You are on page 1of 137



All data is provided fro tr!sted so!r"es
An #appropriate te"$nolo%&' is a et$od ($i"$ a"$ieves a desired o!t"oe in t$e siplest and "$eapest
possi)le (a&* To "$oose t$e appropriate (ater treatent te"$nolo%&+ &o! $ave to ,-. /no( t$e $ealt$
pro)les posed )& &o!r (ater s!ppl&+ ,0. st!d& t$e treatents availa)le+ and ,1. de"ide $o( !"$
p!rifi"ation &o! need or (ant*
The purification of unsafe water requires some trained supervision if it is to be done effectively. Such
supervision is rarely available in the villages, nature or crisis of any kind and the procedure tends to be
neglected sooner or later. Under these circumstances every effort must be made to obtain a source that
provides naturally wholesome water and then to collect that water and protect it against pollution by the
methods described here.
Thus, the necessity for treatment of the water may be avoided, and the practical importance of
managing this can hardly be overemphasized.
If the water needs treatment, this should, if at all possible, be done for the whole community and
certainly before, or on entry to the dwelling so that the water from all the taps in the house is safe.
The practice, common in the Tropics, of sterilizing by filtration and boiling! only the water to be used
for drinking, teeth cleaning, etc., though efficient in itself when carefully done! is frequently nullified
by carelessness. "urthermore, children are likely to use water from any tap.
Water so!r"e
#ater derived from the river$lake system is called surface water. The balance of the rain soaks into the ground
and becomes held in underground aquifers porous rock strata!, from which it may eventually emerge as springs,
if it has not been abstracted as ground water. "rom a puri%cation point of view, the two types of water have very
different needs. &round water is essentially clean of suspended material, by virtue of its prolonged %ltration by
porous rock, but it is heavily loaded with dissolved matter hard water! of many different kinds, and has the
mineral taste accepted by many as the taste of pure water. Surface water, by contrast, has very little in the way of
dissolved salts soft water!, but is frequently loaded with % ne suspended solids and colloidal matter, with a lot of
dissolved organic material, which gives the water a marked brown colour and an unpleasant taste. 'ure water is
rarely found in nature See (!. #ater impurities are classified in accordance with their state) suspended, colloidal
and dissolved. *unning water may pick up and transport solid particles of higher density than water+ the higher
the velocity, the bigger the particles that may be picked up. Surface waters during floods are, therefore, at their
most turbid point+ they have ma,imum loads of suspended matter. "iner particles colloids! may not be visible to
the naked eye but could impart colour and turbidity to the water. -olloids remain in suspension even when the
water is at rest. In its passage over or under the ground, water may pick up substances which are soluble. .mong
these dissolved solids, the most common in natural waters are bicarbonates, carbonates, sulfates, chlorides and
nitrates of -alcium, /agnesium, Sodium, 'otassium, Iron and /anganese. The products of decomposition of
organic wastes such as nitrates and nitrites may be regarded as an indication of organic pollution. The presence
of bacteriological indicators such as Escherichia coli 0. coli! provides positive proof of the faecal origin of such
pollution. .lgae may grow in water under certain conditions and they may impart ob1ectionable tastes and
odours to drinking water+ the removal of algae is essential but often difficult. The presence of Iron or /anganese
may also impart tastes or odours to water and may stain articles that are washed with it. Hardness, caused by
bicarbonates, sulphates and chlorides of -alcium and /agnesium forms insoluble precipitates with soap and
causes the deposition of scale. Sulphates of /agnesium and Sodium, if present in e,cess, act as la,atives+
chlorides, in concentrations higher than 233 milligrammes per litre, give water a salty taste, while fluorides, in
concentrations above 4.2 milligrammes per litre, are undesirable+ in concentrations above ( milligrammes per
litre they may cause mottling of teeth. 5etergents and pesticides may find an easy way into raw water and are
ob1ectionable if present in e,cess. #ater with a high content of dissolved -67, a low p8 value and low alkalinity
is corrosive and is apt to attack metals.
Drin/in% (ater "ontainants
The #orld 8ealth 6rganization has identi%ed 927 substances that can be present in tap water.
#ater authorities are obliged to monitor the levels of only :: of them, and bottled water
manufacturers only 7;. Up until about (3 years ago, lead was thought to be the only dangerous
pollutant in drinking water. Today, in addition to lead, pesticides, bacteria, viruses, coliphage,
nitrates, chlorine, chloro<organics and aluminium must be added to a growing list of health hazards.
-ontaminants typically found in water are shown in "igure =.4(.
>acterial diseases, which are usually spread through water contamination, are a cause for ma1or
concern throughout the world. -holera is very widespread, despite the fact that huge numbers about
one billion! of the cholera bacteria are needed to cause infection. The parasitic protozoan
Cryptosporidium is also widely distributed in nature, infecting a wide range of hosts including farm
animals and pets. The protozoa get into the water supply from animal e,crement. The problem is that
they form a protective stage known as oocysts, which allow them to survive for long periods in water
up to perhaps eighteen months! whilst waiting to be consumed by a suitable host. 6nce infected, a host
becomes a lifetime carrier, liable to relapses ? and it only takes a single Cryptosporidium oocyst to
cause infection. In normal patients, Cryptosporidium gives rise to self<limiting gastro<enteritis, which
can last for up to two weeks. If the patient is immunosuppressed, infection is life threatening.
Cryptosporidium is widespread in drinking water resources. The cysts are only =?9 $m across, and
are very dif%cult to detect and remove from water by conventional water treatment but are easily
removable by micro or ultra% ltration!. They also resist chlorine.
Giardia lamblia is another oocyst that causes disease, but, unlike Cryptosporidium it is treatable
with antibiotics. >oth Giardia and Cryptosporidium can be killed by boiling the water for twenty
minutes, but, unfortunately, boiling concentrates metals such as lead and aluminium in the water.
There are no set rules as to the acceptable quality for potable water, but certain guidelines have been
laid down See If these guidelines are not e,ceeded, no action is necessary. Short<term deviations above
the guideline values do not necessarily mean that the water is unsuitable for human consumption+ the
amount by which, and the period for which, any guideline value may be e,ceeded without affecting
public health depends on the specific substance involved. In developing drinking water standards for
any refugee situation it is necessary to take into account its geographical, socio<economic, dietary and
environmental conditions.
. common treatment plant consists of many processes) screening, coagulation, flocculation,
sedimentation, filtration and disinfection, each of which performs one main function although it may
incidentally assist with some other. #ater impurities are removed in order of size, the bigger ones being
eliminated first. @ot all water contains all the impurities, therefore not all water requires all the
treatment processes. #henever necessary, impurities are removed as follows)
i! floating ob1ects by screening+
ii! algae if present! by straining+
iii! e,cessive iron, manganese and hardness by chemical precipitation+
iv! normal suspended solids by settling sedimentation!+
v! e,cessive bacterial pollution by pre<chlorination+
vi! the remaining fine particles and some more bacteria by filtration+
vii! final bacteria, surviving filtration, by chlorination.
.ll these processes overlap to some e,tent and many need au,iliary processes to be fully effective.
Eer%en"& (ater stora%e
0ven with the use of a hygienic connecting hose, there is a strong chance that the initial Aow
from the source will be contaminated. It is thus recommended that an in<line %lter be used in the
feed line from the mains to the tank.
"urther contamination is then inevitable on standing. Tests conducted with a large plastic
container %lled with tap water showed an initial bacterial count of 4;$ml, which, after only a week,
had risen to =2,333$ml.
Thus, the equipment commonly used to store and dispense fresh water in mobile environments
leaves much to be desired, unless it is sub1ect to additional treatment. This may range from the
simple use of purifying chemicals added to the water, or %ltration or a complete water treat ment
system. 6f these, %ltration is by far the most attractive of the low cost methods.
'ouring water through a clean piece of cotton cloth will remove a certain amount of the suspended
silt and solids. It is important that the cloth used is clean, as dirty cloth may introduce additional
pollutants. Specifically made monofilament filter cloths may be used in areas where guinea<worm
disease is prevalent.
Such cloths remove organisms known as copepods, which act as intermediate hosts for the guinea<
worm larvae. The cloth must always be used with the same surface uppermost. The cloth may be
cleaned using soap and clean water.
2tora%e and settleent
#hen water is stored for a day in safe conditions, more than 23B of most bacteria die.
"urthermore, during storage, the suspended solids and some of the pathogens will settle to the
bottom of the container. The container used for storage and settlement should have a lid to avoid
recontamination, but should have a neck wide enough to facilitate periodic cleaning. "or e,ample a
bucket with a lid could be used for this purpose.
#ater should be drawn from the top of the container where it will be cleanest and contain less
pathogens. Storage and settlement for at least =; hours also eliminates organisms called the
cercariae, which act as intermediate host in the life cycle of bilharziasis schistosomoasis), a water<
based disease prevalent in some countries. Conger periods of storage will leader to better water
The process to eliminate all impurities present as suspended particles which are carried along by
flowing raw waters but which will settle in quiescent or semi<quiescent conditions is called
sedimentation. It is usually considered the minimum treatment for turbid surface waters+ if 7= hours can
be alotted for sedimentation, clarified waters can be directly chlorinated. The sunDs bactericidal effect
has also been documented. >elow a certain particle size, depending on the material concerned, settling
velocity becomes very small and therefore sedimentation becomes unfeasible. This is the case for
colloidal matter, which, as it has been discussed, requires coagulation and flocculation before the
sedimentation process. Sedimentation facilities normally operate under continuous flow to)
i! achieve quiescent conditions in the settling zone+
ii! ensure uniform flow across the settling zone+
iii! obtain uniform concentration of solids as flow enters the settling zone+
iv! ensure that solids entering the sludge zone are not re<suspended.
The efficiency of these structures basically depends on the ratio between the influent flow rate and the
surface area of the tank+ their design should be based on the settling velocities of the particles to be
removed, a factor that should be assessed by qualified technicians and requires the collection of water
samples and specialized laboratory equipment See 47.;!. The main tanks found in practice are shown
in "ig. (3. 8orizontal tanks are compact+ sludge is removed from them under hydrostatic head. -ircular
tanks offer the advantage of simpler scraping mechanisms but are not so compact. The vertical flow
tanks, like the one shown in the figure, operate with a sludge blanket which serves to strain out particles
smaller than those that could be removed by sedimentation alone at the flow rates employed.
Fig. 30a Type o !edimentation Tan"s #hori$ontal lo%)
Fig. 30b Type o !edimentation Tan"s #vertical lo%)
Fig. 30c Type o !edimentation Tan"s #radial lo%)
Free3in% of (ater
-ontrary to an all too common opinion, ordinary freezing of water, though it may retard the
multiplication of bacteria, does not kill them, and ice from a household refrigerator is no safer than
the water from which it was made. The principal methods of purifying water on a small scale are
boiling, chemical disinfection, and filtration. These methods may be used singly or in combination,
but if more than filtration is needed the boiling or chemical disinfection should be done last. 0ach
method is discussed briefly below.
"ollowing this general introduction are descriptions of a variety of water purification technologies)
boiler for drinking water, chlorination of polluted water, water purification plant, and sand filter.
2olar disinfe"tion
Ultra<violent rays from the sun are used to inactivate and destroy pathogens present in water. "ill
transparent plastic containers with water and e,pose them to full sunlight for about five hours or
two consecutive days under 433B cloudy sky!. 5isinfection occurs by a combination of radiation
and thermal treatment. If a water temperature of least 23
- is achieved, an e,posure period of one
hour is sufficient. Solar disinfection requires clear water to be effective.
.n enhanced e,ample is the S65IS system, whereby half<blackened bottles are used to increase the
heat gain, with the clear side of the bottle facing the sun, as shown above.
>oiling is the most satisfactory way of destroying disease<producing organisms in water. It is
equally effective whether the water is clear or cloudy, whether it is relatively pure or heavily
contaminated with organic matter.
>oiling destroys all forms of disease<producing organisms usually encountered in water, whether
they be bacteria, viruses, spores, cysts, or ova. To be safe the water must be brought to a good
ErollingE boil not 1ust simmering! and kept there for 42<73 minutes.
>oiling drives out the gases dissolved in the water and gives it a flat taste, but if the water is left for
a few hours in a partly filled container, even though the mouth of the container is covered, it will
absorb air and lose its flat, boiled taste.
It is wise to store the water in the vessel in which it was boiled. .void pouring the water from one
receptacle to another with the ob1ect of aerating or cooling it as that introduces a risk of
Boiler for Drin/in% Water
The boiler described here "igure 4! will provide safe preparation and storage of drinking water in
areas where pure water is not available and boiling is practical.
#hen the unit was used in work camps in /e,ico, a 73;<liter 22<gallon! drum supplied 73 persons
with water for a week.
Tools and /aterials
73;<liter 22<gallon! drum
43mm ($=E! pipe nipple, 2cm 7E! long
>ricks for two (3cm 4D! layers to support drum
Sand and 4 sack of cement for mortar and base of fireplace
Carge funnel and filter medium for filling drum
/etal plate to control draft in front of fireplace 4Fmm ($=E! valve, preferably all metal,
such as a gate valve, that can withstand heat.
The fireplace for this unit see "igure 7! is simple. It should be oriented so that the prevailing wind
or draft goes between the bricks from the front to the back of the drum. . chimney can be provided,
but it is not necessary.
#hen filling the drum, do not fill it completely, but leave an air space at the top as shown in "igure
4. *eplace the funnel with a filler plug, but leave the plug completely loose.
#ater must boil at least 42 minutes with steam escaping around the loose filler plug. /ake sure that
the water in the pipe nipple and valve reach boiling temperature by letting about 7 liters 7 quarts!
of water out through the valve while the drum is at full boil.
.eration is a treatment process in which water is brought into close contact with air for the primary
purpose of increasing the o,ygen content of the water. #ith increased o,ygen content)
volatile substances such as hydrogen sulphide and methane which affect taste and odour are
carbon dio,ide content of water is reduced+ and
dissolved minerals such as iron and manganese are o,idized so that they form precipitates,
which can be removed by sedimentation and filtration.
Fi%!re 0* Aerator tra&s
The close contact between water and air required for aeration can be achieved in a number of ways.
.t a household level, rapidly shake a container part<full of water, for about five minutes and then
stand the water for a further (3 minutes to allow any suspended particles to settle to the bottom.
6n a larger scale, aeration may be achieved by allowing water to trickle through one or more well<
ventilated, perforated trays containing small stones, as shown in "igure 7. .gain, the water must be
collected in a container and allowed to stand for about (3 minutes to settle suspended particles.
&eration is practiced to add o,ygen from the atmosphere to water and to liberate undesirable gases
such as carbon dio,ide or hydrogen sulphide. It is commonly done by splashing the water over trays
or by blowing air bubbles through the water. It is a viable and cheap means of controlling tastes,
odours and corrosion but its results may not be considered complete in all cases. .mong the
equipment normally used for aeration, the most common are some special nozzles which direct thin
1ets of water into metallic plates to produce fine sprays e,posing water to the atmosphere+ cascade<
type aerators which create turbulence in thin streams of water flowing down+ tray<type aerators
consisting of some five perforated trays, increasing in size from top to bottom, where water falling
from tray to tray! is e,posed to air+ and diffused air aerators, which are tanks where air is bubbled
upwards from diffuser pipes laid on their floor. The latter method is the most efficient+ the amount
of air needed may be regulated+ the tanks are normally about = metres deep and have a retention
time of about 42 minutes. .mong all the methods, however, trays are the most commonly used
because of their low cost, simple operation and reasonably high efficiency.
Iodine is also a good disinfecting agent. Two drops of ordinary tincture of iodine are sufficient to
treat 3.F2 liter 4 quart! of water. #ater that is cloudy or muddy, or water that has a noticeable
color even when clear, is not suitable for disinfection by iodine.
"iltering may render the water fit for treatment with iodine. If the water is heavily polluted, the dose
should be doubled. Though the higher dosage is harmless it will give the water a medicinal taste. To
remove any medicinal taste add 9 percent solution of sodium thiosulphate in a quantity equal to the
amount of iodine added. Iodine compounds for the disinfection of water have been put into table
form, for e,ample, E'otable .qua Tablets,E E&lobalineE and EIndividual #ater 'urification
TabletsE+ full directions for use are given on the packages.
These tablets are among the most useful disinfection devices developed to date and they are
effective against amoeba cysts, cercariae, leptospira, and some of the viruses.
-hlorination $ at the end of this document.
2and Filter
a. Do(n4flo( sand filter
Surface water from streams, ponds, or open wells is very likely to be contaminated with leaves and
other organic matter. . gravity sand filter can remove most of this suspended organic material, but
it will always let rivus and some bacteria pass through. "or this reason, it is necessary to boil or
chlorinate water after it has been filtered.
>y removing most of the organic matter, the filter)
*emoves large worm eggs, cysts, and cercariae, which are difficult to kill with chlorine.
.llows the use of smaller and fi,ed doses of chlorine for disinfection, which results in
drinkable water with less taste of chlorine.
/akes the water look cleaner.
*educes the amount of organic matter, including living organisms and their food, and the
possibility of recontamination of the water.
.lthough sand filtration does not make polluted water safe for drinking, a properly built and
maintained filter will make chlorination more effective. Sand filters must be cleaned periodically.
The household sand filter described here should deliver 4 liter 4 quart! per minute of clear water,
ready for boiling or chlorinating.
#'(HC) data * +ther types o sand ilter include the pac"ed drum ilter that can be improvised i
drums and sand are available and may be a very good %ay o providing limited ,uantities o saer
%ater ,uic"ly to cover small %ater demands #at household or health post levels, or instance). -n
these ilters, %ater passes do%n through sand on a . cm. layer o gravel and is dra%n o at a rate
that should not e/ceed 00 litres per hour or a 100 litre drum2 iniltrated %ater e,ual to the amount
dra%n o is added to the top.)
Tools and /aterials
Steel drum) at least :3cm wide by 92cm 7D, 7F 4$7E!
Sheet metal, for cover) 92cm 7F 4$7E! square
#ood) 2cm , 43cm 7E , =E!, ( meters F.;D! long
Sand) 3.7 cubic meter 9 cubic feet!
>locks and nails
'ipe, to attach to water supply
6ptional) valve and asphalt roofing compound to treat drum
The gravity sand filter is the easiest type of sand filter to understand and set up. It uses sand to
strain suspended matter from the water, although this does not always stop small particles or
6ver a period of time, a biological growth forms in the top 9.2cm (E! of sand. This film increases
the filtering action. It slows the flow of water through the sand, but it traps more particles and up to
F2 percent of the bacteria. The water level must always be kept above the sand to protect this film.
Sand filters can get partially clogged with organic matter+ under some conditions this can cause
bacterial growth in the filter. If the sand filter is not operated and maintained correctly, it can
actually add bacteria to the water.
The drum for the sand filter shown in "igure = should be of heavy steel. It can be coated with
asphalt material to make it last longer.
The 7mm ($(7E! hole at the bottom regulates the flow) it must not be made larger.
The sand used should be fine enough to pass through a window screen. It should also be clean+ it is
best to wash it.
The following points are very important in making sure that a sand filter operates properly.
Geep a continuous flow of water passing through the filter. 5o not let the sand dry out,
because this will destroy the film of microorganisms that forms on the surface layer of sand.
The best way to ensure a continuing flow is to set the intake so that there is always a small
Screen the intake and provide a settling basin to remove as many particles as possible before
the water goes into the filter. This will keep the pipes from becoming plugged and stopping
the flow of water. It will also help the filter to operate for longer periods between cleanings.
@ever let the filter run faster than (.: liters per square meter per minute = gallons per square
foot per hour! because a faster flow will make the filter less efficient by keeping the
biological film from building up at the top of the sand.
Geep the filter covered so that it is perfectly dark to prevent the growth of green algae on the
surface of the sand. >ut let air circulate above the sand to help the growth of the biological
#hen the flow becomes too slow to fill daily needs, clean the filter) Scrape off and discard
the top 4$7cm 4$=E! of sand and rake or scratch the surface lightly.
.fter several cleanings, the sand layer should be returned to its original thickness by adding clean
sand. >efore doing this, scrape the sand in the filter down to a clean level. The filter should not be
cleaned more often than once every several weeks or even months, because the biological growth at
the top of the sand makes the filter more efficient.
-onventional downAow sand %lters are effective for liquid<solid separation at Aow rates up to about
42 m
of %lter area, although higher rate downAow %lters are available. #ith proper selection
of %lter media, gelatinous as well as granular suspended matter can be %ltered out, without a rapid
differential pressure build<up.
The bed is cleaned by a reverse, upward Aow of %ltrate water, suf%cient to e,pand and Auidize
the granules of the bed. #hen suf%cient backwash liquid has passed through the bed, the bed
particles settle back into place under the inAuence of gravity. If the particles are all of the same
material, then the largest ones will settle at the bottom of the bed and the smallest ones at the top,
which is the wrong way round as far as %ltration is concerned, which is best achieved under downA
ow conditions by having the largest pores created by the largest particles! at the top of the bed, %rst
meeting the incoming raw water.
Typical %lter media for the down Aow %lter consist of selected silica sands, and coal or
anthracite, which are tough inert solids, available in a range of particle sizes. 6ne solution to the
problem of matching the pore sizes in the bed is to use layers of different solids, with different
densities. If the denser material also has the smallest particle size, then the layers will resettle after
backwashing with the %nest at the bottom and the coarsest on top. The effect in a two<layer bed is
shown in "igure =.4 , where the trapped solid concentrations are plotted for a conventional bed as
/aterials for use in multi<layer downAow beds include anthracite, with a speci%c gravity of 4.=,
Aint sand 7.:2! and garnet (.;(!. /agnetite =.F! can be used for a fourth layer if necessary.
. two<stage multi<media %ltering system, shown in "igure =.7 has been developed to treat
turbid surface waters coming from rivers, lakes, reservoirs or the sea, but which are low in colour,
iron and manganese. #hilst operational, o,idizing and coagulant solutions are in1ected into the
primary upper! vessel, and a further coagulant is also in1ected prior to the second lower! vessel.
The primary %lter medium comprises anthracite, supported by a layer of silica sand. The secondary
%lter medium comprises a medium<sized layer of silica sand, supported by the %nal polishing layer
of garnet or barium sulphate. The system has four programmable self<cleaning steps, using the
normal backwash principle.
). Up5o( sand 6lters
Simple household filters may be put together inside clay, metal or plastic containers. The vessels
are filled with layers of sand and gravel and pipe work arranged to force the water to flow either
upwards or downwards through the filter. "igure = shows a modified simple upward rapid flow
filter. . filter such as this could be built from a 733 litres drum. It has a filter bed made up coarse
sand of about 3.(m depth! of grain size between ( and =mm diameter, and supported by gravel
covered by a perforated metal tray. The effective filtration rate of such a filter could be as high as
7(3 litres per hour. Such filters must be dismantled regularly to clean the sand and gravel and
remove any settled silt. The frequency of cleaning is dependant on the level of turbidity of the raw
water. "urthermore, such filters are not effe"tive at reovin% t$e pat$o%ens . Therefore the water
must be disinfected or stored for =; hours in order to make it safe.
The alternative approach to match Aow direction with pore size is to undertake the %ltration in
upwards Aow. #ith an upAow sand %lter, Aow is from the bottom through to the top of the bed. The
result is that the entire bed depth is utilized to trap solids, with the %ne top layer acting as the %nal
cleanup zone. This gives a suspended solids capacity of 7F to =; kg$m
, depending on the density of
the suspended matter, a greater capacity per unit of surface area than in a conventional downA ow
%lter. . bed stabilizer is necessary at the top of the bed, to keep it in place during the high on<
stream Aow rates, in order to take full advantage of the bed Hs capacity to retain trapped solids. The
bed is cleaned by upwards Aow of backwash liquid, but the bed is e,panded by air agitation prior to
washing, in order to achieve ma,imum cleaning ef%ciency. This mode of operation allows the upA
ow %lter to handle turbid waters at high Aow rates with longer cycle lengths, while ensuring good
cleaning cycles.
U@I-0" upflow sand filter
". 7ovin% )ed 6lter
The ma1or development of the deep<bed %lter has been to allow the bed material to move
continuously down through the %lter vessel, and then to be carried back up to the top of the bed,
through a cleansing zone. /anufacturers have concentrated on the development of the moving bed
%lter, which provides better cleansing of the bed, and which does not have to be shut down for back
washing. It is thus a truly continuously operating %lter. "igure =.2 shows an e,ample of a typical
continuous self<cleaning sand %lter.
The sand washer is designed to clean each grain of sand by scouring. The dirt particles are
Aoated upward by the air bubble action and leave the %lter over the sludge weir 43!, and are carried
away by some of the wash water. This re1ect water is a small fraction of the total water fed to the
%lter, and is returned to the %lter inlet after the sludge is dewatered.
It is recommended that a %lter screen be used ahead of and in con1unction with %lters of this type.
-ontinuous self<cleaning %lters of this kind are considered to be one of the most reliable types of
bulk water % lter available, with low plant costs and high clari%cation ef%ciencies.
It has no mechanical moving parts yet it is
possible to obtain feed Aow rates of up to 72
$h per square metre of %ltering area. The
main structure of the %lter is a cylindrical tank
with a conical bottom. In operation, the raw
water is fed in 4! where an inlet system, and
7! evenly distributes the Aow into the %lter
bed. This bed is made up of sand of a
predetermined grain size, selected according
to the nature and quantity of suspended solids
in the raw water. The water Aows through the
sand bed and leaves the %lter at the overAow
weir 2!. The sand bed moves continuously
downwards, being sucked from the bottom by
an airlift pump :!, which carries the dirty
sand upward to the top of the %lter into the
sand washer 9!. The washed sand falls back
into the %lter through the chamber ;!. The
pressure drop across the %lter remains at a
constant low value by virtue of the continuous
washing process that keeps the %lter bed clean.
The heart of the system is the sand washer,
which, via the airlift, receives a concentrated
mi,ture of sand, water and dirt particles+ the
dirt is separated from the sand by Aotation due
to the action of micro air bubbles generated at
the air diffuser F!.
2lo( sand filters
Slow sand filters are used in water purification for treating raw water to produce a potable product.
.lthough they are often the preferred technology in many developing countries, they are also used
to treat water in some of the most developed countries such as the UG where they are used to treat
water supplied to Condon and other 0uropean towns like 'aris. Slow sand filters now are also being
tested for pathogen control of nutrient solutions in hydroponic systems. Slow sand filtration is a
water purification process in which water is passed through a porous bed of filter medium. Slow
sand filters are typically characterized by certain design components) the supernatant water above
the filter sand that provides hydraulic head for the process!, filter sand varying in depth, the
underdrain medium usually consisting of graded gravel!, and a set of control devices Sims!. In a
mature sand bed, a thin upper sand layer called a Schmutzedecke forms. The Schmutzedecke
consists of biologically active microorganisms that break down organic matter while suspended
inorganic matter is removed by straining Ian 5uk!.
Slow sand filters are distinguished from rapid sand filters by the biologically active sand medium
including the Schmutzedecke!, and slow detention times.
*apid sand filters utilize primarily a physical removal process, are periodically backwashed for
cleaning, and operate with long detention times. Slow sand filters are cleaned by periodically
scraping the e,isting Schmutzedecke Ian 5uk!. . distinguishing feature of slow sand filters is the
presence of a thin layer, called the schmutzdecke, which forms on the surface of the sand bed and
includes a large variety of biologically active microorganisms. . slow sand filter comprises
appro,imately 4.7 m depth of fine sand supported on two or three gravel layers. It is a very simple
and effective technique for purifying surface water. It will remove practically all of the turbidity
from the water as well as most of the pathogens without the addition of chemicals.
Effe"tiveness of t$e Te"$nolo%&
This is a proven technology that is very effective in removing both suspended materials and
bacteria. 8owever, the improvement in water quality brought about by slow sand filtration will
differ from place to place due to raw water quality, sand grain size, rate of filtration, temperature,
and the o,ygen content of the water. .n indication of the purification effect of a slow sand filter is
summarized in Table 7 and shown in "igure 44.
TABLE 0* Perforan"e of 2lo( 2and Filters*
Paraeter of (ater 8!alit& P!rifi"ation effe"t of slo( sand filtration
-lour (3B to 433B reduction
Turbidity is generally reduced to less than 4
"aecal coliforms
F2B to 433B, and often FFB to 433B,
reduction in the level of faecal coliforms
Iirtual removal of cercariae of schistosomes,
cysts and ova
Iiruses Iirtually complete removal
6rganic matter :3B to 92B reduction in -65
Iron and manganese Cargely removed
8eavy metals (3B to F2B reduction
.! Desi%n n!)er - < They are typically 4 to 7 metres deep, can be rectangular or cylindrical in cross
section and are used primarily to treat surface water.
The length and breadth of the tanks are determined by the flow rate desired by the filters, which typically have a
loading rate of 3.4 to 3.7 metres per hour or cubic metres per square metre per hour!.
#hile many municipal water treatment works will have 47 or more beds in use at any one time, smaller
communities or households may only have one or two filter beds.
In the base of each bed is a series of herringbone drains that are covered with a layer of pebbles which
in turn is covered with coarse gravel. "urther layers of sand are placed on top followed by a thick layer
of fine sand. The whole depth of filter material may be more than 4 metre in depth, the ma1ority of
which will be fine sand material. 6n top of the sand bed sits a supernatant layer of raw, unfiltered water.
B. Desi%n n!)er 0
"igure 7 is a schematic of a common cross section of a slow sand filter.
"igure 7. Typical cross section of a slow sand filter.
3ra%ing by 4ary )ust
The supernatant serves two distinct purposes. "irst, it provides a head of water sufficient to pass the
raw water through the filter bed. Second, the supernatant creates a detention time of several hours
for the treatment of the raw water. The supernatant should not be considered as a reservoir for
sedimentation. If the raw water has a high content of suspended mater, then pretreatment should be
considered to prevent rapid clogging of the filter bed. The supernatant depth is typically a meter
Ian 5uk!.
5epth of a filter bed ranges between 4.3 and 4.= meters Ian 5uk!.
The clean filter medium from a slow sand filter in a treatment plant near 'aris, "rance can be seen
in "igure (.
The underdrain system serves two purposes. It provides unobstructed passage for the collection of
treated water and it supports the bed of filter medium. It is important that the underdrain system
provide a uniform velocity over the entire filter area Ian 5uk!. The underdrain gravel is placed so
that the finest gravel is directly underneath the sand and the coarsest gravel is surrounding the
underdrain pipes or covering the underdrain block 'yper!. This prevents the filter sand grains from
being carried into the treated water system. .n e,ample of support media and underdrain system is
shown in "igure =.
C. Desi%n n!)er 1
The walls of the filter can be constructed of concrete or stone. Sloping walls dug into the earth,
supported or protected by chicken wire reinforcement, and lined with sand or a sand<bitumen mi,ture
could be a cost<effective alternative to concrete. Inlets and outlets should be provided with controllers to
keep the raw water level and the filtration rate constant. >ottom drains consist of a system of manifolds
and lateral pipes "igure 43!. The filtration rate usually employed in developing countries is between
7.2 and :.3 m
$day. 8igher rates may be used, but should be tested to ensure that the higher rates
yield a good quality product water. The system should be designed for fle,ibility. 8ighly turbid water
may need some form of pre<treatment such as settling or rough filtration.
8ome filters using the principles of slow sand filtration can be constructed using a 733 l drum. Such
systems are a common way of treating water for household drinking. The drum has a layer of ( mm
sized gravel on the bottom surrounding the outlet pipe, :33 mm of fine sand on top of the gravel, and a
space for the inlet water. .n overflow and a lid to prevent mosquito breeding and algal growth should
be provided. 6ften large stones are placed on top of the sand to prevent the flow from the inlet pipe
from scouring the sand. This type of filter is often used in rainwater systems see above! to remove
sediment before the rainwater is conveyed to a storage tank.
"igure 43. Simple drawing of slow sand filter Schultz, 4F;=!.
Slow sand filters are often used in both urban and rural water supply systems on larger islands. 6n Ca
5igue, Seychelles, the installation of slow sand filters has proven to be the most cost<effective method
for the treatment of raw water. #ater supplied to .pia and part of the west coast of Upolu, #estern
Samoa, is treated using slow sand filters which work well under normal flow conditions. In *arotonga,
-ook Islands, water from many of the stream sources is treated with small versions of a slow sand filter,
while, in "rench 'olynesia, slow sand filters have been operated successfully in small communities
supplied by stream catchments on the high islands. . small<scale slow sand filter unit has been
designed and operated in a 233 mm steel pipe.
2and for slo( sand (ater filter
From source num. 5 $ Slow sand filters can frequently be constructed largely from locally<available
materials. The effective size of the sand used in slow sand filters is about 3.7 mm, but may range
between 3.42 mm and 3.(2 mm, and with a coefficient of uniformity of between 4.2 and (.3. In
contrast, the range of effective size for the rapid sand filters described in the preceding section is
3.(2 mm to 4.3 mm, with a coefficient of uniformity of 4.7 to 4.9.
"igure 44. 0ffect of sand size on removal of total coliform bacteria in slow sand filtration Iisscher, et
al., 4F;9!.
The physical characteristics of a sand bed are important in maintaining the slow sand filters
efficiency. The effective size is the size opening that will pass ten percent by weight of the filter
material 8aarhoff!. 0ffective sizes in the range of 3.42 mm to 3.(2 mm are used Ian 5uk!. The
uniformity coefficient is the ratio of the size openings that pass si,ty percent of filter material to the
size openings that pass ten percent of filter material, e.g. the effective size 8aarhoff!. Uniformity
coefficients range between two and five+ most facilities maintain uniformity coefficients less than
three 8aarhoff!. The filter medium itself should consist of inert and durable grains+ sand should be
washed so that it is free of clays, loams, and organic matter.
Biolo%i"al and P$&si"al 7e"$aniss
Slow sand filters work through the formation of a gelatinous layer or biofilm! called the $&po%eal
layer or !chmut$dec"e in the top few millimetres of the fine sand layer.
The !chmut$dec"e is formed in the first 43?73 days of operation
and consists of bacteria, fungi,
protozoa, rotifera and a range of aquatic insect larvae. .s a !chmut$dec"e ages, more algae tend to
develop and larger aquatic organisms may be present including some bryozoa, snails and .nnelid
worms. The !chmut$dec"e is the layer that provides the effective purification in potable water
treatment, the underlying sand providing the support medium for this biological treatment layer. .s
water passes through the !chmut$dec"e, particles of foreign matter are trapped in the mucilaginous
matri, and dissolved organic material is adsorbed and metabolised by the bacteria, fungi and
protozoa. The water produced from a well<managed slow sand filter can be of e,ceptionally good
quality with F3<FFB bacterial reduction.
>iological activity in the sand bed is not well understood. Scientists have a vague idea of the
processes involved, but specific interactions are still unknown. Suggested biological removal
mechanisms are predation, scavenging, natural death and inactivation, and metabolic breakdown
8aarhoff!. In the Schmutzedecke, algae, plankton, diatoms, and bacteria break down organic
matter through biological activity. It has been hypothesized that as the raw water passes through the
bed, it constantly changes direction. Thus, the sand grains develop a uniform sticky layer of organic
material that absorbs to the particles by various attachment mechanisms. The sticky layer around
the sand grains is biologically active bacteria, protozoa, bacteriophages! and the organic impurities
are biologically converted to water, carbon dio,ide and harmless salts. .ccording to a study by
-ollins, the bacterial concentrations in the Schmutzedecke were a function of the elapsed time and
potential for cell growth rather than the filtration of free<living bacteria from the source water
-ollins!. The biologically active section of the entire filter bed e,tends 3.=<3.2 m downward from
the surface of the Schmutzedecke Ian 5uk!.
'hysical processes are also inherent to slow sand filter mechanisms. .s the biological activity of
the filter bed decreases, the physical processes of adsorption and chemical o,idation are the primary
mechanisms Ian 5uk!. .dsorption accounts for removals that were traditionally thought to be
purely biological. "or e,ample, the removal of chlorinated organics and the distribution of viruses
are thought to follow adsorption isotherms 8aarhoff!. "urthermore, suspended inorganic matter
may be removed by the physical process of straining Ian 5uk!.
Or%ani" Car)on Reoval 9
.dsorption and biodegradation are considered to be the primary natural organic matter removal
mechanisms -ollins!. Citerature cited by -ollins suggests that large hydrophobic<humic organic
molecules are removed by adsorption, and smaller organic molecules are removed by both
adsorption and biodegradation. The smaller hydrophilic material carbohydrates, aldehydes, and
simple organic acids! are considered to be primarily removed by biodegradation. . common
o,idant for the treatment of water in the United States is chlorine+ the hydrophobic<humic organics
considered to be the more trihalomethane reactive! were removed in greater than ;3B of all
comparisons of organic parameters cited in the -ollins study.
.nother interesting aspect of the -ollins paper was the fact that natural organic matter and organic
precursor material were a function of filter media biomass) the greater the biomass, the greater the
organic carbon removals. Three US sand filters were compared+ the #est 8artford filters use a
unique Schmutzedecke cleaning procedure called the filter<harrowing cleaning technique discussed
in more detail under "ilter Scraping!. This procedure allows for the minimization of biomass
removal from mature sand filters resulting in increased removals of biodegradation and
bioadsorption -ollins!.
*emoval of &iardia and -ryptosporidia $ In the past decade the protozoan parasite -ryptosporidium
parvum has been recognized as a significant threat to public water supplies. The resistant stage of
-ryptosporidia is called an oocyst+ this stage is relatively untouched by a chlorination disinfection
process. Slow sand filtration has been looked at in numerous studies to determine the viability of
this treatment process for the removal of -yrptosporidia. . study in 0ngland by Timms found
reductions of oocysts greater than FF.F9B+ the oocysts were found in the filter media above 7.2 cm.
.nother study in >ritish -olumbia by "ogel contradicts the aforementioned study. "ogel found
removal efficiencies of =;B+ this figure is significantly different than the 433B removals "ogel
cites from previous literature. 8owever, a point to note concerning the >ritish -olumbia filters is
that they were operating well out of the range of the recommended design limits for the uniformity
coefficient at (.2 "ogel!. "urthermore, temperature can adversely affect the performance of a slow
sand filter+ the >ritish -olumbia filters were operating at e,tremely low temperatures of less than
4- "ogel!. 6verall, the literature supports data that strongly suggests slow sand filtration is a
viable alternative for -ryptosporidia removals.
Slow sand filters have also been proven highly efficient in removing Giardia lamblia, a frequently
identified pathogenic intestinal protozoa. The same study by "ogel found that despite the
uniformity coefficient parameter and the low temperatures, Giardia removals were complete. This
data was further supported by literature cited by "ogel. "urthermore, fecal and total coliform counts
were below the detection limit, and the removal rates were similar to Giardia removals "ogel!.
2"$!t3ede"/e 2"rapin% Operations
Source for Desi%n n!)er -
Slow sand filters slowly lose their performance as the !chmut$dec"e grows and thereby reduces the
rate of flow through the filter. 0ventually it is necessary to refurbish the filter. Two methods are
commonly used to do this.
In the first, the top few millimetres of fine sand is scraped off to e,pose a new layer of clean sand.
#ater is then decanted back into the filter and re<circulated for a few hours to allow a new
!chmut$dec"e to develop. The filter is then filled to full depth and brought back into service.

The second method, sometimes called wet harrowing, involves lowering the water level to 1ust
above the !chmut$dec"e, stirring the sand and thereby suspending any solids held in that layer and
then running the water to waste. The filter is then filled to full depth and brought back into service.
#et harrowing can allow the filter to be brought back into service more quickly.
Source for Desi%n n!)er 0
Scraping typically involves the removal of the Schmutzedecke and the operation is site specific.
"requency of scraping depends on the available head, the media grain<size distribution, the influent
water quality, and the water temperature Cetterman!. 8igher frequencies of scraping are associated
with increased water temperature, high solids concentrations in the influent, low head, and small
media pore size. . typical operation involves draining the supernatant usually by continuing
filtration with no influent! to 73 cm below the sand surface, skimming off one inch of the
Schmutzedecke and associated sand, and then filling the filter from the bottom of the bed using
filtered water to prevent air entrapment. The bed should be refilled until depth is sufficient to
continue normal operations Cetterman!.
The study by -ollins at #est 8artford Treatment 'lant suggests a unique way to scrape the
Schmutzedecke that minimizes the amount of biomass removed. The supernatant is drained to a
height roughly (3 cm above the bed. . rubber<tired tractor equipped with a comb<tooth harrow is
place on top of the filter to rake the sand+ simultaneously the filter drains are opened, causing a
steady discharge of overlying water. .s the Schmutzedecke is loosened, the colloidal debris is
caught by the moving water and discharged at the filter surface drain. The process is repeated as
necessary by backflushing until the entire ilter surace has been harrowed -ollins!.
Source for Desi%n n!)er 1
6 "ilter cleaning, by scraping off a 72 mm surface layer of sand, is required at intervals of between
(3 and 433 days depending on the turbidity of the water being filtered. Trials in India found that, at
a turbidity of 43 @TU, the filter could be run for up to F3 days, whereas, at a turbidity of (3 @TU,
the filter run time would be reduced to (3 days. "ilter surface fabrics have been developed for slow
sand filters to improve and ease the cleaning process. 8owever, because aliquots of the sand are
removed in the cleaning process, the sand layer will need to be topped up from time to time. The
depth of sand should not be allowed to become less than :33 mm. Topping up can be done using
washed sand, removed during previous filter cleanings, or fresh sand. 'athogen removal is achieved
within the biological slime layer schmutzdecke! which forms on the surface of the sand. This will
take a few days to re<form after each cleaning and there will be a slight decline in the performance
of the filtration system during this time+ usually, this decline is not significant.
.s they require little or no mechanical power, chemicals or replaceable parts, and they
require minimal operator training and only periodic maintenance, they are often an
appropriate technology for poor and isolated areas.
Slow sand filters, due to their simple design, may be created diy do it yourself.
The most attractive aspect of slow sand filtration is its simplicity of operation and control.
.fter a short training period, the operation of the system can be mastered by a local
Unlike other filtration methods, slow sand filters use biological processes to clean the water,
and are non<pressurized systems. Slow sand filters do not require chemicals or electricity to
Slow sand filters are recognized by the #orld 8ealth 6rganization J4K, 6,fam, United
@ations J7K and the United States 0nvironmental 'rotection .gency J(K as being superior
technology for the treatment of surface water sources. .ccording to the #orld 8ealth
6rganization, EUnder suitable circumstances, slow sand filtration may be not only the
cheapest and simplest but also the most efficient method of water treatment.E
The cost of construction is low, and its simplicity of design and operation means that slow
sand filters can be built and used with limited technical supervision. Cittle special pipework,
equipment, or instrumentation is needed, and the labour required for maintenance can be
unskilled as the ma1or labour requirement is in cleaning the beds, which can be done by
hand. Imported materials and equipment is usually negligible and no chemicals are required.
Cikewise, power is not required if a gravity head is available, and there are no moving parts
or requirements for compressed air or high<pressure water. Iariations in raw water quality
and temperature can be accommodated, provided turbidity does not become e,cessive, and
overloading for short periods does no harm.
-ompared to rapid sand filtration, there is a net savings of water as large quantities of
backwash water are not required.
Costs :Slow sand filter costs can be low if appropriate local materials are available. "or
large slow sand treatment plants, the land area required for constructing the system may add
significantly to the cost. 8ence, the technology is best adopted by small communities where
land cost may not be a problem.>ecause slow sand filters usually require no chemicals or
energy inputs, the capital and operational costs can be very low in comparison to rapid sand
filters. /aintenance costs will include minor repairs to the filters, and replacement of the
sand washed out or removed during the scraping of the silt from the filter surface. 6ther
maintenance costs relate to the replacement of the few moving parts present in the filter
system. The costs will be higher in pumped schemes. 0,amples of specific costs for small
islands are not available.
5ue to the low filtration rate, slow sand filters require e,tensive land area for a large
municipal system.
. slow sand filter may take up to five times the equivalent land area
of a rapid sand filtration plant. /any municipal systems in the U.S. initially used slow sand
filters, but as cities have grown they subsequently installed rapid sand filters, due to
increased demand for drinking water.
Slow sand filtration is normally used for the treatment of surface water supplies only, and,
thus, would not be applicable to many of the small low<lying islands where surface waters
do not e,ist. Slow sand filtration is most suited for use with gravity operated, surface water
supply systems, or systems with large, clear water storage facilities, as the filters require a
continuous flow of water.
Unlike other water filtration technologies that produce water on demand, slow sand filters
produce water at a slow, constant flow rate and are usually used in con1unction with a
storage tank for peak usage. This slow rate is necessary for healthy development of the
biological processes in the filter.

Slow sand filters require relatively low turbidity levels to operate efficiently. In summer
conditions and in conditions when the raw water is turbid, blinding of the filters occurs more
quickly and pre<treatment is recommended.
-logging may occur if the source water is e,cessively turbid or if certain filamentous! types
of algae are present in the raw water. 're<treatment with roughing filters or settling tanks
may be necessary if such clogging occurs frequently.
.lso, to,ic chemical contamination of the raw water may affect the biological surface layer
this could be a good indication of water source problemsL!
A Bio2and Water Filter
is a technological adaptation of the centuries old slow sand filtration process. 5avid /anz designed the
system and hold US patents.
The system is intended for use in rural homes where naturally safe or
treated water sources are not available.
>ioSand "ilters remove F2.3 to FF.3 percent of organic contaminants, including bacteria, viruses,
protozoa, worms, and particles.
Safe water produced by the filters is free of discoloration, odor, and
unpleasant taste
, and can be used for drinking, food preparation, personal hygiene, and sanitation.
/ost common home<based models can produce 73 to :3 litres =.= to 4( imp gal+ 2.( to 4: US gal! of
water per hour.
Healt$ )enefits
. 7339 study conducted by the School of 'ublic 8ealth at the University of @orth -arolina at -hapel
8ill indicates that >ioSand #ater "ilters can reduce the incidence of diarrheal illness by up to
=3 percent.
6ver time, the top layers of sand may become clogged with material, causing flow rates to drop. .
simple stirring or skimming of the top layer of sand is usually sufficient to restore optimal flow.
"requency of needed maintenance is dependent on the quality of the source water.
longitudinal studies have not been completed, SamaritanDs 'urse reports that filters have remained in
effective operation for over ten years.
/ost >ioSand "ilters are constructed from concrete or
&ravel and sand are layered inside the filter with a
'I- collection pipe situated at the base of the filter.
-ontaminated water from rain, surface, or ground sources is
poured through the top of the filter and passes through a
plate that diffuses the stream and blocks large contaminants
e.g. stones, large twigs, leaves!.
The layer of sand must remain undisturbed by the flow of
poured water. These conditions are achieved respectively by
the installation of a diffusion plate on the top reducing the
impact of the poured water flow!, filtration occurs in the
lower layers of sand and gravel, which removes
contaminants that cause odour, cloudiness, and taste. Then,
filtered water comes through the 'I- pipes.
2oe (ater so!r"es "ontain disease4"a!sin% or%aniss ($i"$ need to )e reoved or /illed )efore t$e
(ater is safe to drin/* If "aref!ll& !nderta/en and onitored+ disinfe"tion is an effe"tive eans of
reovin% s!"$ or%aniss* C$lorine is t$e ost (idel& !sed disinfe"tant+ and one ($i"$ is often t$e ost
readil& availa)le* T$is te"$ni"al )rief des"ri)es a et$od of "al"!latin% t$e dose of "$lorine re8!ired to
disinfe"t sall "o!nit& (ater s!pplies*
W$& disinfe"t;
#ater treatment processes such as storage, sedimentation, and sand filtration will reduce the content of
disease<causing organisms in water, but will not leave it completely free of such organisms.
5isinfection, when applied and controlled properly, is the most practical and effective means of
removing such organisms.
7et$ods of disinfe"tion
>oiling water may be effective as a method of disinfection, but it is not practicable for large quantities.
Sunlight can also act as a natural method of disinfection, but it is difficult to control and manage. "or
these reasons, chemical disinfectants especially chlorine and chlorine compounds! are used. Iodine
may also be used as a disinfectant, but it is usually more e,pensive than chlorine compounds.
-hlorine compounds will destroy disease<causing organisms quickly M usually after (3 minutes. They
are widely used and are relatively ine,pensive. If carefully applied, chlorine has the advantage that a
measurable residual of chlorine in solution can be maintained in the water supply. This residual
provides further potential for disinfection and is also an important indicator of successful application.
W$en to !se "$lorine
-hlorine may be used)
when suitable compounds are available and their application can be strictly controlled+
when there is enough time between the addition of chlorine to water and the consumption of the
where a community has a continuous supply of water, with storage capacity+ and
by individuals to provide additional protection.
W$en NOT to !se "$lorine
-hlorine should not be used)
when a regular supply of chlorine compunds cannot be guaranteed+
where chlorine may react with other chemicals in the water creating undesirable or dangerous
to attempt to kill cysts or viruses+ or
when careful monitoring cannot be provided.
C$lorine deand and resid!al
#hen chlorine is added to a water source, it purifies the water by damaging the cell structure of
bacterial pollutants, thereby destroying them. The amount of chlorine needed to do this is called the
"$lorine deand of the water. The chlorine demand varies with the amount of impurities in the water.
It is important to realise that the chlorine demand of a water source will vary as the quality of the water
The aim of chlorination is to satisfy the chlorine demand of the water source. 6nce the demand has
been satisfied, any e,cess chlorine above the level needed to satisfy the demand remains as a resid!al
of chlorine chlorine residual! in the supply.
If a supply is to be adequately disinfected, therefore, there should be a chlorine residual in the supply,
so that there is the capacity to cope with any subsequent bacterial contamination. The chlorine residual
should generally be in the range 3.( to 3.2mg of chlorine per litre of treated water. .ny more than this
and the supply may taste bad and be harmful, and people may refuse to use it. .ny less, and there is no
guarantee that the supply is adequately protected. .n e,ample is given in "igure 4 below
Testin% for "$lorine resid!al
The most common test is the dpd diethyl paraphenylene diamine! indicator test, using a comparator.
This test is the quickest and simplest method for testing chlorine residual.
#ith this test, a tablet reagent is added to a sample of water, colouring it red. The strength of colour is
measured against standard colours on a chart to determine the chlorine concentration. The stronger the
colour, the higher the concentration of chlorine in the water
Several kits for analysing the chlorine residual in water, such as the one illustrated in "igure 7, are
available commercially. The kits are small and portable.
C$lorinatin% (ater s!pplies
-hlorine is available in many forms M as chlorine gas and in compounds such as bleaching powder,
high test hypochlorite 8T8!, tablets, granules, and liquid bleach.
0ach product contains a different amount of usable chlorine, so different quantities of each will be
required for the same purpose. In addition, the chlorine content of each product will reduce over time as
the source is e,posed to the atmosphere. .ll products should be carefully stored to minimize
The best practical method of chlorinating a supply of water is to use two storage tanks of suitable size
alternately, one filled from the source, while the other is used for supply.
A "$lorination "$e"/list
< C$lorine needs at least $alf an $o!r "onta"t tie (it$ (ater to disinfe"t it*
The best time to apply it is after any other treatment process, and before storage and use.
< Never appl& "$lorine )efore slo( sand filtration or an& ot$er )iolo%i"al pro"ess+ as t$e "$lorine
(ill /ill off t$e )a"teria ($i"$ assist treatent+ a/in% t$e treatent ineffe"tive*
< Never add an& solid for of "$lorine dire"tl& to a (ater s!ppl&+ as it (ill not i= and dissolve*
Al(a&s a/e !p as a paste first+ i=in% t$e "$lorine "opo!nd (it$ a little (ater*
N 5isinfection is only one defence against disease. 0very effort should be made to protect water sources
from contamination, and to prevent subsequent contamination during collection and storage.
N The correct procedure for applying a disinfectant to water should be strictly adhered to, and water
supplies should be monitored regularly to ensure that they are free from bacteria. 6therwise, people
may be misled to believe that the water is safe to drink when, in fact, it is hazardous to do so.
N The optimum chlorine residual in a small, communal water supply is in the range of 3.( to 3.2mg$l.
N The chlorine dose required to disinfect a supply will increase if the water is very turbid. In such
circumstances, it is best to treat the water to reduce turbidity before chlorination.
Ca!tion All fors of "$lorine are $arf!l to $ealt$ > avoid s/in "onta"t and do not in$ale t$e
f!es* C$lorine s$o!ld )e stored in "ool+ dar/+ dr& and sealed "ontainers+ and o!t of rea"$ of
C$lorinatin% Wells+ 2prin%s+ and Cisterns
-hlorination, when properly applied, is a simple way to ensure and protect the purity of water.
&uidelines given here include tables to give a rough indication of the amounts of chlorine<bearing
chemical needed. Instructions are also given for super<chlorination for disinfecting newly built or
repaired wells, spring encasements, or cisterns. -hlorine<bearing compounds, such as ordinary laundry
bleach made with chlorine are used because pure chlorine is difficult and dangerous to use.
Deterinin% t$e Proper Ao!nt of C$lorine
The amounts of chlorine suggested here will normally make water reasonably safe. . water<treatment
system should be checked by an e,pert. In fact, the water should be tested periodically to make sure that
it remains safe. 6therwise, the system itself could become a source of disease.
Tools and 7aterials
-ontainer to mi, chlorine
-hlorine in some form
Scale to weigh additive
The safest way to treat water for drinking is to boil it see E>oiler for 5rinking #aterE!. 8owever,
under controlled conditions, chlorination is a safe method+ it is often more convenient and practical than
boiling. 'roper treatment of water with chlorine requires some knowledge of the process and its effects.
#hen chlorine is added to water, it attacks and combines with any suspended organic matter as well as
some minerals such as iron. There is always a certain amount of dead organic matter in water, as well as
live bacteria, viruses, and perhaps other types of life. 0nough chlorine must be added to o,idize all of
the organic matter, dead or alive, and to leave some e,cess uncombined or EfreeE chlorine. This residual
free chlorine prevents recontamination. Too much residual chlorine, however, is harmful and e,tremely
Some organisms are more resistant to chlorine than others. Two particularly resistant varieties are
amoebic cysts which cause amoebic dysentery! and the cercariae of schistosomes which cause
bilharziasis or schistosomiasis!. These, among others, require much higher levels of residual free
chlorine and longer contact periods than usual to be safe. 6ften special techniques are used to combat
these and other specific diseases.
It always takes time for chlorine to work. >e sure that water is thoroughly mi,ed with an adequate dose
of the dissolved chemical, and that it stands for at least (3 minutes before consumption.
'olluted water that contains large quantities of organic matter, or cloudy water, is not suitable for
chlorination. It is best, and safest, to choose the clearest water available. . settling tank and simple
filtration can help reduce the amount of suspended matter, especially particles large enough to see.
"iltration that can be depended upon to remove all of the amoebic cysts, schistosomes, and other
parthogens normally requires professionals to set up and operate.
. home<made slow sand filter is an e,cellent way to prepare water for chlorination.
5epending on the water to be treated, varying amounts of chlorine are needed for adequate protection.
The best way to control the process is to measure the amount of free chlorine in the water after the (3
minute holding period. . simple chemical test, which uses a special organic indicator called
orthotolidine, can be used. 6rthotolidine testing kits available on the market come with instructions on
their use.
#hen these kits are not available, the chart in Table 4 can be used as a rough guide to how strong a
chlorine solution is necessary. The strength of the solution is measured in parts by weight of active
chlorine per million parts by weight of water, or Eparts per millionE ppm!.
Water Condition Initial C$lorine Dose in Parts Per 7illion ,pp.
@o hard<to<kill
8ard<to<kill organisms present or suspected
Iery -lear, few minerals 2ppm &et e,pert advice+ in an emergency boil and cool water
first, then use 2 ppm to help prevent recontamination. If
boiling is impossible, use 43 ppm.
. coin in the bottom of 4$=
liter ; ounce! glass of the
water looks hazy,
l6ppm &et e,pert advice+ in an emergency boil and cool first. If
boiling is impossible usel2 ppm.
O 7arts per million #ppm) is the number o pans by %eight o chlorine to a million parts by %eight o
%ater. -t is e,uivalent to milligrams o per liter.
The chart in Table 7 gives the amount of chlorine<compound to add to 4,333 liters or to 4,333 gallons of
water to get the solutions recommended in Table 4.
C$lorine Copo!nd Per"ent )&
Wei%$t A"tive
B!antit& to add to -CCC U2*
%allons of (ater re8!ired
B!antit& to add to -CCC liters
to %et stren%t$ re8!ired
2 ''/ 43 ''/ 42 ''/ 2 ''/ 43 ''/ 42 ''/
8igh test -alcium

93B 4oz 7oz (oz ;gms 42gms 7(gms
-hlorinated Cime 72B 7 4$7oz 2oz 9 4$7oz 73gms =3gms :3gms
Sodium hypochlorite
4=B 2oz 43oz 42oz (;gms 92gms 44(gms
Sodium hypochlorite 43B 9oz 4(oz 73oP =;gms F2gms 4=(gms
>leachM. Solution
of -hlorine in water
usually 2.72B 4(T6 7:oz (Foz F2gms 4F3gms 7;2gms
Usually it is convenient to make up a solution of 233 ppm strength that can then be further diluted to
give the chlorine concentration needed. The 233 ppm solution must be stored in a sealed container in a
cool dark place, and should be used as quickly as possible since it does lose strength. /odern
chlorination plants use bottled chlorine gas, but this can only be used with e,pensive machinery by
trained e,perts.
Super<chlorination means applying a dose of chlorine that is much stronger than the dosage needed to
disinfect water. It is used to disinfect new or repaired wells, spring encasements, and cisterns. Table (
gives recommended doses.
T.>C0 (
Appli"ation Re"oended
@ew or repaired
23 ppm 4. #ash casing, pump e,terior and drip pipe with solution.
7. .dd dosage to water in well.
(. 'ump until water coming from pump has strong chlorine odor for
deep wells, repeat this a few times at 4 hour intervals.!
=. Ceave solution in well
at least 7= hours.
2. "lush all chlorine from well.
23 ppm Same as above.
-isterns 433 ppm 4. "lush with water to remove any sediment.
7. "ill with dosage.
(. Cet stand for 7= hours.
=. Test for residual chlorine. If there is none, repeat dosage.!
2. "lush system with treated water.
O To ind the correct amounts o chlorine compound needed or the re,uired dosage, multiply the
amounts given under 50ppm in Tables 1 or 3 to get .0ppm and by 50 to get 500ppm.
E=aple -D
. water<holding tank contains ;,333 U.S. gallons. The water comes from a rapidly moving mountain
stream and is passed through a sand filter before storage. 8ow much bleach should be added to make
this water drinkableP 8ow long should the water be mi,ed after addingP
In this case 2 ppm are probably sufficient to safeguard the water. To do this with bleach requires 4(
ounces per 4,333 gallons. Therefore the weight of bleach to be added is 4( , ; or 43= ounces.
.lways mi, thoroughly, for at least a half hour. . good rule of thumb is to mi, until you are certain that
the chemical is completely dissolved and distributed and then ten minutes longer. In this case, with an
;,333<gallon tank, try to add the bleach to several different locations in the tank to make the mi,ing
easier. .fter mi,ing, test the water by sampling different locations, if possible. -heck the corners of
tank especially. 0,ample 7)
. new cistern has been built to hold water between rainstorms. 6n its initial filling it is to be super<
chlorinated. 8ow much chlorinated lime should be addedP The cistern is 7 meters in diameter and (
meters high.
"irst calculate the volume of water. "or a cylinder, Iolume is
<<< Q 8
5 is diameter, 8 is height and is (.4=.!

8ere 5 Q 7 meters 8 Q ( meters.
<<<< , 7 meters! , 7 meters! , ( meters!
I Q F.=7 cubic meters Q F,=73 liters 0ach cubic meter contains 4,333 liters.!
"rom Table ( we learn that a cistern should be super<chlorinated with 433 ppm of chlorine. "rom Table
7, we learn that it takes =3 grams of chlorinated lime to bring 4,333 liters of water to 43 ppm -4. To
bring it to 433 ppm, then, will require ten times this amount, or =33 grams.
=33 grams
<<<<<<<<<<<<<<<<<<<< , F.=7 thousand liters Q (,9:; grams.
thousand liters
#8.T IS .-TII.T05 -.*>6@P
A"tivated "ar)on, also called a"tivated "$ar"oal, a"tivated "oal or "ar)o a"tivat!s, is a form of
carbon that has been processed to make it extremely porous and thus to have a very large surface
area available for adsorption or chemical reactions.
The word activated in the name is sometimes replaced with active.
.lmost all materials containing a high fi,ed carbon content can potentially be activated. The most
commonly used raw materials are coal anthracite, bituminous and lignite!, coconut shells, wood both
soft and hard!, peat and petroleum based residues. /any other raw materials have been evaluated such
as walnut shells, peach pits, babassu nutshell and palm kernels but invariably their commercial
limitation lies in raw material supply. This is illustrated by considering that 4,333 tons of untreated shell
type raw material will only yield about 433 tons of good quality activated carbon.
/ost carbonaceous materials do have a certain degree of porosity and an internal surface area in the
range of 43<42 m7$g. 5uring activation, the internal surface becomes more highly developed and
e,tended by controlled o,idation of carbon atoms < usually achieved by the use of steam at high
Car)oni3ed Co"on!t A"tivated Co"on!t
.fter activation, the carbon will have acquired an internal surface area between 933 and 4,733 m7$g,
depending on the plant operating conditions.
The internal surface area must be accessible to the passage of a fluid or vapor if a potential for
adsorption is to e,ist. Thus, it is necessary that an activated carbon has not only a highly developed
internal surface but accessibility to that surface via a network of pores of differing diameters.
.s a generalization, pore diameters are usually categorized as follows)
micropores R=3 .ngstroms
mesopores =3 < 2,333 .ngstroms
macropores S2,333 .ngstroms typically 2333<73333 .!
5uring the manufacturing process, macropores are first formed by the o,idation of weak points edge
groups! on the e,ternal surface area of the raw material. /esopores are then formed and are, essentially,
secondary channels formed in the walls of the macropore structure. "inally, the micropores are formed
by attack of the planes within the structure of the raw material.
.ll activated carbons contain micropores, mesopores, and macropores within their structures but the
relative proportions vary considerably according to the raw material.
. "o"on!t s$ell based carbon will have a predominance of pores in the micropore range and these
account for F2B of the available internal surface area. Such a structure has been found ideal for the
adsorption of small molecular weight species and applications involving low contaminant
In contrast (ood and peat )ased "ar)ons are predominantly meso$macropore structures and are,
therefore, usually suitable for the adsorption of large molecular species. Such properties are used to
advantage in decolorization processes.
Coal )ased carbons, depending on the type of coal used, contain pore structures somewhere between
coconut shell and wood.
In general, it can be said that macropores are of little value in their surface area, e,cept for the
adsorption of unusually large molecules and are, therefore, usually considered as an access point to
/esopores do not generally play a large role in adsorption, e,cept in particular carbons where the
surface area attributable to such pores is appreciable usually =33 m7$g or more!.
Thus, it is the micropore structure of an activated carbon that is the effective means of adsorption.
It is, therefore, important that activated carbon not be classified as a single product but rather a range of
products suitable for a variety of specific applications.
T80 /.@U".-TU*I@& '*6-0SS 6" .-TII.T05 -.*>6@
*aw /aterials
It has already been stated that essentially any carbonaceous material can potentially be activated. In
addition to the more common raw materials discussed earlier, others can include waste tires, phenol
formaldehyde resin, rice husks, pulp mill residues, corn cobs, coffee beans and bones.
/ost of the developed nations have facilities to activate coconut shell, wood and coal. Third world
countries have recently entered the industry and concentrate on readily available local raw materials
such as wood and coconut shell. -oconut shell contains about 92B volatile matter that is removed,
largely at source by partial carbonization, to minimize shipping costs. The cellulosic structure of the
shell determines the end product characteristics, which at (3<=3B yield on the carbonized basis! is a
material of very high internal surface area consisting of pores and capillaries of fine molecular
dimensions. The ash content is normally low and composed mainly of alkalis and silica.
Coal is also a readily available and reasonably cheap raw material. In the case of coal based carbons, pre<
treatment of the raw coal is necessary in order for it to be processed, since raw coal swells during heating to
produce coke<like structures. -ontrol of this is achieved by first grinding the raw coal and mi,ing it with various
additives, such as pitch, before it is introduced to the activation furnace. 8owever, the grinding process destroys
the mechanical strength of coal < therefore, ground coal is reconstituted into briquettes prior to processing.
The activate obtained depends on the type of coal used and its initial processing prior to carbonization
and activation.
It is normal procedure to grind the coal and reconstitute it into a form suitable for processing, by use of
a binder such as pitch, before activation. This is typical for e,truded or pelletized carbon!. .n
alternative method is to grind the coal and utilize its volatile content to fuse the powder together in the
form of a briquette. This method allows for blending of selected materials to control the swelling power
of the coals and prevents coking. If the coal is allowed to TcokeU it leads to the production of an activate
with an unacceptably high proportion of large pores. >lending of coals also allows a greater degree of
control over the structure and properties of the final product.
Wood may be activated by one of two methods, i.e. steam or chemical activation, depending on the
desired product. . common chemical activator is phosphoric acid, which produces a char with a large
surface area suitable for decolorization applications. The carbon is usually supplied as a finely divided
powder which since produced from waste materials such as sawdust, is relatively cheap and can be used
on a Tthrow<awayU basis.
Since activated carbon is manufactured from naturally occurring raw materials, its properties will
obviously be variable. In order to minimize variability it is necessary to be very selective in raw
material source and quality and practice a high level of manufacturing quality control.
7et$ods of 7an!fa"t!re
.ctivated carbon is carbon produced from carbonaceous source materials like nutshells, peat, wood,
coir, lignite, coal and petroleum pitch. It can be produced by one of the following processes)
4. P$&si"al a"tivation) The precursor is developed into activated carbons using gases. This is
generally done by using one or a combination of the following processes)
o Carboni$ation) /aterial with carbon content is pyrolyzed at temperatures in the range
:33?F33 V-, in absence of o,ygen usually in inert atmosphere with gases like argon or
o &ctivation6+/idation) *aw material or carbonized material is e,posed to o,idizing
atmospheres carbon mono,ide, o,ygen, or steam! at temperatures above 723 V-, usually
in the temperature range of :33?4733 V-.
C$ei"al a"tivation) 'rior to carbonization, the raw material is impregnated with certain
chemicals. The chemical is typically an acid, strong base, or a salt phosphoric acid, potassium
hydro,ide, sodium hydro,ide, zinc chloride, respectively!. -ompared with chemical activation
using zinc chloride, phosphoric acid activation brings no heavy metal to the carbon.
7. Then, the raw material is carbonized at lower temperatures =23?F33 V-!. It is believed that the
carbonization $ activation step proceeds simultaneously with the chemical activation. -hemical
activation is preferred over physical activation owing to the lower temperatures and shorter time
needed for activating material.
.ctivated carbon can be produced by either physical or chemical activation, both of which require the
use of elevated temperature.
-hemical activation is achieved by degradation or dehydration of the, usually cellulosic, raw material
structure. Steam activation, however, initially involves the removal of volatiles, followed by o,idation
of the structureHs carbon atoms.
2tea A"tivation
The use of steam for activation can be applied to virtually all raw materials.
. variety of methods have been developed but all of these share the same basic principle of initial
carbonization at 233<:33 degrees - followed by activation with steam at ;33<4,433 degrees -.
Since the overall reaction converting carbon to carbon dio,ide! is e,othermic it is possible to utilize
this energy and have a self<sustaining process.
- W 876 steam! <<<S -6 W 87 <(4 Gcal!
-6 W X 67 <<<S -67 W:9 Gcal!
87 W X 67 <<<S 876 steam! W2; Gcal!
- W 67 <<<S -67 WF= Gcal!
. number of different types of kilns and furnaces can be used for carbonization$activation and include
rotary fired directly or indirectly!, vertical multi<hearth furnaces, fluidized bed reactors and vertical
single throat retorts. 0ach manufacturer has their own preference.
.s an e,ample, production of activated carbon using a vertical retort is described below.
*aw material is introduced through a hopper on top of the retort and falls under gravity through a
central duct towards the activation zone. .s the raw material moves slowly down the retort the
temperature increases to ;33<4333 degrees - and full carbonization takes place.
The activation zone, at the bottom of the retort, covers only a small part of the total area available and it
is here that steam activation takes place. .ir is bled into the furnace to convert the product gases, -6
and 87 into -67 and steam which, because of the e,othermic nature of this reaction, reheats the
firebricks on the downside of the retort, enabling the process to be self<supporting.
0very 42 minutes or so, the steam in1ection point is alternated to utilize the Tin situU heating provided
by the product gas combustion. The degree of activation or quality! of the product is determined by the
residence time in the activation zone.
The resulting product is in the form of 4U to (U pieces and requires further processing before being
suitable for its various end uses. This entails a series of crushing and screening operations to produce
specific mesh ranges.
-ertain products may undergo further processing such as drying, acid washing or chemical
impregnation to satisfy particular requirements.
Steam activation produces carbon with an alkaline p8 F to 44!.
To optimise its p8, our steam activated carbon can be neutralised with acid and thereby maintained at
between = and :.
C$ei"al A"tivation
The raw material used in chemical activation is usually sawdust and the most popular activating agent is
phosphoric acid, although zinc chloride and sulfuric acid are well documented. 6thers used in the past
include calcium hydro,ide, calcium chloride, manganese chloride and sodium hydro,ide, all of which
are dehydrating agents.
The raw material and reagent are mi,ed into a paste, dried and carbonized in a rotary furnace at :33
degrees -. #hen phosphoric acid is the activating agent the carbonized product is further heated at ;33<
4333 degrees - during which stage the carbon is o,idized by the acid. The acid is largely recovered
after the activation stage and converted back to the correct strength for reuse.
The activated product is washed with destiled water and dried.
.ctivity can be controlled by altering the proportion of raw material to activating agent, between the
limits of 4)32 to 4)=. >y increasing the concentration of the activating agent, the activity increases
although control of furnace temperature and residence time can achieve the same ob1ective.
.ctivated carbons are comple, products which are difficult to classify on the basis of their behaviour,
surface characteristics and preparation methods. 8owever, some broad classification is made for general
purpose based on their physical characteristics.
Po(dered a"tivated "ar)on ,PAC.
. micrograph of activated charcoal under bright field illumination on a light microscope. @otice the
fractal<like shape of the particles hinting at their enormous surface area. 0ach particle in this image,
despite being only around 3.4 mm wide, has a surface area of several square metres. This image of
activated charcoal in water is at a scale of :.7(: pi,els$Ym, the entire image covers a region of
appro,imately 4.4 by 3.9 mm.
Traditionally, active carbons are made in particulate form as powders or fine granules less than 4.3 mm
in size with an average diameter between .42 and .72 mm.
Thus they present a large surface to volume
ratio with a small diffusion distance. '.- is made up of crushed or ground carbon particles, F2?433B
of which will pass through a designated mesh sieve or sieve. &ranular activated carbon is defined as the
activated carbon being retained on a 23<mesh sieve 3.7F9 mm! and '.- material as finer material,
while .ST/ classifies particle sizes corresponding to an ;3<mesh sieve 3.499 mm! and smaller as
'.-. '.- is not commonly used in a dedicated vessel, owing to the high head loss that would occur.
'.- is generally added directly to other process units, such as raw water intakes, rapid mi, basins,
clarifiers, and gravity filters.
?ran!lar a"tivated "ar)on ,?AC.
&ranular activated carbon has a relatively larger particle size compared to powdered activated carbon
and consequently, presents a smaller e,ternal surface. 5iffusion of the adsorbate is thus an important
factor. These carbons are therefore preferred for all adsorption of gases and vapors as their rate of
diffusion are faster. &ranulated carbons are used for water treatment , deodorization and separation of
components of flow system. &.- can be either in the granular form or e,truded. &.- is designated by
sizes such as ;Z73, 73Z=3, or ;Z(3 for liquid phase applications and =Z:, =Z; or =Z43 for vapor phase
applications. . 73Z=3 carbon is made of particles that will pass through a U.S. Standard /esh Size @o.
73 sieve 3.;= mm! generally specified as ;2B passing! but be retained on a U.S. Standard /esh Size
@o. =3 sieve 3.=7 mm! generally specified as F2B retained!. .##. 4FF7! >:3= uses the 23<mesh
sieve 3.7F9 mm! as the minimum &.- size. The most popular aqueous phase carbons are the 47Z=3
and ;Z(3 sizes because they have a good balance of size, surface area, and head loss characteristics.
E=tr!ded a"tivated "ar)on ,EAC.
0,truded activated carbon combines powdered activated carbon with a binder, which are fused together
and e,truded into a cylindrical shaped activated carbon block with diameters from 3.; to 4(3 mm.
These are mainly used for gas phase applications because of their low pressure drop, high mechanical
strength and low dust content.
Bead a"tivated "ar)on ,BAC.
>ead activated carbon is made from petroleum pitch and supplied in diameters from appro,imately 3.(2
to 3.;3 mm. Similar to 0.-, it is also noted for its low pressure drop, high mechanical strength and low
dust content, but with a smaller grain size. Its spherical shape makes it preferred for fluidized bed
applications such as water filtration.
Ipre%nated "ar)on
'orous carbons containing several types of inorganic impregnant such as iodine, silver, cations such as
.l, /n, [n, "e, Ci, -a have also been prepared for specific application in air pollution control
especially in museums and galleries. 5ue to antimicrobial$antiseptic properties, silver loaded activated
carbon is used as an adsorbent for purification of domestic water. 5rinking water can be obtained from
natural water by treating the natural water with a mi,ture of activated carbon and .l68!
, a
flocculating agent. Impregnated carbons are also used for the adsorption of 8
S and thiols. .dsorption
rates for 8
S as high as 23B by weight have been reported.
Pol&er "oated "ar)on
This is a process by which a porous carbon can be coated with a biocompatible polymer to give a
smooth and permeable coat without blocking the pores. The resulting carbon is useful for
hemoperfusion. 8emoperfusion is a treatment technique in which large volumes of the patientDs blood
are passed over an adsorbent substance in order to remove to,ic substances from the blood.
.ctivated carbon does not bind well to certain chemicals, including alcohols, glycols, strong acids and
bases, metals and most inorganics, such as lithium, sodium, iron, lead, arsenic, fluorine, and boric acid.
-ontrary to a claim repeated
Jcitation neededK
throughout the web, activated carbon does not adsorb ammonia.
.ctivated carbon can be used as a substrate for the application of various chemicals to improve the
adsorptive capacity for some inorganic and problematic organic! compounds such as hydrogen sulfide
S!, ammonia @8
!, formaldehyde 8-68!, radioisotopes iodine<4(4
I! and mercury 8g!. This
property is known as chemisorption.
-arbon mono,ide is not well absorbed by activated carbon. This should be of particular concern to
those using the material in filters for respirators, fume hoods or other gas control systems as the gas is
undetectable to the human senses, to,ic to metabolism and neuroto,ic.
.ctivated carbon does adsorb iodine very well and in fact the iodine number, mg$g, .ST/ 57;
Standard /ethod test! is used as an indication of total surface area.
Iodine n!)er
/any carbons preferentially adsorb small molecules. Iodine number is the most fundamental parameter
used to characterize activated carbon performance. It is a measure of activity level higher number
indicates higher degree of activation!, often reported in mg$g typical range 233?4733 mg$g!. It is a
measure of the micropore content of the activated carbon 3 to 73 \, or up to 7 nm! by adsorption of
iodine from solution. It is equivalent to surface area of carbon between F33 m]$g and 4433 m]$g. It is the
standard measure for liquid phase applications.
Iodine number is defined as the milligrams of iodine adsorbed by one gram of carbon when the iodine
concentration in the residual filtrate is 3.37 normal. >asically, iodine number is a measure of the iodine
adsorbed in the pores and, as such, is an indication of the pore volume available in the activated carbon
of interest. Typically, water treatment carbons have iodine numbers ranging from :33 to 4433.
"requently, this parameter is used to determine the degree of e,haustion of a carbon in use. 8owever,
this practice should be viewed with caution as chemical interactions with the adsorbate may affect the
iodine uptake giving false results. Thus, the use of iodine number as a measure of the degree of
e,haustion of a carbon bed can only be recommended if it has been shown to be free of chemical
interactions with adsorbates and if an e,perimental correlation between iodine number and the degree of
e,haustion has been determined for the particular application.
Some carbons are evaluated based on the dechlorination half<value length, which measures the chlorine<
removal efficiency of activated carbon. The dechlorination half<value length is the depth of carbon
required to reduce the chlorine level of a flowing stream from 2 ppm to (.2 ppm. . lower half<value
length indicates superior performance.
.sh content
It reduces the overall activity of activated carbon. It reduces the efficiency of reactivation. The metal
o,ides "e
! can leach out of activated carbon resulting in discoloration. .cid$water soluble ash
content is more significant than total ash content. Soluble ash content can be very important for
aquarists, as ferric o,ide can promote algal growths. . carbon with a low soluble ash content should be
used for marine, freshwater fish and reef tanks to avoid heavy metal poisoning and e,cess plant$algal
7edi"al appli"ations
.ctivated carbon is used to treat poisonings and overdoses following oral ingestion.
It is thought to bind to poison and prevent its absorption by the gastrointestinal tract. In cases of
suspected poisoning, medical personnel administer activated charcoal on the scene or at a hospitalDs
emergency department. 5osing is usually empirical at 4 gram$kg of body mass for adolescents or
adults, give 23?433 g!, usually given only once, but depending on the drug taken, it may be given more
than once. In rare situations activated charcoal is used in Intensive -are to filter out harmful drugs from
the blood stream of poisoned patients. .ctivated charcoal has become the treatment of choice for many
poisonings, and other decontamination methods such as ipecac<induced emesis or stomach pumping are
now used rarely.
.ctivated charcoal for medical use
#hile activated carbon is useful in acute poisoning, it has been shown to not be effective in long term
accumulation of to,ins, such as with the use of to,ic herbicides.
/echanisms of action)
>inding of the to,in to prevent stomach and intestinal absorption. >inding is reversible so a
cathartic such as sorbitol may be added as well.
It interrupts the enterohepatic and enteroenteric circulation of some drugs$to,ins and their
Incorrect application e.g. into the lungs! results in pulmonary aspiration which can sometimes be fatal
if immediate medical treatment is not initiated.
The use of activated charcoal is contraindicated when
the ingested substance is an acid, an alkali, or a petroleum product.
Ingestion of activated charcoal prior to consumption of alcoholic beverages appeared to reduce
absorption of ethanol into the blood. 2 to 42 milligrams of charcoal per kilogram of body weight taken
at the same time as 493 ml of pure ethanol which equals to about 43 servings of an alcoholic beverage!,
over the course of one hour, seemed to reduce potential blood alcohol content.
^et other studies
showed that this is not the case, and that ethanol blood concentrations were increased because of
activated charcoal use.
-harcoal biscuits were sold in 0ngland starting in the early 4Fth century, originally as an antidote to
flatulence and stomach trouble.
Tablets or capsules of activated charcoal are used in many countries as an over<the<counter drug to treat
diarrhea, indigestion, and flatulence.
'revious versions of this article have claimed that evidence
e,ists that it is effective in treating irritable bowel syndrome I>S!,
, but the reference study given did
not use activated carbon or activated charcoal!, rather tablets of non<activated charcoal. It has also been
used to prevent diarrhea in cancer patients who have received irinotecan.
It can interfere with the
absorption of some medications, and lead to unreliable readings in medical tests such as the guaiac card
.ctivated charcoal is also used for bowel preparation by reducing intestinal gas content before
abdominal radiography to visualize bile and pancreatic and renal stones. . type of charcoal biscuit has
also been marketed as a pet care product.
Distilled al"o$oli" )evera%e p!rifi"ation
See also) Cincoln -ounty 'rocess
.ctivated carbon filters can be used to filter vodka and whiskey of organic impurities which can affect
color, taste, and odor. 'assing an organically impure vodka through an activated carbon filter at the
proper flow rate will result in vodka with an identical alcohol content and significantly increased
organic purity, as 1udged by odor and taste.
Jcitation neededK
Re%eneration of t$e spent "ar)on
The regeneration of activated carbons involves restoring the adsorptive capacity of saturated activated
carbon by desorbing adsorbed contaminants on the activated carbon surface.
T$eral re%eneration
The most common regeneration technique employed in industrial processes is thermal regeneration.

The thermal regeneration process generally follows three steps
.dsorbent drying at appro,imately 432 V-
8igh temperature desorption and decomposition 233?F33V-! under an inert atmosphere
*esidual organic gasification by an o,idising gas steam or carbon dio,ide! at elevated
temperatures ;33V-!
The heat treatment stage utilises the e,othermic nature of adsorption and results in desorption, partial
cracking and polymerization of the adsorbed organics. The final step aims to remove charred organic
residue formed in the porous structure in the previous stage and re<e,pose the porous carbon structure
regenerating its original surface characteristics. .fter treatment the adsorption column can be reused.
'er adsorption<thermal regeneration cycle between 2?42 wtB of the carbon bed is burnt off resulting in
a loss of adsorptive capacity.
Thermal regeneration is a high energy process due to the high required
temperatures making it both an energetically and commercially e,pensive process.
'lants that rely on
thermal regeneration of activated carbon have to be of a certain size before it is economically viable to
have regeneration facilities onsite. .s a result it is common for smaller waste treatment sites to ship
their activated carbon cores to a specialised facility for regeneration, increasing the processD already
significant carbon footprint.
.ctivated carbon used in consumer devices such as oil deep fryers or air and water filters can similarly
be reactivated using commonly available heating appliances such as a baking oven, toaster oven, or
simply a propane torch.
Jcitation neededK
The carbon is removed from any paper or plastic containers that could
melt or ignite, and heated to vaporize and$or burn off contaminants.
The EspentE carbon, as it is called, is removed and sent for re<activation treatment. This is done primarily with
granular activated carbon because '.- particles are too small to be effectively re<activated. This process allows
for recovery of appro,imately 93B of the original carbon. This number also allows for any physically lost in the
shipment process. The re<activated carbon is then mi,ed with a portion of new carbon for higher effectiveness
and is then returned to its place in the plant process -lark, 4F;F!.
Reactivation Process Specifics
# Stage
(degrees C)
1 Drying < 100 GA dewatered to !0" of original weight
# Desorption 100 $ %&' volatile materials driven off
( )yrolysis 100 $ %&' heavy organics burnt leaving residue
**%&' and **10(+
vapors and residues from previous stages driven out of
Ot$er re%eneration te"$ni8!es
-urrent concerns with the high energy$cost nature of thermal regeneration of activated carbon has
encouraged research into alternative regeneration methods to reduce the environmental impact of such
processes. Though several of the regeneration techniques cited have remained areas of purely academic
research, some alternatives to thermal regeneration systems have been employed in industry. -urrent
alternative regeneration methods are)
-hemical and solvent regeneration
/icrobial regeneration
0lectrochemical regeneration
Ultrasonic regeneration
#et air o,idation
2oe ind!strial re"ipes for a"tivated "ar)on a/in%
Re"ipe n!*- process of activation is carried out in two stages. "irstly the coconut shell is converted
into shell charcoal by carbonization process which is usually carried out in mud<pits, brick kilns and
metallic portable kilns. The coconut shell charcoal is activated by reaction with steam at a temperature
of F33o- <4433o- under controlled atmosphere in a rotary kiln. The reaction between steam and
charcoal takes place at the internal surface area, creating more sites for adsorption. The temperature
factor, in the process of activation is very important. >elow F33o- the reaction becomes too slow and is
very uneconomical. .bove 4433o- the reaction becomes diffusion controlled and therefore takes place
on the outer surface of the charcoal resulting in loss of charcoal.
Re"ipe n!*0 The precursors stones or shells! were first ground and sieved to a particle size of
73?=3 mesh and then dried at 433V- for 2 h. The precursors were pyrolysed at =23?=;3 V-, at
atmospheric and reduced pressure 933 mm 8g!.The resulted chars were pulverized to a powder form.
The preparation of the activated biochars was based on Cinares method J2K which involved soaking
each gram of the carbonaceous material with a solution of 7g @a68 W43 ml 5#! for 7= h, then, the
mi,ture was filtered and washed with a solution of 2/ 8-l to neutralize the base and finally washed
several times with 5# to remove any traces of -l? ion, ne,t , the product was sub1ected to air drying
for 7= h, and thermally activation was performed at 923?;33 V- for 4h by means of a tubular oven to
obtain the .-s. The adsorption capacities of the prepared samples have been determined by standard
methods using the iodine number, /> and '@' adsorption values, and humidity content. The iodine
number was determined by the .##. >:3=9=! using the sodium thiosulfate volumetric method J:K.
The adsorption of /> and '@' was performed by agitating a known quantity of the adsorbent 3.4 g!
with 433ml of 23 mg$l />! stock solution and 23 ml of 733 mg$l! stock solution of '@' until
equilibrium was obtained. . calibration curves were prepared in the concentration range of 4?:3 mg $l
and 2?733 mg$l for /> and '@' respectively in order to determine the dye concentration in solution
after adsorption.
Re"ipe n!*1 This work focused on the preparation of activated carbon from eucalyptus and wattle
wood by physical activation with -67. The preparation process consisted of carbonization of the wood
samples under the flow of @7 at =33V- and :3 min followed by activating the derived chars with -67.
The activation temperature was varied from :33 to F33V- and activation time from :3 to (33 min,
giving char burn<off in the range of 73$7<;(B. The effect of -67 concentration during activation was
also studied. The porous properties of the resultant activated carbons were characterized based on the
analysis of @7 adsorption isotherms at _4F:V-. 0,perimental results showed that surface area,
micropore volume and total pore volume of the activated carbon increased with the increase in
activation time and temperature with temperature e,erting the larger effect. The activated carbons
produced from eucalyptus and wattle wood had the >0T surface area ranging from =:3 to 4,=F3 m7$g
and =(3 to 4,3(3 m7$g, respectively. The optimum activation conditions that gave the ma,imum in
surface area and total pore volume occurred at F33V- and :3 min for eucalyptus and ;33V- and (33
min for wattle wood. Under the conditions tested, the obtained activated carbons were dominated with
micropore structure ;3B of total pore volume!
Re"ipe n!* G
a. Separate and clean coconut shell from other materials, such as coconut fiber or soil.
b. Sun dry.
c. >urn dried coconut shell at burning sink or drum at (33<233 o- for (<2 hours.
d. Soak charcoal in chemical solution -a-l7 or [n-l7 72B! -alcium -hloride or [inc -holride 72B!
for 47<4; hours to become activated charcoal.
e. #ash charcoal with distilled$clean water.
f. Spread on tray at room temperature to be drained.
g. 5ry in oven at temperature 443 o- for ( hours.
h. -rush or refine activated charcoal with crusher wood$iron into size of 433 mesh.
i. 'ack activated charcoal in plastic.
Re"ipe n!*H zink chloride activation $ dried in oven at 443c for 7=h, crushing mechanically provided
smaller particles with increased surface area and also enabled more efficient chemical activation of raw
material. 'article size was :33um Omicrometar. This size is most suitable for chemical activation using
[n-l7 and particles of coal .To ensure complete activation [n-l7 and mass $ they were mi,ed at ;2-
for 9 hours. .fter chemical activation samples was dried at 443c for 7=$(:hours. Time for drying varied
depending on amount of zinc chloride used. 8igher amount higher period of time. It was crushed again.
.s #alhof propose it was e,pose to light and humidityOC8 for 77h to enhance the development of
pore structure during pyrolysis.'yrolisis of activated carbon, and C8 treated was out in an inert
environment nitrogen flow 93ml$min! at ;33- for 7 hours. To remove e,cess zink choride and residual
inorganic matter , mass was washed in 233 ml of 4.7/ of 8-C and 233ml destiled water. Upon drying
it was storage to conducting physical activation process 7h at ;33c in mi,ture of carbon mono,ide and
carbon dio,ide 92B -6 and 72B -67!
Re"ipe n!*I >ecause the wood charcoal has the form of carbon, so it doesnHt need carbonization but
directly activated by phosphoric acid and then heated in a furnace at temperature of 233 V -, :33 V -,
933 V - and ;33 V - with 2B, 43B, 42B, 73 B solution concentration, then analyzed the adsorption
capacity of iodine solution. "rom the results of this e,periment gotten the best activated carbon is the
activated carbon which heating at temperature of 933 V - and a concentration of 43B activation with
adsorption power of ;;; mg $ g
Re"ipe n!*J 5epartment of -hemical 0ngineering, University of Sydney, >ldg. `34, Sydney @S#
733:, .ustralia. mvali,
.ctivated carbons were prepared from bagasse through a low temperature 4:3 degrees -! chemical
carbonisation treatment and gasification with carbon dio,ide at F33 degrees -. The merit of low
temperature chemical carbonisation in preparing chars for activation was assessed by comparing the
physical and chemical properties of activated carbons developed by this technique to conventional
methods involving the use of thermal and vacuum pyrolysis of bagasse. In addition, the adsorption
properties acid blue dye! of these bagasse activated carbons were also compared with a commercial
activated carbon. The results suggest that despite the high ash content of the precursor, high surface
areas :4=<4=(( m7 g<4!! and microporous median pore size from 3.=2 to 4.7 nm! activated carbons
can be generated through chemical carbonisation and gasification. The micropore area of the activated
carbon developed from chars prepared by the low temperature chemical carbonisation provides
favourable adsorption sites to acid blue dye (F4 mg g<4! of carbon!. The alkalinity of the carbon
surface and total surface area were shown to have complementary effects in promoting the adsorption of
acid blue dye. .dsorption of the anionic coloured component of the acid dye was shown to be promoted
in carbon e,hibiting alkaline or positively charged surfaces. This study demonstrates that activated
carbons with high acid dye adsorption capacities can be prepared from high ash bagasse based on low
temperature chemical carbonisation and gasification.
,urface -ater .reatment by /oughing 0ilters $ A Design1
onstruction and 2peration 3anual 4,A5D6 $ ,7A.1 1''%1
1+0 p.8
)art #9 Design1 construction and operation of roughing
+. lassification of roughing filters
'. General aspects of roughing filter design
'.1 3ain features
'.# :asic filtration theory
'.( Design variables and guidelines
'.& 0low and headloss control
'.! 0ilter drainage system
'.% General design aspects
10. Detailed filter design
10.1 ;ntake 0ilters
10.# Dynamic filters
10.( <ertical$flow roughing filters
10.& =ori>ontal$flow roughing filters
11. /oughing filter efficiency
11.1 )ractical experience
11.# )ilot plant tests
1#. ,election criteria for roughing filters
1#.1 /aw water ?uality as selection
1#.# @ayout and operational aspects as
selection criteria
1(. onstruction of roughing filters
1(.1 0ilter box
1(.# 0ilter material
1(.( ;nlet and outlet structures
1(.& Drainage system
1(.! Gravel and sand washing facilities
1&. 2peration and maintenance of roughing
1&.1 aretaker training
1&.# .reatment plant commissioning
1&.( 0low control
1&.& -ater ?uality control
1&.! 0ilter cleaning
1&.% 0ilter maintenance
1!. 6conomic aspects
1!.1 onstruction costs
1!.# 2perating and maintenance costs
1!.( 2verall costs of water supply
1%. Design examples
1%.1. .reatment of an upland river
1%.# .reatment of a lowland stream
1%.( .reatment of reservoir water
1%.& /ehabilitation of a slow sand filter
1%.! ,tandard designs for compact water
treatment plants
1A. 0inal remarks
Surface Water Treatment by Roughing Filters - A Design, Construction an !peration "anual #SA$D%C
- S&AT, '((), '*+ p,-
Part ./ Design, construction an operation of roughing filters
*, Classification of roughing filters
.s shown in Table 4, filters "an )e "lassified a""ordin% to filter aterial si3e and filtration rate into
the following categories) rock filters, roughing filters, rapid sand filters and slow sand filters. *oughing
filters, using mainly gravel as filter medium, are operated without chemicals and do not require
sophisticated mechanical equipment for operation and maintenance. @evertheless, their design and
application vary considerably. T$e different ro!%$in% filter t&pes are "lassified a""ordin% toD
< lo"ation within the water supply
< s"$ee ain appli"ation p!rpose
< flo( dire"tion
< filter desi%n
< filter "leanin% te"$ni8!e
-onstruction of infiltration %alleries in the river bed or ne,t to the river embankment is an old
technique used to draw surface water and pretreat it at the same time. Infiltration galleries basically
consist of an e,cavated trench filled with gravel and sand layers surrounding a perforated pipe.
8owever, construction work in water bearing aquifers might prove difficult unless e,tensive alluvial
deposits and high seasonal flow variations prevail as may be the case in South<0ast .sia. There, the
infiltration gallery could be the inlet technology of choice as it is almost maintenance free and may
easily be installed during the dry season, provided the aquifer characteristics prevent clogging as well as
breakthrough of fine material. /aintenance and cleaning of infiltration galleries are hardly possible
unless installed in the river banks or in the bed of irrigation canals where the flow can be regulated or
even interrupted. Such flow regulations allow a controlled operation, protect the installed gravel layers
from being washed away, and enable maintenance work, e.g. cleaning or replacement of the filter layer.
Ta)le - Filter Classifi"ation
Filter T&pe 2i3e of Filter 7aterial di% KL Filtration Rate EF K:$L
rock filter S 23 mm 4 < 2 m$h
roughing filter 73 < = mm 3.( < 4,2 m$h
rapid sand filter = < l mm 2 < 42 m$h
slow sand filter 3.(2 < 3.42 mm 3.4 < 3.7 m$h
8owever, due to the limited application and briefly mentioned operational inconveniences, infiltration
galleries will not be further presented in this manual. Inta/e and d&nai" filters are usually the first
components of a treatment scheme. Similar to infiltration galleries, their structure often forms part of
the water intake installation. Intake filters are used as first treatment step, mainly for separation of
solids. The reduced solids concentration in the pretreated water allows a more economical layout and
operation of the subsequent filter units. 5ynamic filters are applied to safeguard the treatment plant
from sudden solids concentration peaks. 8ence, they are usually used not so much for water quality
improvement, but to protect the treatment plant from heavy silt loads and cumbersome filter cleaning
Ro!%$in% filters are generally located at the treatment plant and used as last pretreatment process prior
to slow sand filtration. These filters can be operated either as upflow, downflow or horizontal<flow
filters. The different gravel fractions of roughing filters are installed either in separate compartments
and hence operated in series, or the differently sized gravel is placed in succeeding layers in the same
Filter "leanin%+ which is carried out manually or hydraulically, is dependent on the pattern of the
retained solids in the filter. Intake and dynamic filters separate the solids usually within the inlet zone of
the filter and thus act as s!rfa"e filters* The relatively fine gravel of these filters is cleaned manually by
scouring the top of the filter with a shovel or rake, and flushing the resuspended solids from the filter
bed. *oughing filters containing differently sized filter material act as deep )ed filters and allow deep
penetration of the solids into the filter medium. *emoval of the accumulated solids is carried out by
periodic filter flushing. *oughing filters might gradually get silted up if the retained solids are not
completely removed by hydraulic filter cleaning. Such undesirable filter clogging calls for tedious
manual cleaning and should thus be avoided whenever possible by regular and efficient filter drainages.
-onstruction of shallow beds in upflow roughing filters minimises construction and maintenance work
and, finally, also manual cleaning. 8owever, such shallow upflow roughing filters should be used only
with raw water of moderate turbidity.
. general layout of different prefilters is given in "ig. 7=, the main differences in use and
configurations of the prefilters are summarised in Table 7, and the detailed filter design and operation
e,plained in -hapters 43 and 4= respectively.
"ig. 7= -lassification of 'refilters
Table 7 Use and Cayout of 'refilters
Children as Promoters of an Infiltration Gallery Project
Children as Promoters of an Infiltration Gallery Project
Different East African countries suffer from civil war. This situation causes enormous misery and
drives thousands of people away from their native villaes and towns. They !ecome helpless refuees
fryin to escape cruelty and starvation. The homeless often settle ne"t to larer settlements where they
try to survive# as in the case of an old tracin and !order town. In the last few years# the population of
this town has rapidly rown to $%#&&& people due to refuees and mirants. 'win to the totally
inade(uate water supply and sanitation facilities# the town council faces enormous pro!lems which
could suddenly lead to cholera epidemics.
The town uses the rather lare river as its water supply. The raw water used to !e pumped to a hill
some $%&& m away where a compact treatment plant# donated !y a European aid aency# was never
put to use due to its hih deree of sophistication. The paint is peelin off the containers and some of
the movin parts are radually disappearin. This is why the population collects water from the river
heavily polluted !y human e"creta.
)esfin# a dynamic local sanitary enineer# aware of the inadmissi!le and danerous situation#
approached e"ternal support aencies and as*ed them to finance a water supply project in his town.
+e invited des* officers from the capital to the town and went with them to the !ride across the river.
,rom there# the deleation saw children diin a hole into the river !ed. Clear water started to seep
into the shaft when the hole reached a depth e(ual to the lenth of a child-s arm. This ave )esfin the
idea of installin an infiltration allery in the ./m deep ravel and sandy river !ed and of
constructin a pumpin platform ne"t to the river !an*. After witnessin this simple display# the
deleation approved )esfin-s unconventional !ut appropriate project.
A year later the project was inauurated !y the officials from the reion and invited uests from the
capital. The %&& 0T1 tur!idity of the raw river wafer was reduced to $% 0T1 !y the infiltration
allery# a value similar to the one achieved !y the children and their hole.
(, 0eneral aspects of roughing filter esign
(,' "ain features
5ifferent installations are required for controlled and adequate filter operation and maintenance.
8owever, the main part of the filter is the section containing the filter material. A filter "oprises t$e
follo(in% si= eleents as schematically illustrated in "ig. 72)
4 inlet flo( "ontrol
4 ra( (ater distri)!tion
4 a"t!al filter
4 treated (ater "olle"tion
4 o!tlet flo( "ontrol
4 draina%e s&ste
T$e inflo( to a filter has to be reduced to a given flow rate and maintained thereafter at this rate as
constant flow conditions are essential for efficient filter operation. In order to simplify flow control
during operation, the ad1usted constant flow rate can remain unchanged even during filter cleaning.
8owever, intake filters require a controlled increase of the flow rate in order to provide sufficient
washwater to flush the resuspended solids out of the filter surface.
T$e ra( (ater distri)!tion on a filter should be homogeneous to achieve uniform flow conditions in
the filter bed. Therefore, the flow emerging from a pipe or a channel ought to be evenly distributed over
the entire filter surface. Submerged filter beds, inlet weirs covering the full filter width, or perforated
walls supplying the entire filter cross section are used for this purpose. To avoid scouring effects of the
filter material, the hydraulic energy of fast flowing water has to be reduced by baffles positioned in the
inlet zone. "or the same purpose, concrete slabs or large flat stones should be placed on top of the filter
material ne,t to overfalls.
T$e a"t!al filter consists of a watertight structure containing filter material. The shape of the filter bo,
is normally rectangular and the walls vertical. 8owever, depending on the local construction techniques,
circular tanks and inclined walls may also be built. *ound river bed gravel or broken stones with
sharper edges are generally used as filter material, although any type of inert material resistant to
mechanical forces, insoluble, and not impairing the water quality with respect to odour or colour, can be
used as filter material.
Colle"tion of t$e treated (ater also has to be uniform over the entire filter bed. Uneven water
abstraction would reduce the overall filter efficiency and create undesirable hydraulic short circuits.
'rovision of a free water table on top of the filter bed is the best option to achieve even collection of the
treated water for upflow filters, or the construction of a false filter bottom see "ig. =9! for downflow
filters. . second but less favourable option is the installation of perforated pipes in downflow filters.
"or horizontal<flow filters, construction of a perforated wall in the outlet chamber is necessary for even
abstraction of the treated water.
"ig. 72 /ain "eatures of a "ilter
T$e o!tlet flo( "ontrol prevents the filter bed from drying out. 8ydraulic cleaning of a dried up
roughing filter filled with accumulated solids is a very difficult if not impossible task. Therefore, all
roughing filters must be operated under saturated conditions. . weir or a raised and aerated effluent
pipe maintains the water above the filter bed level. "urthermore, a I<notch weir might be installed to
allow flow rate measurements at the filter outlet.
T$e draina%e s&ste of roughing filters serves two purposes) it is used for hydraulic filter cleaning and
allows complete during maintenance or repair work. 8ydraulic filter bed cleaning calls for high
discharge rates and, therefore, requires rather large pipes and fittings. "or complete water removal,
additional but smaller drains in inlet and outlet compartments can be installed.
"ig. 7: Solid Separation /echanisms in *oughing "ilters
(,. 1asic filtration theory
The following e,planations aim at providing some information about the filtration mechanisms and at
elucidating the process in more details. Reoval of s!spended solids )& ro!%$in% filters is a rat$er
"ople= pro"ess that includes sedimentation, adsorption and biological as well as biochemical
activities. >asically, as illustrated in "ig. 7:, solid particles have to be transported to a surface and
remain atta"$ed to that surface before they are possibly transfored by biological and biochemical
processes. The latter are also important for the removal of dissolved impurities.
8et us no% ollo% the path o a small 9 :m #0. 009 mm) clay particle through a roughing ilter. 7lease
note that the ollo%ing described ;ourney o our clay particle through a roughing ilter is not a science
iction story but a popular description o particle removal mechanisms ta"ing place in roughing ilters.
&nne/ 9 provides additional analytical details o the processes described more scientiically in <3=>.
The small clay particle is e/posed to dierent transportation, attachment and transormation
Transportation mechanisms
2"reenin% removes particles larger than the pores of the filter bed. The smallest pore sizes are roughly
one si,th of the gravel size.
!ince our clay particle is travelling unhindered through the large pores o the coarse, medium and even
ine ilter gravel, as sho%n in Fig. 1?, it %ill never be retained by screening mechanisms.
2edientation separates settleable solids by gravity. The settling velocity is influenced by mass
density, size and shape of the particle, as well as by viscosity and hydraulic conditions of the water.
8et us no% assume that our clay particle has reached the iner gravel raction o our roughing ilter
operated at 0.. m6h iltration rate. Even at this lo% iltration rate, the time o lo% through a pore o 9
mm length and 5.1. mm height is only 50 seconds %hereas the settling velocity o the clay particle
amounts to 0.05 mm6s. Hence, our clay particle %ould need 51. seconds to overcome only hal o the
pore height and %ill, thereore, hardly touch the surace o a gravel grain but continue to drit deeper
into the ilter bed as sho%n in Fig. 1=.
Inter"eption is described as the process which enhances particle removal through gradual reduction of
the pore size caused by accumulated material.
+ur loating clay particle %ants to settle on the ilter medium and desperately calls his riends already
resting on the gravel surace or help. He "no%s that hundreds o million o colleagues have already
entered the roughing ilter beore he started his hopeless ;ourney, and that the ilter load #%eight o
accumulated solids per unit ilter volume) has reached a value o . g6l. Ho%ever, the initial ilter bed
porosity o 3.@ can only be reduced by 0.1.@ i his resting clay particle colleagues are pac"ed li"e
sardines. Fortunately, they are building up on the gravel grains and orm structured pyramids, thereby
increasing the occupied ilter volume by a actor 50. &s illustrated in Fig. 1A, they are thus able to
reduce porosity by 1..@. Ho%ever, our desperate clay particle is missing his colleagues since the
settling distance is still ar too big.
"ig. 79 Screening
"ig. 7; Sedimentation "ig. 7F Interception
"ig. (3 8ydrodynamic "orces
H&drod&nai" for"es are responsible for the water in the filter to flow continuously through the pore
system and not to turn stagnant. The water has to surround each single gravel grain on its way through
the filter. The flow lines are, therefore, not straight but curved around the gravel grains. The water has
to change even its velocity, since restrictions require flow accelerations, and large pore volumes even
force the water to take a short rest.
.s illustrated in "ig. (3, our old clay particle is also e,posed to this flow pattern and hydraulic shear
forces which drive him on a twisted trail. 8e gets thrown off track by these hydrodynamic forces that
lead him into a filter compartment with stagnant water where he has time to settle and 1oin his waiting
8ence, our clay particle had to be grateful to the flow pattern and hydrodynamic forces which
transported him closer to the filter grains or to a quiescent zone where he could settle on the filter
material. 6n his way through the filter he noticed that some very tiny particles, known as colloids, had
slightly changed their direction when compared to the bulk of the solids, and had diffused due to
molecular forces >rownian movement! in some other directions. 8owever, these forces did not affect
him in the least.
Atta"$ent e"$aniss
6ur clay particle, glad to have escaped the flow is still e,posed to the water current which tries to drag
him away. In this delicate situation, the clay particle can count on the help of his colleagues and on the
support of the grain surface.
7ass attra"tion and ele"trostati" for"e < a combination of these two forces is frequently called
adsorption, enable the particles to keep in contact with other solids and the filter material. /ass particle
attraction van der #aals force! and the attraction between opposite electrically charged particles
double layer forces! very much decrease with increasing distance between the particles. In roughing
filters, these forces are important only to hold the settled particles together on the grain surface.
Biolo%i"al a"tivit& will develop in the filter when particles of organic origin are deposited on the filter
material. >acteria and other microorganisms will form a sticky and slimy layer around the gravel or
may build a large chain of organic material floating in the pores of the filter material.
This biological microcosm is alive. . forest composed of microbes is inhabited by monsters in the form
of larvae and smaller microorganisms such as bacteria. The forest is sub1ected to constant changes, the
micro<inhabitants are eaten up by their macro<residents, thus making a prediction of the behaviour
pattern almost impossible. 'articles readily adhere to this organic material and are retained in the filter.
Ele"trostati" and ass attra"tion as (ell as )iolo%i"al a"tivit& allo( parti"les to reain on t$e
deposited aterial*
Transforation e"$aniss
.s time goes by, new particles settle on top of our small clay particle and slowly turn it into a firm
structure of accumulated material. 8owever, he is no longer alone with his clay particle colleagues
since other material of organic origin and biological matter have started to invade the pyramid<like
structure. 8e also notices the change in water quality within the structure as he is no longer e,posed to
fresh water flowing on the surface of the accumulated material.
>iochemical o,idation starts to convert organic matter into smaller aggregates and finally into water,
carbon dio,ide and inorganic salts. .lso part of the dissolved matter is sub1ected to these chemical and
biochemical reactions. Turbidity and colour also undergo changes, while iron and manganese traces are
precipitated and removed.
/icrobiological activity also play an important part in roughing filters. The back of our old clay particle
started to itch and he realised that three tiny microorganisms had attached themselves to his surface. 8e
then remembered that some faecal coliforms, which were tired of swimming around, had asked him for
a lift before entering the roughing filter 8e $lad agreed to give them a ride and they therefore started
their 1ourney through the filter like boat<people on the back of the clay particle. They also remained
together as the particle settled on the filter material. >ut as time went by, these faecal coliforms started
to starve and were attacked by other microorganisms. It was their last twitch 1ust before they passed
away which disturbed our clay particle colleague.
8ence, )iolo%i"all& a"tive ro!%$in% filters are not onl& effi"ient in reovin% solid atter )!t also
in iprovin% si%nifi"antl& t$e "$ei"al and i"ro)iolo%i"al (ater 8!alit&*
T$e M-:1 4 0:1 Filter T$eor&M
6ur old clay particle still did not feel comfortable even after the death of the three faecal coliforms as
he was embedded in a large clay deposit. -onversation with his colleagues became boring and he
started to ponder on how to change his present unsatisfactory situation. 8e remembered seeing a lot of
clay fellows sitting cheerfully on the larger grains while he had to travel through the filter
uncomfortable and squeezed in. Thanks to his quick mind he developed the E4$( < 7$( filter theoryE.
8e knew by e,perience that a particle can bypass a gravel grain either on the left or on the right or settle
on its surface. 8ence, the chances to fall on the grain is 4$(. 8owever, the game continues as there is a
second, third and many other gravel grains to settle on. .t this point our clay particle started making
some calculations. 8e assumed that if about (33 clay particles enter the filter 433 clay particles would
settle on the first layer of grains and 733 clay particles would have to continue their 1ourney to reach the
second layer of grains. 8ere again, 4$( or :9 particles would attach themselves to the second line, and
7$( or 4(( particles remain in the water flow.. The ne,t line of particles would be split in == to ;F
particles. 8e continued his mental arithmetic and was glad that someone was writing it down in .nne,
'roud of himself, the clay particle evaluated his calculation and came to the conclusion that F3B of the
(33 particles which enter a filter are removed already after the 2th or :th gravel layer. The remaining
43B have to travel through another five to si, gravel layers in order to achieve a FFB particle
separation. 8ence, compared to an efficiency of only 4.2B per layer in the second filter section, the first
part of the filter is apparently more efficient in particle removal, since every layer of this filter section
retains about 4:B of the particles. The following filter sections are obviously less efficient in particle
removal. 8owever, our clay particle found it hard to believe that a gravel grain in the inner part of the
filter with the e,act same size and shape as a gravel grain located in the filter inlet should be less
efficient in particle removal. 8e then remembered that filter efficiency is dependent on particle
concentration) the higher the concentration of impurities, the greater the apparent efficiency.
@evertheless, the clay particle was not yet satisfied with his filtration theory as it would mean that the
number of particles found on the gravel would continuously decline on his 1ourney through the filter.
@evertheless, he remembered that the has encountered a sudden increase of settled particles at some
specific points in the filter. 8e quickly concluded that these places were identical with the changes in
the gravel fractions. "urthermore, he recalled that his finer clay particle fellows did not have the same
settling pattern+ i.e., they penetrated deeper into the roughing filter.
.nd finally he had the feeling that the settling conditions were not always the same. There was a crowd
at the filter inlet and it was hard to find a free space to rest. 8owever, the inner filter part provided more
space, and settlement on the gravel grains was even supported by a sticky layer built on the gravel
surface. 8ence, filter efficiency not only depends on the size of the filter medium and particles, but also
on the actual filter load and biological filter activity.
6ur clay particle was pleased that his filter theory and conclusions were endorsed by some researchers
carrying out filtration tests. Their results and developed correlations between the different parameters
on horizontal roughing filtration are summarised in .nne, =. 8owever, based on his own e,perience
and reflections, our small filtration e,pert was convinced that this filter theory is basically also valid for
upflow and downflow roughing filters. 8ydrodynamic forces are present in all roughing filters where
quiescent conditions allow particles to settle. @evertheless, he was frustrated that his E4$( < 7$( filter
theoryE did not fit this ratio. *ecorded filter efficiencies are much smaller on account of the numerous
flow lines curving around the gravel grains.
2!dden li)eration
6ur puzzled clay particle became aware that his environment was deteriorating during his reflections on
filtration. Together with his colleagues he was densely packed in a structure made of decomposing
organic matter and hungry microorganisms. The water around them was also foul<smelling. 8owever,
this liquid hardly mi,ed with the water that was flowing gently at a very constant rate and under laminar
flow conditions through the filter. The clay particle had the strong urge to dive into this fresh water, or
even to swim away.
.ll of a sudden, the quiescent and dull conditions stopped. The filter was filled with tremendous noise
which felt worse than a giant earthquake. The water was shooting downwards through the gravel to the
filter bottom, and all the clay particles were an,iously holding each other. . side of his structure was
flushed away. 6ur terrified particle then saw a clump of particles roll down like an avalanche. The
dragging forces grew stronger, his structure collapsed and then everything went very fast. Under very
turbulent flow conditions, he was flushed to the filter bottom, pressed through a pipe where a valve
nearly broke his neck, and was finally discharged into a lagoon. 6ur old fellow gained his liberty and
felt like a new<born. 8owever, he was not used to the bright sunshine and decided to settle again, but
this time under different conditions.
-losing remark) since the comple, mechanisms of hydraulic filter cleaning are not yet fully e,plored,
our clay particle had no time for further philosophical contemplation on this matter. This is where our
e,cursion through the basic filter theory comes to an end.
N*1 Desi%n varia)les and %!idelines
The main ob1ective of roughing filters is the reduction of solid matter in the raw water from a specific,
in many cases however unknown concentration, to a level which allows a sound slow sand filter
operation. . turbidity value of about 43 73 @TU, or a suspended solids concentration of 7 < 2 mg$l, is
generally considered an adequate pretreated water standard for slow sand filtration. "urthermore, since
the roughing filters have to treat a certain volume of water per day, a reasonable operational period is
necessary between two filter cleanings. 0,plicitly, roughing filters have to meet the following t$ree
desi%n tar%etsD
4 red!"e t!r)idit& and s!spended solids "on"entration )& DC ,%:l. to a level re8!ired for
ade8!ate slo( sand filter operation
4 prod!"e a spe"ifi" dail& o!tp!t B ,O:d.
4 allo( ade8!ate operation d!rin% a deterined filter r!nnin% period T
,da&s or (ee/s.*
"ilter design has to meet these targets and is defined by the following si= desi%n varia)les which can
be selected within a certain range)
4 filtration rate or filter velo"it& v
4 avera%e si3e d%
,. of ea"$ filter edi!
4 individ!al len%t$ I
,. of ea"$ spe"ifi" filter edi!
4 n!)er n
of filter fra"tions
4 $ei%$t H ,. and (idt$ W ,. of filter )ed area A ,P.*
Filtration rate E
%enerall& ao!nts to C*1 4- :$* "ilters are occasionally operated at a filtration rate
of up to 4.2<7 m$h or even F m$h as in intake and dynamic filters. 8owever, the applied filtration rate
significantly influences filter performance although removal efficiency does not seem very much
affected in between varying filtration rates of 3.( and 3.: m$h J=;K. "iltration rate or filter velocity
e,pressed in m$h! is defined as the hydraulic load mb$h! applied to the filter and divided by the area
m]! of the filter bed perpendicular to the flow direction.
Ta)le 1 ?ravel Fra"tion 2i3es for Ro!%$in% Filters
Filter 7aterial 2i3e of Filter 7aterial d% KL
C$ara"teristi"s -st Fra"tion 0nd Fra"tion 1rd Fra"tion
coarse filter 7= < 4: 47 < 4; ; < 47
normal filter 47 < 4; ; < 47 = < ;
fine filter ; < 47 = < ; 7 < =
2i3e d%
of t$e filter aterial !s!all& ran%es )et(een 0C and G * The gravel should be rather
uniform to achieve large porosity. The uniformity coefficient U, defined here as quotient between the
largest and smallest size of a filter fraction U Q dg
i ma,
i min
!, should be in the order of 7 or less. "ilter
medium fractions as listed in Table ( would meet the recommended uniformity.
2in"e len%t$ I
of t$e filter aterial is dependent on t$e filter t&pe+ it a& var& %reatl&* 5ynamic
and intake filters acting as surface filters require a smaller filter depth of about =3 < :3 cm, compared to
roughing filters operated as deep bed filters. The depth of upflow and downflow roughing filters is
limited by structural constraints, however, it is generally between ;3 and 473 cm. The length of
horizontal flow roughing filters is, in this respect, not limited. 8owever, overall length normally lies
within 2 and 9 m.
N!)er n
of filter fra"tions is also dependent on filter t&pe* Surface filters might only need one
fraction whereas roughing filters are usually composed of three gravel fractions. The required overall
filter length can substantially be reduced with the use of differently graded filter fractions as illustrated
in "ig. (4. The bulk of the solid matter is removed by the coarse filter fraction, the medium sized gravel
has a polishing effect, and the finest gravel ought to remove only the remaining traces of solid matter.
Therefore, individual filter length l
of roughing filters are often designed in a ()7)4 ratio.
"ig. (4 Turbidity *eduction along a *oughing "ilter
Hei%$t H ,. and (idt$ W ,. are dependent on str!"t!ral and operational aspe"ts* Shallow
structures of about 4 < 7 m are recommended to avoid problems with respect to water tightness. In view
of a possible manual filter cleaning, 4<m deep structures are even recommended for easy removal of
filter material. The width of the filters should also be limited to allow efficient hydraulic cleaning and
avoid washwater disposal problems. Therefore, filter width should generally not e,ceed = < 2 m, and
filter surface area . for vertical flow filters s$o!ld not )e lar%er t$an 0H 4 1C P or G 4 I P ,"ross
se"tion area. for $ori3ontal4flo( ro!%$in% filters* Since these recommended ma,imum sizes limit
hydraulic filter capacity, several filter units operated in parallel are required to meet the requested
treatment plant output.
-onstruction of at least two parallel filter units is anyhow advisable to allow continuous treatment plant
operation even during ma1or maintenance and repair work.
(,2 Flo3 an healoss control
T$e $&dra!li" "onditions in ro!%$in% filters are deterined )& t$e filtration rate E
calculated as flow rate c mb$h! divided by the active cross<sectional filter area . m]!, e.g.)
m$h! Q c mb$h! $ . m]!
"or adequate filter performance, flo( "ontrol is essential and !st eet t$e follo(in% tar%etsD
< a=i! flo( through the treatment plant should be limited in general, and through the filter units in
< total flo( should be distributed evenly over the parallel running filter units
< "ontrolled (ater levels should be maintained within the filter units.
Weirs+ overflo( pipes and valves are !sed to "ontrol t$e flo( through the treatment plant and the
different filter units. /a,imum flow through the treatment plant is limited by an overflow generally
located at the intake. .t the treatment plant, the flow is equally divided by a distributor bo, or channel
into the different filter units. T$e siplest flo( "ontrol devi"e is a E4not"$ (eir* "inally, ma,imum
flow through the filter unit is limited by an overflow located upstream of the I<notch weir.
T$e o!tlet str!"t!re "ontrols t$e (ater level in roughing filters. Installation of a I<notch weir to
maintain a fi,ed water level is the simplest flow control option. 0ven a normal effluent pipe can keep
this water table at a constant level. 8owever, such an effluent pipe, without weir but connected to
ad1acent pipe installations, does not allow discharge measurements necessary for detection of possible
leaks in the filter structure.
"ilter resistance increases with progressive filter operation. "inal headloss in a roughing filter is usually
small+ i.e., 43 to 73 cm, or (3 cm at the most. Headloss variation in t$e filter "an )e re"orded )& t$e
(ater level in t$e inlet filter "opartent* . general layout of inlet controlled filter is illustrated in
"ig. (7. /ore details on discharge measurements are contained in .nne, 7.
A varia)le (ater level on t$e effl!ent side is achieved with the installation of a manually operated
valve, a self<regulating floating weir or a constant flow device as suggested in J=FK. 8owever, since
final headlosses for horizontal<flow roughing filters are relatively small, use of a variable effluent level
is not re"oended*
"ig. (( contains the main details of self4re%!latin% flo( rate devi"es* Such installations are !sef!l to
aintain a "onstant flo( t$ro!%$o!t t$e treatent plant in p!ped ra( (ater s!ppl& s"$ees+
particularly at night, provided a raw water balancing tank is installed for continuous raw water supply.
"ig. (7 Cayout of an Inlet<-ontrolled *oughing "ilter
(,4 Filter rainage system
.ccumulation of large volumes of solids in the filter media decreases filter porosity and ultimately also
filter efficiency and increases filter resistance. To maintain adequate filter performance and limit filter
headloss, periodi" reoval of t$e a""!!lated solids fro t$e filter edia is essential*
"ig. (( Cayout of "low *ate -ontrolling 5evices
*oughing filters are cleaned either manually or hydraulically. /anual filter cleaning e,cavation,
washing and refilling of the filter media! is labour<intensive and cumbersome. T$erefore+ $&dra!li"
filter "leanin% pla&s a /e& role in lon%4ter and effi"ient ro!%$in% filter operation*
8ydraulic filter cleaning entails fast filter drainage of the accumulated solids, which are flushed down
to the filter bottom, dragged to the drainage system and washed out of the filter. The following are most
iportant desi%n varia)les for $&dra!li" filter "leanin%D
4 filter draina%e velo"it& v
4 inlet area A
of t$e draina%e s&ste
4 $ori3ontal distan"e L
,. )et(een t$e drains or openin%s in t$e filter )otto
4 (as$(ater vol!e E
4 "leanin% fre8!en"& -:T
or filter r!nnin% period T
"ilter drainage velocity is identical with the dropping rate of the water table in the filter. Hi%$ initial
filter draina%e velo"it& v
is re"oended for effi"ient "leanin%* Turbulent flow conditions are
absolutely necessary for resuspension and transport of the accumulated solids through the filter.
Therefore, a drainage velocity of at least (3 m$h, or preferably :3 < F3 m$h, is required for efficient
hydraulic cleaning.
/a,imum drainage rate velocity is very much influenced by minimum cross section area avail able for
washwater flow. The cross section of the drainage pipes constitutes a bottleneck, the other limiting
factor is t$e overall inlet area A
of t$e draina%e s&ste+ ($i"$ s$o!ld )e desi%ned as lar%e as
possi)le* 'erforated false filter bottom systems provide a larger inlet area than a perforated drainage
pipe system.
.fter having been flushed to the filter bottom, the resuspended solids have to be transported to the inlet
of the drainage system. T$e $ori3ontal distan"e L
)et(een t$e openin%s of t$e draina%e s&ste
s$o!ld )e as sall as possi)le to prevent gradual accumulation of sludge at the filter bottom. 8ere
again, the installation of a false filter bottom is recommended since the washwater can be collected
more evenly than in a perforated drainage system, which should be installed with a small horizontal
distance C
of ma,imum 4 < 7 m between the drains.
H&dra!li" filter "leanin% is "arried o!t (it$ t$e (as$(ater vol!e E
stored in t$e ro!%$in%
filter* @ormal filter operation is interrupted and the drains opened. 8ence, compared to rapid sand
filters, hydraulic roughing filter cleaning does not require additional equipment, such as backwash
pumps or even air compressors. To prevent loss of washwater, fast openin% valves and %ates are
ne"essar& to make best use of the washwater stored in the filter bed. Since such devices of relatively
large diameter about 423 < 723 mm! are rather e,pensive, alternative installations, as presented in "ig.
(=, have been developed by local institutions. Fast openin% devi"es !st )e siple in desi%n+ st!rd&
and eas& to operate as (ell as (aterti%$t* "urthermore, they should be fitted with a closing device to
save washwater during drainage. Cocally manufactured devices need to be carefully field<tested prior to
their use in full<scale filter units.
"ig. (= Cayout of "ast 6pening 5evices for "ilter 5rainage
Cleanin% fre8!en"& or filter r!nnin% period T
of ro!%$in% filters is dependent on solid atter
load and )iolo%i"al a"tivit& in t$e filter* &eneral recommendations are not possible since each filter
operates under specific local conditions. @evertheless, periodic hydraulic cleaning is advisable to
prevent gradual accumulation of solids in the filter that filter drainage cannot remove due to
agglomeration and consolidation. -leaning frequencies could amount to once every one to two weeks
during the rainy season, and once every one to two months during the dry season. 8owever, very high
solid loads with turbidities S 4333 @TU call for daily hydraulic flushes. "urthermore, e,cessive
biological activities could hinder efficient hydraulic cleaning or affect taste and odour of the water.
Such conditions would also require frequent hydraulic cleaning. *esearch in the laboratory J44K with
biologically ripened roughing filters suggests that a drying period will have a positive impact on
hydraulic filter cleaning. 8owever, this observation is in contrast with the general recommendation on
keeping roughing filters always wet.
2afe disposal of t$e (as$(ater is iportant* "ilter flushing generates relatively large washwater
volumes up to 43 mb! within a short time about 4 < 7 minutes!. To prevent erosion in steep regions,
intermediate storage in a small, separately constructed pond may be necessary. Such an installation
would allow gradual and controlled discharge of the washwater in a water course or its agricultural use.
"urthermore, the solids washed out of the filter and settled in the pond are a valuable soil conditioner
and fertiliser.
The drainage system used for hydraulic filter cleaning might not be designed to drain the entire filter
structure. 8owever, complete drainage is required during maintenance and repair of the filter.
Additional sall draina%e installations are t$!s ne"essar& for "oplete reoval of t$e (ater
stored in t$e filter* "or this purpose, small drainage pipes equipped with taps or plugs can be used in
large roughing filters. Small structures can, however, be dewatered with buckets or a tube used as
(,) 0eneral esign aspects
Treatent fa"ilities $ave to )e diensioned for e=tree loadsQ i.e., in terms of solids removal for
ma,imum solids concentration in raw waters. 8owever, it is preferable to pretreat the ra( (ater in a
se8!en"e of different treatent !nits* &radual reduction of suspended solids, turbidity or pathogenic
microorganisms by a sequence of different pretreatment units probably offers the most economic option
with respect to investment and operating costs. Small pretreatment units, such as intake filters or
sedimentation tanks, can significantly reduce solid matter concentrations or turbidity peaks.
"urthermore, cleaning of these installations is generally easier than roughing filters. 8ence, roughing
filters should preferably not be designed to handle ma,imum water peaks, but pre<conditioned water
that has already been sub1ected to pretreatment.
Filter len%t$ and perissi)le filter r!nnin% period are "orrelated* 8orizontal<flow roughing filters
in particular were originally designed to provide a large silt storage capacity at low headloss, as filter
cleaning was carried out manually. *elatively important filter lengths of F to 47 m were the
consequence of this original design approach, permitting filter runs of several months, similar to those
of well<operated slow sand filters. 8owever, importance and benefits of hydraulic cleaning have
meanwhile been recognised.
C!rrent desi%n pra"ti"e tends to red!"e filter len%t$s and in"orporate effi"ient $&dra!li" "leanin%
fa"ilities* *egeneration of filter efficiency through frequent hydraulic cleanings has to counterbalance
shorter filter lengths.
T$e !se of saller filter aterial "an iprove filter effi"ien"&* 8owever, besides efficiency in
suspended solids separation, other criteria such as terminal headloss, filter running time and filter
cleaning aspects have to be taken into consideration. Use of only a uniform and fine filter material
allows sufficient pretreatment of the raw water, but at the e,pense of high head losses, short filter runs
and filter cleaning difficulties. T$e ro!%$in% filter te"$nolo%& re8!ires t$e !se of "oarse filter
aterial si3es )et(een 0C 4 G %raded in different fra"tions* 8owever, the use of filter material
coarser than 73 mm with lower removal efficiencies is not advisable as it would require longer filters to
achieve the same treatment efficiency. "urthermore, the filter material should not be smaller than about
= mm to facilitate hydraulic filter cleaning. These recommendations are not applicable to intake and
dynamic filters as these operate differently. 2in"e inta/e and d&nai" filters are )asi"all& s!rfa"e
filters+ t$e& re8!ire sall filter aterial si3es )et(een 0 4 R * These filters act as surface filters,
and their filter depth therefore has no great influence on the overall efficiency.
Filtration rate %reatl& infl!en"es filter effi"ien"&* Sedimentation is the main solids separation process
in roughing filters. Therefore, roughing filters must be operated under laminar flow conditions to
achieve adequate solids removal efficiencies. "low conditions are described by the *eynolds @umber.
.t a value of less than 43, laminar flow can be e,pected see also "ig. (2!. .s the *eynolds @umber is
directly proportional to the filter material size, ma,imum allowable filtration rate for laminar flow
conditions will be determined by the coarsest gravel fraction in a roughing filter. 8ence, for opti!
filter !se+ "oarser filter aterial re8!ires saller filtration rates* 8owever, filtration rate can only
partly be increased by applying smaller filter material, as particle size distribution of the solids and
suspension stability also determine the fillersD solids separation efficiency.
"ig. (2 *oughing "ilter 0fficiency in -orrelation to "low -onditions
African 2iller 3ees Supported Sustaina!ility
The rapid sand filters of Salaa# a district centre located in 4est Africa# were never filled with filter
media and the population of the town was therefore e"posed to unsafe water. The old treatment plant
had to !e e"tended and replaced !y an appropriate treatment scheme. 5ouhin and slow sand filters
were thus tested in a pilot plant shaded !y a lare !ao!a! tree. The location was chosen !y the
e"ternal e"perts sufferin from the Ghanaian sun. The pilot plant was constructed !y the local water
authority# and the e"patriates supervised the start of the field tests. +owever# African *iller !ees also
chose the same lare !ao!a! tree as their ha!itat and were very much attracted !y the white s*in of
the foreiners. 1nder remote supervision !y the e"patriates# the field tests were continued !y the local
staff which ained e"perience and confidence in a treatment scheme they wit6have to run in future at
full scare and without forein assistance. Thus# the African *iller !ees somehow contri!uted to
developin local sustaina!ility.
'+, Detaile filter esign
'+,' 5nta6e Filters
Inta/e filters are "o)ined (it$ (ater a)stra"tion str!"t!res and installed ne,t to small and narrow
river beds as illustrated in 'hoto 2 and "ig. (:. Inta/e filters are often !sed as first pretreatent !nit
in a water treatment scheme. . small weir regulates the water level of the surface water and channels
part of the flow into an ad1acent filter compartment. This filter bo, is filled with two gravel layers. The
top layer consists of relatively fine gravel of less than si, millimetres in diameter.
'hoto 2 0,ample of an Intake "ilter
"ig. (: Cayout of Intake and 5ynamic "ilters
The lower coarser gravel layer acts as filter support and allows an even abstraction of the prefiltered
water through perforated drainage pipes. The abstracted raw water, after being distributed evenly by a
small weir over the entire width of the filter bo,, flows gently over the gravel bed surface. 'art of this
water percolates through the gravel layers and the remaining water is discharged over an outlet weir
back to the river. Inta/e filters are "onstr!"ted alon% rivers and not directly in the river bed, as the
filter material would be washed out during periods of high river discharge. -onstruction of a separation
wall between river bed and filter bo, is recommended to prevent the filter from being washed out.
Inta/e filters "an also )e installed in t$e )ed of sall "anals* Upland rivers with a steep river bed
and a suitable topography might allow the accommodation of a small diverting canal. The filter bed,
comprising different gravel layers, is installed over a small stretch in the canal. 'art of the canal water is
filtered through the series of fine to coarse gravel layers, while the remaining water is returned to the
river. The prefiltered water is collected by perforated drainage pipes laid at the bottom of the coarse
gravel layer, and the discharge rate regulated by a valve placed at the filter control bo,. T$e flo(
velo"it& in t$e "anal !st )e re%!lated by the canalDs intake structures to protect the filter bed from
being washed out during periods of high river discharge. The flow velocity in the canal should actually
range between 3.43 and 3.(3 m$s to prevent fine matter from settling and remaining on top of the gravel
bed, and also to avoid fine filter material from being washed out. This layout may also be applied to
irrigation canals, provided they are continuously supplied and regulated through out the year. 8owever,
"onstr!"tion of inta/e filters alon% rivers is stron%l& re"oended as t$ese filters allo( a ore
relia)le operation t$an inta/e filters installed in "anal )eds*
"inally, Minta/eM filters a& )e lo"ated dire"tl& at t$e treatent plant site and function as
pretreatment facility. This particular location is recommended in gravity water supply schemes with a
raw water intake located in a remote area of difficuIt access. Such a layout will facilitate monitoring and
regular filter cleaning.
Filtration rates of inta/e filters ran%e )et(een C*1 and 0 :$* 8owever, significant solids removal
rates can be e,pected only at filter velocities smaller than 4 m$h, 5esign of the hydraulic structures
should be based on ma,imum filter resistance of 73 to =3 cm. This figure will not be e,ceeded if regular
filter cleaning, e.g. once a week!, is observed.
Relativel& sall filter aterial of less t$an I is !sed in inta/e filters which act as surface filters
as the solids mainly accumulate on top of the filter bed. Since filter cleaning is carried out manually, the
different gravel layers might be disturbed and mi,ed up if filter material of different sizes is used in
intake filters. . filter cloth is sometimes placed in<between the different gravel layers to avoid mi,ing
of the filter fractions and possibly reducing filter porosity and efficiency. 8owever, coarser gravel
hardly contributes to solids removal, but allows an even abstraction of the pretreated water.
@evertheless, regular filtered water collection is also possible with a single filter bed layer, moderate
filtration rates, medium<sized filter structures and reasonable layout of the perforated drainage pipes.
5esign guidelines are summarised in "ig. (9. Use of a single filter layer and false filter bottom offers a
favourable design alternative for intake filters, as filter material mi,ing is no longer possible, and even
abstraction of the prefiltered water is guaranteed.
"ig. (9 5esign of Intake and 5ynamic "ilters
'+,. Dynamic filters
D&nai" filters prote"t t$e treatent plant !nits fro $i%$ t!r)idit& pea/s* 8ighly turbid surface
water can quickly clog filters, especially slow sand filters. Therefore, during periods of e,tremely high
raw water turbidity, the flow may be interrupted to reduce cumbersome filter cleaning. Separation of
solids is only of secondary importance in dynamic filters.
5ynamic filter performance is, as described by its ad1ective, dynamic. The water quality between filter
inlet and outlet hardly changes during periods of low raw water turbidity. 5uring raw water turbidity
peaks, however, the quantitative change is drastic as no water is available in the filter outletL D&nai"
filters a"t li/e t!r)idit& eters connected to an open<close valve+ i.e., they rapidly get clogged when
raw water of high turbidity passes through the filter.
D&nai" filters are siilar in la&o!t to inta/e filters+ )!t differ in filter aterial si3e and
filtration rate* 0specially the gravel size of the top filter layer is smaller+ i.e., less than : millimetres in
diameter, while filtration rate is usually more than 2 m$h, /a,imum available headloss is still limited
and ranges between 73 and =3 cm in spite of finer filter material and greater filter velocity. "inally, the
horizontal flow velocity over the filter bed surface should be small or non<e,istent+ i.e., less than 3.32
m$s or nil, to prevent removal of accumulated silt during turbidity peaks.
D&nai" filters are "leaned after ea"$ ra( (ater t!r)idit& pea/* -leaning is also carried out
manually and in the same way as in intake filters. 5uring periods of dynamic filter interruptions,
treatment plant operation is reduced e.g. by declining filtration rate operation in slow sand filters!.
@evertheless, d&nai" filters s$o!ld onl& )e !sed (it$ ra( (ater e=perien"in% s$ort t!r)idit&
pea/sQ i.e., from a few hours to ma,imum half a day. 5ynamic filters are preferably located at the site
of treatment plants to facilitate monitoring and cleaning by the caretaker.
*ussian sanitary engineers introduced the idea of dynamic filters to .rgentina J23K, where about 23
filters were installed and operated in the late 4F93s. 8owever, filter design differs from the one
presented here. There raw water flows into a dissipation chamber and from there over an inlet weir on
top of a sand filter. The filter is operated in such a way that a water layer of a few millimetres flows
over the sand surface. This flow washes the solid matter deposited on top of the filter bed into a sand
recovery chamber installed at the end of the filter. 'ilot plant studies conducted by #atertek with #ater
*esearch -ommission funding to test and demonstrate the use of this filter design J24 K are in progress
in South .frica.
'+,7 8ertical-flo3 roughing filters
Ro!%$in% filters "an )e "onsidered a aSor pretreatent pro"ess for t!r)id s!rfa"e (ater since
they efficiently separate fine solid particles over prolonged periods. They are therefore placed at the
treatment plant site and operated in combination with other pretreatment units such as dynamic filters or
sedimentation tanks. *oughing filters precede final treatment processes, such as slow sand filtration and
Eerti"al4flo( ro!%$in% filters !s!all& "onsist of t$ree filter !nits arran%ed in series as shown in
"ig. (;. The water to be treated flows in sequence through the three filter compartments filled with
coarse, medium and fine filter material. The size of the three distinct filter material fractions is generally
between 73 and = mm, and graded, for e,ample, into fractions of 47<4; mm, ;<47 mm and =<; mm.
Eerti"al4flo( ro!%$in% filters operate eit$er as do(nflo( or !pflo( filters* They are hence either
supplied by inflowing water at the filter top or at the filter bottom. The filter material of vertical<flow
roughing filters is completely submerged. . water volume of about 43 cm depth usually covers the
gravel. The top should be covered by a layer of coarse stones to shade the supernatant water and thus
prevent algal growth often e,perienced in pretreated water e,posed to the sun. 5rainage facilities,
consisting in perforated pipes or a false filter bottom system, are installed on the floor of the filter
bo,es. "inally, pipes or special inlet and outlet compartments are required to convey the water through
the subsequent three filter units.
"ig. (; Cayout and 5esign of Iertical<flow *oughing "ilters
Eerti"al4flo( ro!%$in% filters are !s!all& operated at C*1 to -*C :$ filtration rates* Iertical<flow
roughing filters may be sensitive to hydraulic fluctuations, especially if loaded with large amounts of
solids. Settled matter might be resuspended at increased filtration rates, causing solids to break through
the filter. "ilter operation at constant flow rates is, therefore, recommended. *aw water containing
colloidal matter and a high suspension stability should be treated at low filtration rates and preferably
with fine filter material. "ilter resistance is usually less than 73 cm per filter unit and, hence, not a
decisive operational criteria for properly designed and operated roughing filters.
5ue to structural constraints, verti"al4flo( ro!%$in% filters $ave a relativel& sall filter dept$ of
about 4 m. Total filter depth of the three filter units used in series is thus ( m. T$is total availa)le filter
dentin liits verti"al4flo( ro!%$i% filter appli"ation* It can generally and efficiently handle moderate
raw water turbidities of 23 to 423 @TU. *aw water pretreatment by intake filters, reduction of filtration
rate or provision of additional filter bo,es would be required to treat raw water of higher turbidities.
In verti"al ro!%$in% filters in la&ers+ where all three gravel fractions are installed in one filter bo,
with a total filter distance of about 4 m, lo( t!r)idit& ra( (ater "an )e pretreated* 8owever, due to
cleaning aspects this filter design can only be used for upflow operated roughing filters. In such filters,
the coarse filter material is placed at the bottom and the finest material at the top of the filter. The
separated solids, which accumulate mainly in the coarse filter fraction ne,t to the filter bottom, can be
easily flushed out with the water stored in the filter. Therefore, t$e !se of !pflo( ro!%$in% filters in
la&ers is re"oended* 5own flow roughing filters in layers face considerable problems with
hydraulic filter cleaning. The bulk of solids accumulated in the coarse filter material on top of such
filters would have to be flushed through the finer rather clean filter material and would thus soil the
entire filter bed.
Ade8!ate and effi"ient (as$(ater "olle"tion is iportant for reliable roughing filter operation.
'erforated pipes or a false filter bottom can be installed in vertical<flow roughing filters. 'erforated
pipes, which should be laid in a coarse gravel pack to support an even washwater abstraction, would
require the installation of additional filter material in vertical<flow roughing filters. 'reference is given
to false filter )ottos as they allow an even washwater abstraction and do not require additional gravel
layers. .lthough special perforated concrete slabs will be necessary, they may be readily produced
. comparison of do(nflo( (it$ !pflo( ro!%$in% filters reveals the following)
Dire"tion of flo( and sedimentation are obviously the first differences which might interfere or
support solids settling on the filter material. Solid removal efficiency should consequently vary in the
two filter types. Theoretically, downflow filters should have a better performance than upflow filters as
the solid particles are more likely to settle on top of the gravel surface in the direction of flow than
under countercurrent conditions. 8owever, practical field e,perience has shown a siilar effi"ien"&
for )ot$ filters* In dead filter zones, where the water flow is reduced to a minimum, solids settle
regardless whether the roughing filter is operated in upflow or downflow direction. 8ence, filter
efficiency is similar in both filter types.
T$e a""!!lation pattern of retained solids is another difference between downflow and upflow
filters. The bulk of the solids is deposited at the inlet of the filter+ i.e., for downflow filters in the upper
part of the filter, and for upflow filters in the filter medium located ne,t to the filter bottom. This,
however, has a treendo!s ipa"t on $&dra!li" filter "leanin%* In downflow roughing filters, the
buIk of accumulated solids has to be flushed with a relatively small washwater volume from the soiled
filter top through the lower and cleaner filter part to the filter bottom. The opposite is true for upflow
roughing filters. The bulk of retained solids is accumulated ne,t to the drainage system and a relatively
large washwater volume, accommodated in the upper filter part, is available to flush the solids out of
the filter. O(in% to t$e iportant filter "leanin% aspe"t+ !se of !pflo( ro!%$in% filters rat$er t$an
do(nflo( filters is re"oended*
The 3athtu! !y the 5iverside
The 3athtu! !y the 5iverside
7iaojian township# located !y the 8iantan river# the larest stream in the province# faced
tremendous pro!lems with its water supply scheme. The raw water was pumped over a distance of
a!out $ *m from the river to the treatment plant situated to the water tower in the centre of the town.
In $9:;# a pu!lic water supply was constructed usin the <traditional< water treatment technoloy#
e.. coaulants and !leachin powder were added to the raw water# which was then pumped into a
sedimentation tan* and thereafter filtered throuh rapid sand filters +owever# the river is *nown to
carry lare silt loads especially durin the rainy season. These solids# accumulated in the raw water
main# increased the headloss in the pipe and seriously reduced its hydraulic capacity. )oreover# the
water consumers drin*in a lot of tea were always complainin a!out the unpleasant taste of the
distri!uted water.
5eha!ilitation of the water treatment !ecame inevita!le after si" years of operation. Em!arrassed !y
the situation# )r 7u and his team transformed the sedimentation tan* into two upflow rouhin filter
units# and constructed two new slow sand filters. All these installations were !uilt !ehind the
surroundin wall of the e"istin treatment plant and did not attract much pu!lic attention. +owever#
these modifications did not solve the cloin pro!lem in the raw water pipe# and the small raw water
pumpin house could not accommodate any installation larer than the common chemical dosin
e(uipment. The project team admitted their desin errors and decided to e"pose themselves to pu!lic
criticism !y erectin a audy and flashy structure.
To avoid a second pumpin step# two additional upflow rouhin filters were constructed on an
elevated positron ne"t to the dam of the river. The walls were covered with nice white tiles as
commonly used in !athrooms. +owever# )r 7u and his team convinced !y the efficiency of rouhin
and slow sand filters had# on this occasion# made the riht decision. 5aw water tur!idity of %&& 0T1
is now reduced to ;& /$:& 0T1 !y the upflow rouhin filters# to $. / =& 0T1 !y the second prefilters#
and the filtrate of the slow sand filters is now always less than % 0T1. Even more important# the
complaints of the consumers have ceased since chemicals are no loner used to treat the water# and
the pu!lic has no reason to criticise the !athtu! !y the riverside.
'+,2 9ori:ontal-flo3 roughing filters
Unliited filter len%t$ and siple la&o!t are t$e ain advanta%es of $ori3ontal4flo( ro!%$in%
filters* &enerally, the shallow structure does not create structural problems, and the filter length is not
limited to a few metres. "urthermore, its simple layout does not require additional hydraulic structures
and installations as in vertical<flow roughing filters. The raw water runs in horizontal direction from the
inlet compartment, through a series of differently graded filter material separated by perforated walls, to
the filter outlet as illustrated in "ig. (F. Filter aterial also ran%es )et(een 0C and G in si3e+ and
is !s!all& distri)!ted as "oarse+ edi! and fine fra"tion in t$ree s!)se8!ent filter
"opartents* To prevent algal growth in the filter, the water level is kept below the surface of the
filter material by a weir or an effluent pipe placed at the filter outlet.
"ig. (F Cayout and 5esign of a 8orizontal<flow *oughing
Filtration rate in $ori3ontal4flo( ro!%$in% filters ran%es )et(een C*1 and -*H :$* It has been
defined here as hydraulic load mb$h! per unit of vertical cross section area m]! of the filter. "ilter
length is dependent on raw water turbidity and usually lies within 2 to 9 m. 5ue to the comparatively
long filter length, $ori3ontal4flo( ro!%$in% filters "an $andle s$ort t!r)idit& pea/s of HCC to -+CCC
Draina%e fa"ilities+ s!"$ as perforated pipes+ tro!%$s or "!lverts+ allo( $&dra!li" filter )ed
"leanin%* These drainage systems are placed at the filter bottom perpendicular to the direction of flow.
5rainage facilities in flow direction must be avoided as they could create short<circuits during normal
filter operation. 8ence, false filter )otto s&stes "annot )e installed in $ori3ontal4flo( ro!%$in%
filters* Since most of the solids accumulate at the inlet of each filter medium, draina%e fa"ilities
s$o!ld )e pla"ed at t$e inlet of ea"$ filter "opartent to enhance hydraulic cleaning efficiency.
Installation of troughs complicates construction of the filter bo, floor. "urthermore, since the horizontal
distance C
between the troughs is usually large, an even abstraction of the sludge is correspondingly
difficult. Therefore, !se of perforated pipes is t$e )est draina%e s&ste for $ori3ontal4flo(
ro!%$in% filters+ as it allows easy installation of a dispersed system. .lthough prefabricated culverts
may allow a more even solids removal, connection to the washwater effluent pipes is more complicated.
Hori3ontal4flo( ro!%$in% filters $ave a lar%e silt stora%e "apa"it&* Solids settle on top of the filter
medium surface and grow to small heaps of loose aggregates with progressive filtration time. 'art of the
small heaps will drift towards the filter bottom as soon as they become unstable. This drift regenerates
filter efficiency at the top, and slowly silts the filter from bottom to top. 8orizontal<flow roughing filters
also react less sensitively to filtration rate changes, as clusters of resuspended solids will drift towards
the filter bottom or be retained by the subsequent filter layers. 8orizontal<flow roughing filters are thus
less susceptible than vertical<flow filters to solid breakthroughs caused by flow rate changes. 8owever,
they may react more sensitively to short circuits induced by a variable raw wafer temperature.
Periodi" "leanin% is also essential for $ori3ontal4flo( ro!%$in% filters* 8ydraulic cleaning is carried
out by fast drainage of the water stored in the filter. 5uring filter drainage, the small unstable heaps of
accumulated solids collapse and are flushed towards the filter bottom. The solid matter stored in the
filter material is washed out of the filter bo, through the drainage system. Draina%e velo"ities of IC to
NC :$ are ne"essar& to achieve a good hydraulic cleaning efficiency. 5rainage pipes of adequate size
are required to achieve the recommended velocity which drains the filter within 4 to 7 minutes.
5epending on the solids concentration in the raw water, regular hydraulic filter cleaning, at intervals of
every few weeks, is required to avoid deterioration of filter efficiency and development of e,cessive
filter resistance. 8owever, filter resistance will not e,ceed 73 cm if normal filter operation and regular
cleaning are observed. "requent and efficient filter drainages will also defer the need for manual filter
cleaning, which nonetheless becomes unavoidable after some years of filter operation.
'hoto : Inside Iiew of a *oughing "ilter >ed during 8ydraulic -leaning
>earnin from Developin Countries
>earnin from Developin Countries
4ater demand of 3asle-s alomeration has !een increasin radually due to industrialisation and
miration into the prosperin area. 0ewly constructed !uildins and hihways have contri!uted to
surface sealin and have reduced natural roundwater rechare. Increased water demand and
reduced recharin rate have led to an alarmin drop in the roundwater ta!le. The roundwater
wells almost ran dry. In order to reverse this situation# an artificial roundwater rechare plant was
constructed in Aesch in the early $9?&s.
5aw water was pumped from the river 3irs into a laoon for coarse matter separation from where it
was conveyed to hori@ontal/flow rouhin filters for fine solids separation. It was then aerated !y
cascades !efore flowin into the lare laoon which acts as slow sand filter. The treated water was
finally led to rechare wells throuh the impervious top layer into the a(uifer. Althouh the oriinal
capacity of the plant was desined for A&& l6s# operation had to !e reduced to .&& l6s due to
operational pro!lems. Inade(uate solids removal efficiency and radual siltin of the rechare wells
were the reasons for the reduced treatment plant operation and for the repeated drop in the
roundwater ta!le.
The hori@ontal/flow rouhin filters were desined accordin to the layout illustrated in ,i. ... The
$%/m lon rouhin filters# filled with one ravel fraction amountin to %& / :& mm in si@e# were
operated at %/$& m6h filtration rates. This inappropriate desin and operation resulted in poor solids
removal efficiencies. The slow sand filter was rapidly cloed so that the partly treated water ran
throuh the coarse material of the em!an*ment directly into the rechare well. ,urthermore# since
the rouhin filters could not !e cleaned hydraulically# the ravel had to !e replaced every si" years /
a costly underta*in which caused headaches to the manaement of the water authority.
In the last decade# the rouhin filter technoloy has# however# !een developed to a via!le treatment
alternative in the developin countries. The responsi!le water authority had access to the necessary
information and field/tested the new rouhin filters. The comparative test results revealed important
improvementsB i.e.# the small upflow rouhin filters with a total filter lenth of $ m had up to a si"
times hiher solids removal efficiencies than the old filters. Since they could also !e cleaned
hydraulically operatin costs were reduced The hih/tech society was ama@ed !y the efficiency of this
low/cost technoloy.
'', Roughing filter efficiency
'',' Practical e;perience
Treatent effi"ien"& is dependent on ra( (ater "$ara"teristi"s+ la&o!t and operation of ro!%$in%
filters* 6n the one hand, size, concentration, type of particles and suspension stability are the most
important water quality parameters influencing suspended solids removal efficiency. 6n the other hand,
filter material size and filter length, applied filtration rate and cleaning frequency are the key factors
determining filter efficiency. 8ence, roughing filters with identical layout and operation may vary in
filter performance with different raw water sources. 0ven a specific filter will most probably not have a
constant filter efficiency with the same raw water source) high particle removal rates will be recorded
during periods of high raw water turbidity where a slower rates will be e,perienced during periods of
moderate raw water turbidity. Therefore, an e,act indication of filter efficiencies is generally quite
Treatent effi"ien"ies of different ro!%$in% filters $ave )een st!died e=tensivel& )& CINARA KGJL
at a pilot plant in P!erto 7allarino+ Cali+ Colo)ia reported also in J=;K!. These field tests are
considered the most comprehensive pilot plant studies for roughing filter development. The pilot plant
consists of a first pretreatment step using intake$dynamic filters to precondition the raw water drawn
from the -auca *iver. The flow is then split into five lines where the filter performance of different
roughing filters is tested in combination with identical slow sand filters used as reference. The
following types of roughing filters are installed at 'uerto /allarino)
upflow roughing filter in series U*"S
upflow roughing filter in layers U*"C
modified horizontal<flow
roughing filter /8*"
horizontal<flow roughing filter 8*"
downflow roughing filter in series 5*"S
.ll roughing filters have similar gravel fractions but differ in filter length. Total filter length of U*"S,
/8*" and 5*"S, amounts to =.=3 m. Total filter length of the 8*" unitis 9.43 m and 4.:3 m for the
U*"C unit. The slow sand filter units are circular in shape, 7.33 m in diameter and 7.33 m in height.
They were filled with a 4 <m deep sand layer, which was gradually reduced due to subsequent sand
cleanings, but never fell below 3.:3 m. The sand has an effective diameter of 3.7 mm and a uniformity
coefficient of 4.29.
"ig. =3 summarises filter effi"ien"ies of t$e different ro!%$in% filters (it$ respe"t to t!r)idit&
reoval at different filtration rates and two distinct raw water turbidity levels. The graphs show higher
removal rates of generally ;2 < F3B or more for periods of high turbidity 423 < 233 @TU!. "ilter
efficiency is reduced to about ;3 ;2B or less during periods of moderate turbidity (3 < 23 @TU!, and is
hence in accordance with the general filter theory. The different but small filtration rates had no
significant influence on the turbidity removal efficiencies of the filters, as laminar flow prevailed in all
gravel fractions also at the highest filter velocity of 3.:3 m$h, The upflow roughing filter in series and
the horizontal<flow roughing filter unit e,hibited best performance throughout all test conditions. The
smallest turbidity removal efficiency was achieved by the downflow roughing filter unit.
"ig. =4 elucidates t$e effi"ien"& of t$e different pilot plant treatent steps (it$ respe"t to
s!spended solids and fae"al "onfor red!"tion* The intake$dynamic roughing filters reduced the
average suspended solids concentration by 22B from about 733 mg$l to F3 mg$l. This concentration was
further reduced to less than 2 mg$l by the roughing filters and the filtrate of the slow sand filters had an
average suspended solids concentration of 3.7 < 3.( mg$l. The relatively high suspended solids
concentration of the upflow roughing filter in layer might be an indication of a comparatively low
process stability of this filter.
.verage faecal coliform concentration of the raw water of about =3,333 -"U$433 ml was subsequently
reduced to about 7=,333, =33 and to less than 4 -"U$433 ml by the treatment scheme consisting of
intake$dynamic filters, roughing filters and slow sand filters. The modified horizontal<flow roughing
filter had the smallest faecal coliform removal efficiency of F:.2B. This rate also influenced
performance of the subsequent slow sand filter, which produced an average effluent of 7.: -"U$433 ml.
.ll the other slow sand filters had average faecal coliform concentrations of less than 4 -"U$433 ml in
their effluents.
"ig. =3 Turbidity *emoval by 5ifferent roughing filters
Fi%* G- do"!ents t$e $i%$ treatent effi"ien"& of t$e pilot plant* The two pretreatment steps and
the slow sand filters were able to reduce the suspended solids concentration from about 733 mg$l to
about 3.7 mg$l, or by ( log, whereas the faecal coliform concentration was reduced from about =3,333
-"U$433 ml to generally less than 4 -"U$433 ml, which corresponds to a = < 2 log reduction.
"ull<scale treatment plants are not so e,tensively monitored and controlled as pilot plants. @evertheless,
Fi%* G0 do"!ents t$e developent of )iolo%i"al filter a"tivities in t$e filters of a "o!nit&
(ater s!ppl& KH0L* The raw water of Ca `averianaDs treatment plant originates from the 'ance *iver, a
highland river of moderate turbidity. The water is treated by an intake filter, two horizontal<flow
roughing filters, and two slow sand filters operated at 4.(, 3.: and 3.3; m$h filtration rates. "aecal
coliform concentration ranging between about 4,333 and 43,333 -"U$433 ml is proof of relatively high
faecal contamination of the raw water. Turbidity of about 73 @TU is relatively low during the dry
periods, but increases to short turbidity peaks after periods of precipitation. The apparent colour
averaging about 433 -U$I follows a similar pattern as turbidity. /ean faecal coliform concentration in
the pretreated water amounted to about 733 -"U$433 ml and did not decline during the monitoring
period of half a year. .lthough the treated water had initially somewhat elevated faecal coliform
concentrations of more than 43 -"U$433 ml, the effluent concentration levelled out to about 4 -"U$433
ml after three weeks of operation. This corresponds to the period of maturation of the slow sand filter.
The overall turbidity and apparent colour reduction, however, indicated a distinct improvement within
the first si, months of operation. #ith progressive filter operation, the respective treatment efficiencies
also increased in the roughing filter, most probably on account of the gradual development of biological
processes in this filter.
Ta)le G s!arises treatent effi"ien"ies of ro!%$in% filters operated at different flo( dire"tions*
The filter material in the downflow and horizontal<flow roughing filter is rather coarse compared to the
one used in the upflow roughing filter operated, however, at more than double the normal filter velocity.
@evertheless, in all three treatment plants, turbidity reduction by the roughing filters amounts to about
93 < F3B. The bacteriological water quality improvement was about of the same order for these three
treatment plants.
Treatent effi"ien"& of ro!%$in% filters is also liited as they are not capable of treating any type of
water as illustrated by the following e,ample J2:K. -onstruction of a water supply scheme was one
component of the Caka Caka multi<purpose pro1ect. The pro1ect team decided to draw raw water from
the newly constructed water irrigation reservoir to supply their scheme. 8igh raw water turbidity lead to
e,tremely short filter runs of the two slow sand filters. Two horizontal<flow roughing filters were
therefore designed on the basis of the available literature data in order to improve operation of the slow
sand filters. The 4;<m long roughing filters were operated at 7.2 m$h filtration rate. Since filter
efficiency was very poor at this rate, filter velocity was gradually reduced to 3.2 m$h, but without
achieving a substantial treatment efficiency improvement. These problems were encountered mainly
because roughing and slow sand filters have never been used in this area before. 8owever, they could
have been avoided by pilot plant tests which are strongly recommended in such a situation! conducted
during the pro1ect design phase.
)icro!ioloical 3enefits from Gravel ,ilters
)icro!ioloical 3enefits from Gravel ,ilters
In $9??# a water supply was constructed in the community of Cocharcas located in an aricultural
area of the hih Andean Sierra and num!erin a!out $#&&& villaers. Its typical desin comprises an
a!straction from an irriation canal# a sedimentation tan*# two slow sand filters a reservoir# and a
distri!ution system with sinle/tap household connections. 3y $9:%# the water supply scheme had
seriously deteriorated. The system was only wor*in two to three hours a day owin to the fre(uent
use of the irriation canal !y other villaers. 0either the inta*e nor the treatment plant had a flow
control system. The careta*er was only concerned with fillin up the reservoir as (uic*ly as possi!le.
Thus# since the slow sand filters were heavily overloaded and without treatment efficiency# faecally
polluted water was supplied to Cocharcas. 5eha!ilitation of the system was thus necessary and an
areement on water use with the neih!ourin communities was therefore concluded A small weir
installed at the inta*e site revealed the limited need for water of the villae as compared to that
re(uired !y ariculture. To allow ade(uate operation of the slow sand filters and as additional raw
water pretreatment step# two ravel filters were constructed with community participation. The
treatment plant is now runnin at a constant flow and the efforts were rewarded !y the followin
water (uality improvement dataC
,aecal Coliforms mean D reduction D reduction
Ecounts6$&& mlF value per trmt step cumulative
raw water .&#&&& / /
sedimentation tan* $A#%&& .? .?
ravel filter $#.&& 9. 9A
slow sand filter. .& 9: 99#9
The Andean e"perience reveals that ravel prefilters enhance slow sand filter. operation and increase
the overall plant performance. The multiple !arrier system proves to !ean appropriate concept for
rural water supply.
"ig. =4 Suspended Solids and "aecal -oliform *eduction by *oughing and Slow Sand "iltration
'',. Pilot plant tests
.s described in the previous chapter, t$e (orld(ide e=perien"e (it$ ro!%$in% and slo( sand filters
do"!ents t$e si%nifi"ant potential of t$is treatent "on"ept in producing potable water from
polluted turbid surface water. There is no doubt about the general treatment efficiency of slow sand
filters, since a biologically mature filter will consistently reduce the concentration of microorganisms by
7 to = log FF to FF.FFB reduction!. It is more a question of specific treatment efficiency of roughing
and slow sand filters when fed with a local raw water source. 8ence, treata)ilit& of a parti"!lar ra(
(ater is of aSor "on"ern to design engineers, especially if local practical e,perience with the
considered treatment process and raw water source is not available.
"ig. =7 Turbidity, .pparent -olour and "aecal -oliform *eduction at the Treatment 'lant Ca `averiana,
Ta)le G E=aples and Pra"ti"al E=perien"e (it$ Ro!%$in% Filters
Cayout and 'erformance .zpitia, 'eru 0l *etiro, -olombia >lue @ile 8ealth 'ro1ect, Sudan
type of roughing filter downflow upflow
multi<layer filter!
filtration rate 3.(3 m$h 3.9= m$h 3.(3 m$h
design capacity (2 mb$d 9F3 mb$d 2 mb$d
filter length :3 cm, " =3<72 73 cm, " 4; 793 cm, " 72<23
and size mm! :3 cm, " 72<47 42 cm, " 47 ;2 cm, " 42<73
of material :3 cm, " 47< : 42 cm, " : ;2 cm, " 2<43
42 cm, " (
turbidity @TU!
raw water 23 < 733 43 < 423 =3 < 233
prefiltered water 42 < =3 2 < 42 2 < 23
faecal coliforms J$433 mlK
raw water
933 4:,333 S (33O
prefiltered water
4:3 4.:;3 R 72O
reference J2(K J2=K J22K
O as 0.coli
>imits of 5ouhin ,ilters
>imits of 5ouhin ,ilters
The >a*a/>a*a Project in the Andean reion is a multi/purpose project aimin at providin water to a
lare irriation scheme and to Tarata# a small town of =#&&& inha!itants located ne"t to the newly
constructed ravity dam. The water in the reservoir was !rownish in colour and had to !e treated
!efore it could !e used as drin*in water. A local non/overnmental oranisation E0G'F e"perienced
in community wor* !ut lac*in technical *now/how in water supply was responsi!le for project
implementation. 4ater treatment desin was !ased on some >iterature athered !y the local staff.
Two lare slow sand filter. units were constructed in a first phase# however# they were rapidly cloed
!y the tur!id raw water. The literature was aain consulted and the staff decided to construct two
hori@ontal/flow rouhin filters as recommended in the manual Each of these filters had a total filter
lenth of $: m and were operated at a flow rate of more than ..% m6in. +owever# tur!idity could s*ill
not !e reduced sinificantly !y the prefilters. Since treatment plant construction had !een carried out
with community participation# Tarata-s citi@ens e"pected a ood water (uality after their su!stantial
contri!ution towards the treatment plant construction. 5elations !etween field staff and local
population !ecame tense.
As rouhin filters had not !een used in their country so far# the 0G' souht advice from water
treatment e"perts of a neih!ourin country. Ana )ara was contracted as consultant to the project.
Durin her first field visit# she went to inspect the reservoir and saw the same !rownish raw water
flowin out of the rouhin filters operated at reduced flow rates as recommended !efore she visited
Tarata. The last rain had ended three months prior to her visit# however# the raw water stored in the
reservoir had hardly chaned appearance since then. Tur!idity was still around =&& 0T1. Ana
)ara-s preliminary conclusion was confirmed !y la!oratory tests which revealed that the raw water
was carryin a lare amount of colloidal matter which could not settle in this sta!le suspension.
,ield tests had to !e conducted to determine the most ade(uate treatment scheme for this cliff cult raw
water. The local 0G' set up a small pilot plant accordin to the desin of the consultant. It was
located ne"t to the inlet of the old ravel filters consistin in three concrete rins used as upflow
rouhin filters and a larer rin simulatin the slow sand filter. The three columns were filled with
ravel from the old filter# however# alum sulphate and lime were added to the effluent of the first
column. +ence# the second two units were operated as contact filters and their efficiency was
e"traordinary. The raw water tur!idity of a!out =%& / A&& 0T1 was reduced to a!out =A& 0T1 in the
first filter column# to a!out .& 0T1 in the second# and to $.% 0T1 in the third filter column at a
dosae of A& m6l alum sulphate and a p+ of $& adjusted with the addition of lime. The pilot plant
results were used as desin values for the reha!ilitation of the treatment plant.
This e"perience proves that each treatment process has its limitation just as the competence of field
. wide range of raw water qualities can theoretically be treated by conventional water treatment
processes+ i.e., coagulation, flocculation, sedimentation, rapid sand filtration and chlorination, as these
treatment schemes are fle,ible with regard to chemical dosage, detention time, hydraulic loads, and
water pressure. In contrast to such schemes, roughing and slow sand filters are rather limited in
operational fle,ibility but provide high process stability. O(in% to t$e sipli"it& of t$e ro!%$in% and
slo( sand filter te"$nolo%&+ onl& t$e follo(in% t$ree salient points (ill $ave to )e ans(ered )&
pilot plant st!diesD
"an ro!%$in% filters red!"e ra( (ater t!r)idit& to a level re8!ired for reasona)le slo( sand
filter operation;
at ($at rate does t$e $eadloss in t$e slo( sand filter in"rease;
on ($at desi%n val!es s$o!ld t$e proSe"t la&o!t )e )ased;
T$e first 8!estion "entres on pretreatent effi"ien"& of ro!%$in% filters with regard to turbidity
reduction. . comprehensive literature review J49K reveals that an upper turbidity limit amounting
between 2 and 73 @TU will allow a reasonable slow sand filter operation. The most adequate
pretreatment scheme is dependent on the raw water characteristic. 8owever, inforation on ra(
(ater 8!alit& is 8!ite often s"ar"e or not availa)le+ especially for small surface water sources. The
lack of water quality data can be partly offset by a characterisation of the raw water source. . highland
river, lowland stream or a still surface water will carry different types of solid matter. .n inspection of
the water course and the study of sludge deposits in the river bed and on the embankment will provide
information on type and size of solids found in the water during periods of high discharge. Interviews
with the local population might provide some details on seasonal water level fluctuation and type of
turbidity colour, period of turbidity peaks! over the year. "inally, simple tests, such as settleability and
suspension stability tests as described in .nne, 4, will generate valuable information on the separation
characteristics of the solids. The information obtained will help to select the type and number of
pretreatment steps as presented in -hapter 47. Effi"ien"& of t$e "$osen pretreatent s"$ee (ill
t$en )e assessed in a sall pilot plant as o!tlined in Anne= H*
T$e se"ond 8!estion fo"!ses on t$e developent of $eadloss in t$e slo( sand filter* The rate of
filter resistance increase is crucial for the determination of slow sand filter performance in treating a
specific pretreated raw water. Slow sand filter runs time between two consecutive filter cleanings! of at
least one month should be possible. "requent filter cleaning would interfere with the biological
activities concentrated mainly on the filter bed+ i.e., in the so<called ESchmutzdeckeE. Slow sand filters
with short filter runs of a few days act predominantly as physical filters and are thus not able to
substantially improve the microbiological water quality. 5evelopment of the ESchmutzdeckeE is known
as ripening period and greatly depends on the organic load and biological raw water composition. This
biological layer gradually grows on top of the sand bed and becomes more compact with progressive
filtration time. Thus, the water level will start to increase in an inlet controlled slow sand filter. The
filter has to be cleaned the moment the supernatant water level reaches the overflow. Since every water
is unique with respect to the building up of the ESchmutzdeckeE, development of the headloss should be
recorded by simple piezometer readings. Operational "$ara"teristi"s of slo( sand filters s$o!ld
t$erefore )e st!died )& pilot plant tests if practical e,perience is not available locally.
T$e last 8!estion ais to optiise treatent plant desi%n* Size and filter layout may be modified
once the selected treatment scheme has proven its general suitability. The hydraulic filter load has a
direct impact on the size of the structures as it influences them proportionally. 8owever, ro!%$in% and
slo( sand filters s$o!ld not )e desi%ned on t$e )asis of t$eir $i%$est perissi)le filtration rate*
Use of more conservative values is recommended, e.g. a filtration rate of 3.4 3.7 m$h for slow sand
filters. 8igher filtration rates may anyhow be necessary with increasing water demands. 8owever,
installation of a prefilter as a first pretreatment step, use of finer filter material in the roughing filters or
reduction of the depth of the sand could offer "ost savin% alternatives ($i"$ "an )e assessed )& pilot
plant st!dies*
Additional aspe"ts are, however, necessary in the planning and implementation phase of pilot plant
studies. These will include)
4 lo"ation of t$e ra( (ater inta/e should, whenever possible, be identical with that of the planned
treatment plant. 8owever, the raw water supply might create some problems, especially if the water has
to be pumped. In this case, tapping of raw water from another water supply system using the same raw
water source should be investigated. . continuous raw water supply is essential for adequate pilot tests.
'rovision of a raw water tank is necessary if water cannot be pumped or supplied continuously.
4 a""ess to t$e pilot plant should be easy and possible throughout the year. The site should be protected
from vandalism and robbery and allow easy discharge of filter drainage water. .vailability of a small
storage room for field test equipment near the pilot plant would be advantageous.
4 t$e aterial re8!ired for t$e pilot plant should be available locally. 'I- or concrete pipes with a
minimum diameter of (3 cm can be used to simulate vertical flow roughing or slow sand filters. Slow
sand filters can also be installed in ferro<cement tanks. 8orizontal<flow roughing filters however require
the construction of open channels to allow adequate filling of the filter material. *ecommendations and
design e,amples of pilot plant layouts are illustrated in .nne, 2.
4 t$e pilot plant tests s$o!ld last for several months. The field test period should at least cover the
period with the highest raw water turbidity and ideally run over a full year. To gain preliminary
operational e,perience with the pilot plant, field tests should start a few months prior to the e,pected
turbidity peaks.
4 pilot plant operation must be performed at constant flow conditions. &eneration of conclusive test
results requires unchanged operational conditions throughout a complete filter run. Since slow sand
filter test runs might last more than a few months, operation of filter units in parallel are recommended
for the study of different design options.
4 t$e pilot plant s$o!ld )e onitored by local people trained on the 1ob. . daily visit should entail the
control of the flow through the filters, measurement of the raw water turbidity and that of the different
treatment steps, as well as recording of the headloss development in the slow sand filter. .dditional
water quality analyses with field test kits or in the local laboratory are included in the monitoring
programme. . monitoring programme is outlined in .nne, 2.
4 t$e field test report presents the results of the monitoring programme, evaluates the data and draws
conclusions for the design of the full<scale treatment plant. The large number of data are best reported
in tables and in the form of graphs. #ater quality data are usually graphically illustrated on a normal or
logarithmic scale as a function of filtration time plotted on a normal scale.
Field tests with pilot plants not only cover technical issues, they also $ave t$e follo(in% iportant
side effe"tsD
4 presentation of t$e treatent pro"ess to t$e f!t!re )enefi"iaries* Caymen may have little
knowledge about the technical aspects of a treatment plant, but may still have to construct their own
treatment plant through a self<help pro1ect. /otivation of community work could prove very difficult if
the villagers have no idea of the kind of structures they are supposed to build. . pilot plant is best suited
to introduce the prospective water treatment facility to the general public.
4 deonstration of t$e treatent effi"ien"& to t$e f!t!re "ons!ers* Iillagers are far more
motivated to contribute by cash and kind to the proposed treatment plant if they can see for themselves
how the water quality changes. They should also taste the treated water and look at its appearance.
>acteriological water quality improvements could be demonstrated to the public with colony counts on
membrane filters for raw and treated water. .ppreciation and acceptance of the treated water quality by
the consumers are important criteria for long<term use of a treatment plant.
4 trainin% of f!t!re treatent plant personnel* Involvement of local staff in the construction,
operation and monitoring of a pilot plant is an e,cellent training opportunity for future treatment plant
operators. 0,perience with pilot filter operation is directly transferable to full<scale operation.
It "an t$erefore )e "on"l!ded t$at pilot plant tests s$o!ld )e "ond!"ted (it$ ro!%$in% and slo(
sand filters in areas ($ere pra"ti"al e=perien"e is not &et availa)le* Field tests are re"oended
parti"!larl& (it$ ra( (ater e=$i)itin% a $i%$ s!spension sta)ilit&* Pilot plant tests s$o!ld also )e
"ond!"ted prior to t$e desi%n of lar%e4s"ale treatent plants to a"$ieve a ore "ost4effe"tive
la&o!t* Finall&+ field tests (ill deterine ra( (ater treata)ilit& and avoid fail!res (it$ respe"t to
ro!%$in% and slo( sand filter appli"ation*

'hoto ; Slow Sand "ilter -olumn
'hoto 9 Iertical *oughing "ilter -olumn
1ne"pected Iron )ine in 5ouhin ,ilters
1ne"pected Iron )ine in 5ouhin ,ilters
The town of Damono in 4est Africa# which num!ered $.#&&& inha!itants# had a piped water supply
system with a daily water capacity of $:& mG . This capacity was limited !y the pac*ae water
treatment plant which had to !e replaced due to serious corrosion and unavaila!ility of spares. The
reional water authority therefore started to study alternatives for the pac*ae plant. As practical
e"perience with rouhin and slow sand filters was now availa!le in the reion# such a treatment
'ption had to !e field/tested.
Daniel and Charles# two enthusiastic junior enineers were in chare of pilot plant desin. Two %m
lon hori@ontal/flow rouhin filter units were constructed and filled with ravel from a near!y
(uarry identified as potential filter material source for the planned $&&& mG6d treatment plant
,ield tests started soon after completion of the pilot plant. Tur!idity reduction !y the rouhin filters
was noticea!le !ut not sufficient. Tur!idity removal of the slow sand filter. was ood and its effluent
enerally amounted to less than % 0T1. +owever# the prefiltered water increased in tur!idity and the
colour of the wafer did not chare. Therefore# a la!oratory chemist was called to the site to ta*e water
samples for (uality analysis.
The chemical analysts did not disclose spectacularly new facts apart from a total iron content of &.&% /
&.$ m6l in the raw water and &.. / &.; m6l in the filtrate of the ravel filters Daniel and Charles
!lamed the chemist for his inaccurate wor* and accused him of havin mi"ed the samples. The
analysis was repeated !ut produced similar results. Perple"ed !y the situation# the youn enineers
started to investiate the filters. +ow can a filter increase the iron concentration H The red surface of
the ravel led them to the solutionB i.e.# the water flowin throuh the filter dissolved iron from the
lateritic material.
The rouhin filters performed well after the ravel had !een e"chaned !y filter media found
elsewhere. +owever# the e"perience with the <hidden iron mine< saved the project from additional
costs# as replacement of filter material in the $&&& mG6d would have !een (uite an e"pensive
'., Selection criteria for roughing filters
Ro!%$in% filters are priaril& !sed to separate fine solids from the water that are only partly or not
retained at all by stilling basins or sedimentation tanks. *oughing filters mainly act as physical filters
and reduce the solid mass. 8owever, the large filter surface area available for sedimentation and the
relatively small filtration rates also support adsorption as well as chemical and biological processes.
Therefore, besides solid matter separation, ro!%$in% filters also partl& iprove t$e )a"teriolo%i"al
(ater 8!alit& and, to a minor e,tent, change some other water quality parameters, such as colour or
amount of dissolved organic matter.
2in"e several different prefilter t&pes are availa)le+ such as intake and dynamic filters, downflow
and upflow roughing filters, and finally horizontal<flow roughing filters, choice of the most appropriate
pretreatment method becomes difficult. 2ele"tion of an ade8!ate treatent s"$ee s$o!ld )e )ased
on t$e follo(in% "riteriaD
raw water characteristics
type of surface water
topography at the intake and at the treatment plant site
economic aspects
operational aspects
The first two aspects focus mainly on raw water quality and are discussed in the ne,t chapter. The last
three criteria deal mainly with treatment plant layout and operation, and are therefore considered in a
second separate chapter.
'.,' Ra3 3ater <uality as selection criteria
T$e ra( (ater "$ara"teristi"s determine to a large e,tent the type of pretreatment process. Turbidity
and suspended solids fluctuation is thus the most important information required for the selection of the
pretreatment scheme.
Avera%e and a=i! levels of t!r)idit& and s!spended solids "on"entration are of great
importance for the design of pretreatment units. In addition, information on the period of turbidity and
suspended solids concentration peaks is essential. 5o such peaks last for a few hours, some days or a
few monthsP "urthermore, the solids should also be characterised according to their settleability and
size. 5o particles settle easily in the water or remain suspendedP "inally, some information on the
organic matter content would be useful.
Inforation+ espe"iall& on pea/ val!es of solids "on"entration+ is unfortunately often missing. The
sanitary engineer is thus forced to assess peak values or, in case of an available database, he will use the
probability paper analysis before designing a treatment scheme. Some simple sedimentation tests
carried out during periods of high and normal river discharge will allow the study of the settling
properties of the solid matter and characterisation of the suspension stability. Separation of coarse sand
and silt can certainly be achieved by a grit chamber, while finer settleable matter can be removed by
sedimentation tanks. *oughing filters will separate suspended solids, but may only partly remove
colloidal matter.
T$e fae"al poll!tion level should also be taken into account when designing pretreatment units as
they can improve, to some e,tent, the microbiological quality of the raw water. 8ence, they act as first
hygienic barriers and reduce the load of pathogenic microorganisms on the slow sand filters. The faecal
pollution level in a raw water source can be assessed )& )a"teriolo%i"al anal&sisQ i.e., determination of
faecal conform concentration. This type of analysis requires special equipment and e,pertise.
"urthermore, a random water test is only representative of the water quality at the time of sampling.
Several samples have to be taken at different times, especially in flowing surface water, to characterise
the faecal pollution level in surface water. A sanitar& s!rve& of t$e "at"$ent area may roughly
assess the possible magnitude of faecal water contamination. This requires inspection of the water
course and its contributors, investigation of wastewater and faecal disposal practices, assessment of
dilution during minimum flow, as well as study of human activities in the catchment area. The adapted
Table 2 proposed in J:K presents a rough surface water classification system with respect to hygienic
risk. It may be used as preliminary determination of the faecal pollution level in surface water.
Ta)le H 2pe"!lative 2!rfa"e Water Classifi"ation (it$ Respe"t to Healt$ Ris/
$ealt$ ris/ )& "ons!ption of
!ntreated (ater
E*"oli:fae"al "olifor
K"o!nt:-CC lL
possi)le sit!ation for s!rfa"e (ater
no risk 3 draining a well<protected catchment area
low 4 < 43 e,tensive farming in catchment area
intermediate 43 < 433 farming and scattered houses in
catchment area
high 433 < 4333 small settlements and water use upstream
of intake
very high risk S 4333 large settlements with wastewater
discharge upstream of intake
In )rief+ avera%e and a=i! t!r)idit& levels and s!spended solids "on"entration+ solids
settlin% properties+ as (ell as pea/ "on"entration period+ are t$e ost iportant ra( (ater
"$ara"teristi"s for sele"tion and desi%n of pretreatent !nits* Fae"al poll!tion level !st )e ta/en
into "onsideration ($en desi%nin% entire treatent s"$ees*
2!rfa"e (ater t&pe has also a strong impact on the characteristics and amount of solids carried by the
water. Small upland rivers, large lowland streams and still surface water generally differ from each
other as shown in "ig. 49 and described hereafter.
2all !pland rivers draining a catchment area, protected by a manifold and rich vegetation, will
probably have a clear or tinted water colour during periods of mean discharge. -oloured water can be
observed especially in slow flowing waters in contact with organic matter, such as in rivers flowing
through dense forests or swampy areas where water washes out humic substances from decaying
material and turns yellowish or brownish. The colour, which may be partly reduced by roughing filters
due to biological activities, will not affect slow sand filter operation which will further reduce colour.
The small upland rivers will react to heavy but short periods of rainfall with a sudden increase in run<off
and water quality change. T!r)idit& pea/s and:or in"rease in "olo!r are !s!all& "orrelated (it$
river dis"$ar%e* These peaks decrease with falling water level as soon as precipitation stops. In such
cases, either intake or dynamic filters may be used to reduce the e,treme peak values or to protect the
treatment plant from heavy solid loads discharged by the river for a few hours.
Lar%e lo(land streas have a different regime. Cocal showers do not greatly affect their discharge
or water quality. .nnual rainfall distribution, including the wet and the dry seasons, have a greater
influence on the discharge. C$an%es o""!r %rad!all& over a period of soe da&s or (ee/s ($en
in"reased t!r)idit& levels or s!spended solids "on"entration are re"orded for a fe( (ee/s or
ont$s* #ater quality fluctuations, e,pressed as ratio between average and ma,imum values, are
usually smaller than in upland rivers. Therefore, pretreatment is permanently required and use of
roughing filters, possibly combined with intake filters, is recommended. The choice of roughing filters
is dependent, among other aspects, on the level and period of high turbidity or suspended solids
concentration. .s a rule, moderate turbidities can be treated with vertical<flow roughing filters in series
and higher raw water turbidities possibly with horizontal<flow filters.
2till s!rfa"e (ater will probably e,hibit the smallest water quality changes. The influent of
reservoirs, ponds and lakes undergoes natural treatment processes. Suspended matter will settle and
microorganisms will die off with increasing retention time. @evertheless, s!spended and "olloidal
atter a& still reain s!spended and al%ae a& %ro( depending on the degree of eutrophication
and e,tent of solar radiation. To protect the subsequent slow sand filter units from e,cessive fine solids
and algal loads, use of finely graded roughing filters may be appropriate and necessary in such
In s$ort+ sall !pland rivers drainin% erosion4prote"ted "at"$ent areas are li/el& to )e of lo(
t!r)idit& ($i"$ a&+ $o(ever+ in"rease to pea/s of s$ort d!ration d!rin% periods of $eav&
rainfall* 2!"$ "onditions favo!r t$e !se of d&nai" and inta/e filters* Lar%e lo(land streas are
%enerall& ore t!r)id and "$an%e t$eir 8!alit& onl& %rad!all& and a""ordin% to t$e ann!al
seasonal pattern* Use of ro!%$in% filters+ possi)l& "o)ined (it$ inta/e filters+ a& offer an
appropriate option for t$e pretreatent of lo(land river (ater* 2!spended solids+ "olloidal
atter+ and al%ae of a sta%nant (ater so!r"e+ re8!ire as pretreatent et$od t$e appli"ation of
finer %raded ro!%$in% filters*
Fi%* G1 offers a atri= for t$e sele"tion of an ade8!ate s!rfa"e (ater treatent s"$ee* Type and
concentration of solid matter, as well as level of faecal pollution, are the decisive criteria for the
determination of the most appropriate separation process. Surface water greatly differs with respect to
these quality parameters. 8owever, information on these characteristics and values for a specific raw
water source is generally rather basic. 8ence, this matri, provides engineers with some guidelines for
designing water treatment schemes.
T$e desi%n of ro!%$in% and slo( sand filters is !s!all& "onservative+ i*e*+ it allo(s !n"ertainties
(it$ respe"t to ra( (ater 8!alit& and treata)ilit&* 8owever, additional tests will allow a more
accurate design of the treatment units. Settleability and suspension stability tests, characterisation of
solids according to type and size inorganic matter such as silt or clay particles+ organic material such as
plankton or algae! and their concentration in the raw water will assist in selecting the appropriate
filtration rate, gravel size and filter length. In addition, raw water may contain other pollutants, such as
true colour, dissolved organic matter, iron or manganese which need to be separated or reduced. "ield
tests are usually necessary to determine the removal rate of these substances by roughing and slow sand
'.,. =ayout an operational aspects as selection criteria
T$e topo%rap$& has to be taken into consideration in the design of water supply schemes. Cocation of
the intake, topographical conditions of the site and operational aspects are essential criteria for the
conceptional layout of water supply systems.
2!rfa"e (ater inta/es often have to be located in remote places to allow the strongly recommended
construction of gravity schemes. 8owever, access to the remote intake sites is often difficult, usually
time<consuming and regular control and cleaning of the installations not guaranteed and quite often
neglected. In such a case, pretreatment at the intake should be reduced to coarse solid removal, and the
actual water treatment should only be carried out at the treatment plant, generally located in front of the
reservoir and as close to the supply area as possible.
Lo"al topo%rap$& may favour construction of a small canal for controlled diversion of surface water
and, consequently, installation of an intake or dynamic filter. "avourable river bed conditions may also
allow construction of infiltration galleries.
Draina%e aspe"ts should be considered carefully. Intake filters usually operate with drained e,cess
raw water. 8ence, their use in a pumped raw water supply system according to standard design see "ig.
(9! is generally not the best option. Intake filters without raw water discharge should be used instead.
"urthermore, operation of roughing filters requires an adequate topography for washwater disposal.
8igh wastewater discharges must be possible without causing erosion. The installations should be large
enough for runoff to discharge into a receiving water course or, preferably, into a pond especially
constructed for washwater storage.
2!in% !p (e "an sa& t$at %ravit& s"$ees s$o!ld )e "onstr!"ted ($enever possi)le+ alt$o!%$
t$e& a& often re8!ire reote inta/es ($i"$ are diffi"!lt to aintain* Nevert$eless+ favo!ra)le
topo%rap$i"al "onditions s$o!ld )e !sed for t$e installation of inta/e and d&nai" filters to
red!"e (aste(ater dis"$ar%e pro)les at t$e treatent plant siteQ an aspe"t re8!irin% "aref!l
"onsideration (it$ ro!%$in% filters*
"ig. =( .lgorithm for the 5esign of Surface water Treatment Schemes Using *oughing and Slow Sand
Cheap Gravel
Cheap Gravel
,ontanero Tulio# careta*er of Taminano-s water supply scheme# located in a south/western reion of
the Andean rane# led our small e"pedition towards the water inta*e structure. Alon the steep path
windin throuh dry !arely cultivated or used rassland# we passed a few rundown haciendas
providin shade to some cattle. Althouh it was already late afternoon# the sun was still stron A
splendid view of the hilly hihland spread peacefully !elow us and rewarded our tedious ./hour wal*.
The narrow trail flattened after the vantae point and led into a narrow valley covered !y small trees
and !ushes# an indication for the presence of water. After reainin our !reath on the flat stretch I
en(uired when the local people had last visited the inta*e. <A!out two wee*s ao< was Tulio-s
answer# whereas the District Enineer declared to have inspected the site a month after the inta*e had
!een reconstructed within the frame of a field study project. Gentle splashin announced the near!y
water source and# after a final jump throuh small !ushes# we reached the !roo*let. 5eha!ilitation
wor* had !een ade(uately carried out and the inta*e filter placed at the river site as recommended.
4ater was flowin throuh the system and everythin seemed to !e wor*in well. The District
Enineer cut a !ranch off a tree to chec* the ravel level in the filter. +e slowly dipped the !ranch
into the tur!id water and lowered it further without ever touchin round. At first every!ody was
surprised and later em!arrassed. The inflow throuh the inlet pipes was instantly stopped with two
plastic sheets and the filter !o" was drained. The emptied tan* revealed that the ravel had !een
removed or illicitly stolen !y some farmer. 4e discovered horse dun at the site# a clear <finerprint<
of the ro!!ery. The District 'fficer pushed his cap !ac* onto his nec* and stared into the empty filter
!o". To refill this remote filter !o" with ravel would certainly re(uire considera!le transport efforts.
The plastic sheets were removed# the water slowly refilled into the inta*e structure !efore it started
flowin aain / untreated / throuh the lon supply line towards Taminano-s water reservoir.
0ot much was said durin our descent# however# I hoped to find the wheel!arrow recently used for
concrete wor* placed aainst the wall of a dilapidated cottae. Anyhow# this would have not reatly
chaned the situation. Tulio# who lives in the villae# is una!le to properly loo* after this remote
inta*e filter.
E"onoi" and operational aspe"ts also influence the selection of pretreatment schemes. -onstruction
costs are correlated with operating costs. *egular maintenance increases reliability of the water
treatment scheme.
Constr!"tion "osts a& possi)l& )e red!"ed (it$ t$e !lti)arrier "on"ept* .dequate
pretreatment units allow the design of subsequent treatment units operated at higher hydraulic rates or
requiring reduced filter lengths. Overall "onstr!"tion "osts of t$e treatent plant (ill t$!s )e
lo(ered* .part from taking advantage of the natural pretreatment potential of stagnant surface water or
optimally locating surface water intakes, use of grit chambers, sedimentation tanks, as well as intake
and dynamic filters should always be taken into account to allow reduced roughing filter sizes. 'art of
the capital costs might be used for operating costs if the treatment units are reduced in size, and filter
cleaning frequency increased. 8owever, since the additional costs for filter cleaning are generally not
substantial, use of intake filters as the first pretreatment stage is usually recommended.
2edientation tan/s s$o!ld )e !sed if lar%e vol!es of settlea)le atter are carried by the raw
water. The high suspended solids concentration in flowing surface water may be separated to a certain
e,tent by perikinetic self<induced! flocculation, thereby reducing the solid matter load on the
subsequent filter units. -onstruction of sedimentation tanks is recommended in such cases since tan/
"leanin% is easier t$an filter "leanin%* . sedimentation tank can easily be converted into a roughing
filter if the solids removal efficiency of the sedimentation tank is insufficient.
Re%!lar filter "leanin% is not only important to restore the treatment plantDs efficiency, but also to
enhance the caretakerDs responsibility and to keep him on the 1ob. Intake filters requiring weekly
cleaning may be a way to support this aspect. "inally, the (as$(ater deand+ which is dependent on
the type of prefilter used, increases as follows) dynamic filters, intake filters, vertical<flow and
horizontal<flow roughing filters. This could also be a decisive factor for the selection of the
pretreatment scheme, especially if water is scarce or has to be pumped.
In s$ort+ t$e overall "osts a& often )e red!"ed )& a so!nd treatent s"$ee desi%n !sin% a
se8!en"e of different treatent steps* 2edientation tan/s and inta/e filters are ade8!ate
pretreatent steps to red!"e $i%$ solids "on"entration of readil& settlea)le and filtera)le atter*
Ro!tine aintenan"e (or/ is essential for %ood treatent plant perforan"e* T$e (as$(ater
deand and disposal s$o!ld )e ta/en into "onsideration ($en desi%nin% a treatent plant*
T$e solid atter reoval effi"ien"& of sedientation tan/s+ inta/e and d&nai" filters is
ill!strated in Fi%* GG* These pretreatment processes can considerably improve the raw water quality or
protect the treatment plant from heavy sludge loads, providing that the nature and occurrence of the
solid matter allows for easy separation in the settling tank or accumulation on the filter bed.
Pretreatent )& t(in sta%e prefiltration is %enerall& favo!red with regard to treatment efficiency
and in terms of construction and operating costs. *eduction of the solid matter load on the roughing
filters by intake filters enables longer roughing filter runs and, consequently, washwater savings.
Conger filter operation may enhance biological processes and, thus, increase filter efficiency with
respect to for instance true colour and dissolved organic matter reduction. .n e,isting treatment plant
can be increased in capacity by subsequent installation of intake filters. The numerous advantages of
intake filters favour a larger use of this filter type.
"ig. == *ole of Sedimentation Tanks, Intake and 5ynamic "ilters in *aw #ater 'retreatment
5un Down Sedimentation Tan* Impairs 5ouhin ,ilter
5un Down Sedimentation Tan* Impairs 5ouhin ,ilter
The villae of Plum!on lies on the northern coastal plain of a lare Asian island. 'win to the saline
roundwater# surface water from the river Cimanu* is used as raw water source. Since the villae is
situated in the lowlands water has to !e pumped in three staesB i.e.# raw water is pumped to a
sedimentation tan* located ne"t to the river# the presettled water is then lifted to the treatment plant
comprisin two rouhin and one slow sand filter. unit# and the treated water is then conveyed to an
elevated tan* supplyin the =#&&& inha!itants of Plum!on.
The water level of the Cimanu* river increases !y more than . m in the rainy season. Durin this
period# the raw water contains very hih and fluctuatin loads of suspended solids# as well as tur!idity
pea* values of more than A#&&& 0T1. E"tensive erosion causes an important slude flu" on the river
teed# and it is estimated that the Cimanu* river creates every year $&#&&& mI of new land around its
estuary. The raw water inta*e is fi"ed and the inlet is always on the same level. This leads to a hih
load of settlea!le material in addition to the suspended solids always present throuhout the year.
The water is pumped into a sedimentation tan* which is also used for !alancin purposes !etween the
first and second pumpin stae. 'riinally# the tan* was . m deep# however# since it has no drainin
facilities# the slude accumulatin in the tan* has never !een removed. Therefore# the wafer depth in
the tan* currently amounts to only &.% m. ,urthermore# the compartment of the first and second tan*
is !y/passed !y a pipe leadin the raw water directly into the last section of this <sedimentation tan*<.
'win to the lac* of pretreatment# untreated raw water is pumped to the rouhin filters which are
una!le to meet this heavy silt load. The filters are cleaned hydraulically every three days# however#
they had to !e cleaned manually after filter operation of only seven months. This cum!ersome wor*
was repeated a second time# thereafter# the careta*ers refused to carry out this senseless jo!.
)eanwhile# different studies on filter improvement alternatives were carried out !y students at the
treatment plant. +owever# an appropriate desin# construction and operation of a new sedimentation
tan* is the *ey to reha!ilitate Plum!on-s water treatment scheme.
'7, Construction of roughing filters
As a atter of prin"iple+ lo"al aterial+ anpo(er and "o!nit& parti"ipation s$o!ld )e !sed+
($enever possi)le+ in t$e "onstr!"tion of an& (ater s!ppl& s"$ee in developin% "o!ntries* The
initial costs might be higher than for package or conventional water treatment systems, however, with
the construction of roughing and slow sand filters, most capital costs flow back into the local economy.
"urthermore, use of local material and manpower is also important with regard to maintenance, repair
and e,tension work, since these resources will remain available even after completion of the treatment
plant. The filter structures should be simple, sturdy and of good finish to guarantee their long<term use
and reduce future maintenance and repair costs. The layout should facilitate both operation and
Lo"al "liate and a var&in% relations$ip to tie (ill )e t$e deterinin% fa"tors for t$e
"onstr!"tion s"$ed!le* 'ro1ect implementation should be initiated long before site clearance and
e,cavation are started. Ideally, the pro1ect engineer should be called by the village after the
communityDs decision to improve the water quality supplied by the system in use or to construct a new
water supply. cuantity and quality of potential water sources will then have to be assessed. .t this
point, the engineer must be aware that a (ater treatent plant is !s!all& t$e ost "ople=
"oponent of a siple (ater s!ppl& s&ste* 6nce again we refer to the saying) Ethe best treatment is
no treatmentE. Use of a remote spring instead of nearby surface water should be taken into
consideration. If the community has to rely on surface water, the water treatment pro1ect will have to
start monitoring the raw water quality, especially during the rainy season. The subsequent dry season
will be used to design the treatment plant, estimate the construction costs and secure the financial
resources. 6rganisation of construction work and final preparations should be started during the
following rainy season and prior to the actual construction work, which will be initiated during the
following dry season+ i.e., two years after the communityDs request for support regarding improvement
of its water supply. 8owever, this pro1ect preparation largely depends on the way the community
discusses the pro1ect, decides upon it and organises its contribution in cash and kind. ProSe"t
preparation 8!ite often ta/es ore t$an t(o &ears+ a time span which is not lost if the pro1ect can be
realised on a strong communal basis. &ood site organisation and availability of building material will
have a positive influence on the progress of construction work, which can often only be carried out
during the dry season when community participation is most likely and climatic conditions favourable.
Therefore, large structures might require an additional two or more dry seasons.....< calculation of the
total time required by a pro1ect is left to the reader.
2o!nd desi%n+ %ood "onstr!"tion 8!alit& and o(ners$ip developent )& "o!nit& parti"ipation
are ne"essar& prere8!isites to a"$ieve a (ell4operated treatent plant* /uch depends on available
skills, on the quality of the material used, and on supervision during construction. . number of key
issues need to be considered during the planning and design phase, such as type of locally available
material, local construction technique, such as skills with masonry structures, access to the planned
treatment plant site, as well as topographic and soil conditions. .n e,perienced construction foreman
and regular site inspections by the design engineer will contribute to improve the quality of the
structures. 'articular attention should be paid to the mi,ing, compacting and curing of concrete, as this
largely determines sturdiness, water tightness, and durability of the structures. Since a water supply
should last for more than two generations, during which time the structures are likely to be e,posed to
adverse climatic conditions, %ood 8!alit& str!"t!res are a)sol!tel& ne"essar&*
'7,' Filter bo;
Filter str!"t!res "an %enerall& )e lo"ated )elo( or a)ove %ro!nd+ as ill!strated in Fi%* GH* The
respective choice depends on soil characteristics, available construction material and hydraulic profile.
6n a flat surface, gravity flow often requires the structures to be placed below ground. This, however,
might cause some problems or additional costs for adequate drainage of the washwater. . partially
buried filter will require less e,cavation work and provide support to the sidewalls by the back<filled
"ig. =2 Cocation and /aterials of *oughing "ilter >o,es
Ro!%$in% filters are !s!all& s$allo( str!"t!res of about 3.: m intake and dynamic filters! to 4 < 7 m
roughing filters!. The size of the filter bo, is dependent on hydraulic capacity see also Section F.=!
and filter length. The filter bo,es should not be too big ma,imum filter area for vertical<flow roughing
filters should amount to 72 (3 m], ma,imum cross section area for horizontal<flow roughing filters
about = < : m]!, to avoid problems with high washwater discharge rates. "urthermore, the filter bo,
should also not be too high preferably around 4 m! to allow easy removal of the filter material during
manual cleaning.
A tren"$ e="avated in ipervio!s soil+ s!"$ as "la&+ silt or lateriti" %ro!nd+ presents a lo(4"ost
sol!tion for a filter )o=* The trench has sloping sidewalls which do not e,ceed the slope stability of the
water<saturated soil slope less than 4)4!. Cining of the base and sidewalls prevents clean filter material
from mi,ing with the surrounding ground. . layer of sand, prefabricated slabs, in<situ applied coatings
concrete lining, ferro<cement, lime mortar! or in e,ceptional circumstances e.g. refugee camps!,
prefabricated plastic linings or the use of geote,tiles, are the most appropriate materials to use.
A (aterti%$t )o= $as to )e "onstr!"ted in perea)le %ro!nd or if t$e filter is installed a)ove
%ro!nd* In such cases, vertical sidewalls are recommended. >urnt clay bricks with cement mortar
lining, concrete bricks or reinforced concrete should be used for such filter bo,es. The foundation and
floors of the bo, need special attention to avoid cracks caused by uneven soil settlement. "inally,
watertight e,pansion 1oints will probably have to be made for long filter bo,es constructed for
horizontal<flow roughing filters. .lternatively, long filter bo,es resting on difficult ground can be split
into two or more separate units with fle,ible pipes interconnecting the compartments. U<shaped units
can also reduce the total length of filter bo,es. In such a layout, inlet and outlet are to be placed on the
same filter side, and the bo, divided into two equal parts by a longitudinal separation wall.
T$e filter )o= s$o!ld )e tested for (aterti%$t4ness+ preferably before it is filled with filter material.
Ceaks can be detected and repaired more easily in an empty structure. Special attention must be paid to
the 1oints at the floor<wall interface or the inlet and outlet bo,es fi,ed to the filter bo,. #atertight 1oints
require water stoppers made of 'I- or rubber. 6ther weak points in the structure include the pipe
sealings which possibly need additional reinforcement to prevent cracking of the walls, and seep rings
to prevent leaks.
'7,. Filter material
T$e filter aterial s$o!ld $ave a lar%e spe"ifi" s!rfa"e to enhance the sedimentation process taking
place in the roughing filter, and $i%$ porosit& to allow the accumulation of the separated solids.
&enerally speaking, an& inert+ "lean and insol!)le aterial eetin% t$e a)ove t(o "riteria "an )e
!sed as filter edi!* "iltration tests revealed that neither the roughness nor the shape or structure of
the filter material have a great influence on filter efficiency J43K. The following material could therefore
be used as filter ediaD
%ravel from a river bed or from the ground
)ro/en stones or ro"/s from a quarry
)ro/en )!rnt "la& )ri"/s
plasti" aterial either as chips or modules e.g. used for trickling filters! may be used if the material
is locally available. .ttention has to be paid to the uplift forces of the water
possi)l& )!rnt "$ar"oal+ although there is a risk of disintegration when cleaning the filter material, it
should only be considered in special cases e.g. for removal of dissolved organic matter!
possi)l& "o"on!t fi)re+ however, due to the risk of flavouring the water during long filter operation,
it should be used with care.
Every Day and Every 2iloram Count
Every Day and Every 2iloram Count
I was a!out to o for lunch when the phone in my office ran. Peter# a former wor*in colleaue# was
at the other end. 4ith the same enthusiasm as in critical situations# I was informed that his office was
contracted !y the Disaster 5elief 1nit to desin and construct a water supply system for a resettlement
camp in East Africa. The camp should host .&#&&& refuees and the infrastructure should consist of
simple shelters# a feedin centre# a small hospital# water and sanitation facilities to !e set up as soon
as possi!le. Peter needed technical advice on the desin of the water treatment plant which had to
meet special construction criteriaB i.e.# installation time and weiht of !uildin material had to !e *ept
to a minimum.
After the call# I had my lunch in the train headin towards the capital where I met Peter in his office
early afternoon to discuss the water supply project. The raw water had to !e pumped from a lare
irriation canal whose water (uality was un*nown !ut rather tur!id loo*in. Cement is rather
e"pensive and often scarce in that country. 4e sat !etween drawins# reports and computers and
loo*ed at each other waitin for a stro*e of enius. )eanwhile# the water of a small !iotope in the
!ac* arden was reflectin the sunliht of the warm summer afternoon. Peter suddenly ot
enthusiastic aain and proposed to use the <!iotope construction techni(ue<. The rest of the desin
was completed within a short time.
Two wee*s later# Toni volunteer of the Disaster 5elief 1nit and construction foreman# was sittin in a
caro aircraft headin for Africa. The aircraft also carried a num!er of lare plastic sheets and
different plastic fittins which were unavaila!le in the country of destination. Soon after his arrival#
Toni started construction of the treatment plant with a!out $&& casual la!ourers# and within si" wee*s
the pumpin station# two sedimentation tan*s# four hori@ontal/flow rouhin filters and the treated
water reservoir were set up. The tan*s and filters were desined as earth !asins with inclined walls
and earth dams made of !as filled with the e"cavated soil. The !asins were then coated with the
prefa!ricated plastic linins. Perforated pipes were laid in the rouhin filters to allow hydraulic filter
cleanin. Trial operation of the treatment plant revealed a satisfactory efficiencyB i.e.# the raw water
tur!idity of $#&&& .#&&& 0T1 was reduced to half of its initial value !y the sedimentation tan*s# and
tur!idity in the effluent of the rouhin filters was recorded at % / .& 0T1.
The infrastructure of the refuee camp was then handed over to the local Commissioner for 5efuees.
+owever# political disputes a!out implementation and use of the camp started amon the countries
involved soon thereafter. ,inally# %#&&& refuees settled almost two years after completion of the camp
and reminded Toni of an e"perience he had durin construction. A local foreman had told himC <Jou
have a watch !ut I have the time .... <.
?ravel is t$e "oonl& !sed filter aterial but it was replaced by broken burnt bricks in the
horizontal<flow roughing filters constructed by the >lue @ile 8ealth 'ro1ect in the Sudan J22K, by palm
fibre called Ei1ukE in a roughing filter pro1ect in Indonesia J=2K, and by plastic material in laboratory
tests at the University of @ewcastle in 0ngland J=2K. Table : shows that filter efficiency of bricks and
plastic as filter material is similar to gravel with respect to turbidity reduction. The filter filled with
palm fibre has a better respective performance compared to the gravel filter. &reater porosity F7B
versus (9B!, responsible for the reduction of the effective flow velocity, is certainly an e,planation for
this observation. 8owever, since use of the palm fibre caused a considerable drop in the dissolved
o,ygen concentration, odour and taste problems could occur. Therefore, more detailed investigations
are always necessary before palm fibre or any other alternative material is used on a long<term basis.
Ro!%$in% filters are !s!all& "oposed of t$ree filter fra"tions ran%in% in si3e fro "oarse to fine*
The coarse and a large amount of the finer suspended solids are removed by the first filter pack. Since a
large pore volume is required in this part of the filter, a coarse filter material is best installed over a
considerable filter length. The subsequent filter material is finer and the packs installed over a shorter
filter length. The last filter fraction of limited length should assume only a polishing function as it is
supposed to remove the last traces of the finest suspended solids found in the water.
Ta)le I Relative Filter Perforan"e (it$ Different Filter 7aterial parallel tests!
T!r)idit& red!"tion ,si3e of filter aterial.
ProSe"t ?ravel alternative filter aterial
burnt bricks, >@8'$Sudan J22K ;9 B 73<(3, 42<73 and 2<43 mm! 99 B bricks (3<23, 42<73,2<43
palm fibreO 'lumbon$Indonesia
(F B 4:<72 mm! :9 B fibre!
plastic material, University of
@ewcastle J=2K
F7 B broken bricks, (3<23 mm+
gravel 4=<4; and 2<F mm!
F= B rings " (; mm pipes " (3
mm caps width 2 mm!
Oonly filled in first filter compartment
2in"e filter effi"ien"& in"reases (it$ de"reasin% filter aterial+ one is tempted to use the smallest
possible filter material or even to omit the larger filter material and to install only one the finest < filter
medium. 8owever, the roughing filter technology requires coarse filter material as denoted by its name.
The finest filter material should not be smaller than about = mm to ease hydraulic filter cleaning. "ilter
material which is too coarse, however, has a smaller filter efficiency and would therefore require a
longer filter length to achieve the same turbidity reduction. .s already illustrated by "ig. (4, !se of at
least t(o or %enerall& t$ree different filter aterial si3es (ill res!lt in an e"onoi" filter desi%n
and ade8!ate filter operation*
Ta)le J ?!idelines on 2i3e and Len%t$ of Filter 7aterial for Different T&pes of Ra( Water
filter len%t$ of 1 fra"tions
type of solid
rate I

gravel sizes of (
upflow roughing
filter in layers
roughing series
horizontal flow
roughing) filter
4: < 7= mm =3 < ;3 cm
3.: < 4 m$h 47 < 4; mm 73 < (3 cm
; < 47 mm 73 < (3 cm
47 < 4; mm =3 < :3 cm :3 < 473 cm 733 < =33 cm
= < 3.; m$h ; < 47 mm (3 < =3 cm :3 < 473 cm 433 < (33 cm
= < ; mm (3 < =3 cm :3 < 473 cm 23 < 423 cm
plankton, algae ; < 47 mm (3 < 23 cm
3.( < 3.2 m$h = < ; mm (3 < 23 cm
7 < = mm (3 < 23 cm
Ta)le J lists soe %eneral %!idelines on si3e and len%t$ of different filter fra"tions* These
guidelines should not be followed too rigidly. 'ractical aspects, such as e.g. availability of specifically
graded material from a quarry, are more important. If adequately graded filter material is not available,
gravel at the construction site can also be sieved through wire meshes or perforated steel plates used as
5uring the initial development phase of the horizontal<flow roughing filter technology, the .sian
Institute of Technology .IT! in >angkok recommended the installation of si, to eight small gravel
layers J2;K. &ravel size should subsequently be reduced from 73 to 7.2 mm and thereafter increased
again to 72 mm. There is no reason why the smallest gravel fraction should be located in the centre of
the filter bed since the following gravel packs have, by nature, a smaller removal efficiency.
T$e filter edia !sed for ro!%$in% filters $as to )e "lean and free fro or%ani" aterial* It is
therefore important to wash the aggregates thoroughly in order to remove all loose and dirty material
from the surface of the filter media. If this recommendation is not followed, the effluent quality of the
roughing filter will be poor and result in rapid clogging of the slow sand filter.
T$e different filter fra"tions s$o!ld )e separated fro ea"$ ot$er to avoid i=in% of t$e
a%%re%ates d!rin% an!al filter "leanin%* The filter material of an upflow roughing filter in layers is
preferably separated by a plastic wire mesh. The different filter media of roughing filters in series are
separated by the different filter bo,es, and perforated separation walls are used for that purpose in
horizontal<flow roughing filters. >urnt brick or cement block walls with open vertical 1oints are best
suited for separating the filter fractions. The total area of the open 1oints should ideally amount to 73 to
(3B of the total filter cross section and be equally distributed over the entire cross section to maintain
an even flow throughout the horizontal<flow roughing filter. 'refabricated perforated bricks or blocks
e.g. holes " ( cm, spacing 2 , 2 cm! or loose rubble could be installed instead of open 1oints.
Simultaneous filling of filter material in layers must be observed with loose or weak separation wall
'7,7 5nlet an outlet structures
Inlet and o!tlet str!"t!res are ne"essar& to re%!late t$e flo(+ evenl& distri)!te and a)stra"t t$e
(ater+ and to "ontrol t$e (ater level in t$e filter* The inlet and, preferably, also the outlet are
equipped with I<notch weirs for flow control if weir overfalls of about 73 < (3 cm can be
accommodated in the hydraulic profile of the treatment plant. The I<notch weir at the outlet can be
omitted in treatment plants with a small available hydraulic head and replaced by an effluent pipe which
will maintain the water level in the filter at a minimum height. A flo( rate "ontrol allo(in% a""!rate
flo( adS!stents s$o!ld al(a&s )e installed at t$e inlet of ro!%$in% filters* . flow rate control at
the filter outlet is not recommended as backwater effects create flow ad1ustment difficulties. .n inlet
weir can control the headloss and the water level increase in the inlet bo, located after the weir also
indicates filter resistance development.
Even flo( distri)!tion t$ro!%$ t$e filter )ed is a"$ieved eit$er )& a false filter )otto or )&
perforated pipes in !pflo( ro!%$in% filters+ and )& an inlet "$a)er (it$ a perforated separation
(all in $ori3ontal4flo( ro!%$in% filters* "ig. =: illustrates the inlet and outlet structures of a
horizontal<flow roughing filter. The middle part of the separation wall ne,t to the inlet chamber should
be perforated. . full wall at the bottom and at the top respectively, prevents coarse settled solids or
floating matter from penetrating into the filter. The minimum width of the inlet chamber should be at
least :3 to ;3 cm for easy removal of settled matter. . similar outlet chamber is installed in horizontal<
flow roughing filters at the effluent side. 8owever, the openings in the separation wall after the last
filter pack are distributed throughout the filter cross section. It is iportant to prote"t t$e pretreated
(ater fro al%al %ro(t$ or fro air)orne poll!tion* "or that purpose and to avoid mosquito
breeding, horizontal<flow roughing filters have to be filled to about (3 to =3 cm above the effluent weir
level with filter material, and the outlet chamber must be covered. In upflow roughing filters, the
effluent pipe located roughly 2 cm above the top of the filter bed, controls the water level. This free
filter water surface should also be protected by a layer of large stones about " 23 < ;3 mm, height 73 <
(3 cm! or by a removable roof cover.
"ig. =: Inlet and 6utlet Structure of a 8orizontal<flow *oughing "ilter
'7,2 Drainage system
Ro!%$in% filters "an )e "leaned eit$er $&dra!li"all& or an!all&* 5rainage systems with a high
hydraulic capacity and capable of abstracting the wash water evenly from the filter bed are necessary
for hydraulic filter cleaning. "ig. =9 contains different drainage layouts. T$e installation of a false
filter )otto is t$e )est option for verti"al4flo( ro!%$in% filters* .bout 43<cm high concrete blocks
support the perforated filter bottom made of roughly 2<cm thick concrete slabs and perforation holes of
about : < ; mm diameter. These slabs are usually installed with open 1oints of about = < ; mm clear
width. Perforated draina%e pipes or perforated "!lverts $ave to )e !sed in $ori3ontal4flo(
ro!%$in% filters and could be a possible alternative to a false filter bottom in vertical<flow roughing
filters. 8owever, a false filter bottom cannot be used in horizontal<flow roughing filters as it would lead
to water short circuits. Therefore, perforated drainage pipes and culverts will also have to be installed
every 4 to 7 m perpendicular to normal flow direction. Since intake and dynamic filters are surface
filters, the sludge which mainly accumulates on top of the filter bed is cleaned manually. Therefore,
these filters do not require drainage installations with a high hydraulic capacity.
Pipes and s$!toff devi"es are re8!ired for $&dra!li" filter "leanin% and for "oplete de(aterin%
of t$e filter )o=* Carge pipe diameters of 423 to 723 mm are necessary for efficient hydraulic cleaning.
The hydraulic capacity of these installations should allow an initial filter drainage velocity of =2 < F3
m$in. The outlet of drainage pipes should be located at the lowest possible level in order to make
optimal use of the available hydraulic head. "or cost reasons, these large diameter drainage pipes should
be as short as possible and firmly fi,ed to the structures in order to withstand the considerable dynamic
pressures generated by the flushing cycles. . manhole, as shown in "ig. =;, can be used as
interconnection between filter bottom and drainage pipe to alternatively reduce the length of the
hydraulic drainage pipes. In contrast to these large pipes, small tubes of 4 < 7 inches in diameter sealed
by nipples will adequately dewater the different compartments inlet, filter and outlet bo,es! of a
roughing filter. Small structures, however, can also be dewatered with the help of buckets or a tube used
as a siphon.
"ig. =9 Cayout of 5ifferent 5rainage Systems for *oughing "ilters
Fast openin% devi"es are re8!ired to initiate a fast $&dra!li" "leanin% "&"le in order not to lose too
much washwater during cleaning. T$ese devi"es s$o!ld )e siple in desi%n+ st!rd& and eas& to
operate* In the long run, they must be watertight and equipped with a shut<off device to interrupt the
drainage process. Use of butterfly valves is the best but most e,pensive option. To reduce construction
costs, different local designs of fast opening devices have been developed as illustrated in "ig. (= on
page Id<F. . good e,ample of an appropriate technology is the modified milk can cover developed by
-I@.*. in -olombia. [8.S in -hina successfully uses a self<designed plug valve installed in a steel
bo,, and 8elvetas in -ameroon applies carefully shaped and firmly installed plugs held by a removable
"ig. =; Cevelling of the 5rainage System
Finall&+ str!"t!res for fast draina%e and safe (as$(ater disposal !st )e provided* The
washwater is generally discharged into an open canal used to convey the hydraulic flush to a nearby
surface water or to a small pond used for intermediate storage. -onstruction of a small lagoon is
recommended to recover the solid matter washed out of the filter for use in agriculture. 5irect discharge
of the washwater in stagnant surface water may gradually silt up the reservoir and adversely affect its
water quality.
'7,4 0ravel an san 3ashing facilities
Filter aterial needs to )e "aref!ll& (as$ed )efore it is pla"ed into t$e ro!%$in% and slo( sand
filters* The gravel and sand must be free from organic material, silt and clay particles, as these
impurities cause serious operational problems. The organic matter could decompose and affect taste and
odour, and the particles slowly washed out of the filters could thus increase turbidity of the treated
water. It must be noted that provision and cleaning of filter material by the community require
considerable effort and time.
Filter aterial (as$in% is )est a"$ieved )& e"$ani"al stirrin% of t$e a%%re%ates in a (as$(ater
)asin+ as mechanical friction rubs the impurities off the aggregatesD surface. The washing site should be
located in the centre of the treatment plant, preferably ne,t to the slow sand filters which are cleaned
manually and more frequently to reduce the transport distance of the filter material. The site should also
not be too small and allow = < : men to wash the filter medium simultaneously. #ashwater can be
saved and gross impurities eliminated if small filter material loads are stirred and removed with a shovel
to a first tank before they are transferred to a second tank for final washing. Such a washing installation
is illustrated in "ig. =F. 8owever, centralised cleaning involves transport of the filter material. Use of
the open drainage channel located along roughing filters is an alternative to the washing site as it
requires less gravel movement.
Ra( (ater of oderate t!r)idit& "an )e !sed as (as$(ater+ and prefiltered water for final sand
cleaning. . fle,ible hose may also be used to wash the filter material, and should thus also be available
if raw water is supplied by gravity to the treatment plant. The washing site may also be equipped with a
shower for the treatment plant operator if the washwater is adequately disposed.
"ig. =F Cayout of &ravel and Sand #ashing Installation
A Precious Plastic Sheet
A Precious Plastic Sheet
This was already my fourth visit to >a Kaveriana treatment plant which forms part of the water supply
system of a private education centre in a rather lare and well/*nown city in >atin America. I have
always loo*ed forward to meetin Alci!iades# the dedicated careta*er of the treatment plant. I enjoyed
our interestin discussions and Alci!iades would (uite often relate some of his new practical
e"perience he ained since my last visit# or we discussed some pendin operational pro!lems for
which we found appropriate solutions. 4e learned a lot from each other# and our wor*in
relationship rew into a personal friendship.
Since my visit last year# new houses had !een constructed in the catchment area. This further
enhanced soil erosion in the rainy season# and uncontrolled wastewater dischares into the rivulet
deteriorated the raw water (uality especially in the dry season. Alci!iades had to install a small rit
cham!er in front of the inta*e filter to reduce the silt load on the filters. +e also tried to reduce the
increasin !acterioloical pollution !y the installation of a sand layer in the rouhin filter. The first
alteration !rouht the e"pected amelioration# however# he faced considera!le pro!lems with the
second modification. The sand layer# which developed a sinificant filter resistance in the rouhin
filter# resulted in filter cloin. Since hydraulic cleanin was no loner possi!le# the rouhin filter
had to !e cleaned manually.
>a Kaveriana-s treatment plant is located on a slihtly slopin meadow in a pictures(ue environment /
reason enouh to ma*e some photos for my project documentation. I had intended to document the
recent chanes at the inta*e filter with some snapshots. 4hat I discovered was a carelessly disposed of
plastic sheet near the inta*e filter# and I was surprised !y this litter that did not at all correspond to
Alci!iades- wor*in manner. +e normally *ept the treatment plant and its surroundin in very clean
condition. I thouht of several reasons for his nelienceB i.e.# his salary may no loner meet the
current standard of livin# or he may have personal pro!lems with the administration vat home.
I was a!out to ta*e away the unpleasant plastic sheet which would had spoiled my photo when
Alci!iades arrived on the scene. I told him my pro!lem and as*ed him to remove the plastic. +owever#
Alci!iades was stronly opposed to my suestion since this piece of plastic was used to clean the
entire filter !o" durin the !iannual removal of the ravel from the inta*e filter. The plastic sheet was
used as temporary sla! in order not to soil the cleaned ravel. Ashamed of my inorance I too* a
picture / alon with the plastic sheet riht in the middle of the photoL
'2, !peration an maintenance of roughing filters
No($ere in t$e (orld (ill a (ater treatent r!n )& itself* .n input of manpower will always be
necessary. *ight from the beginning of pro1ect implementation, operational and maintenance aspects
require careful elaboration and approval of the different parties involved, such as the /inistry of #ater,
support agencies, and the community. . village water committee is usually formed to manage a water
supply scheme. 8owever, t$e "areta/er pla&s a /e& role in t$e operation and aintenan"e of a
(ater s!ppl&* 8e and his colleagues are normally elected by the water committee which also clearly
defines their duties and working conditions, such as salary, fringe benefits, etc. The ma1or tasks of a
caretaker at the treatment plant is to control the water flow, monitor the quality, clean the filters and
carry out general maintenance work. These activities will be described in the following sections.
7ain responsi)ilit& for operation and aintenan"e of a (ater s!ppl& s"$ee !st )e %iven to t$e
"o!nit& "on"erned+ since reliability of a water supply primarily affects its inhabitants. In other
words, the water supply system has to be operated and maintained at village level as much as possible,
using locally available resources and infrastructure. *oughing and slow sand filters meet these criteria
as they do not require chemicals, mechanical spare parts or highly trained staff. 6wnership and self<
management of a water supply by a committed community prevents pro1ect failures and waste of public
'2,' Careta6er training
Copre$ensive trainin% of lo"al staff is essential since technical installations must be run by
operators. #hile most technical problems can be eliminated or reduced by appropriate design and
construction, human aspects which might affect treatment plant operation are more difficult to control.
-areful, formal training of caretakers and e,tensive support, guidance, and supervision in the first years
of operation are necessary.
Ade8!ate pa&ent of t$e lo"al staff is e8!all& iportant* The caretaker of a rural water supply
usually has numerous duties+ he is often in charge of treatment plant operation, maintenance of the
water distribution system, collection of water ta,es, clearing irregularities and complaints, etc.
/otivation alone will not keep him on the 1ob+ his or her work must be adequately compensated.
T$e "areta/er s$o!ld )e re"r!ited fro t$e lo"al villa%e+ he should preferably be settled, married,
land owner and well<recognised by the community. Ideally, he should also have some technical skills,
however, interest and motivation to run a water supply system are more important criteria for his
election. 8e should show his interest and skills during construction of the treatment plant. 8owever, the
future caretaker should preferably not be selected during the construction phase of a pro1ect but after
assessment of the performance of different candidates during this period by the construction foreman
and the design engineer who could then propose qualified candidates to the water committee for
election. Ea"$ "areta/er !st also $ave a dep!t& who can assume operation of the water supply
during his absence for reasons of illness, personal matters or other obligations. 8owever,
responsibilities of the different staff must be clearly defined and separately developed for each
Foral "areta/er trainin% is )est "arried o!t )& t$e e=e"!tin% a%en"& and "ond!"ted in t$e lo"al
lan%!a%e* .nne, 9 outlines a possible training programme. "ormal training will be complemented by
on<site instructions given by the design engineer. 0ach treatment plant is different and thus requires the
development of individual operation schedules that will have to be elaborated 1ointly with the
supervisor in the first year of operation and reviewed regularly on the basis of the treatment plant
performance. Careta/er trainin% is a "ontin!o!s pro"ess* *egular meetings with other caretakers is
an ideal platform to e,change e,perience, enhance reputation and value the importance of the
caretakerDs inconspicuous daily activities.
'2,. Treatment plant commissioning
Filter operation s$o!ld onl& start ($en "onstr!"tion (or/ $as )een properl& "opleted* "or
instance, performance of a horizontal<flow roughing filter only partially filled with gravel will be poor.
The water bypasses the impounded gravel layers in such a way that the unit will not act as a filter but as
an inadequate sedimentation tank. 0mphasis must therefore be placed on a good finish of the
construction work, including the installation of proper flow control and drainage facilities, as well as the
full supply of filter material. 6nce the treatment plant is in operation, provision and installation of
missing filter material have repeatedly proved impossible as the construction partners might refuse to
assume additional work. Subsequent filling of the remaining filter medium could also impair plant
Cleanin% of t$e installed filter aterial )& t$e draina%e s&ste is re"oended before starting
filter operation. The roughing filters should be filled with water up to the effluent weir level at low flow
rates of 3.2<4 m$in. Thereafter, the water is drained off by the first drainage installation situated ne,t to
the inlet. .ny dust on the surface of the filter material is rinsed to the filter bottom. Impurities
accumulated near the drainage system will be flushed out of the filter If necessary, this procedure will
have to be repeated two or three times changing the point of drainage from filter inlet to filter outlet.
Such filter cleaning will prevent dust particles from settling on the fine gravel fraction and increasing
the initial filter resistance. 6perational control of the complete drainage system is a positive side effect
of the described cleaning procedure.
ProSe"t $and over is often "o)ined (it$ t$e ina!%!ration of t$e installations* . supply of clean
and sufficient water on this special day should be guaranteed. #ater treatment operation has to start
about two to three months prior to the official inauguration day to ensure a sufficient supply of good
quality water and to avoid disappointing the community and the invited guests. The treatment plant may
at first not produce the e,pected water quality, as the biological processes, known as filter at!ration+
(ill re8!ire soe tie to develop ,a fe( (ee/s or ont$s. depending on the raw water
characteristics. The treatment plant may be operated at reduced capacity during this period.
'retreatment may be bypassed to accelerate slow sand filter maturation. 8owever, this procedure should
be applied only to slightly turbid raw water carrying dissolved organic matter which will not be reduced
by the prefilters but used for the development of the ESchmutzdeckeE in the slow sand filter.
'2,7 Flo3 control
A 0G4$o!r "ontin!o!s filter operation a/es a=i! !se of t$e installations* -ontinuous and
constant flow conditions usually improve treatment plant performance and reduce the required structure
size. 8owever, gravity flow is usually necessary for such ideal situations. -ontinuous flow might not be
possible in water supply schemes where the raw water has to be pumped. #hen pumping is required,
the treatment plant might be staffed for ; or 4: hours a day, depending on whether one or two shifts are
Interittent slo( sand filter operation is not re"oended for 8!alit& reasons* In order not to
affect the biological activities in the slow sand filter, this filter can be operated at a declining filtration
rate in pumped systems during the unstaffed period of the day. In practice, the stock of supernatant
water is drained through the filter at a continuously declining flow rate during the night and in the
morning hours, the filter is refilled with pretreated water to reassume normal operation. Such an
operation calls for special provisions as pretreated water for the slow sand filter has to be provided at
higher pumping rates.
Ro!%$in% filters are ainl& p$&si"al filters* T$e& are less affe"ted )& flo( interr!ptions as they do
not depend on a continuous supply of nutrients as biological filters. 8ence, intermittent operation can be
applied without causing a significant deterioration of the prefiltrate, provided smooth restarting of filter
operation is observed. 5eclining filter operation rate of roughing filters to supply slow sand filters with
a constant flow is not advisable due to the relatively small water volume stored in the prefilters. T$e
ost favo!ra)le option in a p!ped s"$ee is t$e provision of a ra( (ater )alan"in% tan/ which
allows continuous filter operation. *emoval of the coarse solids is a positive side effect of such a tank.
The different tank volumes required for a 433 mb$d plant are illustrated in "ig. 23.
A Goat and a 3a of Sweet Potatoes
0ond@en was the fourth water supply system I visited on that day and 0d@enshwai was the last
scheme !efore reachin the hotel in 2um!o# a small district town in 4est Africa. Tired from the heat
and dust# I really loo*ed forward to the evenin shower and pro!a!ly also the entire crew in the four/
wheel drive project car. As we entered 0ond@en at rather hih speed# the project coordinator
suested to see the careta*er of the treatment plant at once in order not to lose too much fume.
0either )r 3oniface# the local technician# nor I had any o!jections to his suestion. As we crossed
the junction ne"t to the church we saw a lare crowd of people. 4e assumed that a weddin was !ein
held and drove on to the careta*er-s house. +owever# the chairman of the water committee ran after
our car and tried desperately to attract our attention. This is when our schedule ot mi"ed up.
The crowd in front of the church was not cele!ratin a weddin !ut had !een waitin for us for three
hours. 4e reversed the car and drove it to the church# where a local roup of drummers started
poundin on their instruments as we were directed to chairs in front of the community hall. The
welcomin address was well/prepared# typed on a piece of paper !earin three official stamps# and
read !y )r Patric*# the chairman of the water committee. Two dancin roups performed traditional
dances after his speech. Thereafter# a colourful cap was placed on my head and I was nominated
chief of the villae and loudly applauded !y the entire community. 4hile another dancin roup was
startin its performance# a traditional ift commonly donated to chiefs was conferred on me / a oatL
+eaded !y )rs Dominica# chairperson of the women-s roup# a co@en women started to dance and
deposit a !a filled with sweet potatoes in front of my feet. 'verwhelmed !y so much of honour# I
addressed a word of than*s to the villaers and stressed the importance of maintenance in a water
supply scheme.
)r 3oniface# the local technician# was sittin (uietly ne"t to me durin the entire ceremony. 4e were
then served local dishes and !rews in the community hall and more toasts were e"chaned. It was late
afternoon as we clim!ed into our project car# the oat attached onto the roof rac* and the !a of
sweet potatoes loaded in the rear. After cheerfully wavin ood!ye to the villaers# we left 0ond@en.
+owever# the new chief who was sittin ne"t to )r 3oniface felt em!arrassed !ecause he had won#
durin the two/hour cele!ration# all the praise for the efforts made !y the local technician to mo!ilise
the community# to oranise construction material and to supervise this self/help project / a wor*
which had *ept him !usy for the last two years. 4e loo*ed at each other without e"chanin a word
!ut the messae was clear.
The oat and the !a of sweet potatoes were unloaded in front of his home !efore we drove on to our
hotel under a shinin moon.
T$e flo( t$ro!%$ ro!%$in% filters is "ontrolled )& a flo( "ontrol devi"e at t$e inlet and )& a fi=ed
(eir at t$e o!tlet as illustrated and recommended in "ig. (7 on page Id<9. This is also true for the
recommended inlet controlled slow sand filters which compensate the progressive headloss
development by a gradual increase of the supernatant water level. "or gravity schemes, constant feeding
is maintained by a more or less fi,ed position of the valve in the supply pipe and a subsequent overflow
in the distributor bo,. "or pumped schemes with a raw water tank, the flow to the treatment plant is
regulated by a mechanical flow rate device as shown in "ig. ((. These two main possibilities are
illustrated in "ig. 74 on page Il<F.
E4not"$ (eirs are %enerall& !sed for dis"$ar%e eas!reents* 'ermanently installed I<notch weirs
or transportable equipment used for flow control are described in .nne, 7. The flow rate through each
filter should be routinely controlled once a day if I<notch weirs are installed, and with transportable
equipment at least twice a week according to the monitoring programme outlined in .nne, ;.
Filter resistan"e in ro!%$in% filters is inial and $ardl& red!"es t$e flo( t$ro!%$ t$e filter* The
headloss increases to a few centimetres in well<operated roughing filters. "ilter resistance builds up
along the entire filter bed as roughing filters act as space filters. It is reduced to its initial value by
efficient and regular filter flushing. 8owever, inta/e and d&nai" filters "an )!ild !p "onsidera)le
filter resistan"e as they mainly act as surface filters. Since the headloss in intake filters can increase to
73 < (3 cm within a week J=;K, the flow through the filter has to be ad1usted by gradual opening of the
valve located in the effluent pipe. 5ynamic filters have to produce by definition a high filter resistance
within a short time during periods of high raw water turbidity. This will clog the filter bed and prevent
highly turbid raw water from flowing to the subsequent filters. Intake filters are usually cleaned once a
week, and dynamic filters after every high turbidity peak.
Filter resistan"e "an easil& )e deterined )& eas!rin% t$e level of t$e free (ater ta)les in the
inlet and outlet chamber of roughing filters. The effluentDs weir crest level can be used as reference 3<
level!. &auging rods fi,ed to the walls of these two chambers will facilitate the respective
'2,2 Water <uality control
. water quality monitoring programme usually aims at)
characterising the raw water quality
establishing and monitoring treatment plant performance
developing operational criteria for the roughing filters and slow sand filters i.e., schedule for filter
optimising layout and operation of the filters i.e., e,change of filter material, increase or reduce
filtration rate!.
T$e ost iportant 8!alit& "riteria for drin/in% (ater is its )a"teriolo%i"al 8!alit&* 8owever,
improvement of the bacteriological water quality greatly depends on raw water turbidity, efficiency of
the pretreatment units in reducing this turbidity, and on adequate slow sand filter operation. Turbidity
and bacteriological contamination of the water are, therefore, the key parameters for rural surface water
characterisation. Turbidity measurements play a ma1or role in monitoring the pretreatment steps!, slow
sand filter efficiency is usually established by bacteriological tests.
8owever, )a"teriolo%i"al (ater 8!alit& e=aination re8!ires spe"ial e8!ipent+ specific field test
kits or generally the infrastructure of a laboratory. #ell<trained and e,perienced staff are essential for a
reliable analysis. *outinely performed bacteriological water quality tests of rural water supply schemes
are generally far beyond the capacity of the responsible institution and, therefore, mostly restricted to
random tests. . well<operated slow sand filter is a stable and reliable water treatment unit not requiring
frequent bacteriological tests. In practice, test frequency can be reduced to a minimum once the
bacteriological efficiency of the slow sand filter is established. Headloss developent and len%t$ of
filter r!n are appropriate "riteria to assess t$e )a"teriolo%i"al effi"ien"& of slo( sand filters* #ell<
operated slow sand filters use natural treatment processes, and nature will rebel with a headloss increase
when filters are overloaded, in contrast to chemical and mechanised treatment processes, where
chemical doses and water pressure can often be increased at the cost of the water quality produced. The
caretaker may thus greatly influence and control the quality of the treated water by adequate slow sand
filter operation and by monitoring the headloss development.
"ig. 23 *equired Tank Iolume of a 433mb$d Treatment 'lant in -orrelation to the "low 'attern
T!r)idit& eas!reents are in prin"iple simple and can, therefore, be handled by the local caretaker.
8owever, regular conventional turbidity measurements, although theoretically simple, may be difficult
to carry out in rural areas. Transport and communication problems, the fragile nature of delicate
instruments and the difficuIties with regard to commodities supply e.g. batteries, standards!, are
aspects which may lead to possible failures of the most simple turbidity monitoring programme. 2t!rd&
and siple field test et$ods $ave t$erefore )een developed to allow characterisation of mainly
physical water properties under actual field conditions. .lthough the different methods described in
.nne, 4 will not provide absolute but relative values, they are, however, a useful tool for water quality
description of any specific treatment plant.
A siple t!r)idit& test t!)e developed by 5el.gua J9K replaces the common turbidity meters which
usually require a power supply. Since the visual method is dependent on the sensitivity of the eye, it is
not as accurate as electronic systems, especially in the high turbidity range. The lower practical limit of
the tube can measure five TU Turbidity Units! and therefore meets the turbidity standard required by
slow sand filtration.
T$e filtera)ilit& test roughly indicates the amount of suspended solids in the water and can therefore
be used in place of the standard method for the determination of the suspended solids concentration
which requires a highly accurate scale, a vacuum pump and a drying furnace in an air<conditioned
room. "urthermore, modified Imhoff cones are used for the determination of the settleable solids
T$e sta)ilit& test gives some information on the settling characteristics of the colloidal matter and on
the stability of the suspension. The results of this test not only reflect the size and surface properties of
the solids but also the chemical and organic composition of the water. .dsorption of -a
and /g
on suspended solid surfaces may destabilise a suspension, while humic substances have been reported
to increase, in many instances, the stability of a suspension.
Water saples s$o!ld )e dra(n fro t$e ra( (ater and fro t$e filter inlets and o!tlets as
indicated in "ig. 24. .dditional sampling points may be used to optimise a roughing filter layout e.g.
by possibly e,changing the gravel size!. The efficiency of the individual filter layers can be e,amined
by sampling tubes installed at the end of the different filter layers. #ater sampling from these tubes
should be carried out carefully in order not to resuspend the deposits around the sampling point which
would otherwise lead to inaccurate results. 5rop<wise sampling is recommended, and the first tube of
sampled water must be discarded before starting with the actual sampling.
2iple field test e8!ipent s$o!ld )e allo"ated to ea"$ treatent plant* The caretaker must be
properly trained in order to carry out the different water quality tests and monitoring programme for his
treatment plant. .n e,ample of such a monitoring programme is summarised in .nne, ;. The local
caretaker should be assisted and guided by a supervisor attached to the operation and maintenance
section of the governmental institution responsible for the water supply i.e., district or regional water
administration!. The supervisor will initially carry out monthly and later biannual visits to the treatment
plant in order to support the caretakerDs daily activities and provide a feedback which will be useful for
the design and operation of other treatment plants.
'2,4 Filter cleaning
Filter effi"ien"& is not "onstant )!t a& in"rease at t$e start of filter operation and "ertainl&
de"rease ($en solid atter a""!!lates e="essivel& in t$e filter* 8ence, periodic removal of this
accumulated matter is required to restore efficiency and possibly hydraulic filter performance. "ilters
are cleaned either hydraulically or manually, and the cleaning methods are dependent on the way solids
accumulate in the filter. 8ence, the cleaning procedures will therefore have to be adapted to the
different filters.
In inta/e filters+ t$e solids ainl& a""!!late on top of t$e filter )ed* >y increasing flow velocity
over the filter surface, a fraction of these accumulated solids may be dragged away by the water.
8owever, inta/e filters are !s!all& "leaned an!all& (it$ a ra/e and s$ovel on"e a (ee/* The first
step in the cleaning process is the closing of the valve on the prefiltered water line. Thereafter, the inlet
control valve is opened to increase the horizontal<flow in the filter bo, to about 3.73 m$s to 3.=3 m$s.
The flow over the filter surface may also be increased by closing parallel filter units and directing the
total raw water flow into the unit to be cleaned. This method is particularly advisable in systems with
limited raw water supply such as in pumped schemes or of small hydraulic pipe capacities. T$e solid
atter retained )& t$e filter is first res!spended )& e"$ani"al stirrin% and t$en fl!s$ed )a"/ to
t$e river* /anual cleaning should start at the upper end of the filter and continue in flow direction to
avoid silting of the cleaned gravel. The gravel of the intake filter has to be cleaned completely about
once a year. . flat concrete slab ne,t to the filter should thus be available to deposit and wash the
gravel. . backwash system with a false bottom may be installed in intake filters where a large amount
of raw water at least 43 I$s per m] filter area at a minimum pressure of 7 m water height! is available at
the site. "ilter operation is restarted by draining the prefiltered water into the river or to waste until it
turns clear. Thereafter, the pretreated water can be reconveyed to the subsequent filters of the treatment
2in"e d&nai" filters are also s!rfa"e filters+ t$e& are "leaned an!all&* The cleaning procedure is
similar to that of intake filters, however, dynamic filters have to be cleaned after each high raw water
turbidity event or when filter resistance gradually increases over an e,tended period without turbidity
peaks. -leaning of dynamic filters is easy due to the relatively small filter area as a consequence of the
high filtration rate applied I" S 2 m$h!.
"ig. 24 Cocation of #ater Sampling 'oints
Ro!%$in% filters are ainl& "leaned $&dra!li"all& )!t "an also )e "leaned an!all& if ne"essar&*
*egular cleaning of the filter medium is important for proper filter operation. -ontrary to filter
operation under laminar flow, hydraulic filter cleaning is carried out under turbulent flow conditions.
The water stored in the filter is drained out of the filter compartment at high drainage velocities. In
order not to lose too much washwater stored in the filter, the valves or gates should be opened quickly.
2$o"/ draina%e is a"$ieved )& a fast openin% and "losin% of t$e valves or %ates connected to the
underdrain system of the filter. Starting and stopping the drainage process will induce unstable flow
conditions that will loosen and disintegrate the solid deposits accumulated in the filter. Subsequent high
drainage rates are applied to flush the resuspended solids out of the filter. 8igh peaks of suspended
solids concentrations can be observed in the washwater as illustrated in "ig. 27. 8owever, these high
concentrations rapidly decrease with progressive drainage time and additional drainage cycles. The
suspended solids concentration in the washwater shows a slight increase at the end of filter drainage
when the remaining sludge deposit accumulated at the floor is washed out. In verti"al4flo( ro!%$in%
filters+ ea"$ filter "opartent "an )e drained separatel&* This allows individual cleaning of the
specific filter compartments or of part of a filter if the false filter bottom is divided into segments.
-onventional filter backwashing as applied in rapid sand filtration is not possible since the filter bed of
roughing filters cannot be fluidised. A lar%e vol!e of (as$(ater is availa)le in $ori3ontal4flo(
ro!%$in% filters+ since the different filter compartment are separated by perforated walls, allowing the
stored water of the entire filter to flow into the opened drainage pipe. 8ence, a considerable volume of
washwater is available to flush the sludge accumulated around the drainage pipe out of the filter.
8owever, unless all drainage pipes are opened simultaneously, large vertical drainage velocities
required to flush the deposits accumulated in the entire bed to the filter bottom are more difficult to
achieve. In such a situation, the high washwater discharge may create a disposal problem. In
$ori3ontal4flo( ro!%$in% filters+ it is ver& iportant to start t$e "leanin% pro"ed!re at t$e inlet
side as most of the solids are retained in this part of the filter. .n initial vigorous drainage at the rear of
the filter would wash the bulk of the solids towards this drainage point and enhance the risk of clogging
the fine filter part.
"ig. 27 Suspended Solids -oncentration in #ashwater of Three Subsequent "ilter 5rainage -ycles
Effi"ien"& of $&dra!li" "leanin% "an )e assessed )& $eadloss "oparison )efore and after filter
draina%e* "or this purpose, measurements in the filter inlet and outlet must be conducted under the
same operational conditions, e.g. with similar filtration rates before and after filter cleaning. /anual
cleaning is necessary if initial filter resistance starts to increase and no filter regeneration is observed
after hydraulic cleaning. Installation of transparent plastic tubes, used as piezometers and fi,ed to the
outer wall of the filter bo, at the end of each filter fraction, can be useful for additional headloss
control. The headloss data, recorded at these points, is used to determine regeneration efficiency and
detect premature clogging of the individual gravel fractions. -areful recording of the water table is
important since the difference in head between the subsequent filter layers is usually only within a few
millimetres or centimetres. If the water level reaches the top of a horizontal<flow roughing filter, filter
resistance might become the decisive criteria for manual cleaning. . free water surface on top of such a
filter should never be tolerated since filter efficiency will dramatically drop due to short<circuiting of
the water.
Filter "leanin% fre8!en"& %reatl& depends on t$e ra( (ater "$ara"teristi"s+ filter la&o!t and
operation* /ost of the solid matter ;3<F3B! of tropical surface water is usually composed of stable
inorganic material. Since this type of material does not change the chemical properties of the water
passing through the filter, it can therefore be stored in the roughing filters without negative impacts.
8owever, high levels of organic matter call for frequent and regular cleaning to avoid consolidation of
solid deposits, decomposition of the organics in the filter and to prevent water quality deterioration with
regard to taste and odour. @evertheless, re%!lar $&dra!li" "leanin% is advisa)le sin"e it en$an"es
filter effi"ien"& and red!"es sl!d%e "opa"tin% and an!al "leanin% fre8!en"&*
T$e ann!al $&dra!li" "leanin% s"$ed!le has to be adapted to the annual fluctuation of the raw water
quality. 8igh turbidity loads are preferably treated by relatively clean filters to prevent a breakthrough
of the solid matter that would affect slow sand filter operation. It is therefore recommended to
thoroughly clean the roughing filters before peak loads e.g. before the beginning of the rainy season!.
8ydraulic cleaning can be handled by the caretaker and does not normally require e,ternal assistance
e.g. community participation!. Ea"$ "areta/er (ill+ t$erefore+ $ave to esta)lis$+ t$ro!%$ pra"ti"al
e=perien"e+ t$e optial "leanin% pro"ed!re and fre8!en"& re8!ired )& $is o(n treatent plant*
The caretaker will certainly be most interested in an efficient hydraulic cleaning since manual cleaning
is time<consuming and labour intensive.
7an!al "leanin% !st )e applied ($en t$e solids a""!!lated at t$e filter )otto or+ at (orst+ all
over t$e filter+ "an no lon%er )e reoved $&dra!li"all&* This occurs if proper hydraulic cleaning has
been neglected or if solid matter has cohered to the filter material or at the bottom. . slimy layer may
cover the filter material if there is biological activity in the filter resulting from high loads of dissolved
organic matter in the water. This biological layer will most probably increase filter efficiency at the
beginning, but will subsequently hinder the drift of deposited matter towards the filter bottom.
.ccumulated cohesive matter might also hinder self<regeneration of the filter. "inally, retained material
in silted but drained filter beds will also dry up and form a skin around the filter material. Thus,
ro!%$in% filters s$o!ld never )e /ept dr& !nless t$e filters are properl& "leaned in advan"e*
7an!al "leanin% pro"ed!re ainl& "onsists in e="avatin%+ (as$in% and re4installin% t$e filter
aterial* The filter material is e,cavated from a drained filter. The coarsest filter material is normally
removed first, cleaned and thereafter refilled into the filter section. The first part of the filter material
may be stored for awhile, whereas the remaining material can be washed and directly reinstalled to save
storage space and reduce cleaning operation. .s regards horizontal<flow roughing filters with strong
separation walls, each filter fraction is generally handled separately to avoid mi,ing of material.
Simultaneous e,cavation of the filter material is necessary in upflow roughing filters and in horizontal<
flow roughing filter where separation structures are weak, or where these walls are completely missing.
Resievin% of t$e filter aterial is ne"essar& if i=in% of t$e different fra"tions o""!rred or if the
filter medium has been broken up into smaller pieces due to e,cavation and mechanical cleaning. .
well<specified, uniform size for each filter fraction is essential to maintain high porosity of the filter
bed. In this conte,t, it is clearly more advantageous to install a mechanically<resistant filter material
right from the beginning. *einstallation of the filter material should not create any difficulties.
8owever, the material should preferably be brought into the filter right after having been washed to
avoid contamination with dust or other impurities. 5isintegrated material in roughing filters must be
replaced and filled back to its original level. . stock of additional filter material should therefore be
kept at the treatment plant.
Filter "leanin% involves a %reat deal of an!al (or/+ often )e&ond t$e "areta/erTs a)ilit&*
.dditional manpower must be mobilised either by contracting local casual labourers or by involving the
community. -areful planning and organising is necessary when manual filter cleaning is carried out
with village participation. The cleaning schedule should, for instance, not coincide with a period of
intensive agricultural work. Ade8!ate aterial and tools !st )e provided to allo( effi"ient filter
"leanin%+ otherwise maintenance work will become too tedious and might never be done. /anual filter
cleaning requires shovels, sieves, preferably two to three sturdy wheelbarrows, some wooden boards,
and buckets. The same material already used for construction should therefore remain at the treatment
plant or in custody of the local operator at the end of construction.
'2,) Filter maintenance
7aSor in"idents are often t$e res!lt of inor "a!ses* This saying also applies to roughing filter
maintenance. "ilter maintenance is not very demanding as the prefilters do not include any mechanical
parts apart from the valves. @evertheless, maintenance should aim at maintaining the plant in good
condition right from the beginning. 0,ternal assistance for maintenance work can usually be avoided if
t$e follo(in% (or/ is "arried o!t properl& )& t$e lo"al "areta/erD
periodi" !p/eep of the treatment plantDs premise grass cutting+ removal of small bushes and trees
which could impair the structures by their roots+ removal of refuse!
soil prote"tion against erosion especially surface water intake structures, the washwater drainage
channels and surface runoff!
repairin% fiss!res in the walls of the different structures and replacing the chipped plaster
appli"ation of anti4"orrosive a%ents to e,posed metal parts I<notch weirs, gauging rods, pipes!
"$e"/in% t$e different valves and drainage systems, and occasionally lubricating their moving parts
(eedin% t$e filter aterial
s/iin% off floatin% aterial from the free water table
(as$in% o!t "oarse settled aterial distribution and inlet bo,es!
"ontrollin% t$e an"illaries and replacing defective parts tools and testing equipment!.
T$e ter Mperiodi"M does not onl& appl& to t$e first point in t$is "$e"/ list )!t to all of t$e*
'roper maintenance of the treatment plant guarantees long<term use of the installations at low running
'hoto F 8ydraulic -leaning of *oughing "ilter @ote the Simple 5esign of the "ast 6pening 5evice as
also illustrated in "ig. (=, p. Id<F!
E"citin 0atural Science >ectures
E"citin 0atural Science >ectures
)r. Auustine is a secondary school teacher in 5ehana# a district centre in a Sahelian reion. +is
favourite su!ject is natural science as this su!ject provides the students with a ood !ase for their
adult life. +owever# he was often faced with a passive audience# the pupils also had difficulties in
applyin the su!ject tauht. ,or instance# since temperature is of paramount importance in this hot
country# )r Auustine e"plained that water free@es at & MCelsius E MCF or at =. M,ahrenheit E M,F. The
am!ient temperature was around =. MC which e(uals a!out 9& M,# !ut the water starts to !oil at 9& MC
or at 9& M, H The pupils were confused a!out all the listed fiures and could not et familiar with the
presented theory as )r Auustine had no thermometer to ma*e practical measurements.
+owever# natural science also had its ood sides# especially when the school children could clean the
ravel and the sand of the community water treatment plant. It was always very shrillin to listen to
)r Dic*son-s e"planations. )r Dic*son was the careta*er of the treatment plant and pro!a!ly an
e"cellent enineer as he *new a lot a!out his filters. )r Auustine-s class was invited twice a year to
come and clean the filters of the treatment plant. The school children firstly carried out the manual
wor* lastin for a!out two hours. Thereafter# they could sit under the shade of a palm tree and listen
to the e"citin stories of )r Dic*son. +e told them that very small livin particles are found in the
tur!id raw water# particles which caused diseases such as diarrhoea or the terrifyin uinea worm.
+owever# when the water travelled throuh the filters# these particles# which were retained !y the
ravel and the sand# started to starve in the filters due to lac* of food and eventually died. That was
also the reason why filter cleanin was not at all a ris*y jo!. The students were a!le to understand the
e"cellent e"planations of )r Dic*son.
'nce )r Dic*son showed the students round papers covered with small dots. +e had received these
papers from the la!oratory technician who had recently !een e"aminin the raw and treated water for
!acteria as he termed these livin oranisms. The raw water papers were filled with innumera!le
yellow dots whereas the treated water papers had just two dots# one was even red. The school children
were then really persuaded that the filters were an efficient !arrier to safeuardin their health from
The temperature was s*y/hih and the sun was !eatin down mercilessly on the students who had to
return to their lessons. 'n the way !ac* to school the de!ate on !acteria continued and the pupils
from the neih!ourin communities discussed the idea of introducin a similar treatment system in
their villaes.
'4, %conomic aspects
2in"e n!ero!s fa"tors affe"t (ater treatent plant "osts+ %eneral+ a)sol!te val!es "annot )e
"ited* Such factors include type of treatment plant, local material and labour costs, mode of
implementation construction by private contractors, national institution, or community participation in
a self<help pro1ect!, and geographic location affecting type of structure used to fulfil climatic
conditions, accessibility influencing transport costs, etc.!. T$e overall "osts "oprise t$e initial or
"onstr!"tion "osts+ as (ell as operatin% and aintenan"e "osts* T$e different "osts "an )e
s!)divided into lo"al and forei%n "osts+ an aspect of great importance for developing countries which
may have to import part of the equipment and material required for the treatment plant.
'4,' Construction costs
An eval!ation of t$e "onstr!"tion "ost str!"t!re for different ro!%$in% filter proSe"ts whose design
capacity ranged from 93 to 923 mb$d and located in Tanzania, Genya, Indonesia, and .ustralia, reveal
the following, rather similar breakdown of construction costs)
'ercentage of the construction costs)
< earthwork and structure about 93B
< filter medium about 73B
< piping and accessories about 43B
Topography and soil conditions required e,cavation work and type of foundation!, including type of
structure reinforced concrete or brickwork!, are cost decisive factors for earthwork and structure.
.vailability of local filter material in the required sizes significantly influences the purchase price+ i.e.,
the supply. These first two cost components only have low economies of scale, however, the relative
costs for piping and accessories will decrease with increasing plant size.
T$e spe"ifi" ro!%$in% filter "onstr!"tion "osts per O of installed filter vol!e range between US e
433 and 492, e,cept for the plant in .ustralia where the specific costs amount to US e :33. It is,
however, not only the smallest in capacity and built in reinforced concrete, but it also reflects the prices
of a private contractor in an industrialised country. In developing countries, specific costs ranging from
US e 423 to 733$mb will most probably cover the roughing filter construction costs. In self<help pro1ects
where only building material has to be paid for, construction costs can be reduced by (3<23B.
T$e spe"ifi" ro!%$in% filter "onstr!"tion "osts per O:d (ater o!tp!t are dependent on filter length
and applied filtration rate. "or an assumed total filter length of 2 m and a filtration rate I
of 3.2 m$h for
7= in$d, the resulting specific daily costs amount to)
about US e$ mb$ d
construction costs :3<;3
material costs only
e.g. in self<help pro1ects!
Constr!"tion "osts for slo( sand filters are dependent on t$e respe"tive filter la&o!t and desi%n*
-osts are greatly influenced by type of filter bo, chosen earthen basin and reinforced concrete filter
bo, are the two e,tremes!, and by the sand price. . slow sand filtration cost study made in India J2FK
and based on 4F9F prices revealed specific construction costs of about US e 72<=3$mb$d fore daily
design capacity ranging from 93 to 923 mb. . material cost estimate in the Slow Sand "ilter /anual
J:3K revealed higher specific costs. "or the less e,pensive slow sand filter options with sloping walls or
masonry structure, the material costs were estimated between US e =3:3$mb$d and US e 4:3<7=3$mb$d
respectively. 8owever, the cost estimate for these plants, whose design capacity ranges between 93 and
(23 mb$d, also includes material costs for small clear water tanks of 73 to =3 mb volume. 6wing to the
different material and labour costs, it is difficult to obtain a generally valid cost indication as
demonstrated by the two studies.
A ore "opre$ensive "osts eval!ation (as ade of -H slo( sand filters "onstr!"ted in t$e U2
J49K. 6f these 42 plants, five slow sand filter schemes, ranging in capacity between 4(3 and 4;F,773
mb$d, are gravity operated and have no electrical equipment. . construction cost evaluation of these five
plants resulted in the following subdivision of construction costs)
'ercentage of the construction costs)
site work about 43B
filter media and gravel about 72B
pipes, valves and meters about 73B
filter bo, about =2B
The specific construction costs for uncovered slow sand filters in the US showed the following clear
economies of scale)
- Q F,473f.
- Q construction costs in US e
. Q filter surface area in m]
* Q regression value
. slow sand filter plant with a 23 m] filter area, operated at a 3.42 m$h filtration rate, has a capacity of
4;3 mb$d. -onstruction of such a plant would cost about US e :7,333. >ased on the above equation,
investment costs for a slow sand filter plant with double the filter area and daily capacity would amount
to about US e ;9,333. For slo( sand filters+ this results in the following spe"ifi" "onstr!"tion "osts
per dail& "apa"it&D
for a slow sand filter plant of)
4;3 mb$d about US e (=2$mb$d
(:3 mb$d about US e 7=7$mb$d
These specific costs clearly demonstrate the economies of scale. "urthermore, when applied to the
considered filtration rates in roughing and slow sand filters 3.2 and 3.42 m$h!, the specific construction
costs have a similar order of magnitude. The modified specific costs in the ratio of 3.2$3.42 for slow
sand filters would amount to about US e 433 and 93$mb$d for the 4;3 and for the (:3 mb$d plant, and
can be compared to the costs of US e :3 < ;3$mb$d given for roughing filters.
. construction cost comparison is drawn between slow sand filters and rapid sand filters J4;K.
.ccording to an evaluation of the specific construction costs for seven slow sand filter plants, they
range between about US e (23 and 7,233$mb$d. 8owever, it appears that the construction costs for small
capacity slow sand filter plants are substantially lower than for equivalent rapid sand filter plants on
account of their simple design and minimum mechanical equipment requirement. In addition, roughing
and slow sand filter plants tend to a have long service life which reduces the annual depreciation rates
of the capital costs.
'4,. !perating an maintenance costs
Filter "leanin% "osts are t$e ain operatin% "osts for ro!%$in% and slo( sand filters as the filters
do not require any chemicals. 6n the one hand, sin"e $&dra!li" ro!%$in% filter "leanin% is "$eap+ as
t$e (or/ "an )e "arried o!t )& t$e "areta/er+ t$e r!nnin% "osts (ill reain lo(* 6n the other hand,
$&dra!li" filter "leanin% is ver& "ost effe"tive (it$ respe"t to total operatin% "osts as it reduces the
frequency of manual cleaning known to be labour and cost intensive and generally requiring additional
manpower. The manual cleaning frequency differs for each filter type. Intake filters are usually cleaned
once a week, dynamic filters after every heavy rainfall. /anual cleaning of roughing filters may be
required every three to five years, or may not be necessary with the installation of an efficient drainage
system. "inally, a slow sand filter run may last from one to si, months.
Since salaries of water plant operators vary significantly, t$e tie re8!ired for filter "leanin% is t$e
)est (a& to assess its "osts* /anual cleaning of a 7<m] large intake or dynamic filter may take about
half an hour if the gravel does not have to be removed from the filter bo,. /anual cleaning of roughing
filter media is more time<consuming as the gravel has to be removed from the filter bo,, transported to
the washing place, washed and reinstalled into the filter bo,. .s a rule, about 4.2 mb gravel per man$day
can be cleaned. 8ence, upflow roughing filters run at 3.2 m$h filtration rate for a 7=3 mb$d capacity
treatment plant will require a total labour input of about 4= man$days for manual cleaning or, more
practically, three men could clean 73 mb of gravel in one week. "inally, a manDs daily ability to scrape a
7.2<cm thick layer from a slow sand filter and to transport the sand in buckets to the sand washing place
may be in the order of 433 m] filter area. 23 m] slow sand filter area filtration rate 3.73 m$h! would be
required for the 7=3 mb$d treatment plant, or half a day for one man. /ore realistically, two men would
be able to scrape the sand from a 7=3 mb$d slow sand filter plant in half a day. Sand washing can be
carried out later when the slow sand filters have restarted operation. 8owever, a 7=3 mb$d plant capacity
is split into several roughing and slow sand filter units, t$e different filters are "leaned s!""essivel& in
order to %!arantee an !ninterr!pted s!ppl& of (ater* These specific cleaning capacities and the
annual cleaning time required for the operation of 7=3 mb$d treatment plant are summarised in Table ;.
Ta)le R Ann!al Cleanin% Tie for a 0GC O:d Treatent Plant
Treatent 2"$ee inta/e filter !ppflo( ro!%$in% filter slo( sand filter
Treatment 5esign
filtration rate v

4.72 m$h 3.2 m$h 3.73 m$h
filter bed area .
; m] 73 m] 23 mb
filter bed height 8
3.= m 4 m 4 m
filter material volume I
(.7 mb 73 mb 23 mb
filter units <
7 7 7
filter area per unit
= m] 43 m] 72 m]
-leaning Interval
dry season ; months 7 ,$month 4 ,$month 4 ,$= months
rainy season = months = ,$month 7 ,$month 4 ,$= months
-leaning Time per "ilter
hydraulic cleaning
< 3.72 man hour
manual cleaning
3.2 man hour ; man hours
-leaning Time
dry season 4: man hours = man hours (7 man hours
rainy season 4: man hours = man hours 4: man hours
.nnual -leaning Time (7 man hours ; man hours =; man hours
/anual "ilter -leaning
frequency of manual filter material washing 4 ,$year 4 ,$2 years 4 ,$43 years
cleaning capacity 4.2 mb$man$day 4.2 mb$man$day 4 mb$man$day
cleaning time per filter 4 man$day 9 man$days 72 man$days
contribution to annual cleaning time
4: man hours 77 man hours =3 man hours
Total .nnual -leaning Time =; man hours (3 man hours ;; man hours
2in"e operation of ro!%$in% and slo( sand filters onl& re8!ires la)o!r inp!t+ any community with
a strong interest in treated water can afford the use of such filters. The running costs can be reduced to a
minimum if the community participates in filter cleaning. 8owever, routine operation and maintenance
of a treatment plant will be carried out by a caretaker or operator who significantly influences
efficiency, reliability and water treatment costs. The community should appreciate his work and reward
him accordingly. T$e f!ll& self4reliant treatent pro"esses are t$erefore not dependent on an&
e=ternal finan"ial and te"$ni"al s!pport* Important operating and maintenance costs, which are often
too high for a rural community, can be reduced to an absolute minimum by the installation of self<
reliant treatment processes, such as roughing and slow sand filtration. This is one criteria for long<term
operation of any water supply scheme.
'4,7 !verall costs of 3ater supply schemes
T$e "onstr!"tion "osts of a (ater treatent plant a& a/e !p an iportant part of t$e overall
investent "osts of a (ater s!ppl& s"$ee* -ost comparisons between different water supply
alternatives should therefore be made in the preliminary pro1ect design phase.
2ele"tion of t$e (ater so!r"e s$o!ld )e )ased on te"$ni"al and e"onoi" "onsiderations* 6perating
and maintenance costs of a gravity<operated system are low compared to a slow sand filter plant, which
requires about =3B of the initial construction costs to operate a plant for a period of 73 years. Under
such aspects, it may be more economical to tap a more remote source of good quality water and invest
in a longer water supply main than in a treatment plant.
Constr!"tion "osts for r!ral (ater s!ppl& s"$ees "an )e s!)divided as follo(sD
'ercentage of the construction costs)
< intake works 2 < 72B
< treatment plant 42 < (3B
< distribution system 23<93B
Ann!al operatin% and aintenan"e "osts $ave to )e added to t$e ann!al "apital re"over& "osts to
obtain a correct cost analysis. 8owever, since operating and maintenance costs greatly differ, a general
estimate is difficult. 7B of the construction costs may be an indicative figure to estimate operating and
maintenance costs of roughing and slow sand filters, e,cluding the depreciation costs for the treatment
Finall&+ t$e lo"al and forei%n "!rren"& "ost "oponent "an also )e an iportant fa"tor in proSe"t
)!d%etin%* *oughing and slow sand filters are an essentially self<reliant technology which is largely
reproducible with local means. >ased on the construction cost structure, ;3 < F3B of the investment
costs are e,penditures for construction material, such as gravel, sand, cement, bricks and stones, and for
labour, both readily available in the country. The remaining 43 73B are costs for the purchase of pipes,
valves and accessories I<notch weirs, gauging rods! which may partly have to be imported. Hen"e+
ro!%$in% and slo( sand filters a/e a=i! !se of lo"al reso!r"es+ re8!ire ostl& lo"al
"!rren"& and red!"e t$e need for forei%n "!rren"&*
Padloc*s Enhanced Collection of 4ater ,ees
The tri!e of the Ewes in 4est Africa is *nown for its stron communal spirit and leadership. These
attri!utes also formed the !asis for the successful water supply project implemented as self/help
project si" years ao. The population had !een sufferin from !ilhar@ia and uinea worm which had
spread over the villae throuh ponds used !oth as water source !y the villaers and as !reedin place
!y the two tropical diseases. The community decided to improve its water supply durin the annual
assem!ly held at Easter. 2olly# one of their citi@ens who had !een trained a!road in civil enineerin#
desined the water supply scheme. It consisted of rouhin and slow sand filters fed !y ravity from
the larest laoon# a pumpin station# a reservoir located on top of the central hill# and a distri!ution
system supplyin the nine community settlements of =#%&& inha!itants.
The community# which contri!uted in cash and *ind to the project# could cover half of the total costs
of the project which was completed in three years of hard communal la!our. >eadership was not
always easy durin this period as technical pro!lems had to !e solved and the community motivated to
wor*. +owever# the last tap on the most remote pu!lic standpost was finally installed and the electric
pumps connected to the rist system. At that moment# water should have started to flow !ut it did not.
Since some of the settlements had not yet paid their fees to the water committee# operation of the water
supply was put off until all due payments had !een made. The community went throuh a rouh
period# however# when the last Cedi was finally paid to <)ister )oney<# the nic*named treasurer of
the project# an important cele!ration was oranised !y the entire community to inauurate the new
water supply scheme.
Pu!lic standposts located in every settlement sinificantly reduced the wal*in distance to the water
sources. )ost inha!itants enjoyed the commodity# however# some citi@ens wanted to *now why they
still had to pay for water althouh they had e"tensively contri!uted to the project. The water
committee fi"ed the monthly water rate for each adult at %&& Cedis# the e(uivalent price of a !ottle of
!eer. +owever# the electricity !ill# the salary of the two careta*ers and minor maintenance costs had
to !e covered !y the water fees levied separately in each settlement.. Despite lenthy meetins with the
water committee# a few villaers were still not willin to pay the fi"ed rate. This ave rise to tariff
policy disputes and jeopardised smooth operation of the water supply scheme until the local
!lac*smith suested the installation of padloc*s with steer cylinders on the taps. All the taps were
thus e(uipped with padloc*s which allowed a controlled supply of water to all those who settled their
water !ill reularly.
Today the opponents of the water fees are no loner contestin them# and the padloc*s installed at the
pu!lic standposts hinder people from neih!ourin villaes tryin to collect water illeally.
'), Design e;amples
Five desi%n e=aples are presented $ereafter to ill!strate prefilter and ro!%$in% filter
appli"ation* The population of rural communities is often in the range of 233 to 2,333 inhabitants. The
daily water demand is dependent on the service level+ i.e., whether the population is supplied by public
standposts, yard or house connections, as well as on the water price and water rate system used+ i.e., flat
rate or billed on the basis of water meter readings. In water supply schemes with public standposts,
daily water demands can vary between 73 and (3 litres per person.
The following design e,amples are based on the supply of a village with a current population of about
4,233 inhabitants and annual growth rate of ( B. The current water supply through public standposts
amounts to 73 litres of water per inhabitant. 8owever, water supply installations have to be
dimensioned to meet future water demands of the design population. For a -C4&ear desi%n period+ a
treatent plant operated on a 0G $o!r )asis (ill re8!ire t$e follo(in% desi%n "apa"it&D
current population
4,233 inhabitants
population in ten years 4W3.3(!
f 4,233 Q 7,333 inhabitants
current specific water demand 73 l$p$d
specific water demand in 43 yrs (3 l$p$d
current daily water demand
4,233 f 73 Q (3 mb$d
daily water demand in 43 yrs 7,333 f (3Q :3 mb$d
T&pe of ra( (ater so!r"e and ra( (ater 8!alit& deterine e=tent and t&pe of treatent*
Turbidity level and respective peak values are the most important parameters for pretreatment unit
design. *aw water quality data are often scarce and records of peak values generally unavailable. #ith
time, raw water sources may deteriorate in quality and quantity. #atershed protection is therefore
essential for a long<term use of the source. 8owever, such pro1ects may not prevent water quality
deterioration and, therefore, t$e treatent plant la&o!t s$o!ld allo( for t$e inte%ration of
additional pretreatent steps if t$ese t!rn o!t to )e ne"essar& in t$e f!t!re* "ig. 2( illustrates
possible annual raw water quality variations for the following design e,amples presented hereafter.
'),', Treatment of an uplan river
6ur presumed village is located in a hilly area of a #est .frican country. The area is very rural and
scarcely populated. Temperature and climate are pleasant, with cold nights and annual rainfall of about
7,333 mm, ideal for tea and coffee growth. A sall !pland river flo(s t$ro!%$ tea and "offee
plantations lo"ated on %entl& slopin% $ills s!rro!nded )& dense forests* In future, the farms will be
e,panding their plots further uphill. The central government has recently granted construction of a road
into the catchment area to facilitate the e,ploitation of wood. . water supply is foreseen to compensate
the villagers for the now polluted small upland river.
.fter a sanitary inspection of the catchment area, as well as a discussion with the village chief and with
a recently established water committee, it was decided to construct slow sand filters according to the
respective graph shown in "ig. 2=. 8owever, the slo( sand filters $ave to )e prote"ted fro $i%$ silt
loads "arried )& t$e river d!rin% t$e s$ort )!t $eav& rainfalls* The algorithm discussed and
presented in "ig. =( of -hapter 47 and attached in .nne, : is used to determine the type of pretreatment
required. "ig. 2= shows that t$e planned slo( sand filters are at present s!ffi"ientl& prote"ted )&
d&nai" filters+ $o(ever+ inta/e filters a& )e re8!ired in f!t!re if the raw water quality
deteriorates due to over e,ploitation of the catchment area.
'lease note that "igs. 2= < 29 are duplicates of "ig. =( shown on page dII<= to facilitate its use.
"ig. 2( .nnual Turbidity Iariation of 5ifferent Surface #ater Sources
T$e different filter !nits for t$e treatent of an !pland river (ater are diensioned as follo(sD
treatent plant "apa"it& IC O:d
:3)7=Q 7.2 mb$h
d&nai" filter
filtration rate 2 m$h
total filter area required 7.2 m$h) 2 m$h Q 3.2 m]
number of filter units 4
slo( sand filters
filtration rate 3.472 m]
total filter area required 73 m]
number of filter units 7
filter area per unit 43 m]
Inta/e filter
foreseen in case of future water quality deterioration!
filtration rate 4 m$h
total filter area required 7.2 m]
number of filter units 4
. possible dynamic filter layout is illustrated in .nne, :$4 and general design guidelines are
summarised in "ig. (9 of -hapter 43. Cayout and design of slow sand filters are described in .nne, (.
'),. Treatment of a lo3lan stream
The village in question may be situated in a valley of the Catin<.merican .ndes. .t that high altitude,
grass and farmland may predominate and forests are scarce. The living condition of the population is
difficult and the farmers are forced to use every possible plot of fertile land as a means of subsidence
and for cash crops. T$e onl& perennial (ater so!r"e and (ater s!ppl& of t$e pop!lation is t$e
valle& river loaded (it$ lateriti" soil eroded into t$e strea* The yield of the few springs generally
emerging at the bottom of the hills is low and the springs dry up in the hot season. &roundwater is non<
e,istent in the valley.
The river and its small waterfalls upstream fortunately allow construction of a gravity water supply
scheme. 8owever, t$e river is "onsidera)l& poll!ted )& t$e villa%es lo"ated in t$e !pper part of t$e
valle&* The `unta .dministrativa, a village committee also responsible for community water supply, is
aware of this public health risk and has initiated a water supply pro1ect with the support of a non<
governmental organisation. >ased on the selection criteria illustrated in "ig. 22, the pro1ect team has
designed a (ater treatent s"$ee "onsistin% of inta/e filters+ $ori3ontal4flo( ro!%$in% filters
and slo( sand filters* Up flow roughing filters in series are an economic alternative to horizontal<flow
roughing filters, and their hydraulic cleaning is also easier than the proposed option.
"ig. 2= Treatment option for an Upland *iver
"ig. 22 Treatment 6ption for a Cowland Stream
T$e different filter !nits for t$e treatent of a lo(land river (ater are diensioned as follo(sD
treatment plant capacity
:3 mb$d
:3)7= Q 7.2 mb$h
intake filters
filtration rate 3.; m$h
total filter area required (.7 m]
number of filter units 7
filter area per unit 4.: m]
horizontal<flow roughing filters
filtration rate 3.2 m$h
total filter area required 2 m]
number of filter units 7
filter area per unit 7.2 m]
layout of filter bed
4st gravel fraction " 47 <4; mm (.2 m
7nd gravel fraction " ; < 47 mm 7.2 m
(rd gravel fraction " = < ; mm 4 m
slow sand filters
filtration rate 3.472 m$h
total filter area required 73 m]
number of filter units 7
filter area per unit 43 m]
upflow roughing filters in series
as an alternative to horizontal<flow roughing filters!
filtration rate 3.= m$h
total filter area required :.72 m]
number of filter units (
filter area per unit :.72 m]
layout of filter bed
4st filter unit " 47 <4; mm 4 m
7nd filter unit " ; < 47 mm 4 m
(rd filter unit " = < ; mm 4 m
. possible layout of the intake filters is illustrated in .nne, :$7, and the general design guidelines are
summarised in "ig. (9 of -hapter 43. . design e,ample for horizontal<flow roughing filters is attached
to .nne, :$(, and the respective design guidelines are listed in "ig. (F. Cayout and design of upflow
roughing filters are described in .nne, :$ = and "ig. (;, and those of slow sand filters in .nne, (.
'),7 Treatment of reservoir 3ater
The village may be located in .sia. The landscape there may be very flat and interlaced by many canals
regulating the water table essential for the cultivation of rice. Cand is very scarce and, therefore, even
the smallest plots are used for agriculture. @evertheless, some larger lagoons for duck farming can also
be spotted. In former times, the population used the canal water as raw water source, however, this
source has become increasingly polluted due to increased motorisation of the ships and to industrial
wastewater discharge. 8ence, t$e la%oons alt$o!%$ e=posed to poll!tion fro a8!a"!lt!re and
a%ri"!lt!re 4 are 8!alitativel& t$e )est ra( (ater so!r"e no(ada&s* &roundwater may in the long
run not be used as water source due to the progressing infiltration of sea water which gradually
increases water salinity.
To avoid constructing a water supply scheme with two pumping stages, the water treatment plant will
have to be located at the lowest point ne,t to the dam of a lagoon. The treatment plant will have to be
gravity<fed, which will ensure its continuous operation. D!"/s and $!an a"tivities aro!nd t$e
la%oon deteriorate:$e i"ro)iolo%i"al (ater 8!alit&* "urthermore, use of agricultural fertilisers
enhances eutrophication of this reservoir water. . treatment scheme, as illustrated in "ig. 2:, is
therefore necessary to turn the raw water into drinking water. The populationDs high drinking water
standards will not allow the supply of chlorinated water as it affects the taste of tea. It (as+ therefore,
de"ided to "onstr!"t a treatent s"$ee appl&in% nat!ral p!rifi"ation pro"esses s!"$ as !pflo(
ro!%$in% filters in la&ers and slo( sand filters* . construction brigade started constructing the
treatment plant which was commissioned si, months later. This short construction period was possible
only thanks to the availability of local material.
"ig. 2: Treatment option for a *eservoir #ater
T$e different filter !nits for t$e treatent of a reservoir (ater are diensioned as follo(sD
treatent plant "apa"it&
IC O:d
:3 ) 7= Q 7.2 mb$d
!pflo( ro!%$in% filters in la&ers
filtration rate 3.( m$h
total filter area required ;.(( m]
number of filter units 7
filter area per unit =.7 m]
layout of filters bed
4st gravel fraction " ; < 47 mm 3.= m
7nd gravel fraction " : < 43 mm 3.( m
(rd gravel fraction " 7 < : mm 3.( m
slo( sand filters
filtration rate 3.472 m$h
total filter area required 73 m]
number of filter units 7
filter area per unit 43 m]
.n e,ample of a layout of an upflow roughing filter in layers is illustrated in .nne, :$=, and general
design guidelines are summarised in "ig. (; of -hapter 43. Cayout and design of slow sand filters are
described in .nne, (.
'),2 Rehabilitation of a slo3 san filter plant
"inally, the village in question may have completed its community water supply 72 years ago.
8owever, the situation in this #est .frican country has changed in the meantime. The population has
increased, however, most of its young people have left the village for town. Their income, earned in the
commercial capital, allows them to construct stately houses which they use during weekends and after
retirement, and to support the infrastructural pro1ects in the village. Inflation and a drop in prices for
agricultural products forces the rural population to practice e,tensive land farming. "eeder roads are
constructed to improve transport capacity to the capital and e,port to foreign countries. Indigenous
forests are cut down and tropical wood e,ported. The sturdy structures of the water supply allows a
more or less regular water supply to the village.
8owever, t$e distri)!ted (ater 8!alit& $as deteriorated si%nifi"antl& in t$e last ten &ears and $as
%iven rise to "ons!er "oplaints* T$e e=istin% sedientation tan/s and slo( sand filters "an no
lon%er "ope (it$ t$e in"reased ra( (ater t!r)idit&*
The small upland river, which was well<protected by a dense forest in former times, is now e,posed to
numerous sources of pollution. 0,tensive and careless farming has enhanced soil erosion, cows in the
grassland have considerably increased in numbers and neighbouring villagers have started farming in
the catchment area. .s shown in "ig. 2(, river water turbidity has increased noticeably and springs are
now running dry in the hot season. Slow sand filter runs amount to a few weeks in the dry season and
are reduced to a few days in the rainy season. The sand has therefore been removed from the filter
bo,es and the water now flows unfiltered into the reservoir. As t$e sedientation tan/s "annot "ope
(it$ t$e in"reased t!r)idit&+ t$e& (ill $ave to )e "onverted into ro!%$in% filters to allo( a
reasona)le slo( sand filtration* 8owever, the vault<type sedimentation tanks are not easily converted
into roughing filters. In such situations, part of the slow sand filter area may be used for the installation
of upflow roughing filters. The capacity loss of the reduced slow sand filter area can be compensated by
higher filtration rates of adequately pretreated raw water.
. rehabilitation option for an overloaded slow sand filter plant is illustrated in "ig. 29.
.n e,ample of a possible layout of an upflow roughing filter in layers is illustrated in .nne, :$=, and
general design guidelines are summarised in "ig. (; of -hapter 43. The Integration of an upflow
roughing filter into a sedimentation tank or into a slow sand filter is illustrated schematically in .nne,
"ig. 29 *ehabilitation option for a Slow Sand "ilter 'lant
T$e re$a)ilitated treatent plant is diensioned as follo(sD
ori%inal treatent plant "apa"it&
IC O:d
:3) 7= Q 7.2 mb$h
e=istin% sedientation tan/
length = m
width 4.2 m
depth 4.9 m
tank volume 43 mb
number of units 4
surface load 3.= m$h
retention time = hours
e=istin% slo( sand filters
filter length 2 m
filter width 7 m
filter area 43 m]
number of units 7
filtration rate 3.472 m$h
re$a)ilitation of t$e treatent plant
ne( treatent plans "apa"it&
NC O:d
F3) 7=Q (.92 mb$h
5st option*
transformation of the sedimentation tank into two upflow roughing filter units
!pflo( ro!%$in% filters in la&ers
total filter area required : m]
number of filter units 7
filter area per unit ( m]
filtration rate 3.: m$h
layout of filter bed
4st gravel fraction " 47 < 4; mm 3.9 m
7nd gravel fraction " ; < 47 mm 3.= m
(rd gravel fraction " = < ; mm 3.= m
slo( sand filters
total filter area required 73 m]
number of filter units 7
filter area per unit 43 m]
filtration rate 3.4F m$h
1nd option*
transformation of part of the slow sand filters into upflow roughing filters
e=istin% sedientation tan/
new surface load 3.: m$h
new retention time 7.9 hours
!pflo( ro!%$in% filter
integrated into slow sand filter bo,
filter length 4.72 m
filter width 7 m
filter area 7.2 m]
number of units 7
total filter area 2 m]
filtration rate 3.92 m$h
layout of filter bed
4st gravel fraction " 47 < 4; mm 3.9 m
7nd gravel fraction " ; < 47 mm 3.= m
(rd gravel fraction " = < ; mm 3.= m
red!"ed slo( sand filters
available filter length (.92 m
available filter width 7 m
filter area per unit 9.2 m]
number of filter units 7
total filter area 42 m]
new filtration rate 3.72 m$h
'),4 Stanar esigns for compact 3ater treatment plants
#ater treatment plant pro1ects can be implemented on a standard design basis to reduce design inputs as
well as construction time and costs. This approach is especially appropriate in rural water supply
programmes for the construction of several treatment plants having to treat raw water of similar quality.
In such situations, ro!tine "onstr!"tion pro"ed!res "an )e developed to red!"e "onstr!"tion
periods* F!rt$erore+ "opa"t desi%ns and "aref!l s!pervision of t$e "onstr!"tion (ill lo(er
investent "osts and ena)le t$e !se of e"onoi" "onstr!"tion pro"ed!res s!"$ as t$e ferro"eent
te"$ni8!e* Standard design modules often cover a range of different design capacities. They may be
implemented successively in different construction phases to meet the actual water demand of the
community. 0very water treatment pro1ect will nevertheless have to be carefully adapted to the local
situation and, therefore, calls for a critical evaluation of the prevailing conditions.
. standard design e,ample is illustrated in .nne, :$: . Upflow roughing filters!, slow sand filters!
and a reservoir are integrated in one structure. This e,ample uses a circular design, often applied in
reservoir construction, and takes advantage of locally available construction techniques. A "ir"!lar
rin%+ pla"ed aro!nd t$e reservoir lo"ated in t$e "entre+ provides spa"e for t(o treatent lines
"oprisin% !pflo( ro!%$in% filters and slo( sand filters* 5esign capacity of the illustrated e,ample
amounts to (3 mb$d. 8ence, two such standard design units are required to cover the water demand of
the village in question. These two units may be located in different places to treat different raw water
sources, and may improve the reliability of a water supply system. 5epending on the raw water quality,
the illustrated structure could be used to host alternative treatment systems such as upflow roughing
filters installed in the outer ring and two slow sand filters placed in the centre tank. Such a layout would
require a separate reservoir.
T$e "onstr!"tion of sall standard desi%n !nits also ena)les a p$ased in"rease of t$e treatent
plant desi%n "apa"it&+ satisfying the future water demand development. . further advantage of a
phased implementation is the integration of the operational e,perience in the e,tension design. Filters
"an fre8!entl& )e operated at $i%$er filtration rates without affecting the treated water quality or
without substantially reducing filter running periods. The filtration rate of slow sand filters may for
instance be increased from 3.4 to 3.7 m$h recommended range in the literature!. #ith an efficient
pretreatment and use of sand coarser than 3.42 < 3.(2 mm recommended range in the literature for the
specific sand size d43B!, it may be increased to 3.( and e,ceptionally to 3.= m$h.
T$e filter !nits of a "opa"t (ater treatent plant are diensioned as follo(sD
treatment plant capacity
(3 mb$d
(3 . 7=Q 4.72 mb$h
upflow roughing filters in layers
filtration rate 3.( m$h
total filter area required =.7 m]
slow sand filters
filtration rate 3.472 m$h
total filter area required 43 m]
'>, Final remar6s
If you have reached this part of the manual, you are either an e,perienced reader who first consults the
e,ecutive summary and conclusions of a publication, or a person with a real interest in the roughing
filter technology. .fter having given enough evidence in favour of prefilters and roughing filters, t$is
an!al (ill "on"l!de (it$ soe stron% stateents on r!ral (ater s!pplies in developing countries,
and will emphasise some weak aspects pertaining to the implementation of water treatment plants.
4. No (ater so!r"e (ill rea"$ people ($o erel& read p!)li"ations* Therefore, since this manual
presents a practical technology for field application, it is not 1ust meant for mental pleasure nor to be
filed away in a bookshelf, the reader is kindly requested to ta/e a"tion in his field by promoting and
implementing appropriate technologies.
7. .ppropriate means adapted to the local condition. Therefore, no te"$nolo%& "an )e !niversall&
appropriate* This is also true for slow sand filtration. The often negative e,perience with this treatment
process is frequently the result of an inappropriate raw water quality.
(. Appli"ation of siilar te"$nolo%i"al levels is a "riti"al fa"tor to achieve a sustainable system.
Insufficiently flocculated and settled water pretreated by comple, and unstable processes will create
operational difficulties even for the simple and sturdy slow sand filters. *aw water conditioned by
prefilters and roughing filters will usually meet slow sand filter requirements.
=. A"t!al deand and e"onoi" aspe"ts are decisive factors for the selection of a water supply
system. 'refilters, roughing filters and slow sand filters are fascinating treatment combinations as they
are based on a reliable, sustainable and reproducible technology. 8owever, since these filters require a
considerable structural input, they should only be favoured if no superior water quality source is
available and if water treatment is truly necessary.
2. This manual is mainly a technical document. 8owever, water supplies can be compared to computers
as they both depend on $ard(are and soft(are* The water supply users have to decide, contribute and
operate these facilities. Sociocultural aspects must be integrated in a pro1ect, and institutional aspects
considered carefully. 5egree of training, support and assistance to caretakers greatly influences the
performance and lifetime of a water supply. 8ence, an appropriate and sustainable technology always
requires an interdisciplinary input as illustrated in "ig. 2;.
"ig. 2; /ultidisciplinary Inputs for .ppropriate and Sustainable Technologies
:. Inforation e="$an%e s$o!ld )e re"ipro"al* ^our e,perience with roughing and slow sand filters is
important and your feedback essential. S.@50-, therefore, hopes to receive your views on this
manual, especially your practical e,perience with the presented filter technology.
'rovision of safe water is a great challenge. #e hope that this manual is a step in the direction of the
following policy formulated in @ew 5elhi J:4 K at the end of the International #ater Supply and
Sanitation 5ecade
M2oe for all rat$er t$an ore for soeM
The E"pert-s )irror
The villaers of some town somewhere were ready for the important cele!ration / the inauuration
of their water supply system. ,orotten were the hardships of the past two years and intensive wor*
of the construction of their self/help water supply system. All the villaers and the District Enineer
could !e proud of havin !rouht safe and clean water in the villae. The District Enineer# who
have also supervised construction# trained the careta*ers in operation and maintenance wor* and
had advised the water committee in !oo**eepin# spent the last few days assistin the villaers in
completin their system and oranisin the commissionin ceremony. Everythin was ready in time
for D/day when the <!i shots# the water urus or los sa!ios<# commonly called e"perts# honoured
the villae with their presence. They instantly !ean to inspect the wor* of the villaers# to evaluate
the system and to propose modifications. They also found the small District Enineer and !om!arded
him with (uestions# such asC <what-s your role in project etc.# etc.<. ,ortunately the drummers had
started to play and the District Enineer did not have to answer the (uestion immediately. After the
inauuration act# every!ody had time to visit the installations# the inta*e wor*# the treatment plant#
the distri!ution system with the reservoir and the pu!lic standposts. The atmosphere !ecame more
rela"ed even for the e"perts# and the small enineer had the opportunity to e"plain his role in the
projects. They were standin !y the reservoir overloo*in the villae when he told them that the first
listened to the real needs of the villaers and had to accepted their final decision even if it meant that
they did not want to chane their water supply system.
>ate afternoon# the District Enineer# tired of the hectic days# !oarded the !us to his home town. 'ne
of the e"perts was also in the !us and sat ne"t to an cryin !oy. The enine of the !us was ma*in an
terri!le noise. The villae was already several miles away !ut the small !oy went on cryin.
Em!arrassed !y his unusual situation# the e"pert concentrated on readin his !oo*. The District
Enineer then decided to sit ne"t to !oy in tears. To his surprise he discovered that the !oy was
afraid of the terri!le noise of the !us enine. After a while the !oy slowly calmed don and startin
tal*in with to the District Enineer. The e"perts# ama@ed !y situation loo*ed out of the window and
only saw his own face reflected !y the liht of the sunset in the windowpane.
)ost people tent to !ehave li*e the e"pert instead of the listenin# o!servin and tryin to understand
the situation !efore actin.