rainwater catchment systems in kenya and botswana

8/9/2019 Rainwater Catchment Systems in Kenya and Botswana

1/11

Experience with Rainwater Catchment Systemsin Kenya and Botswana

H. J. McPHERSON AND J. GOULDUniversity of Alberta, Edmonton, CanadaRainwater catchment'is a valuable but frequently neglected alternative a domestic water supply in the developingworld. Inthe arid and semi-arid areas ofKenya and Botswanaa numberofwatersupply programmes whichdependwhollyor inparton rainwater catchment are currently being implemented. This paper discusses this experience and describes the variousconstruction techniques being used. It has been found that rainwater catchments has a number of very unique advantagesand relatively few disadvantages. It is argued that more consideration should be given to the indusion of rainwatercatchment systems. especially in rural water supply programmes.

1. INTRODUCTIONThe last decade has witnessed a growing interest.by international agencies and national govern-"rnents in low-cost water systems aS,solutions to thewater needs of the developing world. There havebeen several reasons for t h i s ~ F i r s t , it has beenrealized that sophisticated solutions, such aspumping systems and treatment plants oftenassociated with a reticulated supply, can simplynot be afforded by many developing countries.Secondly, people are questioning whether hightechnology systems are the appropriate answer tothe needs of the developing world, especially inrural areas where frequently there is no centralizedgovernment maintenance programme.The water supply systems which are increasinglybeing built are those which promise to provideWater at low cost and, more importantly, whichcan be operated and maintained by the usersthemselves.In the water supply sector. major low costtechnologies include handpump wells, gravityWater systems and rainwater catchment. Wellsequipped with handpumps exploit groundwater

resources, and are perhaps the most universaltechnology in use today. Gravity water supplysystems are the second most widely used low-costtechnology, while rainwater catchment has been,to a considerable extent. largely ignored.However, in recent years there has been anincreasing appreciation that rainwater catchmentoffers advantage over the other systems, and thatin certain situations it may be the best answer.Rainwater catchment has been used in EastAfrica for some time and a considerable body ofknowledge about rainwater catchment has beenaccumulated.This paper describes the current technology, thecost of the various alternative systems beingconstructed and the major advantages anddisadvantages of rainwater catchment i'1 EastAfrica.

2. DEFINITIONFirst, what is meant by rainwater catchment? Arainwater catchment is basically any system whichcollects. stores and supplies rain\\;ater runoff forhuman needs. However. because this definition

Dr. McPherson is Professoro fGeography at the University ofAlberta. Edmonton. Canada. and has lIcted liS a consultant on wlItersupply lind sanitation to a number ofmajor international agencies.Mr. Gould is a graduate student at the University o fAlberta lind has worked in Botswana lind Nigeria.Natural Resources Forum United Nations, New York, 1985

BEST AVAILABLE COp,1


2/11

NRFVOL. 9, NO

Fig. 1. Rock catchment system.

Water reservoir ~ - : ,

Rock catchmentsurface

(1) Rock Catchments: systems for the colstorage and supply of rainwateuntreated rock surfaces (Fig. 1) ,(2) Ground Catchments: systems fcollection, storage and supply of rafrom treated and untreated ground(Fig. 2) and(3) Roof Catchments: systems for the colstorage and supply of rainwater frosurfaces (Fig. 3).

H. J. McPHERSON & J. GOULD

Cement or corrugated iron c o v e r i . l i l i l ~

Fig. 2. Ground catchment system.

could include such diverse systems as a puddlesupplying livestock with drinking water or ahydroelectric darn supplying the water and energy,it is helpful to define these systems in a moreprecise manner.Although there is much confusion andduplication of terminology, rainwater catchmentsystems (synonymous with rainwater collectionsystems) normally refer to small-scale systemsproviding individual households or singlecommunities with a primary or supplementarywater supply. In this paper three main types ofrainwater catchment systems will be described.There are:

254

Cemented ortreated eor1h catchment _ . . - s l ~ ~ ~ ~ ~ ~


3/11

NRFVOL. 9, NO.4, 1985 RAINWATERCATCHMENT IN AFRICA 255All of these systems consist of three components, a catchment area, a storage reservoir andan outlet for supply. In the case of rock catchmentand roof catchment , the storage reservoir isnormally above ground and the water is suppliedthrough gravity flow via a pipe or tap. For groundcatchments, however, the storage reservior isgenerally below ground and some method forextracting the water is necessary, either directly byhand or by pump.3. DETERMINING DEMAND, SUPPLY

AND STORAGE

storage.A second method is the graphical mass curvemethod. Figure 4 illustrates the use of this methodto determine the storage needed to provde a yearround supply of water. The graph is produced byplotting the cumulative rainfall runoff (rainfall xcatchment area x runoff coefficient) against timeand by placing the time representing the. s ~ e a d yconsumption of this water tangentially above themass curve. Then the most critical (driest period)in the data can be identified and the storagedemand estimated.

FMAMJ J ASOND

A SON 0

A SON 0

F M A M J

- - - - - - ~ - - - .----eeded ------ Cumulative__ -- water use---------------

F M A M J

Cumulativewaler collected

TOIOI runoff odded cumulalively

Calculalion of storage requiremenl

Mean monlhly roof runoff

Fig. 4. Simplified mass curve analysis for determiningrainwater tank storage capacity.

A more sophisticated method involves 'criticalperiod analysis' where the actual rainfall data forthe part icular locality are analysed and criticalperiods in the rainfall record identified. Thestorage requirement needed to overcome the mostcritical period is determined. Although this

The objective of a rainwater catchment scheme isto satisfy the demand for water as efficiently aspossible.The demand for water can be obtained from auser surveyor a project design value can beestimated. Seven litres per day per capita is, forexample, the minimum amount ofwater needed tosustain life. In many countries a per capita value ofbetween 20 and 30 lid serves a good indicator ofdaily water requirements.Calculating the available supply and the size ofthe storage reservoir needed to meet the estimateddemand is perhaps the most critical step indesigning a -catchment system. I f the storage tankis built too small the system will run dry and theusers will become disenchanted, ifthe storage tankis made too large this will greatly increase the costand reduce affordability. It is therefore essentialthat an accurate estimate of the supply and the sizeof the needed storage tank by determined.Two different estimates should be calculated forsupply. The first estimate is the total amount ofwater available from the annual rainfall: this issimply the catchment area multiplied by the meanannual rainfall multiplied by the runoff coefficient.The second estimate is the most economicallvfeasible rainwater supply system. To determinewhich system to use, it is necessary to compute thecost of the water which could be provided bystorage facilities at different capacities.A crude means of determining the storagecapacity required for any particular rainwatercatchment system is to determine the meanmaximum length of the dry. season (or meanmaximum period without rain) and calculate thevolume of water needed over this period. Thevolume needed is equivalent to the required


4/11

256 H. J. McPHERSON & J. GOULD NRFVOL. 9, NO.4, 1method is slow and laborious when done graphically or when using a calculator, it can be carriedout relatively quickly with the aid of a computer.Computer models for rainwater storage tankdesign have recently been developed by P e ~ e n s(1975) and Latham and Schiller (1984). Thecomputer model approach allows considerableflexibility, as different levels of supply determinedby percentage reliability of occurrence can' becalculated for any given system. Figure 5 shows thelevel of supply associated with any given storagecapacity at 95% and 100% realiabilities. forMahalapye, Botswana. Both the storage andsupply are given as fractions of the total usefulrunoff. Thus, a storage capacity equivalent to 0.4multiplied by the toal available runoff would

1.0

0.8

c. 0.61III010 0.4on

II0.2 IIII0.731

0.2 0.8 1.0

Fig. 5. Storage-supply curves (95% and 100% reliability)forMahalapye. Botswana.

supply 0.73 of this runoff with a 95% level ofreliability but only 0.49 with 100% reliability. Putanother way, this means that a 4 m:! tank shouldyield 7.3 m:! annually for 95% of the time.The computer model approach provides a veryreliable method for estimating the appropriatestorage tank size for any given climatic situation atwhatever level of reliability is desired.The capacity of the storage reservoir having

been determined. the next step is to construct thecatchment reservior.

4. CONSTRUCTIONThe nature of the construction will obviously vdepending on whether a roof catchment, rocatchment or ground catchment scheme is bebuilt. However, in all three an emphasis shouldplaced on building low-cost reservoirs where loinputs of labour and materials can be maximizCost is a key factor, and in East Afrconsiderable experimentation is currently unway to find the least expensive methods of buildcatchment systems.

4.1 ROCK CATCHMENTSReservoirs for rock catchments normally consisa dam wall behind which an open reservoir stothe rainwater. Generally the surface area ofreservoir will be too large to make coverineconomically feasible, despite the consideraevaporat ion losses which may be expected.some cases it may be worthwhile to investigateuse of cetyl alcohol or other evaporatS U p p l ~ s s a n t s . The darn is normally constructemassive concrete, but recent experimentsMutomo in Kenya by Nissen-Petersen (1985) hfound that a series of buttressed, arched darnsthe job just' as well and require less materialsome instances tanks have been constructedstorage of rainwater at small rock catchments.

4.2 ROUND CATCHMENTSDue to the fact that in ground catchment systethe catchment apron is at ground level, the storareservoir always consists of a subsurface taConsequently, some method of extractingwater from the tank is required. Usually a rope abucket or a pump is used.Two low cost and successful methodssubsurface tank construction observed in bKenya and Botswana are the lining of an excavapit with either butyl rubber or ferro-cement.Botswana a 700 m3 excavated butyl rubber lintank. used for irrigation was observed atForesty Brigade in Serowe. It was still functionin 1983 after 7 years of operation, although it nrequires cleaning. The lining of excavated pitsconsolidated soils with butyl rubber is one ofsimplest and cheapest methods of groucatchment tank construction. and this approahas been recommended for application in Afr


5/11

RAINWATER CATCHMENT IN AFRICARFVOL. 9, NO.4, 1985Bateman (1971).However, a more widely used method is theof the ground catchment tank with ferroment. In Botswana more than 400 tanks of thispe have already been constructed, and under the

of Agriculture's Arable Lands Developnt Program (ALDEP) several hundred morere planned. A detailed description of theiris given in Whiteside (1982) in aeaflet produced for local builders to show themow to construct the tanks. Basicallv, the methodvolves lining a cylindrical pit with chicken wirend plastering thiswith three layersofmortar. Theesign also includes a concrete base andrrugated iron cover (Fig. 6). Tanks have beenuccessfully constructed with a capacity ofetween 5 and 30 m:J. The Botswana Technologyis presently experimentingwith tanks of up

o 60 m3 for roof catchment.The ferro-cement tank was developed andesigned for the collection of rainwater fromaditional threshing floors. These plastered mudloors are normally up to 150 m3 in area and areund at most homesteads where crops are grown.e construction of permanent cement catchmentis preferable in terms of the quantity andof water which may be collected, but adds

257considerably to the cost of the system. For a 15 mJtank the cost is currently around US $350. I f thecatchment area is cemented, the cost of thecomplete system, catchment area and the tank isabout US $500.-To help to reduce costs, no pump is included inthe design of these tanks. The water is extractedusing a rope and bucket. A numberof experts haveexpressed concern about the quality ofwater fromthese tanks (Classen, 1980; Maikano and Nyberg,1981; Whiteside, 1982). Bacteriological analysis ofwater from them (Gould, 1984) revealed faecalcoliform counts of between 6 and 1000 in each100 ml s a m p ~ e . Although it has beenrecommended Whiteside (1982) that water fromthe tanks be b o i l e d ~ before drinking, few tankowners have taken this advice due to the lack offirewood and the inconvenience.

4.3 ROOF CATCHMENTTANKSIn contrast to ground and rock catchment systemswhere storage reservoirs always have to beconstructedon-site, for roof tanks the option existsof ei ther purchasing a factory-built tank orconstructing one on the site.

Brick kerb

...-_..:: - ... -_ - = - F R l i l _ t e ~ r; ;qr__ ~ ~ ~ ~ = = = : ~ ~ = c ~ o r : : = ; : r u ~ g ~ a : : t e ~ d ~ i r : : : o : : n~ c o ~ v : : ; : : e r = : : : : : ~ : : : z : ~ ~Treated wooden pole/"(cover support)

Fig. 6. Construction of ground catchment cement tank (cross-section).


6/11

258 H. 1. McPHERSON&1. GOULD NRFVOL. 9, NO.4, 14.3.1 Factory-built catchment tanks adequate.Probably the majority of roof catchment tanks In southeast Asia, one of the successful descurrently in use in Africa are factory-built. in use in Thailand and Indonesia is the bamAlthough the most common type of catchment reinforced tank. In Africa the absence of bamtank is the 200 I oil drum, these are too small to in most areas has led to designs based on feprovide for anything more than a few days of cement, cement blocks and the use of basketwstorage. Most large factory-built catchment tanks frames. Among the dozens of designs whichare of galvanized corrugated iron and are seldom been developed for the construction oflarger than 10 ma. The main disadvantage of catchment tansk, four stand out as havingcorrugated iron tanks is that the tanks themselves successful both in terms of cost effectivenessor the steel from which they are produced is operation. These include cement jar tanks, gimported, and therefore the tanks represent a baskets, concrete ring tanks and ferro-cemcapital-intensive solution which drains developing tanks.countries of foreign exchange. The relatively short Cement jar tanks. Cement Jar Tankslife-expectancy of "these tanks is another constructed using a hessian or cloth bag modisadvantage. Watt (1978) suggests most which is placed on a precast concrete foundacorrugated iron tanks last for 5-10 years. slab, fined with sawdust, grass, sand or any oGalvanized tanks do, however, offer certain appropriate filler, and is then plastereddistinct advantages over other types of roof . mortar. Once the mortar has set, the fillercatchment tanks. The most obvious of these is the mould are removed. These tanks have beenconvenience of installation. However, the structed by the hundred in Kitui district, Kerelatively short life-expectancy, especially in. with the support of the Roman Catholic Chcoastal areas due to salt in the air, does reduce the and other developed agencies. Cement jar teconomic advantage of galvanized tanks in can be built up to 10m:! according to Brelation to those made in cement. Although the (1983), but if they are larger than 3 or 4 malife-span of a galvanized tank depends on the must be reinforced with chicken wire or fendegree of preventat ive maintenance adopted, such wire. More-detai led descriptions ofas painting and cleaning, the gauge and quality of constructi0r:I are. given by Nissen-Petemetal used to construct the tanks also determines (1982), Byrne (1983) andMcPherson et al.their life-span. It seems clear that because many of (1984).the galvanized tanks installed in the 1950s and into Ghala basket tanks. More than 1000 gthe early-1970s are still operatingwhereas maRy of basket tanks have been constructedthose installed in the late-1970s and 1980s are community groups in Kenya, working undealready leaking, that the quality of the tanks being guidance of UNICEF. The design invoproduced by manufacturers in Kenya and South applying mortar to the inside and outsideAfrica (where Botswana's tanks are imported basketwork frame. Tanks up to 6 m:.! canfrom) has deteriorated. easily constructed employing only semi-sk4.3.2 Catchment tanks constructed on-site labour. It is essential that enough cemenStandard engineering wisdom normally results in used and that the outside as well as the insidover design of structures to minimize the risk of the tank is property plastered. In North Kitfailure. However, when small-scale roof catch- Kenya, some poorly constructed tanksment tanks are constructed in this way the result is suffered severe leakage problems. In conta very expensive tankwhich is not worth the added the tanks built by UNICEF in Karai hcost because the tank is unlikely to cause any performed very well. A detailed descriptioserious damage in the' event of failure. For "inis the materials required and the construcreason considerable time and effort has been technique is given by Nissen-Petersen (1invested into the development of low-cost designs. and McPherson et al. (1984).Although these designs do not adhere to the Concrete ring tanks. Concrete ring tanks astringent standards nonnally applied by civil relatively new design which have the advanengineers. they are nevertheless more than of being easy to construct. In Machakos dis


7/11

4.4 COSTCOMPARISONTable I shows the cost of a number of differenttypes of roof catchment tanks in Kenya andBotswana. A number of generalizations can bemade from the table. First, the smaller tank tendsto have a higher cost per unit volume. Ironically,most poor people buy those because they cannotafford the much larger. but more expensive, tankthat they really require. Second, it can be seen thatalthough ferro-cement and concrete ring tanks areamong the cheapest, they are not significantlycheaper than corrugated iron tanks. Although thepotential durability of the ferro-cement tank isgreater than the corrugated iron tank, this is onlythe case where ferro-cement tanks are properlyconstructed and maintained. Where good workmanship cannot be guaranteed it may be cheaperin the long run to install corrugated iron tanks. Incoastal areas, however, due to the rapid rustingwhich is caused by the salty sea air, ferro-cement

NRF VOL. 9, NO.4, 1985 RAINWATER CATCHMENT IN AFRICA 259350 were built in the space of only 12 months Other tank designs recently developed in Kenyaunder guidance from the local Roman C ~ h o l i c which deserve mention are two varieties of ferroDiocese Development Office. The rapid accep- cement tanks which are built using a weldmeshtance and diffusion of the tanks partly reflects framework. The first of these is a design developedthe ease with which they can be constructed as by a commercial enterprise called 'ferrocraft ' inwell as their low cost. A 4 m:l tank costs around Kilifi. The technique consists of making aUS $100 while a 13 m3 tank (the largest so far cylindrical frame of weldmesh and wrapping thisconstructed) costs US $350. with two layers of chicken wire before applying

The construction of these tanks is accom- mortar to the inside and outside of the framework.plished using two concentric corrugated iron More than 100 tanks of capacity 6 rna have beenring moulds made by simply bolting curved produced in this way and, although technically thesegments together. The larger mould has a tanks are very good and extremely durable, theydiameter of 10-20 em more than the smaller one, are more expensive than the other designsso when the two are centred on top of a precast discussed, selling for about US $350each includingconcrete base there is a space into which the transport to the site and a 2-year guarantee.concrete is poured. This design is particularly A second ferro-cement weldmesh design is suitable for areas where aggregate is locally currently being developed by the African Medicalavailable and can be collected and carried to the Research Foundation (AMREF). This designsite by self-help labour. When the first ring of involves the use of papyrus mat shutteringwhich isconcrete has set, the corrugated iron moulds are placed on the inside of the weldmesh frameworkremoved and placed on top of the previously before it is plastered. When the mortar has set thecompleted concrete ring. This process is shuttering is removed. Preliminary results suggestrepeated until a required number of rings that this is an inexpensive and convenient method.(usually three or four) have been completed. Finally, Nissen-Petersen (1982) in his book onThe tank is plastered inside and out with rain catchment and rural water supply in Africa,concrete and a ferro-cement cover cast for the describes a concrete block tank with a corrugatedtank. iron roof which can be built relatively cheaply ifFerro-cement tanks. The advantage of ferro- the blocks are made on-site using communitycement tanks over the three preceding designs is labour.that extremely large tanks can be built using thetechnique. According to Watt (1978) it ispossible to construct surface ferro-cement tankswith volumes of 400 m3 In Kenya, tanks withvolumes of up to 120 m3 have been built atMutomo. For smaller tanks a single cylindricalcorrugated iron mould is used. This is normallytansported in four segments to the site andbolted together. Chicken wire is coiled aroundthe mould. and loops of fencing wire wrappedaround this to act as reinforcement. Mortar isplastered on the outside of the mould and. whenit is finallv set, and mould is unbolted andremoved. Additional mortar is then plasteredon the inside of the tank wall and the base. Ahighly detailed step-by-step account of thisprocedure is given in Watt (1978). For largertanks a moveable mould can be used as thisallows small portions of the tank wall to beconstructed at a time and also avoids the highcosts of a very large mould.


8/11

Typeo/tank Volume Mean cost Comments(ma) (US SIma)Galvanized oil drum (Kenya) 0.2 100 This is the most common type of storage tanGalvanized oil drum (Botswana) 0.2 112.5-225 used. Although cheap, it is too small for mospurposesCorrugated iron (Kenya) 1 60 Must be transported to site5 26 Not very durable

10 24 Rust easily, especially in marine environmenCorrugated iron (Botswana) 2.25 504.5 359 30Brick/cement (Kenya) 1-1000 40-60 This cost assumes a standard engineering desGhala basket (Kenya) 1-8 17-40 2000-3000 have alreadvbeen construc ted inKenya with varying degrees of successCement jars (Kenya) 1-10 33-50 Larger jars require wire reinforcementFerro-cement (Kenya) 1-200 13-26 This new rapidly expanding technology has tFerro-cement (Botswana) 10 40 advantage of producing relatively large tanksConcrete ring (Kenya) 1-25 23 These tanks are simpler to build than ferro-cementSubsurface ferro-cement (Botswana) 20 20 500 have already been built in Botswana.Although these tanks are cheap. problems owater quality exist

Cost comparison of different types of catchment tanks in Kenya and Botswana(in 1984 US $)

260 H. 1. McPHERSON &1. GOULDTABLE I

NRFVOL. 9, NO.4, 1

and concrete ring tanks are greatly superior tocorrugated iron tanks.

5. EXPERIENCEWITHRAINWATERCATCHMENT

Some broad general izat ions can be made withrespect to experience with rainwater catchments inKenya and Botswana. In both Kenya andBotswana, rainwater catchment has been appliedmainly in the arid and semi-arid areas where thereis an urgent need for water and where surfacewater and groundwater are either scarce ornonexistent. An except ion to this is in WesternKenya, where some small individual roof catchments systems are in operation. However, theseemploy the simplest of storage tanks - oil drums or

corrugated iron tanks.Bacteriological analysis of ground catchmtank water in Botswana showed that in the majoof cases the water has unacceptably high levefaecal coliforms. Faecal coliform counts in ecatchment tanks ranged from a low of 6 to a hig1000. In most tanks the faecal coliform countabove 150. In contrast, analysis of water quali13 roof catchment tanks indicated a zero facoliform count and total coliform count windicated that the water presented no health ri

Some problems were noted with the applicaof rainwater catchment systems in both countThese tended to be technical due to a punderstanding of what was involved in the deand construction of rainwater catchments.main problems identified were the following.


9/11

RAINWATERCATCHMENT INAFRICANRFVOL. 9, NO.4, 1985(1) The storage tanks were too small. This wasusually because the householder or projectorganizer did not know how to estimatestorage needs accurately. In Kenya, forexample, the ghala baskets were often muchtoo small, and to assure an adequate supplyseveral basketswere needed.(2) The tanks leaked or did not holdwater. Therewere several reasons for this. A common faultwas that the ferro-cement tanks were notproperly cured. Ferro-cement tanks need tobe kept moist for several days if they are tocure correctly and retain water. In other casesthe cement mix was not correct and leakageoccurred. Some breakage was also noted inferro-cement tanks built at a central locationand transported to the site.(3) Some tanks were not properly covered. Agood cover is essential to reduce evaporationand to protect the water from contamination.5.1 ADVANTAGESANDDISADVANTAGESOF

R A ~ A T E R C A T C H M E N TRainwater catchment has a number of distinctadvantages, which in some situations make it amore acceptable technology than the othersavailable. The advantages can be summarized asfollows.(1) Rainwater is clean and free of disease-causingpathogens. Roof catchment systems, so longas the tank is covered, provide a source of safewater. Ground catchment systems, if bareground is used as the catchment apron, aremuch more likely to yield contaminatedwater, as was found in the ALDEP groundcatchment tanks in Botswana.(2) With roof catchment systems the water isavailable at the home. It can even be pipeddirectly into the house. In western Kenya,where protected springs are being developedas a source of water, the springs are located atthe bottom of deep valleys and the householders must carry the water up severalhundred meters along a steep path. This ismuch more inconvenient than a roof tankbeside the house. A disadvantge of groundcatchment schemes is that they usually serve anumber of families and some walking isnecessary to collect water.

261(3) With roof catchment, the householder ownsthe system. There are tremendous advantagesto this. The householders is much more likelyto look after the tank and maintain it if itbelongs to him.(4) Operation and maintenance costs are low withroof catchment schemes, and usually theowner can do all the work .himself. Thisremoves the need for a governmentmaintenance programme and all the costsassociated with such a service. Experience inboth groundwater projects and gravity watersystems in Africa has shown that some form ofa government central maintenanceprogramme is usually necessary.(5) The technology is simple and all spare parts ormaterials necessary to make repairs can beobtained locally. This contrasts withhandpump wells programmes. In Kenya, forexample, there are no handpumps madelocally except on an experimental basis.Consequently, getting spares and repairinghandpumps is a serious problem and a number

of handpump wells programmes have failedbecause the handpump broke down and werenever repaired. Gravitywater systems can alsobreak down, and in Kenya there are examplesof schemes which no longer function becauseof burst pipes or poorly designed and builtsystems.These are several disadvantages to rainwatercatchment systems. The prime disadvantage iscost. Storage tanks, as has been seen, areexpensive to build so the intial capital outlay islarge. This is the reaspn why so much experimentation is taking place in East Af:ica to developan inexpensive storage tank. Besides the storagetank, there may be other c'osts depending onwhether the system is a ground, roof or rockcatchment. A roof with adequate guttering mayhave to be built because the existing roof is eitherthatched, too small to provide an adequate supply

or has no guttering. In ground catchment systemsan apron may have to be constructed of concrete orplastic and in rock catchment schemes a pumpmaybe necessary.Another disadvantage with rock and groundcatchment schemes is that the water may requiretreatment before it can safely be drunk.


10/11

262 H. J. McPHERSON & J. GOULD NRFVOL. 9, NO.4,6. CONCLUSIONS

In Kenya and Botswana rainwater catchment ismainly practised in the arid and semi-arid areas,and it is still regarded as a technology to be utilizedif water cannot be obtained in other ways. This isunfortunate, as rainwater is convenient and offersgreat promise in humid areas as well.Significantly, in both Kenya and Botswana all ofth e really large projects are exploiting ground orsurface water sources. Rainwater catchment is thepoor relation, and such schemes as there are arelargely the domain of non-governmentorganizations and are relatively small in scope.

However, a positive sign is that some of themajor rural water supply programmes are startingto include some rainwater catchmentas an element.In Kenya, for example, the Gennan TechnicalAidProgramme is supporting roof catchments in theLake Kenyatta resettlement progect; theEuropean Economic Community is interested in

roof catchments in the Machakos districtFinnida has investigated the possibility ofcatchments in their extensive rural water supprogramme in west Kenya. These are hopsigns' and suggest that in the future rainwcatchments may play a more significant rolsupply desperately needed water.In the next several decades in the developworld, and especially in Africa, we are going tofaced with a dramaticwater crisis. In order to mthis crisis, water will have to be providedcheaply as possible from systems that are goinhave to be largelymaintained by the users.

In those efforts to provide water, rainwshou ld not be neglected. Because of its uniadvantages of technical simplicity and cvenience, rainwater catchment can make a mcontribution to supplying th e water needs ofmpeople.

REFERENCESAlcock, P. G. 1983. A water supply strategy for the Inadi Ward. Vulindulela District, Kwazulu: A researchframework. SubsistanceAgriculture StudyGroup, University ofNatal. 13 pp.Bateman; G. H. 1971. Intermediate technology and ruralwater supplies. In Proc. Conf. RuralWater Supply inEast Africa, 5-8 April 1971, Dar Es Salaam, (Ed. Tschenner), pp. 215-8.Byrne, H. 1983. Clean water in Kola. Waterlines, 1(1), pp. 21-3.Classen, G. 1980. Small scale rural water supply.Ministry of Agriculture, Bag003. Gaborone. 80 pp.Farrar, D. M. 1974. Aspects of water supply and conservation in some parts of Africa. Unpublished Ph.DThesis, University ofManchester. .Farrar, D. M. and Pacey, A. J. 1974. Catchment tanks in southern Botswana: A reviewAfrica Fieldwork andTechnology ReportNo.6, 13 pp. (Available from ITDG. 9 King St, London).Feachem, R. G. eta!' 1978. Water, Health and DevelopmenJ: An Interdisciplinary Evaluation. Tri-Med BooksLondon, 218 pp.Fortmann. L. and Roe, E. 1981. Water use in eastern Botswana: Policy guide and summaryof the water pointsurvey. Ministry ofAgriculture, Gaborone, 96 pp.Gould, J. E. 1984. Rainwater catchment systems in Botswana. present, past and future. Waterlines. 2(4)pp. 14-8.Gould, J. E. 1985. An assessment of rainwater catchment systems in Botswana. Unpublished M.Sc. ThesisUniversity of Alberta, 222 pp.Grover, B. 1971. Harvesting precipitation for community water supplies. Unpublished Report, World Bank110 pp.

Hall, N. 1982. Watercollection from thatch. Waterlines. 1(1), pp. 23-6.IIDG 1969. The introduction of rainwater catchment tanks and micro-irrigation to Botswana. IntermediateTechnology DevelopmentGroup, 9 KingSt, London. 74 pp.IIDG 1971. Report on the state of the Botswana catchment tank project after three years of operation.Intermediate Technology Development Group, 9 King St. London. 14 pp. (Contribution No. 173California Water Resources Centre, University ofCalifomia at Davis. 55 pp.)Latham,B. 1984. Rainwatercollection activities in Indonesia and Thailand. Report toCIDA, Ottawa. 132 pp.Latham. B. and Schiller. E. J. 1984. Rainwater collection systems: A literature review. Civil EngineeringDepartment, University ofOttawa. 32 pp.


11/11

NRFVOL. 9, NO.4, 1985 RAINWATER CATCHMENT IN AFRICA 263Maikano, G. J. and Nyberg, A. 1981. Rainwater catchment in Botswana. InRural Water Supply in DevelopingCountries, Proceedings o f Training Workshop August 5-12, 1980, Zomba, Malawi. Ottawa/IDRC.Publication 167e, pp. 13-7.McPherson, H. J. et af. 1984. Low cost appropriate water and sanitation technologies for Kenya. Report toMinistry ofWater Development, Nairobi, Kenya, Cowater International, Ottawa, 225 pp.Nissen-Petersen, E. 1982. Rain Catchment and Water Supply in Rural Africa. Hodder and Stoughton. 83 pp.Nissen-Petersen, E. 1985. Water from rocks.Waterlines, 3(3), pp. 8-13.Omwenga, J. M. 1984. Rainwater harvesting for domestic water supply in Kisii, Kenya. Unpublished M.Sc.Thesis, Tampere University, Finland. 132 pp.Pacey, A. 1977. Water for the Thousand Millions. Pergamon Press, London. 58 pp.Perrens, S. J. 1975. Collection and storage strategies for domestic rainwater supply. InHydrologic SymposiumProceedings, Armidale, NSW, Australia, pp. 168-72,257-65.Thiadens,A. 1983. Catholic Diocese ofMachakos DevelopmentOffice-Water engineer. Unpublished InternalReport. P.O. Box 640, Kachakos, Kenya, 17 pp. '

UNEP 1983. Rain and Stormwater Han'esting in Rural Areas. Published for UNEP by Tycooly InternationalLtd, Dublin. 238 pp.Watt, S. B. 1978. FerrocementWater Tanksand theirConscmction. ITG Publications, 9KingSt, London. 118 pp.Whiteside, M. 1982. How to build a water catchment tank. Mahalapye Development Trust (Botswana).Government Printers, Gaborone. 82 pp.

rainwater catchment systems in kenya and botswana

Documents