Back to Home Page of CD3WD Project or Back to list of CD3WD Publications

Home - English - French - German - Italian - Portuguese - Spanish
                        TECHNICAL PAPER # 32
                     AND SMALL COMMUNITY SYSTEMS
                        By Stephen A. Hubbs
                        Technical Reviewers
                        Dr. F. O. Blackwell
                           Paul S. Fardig
                         Morton S. Hilbert
                 1600 Wilson Boulevard, Suite 500
                   Arlington, Virginia 22209 USA
             Tel:  703/276-1800 * Fax:   703/243-1865
              Understanding Water Supply & Treatment
             for Individual & Small Community Systems
                         ISBN:  0-86619-240-9
            [C] 1985, Volunters in Technical Assistance
This paper is one of a series published by Volunteers in Technical
Assistance to provide an introduction to specific state-of-the-art
technologies of interest to people in developing countries.
The papers are intended to be used as guidelines to help
people choose technologies that are suitable to their situations.
They are not intended to provide construction or implementation
details.  People are urged to contact VITA or a similar organization
for further information and technical assistance if they
find that a particular technology seems to meet their needs.
The papers in the series were written, reviewed, and illustrated
almost entirely by VITA Volunteer technical experts on a purely
voluntary basis.  Some 500 volunteers were involved in the production
of the first 100 titles issued, contributing approximately
5,000 hours of their time.  VITA staff included Maria Giannuzzi
as editor, Suzanne Brooks handling typesetting and layout, and
Margaret Crouch as project manager.
The author of this paper, VITA Volunteer Stephen A. Hubbs, is an
environmental engineer with the Louisville Water Company in Louisville,
Kentucky.  He has worked with the World Health Organization
in Switzerland, Germany, and Holland.   The reviewers are also
VITA volunteers.  Dr. F. O. Blackwell is an associate professor of
environmental health with the East Carolina University School of
Allied Health.  He has worked as a health and sanitation advisor
in Pakistan, and has taught at the American University of
Beirut, Lebanon, School of Public Health.   He is a registered
professional engineer and has worked in the field of environmental
health in 20 countries in Africa, South America, Central
America, and Asia.  Paul S. Fardig specializes in environmental
health and sanitation, with a focus on water supply and sewage
disposal for small towns and villages, including basic sanitary
engineering.  Morton S. Hilbert is Professor and Chairman of the
Department of Environmental and Industrial Health at the University
of Michigan.
VITA is a private, nonprofit organization that supports people
working on technical problems in developing countries.   VITA offers
information and assistance aimed at helping individuals and
groups to select and implement technologies appropriate to their
situations.  VITA maintains an international Inquiry Service, a
specialized documentation center, and a computerized roster of
volunteer technical consultants; manages long-term field projects;
and publishes a variety of technical manuals and papers.
                by VITA Volunteer Stephen A. Hubbs
The design, construction, and operation of small-scale water
treatment systems for individual homes and small communities
represent a significant challenge to public health because of the
wide variety of water quality conditions in developing countries.
Because developing countries often lack expertise for designing
and operating such systems, these systems are often developed
under extreme limitations of both materials and personnel.  For
this reason, any system considered for individual homes or small
communities in developing countries must achieve the basic goals
of water purification through simple design, operation, and maintenance.
For water to be considered suitable for drinking, it should be
aesthetically pleasing; that is, it should look, smell, and taste
good.  It must also be wholesome; that is, it should not contain
any substances that cause sickness or disease (pathogens).  These
two characteristics are mutually important in that water must be
"acceptable" to consumers before they will use it, and free of
harmful agents if it is to be used safely.   It is not uncommon
for consumers to select water that is aesthetically pleasing but
of questionable wholesomeness, over less aesthetically pleasing
water that is free of disease agents.   Consumers tend to judge the
quality of water by the way it looks and tastes, rather than also
taking into account the wholesomeness of the water.
The ideal small-scale water treatment system would be affordable,
simple to design, construct, and operate; and capable of changing
unacceptable water to water that is free of taste, odor, turbidity
(cloudiness or discoloration), and disease agents in a single
process.  Another desirable feature would be for the system to
stop operating automatically if it is producing water that is not
fit for consumption; that is, it should operate only if it is
operating properly.  In reality, however, there is no perfect
system.  Nevertheless, in developing a system, the designer
should always strive to achieve adequate quantity in the least
technically complicated way.
This paper provides guidelines on how to choose a water source,
and how to purify and retrieve water to ensure that it is safe
for human consumption.  Applications are general in nature, relying
on the creativity of the system designer to draw from
whatever resources are available to develop a water treatment
system capable of improving the water supply.
The hydrologic cycle (water cycle) traces the path of water from
the oceans to the atmosphere, rivers, ground, swamps, and eventually
back to the oceans (Figure 1).   As the water progresses

37p03.gif (600x600)

through the various stages of the hydrologic cycle, it is affected
by many factors that determine its ultimate quality.   The water
can be extracted for use at any stage in the cycle; however, the
quantity and quality of water available often limits the user to
only a few choices.  For drinking water, it is important to
select a water source that provides an adequate supply of water
of the highest quality possible.
In areas where air pollution is not a major factor, rain water
can provide a suitable, high-quality source of water.   Typically,
rain is collected from rooftops through gutters and stored in
tanks or cisterns (underground storage vessels).   Because the
roof (or any collection surface) is subject to contamination from
nesting and flying birds and airborne dust, one cannot assume
that this source of water is suitable for consumption.   Underground
storage chambers are subject to infiltration as well as
leakage.  Problems with infiltration can be serious, as water
from nearby outdoor toilets and subsurface sewage disposal systems
can enter the cistern when the water level in the cistern is
low.  For these reasons, rain water must always be disinfected
before it is consumed.  Periodic inspection of the cistern is
recommended, with annual cleaning to remove any sediment that has
The cistern should be sized to provide an adequate supply of
water throughout low rainfall seasons.   In many situations, this
will limit the feasibility of using rain water as a year-round
source of water.  The amount of water available is easily
calculated by multiplying the annual or seasonal average rainfall
(in meters) by the surface area of the collecting surface (in
square meters).  Provisions for screening out large particles
(leaves) and keeping out small animals should be included in any
storage system.
A spring represents a point in the hydrologic cycle where ground
water meets the land surface and flows into a stream.   The water
quality at the point of surfacing is often excellent, as the
water has usually traveled, or percolated, through thick layers
of soil.  In this process of percolation, the water picks up
dissolved minerals (calcium, magnesium, iron, etc.) and is purified
of biological pathogens (disease producing organisms).   The
spring will exhibit varying quantity and quality depending on the
geologic formation in the area.   A continuously flowing spring
that is always clear can provide a good source of drinking water.
In selecting springs as a source of supply, particular caution
should be used in areas of what is called Karst (limestone)
topography.  These areas typically contain many sinkholes, or
depressions, through which surface drainage is transported to the
ground water (Figure 2).  Water entering the ground water by this

37p05.gif (600x600)

path bypasses the percolation process that purifies it.   As a
result, springs in these areas can produce poor-quality water
much like surface water, and must be treated appropriately.
Ground Water
If a stream is located in a sand and gravel stratum, a supply of
suitable drinking water might be obtained easily by drilling or
digging a well into the aquifer that feeds the river (Figure 3).

37p06.gif (600x600)

Since surface streams typically define the lowest hydraulic gradient
in an area, a well dug into the sand and gravel will typically
draw water from highland areas; if these areas have not
undergone extensive development or become contaminated (such as
by dumps and landfills), they will usually provide sanitary
water.  As with spring water, ground water from limestone strata
must be suspect quality.  Depending on local geological conditions,
however, the water may contain unacceptably high levels of
iron, manganese, and/or salt, making it unpalatable.
Ground water can be extracted from any point in the geologic
formation, but the depth and type of cover over the ground water
will determine the feasibility of constructing a well for water
supply.  Water from a well typically exhibits a constant quality.
When the well is properly constructed to eliminate surface contamination,
it can provide an excellent source of drinking water.
Surface Streams
Villages are typically established near a source of water and
transportation routes, as these two factors often determine their
habitability and reason for existence.   The source of water for
villages is typically surface water.   Surface water can be used
to supply water for drinking and washing; it can be a means of
transportation; it can be used for irrigation, livestock watering,
or for sewage disposal.  These multiple uses are often
conflicting, and the water source may not be able to meet all the
demands placed upon it.
While surface streams, rivers, and lakes often represent the most
accessible supply of water to a village, they are also the most
vulnerable to contamination.   Surface water typically has highly
variable water quality, and can be the source of many diseases.
To be suitable for consumption, surface water must always be
treated to remove harmful substances.
For either surface water supply or ground water supply, the point
of water withdrawal should be made as far upstream as possible.
Two principal drawbacks to this concept are (1) people living
below the water source must travel greater distances to obtain
their water; and (2) the higher the source of water, the less
volume of water there is.  A basic understanding of the topography
and geology of the area can help in locating the best
point of water withdrawal.
In selecting a source of water, attention should be given to the
use of land in the immediate watershed and the chance of contamination
to the source.  Problems with unreliable water quality
can be largely eliminated or reduced by avoiding areas that will
likely be contaminated by human waste water, agricultural/livestock
runoff, and industrial discharge.   The most important step
in developing a potable water supply is the selection of the
QRhighest-quality water source possible.
It is difficult to define categorically a particular source of
water as superior to another.   However, ground water supplies and
rain water do have a greater chance of being free of serious
contamination than do surface water supplies.   Of the surface
supplies, springs that provide clear water under all conditions
and that are located in areas that do not have numerous sinkholes
are preferred over surface streams.   Any surface water, including
clear-running mountain streams, can be contaminated by pathogens
and must be treated before use.   No matter what source of water
is being considered, the local factors influencing the water
quality must always be evaluated.   If possible, one should call
upon the local health authorities to analyze the suitability of a
particular water source.
There are many ways of extracting and transporting water from a
source to the point of use.  Water can be taken from streams and
wells by hand and transported in buckets or ceramic vessels.
Where materials and technology are available, water can be pumped
by electric, diesel, or wind-powered pumps and transported
through pipelines.  In situations where the source is located at
an altitude higher than the point of use, the water can be
transported by gravity.  A detailed discussion of these techniques
exceeds the scope of this paper; to obtain this information,
readers are directed to other VITA publications.
Caution should be used in determining how the water will be
extracted and transported.  Extreme care should be exercised to
avoid contamination of the water.   Whenever possible, hand powered
or machine-powered pumps should be installed, and the use of
buckets, which can contaminate the source, avoided.   Pumps also
allow a well to be sealed, eliminating the possibility of foreign
objects or contaminated surface water getting into the well.
This section discusses relatively simple, reliable, and efficient
methods of treating water to remove solids and pathogens.  Methods
for the removal of additional toxic compounds (e.g., heavy
metals, industrial solvents, pesticides) are beyond the scope of
this paper and are not covered here.
Water treatment for any fresh-water system basically involves the
removal of solids, the removal of pathogens (disease-causing
bacteria, viruses, and other microbials), and the removal of
substances that impart bad tastes and odors.   In isolated instances,
additional toxic compounds must be removed before the
water can be drunk.  In supplying water to individual homes and
villages in rural areas, it is therefore far more desirable to
locate a water source free of such toxic agents, because the
removal of such agents can be technically difficult and economically
Solids in water may be of no health concern in themselves.  However,
solids (clay, organic material, etc.) in water can protect
pathogens from disinfection, and result in water quality problems
even in treated systems.  Turbid drinking water is not particularly
appealing, which may lead consumers to select an alternate
source of clear water.  In doing so, however, unknowing consumers
could end up drinking water that is not wholesome, even though it
appears to be of higher quality.   Thus, one goal in the treatment
of water should be the removal of suspended solids.
Solids in water can be divided into three categories:   those that
float, those that sink, and those that are suspended (that is,
they neither float nor sink within reasonable periods of time).
Of these three categories, the suspended solids are the most
difficult to remove.  Floating solids can be avoided by drawing
water from below the surface of the water source.   Solids that
settle without chemical treatment can often be removed by allowing
the water to remain for one day or more in a facility designed
for quiescent conditions (low water velocities).   Suspended
solids, however, must be removed by either chemical or
physical treatment methods.  To remove them in this way involves
more sophisticated equipment and a higher level of maintenance.
Sedimentation; or removal of those solids that sink, was commonly
the only treatment provided to turbid streams through the 1800s.
This process relies on the rate at which the material in the
water settles or sinks, and the retention of water in such a
manner as to allow the material to reach the bottom of the basin.
In sedimentation basins, it is important to remember that the
principal design variable is the surface area of the basin, not
the overall volume.  The basin need only be deep enough to ensure
good hydraulic flow patterns.   Proper design of inlet and outlet
structures is necessary to prevent the system from short-circuiting,
and to avoid the removal of deposits from the floor of the
Sedimentation rates for solids can vary from 10 meters/hour for
heavy silts to less than 0.005 meters/hour (5 mm/hour) for fine
clays.  Thus, the composition of the solids in the water will
determine the feasibility and design criteria for the sedimentation
process.  Fine clay suspensions and water with high color
content can be treated chemically to make the particles settle
more readily.  Such treatment, called chemical coagulation, requires
the availability of chemicals, chemical feed equipment,
and routine sludge removal for proper operation.   Aluminum and
iron salts (alum, ferric sulfate) are typically used when available,
along with organic polymers.   Maintaining these processes
is expensive and requires trained personnel.   Thus, chemical
coagulation is not typically considered for individual/village
water supplies.
A sedimentation basin can be made of any suitable material.  It
can be as simple as a clay pot or as complicated as a concrete
basin with continuous sludge drawoff.   Consideration should be
given to the amount of solids that will be collected in the
basin, and the methods of solids removal that will be used.  If
the solids are to be removed in a batch operation (requiring the
temporary halting of the operation), additional units will be
necessary if a continuous supply of water is required.   In general,
the additional units should be provided if possible, although
this can cause an increase in overall construction costs.
The dimensions of a particular basin are determined by the particles
to be settled, land constraints, the need for long-term
storage, and other physical and economic conditions.   Technical
assistance in designing the facility should be sought whenever
Storing water for extended periods of time can result in the
destruction of bacteria, as well as turbidity removal.   Storage
for two weeks or longer can remove up to 90 percent of disease-causing
organisms.  This process, however, is not effective for
removing all pathogenics, and fine turbidity will remain in
suspension.  In addition, algae may grow in the water during this
time, making the water taste and smell bad.   In general, water
storage is a beneficial pretreatment if algae growth is not a
problem.  Caution must be taken so far as possible to prevent the
contamination of the storage area by human and animal wastes.
Filtration has long been recognized as an effective method of
water purification.  The ancient Egyptians recognized that boiling
and filtering (among other less proven techniques) were
capable of rendering foul water suitable for drinking.   Prior to
1700, it was commonly believed that filtration could remove salt
from sea water.  In the 18th and 19th centuries, many patents
were issued in France and England for various filtration devices,
both small units for in-house use and larger filters for municipalities.
These filters used sand, cinders, charcoal, sponge,
wool, and many other materials.   The earliest mention of the mode
of action in slow sand filters was in the 1840s when an Englishman
noted in a chemistry text that the filter media served to
support "finer materials of mud or precipitate...which...form the
bed that really filters water."   This citation recognizes the
importance of the formation of a filtering layer that must be
allowed to develop on top of the sand before the filter can
operate efficiently.
Slow sand filters (so named because of the relatively slow downward
speed or velocity maintained in the filters) have been noted
as being effective for solids removal and bacterial reduction
for over two centuries.  These early filters were not effective
for highly turbid streams, however, because of the short filter
runs experienced before clogging.   The processes of chemical
coagulation and sedimentation paved the way for the development
of rapid sand filters, which became popular in the early 1900s.
A few modern treatment plants still use slow sand filtration,
although the standard for most large utilities is chemical coagulation,
sedimentation (although direct filtration is becoming
increasingly popular), and rapid filtration through mixed media.
This paper is limited to slow sand filtration only, because it
requires simple operating conditions and generally produces high-quality
water.  Adequate units range from sand-filled drums or
earthen-lined basins to concrete structures with complex under-drain
systems.  Each type of unit suits a particular situation.
A simple filter, designed for domestic use, can be made from a
55-gallon drum and sand.  It can improve the quality of surface
water significantly, as long as initial turbidities are not too
high.  As with any slow filter, the surface of the filter must be
kept wet to maintain the biological growth known as the
"schmutzdecke."  (The schmutzdecke consists of a variety of biologically
active microorganisms that break down organic matter,
while much of the suspended inorganic matter is retained by
straining).  This type of filter can produce 10 to 20 liters of
water per hour if operated continuously, but intermittent operation
is more typical.  In such an operation, the flow rate
through the filter should be limited so as not to exceed optimum
rates (10 to 20 liters per hour).   The filter should be kept
covered to eliminate algae growth and contamination from dust.
For proper filtration, the surface of the filter should always be
kept submerged.
The selection of materials for the construction of a domestic
filter will depend mainly on what resources are available.  If a
55-gallon drum is selected, the interior of the drum must be
protected against rusting.  Containers that have been used for
storing pesticides, herbicides, and other toxic chemicals must
not be used.  The preferred filter medium is sand with an effective
size in the range of 0.15 to 0.35 mm.   Ungraded river sand is
acceptable if nothing else is available.   The sand should be
thoroughly washed by panning to remove very fine sand, clays, and
organic matter.  The sand should be placed in the container in a
layer about 1 meter deep, and arranged with inlet-outlet piping
to allow easy operation.  A typical slow sand filter is shown in
Figure 4.

37p13.gif (486x486)

The design and operation of a slow sand filter for a small village
should be supervised by a qualified person.   The design
criteria should take into account available materials and funds,
as well as the suitability of the water source for filtration.
Regardless of the efficiency of a sand filter for removing turbidity
and reducing bacteria, sand filters alone should not be
considered adequate for the treatment of contaminated surface
waters.  In every case, some form of disinfection should also be
used if the water is to be used for human consumption.
Although sedimentation and filtration can greatly reduce the
amount of bacteria in contaminated water, the reliability of
these two processes to produce water suitable for drinking is
limited.  Many pathogens can survive even after these processes
are operated properly.  Removal of pathogens can be almost negligible
when the processes are carried out improperly.   It is
necessary that any water from a contaminated source be disinfected
before consumption, if at all possible.
Disinfection can be accomplished by mechanical, chemical, and
thermal techniques.  (Other techniques, such as radiation, are
beyond the scope of this paper.)   If the water is sufficiently
free of suspended solids, it can be passed through a small-pore
filter, which is capable of physically blocking the path of
microorganisms.  Certain stone filters have this capability, but
the filtering rate is relatively slow.   Chemical agents, particularly
the halogens (chlorine, bromine, iodine), have been demonstrated
to be highly efficient in killing bacteria.   A universally
recognized method for killing bacteria is boiling, which can
destroy life forms in even turbid suspensions.   Each method of
disinfection has its limitations, which should be recognized
before the technology is adopted.
A recent evaluation of ceramic filters which are capable of
meeting WHO standards for bacterial quality indicated that of all
the strainers tested, only carved stone filters were capable of
yielding acceptable bacterial quality by straining alone.  Other
filters, impregnated with silver, were effective, but the mode of
disinfection was not limited to straining alone.   The carved
stone filter was effective, but it was also relatively heavy and
expensive.  It should be noted that filters that strain out the
test organisms (coliform bacteria) do not necessarily also remove
the pathogenic viruses, which are typically much smaller than
bacteria.  One should be cautious, therefore, in interpreting the
results of straining for pathogen removal based upon indicator
The ability of filtration and straining to remove large numbers
of pathogens should be emphasized.   Properly filtered water is
considered to be far more healthful than unfiltered water.  However,
the complete removal of pathogens cannot be guaranteed.
For this reason, water must undergo further disinfection through
chlorination or boiling.  These two disinfection methods are
discussed in the sections that follow.
Chemical agents such as chlorine, bromine, and iodine have been
used to eliminate waterborne diseases in major water supplies
since the early 1900s.  The most universally supplied agent is
chlorine.  Chlorine combines with water to form hypochlorous acid,
a highly efficient bacteriocide.   The amount of hypochlorous acid
formed by a dose of a chlorine compound will depend on the
amount of organic material and ammonia present, and the pH of the
water.  Typical chlorine amounts in the range of 1.0 mg/1 will
provide adequate protection for fairly clear water; however,
suspended solids can protect pathogens from the disinfectant and
result in incomplete disinfection.   Thus, any water that is
disinfected by chlorine should be free of high levels of suspended
One of the major advantages of the halogen disinfectants is their
ability to form stable residuals, which continue to protect the
water from recontamination.  Depending on the quality of the
water, the residual can persist for as long as a week in the
absence of light.  (The chlorine residual is quickly reduced in
the presence of sunlight.)  One major disadvantage of the residual,
however, is the possibility that the water will develop a
medicinal or chlorinous taste and odor.   The foul taste and odor
are usually not caused by the chlorine (or any other halogen),
but by compounds that have formed with the chlorine.   A common
contaminant, phenol, yields a strong, distinct odor that is
detectable at very low levels.   In certain situations, chlorinous
odors can be removed by increasing the chlorine dosage, which
oxidizes the odor-causing compounds.   In the absence of a sophisticated
laboratory, the suitable amount of chemicals needed for
this purpose can be determined by trial and error.   Table 1
provides instructions for chlorinating drinking water.
Many techniques are available for putting the chemicals into
water, ranging from a single dose into a container to a continuous
feed from some type of storage vessel.   In considering a
technique for small-scale use, the reliability and ease of use
should be given very high regard.   Any technique that is not used
correctly could yield a false sense of security that could be
quite dangerous.
Boiling is perhaps the most well-known and universally applied
method of disinfection.  The common consumption of boiled drinks
(teas) was undoubtedly fostered by the realization that these
drinks were "healthful" (or, more appropriately, non-pathogenic).
Boiling water--even turbid water--for three to five minutes
effectively destroys all pathogens.   However, boiled water often
tastes "flat."  This flat taste can be remedied by allowing the
water to stand for one or more days while exposing it to the air.
Typically, 1 kilogram of wood is required to boil about 1 liter
of water.
Caution should be exercised in storing boiled water, as the
potential for recontamination is quite high.   The water should be
stored in a closed, dark container, preferably in a cool location.
As with any stored water, care should be taken to avoid
contaminating the water when taking water out of the container.
In developing a treatment system for a small water supply, emphasis
should be placed first on securing the highest quality of
water possible (e.g., rain water, ground water, surface water).
Beyond this, any treatment technique that is readily available,
affordable, simple to maintain and operate, and capable of improving
the quality of the water can be used.   In some cases, it
may be impossible to provide chlorination due to the unavailability
of raw material or the unreliability of operation.
Other forms of treatment, although less efficient than chlorination,
may be more reliable and thus provide a consistently better
quality of water than would a less reliable treatment technique.
The most effective treatment technique is one that will not yield
water if it is not operating properly.   To some extent, filtration
systems meet this criterion and thus are very attractive as
a reliable, small-scale form of treatment.   Additional disinfection
is always desirable, however, to ensure pathogen-free drinking
      Table 1.   Amounts of Chemicals Needed to Disinfect
                   Water for Drinking [a]
 Water     Bleaching Powder        High Strength       Liquid Bleach
(m3)       (25-35%) (g)          Cal-Hypochl          (52% sodium
                                   (70%) (g)       hypochlorite (ml)
  1            2.3                     1                    14
  1.2          3                       1.2                  17
  1.5          3.5                     1.5                  21
  2            5                      2                   28
  2.5          6                       2.5                  35
  3            7                       3                    42
  4            9                       4                    56
  5           12                       5                    70
  6           14                       6                    84
  7           16                       7                    98
  8           19                       8                   110
 10           23                       10                  140
 12           28                       12                  170
 15           35                       15                  210
 20           50                       20                  280
 30           70                       30                  420
 40           90                       40                  560
 50          120                       50                  700
 60          140                       60                  840
 70          160                       70                  980
 80          190                      80               1,100
100         230                      100                1,400
120         280                      120                1,700
150         350                      150                2,100
200         470                      200               2,800
250         580                      250                3,500
300         700                      300                4,200
400         940                      400                5,600
500       1,170                      500                7,000
[a]  Approximate dose = 0.7 mg of applied chlorine
     per liter of water.
Note:  For chlorinating drinking water, follow these instructions:
       (1) use one of the chemicals listed in the table, and
       choose the amount according to the quantity of water in the
       distribution tank, cistern, or tanker; (2) dissolve the
       chemicals first in a bucket of water (not more than about
       100 g of calcium hypochlorite or bleaching powder in one
       bucket of water), and pour the solution into the tank (if
       possible, agitate the water to ensure good mixing); and (3)
       repeat this chlorination procedure as soon as the level of
       residual chlorine in the water drops below 0.2 mg per liter.
Source:  S. Rajagopalan.   Guide to Simple Sanitary for
         the Control of Enteric Diseases, (Geneva, World Health
         Organization, 1974.)
Baker, M.N. Quest for Pure Water.   Denver, Colorado:  American
     Water Works Association, 1949.
Mazariegos, J. Fernando, and de Zeissig, Julia Alicia A.  Water
     Purification Using Small Artisan Filters.  Guatemala:  Central
     American Research Institute for Industry, 1981.
Swiss Association for Technical Assistance, ed.  Manual for Rural
     Water Supply.  Zurich, Switzerland:   Swiss Center for Appropriate
     Technology, 1980.
United Nations.  World Health Organization.  "WHO Guidelines for
     Drinking Water Quality," by H.G. Gorchev and G. Ozolins.
     Geneva, Switzerland:   World Health organization, 1982.
United Nations.  World Health Organization.  WHO International
     Reference Centre for Community Water Supply.  The Purification
     of Water on a Small Scale."  WHO Technical Paper no. 3,
     March 1973.
United Nations.  World Health Organization.  WHO International
     Reference Centre for Community Water Supply.  "Preliminary
     List of References on Slow Sand Filtration and Related
     Simple Pretreatment Methods."  The Hague, The Netherlands:
     WHO International Reference Centre for Community Water Supply,
     July 1976.
U.S. Agency for International Development.   Development Information
     Center.   "Methods of Water Treatment."    Technical Note
     no. RWS. 3.M Water for the World.
U.S. Agency for International Development.   Development Information
     Center.   "Determining the Need for Water Treatment."
     Technical Note no. RWS 3.P.1.  Water for the World.
U.S. Agency for International Development.   Development Information
     Center.   "Planning a Water Treatment System."  Technical
     Note no. RWS 3.P:4.  Water for the World.
U.S. Agency for International Development.   Development Information
     Center.   "Designing a Small Community Sedimentation
     Basin."  Technical Note no. RWS. 3.D.2. Water for the
U.S. Agency for International Development.   Development Information
     Center.   "Constructing a Sedimentation Basin."  Technical
     Note no.   RWS. 3.C.2. Water for the World.
U.S. Agency for International Development.   Development Information
     Center. "Designing a Slow Sand Filter."  Technical Note
     no. RWS. 3.D.3. Water for the World.
U.S. Agency for International Development.   Development Information
     Center.   "Constructing a Slow Sand Filter."   Technical
     no. RWS. 3.C.3. Water for the World.
U.S. Agency for International Development. Development Information
     Center.   "Operating and Maintaining Slow Sand Filters."
     Technical Note no. RWS. 3.0.3.  Water for the World.