PDHengineer.

com
Course № M-4009

Filters for Industry

To receive credit for this course

This document is the course text. You may review this material at
your leisure either before or after you purchase the course. To
purchase this course, click on the course overview page:

http://www.pdhengineer.com/pages/M-4009.htm

or type the link into your browser. Next, click on the Take Quiz button
at the bottom of the course overview page. If you already have an
account, log in to purchase the course. If you do not have a
PDHengineer.com account, click the New User Sign Up link to create
your account.

After logging in and purchasing the course, you can take the online
quiz immediately or you can wait until another day if you have not yet
reviewed the course text. When you complete the online quiz, your
score will automatically be calculated. If you receive a passing score,
you may instantly download your certificate of completion. If you do
not pass on your first try, you can retake the quiz as many times as
needed by simply logging into your PDHengineer.com account and
clicking on the link Courses Purchased But Not Completed.

If you have any questions, please call us toll-free at 877 500-7145.

PDHengineer.com
5870 Highway 6 North, Suite 310
Houston, TX 77084
Toll Free: 877 500-7145
[email protected]
 CHAPTER 3
FILTERS FOR THE NUCLEAR INDUSTRY

3.1 Introduction
Filters are widely used in nuclear ventilation, air cleanup, and confinement systems to remove particulate
matter from air and gas streams. Air filters are defined as porous structures through which air is passed to
separate out entrained particulate matter. The word “filter” is derived from a word for the fabric called felt,
pieces of which have been used for air and liquid filtration for hundreds of years. The porous structures of a
filter may also be composed of granular material such as sand or fibers derived from cotton, minerals (glass,
asbestos), metals, or a wide selection of plastic materials. For filtration purposes, the fibers may be woven or
felted into a cloth or formed into a paper-like structure. Filters may also be constructed in the form of highly
porous fibrous beds of considerable depth. Other kinds of air cleaning devices (e.g., adsorbers, liquid
scrubbers, electrostatic precipitators) are sometimes referred to as “filters” because they are capable of
removing particles from an airstream. For clarity, the strict definition of a filter (given above) will be used in
this chapter.

High-efficiency particulate air (HEPA) filters are components of a nuclear treatment system that degrade with
service. The user/owner of the facility shall incorporate written specifications on the service life of the
HEPA filters for change-out criteria. Appendix C provides guidance on determining the acceptable service
life for each application pf HEPA filters.

Air Filter Types

Air filters of many types and materials of construction have been designed, manufactured, and applied to
meet a wide variety of industrial and commercial requirements for clean air (e.g., the nuclear industry makes
full use of all filter types). Commercially available filters are divided into three distinct categories based on
how they operate to remove suspended particulate matter from the air passing through them. The largest
category, often referred to as ventilation or heating, ventilation, and air conditioning (HVAC) filters, is
composed of highly porous beds of resin-bonded glass or plastic fibers with diameters ranging from 1 to
40 micrometers (µm). The fibers act as targets for collecting airborne dust. As their name indicates, HVAC
filters are widely used for air cleaning in mechanical ventilation systems. They are almost all single-use,
disposable items, and are used in all sectors of the nuclear industry, including as prefilters that reduce the
amount of coarse dust reaching more efficient filters located downstream.

A second category also is comprised of single-use, disposable filters called HEPA filters. By definition, a
HEPA filter is a throwaway, extended-medium, dry-type filter with: (1) a minimum particle removal
efficiency of no less than 99.97 percent for 0.3-µm particles, (2) a maximum resistance, when clean, of
1.0 inches water gauge (in.wg) when operated at 1,000 cfm, and (3) a rigid casing that extends the full depth
of the medium1 (Figure 3.1). [Note: Filters of different flows and resistances are allowable by the AG-1
Code.] 2 A filter of identical construction and appearance, but having a filtering medium with a retention of
99.9995 percent for 0.1 µm particles, is referred to as an ultra-low penetration aerosol filter (ULPA). The
filtering medium of HEPA filters is thinner and more compressed, and contains smaller diameter fibers than
HVAC filters. HEPA filters are widely used throughout all phases of the nuclear industry.

3-1
Nuclear Air Cleaning Handbook U.S. Department of Energy

A third category of commercial air filters is

Riveted or known as industrial cleanable cloth filters.
Bolted Corners
As the designation indicates, these filters
have built-in mechanisms for periodically
cleaning the filtering surfaces of accumulated
3/4 in. Wide
Double-Turned
dust. Unlike the first two types, industrial
Flanges Both cleanable cloth filters rely on building a thick
Faces layer of dust on the surface of the cloth to
provide a high-efficiency filtering medium.
Steel-Encased HEPA Filter
This type of filter is used in the nuclear
industry for ore processing and refining and
Corner Joint
Continuous Sheet of Nail or Detail
for similar tasks involving high
Papre Pleated Back Screw from concentrations of coarse mineral dusts.
and Forth Over Each Face
Corrugated Separators
Further, this third category includes special
types of particulate filters for chemical and
Gasket Corner combustion operations. These include deep
3/4 in. Thick Detail beds of sand in graded granular sizes, deep
Exterior Plywood
beds of glass fibers, and stainless steel
3/4 in. membranes formed from compressed and
Corrugated
Separators sintered granules or fibers. Stainless steel
membrane filters operate like industrial
1/4 in.
Wood-Cased HEPA Filter cleanable cloth filters in that they depend on
Notched
a dust layer for high-efficiency particle
removal and must be cleaned periodically,
Figure 3.1 – Filter Casing usually by reverse compressed air jets.

3.2 Filtration
The porosity of air filters has been noted. High porosity is associated with low resistance to airflow (e.g., low-
resistance HVAC filters contain approximately 97 percent voids). In a uniformly dispersed filter medium, the
individual fibers are relatively far apart—so far apart that the gaps between them are larger than the particles
removed from the air. This means that sieving (particle removal via openings that are smaller than the
particle dimensions) is not an important filtration mechanism. In fact, a sieve would make a poor air filter,
even one containing submicrometer openings, because each collected particle closes up a sieve opening so
that very soon no air can pass through. In contrast, filters collect particles from air and gas streams in a
number of well-defined ways that are associated with the dynamic properties of airborne particles. The filters
respond to the physical forces present as an aerosol passes through a porous medium composed of small
granules, fibers, or other shapes.

3.2.1 Particle Collection by Filters

Figure 3.2 shows the streamlines around a spherical granule or a single filter fiber lying normal to the flow
direction. A particle entering the flow field surrounding the fibers must follow the curved path of the
streamlines so it can pass around the obstacle. When particles possess sufficient inertia, they resist following
the curvature of the airstream and come in contact with the fiber because of their higher momentum relative
to that of the conveying gas molecules. The capturing effect of inertial impaction (see I in Figure 3.2) becomes
greater as both aerodynamic equivalent diameter and the velocity of the air approaching the fiber increase.

3-2
DOE-HDBK-1169-2003 Chapter 3

When suspended particles are very small, however, they tend to

follow the curved streamlines closely; that is, they have little
inertia, but are in vigorous, random motion (Brownian motion—see
II drawing in Figure 3.2). Therefore, when a streamline passes
close to the fiber surface, the random movements around the
streamline may result in some of the particles contacting the
fiber and adhering to it. This sets up a concentration gradient
between the zone close to the fiber and the bulk of the aerosol
which, in turn, results in particle diffusion in the direction of the I. The effect of inertial forces
fiber surface. The smaller the particles, the more vigorous their
Brownian motion and the more effective their filtration by
diffusion. Because the rate at which small particles cross
streamlines under the influence of diffusional forces is slow
compared to rate of the effects of inertial force on large particles,
separation of small particles by diffusion is enhanced by slower
velocities through a filter.

Particle collection by interception (III in Figure 3.2) occurs when a

particle traveling in a streamline that approaches a fiber within II. The effect of Brownian Motion

one particle radius makes contact with the fiber and adheres to it.
Interception is independent of flow velocity and is enhanced
when the diameter of the collecting fiber or granule approaches
the geometric diameter of the particle.

The several filtration mechanisms of importance are shown

together in Figure 3.3, where penetration (equal to 100 minus
collection efficiency) is plotted against particle size. The
penetration lines are not cumulative, as particles can be collected III. Particle caught by interception

but once; however, the net effect can be approximated by the

“dashed” summation curve. Figure 3.3 makes it clear there is a Figure 3.2 – Streamlines Around
particle size where both inertial and diffusional forces are a Filter Fiber
minimal and only interception is unaffected. This explains the
concept of a minimum filterable particle size. The exact minimum size depends on fiber diameter, filter
construction, and flow velocity. The minimum filterable particle size for currently manufactured nuclear
grade HEPA filter papers is close to 0.1 µm when operated at the design flow rate of 1 foot per second. The
effect of flow velocity on particle penetration for HEPA filter paper also shows a minimum efficiency point.

3.2.2 Particle Retention in Filters

After an airborne particle contacts a filter element, retention forces prevent re-entrainment under the
influence of the drag of the air. For small particles, the principal retentive force is a surface phenomenon
called the Van der Waals force, which is proportional to the total area of contact. For small spherical
particles, the fraction of the total surface area in contact with a filter fiber will be relatively large, resulting in a
retention force that exceeds the re-entrainment force of the air drag.

3.2.3 Airflow Resistance of Filters

Filter resistance is directly related to airflow rate and filter construction details. Decreasing the diameter of
filter fibers or granules produces higher resistance for the same overall unit volume of the solid fraction of
the filter medium. Greater filter depth at the same porosity increases resistance in proportion to the increase
in depth. Within limits, compressing a highly porous filter medium decreases porosity and increases flow

3-3
Nuclear Air Cleaning Handbook U.S. Department of Energy

resistance, but it does not have

Interception much influence on particle
As Particle Radius Increases, Line Moves removal efficiency until the
As Fiber Diameter Increases, Line Moves
As Interfiber Distance Increases, Line Moves
medium becomes highly
II. Direct Interception
compressed.

Inertia The text in Section 3.2.1 that

describes how fine particles are
As Particle Radius Increases,
Percentage Penetration

Line Moves
As Density Increases, Line collected by filter elements

III.
Moves
As Fiber Diameter Increases,
applies to new clean filters. As

Ine
particles collect on the surfaces
ion

rtia
Line Moves
iffus

of fibers or granules, or become

I. D

entrapped in the interstices

between upstream elements of
Diffusion the filter, the collected particles
As Particle Radius Increases Line Moves
As Fiber Diameter Increases Line Moves
tend to form a coherent dust
As Interfiber Distance Increases Line Moves layer known as a filter cake.
When this occurs, particle
collection gradually shifts from
Particle Size or Linear Velocity
media filtration (i.e., particle
removal by individual filter
Figure 3.3 – The Effects of Inertia, Diffusion, and fibers or granules) to cake
Interception on the Penetration-Velocity Curve filtration, and the filter shares
the characteristics of the
industrial cloth filter because the original structure now has the sole function of providing support for the
filter cake and the filter cake completely takes over the particle separation function. This transformation
produces two important changes: (1) efficiency increases in proportion to the increase in thickness of the
cake; and (2) after formation of a coherent filter cake, resistance of the filter to airflow, which initially
increased at a slow, steady rate as particles accumulated, now increases at an accelerating rate in response to
additional particle deposition and narrowing of the pathways. When cake filtration begins, the filter rapidly
reaches its terminal design airflow resistance. Figure 3.4 shows typical pressure rise curves for two HEPA
filters exposed to atmospheric dust. As shown, the long, slow pressure rise is clearly followed by a rapidly
accelerating increase. The reason for the abrupt change is the onset of sieving, which takes over when the
collected particles form a structure containing less space between the particles than the characteristic diameter
of the particles being collected. When HEPA filters reach this stage, they must be replaced.

3.3 HEPA Filters

The original specifications for HEPA filter media and cased filters were concealed under a veil of military
secrecy because of their use for chemical, biological, and radiological defense purposes. Following World
War II, the Atomic Energy Commission (AEC) chose the military's HEPA filters as their principal device for
particle removal in all exhaust air systems of nuclear facilities. Eventual expansion of the use of HEPA filters
for nonmilitary applications required declassification and release of information about HEPA filter
components and manufacturing methods (see Chapter 1). For this reason, military standards MIL-F-510683,
MIL-F-510794 (filter construction and filter medium preparation), and MIL-STD-2825 (filter testing) were
issued in an unclassified format.

MIL-F-510683 and MIL-F-510794 have now been withdrawn by the Department of Defense and replaced by
the American Society of Mechanical Engineers (ASME) Code On Nuclear Air and Gas Treatment, AG-12 and
U.S. Department of Energy (DOE) Standard (DOE-STD-3020-97).6 While MIL-F-510683 and
MIL-F-51079 were active, the Edgewood Arsenal in Maryland prepared a procurement guide for military and
4

3-4
DOE-HDBK-1169-2003 Chapter 3

nuclear agencies, the Qualified Products List (QPL),

which is based on exhaustive tests of manufacturers’
filter media and filters. The QPL referenced available 7.0

American Society for Testing and Materials (ASTM), Mini Pleat

Technical Association of the Pulp and Paper Industry 6.0 Deep Pleat
(TAPPI), and other standard test procedures and
equipment in its documentation of products. 5.0

Edgewood no longer maintains the QPL, and only

issues letters to manufacturers after qualification 4.0

testing. Standards incorporating the major provisions

of these military specification and qualification 3.0
standards have been issued. Besides AG-1,2 those
most relevant to nuclear service applications include 2.0
two standards administered by the ASME Committee
on Nuclear Air and Gas Treatment (CONAGT), with
participation from DOE and the U.S. Nuclear
1.0

Regulatory Commission (NRC). These standards

relate directly to HEPA filter applications in the 0
DEC JAN FEB MAR APR MAY JUN JUL AUG SEP

nuclear industry (i.e., ASME N509, Nuclear Power Plant

Air Cleaning Units and Components,1 and ASME N510,
Testing of Nuclear Air Cleaning Systems.7) The Figure 3.4 – Pressure Rise With
requirements of Nuclear Regulatory Guide 1.52 have Operating Time for Deep-Pleat and
been incorporated into these standards.8 DOE Mini-Pleat HEPA Filters
prepared a series of filter standards to establish the
performance and physical requirements for the filter media and cased filters used in DOE environmental
protection applications and to set policy and quality assurance procedures for DOE filter test facilities
(FTF).6, 9, 10, 11

While HEPA filters and their properties are discussed in this section, the same facts apply to ULPA filters
(except for differences in penetration, resistance, and media test velocity).

3.3.1 Filter Medium

Filtration theory implies that filter fibers must have diameters that are approximately the same as the aerosol
particles to be removed. Therefore, the standard HEPA filter medium must have fiber diameters of 0.2 to
0.5 µm to remove submicrometer particles, and even smaller fiber diameters are necessary for the ULPA filter
medium. All high-efficiency filters are now made from a mixture of glass fibers with carefully graduated
diameters that provide the required particle retention efficiency without exceeding the maximum airflow
resistance criterion and meet a wide variety of physical and environmental requirements. Typical glass fiber
sizes used to manufacture HEPA filter media are shown in Table 3.1. Small amounts of chemicals are
usually added to the glass fibers at the finish stage or after the medium is formed to impart desirable
properties to the product (e.g., mildew resistance, water repellency, increased tensile strength of the glass
paper). Plastic fibers in amounts less than 7 percent are sometimes added to the glass fibers to increase acid
resistance. The ASME AG-12 Code for the HEPA filter medium is now a universal standard. This is
primarily a performance standard, and the mixture of fiber sizes and specific additives and concentrations
vary among manufacturers. Each filter manufacturer has a proprietary formula that qualifies the product for
nuclear applications. Other nations have well-established criteria for HEPA filter paper that differ only to a
minor degree from the current U.S. standard. Microfibers of plastic materials such as polystyrene,
polycarbonate, and polyvinyl chloride also have been used for manufacturing HEPA filter media. Claims
have been made that triboelectric charge effects, which are induced on these plastic materials during
manufacturing, enhance filtration performance and save energy. Filters from these materials have found

3-5
Nuclear Air Cleaning Handbook U.S. Department of Energy

some acceptance in European markets, but have been rejected by the nuclear industry because of
flammability, high cost, and loss of performance under conditions such as high humidity, ionizing radiation,
and exposure to atmospheric contaminants. A HEPA filter medium made from polyvinyl chloride fibers has
been used in East European installations, but has been found unacceptable elsewhere for the reasons noted
above.

Table 3.1 – Sizes of Glass Fibers for HEPA Medium

Glass Fiber Industry Code Average Fiber Diameter (micrometers)
112 2.60 - 3.80
110 2.50 - 4.00
108B 1.20 - 2.40
108A 0.69 - 1.10
106 0.54 - 0.63
104 0.39 - 0.53
102 0.33 - 0.38
100 0.29 - 0.32
[Note: Glass Fiber Industry Code Numbers 100-110 were determined by the William Freeness
Test. Code 112 was determined by the Manville Micronaire Test FG-436-202 and calibrated by
the Brunauer, Emmett, and Teller Test (BET) Surface area.]

In addition to a limit on the organic material content of these filter papers (for fire and smoke control), other
qualification criteria include:

• Not less than 99.97 percent retention of 0.3-µm test aerosol particles at a flow rate of 32 liters per minute
through a paper area of 100 square centimeters;

• Clean airflow resistance not exceeding 40 millimeters of water at a filtration velocity of 320 centimeters
per minute (0.053 meters per second);

• Average tensile strength of not less than 179 g/cm of width in either direction after exposure to 6.0 to
6.5 × 107 rads;

• Resistance to excessive strength degradation after exposure to high temperature [698 ± 82.4 degrees
Fahrenheit (370 ± 28 degrees Celsius)] for 5 minutes and to wetting by immersion in water for
15 minutes; and

• Paper thickness of approximately 0.38 millimeters.

HEPA filter papers used for nuclear service currently provide collection efficiencies greater than
99.99 percent when tested with a 0.3-µm-diameter aerosol by the official U.S. test method contained in
MIL-STD-282.5 By increasing the fraction of fine glass fibers in the paper that are less than 0.25 µm in
diameter, it is possible to obtain efficiencies in excess of 99.999 percent for 0.1- to 0.3-µm particles with a
modest increase in filter resistance—typically about 25 percent. Performance standards for filter papers that
are acceptable for use in nuclear-grade HEPA filters (as distinguished from performance standards for
fabricated filter units that contain such materials) have not been considered important by some nuclear
authorities. This view is based on the assumption that, unless the glass fiber filter paper has the required
characteristics, the completed filter unit will not meet the acceptance criteria. This approach is reasonable,
provided the filter paper is subjected to equivalent stresses after fabrication (e.g., shock, ionizing radiation,
heat, fire).

3-6
DOE-HDBK-1169-2003 Chapter 3

The filter media production usually constitutes the definition of a batch for HEPA filter manufacturing.
Typically, a batch of media can be used to make a lot of only 6 to ten 24- × 24- × 11 1/2-inch HEPA filters.
Any selective filter testing (as opposed to 100 percent testing at the manufacturers’ or FTF) should be done in
accordance with ASQC-Z 1.4-1993, with the batch size set by the media batch production capability of the
manufacturer. To utilize this standard, the user must also select the appropriate reliability. A value of
90 percent or greater is appropriate for nonsafety class HEPA use.12

3.3.2 HEPA Filter Construction

Most HEPA filter units are constructed the same way—a continuous length of filter paper is folded back and
forth into pleats and corrugated separators are inserted between each fold. The assembly is then sealed into a
rigid, open-faced rectangle. The components of a fabricated HEPA filter include: (1) extensively pleated
filter medium, (2) separators that provide air passages and keep adjacent pleats apart, (3) a rigid filter case that
encloses and protects the fragile filter medium, (4) sealants used to bond the filter pack (consisting of the
assembled pleated medium and separators) to the filter case and to eliminate leak paths between filter pack
components, and (5) gaskets attached to the filter case on one or both open faces to provide an airtight seal
between the filter and the mounting
frame. Some filter construction Filter Casing

methods form the filter paper on

the papermaking machine using an
interval means to keep the adjacent
folds apart, thereby eliminating a Separator
need for corrugated separators.
These filters are called separatorless
HEPA filters (see Section 3.3.3).
Figure 3.5 shows the assembled
components of an open-face, deep-
pleat HEPA filter with corrugated
separators.

3.3.2.1 Separators

The most widely used material for Gasket Seal

the interleaved corrugated Adhesive Bond
Continuous Sheet of
separators is tempered aluminum Between Filter Pack
Flat Filter Medium
and Integral Casing
foil. The aluminum foils currently
used for separators are identified as Figure 3.5 – Open-Face Deep-Pleat HEPA Filter-
ASTM B209, Standard Specification for Type A Filter Pack
Aluminum and Aluminum Alloy Sheet
and Plate,13 alloys 1145-H19, 3003-H19, or 5052-H39, and are a minimum of 0.035 mm thick. When
corrugating the aluminum sheet into separators, edges are often hemmed (turned back on themselves) to
prevent the sharp edges from puncturing or tearing the part of the filter medium folded around the separator.
Examination of disassembled filters aged up to 10 years showed deterioration of uncoated aluminum spacers
to be common to all operating environments. Corrosion leads to adhesion of the spacer to the glass fiber
medium. Levels of radioactive contamination on the evaluated filters appeared not to have affected the aging
process. When greater chemical resistance is required, a plastic coating of an epoxy, thermo-set vinyl (or a
similar compound) is applied to the aluminum sheet. [Note: If significant radiation is a concern, the use of
organic materials may not be appropriate.] A dye is usually added to clear coating materials so that defects in
the plastic coating can be easily detected. After drying to a film, the coating must be 0.0025- to 0.0050-mm
thick, with no cracking, peeling, or delamination after corrugation. Experiments to determine the corrosion-
resistance of certain all-plastic separators have been conducted and have generally found them to be

3-7
Nuclear Air Cleaning Handbook U.S. Department of Energy

unacceptable because the corrugations tend to reflatten due to “plastic memory,” particularly after exposure
to moderately high temperatures. ASME AG-12 details additional requirements for corrugated aluminum
separators.

3.3.2.2 Filter Case

The filter case is constructed of materials that correspond to the specific application, decontamination
requirements, and considerations of disposal ease and cost. Commonly used case materials include fire-
retardant plywood, chromized carbon steel, and alloys UNS S30400 and UNS S40900 stainless steels. The
minimum thicknesses required to maintain rigidity under compressive loads ranging up to 1,400 pounds when
the filter is clamped to a mounting frame, are 3/4 inch for wood and manufacturer’s standard steel sheet
gauge for steel. Grade A-C, American Plywood Association (APA) PS-1 fire-retardant-treated plywood is
acceptable, but the “A” face must be on the inside, facing the pack, and should be assembled with this face
completely coated with a sealant to close off any leak paths. The outer face should be filled and sanded as
smooth as possible (for plywood). This is particularly important for nuclear plant workers whose gloved
fingers and hands must not be punctured by splinters from a wooden frame when replacing filters in a
contaminated area. For wooden case filters, case panels are to be joined with rabetted joints, which are
assembled by gluing with an adhesive and double nailing or doubling screwing with coated box nails,
corrosion-resistant plated screw nails, or flat-head wood screws. The end points of the fasteners must not
penetrate the inside or outside surfaces of the case. Metal cases should be used in instances of potential
wetting or high humidity at elevated temperatures and when the filter will be exposed to corrosive chemicals.

3.3.2.3 Sealants

Sealants used to provide a leak-free bond between the filter pack and case must be resistant to heat and
moisture, noncombustible, fire-resistant, or self-extinguishing, as well as capable of maintaining a reliable seal
under continuous exposure to design operating conditions. Rubber-based adhesives compounded with
chlorine or bromine to ensure self-extinguishing when exposed to ignition are acceptable, but catalytically
cured solid and foamed polyurethanes containing additives for combustion suppression are the sealants of
choice for most filter manufacturers. Sealants should maintain their integrity over a wide temperature range.
Filters designed to operate at temperatures above 392 degrees Fahrenheit (200 degrees Celsius) have been
sealed with compression-packed glass fibers and with ceramic cements reinforced with glass fibers, and have
been hardened thermally. Compression-packed glass fiber seals are sometimes found to be damaged after
shipment. The ceramic seal is often too brittle to withstand commercial shipment. Room temperature
vulcanizing silicone rubber sealants have been used successfully at operating temperatures only slightly lower
than 392 degrees Fahrenheit (200 degrees Celsius).

3.3.2.4 Gaskets

Filters must be installed so that even the smallest volume of air or gas does not escape filtration; therefore,
gaskets and alternative methods of sealing filter units to the mounting frames play a critical role in the
satisfactory operation of HEPA filters. The most widely used sealing method is a flexible gasket attached to
the open face of the filter case and pressed against the flat face of the mounting framework. The second
most popular method is referred to as a “fluid seal.” This method uses a channel formed or routed in the
peripheral face of the filter case that is filled with a highly viscous, very low volatility, nonflammable (or self-
extinguishing), odor-free, non-Newtonian fluid such as a silicone. The fluid flows around and over
imperfections, but does not relax or separate from the surfaces it contacts. For installation, the matching
framework face is equipped with a continuously protruding knife-edge that mates with the fluid-filled channel
in the filter case. The reverse arrangement of a protruding knife-edge on the filter and a fluid-filled channel
on the mounting frame also may be employed. These two mounting methods do not have interchangeable
parts, so hybrid sealing systems are not feasible.

3-8
DOE-HDBK-1169-2003 Chapter 3

Gaskets must be oil- and ozone-resistant.14 Closed-cell sponge gaskets composed of synthetic rubber
(neoprene) that conforms to grade 2C3 or 2C4 of ASTM D1056, Sponge and Cellular Rubber Products15 have
been widely used. Gaskets should have a minimum thickness of ¼ inch and width of ¾ inch. The gasket
face attached to the filter case should be free of any adhesion-resistant mold-release contaminant that may
have been acquired when the gasket material was molded. To ensure an absence of residual mold release
chemical, only cut surfaces are permitted on both gasket faces. Gaskets may be cut out of a sheet of stock as
a single piece or may be made of strips joined at the corners by dovetail or other interlocking arrangement.
Joints are sealed against air leakage with a rubber-base adhesive, usually the same adhesive used to attach the
gasket to the filter case. Manufacturers of neoprene gaskets recommend a shelf life not to exceed 3 years.

3.3.2.5 Faceguards

To guard against damage from careless handling and faulty installation procedures, a recessed faceguard
should be installed across both faces of the filter during fabrication. Woven or expanded metal with square
openings approximating 1/3 inch to 1/2 inch on a side have proven satisfactory in largely preventing the
inadvertent intrusion of hands or other objects into the filter pack. In addition, a metal mesh faceguard
provides added strength to the filter unit, increasing resistance to transportation damage and shock
overpressure. Faceguards should conform to either galvanized steel ASTM A74016 or 304 stainless steel
ASTM A580.17

3.3.3 Separatorless HEPA Filters

A separatorless HEPA filter design,18 shown in Figure 3.6, is constructed without corrugated spacers
inserted between the folds of the filter paper. Instead, a continuous sheet of filter paper is molded on the
papermaking machine with corrugations at intervals. When it is folded back and forth upon itself, it becomes
a self-supporting pack where the peaks of the interval corrugations of successive layers contact each other to
form a honeycomb-like filter pack.
Filter Casing
For the same filter frame size, a
separatorless filter contains more
useful filter paper surface than the
corrugated separator type, and thus
provides greater airflow capacity at
equal resistance.
Continuous Sheet
of Corrugated
Filter Medium
3.3.4 Mini-Pleat HEPA
Filters
Mini-pleat filter construction
methods utilize 7/8 to 1 1/4-inch-
deep pleats with very narrow air
spaces (1/8-inch) between, making
it possible to pack more filter paper
into the standard frame sizes than
can be done with deep-pleat, Adhesive Bond
corrugated separators, or even by Between Filter Pack
and Integral Case
using separatorless construction
methods. Abutting folds are
Figure 3.6 – Separatorless Style Filter-Type C
separated by threads, ribbons,
Filter Pack
tapes, strips of medium, or
continuous beads of glass, foam, or

3-9
Nuclear Air Cleaning Handbook U.S. Department of Energy

plastic spaced across the width of the medium. Mini-pleat filters contain almost twice as much filter paper as
deep-pleat, corrugated separator filters of equal frame size (Figure 3.7) (see Section 3.3.2.3). They are rated
to have an airflow resistance of 0.25 Kilopascals (kPa) when operated at 3,060 cubic meters per hour (m3/hr),
compared to the same resistance for a
flow rate of 1,700 to -2,040 m3/hr for
deep-pleat corrugated separator filters.
This gives the user of mini-pleat filters
the option of utilizing space-saving
higher airflow rates or extending filter
life by operating at lower than rated
airflow capacity. This is called
downrating a filter.

When a mini-pleat filter rated for

3,060 m3/hr is downrated to service at
1,700 m3/hr, it theoretically should
extend service life more than threefold
before it reaches its final permissible
resistance increase. In practice, filter
life extension was found to be merely
1.6-fold because of dust bridging
across the very narrow air passages
between the paper pleats to form a
filter cake covering the face area. An
efficient prefilter might be used to
prevent the formation of a surface
filter cake and extend the service life of
the mini-pleat filter.

Cased mini-pleat HEPA filters are

Figure 3.7 – Mini-Pleat (Thread Separator) Filter- formed from subcomponents
Type B Filter Pack assembled in a continuous “V” array.
The subcomponents are panels that
hold the pleated filter paper in metal frames approximately 23.62 inches wide, 11.81 inches high, and the
depth of the paper pleats. A seal is made between framed filter packs and the standard frame using rubber-
based adhesives, polyurethane, or some other plastic-based material, all of which are chemically compounded
to inhibit their support of combustion.

Another mini-pleat filter design is formed by molding narrow longitudinal ridges into the wet filter paper at
approximately 1-inch intervals while the paper is still on the papermaking machine, then folding the paper as
it comes off the machine into mini-pleats that may be 2, 4, or 6 inches deep.18 The filter pack is mounted
into the filter case perpendicular to the airflow direction instead of mounting a number of shallow panels
arranged inside the filter frame in a series of “V” formations The 6-inch-deep mini-pleat separatorless filter
contains the same area of filter paper as the 12-inch-deep separator type. This type of filter has been placed
into service, but there is no experience to report for nuclear applications.

3.3.5 HEPA Filter Classes and Sizes

In addition to being the workhorse filter for the nuclear industry, HEPA filters have found many important
applications in the industrial, medical, pharmaceutical, and microelectronic sectors. These diverse
applications have resulted in a number of industrial and governmental specifications. In general, these

3-10
DOE-HDBK-1169-2003 Chapter 3

specifications can be grouped into five construction grades and three performance types that provide a range
of materials, manufacturing techniques, performance characteristics, and costs for different applications and
user preferences. A standard covering the grades and types of HEPA filters has been issued as
IEST-RP-CC001.3 by the Institute of Environmental Sciences and Technology.19 This standard lists the
following classifications.

3.3.5.1 Filter Construction Grades

Grade 1 – Fire-Resistant Filters. Filters of this grade must contain fire-resistant materials that may ignite
when the filter is exposed to hot air or fire, but will not continue to burn once the ignition source is removed.
The filter must exhibit a specified retention efficiency after exposure to no more than 700 ± 50 degrees
Fahrenheit (371 ± 10 degrees Celsius). These filters comply with ASME AG-1, Section FC.2

Grade 2 – Semicombustible Filters. This grade costs less, but provides a lower level of protection against
elevated temperature than Grade 1. For this reason, the user should evaluate application of this filter grade
with the individual fire propagation hazards in the area of use. This filter type will fail at temperatures much
lower than Grade 1. These filters comply with UL 586.20

Grade 3 – Combustible Filters. This grade covers filters required for certain service requirements that
permit acceptance of the combustibility hazard. Grade 3 filters are readily combustible and are used only
where high-value product recovery by incineration is desirable, disposal of volumes are critical, or exposure to
chemical atmospheres might be incompatible with the use of a HEPA filter incorporating a medium of glass
fibers. It should be noted that manufacture of a combustible HEPA filter medium formulated from asbestos
and cellulose has been discontinued for more than a decade because of the hazards associated with its use and
the resulting low demand. Specialty filter media for recovery of precious metals by incineration are still
available. These filters comply with UL 900, Class 1.21

3.3.5.2 Filter Performance Levels

IEST-RP-CC001.319 classifies filter performance levels as:

Type A Filter Performance. Sometimes referred to as industrial types, these filters are tested for overall
penetration at rated flow only. The filter retention (inverse of penetration) must exceed 99.97 percent for
0.3-µm particles. ULPA filters greater than this value can be obtained upon agreement between the buyer
and seller.

Type B Filter Performance. In addition to the basic requirements for Type A filters, Type B units are
certified free of significant pinhole leaks that would cause penetration at low flow rates. This type is tested at
20 percent of rated airflow with the filter encapsulated to disclose casing or gasket leaks. This type is
sometimes referred to as “nuclear-type.”

Type C Filter Performance. In addition to the performance required of Type A filters, Type C filters, are
tested with the use of air-generated test aerosols at 80 to 100 feet per minute (fpm) face velocity. The units
are fully face-scanned to detect and eliminate all significant leakage streams greater than 0.01 percent of the
upstream test aerosol concentration to which the filter is subjected. This type is infrequently called “laminar-
flow type.”

Type D Filter Performance. In addition to the testing required for Type C filters, Type D filters should be
retested at their rated airflow and penetration, which should be no more than 0.001 percent of the upstream
concentration. The filter unit should be encapsulated so that all components, including the filter pack, frame,

3-11
Nuclear Air Cleaning Handbook U.S. Department of Energy

and gasket, are subjected to testing. In the U.S., laser spectrometers are used to measure efficiencies of
ULPA filters (>99.99999 percent).

Type E Filter Performance. Type E filters are designed, constructed, and tested in strict accordance with
military specifications for HEPA filters intended for biological use.22 This type is for application in air
cleaning or filtering systems involving toxic chemical, carcinogenic, radiogenic, or hazardous biological
particulates. This type is referred to as a “biological unit.”

UL Class 1,21 Type B filters are recommended for most nuclear applications, particularly in single-pass
systems. These units comprise a large part of those manufactured by industry and are used extensively in
nonnuclear industries as well. UL Class 1, Type C filters are common in clean room applications where
laminar flow requirements are coupled with low particle
penetration.23 UL Class 1, Type D filters presently are used
in printed-circuit or microprocessor clean rooms.

3.3.5.3 Enclosed Filters

Most HEPA units are used in the open-face configuration

(Figure 3.1). When used in this manner, the filter is
secured firmly to a rigid framework by a pressure device
such that a leak-free seal exists between the unit and the
framework. The HEPA filter may also be placed
completely within an enclosing casing that is equipped with
nipples at both ends for attachment to existing ventilation
ducts (Figure 3.8). Enclosing casings may be metal or
plywood, but care must be taken to ensure the casing Figure 3.8 – Enclosed HEPA Filter
material is compatible with Underwriters Laboratories, Inc.,
(UL) requirements for resistance of the filter to heated air and flame.22 The enclosing casing forms the leak-
free pressure boundary in addition to the filter case, and care must be taken to ensure that it is treated as an
encapsulated design for both performance and leak-acceptance testing. Enclosed HEPA units have
significantly higher resistance to airflow than the open-faced design because of the added restrictions of the
duct transitions.

Enclosed filters are sometimes referred to as encapsulated (nipple-connected, closed-face, or self-contained)

HEPA filters. They are not recognized by applicable codes (i.e., AG-12) and standards and fail to meet all the
requirements contained in DOE Standard DOE-STD-3020-976. The most
serious deficiency is failure to meet the requirement for uniform velocity across
the filter face. This can invalidate the in-place filter leak test.

The enclosed filter and its casing are often misused as part of a nuclear
ventilation system pressure and confinement boundary. Enclosed HEPA
filters are not specifically designed, analyzed and tested to meet either the
housing or the ventilation ducting containment requirements of nuclear codes.
When designing and constructing new nuclear facilities, enclosed HEPA filters
should not be used in nuclear ventilation systems. When an installed
ventilation system is being modified or upgraded, consideration should be
given to replacement of enclosed HEPA filters with nuclear grade housings
containing ASME AG-1 certified filters. A technical justification should be
developed where the enclosed filter is not replaced with a housing.
Figure 3.9 – Open-Faced
Cylindrical Axial Flow
HEPA Filter

3-12
DOE-HDBK-1169-2003 Chapter 3

3.3.5.4 Cylindrical Filters

Cylindrical filters may be either open-faced cylindrical axial

(Figure 3.9) or radial flow (Figure 3.10). Filters fabricated
with cylindrical cases appear to offer substantial advantages
such as easier mounting in circular ducts, but in practice they
have been found to have disadvantages attributable to
manufacturing difficulties, escalated costs, and increased
susceptibility to leakage. However, cylindrical filters offer
significant advantages regarding simplified gasketing and
automated filter-changing techniques. In the United Kingdom,
a “push-through filter system” has been developed that permits
changing of cylindrical filters by loading a clean filter that has
gaskets on the top and bottom filter flanges into the filter
housing tube from the “clean side,” then pushing it through
until it ejects the old contaminated filter into the “dirty side” of Figure 3.10 – Radial Flow HEPA Filter
a cell or glovebox. A cylindrical filter of somewhat different
design, but with similar characteristics, has been developed in
the United States.

3.3.5.5 Filter Sizes

The physical dimensions shown in Table 3.2 have been standardized for the HEPA filters currently used in
nuclear service and by U.S. Government agencies. [Note: DOE STD-3020-97 addresses more sizes than are
indicated here, and may be used in addition to the table shown below.] Other sizes can be manufactured and
purchased, but are considered “special orders.” Nonnuclear applications (clean rooms, biological safety
cabinets, medical facilities) generally use the same filter height and depth dimensions shown in Table 3.2, but
may have lengths up to 72 inches. Special HEPA filter configurations for computer applications use many
different sizes and shapes depending on the volume available within the computer cabinet. As many as 1,000
different configurations exist, each specific for a respective manufacturer, model, type, or size of computer.

Table 3.2 – Nominal Sizes and Ratings

Size Minimum Rated Airflow Maximum Resistance
Standard Cubic
Number Feet per Minute Inches Water Pascal
Designation Inches Millimeters (scfm) m3/hr Gauge (in.wg) (Pa)
1 8 × 8 × 3 1/16 203 × 203 × 78 25 42 1.3 325
2 8 × 8 × 5 7/8 203 × 203 × 149 50 85 1.3 325
3 12 × 12 × 5 7/8 305 × 305 × 149 125 212 1.3 325
4 24 × 24 × 5 7/8 610 × 610 × 149 500 850 1.0 250
5 24 × 24 × 11 1/2 610 × 610 × 292 1,000 1,700 1.0 250
6 24 × 24 × 11 1/2 610 × 610 × 292 1,250 2,125 1.3 325
7 24 × 24 × 11 1/2 610 × 610 × 292 1,500 2,550 1.3 325
8 24 × 24 × 11 1/2 610 × 610 × 292 2,000 3,400 1.3 325
9 12 × 12 × 11 1/2 305 × 305 × 292 250 424 1.3 325

[Note: AG-1 currently allows for the qualification of the largest size to apply smaller size filters, i.e., a size
5 filter can be used to qualify a size 4 filter. It has been brought to the attention of the CONAGT that the

3-13
Nuclear Air Cleaning Handbook U.S. Department of Energy

qualification of the size 4 filter listed above may need to be independent of the size 5 qualification. Readers
should check revisions to AG-1 post 2003].

3.3.5.6 Filter Weight

The weight of a filter unit is an important factor in design and maintenance. Table 3.3 lists the weight of
clean, open-faced filters and enclosed filters of rectangular design. For design purposes, the weight of a dirty
filter that is ready for change-out is approximately 4 pounds heavier per 1,000 cfm of rate capacity. Because
many applications employ multiple filter units in banks that are as many as 6 to 10 units in height, minimal
filter weight, without loss of performance, is critical to the ease of original installation and replacement.

Table 3.3 – Weight of Unused HEPA Filters

Approximate Weight (pounds) of Filters With
Filter Size (inches) Nominal Airflow Capacity (cfm) Wood Case Steel Case
Open-face
8 × 8 × 3 1/16 25 2 3
8 × 8 × 5 7/8 50 3.6 5.8
12 × 12 × 5 7/8 125 4.8 7.3
500 17 22
24 × 24 × 5 7/8
1000, 1250, 15000 32 40
24 × 24 × 11 1/2
Enclosed 25 5 9
8 × 8 cross-section 50 7 10.5
8 × 8 cross-section 125 17 20
12 × 12 cross-section 500 64 72
24 × 24 cross-section 1000 78 95
24 × 24 cross-section

3.3.6 HEPA Filter Performance Characteristics

3.3.6.1 Airflow Resistance

Resistance to airflow (pressure drop) of a nuclear-grade, 1,000 cfm capacity filter should not exceed 1 in.wg
when tested at rated airflow (see Table 3.2 for additional filter capacities and pressure drops). The pressure
drop for ULPA filters is frequently greater than for standard HEPA filters, and this feature is subject to
negotiation between customer and vendor. Resistance increases with particulate loading. A new nuclear-
grade filter is qualified by a wet overpressure test up to 10 in.wg for 1 hour; however, this should not be
confused with normal in-service operating pressures. Normal in-service pressures should be limited to 3 to
5 in.wg above startup pressure.

3.3.6.2 Dust-Holding Capacity

The dust-holding capacity of a filter is a function of the type, shape, size, and porosity of the filter as well as
the aerosol size, shape, and concentration characteristics to which the filter is exposed. As HEPA filters are
designed to filter out the smallest particles, they can accommodate only extremely light particulate loadings
without experiencing a rapid pressure drop increase. HEPA filters are affected particularly adversely by
fibers, lint, and other materials that exhibit a large length-to-diameter ratio because they tend to bridge the air
entrance gaps between the adjacent pleats of medium, thereby preventing particles from accessing the full
depth of the filter. A HEPA filter can be protected by a prefilter capable of removing the bulk of large
particles and fibers, thereby extending its useful lifetime. As noted earlier, a dust-holding capacity of
4 pounds per 1000 cfm of rated airflow capacity may be assumed for design purposes. This is probably a
conservative figure for granular dusts, but may overestimate the filter’s dust-holding capacity for metal fumes.

3-14
DOE-HDBK-1169-2003 Chapter 3

An increase in dust accumulation on the filter medium both improves filtration efficiency and increases
resistance to airflow. One of the limitations of HEPA filters is their low-dust-holding capacity and their need
for frequent replacement when exposed to high aerosol concentrations. The pressure rise curve experienced
by HEPA filters also depends on the particulate composition of the atmosphere to which it is exposed. A
filter installed in a moderately contaminated urban area will show as much as a six-fold increase in resistance
in a year's time, whereas a unit in a clean room application may last ten years or longer before reaching a six-
fold pressure increase. The use of a prefilter (described in Section 3.4) increases the service life of HEPA
filters and helps make the combined filtration system cost effective.

Tests conducted at the Harvard Air Cleaning Laboratory24 explored the pressure buildup of filter units under
urban conditions. During testing, commercial deep-pleat, aluminum-corrugated separator HEPA filters and
mini-pleat HEPA filters, all 24 × 24 × 11.4 inches in size, were exposed side-by-side to an urban atmosphere
while being operated continuously at rated and downrated airflow without prefiltration. The downrated mini-
pleat HEPA filters did not fulfill the theoretical prediction of three times the service life of a deep-pleat
U.S. HEPA filter when both were operated at 1,700 m3/hr; instead, an extended service life of about
1.6 times was achieved. This shortfall was attributed to dust and lint bridging the narrow openings between
the pleats of the mini-pleat unit (the pressure rise curves of the two filter types are illustrated in Figure 3.4).
Extremely high concentrations of soot and dense particular matter from fire conditions may overwhelm both
the prefilters and the HEPA filters, thereby inactivating the total system. For this reason, some practical
means of suppressing smoke before it reaches the filters is required. Water curtains, electrostatic precipitators,
inertial separators, or other devices have been utilized for this purpose with varying success.

3.3.6.3 Shock and Blast Resistance

The resistance of HEPA filters to shock and blast is important because these filters are often the final barrier
between a highly contaminated enclosure and the environment. Shock stress may occur from disruptive
natural phenomena (e.g., earthquakes) or from internal and external explosions.

Early tests at the Harvard Air Cleaning Laboratory showed that filter units of 1950s vintage sustained
moderate damage at 6-inch-mercury [2.95 pounds per square inch (psi)] overpressure, and complete
destruction at 10-inch-mercury overpressure (4.91 psi). The U.S. Navy determined that filter units subjected
to an overpressure simulating an atomic explosion (50-millisecond duration) failed at variable values
depending on the face and depth dimensions. The values listed in Table 3.4 are the maximum shocks that
can be tolerated without visible damage or loss of filtration efficiency. Specific conclusions reported from the
Harvard study included: (1) filters with faceguards on both faces had about a 40 percent greater resistance to
shock than those without faceguards; (2) dirt-loaded filters had 15 percent less shock resistance than clean
filters; (3) the smaller the filter face area, the greater the resistance to shock; (4) the greater the filter depth,
the greater the resistance to shock. At overpressures exceeding those listed in Table 3.4 by 0.5 to 1.0 psi, the
filter medium ruptured or experienced cuts on the downstream face. At pressures 2 psi greater than those
listed in Table 3.4, extensive damage occurred. At pressures above 5 psi, the entire filter pack within the
frame was dispersed. No significant differences were found between successive tests of increasing shock
force on the same filter and a one-shot test of the same force—both procedures produced the same failure
modes. Using the data on shock overpressure resistance versus face depth and dimensions, Burchsted18
produced the chart shown in Figure 3.11. Los Alamos National Laboratory (LANL) repeated some of the
Navy shock tests and arrived at similar values for loss of structural integrity. In addition, the researchers
discovered that, although the break point for the units was similar in value, the specific values for rupture
were highly dependent on the filter source. Tests on HEPA filters constructed with a special scrim-backed
glass-fiber filter medium showed that this filter retained an efficiency in excess of 99.92 percent for the test
aerosol after exposure to a differential pressure of 7.5 kPa and a temperature of 932 degrees Fahrenheit
(500 degrees Celsius).

3-15
Nuclear Air Cleaning Handbook U.S. Department of Energy

LANL also conducted tests on filter units under simulated tornado pressure loadings (represented by a slower
pressure buildup, but sustained for a longer period of time). Damage levels in these tests were identical to
those found for shock overpressures of the same level, but shorter duration. Filters of U.S. and European
manufacture gave comparable results. LANL found separatorless filters had only two-thirds the structural
strength of their separator-containing counterparts when subjected to tornado conditions, and only one-half
the strength under shock overpressure exposures. However, another series of seismic simulation tests
conducted by Wyle Laboratories found that separatorless filters successfully withstood seismic shocks
equivalent to 12 moderate (less than 4.0 Richter scale) earthquakes when correctly mounted in well-designed
housings. During these tests, the filters were operated at design flow rate of 1,700 m3/hr, but under
cumulative (multiple earthquake) worst-case conditions. The units were challenged continuously with
heterogeneous test aerosol, with no demonstrated resulting loss of efficiency for the filter, housing, or fluid
seal between the filter and housing. Current NRC regulations do not require seismic testing for filters, but do
allow mathematical analysis of the housing, with the sole consideration being the weight of the filter(s) in the
housing.

Table 3.4 – Shock Overpressure Resistance of Open-face HEPA Filters25

Overpressure (psig)
Filter Dimensions (inches) Recommended Design Limit for Used Filters
Face Depth Overpressure at Failure a With Faceguards Without Faceguards
8×8 3 1/16 3.7 b 2.0
8×8 5 7/8 4.5 b 2.5
12 × 12 5 7/8 3.6 b 2.0
5 7/8 2.2 1.7 1.2
24 × 24
11 1/2 3.2 1.7 1.8
24 × 24
a Clean filter with 4 by 4 mesh faceguards on both faces.
b Faceguards not available.

3.3.6.4 Heat from Fire and Explosion

Grade 1, fire-resistant filters are 3.5

fabricated from a glass medium Depth
11 1/2"
with flame-inhibited or self- 12"x12"x11 1/2"
5 7/8"
3.0
extinguishing adhesive or 3 1/16"

sealant, aluminum alloy 12"x12"x5 7/8" 8"x8"x5 7/8"

separators, and fire-retardant 2.5
wood or metal frames. 24"x24"x11 1/2"
Shock-Overpressure

8"x8"x3 1/16"
Nevertheless, the material that
Resistance (psi)

2.0
collects on the filters poses
special fire and explosion 12"x12"x3 1/16"
hazards when it contains 1.5

substantial amounts of organic 24"x24"x6 7/8" A

or pyrophoric substances. Fires 1.0
from this source can produce
24"x24"x3 1/16"
undiluted hot gases that attain 24"x48"x5 7/8"
B d
.5
temperatures as high as 1,830
24"x72"x5 7/8"
degrees Fahrenheit. The
softening point of glass fibers 0
1 2 4 6 8 10 20 40 60 80 100
used in currently manufactured 1000 d
2 2
A +B
HEPA filter media is about
1,250 degrees Fahrenheit, and Figure 3.11 – Shock-Overpressure Resistance of Clean HEPA Filters
direct impingement of a (Separator Type) as a Function of Size

3-16
DOE-HDBK-1169-2003 Chapter 3

1,700 degrees Fahrenheit flame will cause immediate melting. A glowing solid particle that lands on HEPA
filter media will perforate it if it continues to burn. Explosions that could destroy or seriously damage the
filter from high pressure, shock waves, or an excessive temperature excursion can also occur from ignition of
organic or pyrophoric dusts, vaporized organics, or combustible gas products of combustion. The spark and
flame arresters installed upstream of the filters are designed to alleviate this problem. Spark arresters
constructed of coarse glass fibers provide reasonable protection at low cost. Spark and flame arresters
constructed of grids or heavy wire mesh that provide graduated openings are required to provide a 2-minute
delay before flame penetration.

The recommended limitation for filter operating temperature is 250 degrees Fahrenheit.19 The filter media
binder is assumed to be the HEPA filter component that is most susceptible to failure resulting from elevated
temperature. The binder begins burning off at 350 degrees Fahrenheit.

Commonly used sealants are also highly susceptible to elevated temperatures. Tables 3.5 and 3.6 list
continuous-service temperatures for wood- and steel-cased filters. At temperatures well below the char point
of an elastomeric sealant, the sealant loses its shear strength, resulting in a reduction from approximately
6,000 kPa at room temperature to a low of 100 kPa at 300 degrees Fahrenheit. HEPA filters exposed to
thermal stress will begin to release contaminates at temperatures above 300 degrees Fahrenheit.

Table 3.5 – Recommended Limited Service Temperatures for Steel-Framed Fire-Resistant HEPA
Filter Units Sealed with Elastomeric Adhesives
Temperature to Which Filter was Exposed (degrees Fahrenheit)
Sealant Used Up to 10 Min a Up to 2 Hours Up to 48 Hour Up to 10 Days > 10 Years
HT-30-FR b 750 350 325 300 260
Z-743 c 750 325 300 275 200
EC-2155 d 750 250 220 200 200
Polyurethane foam 750 325 300 275 230
a Some reduction in efficiency may occur after 5 minutes of exposure.
b Goodyear.
c Pittburgh Plate Glass.
d Minnesota Mining and Manufacturing (3M).

Table 3.6 – Recommended Limited Service Temperatures for Wood-Framed Fire-Resistant

HEPA Filter Units a
Temperature to Which Filter was Exposed (degrees Fahrenheit)
Frame Material Up to 10 Min Up to 2 Hours Up to 48 Hours Up to 10 Days b > 10 Years b
3/4-inch-thick plywood a, c 750 300 275 200 180
a Subject to sealant limitations given in Table 3.5.
b Maximum temperature of 120 degrees Fahrenheit where relative humidity is 75 percent or higher.
c Exterior grade, fire-retardant-treated.

3.3.6.5 Moisture and Corrosion Resistance

Moisture

Water exposure is unquestionably an important factor leading to the deterioration of HEPA filters and their
degradation to 0 percent efficiency when coupled with higher pressure drop. HEPA filters become weak and
plug with water. One of the most common events is when people think no detrimental effects occur as a
result of repeatedly wetting the filter and drying it. Tests have shown that repeat wetting and drying of a
HEPA filter will cause the loss of half its strength. There also are very strong effects of operational time on
the behavior of HEPA filters under wet conditions. Tests have shown that the binder starts to get soft and

3-17
Nuclear Air Cleaning Handbook U.S. Department of Energy

dissolves at high differential pressures. One of the most serious issues dealing with HEPA filters in DOE
facilities is their potential for rupture during accidental fires and the resulting release of radioactive smoke.
The water spray systems in the HEPA filter housings used in nearly all DOE facilities for protection against
fires were designed under the assumption that the HEPA filters would not be damaged by the water spray.
The most likely scenario for filter damage in these systems involves filter plugging by the water spray,
followed by fan blowing out of the medium.

Water repellency is important for units that are used in laboratory and industrial applications. Repellency is
measured by the height of a water column that does not leak through the paper. A water repellency of
20 in.wg is required for filters that are operated in high-humidity conditions and stream-containing
atmospheres. In the absence of adequate water repellency characteristics, liquid contaminants that collect on
the filter paper can be carried through it by air pressure or capillary action and become re-entrained into the
downstream air.

Humidity

Numerous German studies from the Nuclear Air Cleaning Conferences during the 1970s and 1980s showed
that high humidity can result in high pressure drop and a corresponding decrease in media strength, the
combination of which can lead to structural damage and a loss of filter efficiency. These tests showed the
most frequent failure mode is rupture of the downstream pleat. With particle deposits, the filter would
absorb water at a lower relative humidity (RH) and would rupture even with a demister installed to protect
the filter. The tests further showed that filter failure under the humid air condition occurred at differential
pressures that were one-third to one-fourth the comparable values for filter failure under dry conditions. The
tests also showed that the tensile strength of a new filter is reduced by a factor of three due to humidity
exposure.

Previous studies have shown serious problems exist with HEPA filter wetting 22, 26, 27, 28 (Bergman, Fretthold).
HEPA filters exposed to wetting or high humidity must be removed from service before an accident can
happen because the strength of the filter may be seriously compromised (see Appendix C).

Corrosion

For many industrial applications, a moisture- and chemical-resistant filter should be capable of withstanding
attack by acids, most gas-phase alkalis, and solvent droplets and vapors. However, fine glass fibers have poor
resistance to hydrogen fluoride (HF), only moderate resistance to other concentrated acids, and fair resistance
to water and milder chemical corrosive agents. On occasion, corrosive chemicals in the airstream will
condense on the filter medium, accelerating the attack on the finest fibers. Airstreams containing some
residual HF and droplets of liquid carryover after treatment by an alkali scrubber produce a severe attack on
the glass fiber filter medium.

In AEC-sponsored research to develop an HF-resistant filter medium, Johns-Manville Corporation

formulated a special glass fiber for the purpose. However, the high costs associated with the finished paper,
together with a high shot content, large fiber diameter, and production difficulties, resulted in only marginal
benefits and precluded the glass fiber’s adoption for industrial use. Media made from ceramic fibers
(a combination of silicon dioxide and aluminum oxide) were found to have higher HF resistance than glass,
but in this case as well, the fibers have not been produced with diameters small enough to provide the
required efficiency characteristics. A U.S. filter manufacturer has developed an HF-resistant, high-efficiency
glass fiber paper containing up to 7 percent of a temperature-resistant polyamide (nylon). Filter units
incorporating this medium were exposed to 2 to 3 parts per million (ppm) of HF and 100 ppm of nitric acid
in a humid atmosphere. The test results were considered successful, and the medium was incorporated into
filters used at a nuclear energy plant. The service life of the new filters was three to four times longer than

3-18
DOE-HDBK-1169-2003 Chapter 3

that of previously used filters that were manufactured with a glass-asbestos filter medium. The adoption of
plastic-coated separators has contributed significantly to extending the life of HEPA filters under corrosive
service conditions.

A wooden case is more resistant to chemical attack than is a steel case. Exterior-grade material should be
specified, however, because interior-grade plywood is unsuitable for outdoor filter operation or for
continuous interior operation in very humid (90 to 100 percent RH) environments at temperatures above
131 degrees Fahrenheit (55 degrees Celsius), particularly when operation and shutdown periods alternate and
the environment returns to room temperature. During cooling, moisture may condense on the surfaces of
the wooden case and infiltrate the structure, causing swelling of the elements and a separation between the
seal and frame. Most exterior-grade wood products employ a moisture-impermeable phenolic resin bonding
agent, while water-soluble urea-formaldehyde resins are used as bonding chemicals for interior-grade
products. Stainless steel is recommended when a metal frame is required. Mildew growth may occur on the
sealant and frame interface in high humidity while the filter is in storage, causing filter degradation.

Seepage of particles collected on HEPA filters never occurs unless the filter paper becomes thoroughly wet.
For this condition, different entrainment mechanisms are involved.

3.3.6.6 Radiation Resistance

Most applications for HEPA and ULPA filters in the electronics and other industries do not involve exposure
to high levels of ionizing radiation. However, post-accident cleanup by nuclear reactor containment systems
and some fuel reprocessing applications of facilities can involve exposure of filters to high levels of radiation.
One reactor accident scenario estimates an integrated beta-gamma dose to the engineered safety feature (ESF)
filters of 3.5 × 107 rads. This radiation level can result in a significant reduction in tensile strength, an increase
in penetration, and an impairment of water repellency. Tests of commercial HEPA filter media before and
after radiation exposures up to a level of 4.5 × 107 rads were made at the Savannah River Site. The filter
papers were tested at a face velocity of 28.2 feet per minute, which is more than five times the design service
velocity and greater than any velocity anticipated under post-accident conditions. Test results showed up to
64 percent loss of strength and penetration increases of 4 to 50 percent. When samples were tested for
degradation of water repellency as a function of gamma dose, half of the samples showed hydrophilic action
in less than 10 seconds and the remainder in 60 to 100 seconds. The current code, ASME AG-12, calls for
filter papers to support a 6-inch column of water after exposure to an integrated gamma dose of 6.0 to
6.5 × 107 rads. Other tests exposed small HEPA filters to a range of radiation doses, and then exposed them
to a flowing steam-air mixture to determine the residual resistance to plugging and rupture. Plugging was
found to be inversely proportional to radiation dose (e.g., filters exposed to 6 × 108 rads ruptured in
100 seconds) but a sample irradiated to only 1 × 108 rads withstood the steam-air mixture for 250 seconds
before failure. Despite some blinding (water vapor interference with particulate capture), unirradiated
samples did not rupture under the same flow regimen. These tests verified the need to provide filter systems
with reliable protection from wetting wherever exposure to spray or condensing steam is possible, particularly
when water exposure may be coupled with high levels of radiation.

3.3.7 HEPA Filter Performance Testing for Nuclear Service

HEPA filters for nuclear service undergo a qualification procedure and two testing regimens. The first
regimen consists of a stringent visual examination and penetration tests at the place of manufacture. The
second regimen is an in-place leak test performed at the place of utilization. DOE requires independent
inspection and penetration tests at the designated DOE FTF prior to installation at its final destination. [For
a detailed discussion of qualification procedures, see Section 8.2, “Proof of Design – HEPA Filter

3-19
Nuclear Air Cleaning Handbook U.S. Department of Energy

Qualification for Nuclear Service.”] The state of DOE testing and the test facility are discussed in DNFSB
Tech-23.29

The manufacturer’s testing regimen involves two distinct phases: (1) a quality control routine to ensure
careful manufacture of the product, and (2) a series of tests to verify filter compliance with standards and
performance criteria related to collection efficiency and resistance to airflow. When all factors are within the
tolerance limits set by applicable specifications, the manufacturer certifies that each filter unit meets the
specification acceptance criteria.30

In addition, DOE mandates independent inspection and penetration testing for all filters purchased. Testing
is currently required for filters installed in hazard Category 1 and 2 facilities that perform a safety function,
and a statistical approach for the balance.31 The filters are tested for compliance with the requirements for
physical characteristics, efficiency, and airflow resistance. This testing is conducted at the DOE-supported
FTF before the filters are released to the customer’s facility. Filters failing to meet the FTF specification
acceptance criteria are rejected and turned over to the purchaser for disposition; typically, they are returned to
the manufacturer for credit. Both DOE and the NRC do not permit repairs of HEPA filters intended for
nuclear service.

3.3.7.1 Manufacturers’ Filter Qualification Test Protocols

Penetration (Efficiency)

For HEPA filters, particle removal is usually expressed as collection penetration (treated air concentration
÷ untreated air concentration × 100) or as penetration (100 - efficiency). Concentration may be expressed by
particle count per unit air volume (emphasizing the smallest particles present), particle weight per unit air
volume (emphasizing the largest particles present), ionizing radiation intensity per unit volume of air (particle
size effect indeterminate), or by light-scattering intensity per unit air volume (emphasizing small particle
sizes). Sometimes filter penetration is expressed as a decontamination factor (DF), the ratio of the untreated
air concentration to the treated air concentration, (e.g., a 99 percent collection efficiency is the same as a DF
of 100, and is equal to a penetration of 1 percent). The DF descriptor is most frequently used when ionizing
radiation is the concentration descriptor.

Airflow Resistance

The resistance of a filter to airflow, often expressed as “pressure drop” and “back pressure,” is almost always
measured as the height of a water column that exerts an equal pressure. This practice probably was borrowed
from hydrology, where the unit has a more direct relationship, as well as the use of water-filled manometers
to measure air filter resistance. The characteristic flow regime through HEPA filter media is aerodynamically
described as laminar. For this reason, the airflow resistance of these filters changes in direct proportion to
changes in air volume throughput (expressed as feet per unit area), even though the air approaching the filter
may be turbulent. The direct proportionality of resistance to flow rate is not a characteristic of prefilters. For
prefilters, resistance is a power function of airflow rate with an exponent larger than 1, but not exceeding 2.

The test protocols used to qualify HEPA filters for nuclear service are described below. Testing of all new
filters intended for nuclear service in the United States is conducted with a 0.3-µm test aerosol in a rig called a
Q107 penetrometer that was designed by the U.S. Army Chemical Corps during the 1950s. Construction and
operation are described in MIL-STD-282, Method 102.9.5 The complete penetrometer consists of test
aerosol generator, an instrument that measures the size and uniformity of the particles formed, a clamping
device to seal the filter under test into the test rig, a total scattering photometer to measure test aerosol
penetration, and a manometer to measure filter resistance at rated airflow rate.

3-20
DOE-HDBK-1169-2003 Chapter 3

The Q107 penetrometer, used for filters of 1,700 m3/hr rated capacity, exceeds 40 feet in length
(Figure 3.12). The Q76 penetrometer, which tests smaller filters and is based on the same principle of
operation, is considerably smaller. When testing a 1,700 m3/hr filter, about 2,400 m3/hr of outside air is
drawn into the system and divided into 3 parallel ducts that carry approximately 170, 500, and 1,350 m3/hr,
respectively. The remainder, approximately 350 m3/hr, is exhausted through another path. The 170 m3/hr
duct contains electric heaters that raise the temperature of the air to 374 degrees Fahrenheit (190 degrees
Celsius). Other electric heaters keep the liquid test aerosol reservoir heated to approximately 392 degrees
Fahrenheit (200 degrees Celsius). The test aerosol is vaporized from the reservoir into the heated airstream as
it sweeps across the liquid surface and is mixed with the air in the 500 m3/hr duct that contains both cooling
units and reheaters to provide partial dilution and temperature control of the test aerosol vapor stream. The
temperature of the test aerosol liquid reservoir establishes the mass concentration of the aerosol; a liquid
temperature of 392 degrees Fahrenheit (200 degrees Celsius) produces 80 to 100 µg/L of test aerosol when
diluted with 2,400 m3/hr of air. The particle size of the aerosol is determined by the temperature differential
between the evaporated test aerosol vapor stream and the much cooler diluting stream—the greater the
temperature differential, the smaller the resulting particle size. Temperature fluctuations in both airstreams
influence particle size distribution; the greater the fluctuation, the wider the size distribution. The combined
flows from the 170- and 500-m3/hr ducts are diluted further with the air in the 1,350-m3/hr duct to produce
the final aerosol concentration used for filter testing. Baffles are placed upstream and downstream to help
mix the aerosol entering and leaving the filter being tested.

Figure 3.12 – Q107 Penetrometer

The test aerosol particle size is determined by passing a sample through an optical particle-sizing instrument
called an OWL32 and noting the degree of polarization of a light beam. A polarization angle of 29 degrees
indicates a particle diameter of 0.3 µm when the aerosol is monodisperse.33

The optical device used to measure particle concentration is a forward-angle, light-scattering photometer
capable of measuring scattering intensity over a range of at least five orders of magnitude. Current
commercial instruments can give a useful signal with a concentration as low as 10 particles/cm3 when finely
tuned and used by a skilled operator. For routine testing, a downstream concentration of 10-4 mg/m3 can be
measured with reliability when the upstream concentration is 10 mg/m3, indicating a filter efficiency of
99.99 percent for the test aerosol. This level of measurement is considered adequate for nuclear applications
(in view of the lesser efficiency credit regularly assigned to filters by regulatory authorities), however,
manufacturers of microelectronic chips have sought filters with much higher retention efficiency.

ULPA filters have an efficiency of 99.9995 percent for particles in the 0.1-µm range, which is the minimum
filterable particle size for currently manufactured HEPA filters operating at their design airflow rate. This

3-21
Nuclear Air Cleaning Handbook U.S. Department of Energy

degree of efficiency is beyond the range of the Q107, but a laser spectrometer has been developed that can
measure filter performance at much higher efficiencies and for smaller particle diameters. This device
measures the sizes of individual particles in an aerosol and displays the particle-size distribution on a screen
and a printout. When used with a polydisperse aerosol challenge, it can measure penetration values as low as
1 × 10-9 in a range of particle diameters from 0.07 to 3.0 µm. Use of duplicate instruments upstream and
downstream permits the determination of a “particle size-collection efficiency” table or chart for individual
filters at a modest cost and within a reasonable period of time. Laser spectrometers can also be used to
determine such important filter performance parameters as maximum penetrating size, efficiency of filters in
series, and the optimum formulation of filter fibers. The laser spectrometer has been used experimentally for
in-place filter testing, but an inability to detect and isolate small leaks in a filter bank at low upstream aerosol
concentrations is unresolved. [Note: Lasers are currently being used routinely for high-efficiency filters
(HEPA and ULPA) with acceptable results. Operator training is still an important issue, as is recognition that
most lasers are calibrated using polystyrene latex (PSL) rather than the test aerosol. The properties of PSL
(e.g., refractive index) are not identical to the test aerosol. This can produce inaccurate results unless
operators understand the differences and set up the equipment properly. Upstream concentration is also
critical because lasers can be blinded by the passage of too many particles to the counter. Most successful
applications use calibrated particle diluters to ensure the laser is not overwhelmed.]

An international sampling of laser use for filter efficiency testing was conducted in 1985 by the Institute of
Environmental Sciences and Technology (IEST) Working Group RP7 (IEST-RP-CC007.1)34. Samples of
14 different high-efficiency filter media were sent to interested parties with recommended protocols for
instrument calibration and test performance. Results from eight participants showed wide variation in
particle size efficiency results for identical filter papers. Incorrect calibration of laser spectrometers and
incomplete knowledge of laser operation were contributing factors.

Based on the 1985 IEST findings, standards-writing groups organized at DOE since 1980 have established
rigid procedures for spectrometer calibration and use for filter testing. The operating policy of DOE’s filter
testing program, contained in DOE-STD-3022-989, calls for testing of all HEPA filters intended for
environmental protection at a DOE-operated FTF. Delivery of filters to a test facility for quality assurance
review is mandatory for all DOE facilities, and the service is also available to the public for a fee. When the
filter manufacturer’s test data are confirmed, the FTF test results are added to the information on the filter
case. The test procedures at the FTFs call for “penetration and resistance tests…visual inspection for
damage and visible defects…[and other]…visually verifiable requirements.” Except for the smallest filter
sizes, penetration tests are required to be conducted at 100 percent and 20 percent of rated airflow capacity,
and the maximum penetration of 0.1- to 0.2-µm particles at both airflow rates is 0.03 percent, in accordance
with draft DOE-STD-3025-99, Quality Assurance Testing of HEPA Filters.10 Penetration tests may be
conducted using a monodispersed aerosol and a total light-scattering photometer, or a polydisperse aerosol
with a single particle counting and sizing instrument.10 A quality assurance program for DOE’s FTFs is
contained in draft DOE-STD-3026-99,11 and specifications for HEPA filters to be used by DOE contractors
are contained in draft DOE Standard DOE-STD-3020-97.6 The HEPA filter specifications in
DOE-STD-3020-97 are the same as those in the previously cited military specifications, except that the size
and size distribution of monodispersed aerosols, when measured by the OWL, must be verified by a single
particle counter.6

3.3.7.2 Quality Control/Assurance Considerations

Systematic quality control and quality assurance testing are conducted at all stages of the product cycle from
development to use. The filter medium receives the most rigorous and extensive control and evaluation,
perhaps because its development and manufacture necessarily demand a degree of art as well as science.
Performance of the filtration medium is determined by a thermally generated monodispersed aerosol
generated by a Q127 penetrometer,35 a smaller version of the Q107 used to test cased filters. The physical

3-22
DOE-HDBK-1169-2003 Chapter 3

characteristics of the medium are controlled by a battery of standard test protocols developed by the TAPPI,
ASTM, and ASME AG-1.2 The use of ASME AG-1 requires an ASME NQA-136 program. After
fabrication, in addition to measuring the efficiency and airflow resistance of the filter assembly with a Q107
or a Q76 penetrometer (depending on the rated airflow capacity and physical size of the filter), a series of
physical tests described in ASME AG-1, Section FC,2 are applied to filter prototypes for qualification. These
include tests of dimension tolerances and resistance to rough handling, pressure, heated air, flame, and
unfavorable environments (simulated desert, tropical, and Arctic conditions).

Filter Test Facilities were established in the early 1960’s (see Chapter 1, Section 1.1.8). The last remaining
FTF is at Oak Ridge, Tennessee, and continues to inspect and test HEPA filters destined for safety class or
safety significant service at DOE facilities. The FTF continues to routinely find problems with HEPA filters
sent by the various manufacturers. Problem HEPAs are returned to the manufacturer at no cost to DOE.
Problems encountered occur in two categories: (1) flow/resistance/penetration amounting to approximately
1 to 2 percent per year, and (2) obvious defects in workmanship (which do not get flow tested) such as
splinters, protruding nails, improper gaskets, etc.) amounting to an additional 2 to 3 percent per year. There
have been major spikes (up to 20 percent) when a media making or packaging process was changed. The FTF
serves its function well.

3.3.7.3 Other Historical Methods of Testing New HEPA Filters

Nebulized Paraffin Oil

In Germany, new HEPA filters are tested according to German Standard, DIN 24-18437 The aerosol used is
generated from a distillate oil fraction (paraffin oil) with a viscosity of 3 to 3.8 × 10-5 m2/sec by heating the oil
to 212 degrees Fahrenheit (100 degrees Celsius) and nebulizing it with compressed air. The oil mist
concentration is about 10 mg/m3, with a droplet size median diameter of 0.36 µm and a geometric standard
deviation of about 2.0. A 45-degree angle, light-scattering aerosol photometer is used to measure the light-
scattering concentration of the aerosol entering and leaving the filter undergoing a penetration test. The
DIN 24-18420 test method differs in details, but is very close in principle to the U.S. test method.

Nebulized Sodium Chloride

The standard test method used in Great Britain for new HEPA filters38 utilizes a dried sodium chloride
aerosol generated from solution with a compressed air nebulizer. An emission-flame photometer is used to
measure the quantity of sodium chloride entering and leaving the filter being tested. The dried aerosol
particles have a concentration of about 3 mg/m3, a mass median diameter of 0.65 µm and a geometric
standard deviation of 2.1. The test rig and test procedures employed do not differ significantly from those
used in the United States, Germany, and a number of other countries.

Nebulized Uranine

The French standard test method, AFNOR NFX 44.011,39 uses dried particles of uranine, a fluorescent
material generated from a solution with a compressed air nebulizer. The aerosol concentration for the test is
about 8 × 10-3 mg/m3. The mass median diameter of the particles is 0.15 µm, with a geometric standard
deviation of 1.55.

Aerosol samples are extracted from the test apparatus upstream and downstream of the filter being tested and
are collected on filter papers. After the sampling period has expired, the filter papers are extracted in water
and analyzed by fluorimetry. Filter efficiency is expressed as the percent by weight of fluorescent particles
collected by the filter. Because of the need to collect samples over some averaging period (e.g., 10 minutes)
and then to extract the uranine quantitatively from the filters and read the fluorescence intensity in a

3-23
Nuclear Air Cleaning Handbook U.S. Department of Energy

fluorimeter, about 30 minutes is required for an analysis. Direct readout of filter efficiency is characteristic of
most other standard test procedures.

Interrelationships Between Test Methods

A number of comparative analyses have been conducted for the purpose of establishing ratios between the
several standard test methods, with indifferent results. This is understandable because different test methods
use different test aerosols, very different analytical processes, and are applied to filters that respond differently
to aerosols that have variable fractions of large and small particles. So, it is wise to view a filter’s ability to
pass the formal test protocols as simple assurance that the filter is constructed of quality components and was
assembled in a sufficiently careful manner to make it free of unacceptable defects. In short, passing any one
of the tests establishes that the filter is satisfactory for nuclear service—nothing more.

3.3.8 The Impacts of Aging, Wetting, and Environmental Upsets on HEPA Filter
Performance
Intuitively, the aging of filters in storage or in use inplace should lead to a higher probability of media or
structural failure. At least five experimental studies 22, 40, 41, 42, 43 have shown that with aging, HEPA filters lose
strength and water repellency but do not necessarily become less efficient. Logically, it follows that filter
efficiency depends on the physical geometry of the filter media, and is not significantly affected when the
organic binders and sealants become brittle or degrade with age. Filter strength prevents structural failure
during events that produce high stress across filter media, e.g., when particle deposits and water accumulation
cause filter plugging. Historical measures of filter strength are: (1) the tensile strength of the paper in
combination with a 10-inch overpressure test on the filter, and (2) burst strength. Burst strength (the
pressure required to tear open the media) quantitatively measures two-dimensional stretches as compared to
the one dimension used to measure the tensile strength. The brittleness of the media, which is measured by
flexing it, is a third major strength measurement, although it is not generally measured in aging studies.
Several authors have noted that aged HEPA filters are very brittle.

Decreasing water repellency produces filter plugging as accumulated moisture plugs filter media and decreases
tensile strength. Critical filter parameters such as media tensile strength and water repellency unfortunately
vary widely by manufacturer and types of particulate deposits. These varying parameters frequently mask the
effects of aging, often making it difficult to derive an age limit using the available experimental data.
M.W. First43 qualitatively described the deterioration mechanisms involved in HEPA filter aging as:

• Aging and weakening of glass fibers;

• Aeterioration of the resin binder and the organic sealant;

• Corrosion of the aluminum separators;

• Moisture damage; and

• Mechanical stresses caused by handling the filter and airflow pulses.

Johnson, et al.,41 were unable to measure the tensile strength across the media folds for aged HEPA filters
because the brittle media cracked; they also observed that the media had lost most of its water repellency.

Following issuance of the Defense Nuclear Facilities Safety Board’s Technical Report 23, HEPA Filters Used
in the Department of Energy’s Hazardous Facilities, 29 DOE initiated efforts to update ERDA 76-21, The Nuclear Air

3-24
DOE-HDBK-1169-2003 Chapter 3

Cleaning Handbook,44 to present new guidelines for root causes and factors that would dictate replacement of
HEPA filters within DOE nuclear facilities. However, as publication of this revision was delayed, increasing
risks identified with aging HEPA filters at many DOE sites required the development of interim criteria for
replacing safety-related HEPA filters to address wetting and environmental conditions, as well as aging
considerations.

Many of these issues have been reviewed throughout the DOE complex in response in part to the Defense
Nuclear Facilities Safety Board (DNFSB) Recommendation 2000-2, Configuration Management, Vital Safety
Systems.31, 45 DOE reviewed several facilities for their conformance to regulations, Orders, and standards
concerning confinement ventilation systems (CVS). These reviews identified both strengths and weaknesses
in the sites’ filter programs in the following areas: (1) independent quality assurance testing/inspection by the
FTF; (2) receiving inspection; (3) storage of HEPA filters; (4) in-place testing; (5) system bypass testing, and
(6) service life. They also identified the need for more periodic CVS reviews. These have typically been
woven into ongoing periodic assessments.

3.3.8.1 Aging

Bergman45 stated that, “a conservative interpretation of my experimental results indicates that the maximum
total life (storage and in-service) of HEPA filters for consistently removing greater than 0.9997 of 0.3 micron
particles from highly hazardous aerosols is 10 years from the date of manufacture for applications in dry
systems, and 5 years in applications where the filter can become wet more than once for short periods of
time.” If a filter gets wet it should be replaced expeditiously. At Oak Ridge National Laboratory (ORNL),
Lawrence Livermore National Laboratory (LLNL), Idaho National Engineering and Environmental
Laboratory (INEEL), and Savannah River Site (SRS) for “dry service” at normal relative humidity, the 10-year
criterion is applicable to HEPA filters for aging. The date of installation is available for most safety-related
HEPA filters. Historically, the date of manufacture has not been documented in a readily accessible manner,
but will be under the new Standards Based Management System (SBMS). Clearly, however, the date of
manufacture may not be retrievable for currently installed filters. If this information is available (without
having to remove the filter to retrieve the data on its frame), the filter service life will be determined based on
the date of manufacture. If the date of manufacture is not available, the date of installation will be used. If
neither is available, the filter will be assumed to be over 10 years old and subject to immediate replacement.

3.3.8.2 Wetting

In his experiments, Fretthold42 demonstrated that “previous water exposure weakened the filter media
irreversibly,” and that the “burst strength of the filter media decreased significantly with each wetting and
drying.” The replacement criteria will be exposure to a single occurrence of filter wetting. Potential sources
of filter wetting are entrained droplets from actuation of sprinklers in areas that are upstream of the airflow to
the filters, rain or groundwater inleakage into the filter system, or condensation from a leak of steam or hot
water.

3.3.8.3 Upset Environmental Conditions

Section 12.05 of the Lawrence Livermore National Laboratory Health and Safety Manual,47 High Efficiency
Particulate Air (HEPA) Filter System Design Guidelines for LLNL Applications, stated that continuous exposure to
the following operational environments will permanently damage or compromise HEPA filters:

• Moisture and Hot Air: 95 to 100 percent RH at temperatures higher than 130 degrees Fahrenheit.

• Fire: Direct fire or high concentrations of particulate matter produced by fire.

3-25
Nuclear Air Cleaning Handbook U.S. Department of Energy

• High Pressure: 6.0 in.wg or more, internal or differential across the filter media. Filters should be
changed if the differential pressure [adjusted for rated flow] exceeds 4.0 in.wg.

• Corrosive Mist: Dilute moist or moderately dry concentrations of acids and caustics.

• Shock Pressures: More than 1.7 psig.

The following criteria were modified for conservatism and simplification for use in an SBMS.

• Wetting: A single occurrence of filter exposure to water including entrained droplets from actuation
of sprinklers in the area upstream of the filters, rain or groundwater, or condensation from a leak of
steam or hot water.

• Moisture and Hot Air: HEPA filters may be operated continuously at 180 degrees Fahrenheit and
between 5 and 75 percent RH, or at 120 degrees Fahrenheit and between 75 and 95 percent RH.
HEPA filters are not to be used for installations where there is a possibility of condensation forming
on them. They will provide maximum service life when operated below 100 degrees Fahrenheit and
75 percent RH. Top
Start Gasket
• Fire: A single occurrence of direct flame
impingement. [Note: Filters subjected to
smoke from fires must have an in-place
leak test performed on them immediately
by the responsible in-place testing group
(i.e., within 24 hours) and must be
replaced if the filter fails the in-place leak
test.] Left Right

• High Differential Pressure: A single

occurrence of a differential pressure
across a single filter of 8.0 in.wg or more.

• Shock Pressure: A single exposure to

more than 1.7 psig.
Gasket

• Corrosive Mist: Prolonged exposure Finish

Not to Scale
Bottom
(more than 4 weeks) to dilute moist or
moderately dry concentrations of acids Figure 3.13 – Suggested Filter Probe Traverse
and caustics. Diagram

3.3.8.4 In-Place Testing of Filter Installations

An in-place leak test is done after filters are installed at a DOE nuclear facility to ensure the performance of
the confinement ventilation system. The in-place leak test is used both for an acceptance and for surveillance
leak testing of the installed HEPA filter bank. An in-place leak test and visual inspection of HEPA filters are
performed initially upon installation to detect bypasses and damage to filters and periodically to establish
current condition of a nuclear air cleaning system and its components. Specific objectives of in-place filter
testing are (1) to test the aggregate performance to filters in a filter bank, (2) to evaluate the effectiveness of
seals between the filter gasket and the filter housing, (3) to assess the leak-tightness of the filter housing, and
(4) to determine whether bypasses exist around the filter housing. Each time repairs are made, the system

3-26
DOE-HDBK-1169-2003 Chapter 3

must be retested until it meets the established criteria for leaktightness.48 Detailed information on in-place
filter testing is included in Chapter 8.

3.3.8.5 Packaging, Storage, and Handling of HEPA Filters

The manufacturer should have a quality program for the packaging, shipping, handling, and storage of HEPA
filters (e.g., NQA-1). HEPA filters are normally packaged in corrugated cardboard cartons that conform to
shipping regulations. Additional internal pieces are inserted to protect the filter faces from damage during
handling and transit. Palletizing crating should be constructed for ease of disassembly (see Figure 3.14). For
multiunit shipments, individual cartons should be crated and palletized to minimize handling, particularly at
trans-shipment points when using public carriers. For very large shipments, sealed and dedicated trailers are
recommended. [Note: Filters shipped in less-than-truckload amounts using common carriers are often
rearranged incorrectly by the carriers,
resulting in damaged filters.] Upon
delivery at the destination, mechanical
warehousing equipment should be used
for unloading and transferring the
shipment. Cartons should be placed in
clean, dry, interior storage until used.
They should be positioned as directed
on the carton exterior, and no more
than three filter cartons should be
stacked atop each other.

When a filter is inserted in the

cardboard shipping container, the
pleated folds should be oriented in the
vertical direction (except Type B filters),
and both the filter frame and the
enclosing carton should be labeled with
a vertical arrow or the notation, “This
Side Up” (including Type B filters).
Figure 3.14 – Filter Crating and Palletizing
When handling a filter inside a carton,
the box should be tilted on one corner,
picked up, and carried by supporting it
at diagonally opposing corners. Removing the filter from its shipping carton without damaging the medium
is best accomplished by opening and folding back the top flaps of the carton, inverting the carton onto a
clean surface, and lifting the carton off the filter. Then the filter unit can be grasped by the outer frame
surfaces without the danger of personnel coming into contact with the filter pack enclosed within the frame.
Additional details can be found in Appendix B.

3.4 Prefilters for HEPA Filters

3.4.1 Filter Descriptions

The service life of HEPA filters can often be extended by using less efficient filters that selectively remove
the largest particles and fibers from the incoming airstream. In some cases, HEPA filter lifetimes can be
increased by as much as four times with multiple prefilter changes during the interval between HEPA filter
changes. It is recommended that HEPA filters be protected from: (1) particles larger than 2 µm in diameter,
(2) lint, and (3) particle concentrations greater than 2.3 mg/m3. Selection of an appropriate prefilter includes

3-27
Nuclear Air Cleaning Handbook U.S. Department of Energy

consideration of: (1) the rapidity of filter resistance buildup and associated energy costs, (2) the size and
complexity of the resulting filtration system, (3) the fact that replacement filters and associated costs generally
increase with increasing prefilter efficiency, and (4) the disposal costs for contaminated HEPA filters and
potentially uncontaminated prefilters. It has been estimated that, with frequent prefilter replacements, savings
in filter system operation could be as much as one-third the cost of operating without prefilters. Assessment
of an acceptable combination of prefilters and HEPA filters depends on the dust-loading and efficiency
characteristics of the different filter types available for the particular aerosol to be filtered. The clogging
susceptibility of HEPA filters will vary with the dust and filtration characteristics of the prefilters.

The types of filters used as prefilters are also widely used for cleaning ventilation supply air in conventional
HVAC systems. The important advantage of filtering ventilation supply air for many operations that generate
radioactive particles is a reduction in the dust load that reaches the final contaminated filters. This helps
extend the service life of the exhaust filters, thereby reducing overall system costs because the supply air
filters can be changed without resorting to radiation protection measures—often the most costly aspect of a
contaminated exhaust filter change. These filters have a wide range of efficiencies, including 5 to 10 percent
for warm air residential heating systems; 35 to 45 percent for ventilation of schools, stores, and restaurants;
and 85 to 95 percent for fully air-conditioned modern hotels, hospitals, and office towers.

3.4.2 Classes, Sizes, and Performance Characteristics of Prefilters

For prefilters intended to remove only the largest airborne particles, a reverse relationship between retention
and re-entrainment forces occurs, causing collected particles to seep through the filter under prolonged
airflow unless the filter fibers are coated with viscous liquids to wet the collected particles and increase the
area of contact between them and the filter surfaces.

The most widely used test methods for ventilation air filters are published by the American Society of
Heating, Refrigerating, and Air Conditioning Engineers (ASHRAE) as Standard 52.1-92,49 which contains
two different protocols. One uses a prepared “test dust” consisting of road dust, carbon black, and cotton
fibers. In this procedure, the test dust is aerosolized by compressed air and blown into the filter at a
concentration many times that normally found in ambient air. The filter is rated by the weight percent of
dust retained. This obsolete test method originated in the days when coal was the only fuel and has little
relevance to today’s air filter requirements. The second test method uses unaltered atmospheric air as the test
medium and rates filter efficiency on the basis of the percent reduction in discoloration of simultaneous
samples taken on white filter papers upstream and downstream of the filter being tested. Reductions in
discoloration cannot be related to weight percent efficiency. In addition to dust-collecting efficiency, the first
test procedure measures filter resistance increase with dust deposition and dust-holding capacity. Ventilation
filters in the 35 to 95 percent efficiency range are evaluated by the atmospheric dust discoloration test.

Table 3.7 (from ASHRAE 52.2)50 shows cross-reference and application guidelines for air cleaners with
particulate contaminants. For comparison purposes, the HEPA filter is rated at 100 percent for both the
stain-efficiency and artificial dust arrestance tests. Because the atmospheric dust test is based on the staining
capacity of the dust that penetrates the filter, compared to the staining capacity of the entering dust, it is not a
true measure of particle-removal efficiency for any one particle-size range.

3-28
DOE-HDBK-1169-2003 Chapter 3

Table 3.7 – Cross-reference/Application Guidelines for Air Cleaners with Particulate Contaminants
Std. 52.2 Approximate Std. 52.1
Minimum Results Application Guidelines
Efficiency
Reporting
Value Duct Spot Typical Controlled Typical Applications Typical Air Filter/Cleaner
(MERV) Efficiency Arrestance Contaminant and Limitations Type
≤0.30 µm Particle Size HEPA/ULPA Filters
20 n/a n/a Virus (unattached) Cleanrooms ≥99.999% efficiency on 0.1-
Radioactive materials 0.2 µm particles, IEST Type F
19 n/a n/a Carbon dust Pharmaceutical ≥99.999% efficiency on 0.3
Sea salt manufacturing µm particles, IEST Type D
18 n/a n/a All combustion smoke Carcinogenic materials ≥99.99% efficiency on 0.3 µm
particles, IEST Type C
17 n/a n/a Radon progeny Orthopedic surgery ≥99.97% efficiency on 0.3 µm
particles, IEST Type A
16 n/a n/a 0.3-1.0 µm Particle Size Hospital inpatient care Bag Filters: Nonsupported
All bacteria General surgery (flexible) microfine fiberglass
15 >95% n/a Most tobacco smoke Smoking lounges or synthetic media, 12 to
Droplet nuclei (sneeze) Superior commercial 36 inches deep, 6 to
14 90-95% >98% Cooking oil buildings 12 pockets
Most smoke Box Filters: Rigid style
13 80-90% >98% Insecticide dust cartridge filters 6 to 12 inches
Copier toner deep may use lofted (air laid)
Most face powder or paper (wet laid) media.
Most paint pigments
12 70-75% >95% 1.0-3.0 µm Particle Size Superior residential Bag Filters: Nonsupported
Legionella Better commercial (flexible) microfine fiberglass
11 60-65% >95% Humidifier dust buildings or synthetic media, 12 to
Lead dust Hospital laboratories 36 inches deep, 6 to
10 50-55% >95% Milled flour 12 pockets.
Coal dust Box Filters: Rigid style
9 40-45% >90% Auto emissions cartridge filters 6 to 12 inches
Nebulizer drops deep may use lofted (air laid)
Welding fumes or paper (wet laid) media.
8 30-35% >90% 3.0-10.0 µm Particle Size Commercial buildings Pleated Filters: Disposable,
Mold Better residential extended surface, 1 to 5 in.
7 25-30% >90% Spores Industrial workplaces thick with cotton-polyester
Hair spray Paint booth inlet air blend media, cardboard
6 <20% 85-90% Fabric protector frame.
Dusting aids Cartridge Filters: Graded
5 <20% 80-85% Cement dust density viscous coated cube or
Pudding mix pocket filters, synthetic media
Snuff Throwaway: Disposable
Powdered milk synthetic media panel filters
4 <20% 75-80% >10.0 µm Particle Size Minimum filtration Throwaway: Disposable
Pollen Residential fiberglass or synthetic panel
3 <20% 70-75% Spanish moss Window air filters
Dust mites conditioners Washable: Aluminum mesh,
2 <20% 65-70% Sanding dust latex coated animal hair, or
Spray paint dust foam rubber panel filters
1 <20% <65% Textile fibers Electrostatic: Self charging
Carpet fibers (passive) woven
polycarbonate panel filter

3-29
Nuclear Air Cleaning Handbook U.S. Department of Energy

ASHRAE Standard 52.1-9249 tests have replaced those sanctioned formerly by the Air Filter Institute and the
Dill Dust-Spot Test of the National Institute for Standards and Technology. Care must be taken in the
interpretation of data from the ASHRAE tests. Arrestance test results depend highly on particles that exceed
1 µm in diameter, but the ambient atmospheric dust test results depend on the nature and concentration of
aerosol particles at the testing location. The average particle size of the urban atmosphere is assumed to be
0.5 µm. The results of the various tests are not comparable, and a filter determined to be efficient by one test
may be determined to be inefficient by another. Users should examine the test used to evaluate a filter’s
efficiency to properly understand the results. Efficiency tests are made on prototype filters, and the results
are extrapolated to other units of similar design (certification of every prefilter by testing would be too costly).

Values stated in Table 3.7 for dust-holding capacity were determined with resuspended synthetic dust
mixtures. Dust-holding capacity varies with the nature and composition of the particles (e.g., carbon black,
cotton linters). Dust-holding capacity under service conditions cannot be predicted accurately on the basis of
manufacturers’ data. Air resistance is the primary factor in prefilter replacement. Although manufacturers
recommend specific values of resistance for prefilter replacement, loss of adequate airflow is often a more
reliable indicator of system performance and is also more cost effective. Panel filters will plug rapidly under
heavy loads of lint and dust. An accumulation of surface lint may increase the efficiency of an extended-
medium filter by adding “cake” filtration principles to the existing physical mechanisms. The extended-
medium prefilter will plug readily in an airstream carrying profuse smoke and soot from a fire. Operation at
airflows below rated capacity will extend the service lives of filters and be more cost effective by reducing the
frequency of filter replacement. On the other hand, when airflow exceeds rated values, dust-loading rate and
system costs begin to increase exponentially along with proportional increases in airflow. [ASHRAE also
publishes Standard 52.2-99,50 which gives methods for testing filter efficiency by particle size using optical
particle counters, including lasers.]

3.4.3 Construction of Prefilters

Prefilters are classified by the American Refrigeration Institute (ARI) 850-9351 as follows:

• Group I - Unit or panel.

• Group II - Self-cleaning, self-renewable, or any combination thereof.

• Group III - Extended surface.

• Group IV - Electronic air cleaner.

• Group V - Air filter media.

Group I panel filters (viscous impingement filters) are shallow, tray-like assemblies of coarse fibers (glass,
wool, vegetable, or plastic) or metal mesh enclosed in a steel or cardboard casing. The medium is usually
coated with an inhibited viscous oil or adhesive to improve trapping and retention of particles. Single-use
disposable and cleanable-reusable types are available. The latter have metal mesh and generally are not used
in nuclear applications for effluent or process air cleaning because of the high labor costs associated with
cleaning and disposal of entrapped radioactive materials. A disposable panel filter has a fairly high dust-
holding capacity, low airflow resistance, low initial and operating costs, and high removal efficiency for large
particles. It is particularly effective against fibrous dust and heavy concentrations of visible particles, but is
ineffective for smaller particles. For nuclear service, it is less cost-effective than the more costly Group II or
III filters that provide better protection for the HEPA filter.

3-30
DOE-HDBK-1169-2003 Chapter 3

Group II (moderate-efficiency) and Group III (high-efficiency) filters are usually comprised of extended-
medium, dry-type, single-use disposable units. The filter medium is pleated or formed into bags or socks to
provide a large filter surface area with minimal face area. They are not coated with adhesive. The particle size
efficiency of Group II filters is moderate to poor for submicrometer-sized particles, but often approaches
100 percent for particles greater than 5 µm. In most cases, the pressure drop of extended-media Group II
filters varies directly with efficiency. Group II filters are recommended for high lint- and fiber-loading
applications. The large filter area relative to face area permits duct velocities equal to or higher than those of
panel filters.

Group III filters are preferred when higher efficiency for smaller particles is desired. The dust-holding
capacity of Group III filters usually is lower than that of Group II filters.

3.4.4 Electrostatic and Electrified Filters

An electrostatic charge may be induced on filter fibers by triboelectrification and by sandwiching the fiber
bed between a high voltage and a grounded electrode. Triboelectrification can be used to induce a high
electrostatic charge on suitable high dielectric materials, but under practical-use conditions, the charge is
subject to rapid dissipation due to air humidity, oily particles, fiber-binding particles, and other interference.
Continuously activated electrodes can induce a more permanent charge.

A program to develop electrofibrous filters, undertaken by DOE at LLNL, has proved them effective in
providing greater efficiency and longer service life for the prefilters used to protect HEPA filters. They have
been used in gloveboxes and for other applications. Laboratory tests using test and sodium chloride aerosols
have shown that an “electrofibrous prefilter increases in efficiency from 40 to 90 percent as 10 kV is applied
to the electrode.” A comparison of uncharged, triboelectrically charged, and permanently charged fibrous
filters demonstrated the higher collection efficiency of the permanently charged filter design for
submicrometer particles. When continuously charged electrofibrous filters were applied as prefilters for
HEPA filters in exhaust air systems or gloveboxes used to burn uranium turnings, they significantly
prolonged the life of the final filters.

3.4.5 Operation and Maintenance of Prefilters

All prefilter construction materials must be compatible with those of the downstream HEPA filters they are
designed to protect. Therefore, they must conform to the rigorous physical properties prescribed for HEPA
filters (e.g., resistance to shock, vibration, tornado, earthquake, moisture, corrosion, and fire). Survivability
under the specific operational conditions and requirements must be addressed when prefilters are selected
because moisture or corrosive products in the airstream may limit the choice of filter. Although many filter
media will not withstand acid or caustic attack, glass fibers are corrosion-resistant except for fluorides.
However, the casing and face screen materials may be less so. Aluminum may deteriorate in marine air, from
caustics, or from carbon dioxide. Plastics have poor heat and hot air resistance and generally will not satisfy
UL requirements. Condensation from high humidity and sensible water may plug a prefilter and result in
more frequent replacement. In general, a prefilter made of construction materials identical to those in the
HEPA filter will have equivalent corrosion and moisture resistance. Any increase in resistance from moisture
accumulation will be greater for MERV 17-20 filters than for MERV 9-16 filters (ASHRAE 52.2 Table E-1)50.
UL classifies ventilation air filters in two categories with respect to fire resistance.50 When clean, UL Class 1
filters do not contribute fuel when attacked by flame and emit a negligible quantity of smoke. UL Class 2
filters are permitted to contain some small amount of combustible material, but they must not contribute
significantly to a fire. The collected material on inservice UL-approved Class 1 and 2 filters may burn
vigorously and create a fire that is difficult to extinguish. Therefore, use of an UL-rated prefilter should not
lead to an unwarranted sense of security on the part of the user. The UL maintains a current listing of filters
that meet the requirements of their standards.21

3-31
Nuclear Air Cleaning Handbook U.S. Department of Energy

Most types of prefilters are suitable for continuous operation at temperatures not exceeding 149 to
248 degrees Fahrenheit (65 to 120 degrees Celsius). Other types with glass-fiber media in steel or mineral
board frames may be used at temperatures as high as 392 degrees Fahrenheit (200 degrees Celsius). Users of
high-temperature prefilters should take a conservative view of performance claims, particularly claims related
to efficiency at operating temperature.

Because of waste disposal requirements, the preferred choice of a prefilter for nuclear applications is the
single throwaway cartridge. A replaceable-medium filter offers an advantage over the throwaway because the
bulk of material that needs to be discarded is smaller and handling and disposal costs are minimized.
However, re-entrainment of contaminants and contamination of the peripheral area are possible because the
medium is removed from the system and prepared for disposal. The replaceable-medium type is not
recommended for toxic exhaust systems. The cleanable-medium filter is undesirable for nuclear systems
because of the extensive downtime of the system that is required for changing and decontaminating areas in
proximity to the filter installation.

3.5 Deep-Bed Filters

Deep-bed filters were designed, built, and placed in service early in the development of nuclear technology
for treating offgasses from chemical processing operations. The first, a sand filter, was constructed at the
Hanford, Washington, nuclear facility in 1948, and deep-bed glass fiber filters were constructed soon after.
These were not considered competitive with then-current versions of the HEPA filter (the CWS-Type 6 or
AEC-Type 1), but were thought to have a different function. With the thin-bed filters, the intent is usually to
replace or clean the filter medium periodically. The deep-bed filter, on the other hand, usually has as its
objective the installation of a unit which will have a long life, in the dust capacity sense, of say 5 to 20 years,
corresponding to either the life of the process or the mechanical life of the system. Thus, when resistance
starts increasing rapidly, instead of replacing or cleaning the filter medium, the entire filter installation would
be abandoned and replaced with a new unit. In fact, the life span of some deep-bed filters constructed during
the early 1950s has not yet been entirely expended. A partial explanation for this longevity is the original
design concept that deep-bed filters would be used where the total aerosol concentration was usually on the
order of or less than normal atmospheric dust concentrations. An important reason for selecting sand for the
initial bed material was a need to filter large volumes of wet corrosive aerosols for which more usual filter
materials would prove unsatisfactory. Deep beds of crushed coke had been used by the chemical
manufacturing industry for many years to remove sulfuric acid mist from the effluent gas of sulfuric acid
manufacturing plants prior to 1948. Silverman cited efficiencies as high as 99.9 percent by weight for a
crushed-coke bed against a sulfuric acid mist of 0.5 to 3.0 µm in diameter.52 Perhaps a carbon-filled bed was
considered unsuitable for filtering an aerosol that might contain fissile material, and sand was selected for the
first deep-bed filter for nuclear fuel processing facility ventilation air.

3.5.1 Deep-Bed Sand (DBS) Filters

Some of the following material is taken directly from ERDA 76-2144. Although dated, it is still relevant today.
It has been updated where appropriate. Initially, sand filters were installed at the Hanford, Washington,
nuclear facility and at the Savannah River nuclear plant. Following their success, more were added at
Hanford and Savannah River and others were constructed at plants in Morris, Illinois, and Idaho Falls, Idaho.
The Argonne National Laboratory compiled a bibliography of DBS filters. These DBS filters had collection
efficiencies for particles greater than or equal to 0.5 µm that compared favorably with the HEPA filters of
that era. Their advantages for the nuclear programs at these sites included large dust-holding capacity, low
maintenance, chemical resistance, high heat tolerance, fire resistance, and a capability to withstand large shock
and gross pressure changes without operational failures. They also had disadvantages such as high capital

3-32
DOE-HDBK-1169-2003 Chapter 3

costs, need for large areas and volumes, inability to maintain the granular fill, and lack of a reasonable means
of disposing of the contaminated fill.

DBS filters contain up to 10 feet of rock, gravel, and sand constructed in graded layers that diminish granule
size by a factor of 2 as the layers go from bottom to top. Airflow direction is upward so that granules
decrease in size in the direction of flow. A top layer of moderately coarse sand is generally added to prevent
fluidization of the finest sand layer underneath. The rock, gravel, and sand layers are positioned and sized to
provide the desired structural strength, particle collection ability, dirt-holding capacity, and long service life.
Ideally, the layers of the largest granules, through which the gas stream passes first, remove all the large
airborne particles, whereas the fine sand layers on top retain the finest smallest particles at high efficiency.
Below the granular bed there is a layer of hollow tile that forms passages for air distribution. The total bed is
enclosed in a concrete-lined pit. The superficial velocity is about 5-feet per minute, and pressure drop across
the seven layers, sized 3 1/2-inch average diameter down to 50 mesh, is from 7 to 11 in.wg. Collection
efficiencies as high as 99.98 percent for test aerosols have been reported. Some DBS filters have experienced
premature plugging at relatively low dust loadings. Another suffered partial collapse from disintegration of
grout between the tiles supporting the overhead filter structure. These failures were caused by moisture
leaking through voids in the system perimeter or by chemical corrosion and erosion of system components
from nitric acid fumes in the effluent air. Disposal of inoperable DBS filters, usually contaminated, is
generally accomplished by sealing and abandonment. Replacement systems normally are constructed nearby
to accommodate the same air intake duct system.

Currently, there is renewed interest in sand filters for ESF applications (e.g., the plutonium Pit Disassembly
and Conversion Facility in Savannah River, South Carolina; emergency confinement venting for light-water
reactors). The Swedish confinement venting system, known as FILTRA, features large concrete silos filled
with crushed rock. It is designed to condense and filter the stream blown from the confinement and to
release to the atmosphere less than 0.01 percent of the core inventory.

3.5.2 Deep-Bed Glass Fiber (DBGF) Filters

The rapidly emerging glass fiber technology of the late 1940s shifted attention to the use of very deep beds
(1 or more meters thick) of graded glass fibers as a satisfactory substitute for sand filters when treating
gaseous effluents from chemical operations. They proved to be more efficient, less costly, and to have a
lower airflow resistance than the DBS filters they replaced. In addition, these DBGF filters employ a medium
that has more controllable physical features and more assured availability than the DBS to permit a larger
airflow per unit volume at lower pressure drop, lower operating costs, and potentially lower spent-filter
disposal costs. DBGF filters have been used at Hanford for several decades on their Purex process effluent
streams. However, the DBGF filters do not have the corrosion resistance of the DBS, particularly from HF,
and are less fire-resistant. The DBGF is also less of a heat sink and has less capability to resist shock and
high-pressure transients.

The intake segment of the DBGF filter system was designed with layered beds of uniform-diameter glass
fibers to a total depth of 8 to 84 inches. Each layer in the direction of airflow was compressed to a higher
density and enclosed in a stainless steel tray with impermeable walls and a perforated screen above and below.
Capacity varied from 200 to 200,000 cfm (350 to 350,000 m3/hr). Although the first unit constructed at
Hanford was small (400 m3/hr (235.4 cfm), many of the 25 subsequent units were much larger and
experienced extensive usage from nuclear fuel processing to hot cell ventilation. The glass fiber of preference
for this application was Owens-Corning's 115-K, a 29-µm-diameter, curled glass fiber that resisted clumping,
settling, and matting. A system that was designed for downward airflow became inoperative from
precipitation of ammonium nitrate at the filter face. Subsequent units were designed with airflowing upward
and were equipped with water sprays directed from below to dissolve salt precipitation on the intake face to
reduce pressure drop buildup. The design airflow velocity of a typical DBGF was 50 feet per minute, and

3-33
Nuclear Air Cleaning Handbook U.S. Department of Energy

clean pressure drop was close to 1.5-in.wg. The final pressure drop, after a total particle loading estimated at
10,500 pounds, was 8-in.wg. The final stage of a second-generation DBGF filter system employed two
12-mm blankets of 3.2-µm- and 1.2-µm-diameter glass fibers fabricated as a twin-layer bag stretched over a
stainless steel framework. Airflow from the first stage passed through the filtration blankets from the outside
to the inside, then was exhausted from inside the metal framework. The number of bag filters was
proportional to the capacity of the intake segment of the DBGF filter. Later designs of the DBGF filter's
cleanup stage substituted HEPA filters in a group of manifolded caissons (encapsulating filter holders), and a
comparable increase in collection efficiency was realized. The most recent installation of a DBGF filter
system required more than 100 HEPA filters downstream of a deep bed containing more than 38,000 pounds
of 115-K fiber. By carefully selecting the packing density, bed depth, and airflow velocity, collection
efficiencies greater than 99 percent for 0.5 µm particles were attained.

Provision for periodic backflushing will often extend the life of the total filter. Most DBGF filter systems,
contained in vaults below ground, are resistant to shock and overpressure from natural phenomena. The
dust-holding capacities of DBGF filters are very large, and many units have operated for years without
attendance or maintenance. Pressure drop sensors can often predict evolving difficulties and indicate when it
is time for backflushing, precipitate dissolution, or other preplanned remedial actions. Just as for DBS filters,
decontamination and disposal is difficult for small systems and nearly impossible for the larger systems.

3.5.3 Deep-Bed Metal Filters

Deep beds of metal fibers have a number of applications in the nuclear industry, particularly where maximum
resistance to fires, explosions, and overpressure shocks are essential. In offgas systems containing substantial
concentrations of HF, use of stainless steel metal fibers has been studied as a substitute for glass.

In most cases, the objective when using metal fiber filters is to obtain particle collection efficiencies that
duplicate those obtainable with HEPA filters. However, the unavailability of metal fibers with diameters
close to or below 1 µm makes it necessary to provide great filter depth as a substitute for small fiber
collection efficiencies. For sodium fire aerosols, high collection efficiency can be obtained with relatively
large diameter metal fibers because the combustion products in air, sodium oxide, and carbonate rapidly form
large flocs that are easily filtered. The ease of filtration results in the extremely rapid formation of a high-
resistance filter cake that severely limits the amount of sodium aerosol particles that can accumulate in the
filter before the limit of the fan's suction pressure is reached. Here, the requirement is for a graded-efficiency,
deep-bed, metal filter with a large storage capacity in the initial layers of the filter for the fluffy sodium aerosol
particles, a high efficiency for small particles in most downstream layers of the filter, and the elimination of
abrupt interfaces between graded fiber layers where a filter cake might form. This is a different filtration
requirement than obtaining high efficiency for low concentrations of small, nonagglomerating particles—
instead, the requirement is for uniform particle storage throughout the depth of the filter. Here also, uniform
diameter fibers can be used in great depths, as in the DBGF filters, to substitute for the presence of very
small-diameter filter fibers.

Other types of metal filters have been constructed by sintering stainless steel powders or fine fibers into a
sieve-like structure that function very much like a conventional pulse-jet-cleaned industrial cloth filter. The
metal membrane has an inherent high efficiency for particles greater than a few micrometers, but depends on
the formation of a filter cake to obtain high efficiency with submicrometer particles. Clean airflow resistance
is high and increases rapidly as cake thickness builds up. It is cleaned periodically by backflow jets of
compressed air. Efficiencies are comparable with those of HEPA filters when the sintered metal filters are
precoated with filter aids. Because of their high-temperature resistance and ability to handle high
concentrations of mineral dusts, these types of filters have been used in nuclear incinerator offgas cleaning
systems, particularly when heat recovery from the hot filtered gases is desired. However, care must be
exercised to avoid releasing tar-like combustion products to sintered filters that are operated at high

3-34
DOE-HDBK-1169-2003 Chapter 3

temperatures because the tarry material tends to lodge in the pores and turn to cake that cannot be removed
by chemical means or by elevating the temperature to the limit of the metal structure.

Another type of sintered filter construction for high-temperature applications has been prepared from a
mixture of stainless steel and quartz fibers. The composite material has the same efficiency and pressure drop
as HEPA filter glass paper, but has 4 times the tensile strength and can operate continuously at temperatures
up to 932 degrees Fahrenheit (500 degrees Celsius). Applications of the stainless steel and quartz fiber HEPA
filter medium have not proceeded beyond the laboratory stage.

3.6 Demisters
Liquid droplet entrainment separators are required in the standby air treatment systems of many water-cooled
and -moderated power reactors to protect the HEPA filters and activated-charcoal adsorbers from excessive
water deposition should a major high-temperature water or stream release occur as a result of an incident
involving the core cooling system. Droplet entrainment separators are also used in fuel processing operations
to control acid mists generated during dissolving operations and subsequent separation steps.

Entrainment separators consisting of a series of bent plates are widely used in HVAC applications for
controlling water carryover from cooling coils and humidifiers; but for nuclear applications, their droplet
removal efficiency is inadequate. Therefore, fiber-constraining demisters with a much greater efficiency for
small droplets are standard for nuclear service. Entrainment separators utilizing fiber media remove droplets
by the same mechanisms that are effective for dry fibrous filters, but they must have the additional important
property of permitting the collected water to drain out of the cell before it becomes clogged. Should clogging
occur and the pore spaces fill with water, the pressure drop across the separator will rise and some of the
water retained in the pore spaces will be ejected from the air discharge side to create sufficient passages for air
to pass through. The ejected water can become airborne again by this mechanism.

Droplets from condensing vapors originate as submicrometer-sized aerosols, but the droplets may grow
rapidly to multimicrometer size by acting as condensation centers for additional cooling vapors and by
coagulation when the concentration of droplets exceeds 106 droplets/ml. Firefighting spray nozzles,
confinement sprays, and other devices that mechanically atomize liquid jets yield droplets that predominantly
range from 50 to more than 1,000 µm in diameter. This range means that entrainment separators must not
only be capable of removing the smallest droplets, but also must resist becoming flooded by the largest
droplets and releasing the collected liquid as entrained water.

The NRC recommends the use of entrainment separators for engineered safety systems when the air may be
carrying entrained liquid droplets or a cooling and condensing vapor. 8, 31, 45, 50 Although HEPA filter paper is
treated for water repellency, high-water loadings rapidly saturate the paper and raise its airflow resistance to a
point where gross holes can result. Hot water and steam cause paper to lose its strength and to fail even
more rapidly. Therefore, the criteria for entrainment separators used for nuclear service call for: (1) at least
99.9 percent retention by weight of entrained water and condensed steam in the size range 1 to 2,000 µm
diameter, at a duct velocity from 250 to 2,500 linear feet per minute, and water delivery rate of 8 gallons per
minute (gpm) per 1000 cfm of installed HEPA filter capacity; (2) at least 99 percent retention by count of
droplets in the 1- to 10-µm-diameter range, at a duct velocity from 250 to 2,500 linear feet per minute; (3) no
flooding or water re-entrainment at a water-steam delivery rate of 8 gpm at a duct velocity of 2,500 linear feet
per minute; and (4) a temperature tolerance at least to 320 degrees Fahrenheit (160 degrees Celsius) and
gamma radiation exposure up to 106 rads integrated dose without visible deterioration or embrittlement of the
materials of construction. An entrainment separator with these characteristics will provide long-term
protection for a downstream HEPA filter that would be destroyed in a few minutes without it. Entrainment
separators are usually constructed of deep layers of high-porosity metal and glass fibers, either packed or

3-35
Nuclear Air Cleaning Handbook U.S. Department of Energy

woven into stable batts, and arranged in graded sizes and packing density to give the desired small droplet
collection capability with excellent resistance to flooding and re-entrainment.

3.7 Filter Design Selection

Nuclear-grade HEPA filter papers are distinguished from otherwise identical products by their proven
resistance to deterioration by radiation. This requirement is spelled out in ASME AG-1,2 which calls for
50 percent retention of original strength and water repellency after exposure to an integrated dose of
6.0 × 107 to 6.5 × 107 rads at a dosage rate not to exceed 2.5 × 106 rads per hour. Because all fabricated filters
destined for nuclear service will contain identical or equivalent paper, selection can be based solely on the
type of filter construction.

Deep-pleat filters with corrugated aluminum separators have dominated nuclear service both by numbers and
years of use, and therefore have the longest and most thoroughly documented performance record. They
appear to be stronger than other filter designs, although mini-pleat and separatorless filters are able to meet
existing strength requirements in applicable filter standards. Mini-pleat construction has the desirable
advantage of packing twice as much paper into a given volume of filter. A disadvantage of the mini-pleat
design is the narrowness of the air passages between adjacent pleats, which make it susceptible to premature
clogging of the openings by large particles and fibers. This may not be a difficulty when the air being filtered
is exceptionally dust-free or when efficient prefilters are employed. Nuclear service experience is sparse or
totally lacking for types of filter construction other than deep-pleat filters with corrugated separators,
although there may be equivalent experience in nonnuclear applications.

Special nuclear filters are needed when service conditions involve exceptional physical or chemical stress.
Although the usual run of filters for nuclear service must provide resistance to short-term exposure to heated
air and flame, they are not designed for long-term operation at temperatures exceeding 250 degrees
Fahrenheit (120 degrees Celsius). Because the organic sealant between filter pack and filter frame is the least
temperature-resistant component, it is possible to increase temperature resistance by substituting a tightly
compressed fine-fiber batt for the organic adhesive. In addition, substituting a metal frame for a plywood or
composition board increases temperature resistance to the melting point of the glass fibers in the filter
medium [932 degrees Fahrenheit (500 degrees Celsius)]. Before this temperature is reached, the organic
binder and water-repellent chemicals in the paper will be lost, but this may not adversely affect filtration
efficiency or airflow resistance, but does reduce the filter strength.

The chemical resistance of low-temperature nuclear filters is generally excellent for all dry gases. With high
humidity, the presence of HF will cause etching and embrittlement of the glass fibers and ultimate failure of
the filter. When droplets of HF or condensed water plus HF gas are present in the airstream, rapid failure of
the glass filter paper may be anticipated. Rapid failure (within hrs) also occurs when hydroscopic salts from
chemical processing collect on the filter surface and form a moist, slush-like cake that absorbs HF and
infiltrates the pores of the filter paper. Special filter papers have been formulated with 7 percent Nomex
fibers to provide extra chemical resistance for this type of service.

Aluminum separators are especially susceptible to chemical attack by many substances other than HF. United
States requirements call for vinyl-epoxy coatings of 0.2 to 0.3 µm in thickness on both the sides and edges of
aluminum separators when the presence of acid is predicted. Stainless steel separators are a more costly
alternative.

Deep-bed filters of sand, gravel, and crushed stone do not compete directly with HEPA filters, except at a
few installations involved in chemical operations associated with fuel reprocessing, but they have recently
come under intense study as a means of mitigating core meltdown events by providing a filtration capacity for
venting confinement vessel overpressures and for coping with a possible hydrogen burn inside the

3-36
DOE-HDBK-1169-2003 Chapter 3

confinement. DBS filters have also been studied extensively for a potential role in mitigating loss of coolant
accidents for metal-cooled reactors.

3-37
Nuclear Air Cleaning Handbook U.S. Department of Energy

3.8 References
1. ASME (American Society of Mechanical Engineers), 1989, Nuclear Power Plant Air Cleaning Units and
Components, ASME N509, New York, NY.

2. ASME (American Society of Mechanical Engineers), 2003, Code on Nuclear Air and Gas Treatment, ASME
AG-1, New York, NY.

3. DoD (U.S. Department of Defense), 1986, Filters, Particulate, High-Efficiency, Fire-Resistant, U.S. Military
Specification MIL-F-51068F, Aberdeen Proving Ground, MD, June.

4. DoD (U.S. Department of Defense), 1963, Filter Medium, Fire-Resistant, High-Efficiency, Military
Specification MIL-F-0051079, U.S. Army Armament, Munitions and Chemical Commands, Aberdeen
Proving Ground, MD.

5. DoD (U.S. Department of Defense), 1995, Filter Units, Protective Clothing, Gas Mask Components and Related
Products: Performance Test Methods, U.S. Military Specification MIL-STD-282(4), Edgewood Biological
Center, MD, January 12.

6. DOE (Department of Energy), 1997, Specification for HEPA Filters Used by DOE Contractors,
DOE-STD-3020, Washington, DC.

7. ASME (American Society of Mechanical Engineers), 1989, Testing of Nuclear Air-Cleaning Systems, ASME
N510, New York, NY.

8. NRC (U.S. Nuclear Regulatory Commission), 1976, Design, Testing, and Maintenance Criteria for Engineered
-Safety-Feature Atmospheric Cleanup System Air Filtration and Adsorption Units of Light-Water-Cooled Nuclear Power
Plants, NRC Regulatory Guide 1.52 (Rev. 1), Washington, DC.

9. DOE (U.S. Department of Energy), 1998, DOE HEPA Filter Test Program, DOE-STD-3022,
Washington, DC.

10. DOE (U.S. Department of Energy), 1999, Quality Assurance Inspection and Testing of HEPA Filters,
DOE-STD-3025, Washington, DC.

11. DOE (U.S. Department of Energy), 1999, Filter Test Facility Quality Program Plan, DOE-STD-3026-99,
Washington, DC.

12. ANSI/ASQC (American National Standards Institute/American Society for Quality Control), 1993,
Sampling Procedures and Tables for Inspection by Attributes, ASQC-Z 1.9; DoD (Department of Defense)
Sampling Procedures and Tables for Inspection by Attributes, MIL-STD-105E, May 10, 1989, superceded by
ANSI/ASQC Z1.9-1993.

13. ASTM (American Society for Testing and Materials), 2002, Standard Specification for Aluminum and
Aluminum Alloy Sheet and Plate, ASTM B209, West Conshohocken, PA.

14. Adley, F. E., Progress Report, 1996, “Factors Influencing High Efficiency Gasket Leakage,” 9th Air
Cleaning Conference, Atomic Energy Commission Report CONF-660904, National Technical
Information Service, Springfield, VA.

3-38
DOE-HDBK-1169-2003 Chapter 3

15. ASTM (American Society for Testing and Materials), 2000, Standard Specification for Flexible Cellular
Materials-Sponge or Expanded Rubber, ASTM D1056-2000, West Conshohocken, PA.

16. ASTM (American Society for Testing and Materials), 1998, Standard Specification for Hardware Cloth (Woven
or Welded Galvanized Steel Wire Fabric), ASTM A740, West Conshohocken, PA.

17. ASTM (American Society for Testing and Materials), 1998, Standard Specification for Stainless or Heat-Resisting
Steel Wire, ASTM 580, West Conshohocken, PA.

18. Burchsted, C. A., 1968, “Environmental Properties and Installation Requirements of HEPA Filters,”
Symposium on Treatment of Radioactive Wastes International, Atomic Energy Commission, Vienna,
Austria.

19. IEST (Institute of Environmental Sciences and Technology), 1993, HEPA and ULPA Filters,
IEST-RP-CC001.3, Mt. Prospect, IL.

20. UL (Underwriters Laboratories Inc.), 1990, Standard for Safety for High Efficiency, Particulate, Air Filter Units,
UL 586, Northbrook, IL.

21. UL (Underwriters Laboratory, Inc.), 1994, Safety Standard for Air Filter Units, UL 900, Northbrook, IL,
November 9.

22. Gilbert, H., J. K. Fretthold, F. Rainer, W. Bergman, and D. Beason, February 1995, “Preliminary Studies
to Determine the Shelf Life of HEPA Filters,” 23rd Department of Energy/Nuclear Regulatory
Commission Nuclear Air Cleaning Conference, CONF-940738, pp. 613-638, National Technical
Information Services, Springfield, VA.

23. Flanders Filters, Inc., 1988, Laminar Flow Guide HEPA, ULPA, and BLSI Filters, Washington, NC.

24. First, M. W. (Harvard University), 1980, “Performance of 1,000 and 1,800 CFM HEPA Filters on Long
Exposure to Low Atmospheric Dust Loadings II,” 16th Department of Energy Nuclear Air Cleaning
Conference, DOE Report CONF-801038, National Technical Information Service, Springfield, VA.

25. Anderson, W. L and T. Anderson, 1966, “Effect of Shock Overpressure on High Efficiency Filter
Units,” 9th Atomic Energy Commission Air Cleaning Conference, USAEC Report CONF-660904,
National Technical Information Services, Springfield, VA.

26. Bergman, W. and Hune, 1999, “Maximum HEPA Filter Life,” UCRL-AR-134141, Lawrence Livermore
National Laboratory, Oakland, CA.

27. Fretthold, J. K., 1997 “Evaluation of HEPA Filter Service Life,” Rocky Flats Environmental Technology Site
Report, RFP-5141, Boulder, CO, July 14.

28. Ricketts, C. I., V. Rüdinger, and J. G. Wilhem, 1987, “HEPA Filter Behavior Under High Humidity
Airflows,” 19th DOE/NRC Nuclear Air Cleaning Conference, Springfield VA, CONF-860820, NTIS,
pp. 319-52, May.

29. Zavadoski, R. and D. Thompson, 1999, HEPA Filters Used in the Department of Energy’s Hazardous Facilities,
DNFSB Tech-23, Washington, DC.

3-39
Nuclear Air Cleaning Handbook U.S. Department of Energy

30. DOE (U.S. Department of Energy), 1961, Recommended Minimal Specification Revised for the High-Efficiency
Particulate Filter Unit, Atomic Energy Commission Health and Safety Information Issue No. 120,
Washington, DC, June 30.

31. Conway, J. (Defense Nuclear Facilities Safety Board), 2000, Letter to B. Richardson, Secretary of Energy,
Recommendation 2000-2, Configuration Management, Vital Safety Systems, March 8.

32. LaMer, V. K. and Sinclair, D., 1943, A Portable Optical Instrument for the Measurement of Particle Size in Smokes,
(The OWL); and, An Improved Homogeneous Aerosol Generator, OSRD 1668, Office of Technical Services,
Washington, DC.

33. Rüdinger, V., C. I. Ricketts, and J. G. Wilhelm, 1985, “Limits of HEPA Filter Application Under High
Humidity Conditions,” 18th Department of Energy/Nuclear Regulatory Commission Nuclear Air
Cleaning Conference, CONF-840806, -1058, -1084, March.

34. IEST (Institute of Environmental Sciences and Technology), 1992, Testing ULPA Filers, IEST-RP-
CC007.1, Mt. Prospect, IL.

35. Hinds, W., M. W. First, D. Gibson, and D. Leith, 1979, Size Distribution of ‘Hot DOP’ Aerosol Produced by
ATT Q-127 Aerosol Generator, 15th Department of Energy Nuclear Air Cleaning Conference, National
Technical Information Service, Springfield, VA. p. 1127-1130.

36. ASME (American Society for Mechanical Engineers), 2000, Quality Assurance Requirements for Nuclear
Facilities, ASME NQA-1, New York, NY.

37. German Standards Institute (DIN), 1974, Typprüfung von Schwebstoffiltern, German Standard DIN 24, 184,
October.

38. BSI (British Standard Institute), 1969, Method for Sodium Flame Test for Air Filters, BSI-3928.

39. Dupoux, J. and A. Briand, 1977, Méthode de Mesure de L’efficacité des Filtre au Moyen d’un Aérosol d’Uranine
(Fluorescéine), Seminar on High Efficiency Aerosol Filtration, AFNOR NF X 44-011 Standard,
Commission of the European Communities, Luxembourg, pp. 249.

40. Robinson, K. S., C. Hamblin, R. C. Hodierne, and M. J. S. Smith, 1986, “In-service Aging Effects on
HEPA Filters,” Gaseous Effluent Treatment in Nuclear Installations, Graham and Trotman, London, England.

41. Johnson, J. S., D. G. Beason, P. R. Smith, and W. S. Gregory, 1989, The Effect of Age on the Structural
Integrity of HEPA Filters, 20th Department of Energy/Nuclear Regulatory Commission Nuclear Air
Cleaning Conference, CONF-880822, National Technology Information Service, Springfield, VA, May.

42. Fretthold, J. K., 1997 Evaluation of HEPA Filter Service Life, Rocky Flats Environmental Technology Site
Report, RFP-5141, Boulder, CO, July 14.

43. First, M. W., 1996, Aging of HEPA Filters in Service and in Storage, Journal of the American Biological Safety
Association, 1 (1) pp. 52-62.

44. ERDA (Energy Research and Development Administration), 1976, Nuclear Air Cleaning Handbook,
ERDA 76-21, Oak Ridge, TN.

3-40
DOE-HDBK-1169-2003 Chapter 3

45. Bill Richardson, Secretary of Energy, 2000, Letter to John Conway, Defense Nuclear Facilities Safety
Board Chairman, Implementation Plan for Recommendation 2000-2, Configuration Management, Vital Safety Systems,
October 31.

46. Bergman, V., “Maximum HEPA-filter Life,” UCRL-AR-134141, Lawrence Livermore National
Laboratory, Oakland, CA.

47. LLNL Health and Safety Manual, Adverse Operating Conditions, Section 12.05, High Efficiency Particulate
(HEPA) Filter System Design Guidelines for LLNL Applications, Lawrence Livermore National
Laboratory, Livermore, CA.

48. Linck, F. J. and J.A. Greer, 1975, “In-Place Testing of Multiple Stage HEPA Filter Plenums,” 13th AEC
Air Cleaning Conference, Energy Research and Development Administration Report CONF-740807,
National Technical Information Service, Springfield, VA, March.

49. ASHRAE (American Society of Heating, Refrigerating, and Air Conditioning Engineers), 1992,
Gravimetric and Dust Procedures for Testing Air Cleaning Devices Used in General Ventilation for Removing Particulate
Matter, Standard 52.1-92, Atlanta, GA.

50. ASHRAE (American Society of Heating, Refrigerating, and Air Conditioning Engineers), 1999, Method of
Testing Ventilation Air Cleaning Devices for Removal Efficiency by Particle Size, Standard 52.2-99, Atlanta, GA.

51. ARI (American Refrigeration Institute), 1993, Standard for Commercial and Industrial Air Filter Equipment,
ARI 850, Arlington, VA.

52. Silverman, L., 1951, Chemical Engineering Program, No. 9.

3-41