Document text
SuuiOXfl
Filtration in the Use of
Individual Water Purification Devices
Technical Information Paper #31-004-0306
PURPOSE
This information paper provides an in-depth review of filtration (including adsorption and ion
exchange) as a pathogen and particulate reduction mechanism when treating natural waters.
This paper is intended to assist the reader in evaluating the capabilities of Individual Water
Purification Devices (IWPDs) using size exclusion, adsorption, and/or ion exchange to reduce
disease-causing bacteria, virus, and protozoan cyst populations, as well as turbidity causing
particulate matter.
REFERENCES
Appendix A contains a list of references.
INTRODUCTION
Background
Understanding the ability of filtration to reduce disease-causing microorganisms is important in
protecting Soldiers, who are considering using this technology, from acute health threats posed
by these microorganisms. Soldiers deployed beyond traditional field drinking water supplies
must have access to potable water. Using IWPDs is one way to provide microbiologically safe
water in these situations. These IWPDs must protect the Soldier from acute microbial health
threats. The U.S. Environmental Protection Agency (EPA) Guide Standard and Protocol for
Testing Microbiological Water Purifiers (reference 1) provides perfonnance standards by which
an IWPD using filtration can be evaluated. The performance standards are a minimum 6-log
reduction/inactivation of bacteria, 4-log reduction/inactivation of viruses, and 3-log reduction/
inactivation of protozoan cysts (typically Giardia or Cryptosporidium ). IWPDs meeting these
standards are considered effective at reducing disease causing bacteria, viruses, and protozoan
cysts. Some IWPD manufacturers test their devices using this protocol. This is considered the
best way to evaluate the IWPDs pathogen reduction capabilities. In the absence of that testing
data, this information paper can be used to gain an understanding of the advantages as well as
limitations of filtration and help determine if an IWPD using filtration could successfully meet
the EPA Guide’s minimum performance standards.
Origin of Filtration for Water Treatment
For the purpose of this paper, filtration will be used broadly to incorporate separation by (1)
granular media, (2) size exclusion (e.g., membranes), (3) electrochemical adsorption (e.g.,
activated carbon), and (4) ion exchange (e.g., anion, cation exchange). Filtration is a well-
Report Documentation Page
Form Approved
OMB No. 0704-0188
Public reporting burden for the collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and
maintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information,
including suggestions for reducing this burden, to Washington Headquarters Services, Directorate for Information Operations and Reports, 1215 Jefferson Davis Highway, Suite 1204, Arlington
VA 22202-4302. Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to a penalty for failing to comply with a collection of information if it
does not display a currently valid OMB control number.
1. REPORT DATE
MAR 2006
2. REPORT TYPE
technical
3. DATES COVERED
00-10-2005 to 00-03-2006
7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES)
U.S. Army Center for Health Promotion and Preventive Medicine
(USACHPPM),5158 Blackhawk Road, APG,MD, 21010
9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES)
4. TITLE AND SUBTITLE 5a. CONTRACT NUMBER
Filtration in the Use of Individual Water Purification Devices 5b GRANT NUMBer
5c. PROGRAM ELEMENT NUMBER
6. AUTHOR(S) 5d. PROJECT NUMBER
Arthur Lundquist; Steven Clarke; William Bettin TIP 31-004-0306
5e. TASK NUMBER
5f. WORK UNIT NUMBER
7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) 8. PERFORMING ORGANIZATION
U.S. Army Center for Health Promotion and Preventive Medicine report number
(USACHPPM),5158 Blackhawk Road, APG,MD, 21010 TIP 31-004-0306
9. SPONSORING/MONITORING AGENCY NAME(S ) AND ADDRESS(ES ) 10. SPONSOR/MONITOR' S ACRONYM(S)
11. SPONSOR/MONITOR'S REPORT
NUMBER(S)
12. DISTRIBUTION/AVAILABILITY STATEMENT
Approved for public release; distribution unlimited
13. SUPPLEMENTARY NOTES
14. ABSTRACT
Soldiers deployed beyond traditional field drinking water supplies must have access to potable water.
Using Individual Water Purification Devices (IWPDs) is one way to provide microbiologically safe water in
these situations. Understanding the ability of filtration to reduce disease-causing microorganisms in water
is important in protecting Soldiers, who are considering using this technology, from acute health threats
posed by these microorganisms. This information paper provides an in-depth review of filtration (including
adsorption and ion exchange) as a pathogen and particulate reduction mechanism when treating natural
waters. This paper is intended to assist the reader in evaluating the capabilities of IWPDs using size
exclusion, adsorption, and/or ion exchange to reduce disease-causing bacteria, virus, and protozoan cyst
populations, as well as turbidity causing particulate matter.
15. SUBJECT TERMS
water; drinking water; water treatment; purification; filtration; military; IWP (individual water purifier);
pathogen removal;
16. SECURITY CLASSIFICATION OF:
a. REPORT
unclassified
b. ABSTRACT
unclassified
c. THIS PAGE
unclassified
17. LIMITATION OF
18. NUMBER
ABSTRACT
OF PAGES
19
RESPONSIBLE PERSON
Standard Form 298 (Rev. 8-98)
Prescribed by ANSI Std Z39-18
TIP #31-004-0306
studied process for drinking water treatment. Naturally, as groundwater migrates in the
subsurface, contaminants are removed from the water due to ionic attraction as well as sieving
based on size. Concurrently, contaminants such as iron and manganese may be dissolved into
the groundwater and often remain in the dissolved form until pumped to the surface. Similarly,
microorganisms are imparted to and extracted from the groundwater during subsurface
movement. Surface water (e.g., ponds, lakes, rivers), like groundwater, has ever-changing
quality with respect to microorganisms, particulates, chemistry, etc., but is more exposed to
human activity, often degrading water quality. To reduce water contaminants and create potable
water safe for human consumption, water treatment has included filtration to mimic and better
the natural removal of water contaminants. Filtration for water treatment dates back to 2000
b.c.e., where crude sand and charcoal filters were used to provide better tasting water (reference
2). Centuries later Hippocrates designed a cloth bag known as the Hippocrates Sleeve, used to
remove sediments from water after boiling. By the end of the Middle Ages water quality began
to be linked with disease. In the mid 19th century the spread of Cholera was noticeably
decreased where sand filtration was utilized (reference 2). The benefits of water filtration for not
only increasing water aesthetics, but decreasing the spread of disease, lead to the widespread use
of filtration seen today when purifying water for potable use.
Current Use of Filtration for Water Treatment
The original slow sand filtration developed centuries ago has now been replaced with rapid sand
filtration using multi-media beds, adsorption, utilizing electrochemical forces to attract
contaminants to the media surface, natural and synthetic membranes engineered with distinct
pore sizes, and ion exchange, where one ion is removed from the water and replaced with a less
offensive ion. Current U.S. Army field water treatment includes several filtration devices such
as the Reverse Osmosis Water Purification Unit, Tactical Water Purification System, and
Lightweight Water Purifier, designed for large volume water purification. An industry challenge
has been to reduce the size of full-scale filtration processes down to individual units, while
maintaining treatment efficacy against pathogens and particulate matter, but without excessive
maintenance. To date, there have been no IWPDs fielded to the Soldier that have used filtration
as the primary mechanism of water purification. Currently fielded emergency drinking water
products include an iodine-based disinfection tablet (Globaline ) and a flocculant-chlorine
TM
disinfectant based product (Chlor-Floc ). Today, there are several Commercial-Off-The-Shelf
(COTS) IWPDs that use filtration as the primary pathogen reduction mechanism.
Globaline is a trademark of Wisconsin Pharmacal Company, Jackson, WI.
Chlor-Floc is a trademark of Control Chemical, D/B/A Deatrick and Associates Inc., Alexandria, VA. Use of
trademarked products does not imply endorsement by the U.S. Army, but is intended only in identification of a
specific product.
2
TIP #31-004-0306
SEPARATION MECHANISMS
The mechanisms of separation during filtration vary depending on material and design. Overall,
several mechanisms may be simultaneously rejecting contaminants. For example, during
filtration primarily incorporating size exclusion, adsorption and depth filtration mechanisms are
likely aiding in particle retention.
Straining
Straining entails the removal of particles by size exclusion when particles are larger than the void
spaces in the filter. Straining is a removal mechanism for virtually all filtration technologies
with the importance of this mechanism related to raw water quality and size of particulate matter
in reference to pore size.
Straining by Granular Media
For spherical granular media, close-packed arrangement will remove particles when the ratio of
particle diameter to grain diameter is greater than 0. 15 (reference 3). For typical slow sand
filters, this equates to the removal of particles down to about 15 pm, increasing to 30-80 pm for
rapid sand filtration. It should be noted that other mechanisms aid in the removal of smaller
particles for these filtration techniques. Specifically, for slow sand filtration a thin slimy layer
of particulate sludge forms, termed smutzdecke, effective in trapping particulates and
microorganisms at the surface. When particulates fonn a layer during granular media filtration it
may also be termed a cake. Cake filtration is often used to describe straining out particles, often
smaller than the media pore size, by this top layer, or build-up, when evaluating granular carbon
filtration.
Straining by Membrane Filtration
Porous membranes contain varying size pores and are rated by their pore size based on nominal,
average, and absolute size. Absolute pore size is the size of the largest particle (e.g., glass bead)
that will pass through a membrane under specific testing conditions. For membranes with
uniform cylindrical pores this rating has meaning, but only under the low pressure conditions
tested during pore size determination. Membranes with cylindrical pore structures are called
capillary-pore membranes. Conversely, some membranes are manufactured to create a tortuous
path (sponge-like appearance, tenned tortuous-pore membranes) where pores of varying size
create a path by which depth filtration mechanisms arise as well as size exclusion. In this case,
the term absolute pore size has little meaning, and nominal ratings are used. Nominal pore
ratings specify the percentage of particles removed of a certain size particle, again usually tested
with glass beads (e.g., 80% of 1 pm particles retained). Lastly, membrane pore size can be rated
as the average size of all pores. Different pore size testing techniques, as well as varying
definitions, create a questionable pore rating system unless proper information on the membrane
3
TIP #31-004-0306
is noted. For example, it has been noted that certain manufacturers state absolute pore sizes
when a membrane can remove 85% of a certain size particle, contrasting the historical definition
of an absolute pore rating. Caution, therefore, must be used when evaluating membrane efficacy
based solely on stated pore size.
Depth Filtration Theory
Particle removal and retention within depth filters involves Van der Waals forces where two
surfaces have attractive forces, in this case between the particle and the media surface. Van der
Waals forces are short-ranged, and only become effective when the two surfaces are in close
proximity. For particle-media surfaces to come close enough together for these forces to become
effective, transport mechanisms must be present. These mechanisms are represented by three
different processes, which include interception, inertia and sedimentation, and diffusion. These
processes are attributed with most particle removal. As a particle is transported through a filter,
if the streamline is within one half or less of the diameter of the particle from the media surface,
the particle will be intercepted. Second, as streamlines curve around the media, particles can
deviate from the streamline and continue towards the media due to inertia forces. Particles may
also deviate from streamlines due to gravitational forces and settle onto the media surface. In
both cases, particle will be retained at the media surface. Lastly, particles may deviate from
streamlines due to Brownian motion and diffuse to the media surface. The following diagram,
Figure 1 (borrowed from reference 3), illustrates the different filtration mechanisms described.
Depth filtration is not limited to granular media, but can be applied to microfilters, membranes
and carbon filtration as well.
Figure 1. Filtration Mechanisms.
Particle transport mechanisms in
fundamental filtration theory: (a) inter¬
ception, particle A follows streamline but
collides with the collector because of the
proximity between the streamline and
the collector; (b) inertia, sedimentation,
particle B deviates from the streamline
and collides with the collector because of
inertial or gravitational forces; (c) diffusion,
particle C collides with collector due to
random Brownian motion.
Diagram borrowed from reference 3.
4
TIP #31-004-0306
Rejection by Osmotic Membranes
Two solutions in contact with one another with varying solute concentrations naturally try to
equilibrate. In water treatment we can use this driving force to equilibrate, by placing a semi-
permeable membrane between the two solutions. By engineering the membrane to allow passage
of the water molecules through the membrane, yet reject the solutes, the two solutions will
naturally equilibrate as the water dilutes the more concentrated side. Flux through the membrane
will vary based on solute gradient, temperature, and membrane properties. Common practice in
water treatment is to reverse the natural osmotic tendency by pressurizing the influent side,
forcing water molecules through the membrane and rejecting the solutes, termed reverse osmosis
(RO). Despite use in water treatment for many years, the exact mechanism of water transport
and solute rejection is still debated. The underlying question is whether these membranes are
non-porous and diffusion driven, or whether they contain very small pores for preferential (size
exclusion) convective transport of the solvent. There are several theories, or models, on the
rejection mechanisms of osmotic membranes of which three are most commonly accepted.
Solution-Diffusion Model
The solution-diffusion model describes penneation through a dense membrane that is permeable
but non-porous. Water and solutes dissolve into the membrane, diffuse through the solid
material, and re-liquefy on the permeate side. In this model, separation occurs due to the
different flux of solutes.
Pore Flow Model
This model considers convective flow through a porous membrane. Water and solute flux is
coupled with separation occurring due to sieving. Since many solutes, namely salt, are similar in
size to water molecules, physical sieving would not be efficient. An apparent limitation of this
model is the small pore size required, less than 0.1 nm, for separation to occur.
Preferential Sorption-Capillary Flow Model
This model describes a porous membrane where water is preferentially sorbed to the surface and
transported through the membrane due to concentration gradient. Membranes with low dielectric
constants prefer water molecules, creating a layer of low solute concentration, in essence
blocking the solutes from contact with the membrane surface and therefore preventing passage.
Osmotic potential, to pull water across a membrane from a less to more solute concentrated side,
has also been applied to IWPDs in a passive form. By using a non-offensive solute on the
membrane product side, water will naturally pass across the membrane to the higher solute
concentration. Sometimes termed forward osmosis, this process, simply termed osmosis (O) for
this paper, utilizes the same pathogen reduction mechanisms as that of conventional RO.
5
TIP #31-004-0306
Adsorption
Adsorption is a mass transfer operation in which contaminants present in a liquid phase are
accumulated on a solid phase, thereby being removed from the liquid. The constituent being
adsorbed is referred to as the adsorbate and the solid onto which the constituent adsorbs is the
adsorbent. The degree of adsorption is affected by attraction of the three following interfaces:
adsorbate/adsorbent, adsorbate/water, water/adsorbent. The strength of the adsorbate/adsorbent
interface as compared to the others will determine adsorption efficacy. Dissolved species are
concentrated onto the surface by physical attraction or chemical reaction. Physical adsorption is
by nonspecific binding mechanisms such as Van der Waals forces. This binding is reversible,
where adsorbates may desorb in response to a decrease in solution concentration. Chemisorption
entails specific attraction where chemical binding transfers electrons between the adsorbent and
adsorbate. Physical adsorption has weaker forces and bonding energies, operates over longer
distances, and is more reversible than chemical adsorption. Chemical adsorbates, which can only
form a layer one molecule thick due to specific bonding, may have several different attractive
forces. Polar compounds having a slightly positive and negative end and molecules orient
themselves to lower their combined free energy, creating a dipole attraction. The negative end
attracts the positive end of another molecule fonning a dipole-dipole bond. More important to
water treatment is the dipole-dipole bond with water, termed hydrogen bonding. These bonds
are very strong and are responsible for water being a liquid at room temperature. Hydrogen
bonding between the water molecule and adsorbate competes with adsorbate/adsorbent
attraction. By maximizing physical attraction, covalent bonding and Coulombic forces, all of
which are not involved in adsorbate/water, water/adsorbent interaction, we can increase
adsorption efficacy. Water pH, molecule size, and adsorbate solubility all play roles in
adsorption and affect species (polar, neutral, ionic) differently. Since adsorption is not a primary
mechanism for pathogen reduction these interactions will not be further discussed but can be
found elsewhere (references 3-5). During the adsorption process, dissolved species are
transported into the porous structure of the adsorbent material by diffusion, then adsorbed onto
the interior surface of the grain. Porous adsorbent materials have very large internal surface
areas (400 - 1500 m2/g), and pore volume (0.1 - 0.8 mL/g) (reference 3) creating many sites for
adsorption to occur. Three commonly used commercial adsorbents include zeolites
(aluminosilicates), synthetic polymeric adsorbents, and activated carbon. A notable affect on
adsorption with the most common adsorbent, activated carbon, is water pH. In order for
electrostatic interactions to contribute to removal by adsorption, particle-media charges must
attract the particle to the media surface. Since most particles in natural waters posses a negative
charge, media should posses a positive charge. As pH increases, activated carbon becomes less
positive until a point of zero charge (PZC) is reached (reference 4). At a pH above this point,
electrostatic interactions repel particles from the surface, inhibiting adsorption. Depending on
the carbon used the PZC may range from a pH of less than 4 up to greater than 10 (reference 4).
6
TIP #31-004-0306
Ion Exchange
Ion exchange for drinking water is a process in which ions within the water stream are adsorbed
to the surface of resins and exchanged for a less offensive ion that is then imparted into the
finished water. A generic representation of softening using a sodium resin is shown below, with
R representing the exchange resin.
R-(Na+)4 + Ca+2 - — ► R-(Ca+2) + (Na+)4
Similar to adsorption, ion exchange is powered by electrostatic/electrochemical attraction in
which ions of opposite charge attract, however, with ion exchange, the presaturant ions on the
resin are released into the water. For ion exchange to occur, the presaturant ions cannot be
present in the bulk fluid. Natural tendency to equilibrate will favor ions both in the bulk fluid as
well as on the resin surface, therefore equilibrium will occur if given enough time (reference 6).
Resin beads are usually 0.04 to 1.0 mm in diameter and made by materials such as polystyrene
divinylbenzene. Favorable ion exchange resins are reversible, and once all exchange sites are
exhausted they can be restored through regeneration, although eventually irreversible fouling
will occur. Regeneration usually consists of several bed volumes of highly concentrated
regenerant followed by rinse water. To date, the most common use of ion exchange has been for
softening, although heavy metal reduction and resins designed for specific ion reduction are also
becoming more commonplace. There are four common ion exchange resins, classified as either
strong-acid cation, weak-acid cation, strong-base anion, or weak-base anion. The cation
exchange resins are negatively charged resins often used for calcium and magnesium removal,
while the less common anion resins are positively charged for the removal of nitrate and other
anions. Both strong-acid and strong-base resins are effective throughout all pH ranges, with the
weak-acid and base resins effective only within narrow alkaline and acidic pH regions,
respectively. The preference of the ion exchange resin to attract one ion over another is termed
its selectivity sequence. Ions are ranked based on separation factors, or the ratio of the affinity of
the resin to favor the ion compared to the presaturant ions already attached to the resin. In
general, with dilute solutions, ion exchange resins prefer ions with the highest charge and lowest
degree of hydration. If both anion and cation removal is required, different resins can be run in
series or mixed bed resin columns can be used to produce deionized water. In this case, strong-
acid resin of the H+ form and strong-base resin of the OH' form are mixed with the resultant
presaturant ions released forming water. In this case no ions are imparted to the finished water.
A major drawback of mixed bed resins is that the resin must be separated before regeneration can
occur. Since IWPDs are not designed to be regenerated, these drawbacks are not applicable.
ROLE OF PATHOGEN IN FILTRATION SEPARATION MECHANISMS
The primary difference between pathogens for reduction during filtration is size. Approximate
sizes are as follow: viruses 0.005 - 0.3 pm, bacteria 0.1 -10 pm, Cryptosporidium oocysts
4-6 pm, Giardia cysts 8-12 pm. Common filters used in IWPDs have pore sizes between
7
TIP #31-004-0306
0.2 and 2 jam, although some exist outside of this range. Primary reduction mechanisms for each
pathogen vary with purification technology, with generalizations based on pathogen morphology
as follows. (1) Based on size exclusion alone, filter retention of Cryptosporidium oocysts and
Giardia cysts is likely for properly functioning devices. It is generally assumed that if a filter
can reduce Cryptosporidium oocysts then Giardia cyst reduction is likely (reference 7).
Utilizing filters where the primary means of reduction is by size exclusion, latex microspheres
have been used as surrogates, demonstrating the lack of importance of other mechanisms for cyst
reduction (references 1, 8). (2) Bacterial reduction by filters is based on adsorption as well as
size exclusion (reference 9). Reduction by microporous media with pore sizes of 0.45 pm or less
will likely provide adequate bacterial reduction based on size exclusion alone. Clean bed
filtration, utilizing larger pore sizes will likely not meet the bacterial reduction requirements of
references 1 and 10. (3) Due to the extremely small size of viruses, reduction by size exclusion
to the levels required in references 1 and 10 is unlikely, unless utilizing very tight membranes
such as for osmosis. Extensive literature exists demonstrating viral adsorption onto microporous
filters as well as how water quality affects viral reduction (references 9 and 1 1-24). Particles
immersed in aqueous solutions, including viruses, develop a surface charge by adsorbing ions on
its surface (reference 1 1). The charge of viruses has been shown to play a significant role in
adsorption onto surfaces and this charge changes with pH. Similar to the ZPC of activated
carbon, the pH at which viruses have no net charge is called the isoelectric point (pi). Below this
pH, viruses are positively charged, and above this point they are negatively charged. Coupling
filters that are positively charged at a pH where the viruses are negatively charged, with the
difference in charge minimized (e.g., near both pi) promotes the most efficient adsorption
(reference 12). From this, it is apparent that no single combination of adsorbent/adsorption
conditions exists to give optimum reduction of all viruses for all water qualities (reference 12).
Increasing electrostatic and or hydrophobic interactions by the addition of chemicals such as
magnesium sulfate (reference 13) or by specially treating the filter to promote a positive charge
at natural water pH will increase virus retention (references 14-17). One study investigating
coliphage reduction by a 0.2 pm microporous filter, showed reduction based on adsorption as
well as size exclusion (reference 9). Initial retention on clean bed filters was based on inertial
impaction due to adsorptive forces, resulting in low to moderate reduction and highly affected by
flow rates, water quality, and membrane material. As cake formed on the surface the primary
reduction mechanism changed to direct interception at the surface due to reduction in pore size
(reference 9). Reduction efficacy was less affected by water quality but still showed some
susceptibility to changes in flow rate. Virus reduction by adsorption or size exclusion on
capillary formed membranes is unlikely to consistently meet the requirements of reference 1 .
IWPDs USING MEMBRANE FILTRATION
Membrane Filtration
A membrane is a thin layer of semi-permeable material that is capable of separating materials
when a driving force is applied across the surface. This separation into two phases
8
TIP #31-004-0306
(concentrations) creates a chemical potential between the two sides of the membrane that is
based on the physical and chemical properties of the materials being separated. Membranes are
not considered to be passive materials but are termed functional materials whose performance
characteristics are based on the nature of the elements to be separated and the driving force.
Membranes are classified based on the size or molecular weight cutoff (MWCO) of the solutes
they are capable of rejecting. Membranes used in water treatment, in order of decreasing pore
size/MWCO, are microfilters, ultrafilters, nanofilters, and osmotic membranes. In addition to
the pore size, membranes are also classified based on their structure, either symmetric or
asymmetric. Symmetric membranes contain consistent pores, porosity, and transport properties.
Asymmetric membranes contain complex pore structure with pore size, porosity, and transport
properties changing with depth. Asymmetric membranes contain a thin active layer where
separation occurs, supported by a thicker, more porous support structure to provide membrane
integrity. Currently available IWPDs utilize micro and osmotic membrane filters. Membranes
are complex materials and are often difficult to classify due to minor differences in materials and
structure. The following information gives general information on the most common types of
membranes used in IWPDs. Membrane configurations within IWPDs are commonly oriented as
flat sheet, pleated sheet, or hollow fiber. With respect to pathogen reduction efficacy, membrane
orientation is not a factor. Due to lack of information provided by manufacturers, and the
proprietary nature of IWPDs, not all types of membranes found in IWPDs will be discussed.
Polymer Microfilter Membranes
Polymer microfilter membranes used in IWPDs are thin sheets up to about
200 pm thick or hollow fiber microporous membranes having diameters of 70 to 600 pm and
thicknesses similar to thin sheet membranes. These membranes are engineered with specific
properties for different applications and can be made of many materials. Common materials may
be polycarbonate (PC), cellulose acetate (CA), or polyethersulfone (PES). Each material
contains properties that affect membrane performance. In general, increasing hydrophilicity
(contact angle less than 90 degrees, e.g., does not repel water molecules) will decrease fouling
potential and increase flux. Membranes that are biologically inert, operate over a wide pH and
temperature range, and are chemically resistant are the most desirable for water treatment.
Detailed descriptions on the production of these membranes can be found in reference 25.
Microbial pathogen reduction mechanism by polymer microfiltration membranes is
based on pore structure. Capillary-pore membranes, often made of PC, are thin (about 10 pm)
and consist of uniform cylindrical pores, reject microbes based on size exclusion alone, and are
generally given an absolute pore size rating. In theory, these membranes should reject all
microbes greater than the pore size, but in practice, defects in pore size manufacturing as well as
seams and seals within the device will prevent total rejection of larger organisms. During use,
capillary-pore membranes will build-up rejected solids on the surface of the membrane. This
build-up will decrease the effective pore size of the membrane and increase headloss. As this
clogging increases, so does the ability of the membrane to reject microorganisms. Clean
9
TIP #31-004-0306
capillary-pore membrane micro filters have pore sizes down to 0. 1 pm, which can be expected to
reject bacteria and protozoan cysts, but have minimal effect on virus reduction. In contrast to
capillary-pore membranes, tortuous-pore membranes are thick (about 150 pm), consist of
sponge-like structure where sieving as well as depth filtration mechanisms dominate, and have
increased flux over capillary-pore membranes. These are often made of CA or PES. Pore sizes
vary with depth and spatially with direction. In addition to sieving, microbes are adsorbed onto
the media as described in the above sections on depth filtration theory and adsorption. Due to
more efficient separation mechanisms, these membranes have been shown to retain particles
orders of magnitude smaller than the nominal pore size (reference 25). Tortuous-pore
membranes, like capillary-pore membranes, have pore sizes down to about 0.1 pm, making these
efficient at retaining bacteria and protozoan cysts, but not effective at sieving viruses. Due to the
adsorptive nature of these membranes, it has been shown that several log virus reduction can be
achieved but results are inconsistent and drop with continued production (references 3 and 25).
Polymer microfilter membranes are very effective at reducing particulate matter and based on
pore size should be able to reduce water turbidity to below 1 nephelometric turbidity unit (NTU).
Due to the small pore size of these membranes they are prone to fouling, especially with the
dead-end configurations used in IWPDs. Pre-filtering and a cleanable or backwashable
configuration will reduce fouling.
Osmotic Membranes
Osmosis uses pressure, RO or solute gradient osmosis, to drive the solvent through a
dense, nonporous membrane (some models consider a porous membrane) that will retain salts
and solutes down to very low molecular weights. Natural osmotic pressure induces travel from a
less to a more concentrated solution. A pressure, in excess of the osmotic potential, must be
applied to reverse this flow (RO). Osmotic potential is a function of the molar concentration of
the solute. In essence, smaller molecules create higher osmotic potentials. Pressures to reverse
this natural tendency can be high. Twice the osmotic pressure is common in design with
seawater separations, with pressures of 5 to 8 MP are typically used. The mechanism of
separation for RO is solution/diffusion + exclusion as explained above. Separation is based on
the solubility and diffusivity of materials in the membrane. RO membranes are usually made of
hydrophilic cellulose acetate materials, cellulose ester plastics, or composites such as a cross-
linked polyamide on a polysulfone and fabric base. CA membranes along with other non¬
composite membranes are termed asymmetric. The entire membrane is composed of the same
material with the pore size decreasing as you approach the surface. In nonporous asymmetric
membranes, the surface skin is dense with a porous support membrane underneath of the same
material. Composite membranes are anisotropic where the top layer and sublayer originate from
different material. The top dense layer sits on top of a porous material, usually an asymmetric
membrane. Composites can be designed for certain selectivities, but presently are less common
than CA. CA membranes can resist a low level chlorine residual, but are very susceptible to
biological degradation. RO membranes are very thin ranging from 0.25 to 4 pm to increase flux
through the membrane as flux is inversely proportional to membrane thickness. They operate
10
TIP #31-004-0306
ideally at pH 4 to 6.5 and at temperatures below 30° C. Water flux increases with temperature as
long as temperature remains within the ideal range of the membrane material. Membrane
configuration may be plate and frame, spiral-wound, tubular, or hollow fine fiber. The most
common configuration, spiral-wound, contains sheets of membranes separated by spacer sheets
then rolled together around a feedwater spacer. The hollow fine fiber configuration is similar to
that used for micro filtration but incorporating tighter membranes. Increased surface area,
resulting in higher flux, and less fouling are benefits of the hollow fine fiber design.
Osmotic membranes are classified based on MWCO with mechanisms of removal
described in an above section. Measured in dalton, these membranes are capable of rejecting
molecules with a mass of > 100 dalton regardless of charge. Generally speaking rejection
efficacy favors multivalent ions, branched isomers, and increasing molecular mass. Based on
size exclusion alone, osmotic membranes are capable of retaining species as small as 0.0001 pm
(reference 26). These membranes can remove most all natural water contaminants known,
although no treatment can universally remove everything. Microorganisms, salts, hardness, and
organic chemicals, among many others can be removed, whereas most dissolved gases such as
hydrogen sulfide and carbon dioxide will not be removed (reference 26). IWPDs utilizing
osmotic membranes are historically designed for salt water desalination. With the introduction
of IWPDs using osmosis, application to fresh water has been considered. Currently, IWPDs
using RO or O should be capable of reducing waterborne pathogens (bacteria, cysts, and viruses)
to levels considered acceptable for human consumption, as recommended by the EPA (reference
1). Devices using osmotic membranes will produce the lowest NTU water of all membrane
materials. IWPDs using RO are historically not designed for natural water purification where
turbid water may quickly foul the membrane. RO units will perform most efficient for
desalination were particulate matter is not a concern. RO use in IWPDs for natural waters would
require very efficient pre-filtering, as by another membrane process such as microfiltration, and
is therefore not considered a viable technology. IWPDs using O will also produce extremely low
NTU water and will not be affected by particulate matter regardless of natural water turbidity.
Since O devices do not use pressure to force water through the membrane, no cake is formed at
the media surface and no pre-filtering is required.
IWPDs USING CERAMIC MICROFILTRATION
Ceramic microfilters are made from inorganic ceramic pastes derived from powders of alumina
(AEO3), zirconia (Zr02), and titanium (Ti02). These pastes are extruded and sintered at high
temperature to form membrane supports with macro pores. Subsequently, submicronic powders
are laid on the supports to create smaller pore diameters. This process creates a symmetric
material with high chemical, mechanic, and thennal resistance that can be formed in a variety of
shapes including candles, discs, and tubes (reference 27). Pore structure is tortuous path depth
filtration with symmetric pores throughout the depth of the filter. With pore sizes down to
0.1 pm, ceramic microfilters are efficient at retaining bacteria and cysts through adsorption and
depth filtration mechanisms. At the household level utilizing untreated water sources, ceramic
11
TIP #31-004-0306
filter use has been shown to reduce coliform bacteria resulting in greater than 70% reduction in
cases of diarrhea (reference 28). As with other microfilters, no mechanism exists to adequately
reduce virus concentrations. Commercially available ceramic microfilters are often impregnated
with silver to discourage microbial growth on the media surface. This is intended solely to limit
growth on the media and will have no effect on bulk water pathogen reduction. Ceramic
microfilters are very effective at reducing particulate matter and based on pore size should be
able to reduce water turbidity to below 1 NTU. Due to the small pore size of these filters they
are prone to fouling, especially in dead-end configurations used in IWPDs. For IWPD use,
ceramic filters are designed to be mechanically cleaned by scraping particulate build-up from the
media surface. The ability to clean this media multiple times makes these filters a very effective,
but high maintenance, technology for use with turbid waters. Due to the small pore size of these
membranes, pre-filtering is required.
IWPDs USING FIBER AND FABRIC FILTRATION
Fiber and fabric microfilters can be made of compressed or cast fibers such as cellulose papers,
woven fabrics, and glass, in addition to numerous other materials (reference 29). The most
common to IWPDs are fiber microfilters made of material such as borosilicate glass. These
filters are symmetric depth filters with pores sizes down to about 0.2 pm. Pathogen reduction
follows depth filtration, adsorption, and straining mechanisms. Clean bed pathogen reduction
may entail Van der Waals interaction and electrostatic interactions as well as straining based on
size exclusion. After continued use, cake formation will likely make straining the predominant
rejection mechanism. Consistent reduction of bacteria and cysts based on size exclusion is
expected. No mechanisms exist to consistently reduce virus to the standards of reference 1.
Fiber and fabric microfilters are very effective at reducing particulate matter and based on pore
size should be able to reduce water turbidity to below 1 NTU. Due to the small pore size of
these filters they are prone to fouling, especially in the dead end configurations used in IWPDs.
With proper design, such as allowing for mechanical cleaning by way of scraping the surface,
these filters can be highly effective at treating turbid waters. Non-cleanable filters, requiring
replacement once clogged are not as desirable for turbid waters. Due to the small pore size of
these membranes, pre-filtering is required.
IWPs USING CARBON FILTRATION
Carbon Filtration
Carbon used for water treatment can be of three different fonns; granular, powdered, block.
Granular activated carbon (GAC) for water treatment is often made from wood, peat, lignite,
coal, or coconut shells. Manufacturing consists of carbonization and activation. Carbonization
is conducted in the absence of air at temperatures up to 700° C, while activation, or oxidation, is
accomplished at temperatures of 800 - 900° C in the presence of oxidizing gases such as steam
or CO2. Activation bums off anything volatile, leaving highly porous grains with large surface
12
TIP #31-004-0306
areas. Grain size varies with typical values between 0.4 mm and 2.5 mm. Powdered activated
carbon (PAC) is made of the same materials as the granular form, but activation can entail either
gas or chemical processes. The final product is powder with typical particle sizes ranging from
10 to 100 pm. Carbon block is produced by sintering powdered carbon, thennoplastic binders,
and other additives. Material is extruded or molded under heat and pressure to form a hollow
filter block of just about any shape or size. Absolute control over pore size is possible as well as
engineering for specific contaminant reduction. Carbon blocks, unlike GAC, contain increased
surface area, do not exhibit channeling, and contain an order of magnitude smaller pore size
resulting in increased adsorption capacity (reference 30). Commercially available carbon block
is often impregnated with silver to discourage microbial growth on the media surface. This is
intended solely to limit growth on the media and will have no effect on bulk water pathogen
reduction. When carbon adsorption capacity becomes exhausted, regeneration, involving the
desorption of solutes from the media without affecting the media surface, and reactivation,
entailing partial regeneration affecting the media surface, are conducted to restore the media for
future use.
Pathogen Reduction
GAC has no specific mechanism for pathogen reduction beyond that typical of other granular
media (reference 31). Typically larger in size than most filter media, pathogen and particulate
removal by GAC is poorly accomplished by the straining and depth filtration mechanisms
described in an above section. PAC, like GAC, is used for taste and odor reduction, and is not
considered an effective barrier to pathogens. Carbon blocks have been shown to effectively
reduce pathogens from water (references 32-34). Pathogen reduction by carbon blocks can
follow any of the three generally accepted particle reduction mechanisms for porous media; cake
filtration (surface retention), depth filtration, or adsorptive filtration. Depending on pore size,
pathogens may be retained based on size exclusion alone. As cake forms on the media surface,
exclusion of smaller particles due to decreased pore size is considered a predominant reduction
mechanism (reference 33, 34). Carbon block surface charge may play an important role in clean
bed filtration. The surface charge of carbon block is based on the pH at which the surface is not
charged, called the PZC (reference 4). At pH below this point the surface is positively charged
and above this point negatively charged. Since pathogens generally possess a negative charge, as
pH decreases, reduction should increase due to electrostatic interactions. It has been shown that
initial reduction due to electrostatic or Van der Waals attraction is followed by straining, as the
negatively charged particles neutralize the surface of the carbon block (reference 32). When pH
was above the PZC, pathogen reduction based on adsorption was ineffective. Proprietary
chemically treated carbon blocks are available that have been shown to be capable of reducing
bacteria, cysts, and viruses by the requirements of reference 1 (reference 32). Little is known
about the proprietary chemical treatment and the exact pathogen kill mechanism is unclear. With
respect to available IWPDs, carbon blocks with pore sizes of 1 pm or greater are common.
Based on this, cyst reduction would be likely, and except for specially treated carbon blocks,
consistent bacterial and viral reduction would not be expected to the reduction requirements of
13
TIP #31-004-0306
reference 1 . Granular carbon filtration will retain some particulate matter based on particle size.
As a cake forms on the surface, increased removal will occur. Clean bed granular carbon alone
will not likely reduce water to less than 1 NTU. Carbon block filtration will reduce particulate
matter with efficacy based on block pore size. Again, particulate size will be a factor in retention
within carbon blocks which, as used currently in IWPDs, have a pore size of about 1-2 pm.
Granular carbon will not likely be the limiting treatment technology requiring pre-filtering for
IWPDs, as an additional pathogen reduction mechanism will be present that will dictate required
pre-filtration. To reduce clogging, pre-filtering is beneficial when using carbon block, but not
required as shown by current device configurations.
IWPs USING ION EXCHANGE
Ion exchange is not a proven technology for pathogen reduction. IWPDs utilizing ion exchange
must employ an additional mechanism to adequately reduce microbial contamination. Microbial
growth can occur within ion exchange beds, possibly resulting in increased contamination due to
microbial growth sloughing into the effluent stream. One non-conventional ion exchange
process has shown much promise at inactivating pathogens. Iodine ion exchange resins,
primarily of the tri-iodide or penta-iodide form, have been extensively studied and are considered
effective at pathogen inactivation through disinfection mechanisms (references 29, 35). Ion
exchange is not designed for, and will not be effective at, reducing particulate matter. Pre¬
filtering is necessary to avoid fouling of the resin.
CONCLUSION
The effectiveness of filtration as the primary mechanism to reduce pathogens in IWPDs is based
on the technology used as well as the raw water quality. Filtration utilizing microporous filters
primarily reduces pathogens by size exclusion due to surface or depth filtration mechanisms.
Adsorptive interactions contribute to pathogen reduction during the initial filtration until cake
formation occurs where charge neutralization limits the effectiveness of this mechanism. For
IWPDs using size exclusion as the reduction mechanism, bacteria and cyst reduction is possible
dependant on pore size. The small size of viruses prevents retention by size exclusion to the
reduction requirements for purifying natural water. Adsorption of viruses has also been shown
to be inadequate to consistently meet requirements for producing microbiologically safe water.
Carbon filtration performs similar to granular or microporous filters with equivalent pore sizes.
Proprietary chemically treated carbon surfaces have been shown to meet reduction requirements
for microbiologically safe water but may be sensitive to water characteristics such as pH.
IWPDs using osmotic membranes are the most effective at reducing pathogens although pressure
driven osmotic devices will quickly foul when used with fresh water sources. For IWPDs,
filtration will decrease the particulate matter present in turbid water with efficacy based on pore
size. The ability of the IWPD to perform properly with turbid water sources is dictated by the
pre-filter configuration and ability to clean the media surface.
14
TIP #31-004-0306
Table. Summary of the Pathogen Reduction Efficacy and the Effect of Particulate Matter
on IWP Filtration Technologies.
Technology
Summary
Membrane
Microfilter
Expected effectiveness at reducing bacteria and cysts. Microfilter pore
size too large to adequately reduce viruses, requiring additional treatment.
Common configurations limit the effectiveness of membrane surface
cleaning making this technology susceptible to fouling from particulate
matter. Degree of fouling directly related to efficacy of pre-filter.
Straining as well as depth filtration mechanisms may be involved in
microbial and particulate rejection based on membrane structure.
Ceramic
Microfilter
Expected effectiveness at reducing bacteria and cysts. Microfilter pore
size too large to adequately reduce viruses, requiring additional treatment.
Ability to scrape rejected material from the microfilter surface enables
flow to be restored after fouling. Frequency of cleaning, and length of
filter useful life directly related to efficacy of pre-filter. Straining as well
as depth filtration mechanisms can be involved in microbial and
particulate rejection.
Fiber/Fabric
Microfilter
Expected effectiveness at reducing bacteria and cysts. Microfilter pore
size too large to adequately reduce viruses, requiring additional treatment.
Filters designed to be cleanable should provide some ability to restore
flow after fouling. Frequency of cleaning, and length of filter useful life
directly related to efficacy of pre-filter. Non-cleanable filters highly
susceptible to fouling. Straining as well as depth filtration mechanisms
may be involved in microbial and particulate rejection.
Reverse Osmosis
Effective at reducing bacteria, viruses, and cysts. Technology is not
designed to treat fresh water sources and, therefore, requires very effective
pre-filtering to prevent membrane fouling. Not a feasible IWP technology
for microbial or particulate reduction of fresh water.
Osmosis
Effective at reducing bacteria, viruses, and cysts. Technology is passive,
eliminating the fouling effects of turbid water, and eliminating the need
for pre-filtration. Slow production of fluid, exacerbated by cold
temperatures.
Granular/Powdered
Carbon
Not considered effective at reducing bacteria, viruses, or cysts. Granular
media is often too large to effectively reduce pathogens based on size
exclusion and is not considered effective at depth filtration mechanisms.
Powdered carbon is used solely for taste and odor reduction and is not
effective at pathogen reduction. Particulate matter affects these
technologies similar to conventional granular media.
Carbon Block
Expected effectiveness at reducing cysts. Consistent reduction of bacteria
is not expected due to the pore size of carbon blocks commonly used in
IWPs. Not effective at adequately reducing viruses, although proprietary
15
TIP #31-004-0306
media has shown some promise. Pathogen reduction based on size
exclusion and depth filtration mechanisms. Effects of particulate matter
similar to other technologies of similar pore size. Pre-filtration and
cleanable filters will decrease fouling from particulate matter.
Ion Exchange
Not considered effective at reducing bacteria, viruses, or cysts. Iodine ion
exchange resins have been proven effective at pathogen inactivation
through disinfection mechanisms. Particulate matter fouls ion exchange
resin and therefore prefiltration is necessary.
PREPARED BY: Arthur H. Lundquist, Environmental Engineer
DATED: March 2006
16
TIP #31-004-0306
APPENDIX A
REFERENCES
1. U.S. Environmental Protection Agency (EPA), Registration Division Office of Pesticide
Program, Criteria and Standards Division Office of Drinking Water, 1987. Guide Standard and
Protocol for Testing Microbiological Water Purifiers. Washington, D.C.
2. Baker, M.N. and Taras, M.J., 1981. The quest for pure water: The history of theTwentieth
Century, Volume 1 and 2. American Water Works Association (AWWA), Denver, CO.
3. Crittenden, J.C., et. al., 2005. Water Treatment: Principles and Design, 2nd ed., John Wiley
& Sons, Inc., New Jersey.
4. Sontheimer, H., et.al., 1988. Activated Carbon for Water Treatment, 2nd ed., AWWA,
Denver, CO.
5 Pontius, F.W. (ed), 1990. Water Quality and Treatment., 4th ed. McGraw-Hill, Inc.,
New York.
6 Owens, D.L., 1985. Practical Principles of Ion Exchange Water Treatment. Tall Oaks
Publishing, Inc., New Jersey.
7. Hunter, D., 2000. “Control of Cryptosporidium in Water Systems Using Cartridge
Filtration,” Health Estate Journal, 54(1), pp. 7-9.
8. Long, W.R., 1983. “Evaluation of Cartridge Filters for the Removal of Giardia lamblia Cyst
Models from Drinking Water Systems,” Journal of Environmental Health, 45(5), pp. 220-225
9. Farahbakhsh, K. and Smith, D.W., 2004. “Removal of Coliphages in Secondary Effluent by
Microfiltration-Mechanisms of Removal and Impact of Operating Parameters,” Walter Research,
38, pp. 585-592.
10. National Sanitation Foundation (NSF) International Protocol 231, 2003. Microbiological
Water Purifiers, NSF International, Ann Arbor, MI..
11. Gerba, C.P., 1984. “Applied and Theoretical Aspects of Virus Adsorption to Surfaces,”
Advances in Applied Microbiology, 30, pp. 133-168.
12. Sobsey, M.D. and Glass, J.S., 1984. “Influence of Water Quality on Enteric Virus
Concentration by Microporous Filter Methods,” Applied and Environmental Microbiology,
47(5), pp. 956-960.
17
TIP #31-004-0306
13. Farrah, S.R., 1981. “Chemical Factors Influencing Adsorption of Bacteriophage MS2 to
Membrane Filters ” Applied and Environmental Microbiology, 43(3), pp. 659-663.
14. Shields, P.A. and Farrah, S.R., 1982. “Influence of Salts on Electrostatic Interactions
between Poliovirus and Membrane Filters,” Applied and Environmental Microbiology, 45(2),
pp. 526-531.
15. Gerba, C.P. and Naranjo, J.E., 2000. “Microbiological Water Purification without the Use
of Chemical Disinfection,” Wilderness and Environmental Medicine, 11, pp. 12-16.
16. Sobsey, M.D. and Jones, B.L., 1978. “Concentration of Poliovirus from Tap Water Using
Positively Charged Microporous Filters,” Applied and Environmental Microbiology, 37(3), pp.
588-595.
17. Preston, D.R., Vasudevan, T.V., Bitton, G., Farrah, S.R., and Morel, J., 1988. “Novel
Approach for Modifying Microporous Filters for Virus Concentration from Water,” Applied and
Environmental Microbiology, 54(6), pp. 1325-1329.
18. Sobsey, M.D. and Hickey, A.R., 1984. “Effects of Humic and Fulvic Acids on Poliovirus
Concentration from Water by Microporous Filtration,” Applied and Environmental
Microbiology, 49(2), pp. 259-264.
19. Lipson, S.M. and Stotzky, G., 1983. “Adsorption of Reovirus to Clay Minerals: Effects of
Cation-Exchange Capacity, Cation Saturation, and Surface Area,” Applied and Environmental
Microbiology, 46(3), pp. 673-682.
20. Sobsey, M.D. and Cromeans, T., 1984. “Effects of Bentonite Clay Solids on Poliovirus
Concentrations from Water by Microporous Filter Methods,” 49(4), pp. 795-798.
21. Guttman-Bass, N. and Catalano-Shennan, J., 1985. “Effects of Humic Materials on Virus
Recovery from Water,” Applied and Environmental Microbiology, 49(5), pp. 1260-1264.
22. Jacangelo, J.G., Madec, A., Schwab, K.J., Huffman, D.E., and Mysore, C.S., “Impacts of
Feedwater Quality and Operational Conditions on the Removal of Viruses and Bacteria by
Low-Pressure Membrane Filtration,” AWWA Water Quality Technology Conference, 2002.
23. Lukasik, J., Scott, T.M., Andryshak, D., and Farrah, S.R., 2000. “Influence of Salts on
Virus Adsorption to Microporous Filters,” Applied and Environmental Microbiology, 66(7),
pp. 2914-2920.
24. Hill, W.F., Akin, E.W., Benton, W.H., and Metcalf, T.G., 1972. “Virus in Water - II.
Evaluation of Membrane Cartridge Filters for Recovering Low Multiplicities of Poliovirus from
Water Applied Microbiology, 23(5), pp. 880-888.
18
TIP #31-004-0306
25. Porter, M.C. (ed), 1990. Handbook of Industrial Membrane Technology. Noyes
Publications, New Jersey.
26. Mallevialle, J. (ed), 1996. Water Treatment Membrane Processes. McGraw-Hill, Inc., New
York.
27. Weber, R., Chimiel, H., and Mavrov, V., 2003. “Characteristics and Application of Ceramic
Nanofdtration Membranes,” Annals New York Academy of Sciences, 984, pp. 178-193.
28. Clasen, T.F., Brown, J., Collin, S., Suntura, O., and Cairncross, S., (2004). “Reducing
Diarrhea Through the Use of Household-based Ceramic Water Filters: A Randomized,
Controlled Trial in Rural Bolivia,” American Journal of Tropical Medicine & Hygiene, 70(6),
pp. 651-657.
29. World Health Organization, 2002. Managing Water in the Home: Accelerated Health
Gains from Improved Water Supply, ” Geneva.
30. Rice, P.J., 1994. “The Ins and Outs of Carbon Block Filters,” Water Technology, pp. 64-65.
3 1 . U.S. Environmental Protection Agency, EPA/600/R-01/1 10, 200 1 . Controlling Disinfection
By-Products and Microbial Contaminants in Drinking Water, Washington, DC.
32. Koslow, E.E., Nielsen, S.C., and Rook, M.J., 2002. “The Quest for the Holy Grail:
Microbiological Carbon Block Filters,” Water Conditioning & Purification.
33. Lau, B., Harrington, G.W., Anderson, M.A., and Tejedor, I., 2005. “Physiochemical
Aspects of Cryptosporidium Surrogate Removal in Carbon Block Filtration,” Journal of the
American Water Works Association, 97(2), pp. 92-101.
34. Lau, B., Harrington, G.W., Anderson, M.A., and Tejedor, I. “ Cryptosporidium Surrogate
Removal by Point-of-use Carbon Block Filters,” AWWA Water Quality Technology
Conference, 2002.
35. U.S. Army Center for Health Promotion and Preventive Medicine, 2005. Technical
Information Paper: Iodine Disinfection in the use of Individual Water Purification Devices,
Aberdeen Proving Ground, Maryland.
19