability of the machine.

3.1 Efficiency of Firefighting

The main aim of the design is to put out a fire effectively. More focus will be on the
ability of the automated machine to perform its function efficiently before anything
else. The mobility of the firefighting machine in different terrain scenarios will be of
very high importance. Another area of focus is the delivery of the working fluid from
the reservoir, through the nozzle to the unwanted fire. Making these two aspects
efficient will increase the machine’s effectiveness and enhance its functionality.
For this design research various firefighting mechanisms which already exist or
have been proposed were considered and critically analyzed in a bid to develop a new
and improved machine concept. In order to reach the final solution three possible
solutions were generated and compared so as to obtain the best design. To determine
the best possible design, the Binary Dominance Matrix (takes into account various
factors including cost, functionality and weight) is going to be employed. The best was
shown here and it the following technical specifications.

3.2 Design Data

The following data shows the technical specifications of the automated firefighting
machine.
 Design of a Robotic Firefighting Machine 571

Table 1. Technical specifications of the robot for firefighting.

Specification Description
Weight 800 kg
Height 1400 mm
Length 1600 mm
Width 1200 mm
Power sources 24 V rechargeable lead acid batteries or DC power
via umbilical cables from 24 V vehicle supply
Fluid flow rate 700 litres per minute
Range of fluid from 15 m
nozzle
Track system Steel reinforced rubber tracks (heavy duty)
Climbing slope 25°
Speed 10 km/h

4 Results

4.1 Nozzle (End Effector) and Pipe Selection

Nozzle: The nozzle will be the end effector on the robotic arm. The purpose of the
nozzle is to build fluid pressure by restricting flow. The design of the robotic arm is such
that different nozzle types can be mounted and dismounted easily, depending on the
operators’ requirement. However, due to the smooth bore widespread and common use,
this is the nozzle for use in the machine design. The specifications for this nozzle are:-
• Nozzle type - Smooth Bore
• Bore tip - 1-1/2 in. = ð32  25:4Þ mm = 38.1 mm
• Discharge - 15/16 in. = ð1516  25:4Þ mm = 23.813 mm
• Length - 15 mm
• Nozzle Pressure - 50 psi = (50  0.0689476) bar = 3.477 bar
• Flow rate - 700.301 l/min
• Material - Aluminum 6061 Alloy (Refer to Table 1 for properties).

Hose: The robot will make use of a hose coupled from the pump discharge to the
nozzle. The pipe will be pinned to the robotic arm rather than let lose so as to prevent
kinks and pressure loss of water.
• Length of hose - 0.35 m + 0.26 m + 0.39 m = 1 m
• Hose size - 134 in. = 45 mm
• Material - 100% Polyester (+80 °C heat resistant, lightweight, kink resistant, syn-
thetic rubber coating, oil resistant)
• Mass (with water) - Volume  Density = 1.590 kg.
For the selected nozzle, the reaction is 31.298 kg with which the robotic arm can
handle.
572 R. Munzara et al.

PUMP SIZING: A centrifugal pump will be used to pump water or foam during the
firefighting operation.
The pump’s specifications from calculations are as follows
Flow rate q = 0.0117 m3/s
H = 24.905 m + 26.117 m = 51.022

Power required by Pump ¼ P ¼ qgHQ

P ¼ 5:856 kW

4.2 Determination of Length of the Robotic Arm

The arm will have to hold the nozzle in perfect position to fight fire. This is done
basically by determining the length of the upper arm and forearm. The upper arm will
be rigid at 30° from the horizontal and the forearm at a maximum of 50° from the
horizontal at full stretch (Fig. 1).

Fig. 1. Orientation of the robotic arm

The Sine Rule will be used to find x and y:-

Therefore x ¼ 0:35175 m

Also

y ¼ 0:26108 m
 Design of a Robotic Firefighting Machine 573

The length of the upper arm will thus be 0.26 m (260 mm) and the forearm 0.35 m
(350 mm) (Fig. 2).

Fig. 2. Robotic arm assembly

Kinematics
The kinematics of the robotic arm was done using the RoboAnalyzer software. The
position and orientation of each robotic link, with respect to the previous link, is
represented by a set of four DH parameters.
The results of forward kinematics are as shown (Fig. 3).

Fig. 3. Isometric view of the forward kinematics (“quantic” trajectory of the arm)

The graphical results of the upper arm and forearm trajectory and its configuration
(D-H matrix) are shown below (Figs. 4, 5 and 6):-
574 R. Munzara et al.

Fig. 4. Upper arm trajectory (Purple – y plane; Orange – z plane; Lime – x plane)

Fig. 5. Forearm trajectory (Blue – y plane; Brown – z plane; Brown – x plane)

Fig. 6. Acceleration and velocity of the forearm

Sensors and Control

Encoder: Since the actuators to be used to drive the two moving joints are servo
motors, built-in rotary encoders are to be used as the proprioceptor. The Siemens
Multi-turn Absolute Encoder is the proprioceptor to be used.
Proximity Sensor: The robot will make use of the sharp ranger GP2D12 infrared
sensor that detects from within 10 cm to 80 cm of object proximity. The machine will
 Design of a Robotic Firefighting Machine 575

make use of 5 such proximity so as to provide better movement of the device. The
sensors have relatively a high bandwidth that there is no need to model their dynamics.
Flame Detector: The sensor of choice is the phototransistor. The light (from flame) is
detected through a transistor’s base. If light from the flame is detected, the output
voltage will be high and also vice versa. The SFH 309 PFA manufactured by OSRAM
Opto Semiconductors will be the transistor of choice as it has a large wavelength range
for the different types of flames. At 880 nm, the transistor has an acceptance angle of
150°. It also has a rise or fall time of 6 ls, which is best for the design. The device will
have five flame detectors arranged the same as the proximity sensors. Three of the
sensors will be at the front, and two at either side, none at the rear.
Heat Sensor: A thermistor will be used for thermal sensing. The heat sensor will be
located at the front side of the machine. The thermistor will allow the machine to
operate at temperatures below 90 °C, above which it will send a “high” signal to the
controller and robot will move backwards.

4.3 Robot Control (Controller and Programming)

The controller acts as the computer or brain of the computer. This is where the robot
program is fed and run for all the robot operations (including the arm movements and
pumping system). The controller implements specific functions like motion, timing,
sequencing, counting and arithmetic to regulate and control, through analogue or digital
input and output modules. The control of the robot will be done using Arduino and the
Arduino microcontroller. The Arduino software is used to load the control code.
Signals from flame detectors, proximity sensors and the temperature sensor will be used
to control the device using Arduino code (Figs. 7 and 8).

Fig. 7. Isometric view of the firefighting machine front

576 R. Munzara et al.

Fig. 8. Firefighting machine assembly

4.4 Stress Analysis

The stress analysis of two components was done using SolidWorks (Von Misses
theory). The two components for which the analysis was performed are:-
i. Forearm - If it can withstand the nozzle reaction.
ii. Hose pipe - Pressure analysis to ensure it can withstand 60 PSI.
Forearm: With the nozzle reaction at 369.66 N the stress, strain and displacement
analysis of the robotic forearm as performed (Figs. 9, 10 and 11).

Fig. 9. Stress analysis results (forearm)

 Design of a Robotic Firefighting Machine 577

Fig. 10. Stress analysis – force displacement forearm

Fig. 11. Displacement and strain analysis results (forearm)

Hose pipe: The hose pipe was analyzed to see if it could withstand the set pressure of
60 PSI (Figs. 12, 13 and 14).
578 R. Munzara et al.

Fig. 12. Stress analysis of pipe at 60 PSI

Fig. 13. Hosepipe pressure displacement

Fig. 14. Displacement and strain results of pipe assuming it has no free play
 Design of a Robotic Firefighting Machine 579

5 Recommendations and Conclusion

Despite the undisputed importance of fire, it can also be dangerous and costly if not
handled well. Unwanted fires can result in heavy losses including properties and
important documents. They can also cause injuries or even death. The use of an
automatic machine (robot) to fight fire ensures less risk to firefighters at the same time
increasing efficiency in putting out the fire. The use of robotics and control enhances
accuracy and ease of operation of the machine. Careful design of each component
guarantees fulfilment of the aim, which is to putting out fire, with high efficiency and
effectiveness. The cost of the machine is easily overshadowed by the privileges
assumed due to its manufacture. The production of this machine thus ensures reduced
financial losses and increased human safety.

Acknowledgment. Would like to thank UJ for registering for this paper and funding for
everything.

References
1. American Bureau of Shipping (2005) Guidance notes on fire-fighting systems. ABS, New
York
2. Tan CF, Liew SM, Alkahari MR, Ranjit SSS, Said MR, Chen W, Rauterberg GWM,
Sivakumar D, Sivarao (2013) Fire fighting mobile robot: state of the art and recent
development. Aust J Basic Appl Sci 7:220–230
3. Koren Y (2000) Robotics for engineers. McGraw Hill Book Company, New York
4. Williams B (2014) Mechanical engineering. Ohio University, Ohio
5. Ross LT, Fardo S, Masterson JW, Towers RL (2011) Robotics: theory and industrial
application, 2nd edn. The Goodheart-Willcox Co., New York
6. Chirouze MCP (1983) An introduction to robot technology. McGraw-Hill, New York
7. Poole H (1989) Fundamentals of robotics engineering, 1st edn. Van Nostrand Reinhold,
New York
8. Spong MW, Hutchinson S, Vidyasagar M (2004) Robot dynamics and control, 2nd edn.
Adventure Work Press, New York
9. Massachusetts Academy of Mathematics and Science (2010) Inventions that changed
history. Massachusetts Academy of Mathematics and Science, Massachusetts
10. Meyers J (1993) A Short History of the Computer. www.jeremymeyers.com/img/comp/
comp.pdf. Accessed 16 Sept 2016
11. Shende V (2014) Analysis of research in consumer behaviour of automobile passenger car
customer. Int J Sci Res Publ IV(2):3
12. Merriam-Webster (2016) Merriam-Webster. http://www.merriam-webster.com/dictionary/
sport-utility%20vehicle. Accessed 16 Sept 2016
13. DeMuro D (2013) Autotrader. http://www.autotrader.com/car-shopping/suv-vs-crossover-
whats-the-difference-215843. Accessed 16 Sept 2016
14. The-Blueprints.com (2016) Artist, [Art]
15. The Great Soviet Encyclopedia (1979) The Free Dictionary. http://encyclopedia2.
thefreedictionary.com/steering+system. Accessed 16 Sept 2016
580 R. Munzara et al.

16. Griffin K (2016) About Autos. http://usedcars.about.com/od/glossaryatoe/g/Crossovers.htm.

Accessed 16 Sept 2016
17. Kobelev V (2014) Effect of static axial compression on the natural frequencies of helical
springs. Multidiscip Model Mater Struct 10(3):379–398
18. Stander N, Du Preez R (1992) Vibration analysis of coil springs by means of isoparametric
curved beam finite elements. Commun Appl Numer Methods 8(6):373–383
 Design of a Portable and Sustainable
Bicycle Frame

T. Mushiri1(&), R. Gumbo1, and C. Mbohwa2

1
Department of Mechanical Engineering, University of Zimbabwe,
P.O Box MP167, Mt Pleasant, Harare, Zimbabwe
[email protected]
2
Department of Quality and Operations Management,
University of Johannesburg, Bunting Road Campus,
P.O Box APB524, Johannesburg, South Africa

Abstract. This paper is about designing a bicycle frame which is portable as

well as sustainable for the nation of Zimbabwe. It is aimed at taking advantage
of the properties offered by bamboo in the context of a bicycle frame e.g. ability
to absorb vibrations when riding the bicycle on a rough terrain. Aspects of
stiffness and ride quality are discussed in a detailed manner. There is also
information on the three circumstances that affect a structure to work well. In
this paper, frame materials like steel and carbon fiber are studied in view of how
they have survived in the market and how those aspects can be incorporated in
to the material bamboo to survive the market. Advantages and disadvantages of
steel, titanium, aluminium and carbon fiber are viewed in a way to understand
the impact of introducing bamboo as a material of choice for the nation of
Zimbabwe. What makes up bamboo and its mechanical properties are also
studied. The bicycle frame parts are also given attention to in order to view how
they absorb force during the bicycle’s transit i.e. chain-stays, down tube, bottom
bracket, seat-stays, top tube and seat tubes. How the tubes are to be joined is
narrated in the methodology section. This is to answer the issue that in steel and
other materials there is welding to join the tubes to form the bicycle frame; what
about in bamboo bicycles?

Keywords: Bamboo
Bicycle frame
Zimbabwe
Design
Portable

Sustainable

1 Introduction

Bicycles are a means of transport and use no fossil fuels to be propelled. The human
energy used to propel bicycles is renewable and largely non-polluting. In a way,
bicycles provide a means to lose weight in the saddle [4]. The bicycle is the primary
mode of mobility for millions of people throughout many poorer parts of the world. In
addition to personal transport, they see a vast number of applications including moving
goods to market, the sick to hospital, run errands and delivering of mails, taking
vacations and even the distributing medicines. Bicycles encourage human interaction,
communication, and connectivity [1]. The cost to construct and maintain quality bike
trails is much lower than the cost to construct new roads and/or widen existing roads
© Springer Nature Singapore Pte Ltd. 2020
R. Dissanayake and P. Mendis (Eds.): ICSBE 2018, LNCE 44, pp. 581–594, 2020.
https://doi.org/10.1007/978-981-13-9749-3_51
582 T. Mushiri et al.

[2]. In Zimbabwe, quite a number of people can own cars or even motorcycles and
people without bicycles have to rely on inadequate and relatively undependable buses
for short journey, say less than 30 km. Provision of bicycles to students can prove
significant in reducing travel time and improve educational outcomes [3].
Background and Literature Review
A bicycle frame is the supreme component of a bicycle [11], onto which wheels and
other bicycle accessories are fitted. The current and most common frame design for an
upright bicycle is based on the safety bicycle consisting of two triangles: a frontal main
triangle and a paired rear triangle.
In 1817, one Karl Barn von Drais [12, 13] devised a frame that enabled steering the
front wheel. Heaviness and cumbersome was associated with the then bicycle. For
many years until the present day, inventors and engineers have been busy coming up
with better handling frames, which are lighter-in-weight, stiff capable of travelling long
routes without disintegrating. To this day cyclists, inventors and engineers seek to
make today’s frames better than yesteryears in various ways [10] (Fig. 1).

Fig. 1. a. Aluminum bicycle frame b. Carbon fiber bicycle frame c. Titanium bicycle frame d.
Steel bicycle frame

Aluminum has been used in bicycle frames since the 1890’s. The frame has an
advantage of needing no paint [19]. Being oversized aluminum does have a very un-
forgiving characteristic when overstressed or cracked, it can fail very suddenly. It is of
importance to inspect aluminum frames after an accident for any sign of cracking or
stress, even if the bike rides perfectly. Carbon fiber have a very strong strength to
weight ratio. Custom carbon fiber frames are quite expensive. Machining of titanium is
difficult evidenced by requiring a lot of extra steps in the welding procedure. The
material is not readily available, thus much more expensive to purchase. Titanium is
more durable as a modern bicycle frame material. Steel is popular as a bicycle frame
and very easy to machine. Welding steel doesn’t require all of the special tooling and
steps that welding titanium does. It also takes less time and less electricity than alu-
minum welding [19]. Steel is the standard that majority of frame materials are com-
pared to.
 Design of a Portable and Sustainable Bicycle Frame 583

1.1 Problems Associated with the Current Materials

Steel, aluminum, carbon fiber and titanium frames are very difficult to repair according
to the British cycling organization. The aims and objectives of this research are to
reduce the cost of bicycle frame by at least 10% and to design a sustainable bicycle
frame.

1.2 Bicycle Frame Layout

The frontal, facing down, triangle and rear triangle connection are responsible for the
structure of the bicycle frame. The diamond design has proven to be a great use of
materials and the strength of the design comes from the triangle shapes [20]. Head tube,
top tube, down tube and seat tube comprise the front triangle. Seat-tube, a pair of chain-
stays and a pair of seat-stays make up the rear triangle. The two triangles are connected
by use of a bottom bracket found at the lower end of the seat-tube (Fig. 2).

Fig. 2. The bicycle frame [15]

Every bicycle has a resonant frequency at which it oscillates in a different way.

A metal tube ring can be heard when the tube is tapped with a fingernail. However,
resonances below the audible range only affects the feel of a bicycle. For example, the
ringing of the frame in torsion as the frontal triangle tilts relative to the rear triangle,
excited by the front wheel’s turning slightly to the right and left [5]. According to Allen
and Brown, the front triangle is less stiff in torsion and the resonant frequency is lower.
Resonances affect the feel when going over bumps. This implies that two frames which
are equally stiff feel different when they resonate at different frequencies. According to
the Oxford dictionary, damping is a reduction in the amplitude of a vibration or
oscillation [14]. Every metal frame has a very low damping such that when they ring
long enough a clear tone is produced. Carbon fiber frames will give dull sounds when
tapped because carbon fiber has more damping in contrast to metal. The frame feel is
much less at the low frequencies. Rubber is a material that is highly damped and so is
leather. Brown and Allen alludes that the greatest damping in a bicycle is on the tires
584 T. Mushiri et al.

and the saddle unless the bicycle has suspension [16]. A suspension fork or frame is a
very much damped resonant system [21]. If damping was absent, the frame would
bounce up and down repeatedly after every bump. Suspension adds weight, hence
contribute to the feel obtained from the frame. The materials of the bicycle are as
shown in Table 1.

Table 1. Table of properties of bicycle frame materials [7, 8, 17, 19]

Material Yield strength Ultimate tensile strength Density
(MPa) (MPa) (g/cm3)
Steel, structural ASTM a36 250 400–550 7.8
steel
Steel, 1090 mild 247 841 7.58
Chromium-vanadium steel 620 940 7.8
AISI 6150
Aluminum alloy 6061-T6 241 300 2.7
Carbon fiber N/A 4137 1.75
Bamboo 350–500 0.4
6A1/4V worked Titanium 1085 1220 4.43
[12]

1.3 Bamboo as a Material

Bamboo is a natural composite material comprising cellulose fibers embedded in a
lignin matrix [10]. The cellulose fibers are aligned along the length of the bamboo, thus
providing maximum tensile strength and rigidity in that direction. Furthermore, the
fibers are recognizably more abundant on the outermost part of the bamboo stem while
the inner part of the stem is hollow [10]. This evidently means the bamboo is strong,
stiff, and light (Fig. 3).

Fig. 3. a. Plantation of bamboo b. Bamboo tubes c. Bamboo sectioned tubes and way of
connection [10, 18]

2 Materials and Methods Used for Design

For mechanical drawings, Solid Works 2015 was used. The specifications are shown in
Table 2. Stresses were also determined by the finite element method.
 Design of a Portable and Sustainable Bicycle Frame 585

Table 2. Bamboo technical specifications

Description (N/mm2)
Modulus of elasticity 16.17
Bending strength 20.27
Compressive strength 7.86
Tensile strength 14.96
Longitudinal shear strength 1.41

The designed bamboo frame was drawn using SolidWorks as shown in Fig. 4.

Fig. 4. Bamboo chosen shape

The length and shape of the bicycle frame tubes, together with the angles at which
they are set, describes the importance of the bicycle design. The consequences would
be how well the bike climbs elevated routes, how urgently it accelerates, and how
stable it is, perceived at speed and through bends and corners. Importantly, geometry
will have an immense influence on how comfortable the bike feels beneath different
physiological types. For design the following are considered: There is need to visualize
the stresses the frame will incur in each of the listed scenarios below:
(1) Sudden braking at top speed to prevent a frontal collision - fork stress or hori-
zontal compaction.
(2) Riding uphill - out of a low part of a ridge.
(3) Jumping off a raised path.
(4) Turning aggressively i.e. leaning and shifting your body weight at an intersection.
Using anthropology data for Zimbabwe, the average human as 75 kg, the weight of
frame as 2.5 kg and mass density of bamboo as 350 kg/m3 [22].
586 T. Mushiri et al.

3 Simulation and Results

For the frame: Moments about A and using support reactions in Newton’s Third Law
we get the forces and they are derived as follows: þ ÝMA ¼ 0 (Figs. 5, 6, 7, 8, 9, 10,
11, 12, 13 and 14) and (Tables 3 and 4);

Fig. 5. Frame dimensional design.

Fig. 6. Chain stay

24:5ð0:5625Þ þ 735:8ð0:5625 þ 0:45 cos 75:3 Þ  R2 ð0:5625 þ 0:4975Þ ¼ 0

 R2 ¼ 483:5 N þ " RFy ¼ 0R1 þ 483:5 ¼ 735:8 þ 24:5; R1 ¼ 276:8 N

Let FSS -be force on seat stay

FDT -be force on down tube
FCS -be force on chain stay
FTT -be force on top tube
FHT -be force on head tube
FST -be force on seat tube
Moments about E;

þ ÝME ¼ 0;

FSS sin 53:8 ð0:4625Þ  483:5 cos 5:1 ð0:4625Þ ¼ 0

 Design of a Portable and Sustainable Bicycle Frame 587

 FSS ¼ 596:8 N þ ! RFx ¼ 0; 596:8 cos 53:8 þ 483:5 sin 5:1 þ R3x ¼ 0;

R3x ¼ 395:5 N
þ " RFy ¼ 0;
483:5cos5:1  596:853:8 þ R3y ¼ 0;
R3y ¼ 0 N

Fig. 7. Seat stay

Fig. 8. Seat tube

Moments about D; þ ÝMD ¼ 0; FCS sin 53:8 ð0:525Þ þ 483:5 sin 41:3 ð0:525Þ ¼ 0;

FCS ¼ 395:4 N
þ ! RFx ¼ 0;
394:4 cos53:8  483:5 cos41:3 þ R4x ¼ 0;
 R4x ¼ 596:8 N
þ " RFy ¼ 0;
395:4 sin53:8 þ 483:5 sin41:3 þ R4y ¼ 0;
 R4y ¼ 0 N

Moments about E: þ ÞME ¼ 0; FTT sin75:3 ð0:45Þ  735:8 sin14:7 ð0:45Þþ

596:8 sin56 ð0:45Þ ¼ 0;
588 T. Mushiri et al.

) FTT ¼ 318 N þ ! RFx ¼ 0;

318:5 cos75:3 þ 596:8 cos56  735:8 cos14:7 þ R5x ¼ 0; ) R5x ¼ 297:2 N

þ " RFy ¼ 0;
R5y þ 596:8 sin56  318:5 sin75:3  735:8 sin14:7 ¼ 0; ) R5y ¼ 0 N

Fig. 9. Head tube

Fig. 10. Seat stay

Fig. 11. Down tube

 Design of a Portable and Sustainable Bicycle Frame 589

Moments about A: þ ÞMA ¼ 0; 318sin75:3 ð0:45Þ  FDT sin66:4 ð0:4Þ ¼ 0;

) FDT ¼ 377:6 N þ ! RFx ¼ 0;

276:8 cos14:7  377:6 cos66:4 318 cos75:3 þ R6x ¼ 0;

) R6x ¼ 35:1 N
þ " RFy ¼ 0;
276:8 sin14:7  377:6 sin66:4 þ 318 sin75:3 þ R6y ¼ 0;
) R6y ¼ 31:8 N

Moments about F: þ ÞMF ¼ 0; 318:5 sin48:7 ð0:525Þ þ 735:8 sin41:3

ð0:525Þ þ FST sin56 ð0:525Þ ¼ 0;

FST ¼ 297:2 N þ ! RFx ¼ 0;

R7x þ 735:8 cos41:3 þ 318:5 cos48:7 þ 297:2 cos56 ¼ 0;

) R7x ¼ 979:2 N þ " RFy ¼ 0;

R7y þ 318:5 sin48:7  735:8 sin41:3  297:2 sin56 ¼ 0;

) R7y ¼ 493:1 N

Moments about E: þ ÝMD ¼ 0; FHT sin66:4 ð0:625Þ þ 276:8 sin51:7 ð0:625Þ ¼ 0;

) FHT ¼ 237:1 N þ ! RFx ¼ 0;

R8x þ 237:1cos66:4  276:8cos51:7 ¼ 0; ) R8x ¼ 76:6 N

þ " RFy ¼ 0; R8y  237:1 sin66:4 þ 276:8 sin51:7 ¼ 0; ) R8y ¼ 0 N

Fig. 12. Bicycle frame body diagram

590 T. Mushiri et al.

Fig. 13. Shear Force Diagram (SFD)

Fig. 14. Bending Moment Diagram (BMD).

Table 3. Name of frame and parameters

Name of Length Diameter Thickness Deflection Force
frame (m) (mm) (mm) (mm) (N)
FSS 0.5248 25 3.1 8.2 617.3
FDT 0.6018 37.5 3.1 4.1 404.4
FCS 0.4625 25 3.1 4.9 409.0
FTT 0.5625 37.5 3.1 3.2 336.0
FHT 0.05 50 6.25 1.0 225.9
FST 0.45 37.5 6.25 1.1 281.3

Let rmax = maximum stress, smax = maximum shear stress, e = strain of the bam-
boo tube
D = diameter of the bamboo tube, t = thickness of the bamboo tube,
rallowable = allowable shear stress,
rfailure = ultimate strength of bamboo, Ebamboo = young’s modulus of bamboo,
d = deflection, Choosing a safety factor (FS) of 3, rfailure = 425 MPa, E = 16.2 MPa
¼ 425 MPa ¼ 141:7 MPa
r
FS ¼ failure ; ) r
rallowable allowable 3
 Design of a Portable and Sustainable Bicycle Frame 591

Table 4. Stresses in different areas on the bicycle.

Place on the bicycle Stress Safe or not safe
Chain stay rmax ¼ 1:62 MPa\141:7 MPa ¼ rallowable Safe
Seat stay rmax ¼ 2:45 MPa\141:7 MPa ¼ rallowable Safe
Seat tube rmax ¼ 0:404 MPa\141:7 MPa ¼ rallowable Safe
Down tube rmax ¼ 1:03 MPa\141:7 MPa ¼ rallowable Safe
Head tube rmax ¼ 0:241 MPa\141:7 MPa ¼ rallowable Safe
Top tube rmax ¼ 0:871 MPa\141:7 MPa ¼ rallowable Safe

r  10 6
von Misses Stress, rv : ry rv ¼ Nðderived from factory of safety diagramÞ rv ¼ 20:26
1:402  102 ¼
1:445  105 N=m2
Since rv \ry ; design is safe (Figs. 15, 16, 17 and 18) (Table 5).

Table 5. Costing related to the bamboo

Material Units Thickness Dimensions Cost ($)
Bamboo tube 1 6.25 mm Dia. 50 mm * length 300 mm 3.00
Bamboo tube 3.1 mm Dia. 37.5 mm * length 475 mm 4.00
Bamboo tube 3.1 mm Dia. 37.5 mm * length 575 mm 5.00
Bamboo tube 6.25 mm Dia. 37.5 mm * length 500 mm 5.00
Steel drop outs 2 5 mm Dia. 25 mm 1.00
Bottom-bracket (steel) 1 6.25 mm Dia. 50 mm * 40 mm 3.00
Bamboo tube 3.1 mm Dia. 37.5 mm * 600 mm 5.00
Bamboo tube 3.1 mm Dia. 25 mm * 525 mm 4.00
Wrapping tape 12 15 mm * 5000 mm 0.30
Hemp fiber 20 5000 mm * 10 mm * 2 mm 0.25
Total 39.60

Fig. 15. von Mises Stress

592 T. Mushiri et al.

Fig. 16. Displacement analysis

Fig. 17. Factor of Safety (FOS).

Fig. 18. Strain

4 Discussion and Recommendations for Further Work

People who seek to be their own employers should see this project as an idea to acquire
finances. It is recommended to establish the bamboo frame building factory on the
bamboo plantation to cut down costs mentioned in the bill of materials i.e. tubes will be
acquired for free. For further savings, the government of Zimbabwe should reduce
 Design of a Portable and Sustainable Bicycle Frame 593

importation of bicycle from the Far East and start harvesting more bamboo plants in
Zimbabwe, especially in provinces where rainfall is limited. It is recommended that
maximum loading on the frame is not to exceed 100 kgs. For better enjoyment, store
frame in a dry storeroom. Since some bamboo will be cut, afforestation should be done
more often and policies of plantation should be encouraged. Zimbabwe has a National
Tree Planting Day which will support this project of bamboos as raw materials for
bicycles in abundance and well replaced [23]. To cater for comfort, it will be necessary
to check the reach and make it long, handle height and add spacers under the stem,
make the saddle comfortable by covering with canvas and adjust height and angle and
lastly tyre pressure should not be too much but average.

5 Conclusion

The bamboo bicycle frame is a successful project. Objectives of the frame were real-
ized; being portable as well as the use of sustainable material in bicycle frame make.
The cost was reduced in the forming of the frame including cutting cost on importation.
This can be used but one thing to be noted is that forces are less than those of steel
hence average people should cycle general to make it durable. This research is very
useful to Zimbabwe.

Acknowledgment. Would like to thank UJ for paying the registration fees and avail resources
for paper to be presented in Sri Lanka.

References
1. Agur L (2013) Carbon fiber vs Aluminium vs Steel vs Titanium i love bicycling. http://www.
ilovebicycling.com. Accessed 9 Oct 2016
2. Bamboo bicycle club (n.d.) Frequently Asked Questions About Bamboo Bikes. http://
bamboobicycleclub.org/. Accessed 16 Mar 2017
3. Bamboo Bike Supplies (2015) Composite Lug Wrapping with Hemp Fibers. https://
bamboobikesupplies.com. Accessed 19 Apr 2017
4. Barbour M (2016) 30 reasons to take up cycling. www.bikeradar.com. Accessed 15 Sept
2016
5. Brown S, Allen J (2010) Frame Materials for the Touring Cyclist. http://www.sheldonbrown.
com/framematerials. Accessed 21 Feb 2017
6. DeMerchant C (2017) Youngs Modulus is a Measure of Stiffness. http://www.
christinedemerchant.com/youngmodulus.html. Accessed 21 Mar 2017
7. Gerguri J, Broughton J (2016)
8. Hearn EJ (1997) Mechanics of materials, 3rd edn. Butterworth Heinemann, United Kingdom
9. Herlihy D (2004) Bicycle: the history. Yale University Press, New Haven
10. Instron (2012) Bamboo Bikes - A “Growth” Industry. http://www.instron.us/. Accessed 20
Mar 2017
11. Kolin MJ, de la Rosa DM (1979) The bicycle frame. In: Lepley TA (ed) the custom bicycle, p 1
12. Maleque MA, Hossain MS, Dyuti S (2014) Material properties and design aspects of folding
bicycle frame. https://www.researchgate.net/publication. Accessed 7 Mar 2017
594 T. Mushiri et al.

13. Madyda K, Nixon N (n.d.) Bike history. http://nickandkevin.weebly.com/. Accessed 10 Mar

2017
14. Mancone M (2008) Brief bicycle history. iml.jou.ufl.edu. Accessed 15 Sept 2016
15. Moyo J (2016) Zimbabweans spot green gold in bamboo that spares forests. http://www.
reuters.com/article/us-zimbabwe-environment-forests-bamboo-idUSKCN10L1DP. Accessed
7 Mar 2017
16. Mozer D (n.d.) Bicycle history timeline. www.ibike.org. Accessed 15 Sept 2016
17. Palermo E (2014) Who Invented the Bicycle? http://www.livescience.com/. Accessed 10
Mar 2017
18. Reiss-Davis Z (2012) What are the benefits of a bamboo bicycle frame? www.quora.com.
Accessed 16 Mar 2017
19. Rodriguez D (2009) Custom bicycles and frame materials. http://www.rodbikes.com.
Accessed 22 Sept 2016
20. Salvagione P (n.d.) Science of Cycling: Bicycle Frame Design—Exploratorium. https://
www.exploratorium.edu/cycling/frames4.html. Accessed 21 Feb 2017
21. The Engineering Toolbox (n.d.) Modulus of Elasticity or Young’s Modulus - and Tensile
Modulus for common Materials. http://www.engineeringtoolbox.com/young-modulus-d_
417.html. Accessed 21 Mar 2017
22. Walker R (2011) Mass, weight, density or specific gravity of wood. https://www.simetric.co.
uk/si_wood.htm. Accessed 10 May 2017
23. Bulawayo24, 3 December 2013, Zimbabwe National Tree Planting Day. https://bulawayo24.
com/index-id-news-sc-national-byo-39716-article-Zimbabwe+National+Tree+Planting
+Day.html. Accessed 17 July 2018
 Design for Manufacture and Assembly
of an Automated Dish Washing Machine

Tawanda Mushiri1(&), Rutendo F. Solomon2, and Charles Mbohwa3

1
Department of Quality and Operations Management,
Faculty of Engineering and the Built Environment, University of Johannesburg,
Auckland Park, 2006 Johannesburg, South Africa
[email protected]
2
Department of Mechanical Engineering, University of Zimbabwe,
P.O Box MP167, Mt Pleasant, Harare, Zimbabwe
3
Faculty of Engineering and the Built Environment,
University of Johannesburg, Auckland Park, 2006 Johannesburg, South Africa

Abstract. This paper was a research done at a local university in Zimbabwe.

Dishwashing by hand involves scrubbing, rinsing and drying of dishes which is
time consuming and laborious when done on a large scale. Each dining hall at
the university uses about one thousand plates per meal and the university is
understaffed thus the dishes are not thoroughly cleaned. A dishwashing machine
is thus a necessity for the university. Researching of literature was done and
information on the current designs of dishwashers, their operations, merits and
demerits was gathered. A case study on the university was carried out using
interviews and questionnaire. The dishwasher for the university was designed
and it uses pressurised water jets to clean the plates. The pressurised water jets
are produced by a pump installed in the dishwasher. The plates are put into the
dishwashing using a rake conveyer system which is automatically controlled.
The von Mises stresses of the system were analysed using Solidworks and they
were in the desired range. Deflections were also done using Solidworks.
Specifications and drawings of the dishwashing machine were developed.
Materials of the design were selected concentrating on reduction of cost through
minimising the number of parts retaining its function.

Keywords: Design for manufacture and assembly (DFMA), automated

Dish washing machine

1 Introduction

Dish washing machines are devices that are used to clean and dry dishes and kitchen
utensils. When using a dish washing machine the dishes are put into the racks and the
machine is closed, the dishes are showered by pressurised, hot and soapy water jets
until they are clean. This machine can clean any type of kitchen utensil. The cleaning of
dishes by hand involves scrubbing, rinsing and drying of the dishes. There is use of
detergents in the washing and some of the detergents have long time effects on the
hands [1]. The process of hand dish washing takes a lot of time and consumes a lot of
energy and water [6]. This has resulted in the design and manufacture of dish washing

© Springer Nature Singapore Pte Ltd. 2020

R. Dissanayake and P. Mendis (Eds.): ICSBE 2018, LNCE 44, pp. 595–603, 2020.
https://doi.org/10.1007/978-981-13-9749-3_52
596 T. Mushiri et al.

machines that can clean dishes using minimum energy, minimum water and taking less
time [9]. The existing dish washing machines are expensive and the maintenance costs
are high thus there are not common in Zimbabwe.

1.1 Background of the Study

The first dishwashing machine was manufactured in the middle of the 19th century.
A patent of the first device was signed in 1850 by Joel Houghton [2]. The machine was
crank operated and built from wood. When cranked, water would make its way through
the wood plumbing and spray over the dishes. A woman named Josephine Cochrane
who was very wealthy improved on that invention and manufactured the first suc-
cessful dish washing machine [7]. These dishwashers could not dry the dishes thus the
dishes were dried by hand. The water for washing was also heated manually using
wood thus they were not effective. They also wasted a lot of water.
Dish washing technology continued improving and the recent dish washing
machines are far much better than those in the 19th and 20th centuries. The recent
dishwashing machines are now more helpful than ever. They have a water efficiency of
about 4 to 6 gallons of water per cycle [4]. The electricity usage has reduced by about
15%. These dishwashers produce less sound than the previous ones. Some of them
even have soil sensors and innovative dish rack designs. Due to the increase in water
and energy efficiency, the recent machine take a lot of time as compared to those of the
20th century. The 20th century machines took 60 to 80 min per cycle but the recent
ones take about 120 to 140 min. All the improvements have led to the increase of the
cost of dishwashing machines thus they are not common in developing countries.
Dishwashers are a reproducing ground for possibly unsafe organisms, new research
has uncovered. Scientists found that the warm, soggy condition in dishwashers is
perfect for the development of growths including Exophiala, Rhodotorula, and Candida
parapsilosis – yeast-like species that can be hurtful to individuals. On the off chance
that breathed in, these parasites can colonize the lungs and cause contaminations that
can be hard to treat. The elastic seals in dishwasher entryways are an ideal rearing
ground for organisms and that a few sorts, for example, Exophiala.
About 500 plates, 1000 pieces of cutlery and 500 cups are used per meal in a dining
at the university, there is no enough manpower for dishwashing thus dishes are not
thoroughly cleaned.
The major objectives of the research are:
• To design a dish washing machine for the university that washes 500 plates, 1000
pieces of cutlery and 300 cups per hour.
• To design a machine which requires minimum or zero human assistance.
• To design a machine that can be manufactured by locally available resources.
 Design for Manufacture and Assembly of an Automated Dish Washing Machine 597

2 Literature Review

Dish washing involves preparation, filling, washing, rinsing and drying (Fig. 1).

Fig. 1. a. Parts of a dishwasher b. Dishwasher cleaning cycle

Despite the fact that part of human exercises are robotized in the present aggressive
world. There is a slack in robotized dish washer. A few machines are as of now planned
with the assistance of high speed water jet; henceforth there is a possibility of
uncleanliness and not evacuating extreme strains in dishes. Keeping in mind the end
goal to defeat the above issues an exceptional machine called―Automatic Dishwasher
Machine” with rollers, brush and water fly presented in this work. By consolidating
dishwasher with roller component utilizing brush and water stream, it is conceivable to
clean the articles viably than alternate machines and with no labor [11].

3 Materials and Methods Used for Design

For mechanical drawings, Solid Works 2015 software was used. SolidWorks (stylized
as SOLIDWORKS) is a solid modeling Computer-Aided Design (CAD) and
Computer-Aided Engineering (CAE) computer program that was run on Microsoft
Windows in this research. It is a solid modeler, and utilizes a parametric feature-based
approach to create models and assemblies. The dish washer designed has the following
parameters as in Table 1. The selection was done as steps in SolidWorks. The response
of a section relies upon the material selected to the part. SimulationXpress must know
the flexible properties of the material of the part. You allocate material to the part by
picking a material from a material library. Materials in SOLIDWORKS have two
arrangements of properties, visual and physical (mechanical). SimulationXpress utilizes
the physical properties of the materials characterized in the SOLIDWORKS Material
library. Materials can be isotropic, orthotropic, or anisotropic. SimulationXpress bol-
sters isotropic materials as it were (Table 2).
598 T. Mushiri et al.

Table 1. Parameters used in the design

Parameter Quantity
Water efficiency 10 gallons per cycle
Time per cycle 15 min
Number of plates per cycle 300 plates
Type of plates Melamine
Energy efficiency 10 joules per circuit
Number of rakes 2
Dimension 0.7  0.7  1 m
Plate area 0.0625 m2
Plate thickness 3 mm

Table 2. Components and quantities to be designed

Component Quantity
Centrifugal pumps 2
Spray arms 2
Water collector pipes 1
Racks 2
Housing 1

To Assign/Modify a Material to a Part:

1. Click Choose Material (Material tab of the SimulationXpress wizard).
2. In the Material dialog box, expand the class of materials and select a material.
3. Click Apply.
4. Click Close
5. Click Next.
You can also assign a material by right-clicking the part in the SimulationXpress
study tree and clicking Apply/Edit Material (Fig. 2).
Approximately 12,000 students need to eat three times a day.

Fig. 2. Designed dish washing machine

 Design for Manufacture and Assembly of an Automated Dish Washing Machine 599

4 Simulation and Results

4.1 Water Jet Arrangement
The arrangement of the jet of water for design was shown below in Fig. 3.

Fig. 3. a. Water jets on horizontal plate b. Inclined plate

Fx = qaV2 = 1000  3.14  10−6  3.982 = 0.0497 N (Horizontal plane). This

is the force acting in horizontal plane.
Fx = qAv2 sin2 h = 1000  3.14  10−6  3.982  sin2 60°; Fx = 0.0373 N for
inclined plane.
From the Darcy equations the researchers calculated the losses due to friction,
h ¼ 4flv
2dg ; but Re ¼
2 1030:41:141
1:1410 = 14035 and the flow is turbulent, so the Darcy equation
must be used. To determine the value of friction factor:
Relative roughness = k/d = 0.000 08*0.04 = 0.002. From the Moody chart in
appendix 1, for Re = 1.4  10−4 and relative roughness of 0.002.
f = 0.008; h ¼ 4x0:008x1:65x3:98x3:98
2x9:81x0:030 ; hf = 1.1 m, hi = 1.1 + 0.55 + 1.25 + 1.5; hi =
4.5 m; Energy = qghQ = 1000*9.81*4.5*5  10−5 = 0.720725 KW. 1 HP pump
with a flow rate of 30 litres per minute with a head of 4.5 m and pressure of 0.2 bar was
selected (Fig. 4).

Fig. 4. a. The stress analysis of the shaft b. The displacement analysis of the shaft
600 T. Mushiri et al.

4.2 Stress Analysis

p 100  103 p 120  103

rx ¼ ¼ ¼ 18:7MPa; r y ¼ ; ¼ ¼ 33:3MPa
pr2 4  30  103 2prl 2  30  103
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
rx þ ry ðrx  rx Þ2
r1;2 ¼ þ þ s2xy ;
2 2
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ð16:7 þ 33:3Þ  106 ð16:7  106  33:3  106 Þ2
¼ þ þ 200  1052 ¼ 48:2MPa
2 2
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
rv ¼ r2x þ r2y  rLr2x þ 3s2xy ;
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ð16:6  106 Þ2 þ ð33:6  106 Þ2  ð16:8  106  33:5  106 Þ þ ð3  ð10  106 Þ2 Þ
Sy 350
¼ 49:1MPa; rd ¼ ;¼
n 3
¼ 116MPa:

4.3 Stress Analysis on the Melamine Plate

qffiffiffi qffiffiffiffiffiffiffiffiffiffi
Calculation of slenderness ratio is as follows: S ¼ Ler; But r ¼ l
A; r¼ 0:03
0:0067;
r ¼ 12:33
sffiffiffiffiffiffiffiffiffiffiffi rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
0:03 2p2 E 2p2 24  109
S¼ ; S ¼ 0:0045; Scr ¼ ; Scr ¼ ; Scr ¼ 204:56
12:33 ry 200  106

Calculation of the critical load: Since S = 0.0045 < Scr = 204.6 the melamine plate
is a short column thus the following equation is used for the critical load
r 2 r 2
Pcr ¼ ry Að1  ð4p2yEÞ Ler ; Pcr ¼ ry Að1  ð4p2yEÞ Ler ; Pcr ¼ 4:345 kN.
Since 4.345 is > 1.020 from the water jets the melamine plates will fail due to the
water pressure. The detailed drawings are shown in Fig. 5.
 Design for Manufacture and Assembly of an Automated Dish Washing Machine 601

Fig. 5. a. Detailed full drawing b. Top view c. Front view d. Right hand side e. Left hand side
f. Bottom side g. Back view

5 Discussion and Recommendations

The number of pumps was reduced from four to two since the dirty water from the
dishes will be drained by gravity the water collector then to the sewer system thus no
need for the two pumps to pump dirty water from the washer. The following were
allocated materials as follows and suggested prices to buy the materials were put
across. The dishwasher uses a lot of detergents to disinfect the plates. This can be
improved by using pressurised heated water. Heated water will also resulting in the
dishes drying faster (Table 3).

Table 3. Part list and materials

Part name Description Quantity Unit price (US$) Cost
Drive shaft Carbon steel 1 $2.00 $2.00
Motor 1 $90.00 $90.00
Pump 1 $100.00 $100.00
Bearing Gray cast Iron SN 1 $11.00 $11.00
Belt Steel 1 $30.00 $30.00
Pulley Brass 2 $11.00 $22.00
Spray arms Copper 2 $5.00 $10.00
Pipes Copper 30 $1.00 $30.00
Housing Cast stainless steel 1 $10.00 $10.00
Water collector Gray cast iron SN 1 $2.00 $2.00
Idler Gray cast iron SN 2 $2.50 $5.00
Stand Cast stainless steel 1 $4.50 $4.50
Pipe connectors Copper 20 $0.65 $13.00
Total $329.50
602 T. Mushiri et al.

6 Conclusion

The research was a success, next step is to install this dishwasher at the university. The
university community will benefit a lot in coming up with this dishwasher. A dish-
washer is a mechanical gadget for cleaning dishware and cutlery. Not at all like manual
dishwashing, which depends generally on physical scouring to expel dirtying, the
mechanical dishwasher cleans by showering high temp water, ordinarily in the vicinity
of 45 and 75 °C (110 and 170 °F), at the dishes, with bring down temperatures utilized
for sensitive items. A blend of water and dishwasher cleanser is pumped to at least one
turning shower arms, which shoot the dishes with the cleaning blend. Once the wash is
done, the water is depleted, more heated water enters the tub by methods for an electro-
mechanical solenoid valve, and the flush cycle starts. After the flush cycle completes
and the water is depleted, the dishes are dried utilizing one of a few drying techniques.
Ordinarily a flush guide, a synthetic to decrease surface pressure of the water, is utilized
to lessen water spots from hard water or different reasons. Plates will be very much
clean than manual washing.
Notwithstanding local units, mechanical dishwashers are accessible for use in
business foundations, for example, lodgings and eateries, where a substantial number
of dishes must be cleaned. Washing is led with temperatures of 65–71 °C (149–160 °F)
and sanitation is accomplished by either the utilization of a promoter warmer that will
give a 82 °C (180 °F) “last flush” temperature or using a substance sanitizer.
The new age best-in-class machine beat the earlier age unit by a vast edge. It did
likewise measure of work however utilized significantly less vitality and water. The
advancements on the new designed dishwasher at university are at present available
and accessible for manufacture and use. These advancements increment the produc-
tivity of the dishroom and take care of numerous upkeep issues, and likely increment
the lifetime of the machine while guaranteeing appropriate machine task. The generous
cost investment funds and low straightforward payback time influence a sensible
business to case to resign even high-proficiency machines early.
 Design for Manufacture and Assembly of an Automated Dish Washing Machine 603

Appendix 1: Moody Diagram

References
1. Barro N, Bello AR, Savadogo A, Ouattara CAT, Iiboudo A, Traoré AS (2006) Hygienic
status assessment of dish washing waters, utensils, hands and pieces of money from street
food processing sites in Ouagadougou (Burkina Faso). Afr J Biotechnol 5
2. Emmel J, Parott K, Beamish J (2003) Dishwashing and water conservation: an opportunity
for environmental education. J Ext 41
3. Flannery JP (1956) Dishwashers. Google Patents
4. Gram M, Schweier P (2016) Dishwasher with a dynamic filling sequence. Google Patents
5. Pruehs H, Schenker G, Altenschoepfer T, Piorr R (1990) Use of hydroxyalkyl polyethylene
glycol ethers as surfactants in rinse aids for dishwashing machines. Google Patents
6. Stamminger R (2011) Modelling resource consumption for laundry and dish treatment in
individual households for various consumer segments. Energ Effi 4:559–569
7. von Hippel E, Ogawa S, de Jong JP (2011) The age of the consumer-innovator. MIT Sloan
Manage Rev 53:27
8. William D, Andrews DC (1974) Technology and the housewife in nineteenth-century
America. Women’s Stud Interdisc J 2:309–328
9. Wood G, Newborough M (2003) Dynamic energy-consumption indicators for domestic
appliances: environment, behaviour and design. Energ Build 35:821–841
10. Venkatesh RB, (2017) Design and fabrication of automatic dishwasher machine. In:
International Conference on Emerging trends in Engineering, Science and Sustainable
Technology (ICETSST), E-ISSN: 2348 – 8360, pp 26–31
 Integration of BIM and Construction Supply
Chain Through Supply Chain Management;
An Information Flow Model

A. P. Rathnasinghe(&), M. K. C. S. Wijewickrama, U. Kulatunga,

and H. S. Jayasena

Department of Building Economics, University of Moratuwa,

Moratuwa 10400, Sri Lanka
[email protected]

Abstract. Building Information Modelling (BIM) is a revolutionary stride of

technology in the orthodox construction and procedures in conventional
Architecture Engineering and Construction (AEC) industry which is stressed
upon for its delay on embracing new technologies. The strength of BIM has
been disvalued in the eyes of many professionals, to think of BIM as just
another ‘software’ which is in reality, a ‘process’ of attaining an outstanding
collaboration among each and every stakeholder throughout a project’s life
cycle. Hence, the philosophy of Supply Chain Management (SCM) in a con-
struction project can be achieved effectively with the BIM’s promise of flexible
and transparent interaction among Construction Supply Chain (CSC) contribu-
tors. However, the complexities in CSC of a conservative setup has been more
composite with the intervention of BIM. Consequently, the BIM project
stakeholders have encountered complications on the effective application of
BIM on CSC while reaching the envisioned goals of BIM. Thus, the intention of
this study is to develop a foreseeable information flow model related to con-
struction supply chain in a BIM aided project. In order to attain the aim, an
extensive literature synthesis was piloted to develop a conceptual informational
flow model among CSC stakeholders of a BIM project. This contemporary
research outlines that the CSC of a BIM project is much interactive and flexible
with its collaborative effort of stakeholders when compared to the clashes among
professionals in traditional setups. In a practical context, the roles and duties of
BIM project stakeholders identified in BIM standards, have slightly differed due
to real-world complications in construction industry. Besides that, BIM stan-
dards have laid down charismatic arrangements on information flows among
BIM project stakeholders, which is in reality, a complex and random
setup. Hence, the research outcome has successfully answered the complications
by laying down guidelines for any outsider to the BIM field to recognise what
would be the contribution of each stakeholder throughout the BIM project’s life
cycle.

Keywords: Building Information Modelling (BIM)
PAS1192:2

Construction Supply Chain (CSC)
BIM project stakeholders

Supply chain information flows

© Springer Nature Singapore Pte Ltd. 2020

R. Dissanayake and P. Mendis (Eds.): ICSBE 2018, LNCE 44, pp. 604–614, 2020.
https://doi.org/10.1007/978-981-13-9749-3_53
 Integration oF BIM and CSC Through Supply Chain Management 605

1 Introduction

The success of a construction project is usually due to its effective and efficient
management of various knowledge areas. The areas identified are the major manage-
ment knowledge areas in Organizational Management. Within the modern trends in
global construction, Building Information Modelling (BIM) helps to substitute the
customary procedures and implement innovation to orthodox construction. BIM also
acts as an ‘Integrator’ for all those aspects of management. Moreover, such quality of
BIM would be more effective only with a clearly defined information flow among the
project stakeholders. The development of a BIM model is based on coherent infor-
mation feedings by said stakeholders. However, due to the vagueness in Construction
Supply Chain (CSC), the defined information sharing, feeding or collection is difficult.
This may result in reducing the productivity of BIM application too.
The paper comprehends the preliminary findings of the literature synthesis on how
the information management among the BIM CSC stakeholders has been affected with
the intervention of BIM standards as with special particulars to BIM Level 2 introduced
by the United Kingdom (UK). Accordingly, the outcomes of the literature synthesis is
illustrated in a conceptual model so as to express the information flows of BIM
environment in a much structured manner.

2 Influence of BIM Towards the CSC

Construction projects establish the critical portion of all disciplines in projects focusing
on their budget, multiplicity and cost (Rokooei 2015). Moreover, Rokooei (2015) has
highlighted the necessity of managing it, regardless of the financial scale of a con-
struction project. Rokooei (2015) further extended the success criterion of construction
project management into information management, where the successfulness could
have been assured through aiding BIM.
Accordingly, modern Architecture Engineering and Construction (AEC) trade is
multifaceted due to critical issues such as split nature in resources production, com-
plexity in supply processes, constraints to construction and the outsized number of
project stakeholders as of latest procurement pathways (Zavadskas et al. 2010). In view
of that, the information supply chain concept is being used in investigating the con-
tribution and dependence of stakeholders within project environment.
The concept of Supply Chain (SC) has been defined as “the network of organi-
zations that are involved, through upstream and downstream linkages, in the different
processes and activities that produce value in the form of products and services in the
hands of the ultimate customer” (Chirstopher 2011, p. 3). Consequently, Supply Chain
Management (SCM) envisions the entire supply chain as a process of production
without being restricted to a single level, and expects to increase the transparency and
logical alignment among the SC actors upon their interconnections which will not be
limited on practical or commercial aspects (Cooper and Ellram 1993). Besides, as
Mentzer et al. (2001) further clarifies, SCM is a successful supervisory mechanism to
manage the various flows of information and products within the project boundaries.
606 A. P. Rathnasinghe et al.

The implementation of BIM in the construction industry is appreciated by Robson

et al. (2014) as an expedition through the intelligence of building information up to the
‘maturity’ from the conventional 2D drawings at level 0, to the exclusively amalga-
mated 3D modelling in criterions of design, time and cost. However, Goh et al. (2014)
underlined that BIM was conceived as a technological solution for problems related to
the informational SC processes.
However, the traditional SCM intents at achieving naive and obvious interactions
among the various stakeholders in SC, common profits and controlling the afore-
mentioned project uncertainties (London 2009). With the mandating of BIM in con-
struction projects by the most government authorities, the traditional SC of the
construction projects was drastically changed due to the intervention of the digital
world to orthodox construction and planning techniques. Accordingly, the British
government has made a significant contribution towards introducing uniformed and
consistent specification series, which is commonly practised in the global context.
Therefore, it is worth investigating how the information management among the
stakeholders is performed with this new technological interference of BIM as intro-
duced by the British Standards on BIM implementation. This paper presents a literature
synthesis aimed to identify a suitable information flow model for information man-
agement of a CSC of a construction project.

3 Research Method

This research seeks to address the research problem of “how the information man-
agement among BIM project stakeholders is to be performed with the technological
intervention of BIM in CSC” through a qualitative approach, as qualitative methods
contribute in conducting in-depth investigations on emerging concepts and is more
appropriate, when the research has a trivial base of literature. Accordingly, this research
is mainly focussed on a thorough investigation of literature sources such as; BIM
standards published by The British Standards Institution and other BIM related books,
periodicals, and conference proceedings. Subsequently, the outcomes have been
illustrated in a conceptual model which is totally based on BIM related literature.

4 Standard Practices of BIM and Their Effect on Traditional

Construction Phenomena

As identified by Wickersham (2009), existing BIM-related standards provide a suc-

cessful initiative to establish a collaborative BIM method. Further, McGraw Hill (2014)
surveys suggested that the high implementation of BIM should be effectively managed
and controlled by introducing relevant standards depending on the project context.
However, these BIM standards are mainly paper-based standards (e.g. BS1192:
2007 and PAS 1192-1, 2 and 3), and IT related data exchange standards (e.g. COBie
and IFC) (Smith 2014). BIM implementation in the UK has improved gradually until
the introduction of BIM policy by the UK government’s BIM Task Group in May 2011
(HM Government 2012). This transitional position has been called by the scholars as
 Integration oF BIM and CSC Through Supply Chain Management 607

the BIM ‘light bulb moment’ for UK constructions (NBS 2012). Henceforth, the UK is
presently considered to be the most leading and influencing nation on global BIM
resourcefulness (Saxon 2013).

4.1 Publicly Available Specification (PAS) 1192-2:2013

This UK based specification is formed to manage the information with BIM which may
arise during the supply stages of resources. Accordingly, it provides a guideline to
implement a BIM project and standard requisites on the Employer’s Information
Requirements (EIR) and the BIM Execution Plan (BEP) (Rock 2018). Further, PAS
1192:2 describes the CDE based on its collaborative features to perform as a uniform
stage for information distribution in between the task teams of different professional
codes (Applecore Designs 2016).

4.2 Significant Changes Due to BIM Standards

Arayici et al. (2012) argued on BIM standards as it has made significant changes in
predefined design instances regarding the documentation and management procedures
of those design scenarios. Meanwhile, it is considered that those BIM values denote a
mutual concern on achieving utmost goals of the time, cost and quality triangle and
employer’s gratification (Takim et al. 2013). Lee (2008) highlights a different outlook
on BIM, on reinventing the standard construction procedures, through a simulated
environment which enables greater productivity, corporation and communication.
Therefore, Takim et al. (2013) further appraised the implementation of BIM in the AEC
industry as a turnover point, which has improved the quality of life among stakeholders.
Even though many authors suggest achieving a greater efficiency through BIM
implementation within the AEC industry, some have identified the limitations and
issues in this scenario. Arayici et al. (2012) found out that one such limitation of BIM is
in the information congregation process where the final design proposal is solely based
on the deliberation and user requirements which are poorly collected. As a remedy,
Howard and Bjork (2008) highlighted on the importance of developing BIM standards
which are compatible with the industry’s stakeholder requirements.

5 BIM Common Data Environment (CDE) and Its

Information Flows

The main pathway to achieving a well-organized data collection in a BIM-aided project

is to have a CDE; a virtual platform managed by project teams for assembling, handling
and distributing digital information among themselves upon requirement. The layout of
a CDE is defined in the PAS 1192-2 specification which is the regulatory manual for
public projects done in BIM Level 2. This guidance is based on the core principles of
BS 1192: 2007.
As identified in PAS 1192:2, there are four main phases of a CDE. The first phase
of CDE is ‘Work in progress (WIP)’ where data is currently in fabrication and has not
608 A. P. Rathnasinghe et al.

yet been checked and verified for use outside of the authoring team (Architectural,
Engineering and Construction Council (AEC) 2018).
Once the WIP data has been checked, and approved by the particular stakeholder
who does the uploading those would be moved into the ‘Shared’ area. In this phase,
other stakeholders may have in access and can use that information in the creation and
development of their contributions. Under BIM Level 2, while the other stakeholders
re-use the preliminary information, the ability to alter the information remains with its
creator.
Once an individual WIP has gone over the ‘Shared’ process, all other stakeholders
of the BIM team may have already made their comments where the work has been
matched and detected for any clashes with other personnel. Such pieces of work will be
suitable for publishing. ‘Published’ means the specific piece of information is the
outcome to a particular client’s information requirement. The outcomes of the Pub-
lished or Issued area would deliver as 2D design format or Portable Document Format
(PDF), drawings or data sheets, which would also be kept in the Published Area of the
folder structure.

6 BIM Project Stakeholders

It is the common viewpoint that, BIM has established a rigid commitment based
environment among the project members. Sebastian (2011), explained that the ideal
status in BIM utilization can only be achieved through an effective collaborative
environment consisting of multi-disciplinary professionals. Furthermore, Sebastian
(2011) argued that this collaborative environment requires clear administrative defi-
nitions on key roles rather than modifying them. Moreover, such environments will be
successful through the rescheduled contractual relationships and re-engineered inte-
gration process.

6.1 Characteristics of BIM Project Stakeholders

Latest surveys revealed the necessity of engaging BIM skilled professionals for a
successful BIM implementation (Allen Consulting Group 2010). As mentioned by
Olatunji (2011), experts with required professional skills can be gained with certain
modifications and training to the conventional project professionals. Accordingly, the
training of stakeholders should be compatible to comply with the specialized needs of
BIM software.
Arayici et al. (2012), highlighted that BIM stakeholders should possess modern
management skills with special concentration to external party support. Henceforward,
research studies aim at an effective redevelopment of consultants within the BIM team
who support the completion milestones in the model.
As Ashcraft (2008) suggested, when producing a BIM model, undivided com-
mitment and care from the BIM project stakeholders is most important. The main aim
of such a collaborative environment should be to develop a comprehensive BIM model
with less disputes among the project members. This point of view has been discussed
again under the BIM overlay to the Royal Institute of British Architects (RIBA 2012)
 Integration oF BIM and CSC Through Supply Chain Management 609

as expected job profiles and services to be performed by respective stakeholders within

a BIM project.

6.2 Roles and Duties of BIM Project Stakeholders

The arrangement of the BIM team and its members’ copyrights will be settled by the
end of project. However, on other hand, the expected usage of technologies and
approaches may be considered at the time of creating the model. Consequently, the
administration of BIM projects of such nature will be done by a panel of BIM experts
consisting of BIM architects, BIM managers, model developers, and draftsmen (Ola-
tunji 2011). Even though specific professional titles in BIM environment are still
untried in the conventional setup, traditional professionals with BIM knowledge are
still valuable as substitutes to the BIM job profiles.
As identified by Aranda-Mena et al. (2009), ‘project manager’ is one of the critical
BIM project team members who is distinct with the BIM and technological knowledge.
Moreover, studies of Kassem et al. (2014), highlighted the role of ‘BIM coordinator’ as
the main professional who acts on behalf of the client with professional liability and the
job profile has been recognized by the BIM protocol as a central role in CDE.
CIC BIM protocol (2013) suggested that the lead designer and BIM coordinator
should work together so as to embrace the contractual obligations related to the BIM
model work distribution. Moreover, the protocol recommended a ‘BIM manager’ on
regulating the administrative and procedural rules related to BIM model management
so as to preserve the collaborative nature and the duties attached to the BIM concept.
BIM overlay to RIBA (2012), introduced a number of professional profiles associated
with the BIM project team as in Table 1 which are significantly different when com-
pared to the traditional project team stakeholders.

Table 1. Identified duties of BIM project stakeholders

Job title Key duties within BIM environment
Main Lead architect is the person who ensures inter-relationships among
designer/Architect various designs of task teams used in BIM environment. Moreover, lead
designer is the main regulator on the accuracy of the information fed by
the task team members to the shared model
Client Briefly, a personality who represents the client at each phase of BIM
representative project lifespan. Accordingly, the role is granted with limited decision
making authority on project milestones, and power to assign minor
project staff
Technical Technical advisors facilitate the construction professional staff and the
advisors client in understanding the technological aspects behind the BIM
concept. Furthermore, the title holder is responsible with the upgrading
of intelligent brief on the project plan
Delivery manager Delivery managers are accommodated with the monitoring and
supervising on the supply chain of the project whether the project is
going along the intended progress of work. However, in practical
context, the client’s representative will replace with the role of delivery
manager within the project team
Source: Adapted from (BIM overlay to RIBA 2012)
610 A. P. Rathnasinghe et al.

7 Information Flows in Construction Supply Chain

Partnerships

A successful case study on communication amongst project and SC stakeholders with

the inspiration of BIM was conducted by Papadonikolaki et al. (2017). According to
their outcomes, the communication among the SC stakeholders is directed through the
pre-defined channels as identified in their particular contract. The architect, the engi-
neers and other identified main stakeholders would be in continuous communication
with the contractor, where they are obliged to confer on decision making through their
respective employment contracts.
Meanwhile, the supplier, who is obliged with a ‘chain contract’ (SC framework
agreement) with the contractor, may control a part of direct channels with subsidiary
suppliers or engineers in behalf of the contractor for the ease of solving problems
within an integrated environment. However, such communication is exclusively
practiced among the engineers or suppliers correspondingly, where the consultants
(architect, structural and mechanical) may not be involved in. In contrast, the supplier
and consultant often communicate mostly on fiscal indecisions and build trust.
Further Papadonikolaki et al. (2017) recognised that the contractor’s role towards
successful CSCM information establishment would be impossible if it has been
addressed from the monetary view instead of building trust. Moreover, lead designers
hold a vital role in such information flows as s/he holds the main outcome on clarifying
agreements on design and the material perspectives with the contractor.

8 Integration of BIM and Construction Supply Chain

As identified by Zavadskas et al. (2010), the AEC industry is having a challenging SC

because of the split supply of resources, building challenges and the increasing of
project participants. Further, they suggested the deployment of BIM may neutralize
such executive and bureaucratic impediments in the AEC SC.
Rezgui and Miles (2010) further commented on the use of inter-organisational IT,
such as BIM to perform an effective CSCM. Therefore, they suggested that the inte-
gration among BIM may cause an effective management of information flows of the
SC, because BIM is a designed data model of building information (Eastman et al.
2011).
However, while the integration among those two concepts are vital to the effec-
tiveness of the AEC industry, the organizational challenges due to this integration are
not yet thoroughly reviewed by the scholars. In contemporary studies, BIM imple-
mentation and SCM integration are discussed only by means of the theoretical aspects
while not significantly involved in the research process (Papadonikolaki et al. 2015).
Accordingly, Flows of material, information, and the network of stakeholders are the
main components to be considered in SCM.
 Integration oF BIM and CSC Through Supply Chain Management 611

8.1 BIM Integration with Information Flows and CSC Stakeholders

The definition of BIM by the National BIM Standard Project Committee itself states the
BIM’s relationship in assisting and controlling the information flow of a building
(BuildingSMART 2016). Accordingly, BIM provides with solutions for 3D modelling,
interoperability and transparency in project decision making throughout the SC
stakeholders. Furthermore, it provides an accessible podium for each project member to
improve the harmonization among the stakeholders’ decision making and feedback
mechanisms while improving the traditional data management (Demian and Walters
2013). However, while the impacts of the BIM models are still doubtful, Van Berlo
(2012) recommends workarounds that provide reasonable BIM interoperability.
Therefore, the above certifies BIM’s effort on managing the information flow, a core
feature of SCM.

9 Development of Flow Model for Construction Supply

Chain’s Information Flows

In developing the conceptual model as in Fig. 1, the concern was on the information
management principles offered in PAS 1192:2 specifications, where it clearly presented
the relationships between the documents and stakeholders in information management
of a BIM project.
According to the readings, during the tender stage (Pre-contract stage), a forth-
coming contractor to the project may develop a BIM Execution Plan (BEP) while
validating their competence, and aptitude with evidence to fulfil Employer’s Infor-
mation Requirements (EIR). Accordingly, Pre-contract BEP makes the pathway to
Project Implementation Plan (PIP), which consists of project goals for collaboration,
information modelling and key project milestones and where they fit with the broader
project programme.
However, when the most suitable contractor has been selected, s/he is required to
develop a more comprehensive BEP while confirming SC’s proficiencies. This is
indicated in the conceptual model as the ‘Post-contract BIM Execution Plan’. Fur-
thermore, this Post-contact BEP arouses the need of ‘Master Information Delivery Plan
(MIDP)’ which signifies when and how the information is being produced. Meanwhile,
it represents the responsible party’s involvement in producing a piece of information
and their probable use of standard protocols and procedures.
In general terms, BEP is the set of plans prepared by the respective suppliers of SC
to signify on how the information modelling features of the BIM project are being
sustained. Furthermore, the BEP presents the solutions for the problems raised in EIR;
the document consisting of standards, procedures and information transmissions to be
embraced by the respective contractors (suppliers) as a part of continuing the project.
Furthermore, the final outcome as a MIDP is established on a compacted chain of
individual Task Information Delivery Plans (TIDP), consisting of each team’s tasks
with the layout, date and duties.
Accordingly, the combination of MIDP, BEP and the PIP, will form the BIM model
based on the design and specification data provided by the lead architect and other task
612 A. P. Rathnasinghe et al.

teams to the CDE. Henceforward, the traditional SC relationships among the contractor,
suppliers and BIM task teams commence as illustrated in the bottom part of Fig. 1.

Fig. 1. Conceptual information flow model

10 Conclusion

The comprehensive literature synthesis revealed the complexity of the CSC due to
numerous dynamics and the intervention of BIM as an IT solution to the construction
industry which is a late embracer on such contexts.
Furthermore, the paper discusses the standard roles and duties of BIM project
stakeholders and the composition of BIM project teams as identified by the regulatory
institutions. Accordingly, BIM related literature exposed BIM specific job profiles such
as; lead architect, client’s representative, delivery manager and technical advisors.
Subsequently, as a BIM collaborative environment is technically called as the CDE,
the literature proved that there are different flows among each and every stakeholder
which makes it more complex to examine. Moreover, the standards have suggested the
technical stages of CDE and upon each stage standards have been presented with the
flow of information from the working stage to the published stage. Moreover, the
 Integration oF BIM and CSC Through Supply Chain Management 613

works of scholars proved that the effective management of material and information
flow can be ascertained through the integration of BIM with SC information flows.
Therefore, upon the outcomes of BIM project stakeholders of CSC and the infor-
mation flows among them, a structure of the anticipated information flows for CSC of a
BIM project was proposed in Fig. 1. Consequently, it was further developed with the
assistance of researches in field of SC in BIM projects. As the model was totally based
upon the literature outcomes, it was acknowledged as a conceptual model.
In view of that, the model illustrated with main BIM project stakeholders and their
anticipated information flows accordance with the informational outputs.

References
Allen Consulting Group (2010) Productivity in the buildings network: assessing the impacts of
building information models. Built Environment Innovation and Industry Council. Sydney:
Allen Consulting Group
Applecore Designs (2016) The Eight Pillars of BIM Level 2 - Part One - Applecore Designs.
https://www.applecoredesigns.co.uk/coreblog/pillars-bim-level2-part-one. Accessed 9 Nov
2018
Architectural, Engineering and Construction Council (UK) (2018) AEC (UK) BIM Standard for
Autodesk Revit. https://aecuk.files.word.press.com/2010/06/aecukbimstandardforrevit-v1-0.
pdf. Accessed 9 Nov 2018
Aranda-Mena G, Crawford J, Chevez A, Froese T (2009) Building information modelling
demystified: does it make business sense to adopt BIM? Int J Manag Proj Bus 2(3):419–434
Arayici Y, Egbu C, Coates P (2012) Building Information Modelling (BIM) implementation and
remote construction projects: issues, challenges and critiques. J Inf Technol Constr (ITcon)
17, pp 72–59. http://www.itcon.org/data/works/att/2012_5.content.03794.pdf. Accessed 9
Nov 2018
Ashcraft H (2008) Building information modelling: a framework for collaboration. Construction
Lawyer 28(3). http://www.hansonbridgett.com/About-Hanson-Bridgett/press-and-news/*/
media/Files/News/CL_HowardAshcraft_BIM.pdf. Accessed 9 Nov 2018
BuildingSMART (2016) Frequently asked questions about the national BIM standard-United
States™. https://www.nationalbimstandard.org/faqs. Accessed 9 Nov 2018
Christopher M (2011) Logistics and Supply Chain Management. Prentice Hall, London
CIC BIM Protocol (2013) CIC BIM Protocol. Construction Industry Council, London
Cooper M, Ellram L (1993) Characteristics of supply chain management and the implications for
purchasing and logistics strategy. Int J Logist Manage 4(2):13–24
Demian P, Walters D (2013) The advantages of information management through building
information modelling. Constr Manage Econ 32(12):1153–1165
Eastman C, Teicholz P, Sacks R, Kathleen L (2008) BIM Handbook: A guide to Building
Information Modeling for owners, managers, designers, engineers and contractors. Constr
Econ Build 12(3):101–102
Goh K, Goh H, Toh S, Peniel Ang S (2014) Enhancing communication in construction industry
through BIM. In: Proceedings of the 11th International Conference of Innovation and
Management, pp 23–35. http://eprints.uthm.edu.my/6696/1/Enhancing_Communication_in_
Construction_Industry_through_BIM.pdf. Accessed 9 Nov 2018
HM Government (2012) Building Information Modelling, Industrial Strategy – Government and
Industry in partnership. London: Government Report
Howard R, Bjork B (2008) Building information modelling – Experts’ views on standardisation
and industry deployment. Adv Eng Inf 22(2):271–280
614 A. P. Rathnasinghe et al.

Kassem M, Iqbal N, Kelly G, Lockley S, Dawood N (2014) Building information modelling:

protocols for collaborative design processes. J Inf Technol Constr (ITcon) 19, pp 126–149.
http://www.itcon.org/data/works/att/2014_7.content.00672.pdf. Accessed 9 Nov 2018
Lee C (2008) BIM: changing the AEC industry. In: Proceedings of PMI Global Congress 2008.
Project Management Institute, Denver, Colorado, USA
London K (2008) Industrial organization object-oriented project model of the facade supply
chain cluster. Construction Supply Chain Management Handbook, pp 13-1–13-46
McGraw-Hill Construction (2010) The business value of BIM in Europe. Smart Market Report.
Bedford: McGraw-Hill Construction. Available at: http://images.autodesk.com/adsk/files/
business_value_of_bim_in_europe_smr_final.pdf. Accessed 9 Nov 2018
Mentzer J, De Witt W, Keebler J, Min S, NIx N, Smith C, Zacharia Z (2001) Defining supply
chain management. J Bus Logist 22(2). https://www.biblioteca.fundacionicbc.edu.ar/images/
e/e4/Conexion_y_logistica_2.pdf. Accessed 9 Nov 2018
NBS (2012) BIM Legal Roundtable Discussion. NBS Website. http://www.thenbs.com/
roundtable/legal/bimLegalRoundtable_2012.asp. Accessed 9 Nov 2018
Olatunji O (2011) Modelling the costs of corporate implementation of building information
modelling. J Financ Manage Prop Constr 16(3):211–231
Papadonikolaki E, Verbraeck A, Wamelink H (2017) Formal and informal relations within BIM-
enabled supply chain partnerships. Constr Manage Econ 35(8-9):531–552. https://doi.org/10.
1080/01446193.2017.1311020
Papadonikolaki E, Vrijhoef R, Wamelink H (2015) Supply chain integration with BIM: a graph-
based model. Struct Surv 33(3):257–277
Rezgui Y, Miles J (2010) Exploring the potential of SME alliances in the construction sector.
J Constr Eng Manage 136(5):558–567
Robson A, Boyd D, Thurairajah N (2014) UK Construction Supply Chain Attitudes to BIM. In:
50th ASC Annual International Conference Proceedings. Birmingham: Birmingham City
University. http://ascpro.ascweb.org/chair/paper/CPRT224002014.pdf. Accessed 9 Nov 2018
Rock S (2018) Construction: The eight pillars of Level 2 BIM. Herbert Smith Freehills - Real
estate development notes. https://hsfnotes.com/realestatedevelopment/2016/05/04/the-eight-
pillars-of-level-2-bim/. Accessed 9 Nov 2018
Rokooei S (2015) Building information modeling in project management: necessities, challenges
and outcomes. Proc Soc Behav Sci 210:87–95
Royal Institute of British Architects [RIBA] (2012) BIM Overlay to the RIBA Outline. London:
Author
Saxon R (2013) Growth through BIM. Construction Industry Council, London
Sebastian R (2011) Changing roles of the clients, architects and contractors through BIM. Eng
Constr Archit Manage 18(2):176–187
Smith P (2014) BIM implementation – global strategies. Proc Eng 85:482–492
Takim R, Harris M, Nawawi A (2013) Building Information Modelling (BIM): a new paradigm
for quality of life within architectural, engineering and construction (AEC) industry. In:
Proceedings of AMER International Conference on Quality of Life. Langkawi, Malaysia:
Universiti Teknologi MARA (UiTM), pp 23–32
Van Berlo L (2012) Collaborative engineering with IFC: new insights and technology. In:
Proceedings of 9th European Conference on Product and Process Modelling. Iceland:
Blackwell Publishers
Wickersham J (2009) Legal and Business Implications of Building Information Modelling
(BIM) and Integrated Project Delivery (IPD). Rocket Publishing Ltd, USA
Zavadskas E., Turskis Z, Tamošaitienė J (2010) J Civil Eng Manage 16(1):33–46. https://
journals.vgtu.lt/index.php/JCEM/article/view/5898
 Removal of Heavy Metals from Industrial
Wastewater Through Minerals

U. A. S. L. Muthukalum(&), C. A. Gunathilake, and C. S. Kalpage

University of Peradeniya, Peradeniya, Sri Lanka

[email protected]

Abstract. Heavy metals are toxic to the health and environment and causes
harmful consequences to human beings. It is essential to take effective measures
to remove metals from contaminated water. In this research, naturally available
brick materials with different compositions collected from Kandy (type A),
Gampaha (type B), and Embilipitiya (type C) areas were used to remove Cu, Cr
and Pb ions. These brick materials are cost effective and are a readily available
alternative to conventional heavy metal removal. The characterization of brick
material is performed using X-Ray Fluorescence (XRF) and Nitrogen Adsorp-
tion – Desorption analyser. Calculated surface area and total pore volume
and pore width of type A, B, and C are lie respectively, in the range of
128–154 m2/g, 0.24–0.45 cm3/g, and 5.6–16.7 nm. The main objective of this
research is to investigate how the characteristics of adsorbents influences the
adsorption process and to identify the best model to describe the kinetic and
equilibrium adsorption to purify metal contaminated water. Results indicate that
Pseudo – first – order kinetics model properly described the adsorption of Cu2+
to the brick type C, which has maximum adsorption capacity of 497 mg g−1.
The adsorption process of Pb2+ to the brick type A and Cu2+ and Pb2+ to brick
type B and C were well-fitted with Pseudo – second – order kinetics model. In
equilibrium studies, the Langmuir isotherm showed a better fitness in adsorption
of Cu2+ into brick type A and C, Pb2+ into brick type B, whereas the Freundlich
isotherm well represented the adsorption characteristics of Cu2+ into brick type
A, Pb2+ into brick type B and Cr6+ into all brick types. The comparison of
results indicated that the use of brick types A, B, C can be used as a potential
nan sorbent to remove heavy metals from industrial waste water.

Keywords: Heavy metals
Minerals
Adsorption
Specific surface area

Ion concentration

1 Introduction

In the current world, heavy metals are a major source of pollution in the aquatic and
soil environments. Due to certain Industrial activities, a lot of companies generate
waste water containing different metals; mainly Zn, Cr, Cu, Ni and Cd. According to
previously conducted studies, it has been revealed that high levels of particular heavy
metals in streams and soil can be identified due to the activities such as mining
societies’ effluents, agricultural activities, discharge from Industrial parks and sewage
work. The conventional way of removing heavy metal is usually performed by the

© Springer Nature Singapore Pte Ltd. 2020

R. Dissanayake and P. Mendis (Eds.): ICSBE 2018, LNCE 44, pp. 615–632, 2020.
https://doi.org/10.1007/978-981-13-9749-3_54
616 U. A. S. L. Muthukalum et al.

addition of a base to precipitate hydroxides. But the expected dissolved concentrations

of heavy metals cannot be achieved due to the complexation with inorganic and organic
ligands of the water (Danielsson 2013).
Considering the adsorption capacity of minerals containing oxide, they have the
capability of removing many minerals over a wider pH range and lowering dissolved
levels of heavy metals. As an effective method of removing toxic heavy metals,
adsorption is an important physicochemical process that occurs at the solid liquid
interface (Garcia Sanchez 1999). Natural minerals have been investigated as adsorbents
in many researches for the removal of various toxic and hazardous pollutants. The most
important parameters that control heavy metal adsorption and their distribution have
been identified as mineral type, metal speciation, metal concentration, soil pH, solid:
solution mass ratio and contact time (Pare 2012).
In this research, different types of bricks which are made up of soil compositions
from 3 locations in the country are tested for the ability to remove heavy metals
through adsorption. The adsorption of selected heavy metals such as Pb, Cr, Cu etc. are
investigated for kinetic adsorption and equilibrium adsorption. According to the vari-
ation of the factors in the reaction medium and the characteristics of the adsorbent, the
adsorption process is observed and the best model which fits the observed data is used
to obtain the adsorption capacity for each heavy metal.

2 Materials and Methods

2.1 Materials
2.1.1 Adsorbent
Three different kinds of bricks from different locations in the country (Kandy, Gam-
paha, Embilipitiya) were selected as shown in Table 1 and they were ground into
powder form.

Table 1. Brick materials obtained from different places

Brick type Area
A Kandy
B Gampaha
C Embilipitiya

2.1.2 Adsorbate
Metal ions, subject to study are Pb2+, Cr6+ and Cu2+. Salt solutions of Pb(NO2)2,
CuSO4 and K2CrO4 were used as sources of heavy metal ions. Dilute HNO3 and
NaHCO3 were used for pH adjustment.

2.2 Methods
To determine the ion concentration in each sample, an atomic adsorption spec-
trophotometer is used. In AAS, free atoms are generated by the sample (to be analysed)
 Removal of Heavy Metals from Industrial Wastewater Through Minerals 617

in the atomizer through a flame, and a light beam is directed to the flame. The amount
of light absorbed by the atomized elements is measured eventually and the unknown
concentration is determined through a calibration curve which is obtained using known
concentrations of solutions.
Elementary material composition is investigated using X-Ray Fluorescence
(XRF) method.
The specific surface area is determined using Nitrogen Adsorption Desorption
Analyzer. The specific surface area and the pore size distribution of a sample is
determined by the physical adsorption of a gas on the surface of the solid and by
calculating the amount of adsorbate gas corresponding to a monomolecular layer on the
surface. Initially, the mineral sample is placed in the degassing unit to be cleaned and
removed of impurities. The sample with the actual weight is then moved to the surface
area analyser. Liquid nitrogen is used to make the inert gas adsorbed onto the mineral
surface. The data processing unit observes the sample and the isotherm is obtained with
the parameters under BET summary (Kruk et al. 1997).

3 Adsorption Studies and Isotherms

3.1 Kinetic Studies

3.1.1 Kinetic Adsorption Experiments
Kinetic adsorption studies were carried out using the initial metal ion concentration for
all 3 solutions (Pb2+, Cu2+, Cr6+) of 100 ppm. 0.05 g of brick material samples (ad-
sorbent) were prepared. Initially, for one brick material sample selected and one ion
solution, 50 ml of ion solution was taken into a beaker and the adsorbent mass was
placed while stirring (150 rpm, 25 °C, pH = 5). 0.2 ml of solution was withdrawn
before mixing the adsorbent and ion solution (t = 0 min) at pre-determined intervals
(t = 10 to 120 min) (Gunathilake et al. 2015). The amount of adsorbed ions per unit
mass of the adsorbent at a time t (qt) was calculated by the following equation.

ðC0  Ct ÞV
qt ¼ ð1Þ
m
Where C0 (ppm) is the initial concentration of metal ions, V (mL) is the volume of
metal ion solution and m (g) is the adsorbent mass. Ct (ppm) is the metal ion con-
centration at a time t. Metal ion concentration was analysed by atomic adsorption
spectroscopy.

3.1.2 Analysis of Adsorption Kinetics Data

Two models were used to analyse the adsorption kinetics data which is mentioned
below. The first order adsorption kinetics model can be expressed by the following
equation.

dQ
¼ K1 ðqe  qt Þ ð2Þ
dt
618 U. A. S. L. Muthukalum et al.

Where qe and qt are the amount of adsorbed metal ions at the equilibrium and time t,
respectively, K1 is the rate constant of the first order adsorption reaction. Integration of
above equation gives the following linear model.

K1 t
logðqe  qt Þ ¼ þ log qe ð3Þ
2:303

The pseudo second order kinetic equation is as following.

dQ
¼ K 2 ð qe  qt Þ 2 ð4Þ
dt
where K2 is the pseudo second order rate constant of adsorption (g mg−1 min−1). By
integrating this equation, following linear model can be obtained.

t 1 t
¼ 2
þ ð5Þ
qt K2 qe qe

3.2 Equilibrium Studies

All the adsorption studies are carried out in 200 ml flasks by subjecting a given 50 mg
dose of adsorbent to a period of stirring with 50 ml of solution on a magnetic stir.
To obtain the adsorption isotherm, solutions with metal concentrations ranging
between 50–300 ppm are used and stirred for 6 h as it was optimized in kinetic studies,
under 25 °C, 5 pH and 150 rpm.
The adsorption isotherms under Langmuir model and Freundlich model are
developed and the most suitable model for each adsorbent is evaluated by comparing
the coefficient of determination and the model constants are also calculated
Langmuir equation,

KQm Ce
Qe ¼ ð6Þ
1 þ KCe

In linear form,

1 1 1 1
¼ þ ð7Þ
Qe Qm KQm Ce

Where,

V
qe ¼ ðC0  Ce Þ  ð8Þ
m
 Removal of Heavy Metals from Industrial Wastewater Through Minerals 619

Freundlich equation,
1
Qe ¼ KCen ð9Þ

The linear model is expressed as,

log Qe ¼ log K þ ð1=nÞ  log Ce ð10Þ

In these models, the used notations are, Qe - Equilibrium concentration of adsorbate

in the solid phase in mg/g, Ce - Equilibrium concentration of adsorbate in the liquid
phase in mg/l, Co - Initial concentration of adsorbate in the liquid phase in mg/l, m -
Initial mass of the adsorbent in g, V - Volume of the liquid phase in ml, K - Langmuir
Constant, Qm - Maximum adsorption capacity and 1/n - Adsorption Intensity.

4 Result and Discussion

4.1 Characterization of Raw Brick Samples
The composition of each brick material was investigated using X-Ray Fluorescence
(XRF) and the results are presented in Figs. 1, 2 and 3. It can be seen that the brick
samples A, B and C contains approximately equal amounts of Fe, but the content of Ti,
Pb and Cu in brick sample A is much less than that of brick sample B. The XRF pattern
for brick sample A reveals that it contains 89.6% Fe, 2.3% Ba, 1.9% Mn and 1.4% Ca
whereas brick sample B contains 87.4% Fe, 4.8% Cu and 4.1% Ti.

Fig. 1. XRF pattern for brick sample A

620 U. A. S. L. Muthukalum et al.

Fig. 2. XRF pattern for brick sample B

Fig. 3. XRF pattern for brick sample C

4.2 Calculation of Surface Properties

The Brunauer–Emmett–Teller (BET) specific surface area (SBET) of each sample was
calculated by following its N2 adsorption isotherm in the relative pressure range of 0.05
to 0.20. The single-point pore volume (Vsp) was determined based on the amount N2
gas adsorbed at a relative pressure of *0.98. The pore size distributions (PSD) were
calculated using adsorption branches of N2 adsorption/desorption isotherms and
applying the improved Kruk–Jaroniec–Sayari (KJS) method developed for cylindrical
pores. The total pore volume (Vt) was calculated by integrating the entire PSD curve
and the volume of micropores (Vmi) was calculated by integrating the PSD curve up to
2 nm. The mesopore volume (Vmeso) was determined by integrating the PSD curve
 Removal of Heavy Metals from Industrial Wastewater Through Minerals 621

from 2 nm to the end of the PSD curve. In addition to the PSD curves that display the
overall distribution of pores, the pore widths (Wmax) corresponding to the maxima of
these distributions are also tabulated to show the characteristic pore sizes of the samples
studied.
Figures 4 shows comparisons of N2 adsorption/desorption isotherms for the type A
brick, type B brick and type C brick. All materials exhibit type IV nitrogen adsorption
isotherms and H1 type hysteresis loop with capillary condensation–evaporation steps
beginning at a relative pressure of around 0.70. Insets in Fig. 4 shows the pore size
distribution (PSD) curves corresponding to the aforementioned adsorption isotherms
obtained by the KJS method.

1000

PSD (cm3 g-1 nm-1)

1
2
800 0.02 2 3
Volume adsorbed (cm3 STP/g )

600
0.00
0 10 20 30 40
Pore width (nm)
400
1 Type A Brick 3
2 Type B Brick
1
3 Type C Brick
200 2

0
0.0 0.2 0.4 0.6 0.8 1.0
Relative pressure

Fig. 4. N2 adsorption isotherms and their corresponding PSD curves (insets) for the type A brick
and type B brick.

Surface properties calculated using N2 adsorption isotherms reveals that the type C
brick sample has better surface properties, as evidenced by higher surface area, total
pore volume (Vt), and pore width (Wmax), when compared to the type B brick sample
(Table 2). The notations used are SBET—specific surface area, Vsp—single point pore
volume, Vt—total pore volume, Vmi—volume of micropores, Wmax—pore width.
622 U. A. S. L. Muthukalum et al.

Table 2. Adsorption properties obtained for the A, B and C brick samples studied.
Samples SBET (m2/g) Vsp (cc/g) Vt (cc/g) Vmi (cc/g) Wmax (nm)
Type A brick 146 0.39 0.42 0.03 12.9
Type B brick 128 0.22 0.24 0.02 5.6
Type C brick 154 0.42 0.45 0.03 16.7

4.2.1 Kinetic Studies of Metal Ion Adsorption

Kinetic studies of metal ion adsorption process revealed the rate of metal ion uptake by
the brick samples. Brick sample – A, brick sample – B and brick sample – C were
considered and each cation solution (Pb2+, Cu2+, Cr6+) was tested for the time
dependence of adsorption, as shown in Figs. 5, 6 and 7.

400

Variable
300 Cu
Pb
Qt (mg/g)

200

100

0 10 20 30 40 50 60 70 80 90
t (min)

Fig. 5. Experimental adsorption kinetics of metal ions on to brick type A

400

Variable
Cu
Cr
300 Pb
Qt (mg/g)

200

100

0 20 40 60 80 100 120
t (min)

Fig. 6. Experimental adsorption kinetics of metal ions on to brick type B

 Removal of Heavy Metals from Industrial Wastewater Through Minerals 623

400

Variable
Cu
Cr
300 Pb
Qt (mg/g)

200

100

0 20 40 60 80 100 120
t (min)

Fig. 7. Experimental adsorption kinetics of metal ions on to brick type C

The linear form of Pseudo – first – order and Pseudo – second – order kinetic
equations were applied to obtain the rate of ion adsorption. These forms of linear
equations were used to model the adsorption process and the validity of each equation
was checked using the correlation coefficient (R2) for each brick type in each ion
solution. Cu2+ ion uptake for brick type A fit well with Pseudo – first – order kinetics
model as shown in Fig. 8 with 95.8% data fitting the model. In the case of Pb2+ ion
uptake, brick type A was more consistent with Pseudo – second – order kinetic model
which represented 96.4% data in the plot as shown in the Fig. 9. But Cr6+ ion con-
centration was observed in the solution as gradually increasing with time passing, and
therefore it could not fit with any kinetic model. The uptake of Pb2+ into brick type B
was investigated and experimental data could be fitted with Pseudo – second – order
kinetics model which had a correlation coefficient of 92.4%. Cu2+ ion adsorption into
brick type B was properly fitted with Pseudo – second – order kinetics with 98.9% data
represented by the plot as shown in Fig. 10. In the case of adsorbing into brick type B,
Cr6+ shows a better fitness of 96.3% with Pseudo – second – order kinetics model with
352 mg/g adsorption capacity. Cu2+ was fit well in adsorbing into brick type C with
Pseudo – first – order kinetics model having 497 mg/g, whereas Cr6+ and Pb2+ showed
a best fit with Pseudo – second – order kinetic model as shown in Figs. 7, 8 and 9 with
adsorption capacities 191.8 mg/g 240.6 mg/g (Figs. 11, 12, 13, 14 and 15).
624 U. A. S. L. Muthukalum et al.

2.8

2.6

2.4
log(qe-qt)

2.2

2.0

1.8

0 10 20 30 40 50 60 70 80
t(min)

Fig. 8. Linear form of adsorption kinetics equation for Pseudo - first - order kinetic model for
Cu2+ adsorption on brick type A

12

10

8
t/qt (min g/mg)

0 20 40 60 80 100 120
t (min)

Fig. 9. Linear form of adsorption kinetics equation for Pseudo - second – order kinetic model
for Pb2+ adsorption on brick type A
 Removal of Heavy Metals from Industrial Wastewater Through Minerals 625

0.31

0.30

0.29

0.28
t/qt

0.27

0.26

0.25

0 20 40 60 80 100 120
t

Fig. 10. Linear form of adsorption kinetics equation for Pseudo - second - order kinetic model
for Cu2+ adsorption on brick type B

0.325

0.300

0.275

0.250
t/qt

0.225

0.200

0.175

0.150

0 20 40 60 80 100 120
t

Fig. 11. Linear form of adsorption kinetics equation for Pseudo - second - order kinetic model
for Cr6+ adsorption on brick type B
626 U. A. S. L. Muthukalum et al.

0.4

0.3
t/qt (min g/mg)

0.2

0.1

0.0

0 20 40 60 80 100 120
t (min)

Fig. 12. Linear form of adsorption kinetics equation for Pseudo - second - order kinetic model
for Cu2+ adsorption on brick type B

2.70

2.69

2.68

2.67
Log(Qe-Qt)

2.66

2.65

2.64

2.63

2.62

0 20 40 60 80 100 120
t (min)

Fig. 13. Linear form of adsorption kinetics equation for Pseudo - first - order kinetic model for
Cu2+ adsorption on brick type C
 Removal of Heavy Metals from Industrial Wastewater Through Minerals 627

1.5

1.4

1.3

1.2
t/Qt

1.1

1.0

0.9

0.8
0 20 40 60 80 100 120
t (min)

Fig. 14. Linear form of adsorption kinetics equation for Pseudo - second - order kinetic model
for Cr6+ adsorption on brick type C

1.5

1.4

1.3
t/Qt

1.2

1.1

1.0

0.9
0 20 40 60 80 100 120
t (min)

Fig. 15. Linear form of adsorption kinetics equation for Pseudo - second - order kinetic model
for Pb2+ adsorption on brick type C

Figures 16, 17 and 18 shows the possibility of up-taken of each metal ion into each
brick type.
628 U. A. S. L. Muthukalum et al.

300

Brick type A
Brick type B
250 Brick type C

200

Cu (mg/g)
150

100

50

0 50 100 150 200

Ce (mg/L)

Fig. 16. Cu2 þ uptake into brick materials

200

Brick type A
Brick type B
Brick type C
150
Cr (mg/g)

100

50

0 20 40 60 80 100 120 140 160

Ce (mg/L)

Fig. 17. Cr6 þ uptake into brick materials

250 Brick type A

Brick type B
Brick type C

200

150
Pb (mg/g)

100

50

0 20 40 60 80 100 120 140 160 180

Ce (mg/L)

Fig. 18. Pb2+ uptake into brick materials

According to the Figs. 16, 17 and 18, brick type B shows a higher possibility of up-
taking Cu2+ while brick type A shows ascending adsorption capacity as it leads to
predict a higher possibility of adsorption in higher Cu2+ concentrations. Each brick type
 Removal of Heavy Metals from Industrial Wastewater Through Minerals 629

shows almost similar adsorption capacities of up-take of Cr6+ in lower concentration

whereas brick type A shows increasing adsorption capacity in higher concentration.
Brick type A shows a significant adsorption of Cr6+ compared to other brick types.
The considered ion uptakes into each brick type is listed in the table. According to
the results, the amount of Cu2+ uptake to brick type A (106.6 mg/g) is significantly
higher compared to the amount of Cu2+ uptake to brick type B (11.04 mg/g). Pb2+
uptake value obtained for adsorbing to brick type A (753 mg/g) is considerably higher
compared to brick type B (25.08 mg/g) and it is relatively competitive with other
values as shown in Table 3. Brick type C shows relatively higher adsorption possibility
compared to brick type B but not as high as brick type A.

Table 3. Kinetic parameters obtained for A, B and C brick samples

Brick type Metal ion K1 (/min) qe (mg/g) R2
2+
A Cu 0.036 106.6 95.8%
C Cu2+ 0.004 497.0 97.2%
Brick type Metal ion K2 (g/ mg min) qe (mg/ g) R2
A Pb2+ 0.001 753 94.5%
B Cu2+ 0.022 11.0 92.4%
B Pb2+ 0.004 25.1 98.9%
C Cr2+ 0.032 191.8 97.8%
2+
C Pb 0.008 240.6 99.6%

4.2.2 Equilibrium Studies of Metal Ion Adsorption

Equilibrium studies were investigated based on two models which are Langmuir iso-
therm and Freundlich isotherm, for the uptake of three metal ions; Cu2+, Cr6+ and Pb2+,
for each brick type. The fitness of each model to the corresponding experiment was
observed and plotted as shown in Figs. 19, 20 and 21.

250

Cu
200
Cr
Pb

150
Qe (mg/g)

100

50

0 20 40 60 80 100 120 140 160

Ce (mg/L)

Fig. 19. Equilibrium studies of different metal ions into brick material A
630 U. A. S. L. Muthukalum et al.

250
Cu
Cr
Pb
200

Qe (mg/g) 150

100

50

0 20 40 60 80 100 120 140

Ce (mg/L)

Fig. 20. Equilibrium studies of different metal ions into brick material B

300
Cu
Cr
250 Pb

200
Qe (mg/g)

150

100

50

0 20 40 60 80 100 120 140 160

Ce (mg/L)

Fig. 21. Equilibrium studies of different metal ions brick material C

These figures show that adsorption of heavy metals initially increases with
increasing metal concentration and then levels off. The high metal ion uptake of brick
type C at the beginning of the equilibrium isotherm may be caused by the availability
higher specific surface area and pore volume. The experimental data were correlated by
the nonlinear form of Langmuir and Freundlich isotherms and its parameters calculated
from the curve fitting the results are listed in Table 4.
Considering Cu2+ ion uptakes measured for each brick type, brick type A shows a
significantly higher value (802.4 mg/g) compared to the other two brick types. Pb2+
uptake is higher into brick type C having 312.3 mg/g of adsorption capacity. In the
case of adsorbing Cr6+ into each brick type, which fits well with the Freundlich
isotherm, shows higher uptake in adsorbing brick type A.
 Removal of Heavy Metals from Industrial Wastewater Through Minerals 631

Table 4. Equilibrium parameters obtained for A, B and C brick samples

Brick type Metal ion Langmuir isotherm R2
K (l/mg) Qmax (mg/g)
2+
A Cu 0.003 802.4 98.9%
A Pb2+ 0.024 522.3 94.5%
B Pb2+ 0.042 187.2 99.8%
C Cu2+ 0.002 300.0 96.4%
C Pb2+ 0.004 312.3 97.9%
Brick type Metal ion Freundlich isotherm R2
K n
6+
A Cr 4.22 0.84 97.3%
B Cu2+ 3.36 0.65 99.7%
B Cr6+ 1.44 0.36 99.4%
C Cr6+ 2.84 0.58 98.2%

5 Conclusion

Brick materials have space inside their porous structure, which is the main reason they
adsorb toxic heavy metals that are presented in the wastewater molecules. The ability of
brick materials to accommodate those ionic species increases their adsorption capacity.
The adsorption capacity of the investigated two brick types for selected metal ions
(Cu2+, Pb2+, Cr6+) show considerably significant features. Considering Cu2+ ion, the
adsorption capacities, brick type A > brick type B > brick type C and for Pb2+ ion, the
adsorption capacities were brick type A > brick type C > brick type B. Cu2+ adsorbed
to brick type A fit well with Pseudo – first – order kinetic model and, Cu2+ adsorbed to
brick type B and Cu2+ and Pb2+ adsorbed to brick type B show a good fit with Pseudo –
second – order kinetic model. Brick type C is not much significantly different from
other brick types which shows relatively average adsorption capacities. In the case of
uptake possibility of the Cu2+, brick type B > brick type A > brick type C. Further
Cr6+ brick type B * brick type A > brick type C. It is evident that the presence of
meso-structure in brick materials allows it to effectively adsorb metal ions leading to a
substantial enlargement of the adsorption capacity.

Acknowledgements. The BET analysis data were obtained at department of chemistry, Cali-
fornia State University, Stanislaus, and the authors thanks to Dr. Amanpreet Manchanda for the
support with BET experiments. Further, thanks to Dr. Buddhika Dassanayake from Faculty of
Science, university of Peradeniya for the technical support with the XRF measurements.

References
Conseil National de Recherches du Canada (1980) Can J 64:1–8
Danielsson L (2013) Removal of chromium in wastewater with natural clays in southern Malawi
Garcia Sanchez A (1999) Sorption of heavy metals from industrial waste water by low-cost
mineral silicates. Clay Miner 34(3):469–477
632 U. A. S. L. Muthukalum et al.

Gunathilake C et al (2015) Adsorption of lead ions from aqueous phase on mesoporous silica
with P-containing pendant groups. ACS Appl Mater Interfaces 7(41):23144–23152
Kruk M, Jaroniec M, Sayari A (1997) Application of large pore MCM-41 molecular sieves to
improve pore size analysis using nitrogen adsorption measurements. Langmuir 13:6267–6273
Pare S (2012) Heavy metal removal from aqueous solutions by sorption using natural clays from
Burkina Faso. Af J Biotechnol 11(45):10395–10406
Rahman MM et al (2014) Removal of heavy metal ions with acid activated carbons derived from
oil palm and coconut shells. Materials 7(5):3634–3650
Rashed MN (2001) Lead removal from contaminated water using mineral adsorbents.
Environmentalist 21(3):187–195
Talaat HA, El Defrawy NM, Abulnour AG (2011) Evaluation of heavy metals removal using
some Egyptian clays, vol 6, pp 37–42
Tripathi A, Rawat Ranjan M (2015) Heavy metal removal from wastewater using low cost
adsorbents. J Bioremediat Biodegradation 6(6)
Uddin MK (2017) A review on the adsorption of heavy metals by clay minerals, with special
focus on the past decade. Chem Eng J 308:438–462
Wingenfelder U et al (2005) Removal of heavy metals from mine waters by natural zeolites.
Environ Sci Technol 39(12):4606–4613
Author Index

A E
Abeygunasekera, S., 479 Edirisinghe, D. I. U., 559
Abeysinghe, S. A., 137 Ekanayake, A. W. M. V., 282
Abeysiriwardana-Arachchige, I. S. A., 91, 107
Alwis, P. L. A. G., 261 F
Amarasinghe, B. M. W. P. K., 75 Fatiguso, F., 311
Ariyarathna, S. M. W. T. P. K., 559 Fawzia, S., 479
Ariyarathne, S. M. W. T. P. K. T., 288 Fernando, Sarah, 323
Arunakumara, K. K. I. U., 261 Fernando, Shiromal, 323

B G
Bakmeedeniya, L. U., 268 Galabada, G. H., 340
Bandara, C. S., 3, 36, 45 Gamage, J. C. P. H., 458, 479, 526
Bandara, H. R. L. C., 131 Ganyani, S. M., 567
Bandarathilake, H. M. D. P., 227 Gumbo, R., 581
Gunathilake, C. A., 615
C Gunawardane, S. D. G. S. P., 268
Cantatore, E., 311
Chandrasiri, G. W. H. A., 11 H
Chandrathilaka, E. R. K., 526 Halwatura, R. U., 340
Chapman, W., 107 Harries, K. A., 447
Chathura Sanjeewa, W. G., 147 Hendawitharana, S. U., 116
Cooray, V. Y., 252 Henkanatte-Gedera, S. M., 91
Herath, D. C., 171
D Herbert, Joselin, 419
Danthurebandara, M., 288 Hewavitharana, F. S. T., 202
De Fino, M., 311 Hippola, H. M. S. S., 36, 45
De Silva, L. K. U., 538
De Silva, Sudhira, 261 I
Delanka-Pedige, H. M. K., 91 Ireshika, G. P. R., 147
Denipitiya, D. N., 57
Dissanayaka, R. H. M., 458 J
Dissanayake, M. C. W., 433 Jain, A. K., 353
Dissanayake, P. B. R., 3, 36, 45 Jayakody, D. S. N., 378

© Springer Nature Singapore Pte Ltd. 2020

R. Dissanayake and P. Mendis (Eds.): ICSBE 2018, LNCE 44, pp. 633–635, 2020.
https://doi.org/10.1007/978-981-13-9749-3
634 Author Index

Jayasena, H. S., 604 O

Jayasena, N., 295 Obara, F., 433
Jayasinghe, T. D., 162 Ohashi, H., 433
Jayasiri, G. P., 3, 36, 45
Jayasooriya, E. M. S. D., 36, 45 P
Jha, K. N., 353 Palamakumbura, G. W. R. M. R., 227
Jinadasa, K. B. S. N., 538 Pallegedara, A., 282
Perera, A. A. D. A. J., 202
Perera, Chamal, 36, 45
K Perera, G. I. P., 239
Kalpage, C. S., 559, 615 Perera, G. M. C. A., 3
Kariyapperuma, K. A. D. D. T., 550 Perera, I. N., 252
Karunananda, P. A. K., 550 Perera, W. W. P. K., 394
Karunarathne, H. K. D. T., 538 Platt, S. L., 447
Karunaratne, S., 193 Poologanathan, K., 470
Karunawardena, A., 433 Priyankara, K. P. M., 239
Kawakami, T., 99 Priyasad, M. K. D. D., 116
Khandan, N., 91
Konthesingha, K. M. C., 470
R
Kularathne, K. A. M., 162, 171
Rajapaksa, R. M. G., 282
Kulathunga, A. K., 36, 45
Rajapakse, R. L. H. L., 116
Kulatunga, U., 295, 604
Rajapakshe, S. M., 99
Kumara, G. R. A., 282
Ranasinghe, E. N., 75
Kumara, K. M. S., 171
Ranathunga, P. G. K. B., 162, 171
Kumarasinghe, K. A. G. G. K., 11
Ranaweera, R. A. C. B., 147
Kumari, L. M. T., 295
Randil, Chameera, 36, 45
Kurukulasuriya, L. C., 365
Randil, O. P. C., 3
Rathnasinghe, A. P., 604
L
Rathnayake, K. A. I. L., 162
Liyanage, Ruwan, 147
Rathnayake, R. M. T., 183
Lowe, W. U. A., 211

M S
Maduranga, K. L. D., 22 Sanjeewani, H. D. M., 550
Madusanka, N., 295 Sathyaprasad, I. M. S., 22, 211
Magallagoda, L. P. G., 458 Scioti, A., 311
Mbohwa, Charles, 581, 595 Seneviratne, V. L. W., 489
McCabe, M. J., 447 Silva, M. A. L., 458
Mendis, H. S. A., 211 Siriwardana, C. S. A., 3, 45
Mendis, Priyan, 323 Siriwardane, Sudath C., 499
Munasinghe-Arachchige, S. P., 91 Siriwardena, C. S. A., 36
Munzara, R., 567 Solomon, Rutendo F., 595
Mushiri, T., 567, 581 Suntharalingam, S., 513
Mushiri, Tawanda, 595 Sylva, K. K. K., 36, 45, 183
Muthukalum, U. A. S. L., 615
Muthumala, C. K., 261 T
Takahashi, Y., 513
N Tchinda, D., 91
Nagaratnam, B., 470 Thambiliyagodage, C. J., 252
Nanayakkara, D., 489 Thennakoon, H. M. K. S. B., 288
Nanayakkara, S. M. A., 470 Thomas, Sajan, 419
Nandalal, H. K., 11 Tilak, S. D. L., 433
Nirmalakhandan, N., 107 Tripathi, K. K., 353
Author Index 635

U Weeraratne, R. P., 550

Udalamaththa, S., 57 Weerasekara, W. B. M. L. I., 99, 131, 137, 162
Udapitiya, G. A. M. A. S. B., 183 Weerasekera, K. S., 411
Upasiri, I. R., 470 Weragoda, S. K., 99, 131, 137, 162, 171
Upeka, S. A. C., 162, 171 Wijekoon, W. B. M. M. W., 147
Wijesekera, R. D., 252
V Wijewickrama, M. K. C. S., 604
Varghese, Jacob Thottathil, 419
Vasanthan, N., 433 Y
Venuja, T., 365 Yasamali, R. G. N., 22

W Z
Weerakoon, H. U., 22 Zhang, Y., 91