i research into cost effective technologies for … · research into cost effective technologies...

201 92RE

II

Water, Engineeringand Development Centre

IIIIIIIIIII1II

•••• ••NAL R E F E R E N C E G ^ M T ; : T

•••MTV W A T E R oU!*r"L.Y A N O

RESEARCH INTO COST EFFECTIVETECHNOLOGIES FOR WATER SUPPLYIN GUINEA WORM ENDEMIC AREAS

Brian Skinner and Richard Franceys

DRAFT FINAL REPORT

Funded by:UNICEF3 United Nations PlazaNew York, New York 10017USA

WEDCLoughborough University of TechnologyLeicestershire LEU 3TU England

Phone: 0 (44) 509 2226IT , , Q o Oc .Fax: 0(44) 509 211079 t 2 ° l ' ^ K b - !

TIIIII

Brian Skinner and Richard Franceys

III

I

IIII

RESEARCH INTO COST EFFECTIVETECHNOLOGIES FOR WATER SUPPLYIN GUINEA WORM ENDEMIC AREAS

November 1992

Water Engineering and Development Centre,Loughborough University of Technology

I Leicestershire, UK

DRAFT FINAL REPORT

.:.;. J 0 : > • . ; : • : , -: , : . : • • ; > A D T N ,

d . (070) 8 ' 4 i j ! i ax;. 141/142

tor ; V J Q J D \UNICEF '- ,—3 United Nations PlazaNew York

EXECUTIVE SUMMARY

The purpose of this study is to investigatealternative low cost technologies for providingimproved drinking water systems that can beused in isolated communities and areas withdifficult hydrogeological conditions in order toreduce per capita costs to below US$30 percapita. These technologies have to be easilyreplicated on a large scale and have to beprovided as quickly as possible to achieve theeradication of Guinea worm within this decade.

Based on a literature survey and direct reportsfrom projects the study clarifies the choicesavailable for small community rural watersupply. Costs collected during the course of diestudy vary widely even for the same technology,partly because they have been based on widelydifferent assumptions regarding, for example,overheads and amortization but also becausefield conditions vary widely. Very little cost datais available for some of the technologies. Theproduction of 'typical unit costs' for thetechnologies studied has therefore not beenattempted.

However, the results of the study indicate thatappropriate CHOICE between the variousoptions for cost effective technology is moresignificant than the particular cost effectivenessof any individual technology. It is not the role ofthis study to investigate programmeimplementation. However, to achieve costeffective technology the implementation processmust be designed to enable consumers to choosethe technology which they find to be mostsuitable.

• None of the technologies investigated willbe effective without adequate consumerinvolvement.

• For sponsors who desire sustainable benefitsas rapidly as possible the only solution is todramatically increase their spending oninformation, education and communication(marketing and selling) in order toencourage the necessary changes intraditional water drawing practices and diefast take-up of the consumer's choice ofimprovement.

The transmission of Guinea worm disease canbe prevented by interventions which do nothingto improve the usual poor bacteriological qualityof the water. Such faecaily contaminated wateris contributing to many odier diseases, and evendeaths in the community.

The report includes sections that givesuggestions for interventions which reduce thetransmission of Guinea worm disease withoutmuch, or any improvement in the bacteriologicalquality of the water. The audiors are of theopinion that while in the very short termmethods which only strain out, or kill, die waterfleas (copepods) which carry the disease may beneeded the provision of a potable water sourcesshould be an urgent priority. It should be notedthat although the straining and use of insecticidemay lead to the eventual elimination of Guineaworm disease these interventions will have to befollowed at some stage by the provision of apotable water supply if the level of health in thecommunity is to rise appreciably. If a potablewater source can be quickly supplied then thecosts of many of the straining and insecticideinterventions can be saved.

In the report die water supply technologies havebeen considered in diree categories. These areconsidered in the order of preference forconveniently supplying sufficient quantities of agood quality of potable water, namely:

GroundwaterRainwaterSurface water

Groundwater exploitation has already playedan important part in the eradication of Guineaworm disease from a number of places in theworld. Groundwater has two main advantages:

• it is usually potable and tree of anycopepods to be infected widi Guinea wormlarvae. It does not usually contain faecalpathogens or other contaminants;

• aquifers are 01 ten extensive and in suchcases it is usually possible to locate a wellvery near to a community needing water.

The mam disadvantages of exploitinguroundwater are:

IIIIIIIIIIIIIIIIIII

IIIIIIIIIIIIIIIIIII1

• it can be difficult to excavate the holenecessary to reach the ground water. Thisproblem increases as the depth to thegroundwater increases and with thehardness of the sub-strata.

• except in the case of springs and artesianwells, some way has to be found to lift thegroundwater up to ground level. Thisproblem obviously increases with the heightof lift required, and extraction using ahandpump is increasingly difficult as thegroundwater depth increases. Fewhandpumps are acceptable to communitieswhere the lift is beyond about 45 metresbelow the surface and because of theincreased forces on the components of suchpumps they need to be very robust.

The most important observations andrecommendations to arise from the section are:

• Where springs occur full use should bemade of them. When they are properlyprotected they can supply potablegroundwater, in most cases without theneed for a pumping device. Cost effectivespring protection designs which do not needspring boxes are available.

• Appropriate groundwater locationtechniques should always be used to avoidas much as possible the cost of wasted bore-holes or wells;

• Hand-drilled boreholes have great potentialwhere the strata is suitable. They are oftenmore cost effective than hand dug wells butunlike draw wells they need sustainablehandpumps for the withdrawal of water.The boring equipment is relatively cheapand usually very portable, and thecommunity can be fully involved in its usefor borehole construction.

• In a limited category of strata the use ofwash-bored tubewells is very cost effectiveand it can lead to the very rapidconstruction of tubewells.

• In many cases low yielding strata,traditionally not considered suitable lorcroundwater exploitation, can be used to

supply water to a community. This isespecially true when hand dug wells areconstructed but very low yielding aquiferscan be exploited by a tubewell if it isdesigned to serve only a small population,such as would be the case when using theBlair bucket pump.

• Where the groundwater is not excessivelydeep, open hand dug wells have greatpotential and such a method of exploitinggroundwater makes maximum use of localmaterials and community participation.Where unprotected wells are already beingused by a community the transmission ofGuinea worm larvae is a risk, and properprotection of those wells, or construction ofnew wells, is the obvious choice oftechnology. The use of a bucket andwindlass, or a 'shaduf reduces the risk ofwater pollution by users.

• Where a sustainable handpump acceptableto the community is available, hand dugwells should be covered and water should bewithdrawn using the handpump becausethis reduces the opportunities for the wellwater to become contaminated. Howeverfacilities for water to be withdrawn from thewell by bucket should also be provided foruse in an emergency. The simple, easilymaintained, direct action type of handpumphas great potential for most hand dug wellsince the pumping lift usually does notexceed 15m. Where the community beingserved by the well is less than 60 the Blairbucket pump is an alternative which shouldbe considered.

• Where there are more than one protectedsource in a community, such that the veryoccasional failure of one handpump on onesource would not mean people had to returnto polluted traditional sources, the 'casedwell' design or 'partially lined well' designshould be considered since these useconsiderably less materials than fully linedwells. Both types of wells are suitable forhandpumps or the Blair bucket pump. The'Modified Chicago Method' of temporarysupport can be very cost effective in theconstruction of such wells.

Powered drilling rigs for constructingboreholes need to be chosen with care tofind a type most appropriate to the fieldconditions. Relatively simple, lightweight,cheaper rigs are now available which havethe potential for much cheaper unitborehole costs. Their manoeuvrability oftenmeans that boreholes can be drilled inremote places inaccessible to the heavierrigs.

There are a vast number of differenthandpumps available on the market but alsoa number of sources of useful advice and theresults of field tests to enable a wise choiceto be made for particular field conditions.Undoubtedly the suitability of a pump forvillage level operation and maintenance(VLOM) is very important. Of equalimportance are the standardisation in anyone country on just a few models ofhandpump, and the in-country manufactureof the pumps, or at least the commonlyneeded spares.

The use of plastics for handpumpcomponents in corrosive groundwatersoffers many advantages and is very costeffective compared to the alternative usestainless steel. The successful use of plasticrising mains, particularly with open tophandpump cylinders has already beenproven in the field, and current researchshould shortly produce improved connectorswhich will allow the pipe sections to beeasily separated when necessary. Directaction handpumps which often have onlyplastic components below ground aresuitable for use in virtually all situationswhere the water lift does not exceed 15m.The use of stainless steel rods and plasticrising mains with deepwell pumps is alsonow well tried.

The simple, relatively inexpensive andeasily maintainable Blair bucket pump isworthy of consideration where the numberof people relying on a source is less man 60and the lift is less than 15m. The low dailydemand from such a group may extendfurther the type of strata which can beconsidered suitable for yieldinggroundwater. If due to the convenientlocation of the borehole people are willing

to spread their demand throughout the day,avoiding sustained peak use, the pumpcould be used on a very slow yieldingborehole.

• The use of a shaduf with the Blair bucket ispotentially very cost effective but to date hasnot been tested in the field.

• Where the aquifer appears to be suitable fortheir use, the potential for directly installingthe types of open top cylinder handpumpswhich have extractable foot valves shouldbe investigated. Such a method ofinstallation dispenses with the need for alarge diameter borehole casing and cantherefore be very cost effective.

• There is a wide variation in the way inwhich different projects calculate unit costs,and caution is needed when consideringcosts in isolation from full details of theirbasis.

Rainwater is pure, and if collected from anelevated surface such as a roof and stored in anhygienic manner it is usually suitable fordrinking without treatment and offers no risks ofthe transmission of Guinea worm disease.

• The main disadvantage of its use is thatthere is usually a need to store a largevolume of it for consumption during periodswhen there is no rain. If only water fordrinking purposes is stored and traditionalsources are used for other purposes thevolume of storage is greatly reduced.

• If rainwater is captured from the roof of theuser's house there is the advantage that thestorage can be placed conveniently close tothe point of use and if the storage is aboveground, the water can be collectedhygienically from a tap or hose.

• Where roofs to buildings are not suitable,cheap temporary catchments of wovenplastic sacks may be feasible.

• The annual rainfall figures for many of theareas of countries in West Africa whereGuinea worm is endemic are high enough


TIIIIIIIIIIIIIIIIII1

to suggest that rainwater catchment is apotential source of potable water.

• Simple cement jars with hygienic draw-offhoses can provide cost effective potablewater supply.

• Where the materials are available in thecountry the use of galvanised iron sheetsand galvanised wires for the DANIDAdesign of suspended gutters is likely to becost effective.

Surface water is usually of poor bacteriologicalquality and whenever possible an interventionwhich removes the risk of faecal pathogens aswell as infected copepods should be used.

Of the interventions which mainly address onlythe problem of copepods the most effective relateto the separation of people from the source.

• There is a need for physical barriers and theacceptance by everyone of a communityrule that no one enters the water but that itis drawn only from a dedicated point.

• There are three preferred methods ofdrawing water at this 'dedicated' point,described in order of preference. The firstcomprises a sustainable handpump andfloating intake. The handpump can eitherbe directly connected to the floating intakeor it can draw water from a well-likereservoir in one of the banks which receiveswater from the intake.

• Secondly the use of a shaduf or other bucketand rope method from a vertical abutmentwith a surface which drains away from thepond is recommended.

• A final option is the use of a hand heldbucket dipped into the water from aplatform or ramp close to the surface of thesource.

• In order to give a good level of protectionagainst copepods the three recommendedoptions can be modified by the use ofstrainer material. Either the floating intake-can be protected by a box of strainingmesh, or, adjacent to the access platform, an

open topped strainer box can be used. Thisbox floats in the water with straining meshon its sides and base so that water drawnfrom within the box when users dip theirbuckets into it is automatically strained(Neither of these two suggestions are knownto have been tried so they both need to befield tested to discover how effective andsustainable they are.)

• Where necessary and feasible chemicalinsecticide (Abate) can be used if itsapplication is properly managed

Of the interventions which provide copepod freewater of reasonable or good bacteriologicalquality, depending on the field conditions, thefollowing methods are recommended.

• Convert stepwells to protected draw wells orto wells equipped with a handpump (if asustainable handpump is available).

• Collect water which is naturally infiltratingnear to water sources or collect water fromman made infiltration systems.

The interventions available after the water hasbeen collected include straining, storage andtreatment.

• The straining of water before consumptionis recommended if other water treatmentmethods which remove most bacteria aswell as copepods are not feasible. Thisstraining is particularly appropriate wherepeople need to consume water away fromtheir usual protected source.

• Settlement, long term storage and strainingusing vessels in the home can lead to agreatly improved bacteriological quality tothe water and this should be field tested.Cement mortar jars are suggested to beideal containers and they could be sold withpurpose made strainers, and taps or pipesfor hygienic withdrawal of water.

• The use of locally available coagulantscould be promoted to lead to an even betterquality of water from the system justmentioned but this needs further study inthe field.

• It is considered doubtful whether domesticor pond side slow sand filters can beproperly sustained to produce water of goodquality but their use would lead to someimprovement in the quality of the water.

• Solar disinfection does work during periodsof bright sunshine and in the absence ofother methods of improving thebacteriological quality of the water it couldbe field tested to see how it works out inpractice

• Chemical disinfection at a users home is notconsidered to be sustainable.

Other points regarding surface water include:

• Improvements to the ground levelcatchments areas which contribute to rainfed ponds will lead to a better quality ofwater in the pond. These improvementsinclude fencing, interception and diversionof surface water from outside the catchment,settlement and the use of gravel filters.

• Where sustainable suction pumps are notavailable it may be possible to convert directaction and deepwell pumps to allow them tobe used as offset pumps for drawing waterfrom floating intakes or bed infiltrationsystems.

• Where man made infiltration systems are tobe used in static water bodies, or where slowsand filtration systems are to be used, it isrecommended that the feasibility of usingone or more mats of open weave filteringfabric on the surface of the filter isinvestigated. This has the potential tosimplify the cleaning process and can leadto improved lengths of filter run and ashorter 'ripening' period after filtercleaning.

The Conclusions And Recommendations of thestudy are that there is a large range of cost-effective technologies available for water supplyin Guinea worm endemic areas.

For the sake of long-term consumerunderstanding and health promotion it isrecommended that wherever possible assistance

should be aimed at providing potable drinkingwater, not just water free from infective Guineaworm larvae.

The results of the research indicate that the issueis not primarily the determination of a cost-effective technology but rather the choice oftechnology sui table for a particular location.This is of vital importance to achieve lifecyclecost effective technologies which are necessaryto ensure sustainability.

Locations differ according to socio-economic,hydro-geological and hydrological conditions.To obtain the most suitable choice thecommunny/household/consumer need to beinvolved in an education, information andcommunication process that will enable users tomake their own choice according to theirunderstanding of improvements in health andconvenience and according to their willingnessto pay for sustaining those improvements.

The desired rate of implementation can beachieved only by innovative use of promotionand marketing campaigns not normallyassociated with 'civil engineering products'.

Cost effectiveness of implementation is likely tobe enhanced by the encouraged involvement ofprivate sector suppliers.

The starting point for upgrading to potablewater free of Guinea worm is normally toimprove existing water sources, though theprotection of potable water is difficult withsurface water sources.

Life-cycle cost effective technologychoices: Groundwater

When making choices, the exploitation ofgroundwater is normally the first priority sinceit can often supply sufficient quantities of goodquality water.

Within this category every effort should be madeto cap any spring sources (and where possible topipe the water to a number of more convenientcollection points)

If there are no springs and communities arescattered but groundwater is relatively shallow,

IIIIIIIIIIIIIIIIIIII

TIIIIIIIIIIIIIIIIII

a number of hand drilled/hand dug wells withBlair type bucket pumps should be promoted.Where population density is higher the sametechnologies can be used with direct actionhandpumps.

If the ground conditions make it necessary,powered mini-rigs are applicable with deepwellopen top cylinder pumps for water collectionfrom beyond 15m depth.

Life-cycle cost effective technologychoices:Rainwater

Where groundwater cannot be exploited,rainwater catchment using roof catchments withsupported "V1 gutters and cement mortar jars canbe promoted to give a good quality of potablewater. Nylon sack catchments can be used wherethere are only thatched roofs.

Life-cycle cost effective technologychoices:Surface water

Where there is no alternative but to use existingsurface water sources, strainer boxes can be usedto prevent Guinea worm as an interim measureeven where there are no pumps.

The next stage of improvement for surface wateris to use hand powered suction pumps withfloating intakes protected by strainers.

Neither of these methods remove pathogenicbacteria so they should be promoted inconjunction with household cement jars forstorage (with strainer covers as an addedprecaution). Any length of storage will begin toreduce the risk from pathogens not removed bystraining.

Surface water sources should be upgraded toinfiltration galleries or river bank wells toensure potable water at the earliest opportunity.

Portable, individual strainers have a vital role toplay as back-up to the other methods ofpreventing the ingestion of Guinea worm larvaeand these can be used by people moving awayfrom protected household sources.

CONTENTS

1. INTRODUCTION

Terms of Reference of StudyWork carried outIntroduction to Guinea worm

2. DIFFERENT WATER SOURCES ANDTHE RISKS OF GUINEA WORMTRANSMISSIONConsideration of various water sources and theassociated risks of Guinea Worm transmission

Recognition of water supply technologies thatalways give copepod free water groundwaterfrom protected sources rainwater from protectedcatchments stored in a protected tank

3. COST EFFECTIVE GROUNDWATEREXPLOITATIONConsideration of cost reductions in exploitationof groundwater:•springs•hand dug wells•lower cost boreholes•jetting•hand augering•appropriately sized rigs•rising mains as permanent borehole casing•lower cost handpumps•plastic rather than stainless steel in aggressivewater•direct action rather than conventional•Blair 'bucket pump', use with shaduf

4. COST EFFECTIVE RAINWATEREXPLOITATIONConsideration of cost-effective technologies forrainwater catchment•community systems or household systems•use only for drinking water•roof catchments•temporary catchment surfaces (e.g. stretchedwoven polypropylene sacking)•mortar jars•wire reinforced mortar•ferrocement tanks•ground level catchments

5. COST EFFECTIVE IMPROVEMENTSTO SURFACE WATER AND

UNPROTECTED GROUNDWATERSOURCESConsideration of the various technologies andcommunity practices which can be used toimprove the water from existing sources•community rules to only allow uninfectedpeople into the water•physical barriers or methods of collectionwhich prevent, or make it unnecessary foranyone to enter the water•fences•platforms•conversion of step wells to draw wells•water lifting devices•straining out the copepods• at the source(s)• at home•chemical treatment• of the source(s)• at home•fish as copepod predators•boiling•infiltration systems• protected wells in river/pond banks• infiltration systems in river/pond beds• groundwater dams and sand-storage dams• infiltration using suction pumps (e.g. SWSfilter)•use of geotextiles in pond infiltration systems•filtration systems• slow sand filter near to source• domestic sand filters• use of geotextiles•long term storage•solar disinfection

6. CONSIDERING THE COMMUNITYConsideration of how the community's viewsaffect the adoption of improvements and newtechnologies

7. CONCLUSION ANDRECOMMENDATIONSRecommendations of cost-effective technologiesto be used on a large scale in selected countries

For die sake of those who do not have abackground in the subject the following sectionsof the report includes some basic backgroundinformation about the occurrence andexploitation of water sources which will be wellknown to other readers.


TIIII

i

1. INTRODUCTION

1.1 Background To The Report

This report has been carried out under terms ofreference drawn up by the Water and SanitationSection, UN1CEF, New York (Appendix I )

The terms of reference brought particularattention to the need

'to investigate alternative low cost technologies[for providing improved drinking water sources]that can be used in isolated communities andareas with difficult hydrogeological conditionsin order to reduce per capita costs which shouldnot exceed USS30 per capita (low cost ruralwater supply systems at present average USS20per capita). These technologies have to be easilyreplicated on a large scale and, if [Guineaworm] eradication is to be achieved this decade,appropriate low cost water supply services haveto be provided as quickly as possible'

The terms of reference give the overall purposeof the report as being

To investigate and propose low cost effectivemethods to provide clean drinking water toguinea worm endemic areas in Africa. Theresults of the findings will be used to developsuitable projects in a selected number of guineaworm endemic countries'

The terms of reference were slightly amendedduring the writing of the report, particularly inthe light of the discussions at the TechnicalSupport Team tor Dracunculiasis meeting inAnnecy, France, 24 - 28 August 1992. Theseamendments include the omission of a study ofsolar powered systems and the omission of (liedetailed project proposal for follow up action tofield test the different technical options (Section3.3 of the terms of reference).

A detailed section on low cost methods ofexploiting groundwater, which was not includedin the areas of technologies to be considered inthe terms, has now been included in die report.The date of submission of the report was alsochanged although the period of the consultancyremained at seven weeks.

Methodology

Guidelines on the suggested methodology for thepreparation of the report were given in Section 4of the terms of reference have generally beenfollowed except that no visit was made to theInternational Reference Centre, The Hague.This was considered unnecessary since literaturereferences were obtained from a search of theircomputerised data base.

In addition to obtaining information fromGlobal 2000, Peace Corps, CDC, USAID (Waterand Sanitation for Health Project) which weresuggested in the terms of reference the authorshave consulted WaterAid, SKAT, ODA, variousUN1CEF staff and a number of manufacturersand researchers. The help of all these people ismuch appreciated and sources of informationobtained from them are referenced in die body ofthe report.

A postal survey was also carried out in WestAfrica among interested parties but full resultsare not yet available.

Problems Experienced Obtaining RelevantCost InformationOne of the biggest problems experienced duringthe short period allowed for the production ofthis report has been die difficulty experienced inobtaining suitable cost data for the technologiesbeing investigated. Unfortunately this meansthat it has not been possible to give indicativecosts for some of them although they appear tobe low cost technologies.

However there is a danger in being too focusedon a typical cost per capita1 figure for any of thetechnologies because in practice these may varyquite widely between programmes and will;dmost certainly vary between countries. Thesevariations are due to a number of factorsincluding the specific site conditions andmethod of construction, the availability ofmaterials and skilled labour, and the numberand density of distribution of sites in any oneprogramme. Such variations can be noted in thewide range of costs which have been obtainedfrom different sources for the drilling ofboreholes, and which are quoted in Section .With some of these costs and those for some ofthe other technologies it is often not clear onwhat basis the original source calculated thefigure. For example different projects are likelyto be using different wavs of:

distributing programme overheads includingcosts of expatriates;

the costs of other very important components ofwater supply projects such as health educationand community mobilisation;

allowing for the 'cost1 of donated materials andlabour;

allowing for the amortisation of plant andvehicle costs;

Such problems have been recognised for sometime and global standardisation of costingprocedures to overcome many of thesedifferences is being addressed by the Water andEnvironmental Sanitation Section at UNICEFNew York. However the WESCOST computerprogramme recently developed to allow realisticcomparisons of unit and per capita costs has notyet been fully field tested and no comparativecosts for any of the technologies examined inthis report are currently available.

In the absence of suitable cost data for a countrythe only option is for a programme manager towork with some communities to set up somesmall pilot projects to test out the technologies ata few sites to establish:

some idea of the suitability of the technology for(lie field conditions;

the likely rates of completion;

die level of health education and communitymobilisation needed;

the requirements for labour, transport andmaterials;

the logistics of managing the programme;

the unit costs and per capita costs;

the community's acceptance of the technologyand their motivation to contribute labour andmaterials or money to the construction of aprotected water source.

Fortunately the capital cost of the pilot testing ofthe technologies discussed in tins report will not

be very high since the equipment necessary isusually not very expensive and in the case of thesimple powered drilling rigs is only moderatelyexpensive.

Life Cycle CostsThe USS30 and USS20 per capita figures quotedin the terms of reference are the initialinvestment costs for water supply technologies.The initial investment costs are only one aspectof the costs of a technology. When making aneconomic choice between different technologiesit is also important that the annual running costsand the replacement costs are also consideredespecially if the community is expected to bearthese costs. Here again the programme managerwill often meet the problem of the usual lack ofinformation about life expectancies andmaintenance costs for much of the hardwareused with low cost water supply technologiesand suitable discount rates to apply to thecalculations, but some attempt needs to be madeto consider this aspect. Unfortunately there hasnot been time nor the raw data to allow this tobe attempted in this report.

Experience From Existing ProjectsFor some of the technologies it may be possibleat an early stage to liaise with programmeselsewhere in the world who have already usedthem. This will enable managers, healtheducators and technologists to gain useful ideaswhich have already been learned by others, andhelp them to avoid mistakes made by others.Training of key workers for a new project bythose working in an existing project will also bevaluable where a new technology is beingintroduced to an area.

1.2 The Importance Of Potable Water

The transmission of Guinea worm disease canbe prevented by interventions which do nothingto improve the usual poor bacteriological qualityof the water. Such faecally contaminated wateris contributing to many oilier diseases, and evendeaths in the community.

The report includes sections that givesuggestions for interventions which reduce thetransmission of Guinea worm disease withoutmuch, or any improvement in the bacteriologicalquality of the water. The authors are of theopinion that while in the very short term



methods which only strain out, or kill, the waterfleas (copepods) which carry the disease may beneeded the provision of a potable water sourcesshould be an urgent priority. It should be notedthat although the straining and use of insecticidemay lead to the eventual elimination of Guineaworm disease these interventions will have to befollowed at some stage by the provision of apotable water supply if the level of health in thecommunity is to rise appreciably. If a potablewater source can be quickly supplied then thecosts of many of the straining and insecticideinterventions can be saved.

Health education is of course needed with both atemporary intervention and with the morepermanent intervention of a clean water supply.However in the case of straining and insecticideinterventions the community can not be taughtto drink water from a potable source becausethere is none. Some difficulties maysubsequently arise if at a later stage if forexample they are told by health educators thatthe strained water from the pond which theywere at first encouraged to drink is not after allgood for them but that they should instead gettheir drinking water from a new potable source.In some parts of India women have beenobserved straining borehole water through cloth,a habit taught to them to strain out copepodswhen drawing water from the old stepwells.There are no copepods in the groundwater fromthe handpump so straining is unnecessary, andindeed it can be detrimental to the quality of thewater if the cloth is not kept clean.

Previous Work On Cost EffectivenessThree previous reports on cost effectiveness andcost-benefit of Guinea worm eradicationinterventions were consulted during thepreparation of this report (Paul et al (1986),(Paul (1988a), Paul (1988b)). One report whichcompares costs of rainwater harvesting in fiveAfrican countries without specific reference toGuinea worm was also consulted (Lee &Visscher(1990)).

Pauls work is based on the development of a PC-based spreadsheet software which was fieldtested in Pakistan. It is reported that it can beadapted by a computer programmer to suit theparticular conditions of any country. The cost-benefit analysis (CB A) compares the full costs ofa method of intervention (over a period of years

for a certain defined disease prevalence) withthe value of the expected benefit. The latter isbased on the product of the number of dayspeople would have previously been out of actionbecause of Guinea worm disease and the percapita per day agricultural productivity figure.However Paul notes that other benefits such asincreased school attendance will also result fromthe eradication of the disease.

Comparison between the theoretical benefit-costratios calculated by the software for differentinterventions can aid a programme manager tochoose between them. Sensitivity analysis allowsthe effect of possible errors in the assumptionsmade when inputting data to be examined.

The study in Pakistan describes the assumptionswhich had to be made and qualifies the results.Tliese show that for the three different provincesstudied the BCR for various differentinterventions varied from 0.48 to 1.53. Howeverimproved drinking water systems onlycomprised a low proportion (0 - 10%) of theinterventions being studied in any province andin each case only 40% of the cost of the watersystem was used in the CBA. The 40% factorwas used the improved water intervention hasmany benefits in addition to the eradication ofGuinea worm disease whereas the otherinterventions (the use of straining mesh andchemical insecticide) related purely to theelimination of the disease.

Lee and Visscher (199(1), when looking atBotswana, Kenya, Mali. Tanzania and Togocalculate the annual equivalent cost (AEC) perm^ supplied from rainwater catchment systemsand water from some omer sources. Lack ofspecific data meant that some very generalisedassumptions had to be made with regard to theamount of water used from each type of system,their life span and their running cost. Their costcomparisons between countries demonstrate awide range of costs, probably partly due to theproblems of differences in the costing basis usedby the sources from which they obtained theirraw data. When comparing costs in just onecountry they conclude that in Botswanarainwater catchment can often compete withother sources, and in Tanzania its cost iscomparable to diesel powered borehole supplysvstems.

1.3 Guinea Worm Disease

BackgroundGuinea worm disease (dracunculiasis) currentlyaffects an estimated 3 to 5 million people peryear (CDC 1991), mostly from seven Africancountries (Figure 1.1) but also in some pans ofIndia and Pakistan. It is caused by a parasiticworm dracunculus medinensis which is carriedin a cyclopoid copepod (a small water flea 1-3mm long sometimes referred to a a cyclop) andit can only infect a person if they drink watercontaining a copepod which is harbouring theinfective stage of the parasite.

People rarely die from the disease but the fullygrown worm incapacitates its victims for four tosix weeks, or sometimes longer. This means thatinfected men and women cannot work in thefield, infected children cannot attend school andinfected mothers can not care properly for younginfants. Earnings, learning and good nutritioncan therefore all be undermined by the disease.

There is no vaccine that can prevent the disease.It can not be treated in a person by use of a drugalthough recently it has been discovered thatsome relief can result from the early removal of,the live worm by making a small incision andcarefully removing it from below the patientsskin.

The disease in theory can be prevented relativelyeasily by interventions that break the yearlycycle of its transmission. All of theseinterventions, other than surgical removal, relateto the source of water. This report considers allthe different forms of intervention either at thewater source or after the infected water has beencarried home.

The cycle of transmissionDuring the one-year incubation or growingperiod in the human host, the adult femaleworm moves to a position under the skin of theafflicted person. Then the parasite causes apainful blister to form, usually on the lower legor foot. When the person immerses the affectedbody part in water, the blister breaks, and theworm is exposed releasing hundreds ofthousands of tiny first-stage larvae into thewater. The adult female worm is capable ofreleasing larvae on repealed exposures to waier.

Some of the larvae deposited in the water areingested by the copepods where they live anddevelop into third-stage larvae after 10-14 days.Only these third-stage larvae are infective topeople.

After people drink water containing infectedcopepods, gastric juices in the stomach kill thecopepods, allowing the infective larvae toescape. These larvae migrate to the smallintestine, penetrate through the intestinal walland live in the abdomen. Male and femalelarvae reach maturity after about 90-120 days,when mating occurs. Thereafter, the femalecontinues to grow into an adult worm. Duringthis time the adult female moves towards thelower limbs and emerges after about 10-14months.

It usually takes several weeks for the afflictedperson to completely extract the worm. Duringthis time the person is disabled or in pain, oftenfrom infection resulting from the worm as itemerges from an abscess or from inflammationof the joints. Secondary bacterial infections arecommon and usually prolong and complicaterecovery. Tetanus can develop, as well as frozenjoints and permanent crippling. The worms donot survive in people for more than one year.They eimer surface through the skin and areextracted, or die inside the body.' (CDC 1991)

No immunity to dracunculiasis develops andrepeat infections are common. No animals otherthan man have been shown to be importantnatural hosts of the Guinea worm althoughmany can be experimentally infected in thelaboratory (Muller, 1985).

Hie life cycle of the Guinea worm is wellillustrated in Figure 1.2.

Interventions in the cycle of transmissionThe six main points of intervention in the cycleare illustrated in Figure 1.3. These are discussedin later sections of the report. Briefly the are:

providing water which can not sustaincopepods. For example this could be a newprotected groundwater source.

Avoiding the contamination of the water byGuinea worm larvae. Iliis can be achieved onlyit the whole community change those practices

T

1

Life Cycle of the Guinea Worm

7. In tvw> w**tt» ttielarva* uno*rqo twomolt* wimin in« wanlisa to Mcomtthira-stag* larva*,which can nrecthumans.

6. Watar (laaconsumesworm larvas,wtiicn rasittdigestion.

S. On contact withwater m« amergngworm releasesimmatura< first-Mao*)larvae r?o the watersource onen a pond orshallow wwi. A(receiving larvasurvives ffirea aavsunless it finas a nosL

1. Person drinks wall orpond watar containing

water flaaa (Cydoos) matara mfactad with mature

(thiro-*tat»| worm larvaa.

2. Gastnejuioas in tha

human *tomacndigast tna watar

tlaaa. Wormlan/aaare

retaaaaaandmovaio

abdoniinaltiiauaa. wnarethay grow and

mat*,

3. Fartlizad iamal**worms rmgrata to

vanous Coov ragnns.usually th* lowaf dmo».

(MaM* dM Mom altarmating).

Source: 1992 Medical and Health Annual © Encyclopaedia Bntannica,illustration oy John Draves

A y»ar attar intactiontn* wonn twain* to

amerg* through th*sKm at tha sit* or a

•ainlul blister.Inc.;

I o D. medinensis Life CyclePoints of Intervention Against Dracunculiasis

Person with, emergentQ. medinenaia ' Provlda safe water source

Apply antiseptic and/orocclusiva Pannage, anaeOucate person not tocontact drink int water source

Ingestlon ofInfective copeoods

educate individualsto •=><•<• a m or boiJall drinking water

Contiainatlon ofdrinking water source

Treat water sourcewith chemicals to kill copepods

maainanala larva* lngaatadby coDepods

which previously led to them coming intocontact with the source water and therebycontaminating it. It needs one person carrying aGuinea worm lesion which comes into contactwith the source to infect a large number ofcopepods and restart the disease cycle. Howeversuch measures are better than nothing.Occlusive bandaging can also reduce the risk oflarvae from an infected person reaching thewater.

Treating the water source with chemicals. Theuse of chemicals, particularly temephos (tradename Abate) to kill all the copepods and larvaein the source water is a widely used intervention.However achieving the required concentration ofTemephos in the source at the regular intervalsneeded during the transmission season is notvery easy. Where the source is very large, orwhere there are a large number of small sourcestreatment by this method often becomesimpractical

Straining the copepods out of the water sourceas it is collected. If a well on the bank of a pondcollects water from an infiltration system theinfected copepods (and many other pathogens)will be strained out of the water before it entersthe well so there is no risk of people ingestinginfected copepods. Likewise if the water passesthrough a fine strainer before it is collected froma source then users are protected from Guineaworm disease (although they may ingest anumber of other pathogens).

Straining the copepods out of the water at homeor killing them by boiling drinking water. Ifpeople always strain their drinking waterthrough a finely woven cloth or through asuitably fine mesh the copepods (although veryfew of the other pathogens) will be removedfrom the water and those who drink the waterwill be protected from infection. They can alsokill the copepods (and other pathogens) byboiling all drinking water.

Removal of adult worms just before theyemerge. As mentioned above the practicality ofthis is only a recent discovery and it is not yetwidely practised and it has been added to Figure1.3 by the present authors. Obviously if theworm is removed before it can release larvaethere is no risk of the copepods ever becominginfected bv that worm.

Important aspects about the life of copepodsGuinea worm disease is a seasonal disease andthe implications of this on various interventionsis discussed further in parts of Section 4 and 5.The seasonal nature is partly due to the waterdrawing habits of users which vary from seasonto season but also in part due to the conditionsunder which the copepods survive. Their biologyhas been studied by a number of researchersparticularly by Onabamiro in Nigeria and byMuller and some aspects of this are relevant tothis report.

Muller (1971) has noted that a larva usuallymoults a second time 12-14 days after beingingested by a copepod if the water temperatureof the pond water is between 25°- 30°C. A fewdays after this moulting the third stage larvaebecomes infective to a person who ingests thecopepod and larva. Apparently outside of thisrange of temperatures the time required formoulting lengthens and the survival rate ofinfected copepods falls off radically (de Rooy,1992). In many areas of Nigeria the harmattanwinds in January - February cool off the pondsand virtually stop transmission (cit.ibid).

As discussed later in this report Muller (1985)has also noted that copepods do not usuallycolonise draw wells provided the diameter is lessthan about 4 metres. This is probably due to thelack of enough light for the growth of the algaeand protozoa on which they feed. Copepods arenot usually found in fast flowing water.

He also notes that copepods which contain third-stage larvae become sluggish and sink nearly tothe bottom of a pond or step-well (Muller,1985). The implications of this for on the way inwhich water is collected from a source isdiscussed in Section 5.

Quality and quantity of water resulting froman interventionAs noted above, a number of the interventionswhich can be used to help eradicatedracunculiasis do nothing to improve thebacteriological quality of the water and althoughafter the intervention the water will not harbourinfected copepods it will most likely contain anumber of other pathogens harmful to those whoconsume the water. Wherever possible anintervention is chosen which will remove or

IIIIIIIIIIIIIIIIIII

certainly drastically reduce these otherpathogens so that instead of only eradicatingGuinea worm disease other illnesses in thecommunity can also be reduced. However, asdiscussed in Section 2, it is recognised that thefaecal pathogens found in water (which are thesource of many common illnesses in developingcountries) are also ingested via other routes soonly addressing the quality of the water is not asufficient intervention to lead to a drasticreduction in many of these other disease.

If the intervention chosen for the eradication ofGuinea worm also leads to a more convenientsource of a greater quantity of water than beforethen considerable health benefits will result eventhough the water is still to some degreebacteriologically contaminated. This is becausea higher quantity of water usually leads to ahigher standard of hygiene which eliminatesmany of the other routes of infection by faecalpathogens (Feachem, 1980). The avoidance ofthe faecal contamination of potable water in ahome is discussed in Section 2.

Convenience of new protected water sourcesOne area that needs careful consideration whenproviding a new or improved source of water toprevent people becoming infected with Guineaworm is the convenience of the new source orthe method of collecting water. Unless extensiveappropriate health education leads to a changein attitudes people will use unprotectedtraditional sources if they are more convenientthan the new source. If the traditional source canbe protected in a way acceptable to thecommunity then this becomes less of a problemand the improvements at traditional watercollection sites is often die best starting point foran intervention.

With scattered communities which are usingnumerous traditional sources the centralisedprovision of an improved source may thereforenot be a solution to the problem of providingcopepod-free water. The provision of a greaternumber of smaller convenient sources (such asshallow hand augered boreholes or shallowwells with simple water lifting devices, ordomesuc rainwater catchment systems) shouldbe seriously considered in these situations, evenif they can only be schemes suitable forproviding just the potable water component ofdomestic demand.

The problems of people who need to consumewater whilst away from home and away fromany protected sources provided for domestic useneeds to be addressed. This applies equally topeople who are away for a few hours as to thosewho seasonally migrate, for example to work indistant fields. This problem can only beaddressed by appropriate health education tomake people aware of the risks of infection fromall unprotected sources to encourage diem wherepossible to carry potable water from home to thefield or to take small straining devices to at leastprotect diemselves from dracunculiasis. Where anumber of fanners recognise the importance ofpotable water it may eventually be possible toconstruct small protected sources in the fields.People visiting friends or markets away fromhome need to understand diat they must at allcost avoid consuming unfiltered water. If peopledo not understand diese risks they may becomeinfected away from home and come to theconclusion mat the interventions carried out athome, or in meir villages, do not work and diatany effort necessary to make die interventionswork may then be considered useless.

It is important mat before any intervention takesplace that die programme designers are fullyaware of the possible variety of existing waterdrawing practices used by people in acommunity throughout the year so diat dieinterventions can be properly targeted.

The Community and Health EducationFrom what has been said so far it should be clearthat technology alone, even if it is cost effective,is never going to lead to me elimination ofGuinea worm disease. All interventions must gohand in hand with appropriate health educationbased on a proper understanding of thecommunity. We will return to this theme manylimes during diis report but as discussed inSection 6 this subject has been well coveredelsewhere and diese aspects are not examined indetail in this report.

14

\ - U- Guinea Worm Disease Transmission Seasons in Africa

Zone B

Zone A: Peak GWD transmission season May-August (the rainy season)Zone B: Peak GWD transmission October-March (the dry season)

Adapted from: Guiguemde. T.R. 1986. Caracteristiques climatiques des zonesd'endemie et modaiites epidemiologiques de la dracunculose en Afrique. Bull.Soc. Path. 79:89-95.


2.0 WATER SOURCES AND THE RISKSOF GUINEA WORM DISEASE(DRACUNCUIJASIS) AND OTHERDISEASES

SECTION SUMMARY

• Rainwater from roofs which is stored incovered containers from which the water isdrawn hygienically offers no risk of Guineaworm transmission and little or no risk oftransmission of other pathogens;

• Groundwater from a protected source, ifcollected in an hygienic manner, offers norisk of Guinea worm transmission and littleor no risk of transmission of otherpathogens;

• Fast flowing surface water sources offer noGuinea worm transmission risks. Othersurface water sources have a varying levelof Guinea worm transmission risk but thiscan be made minimal if the water can becollected without people entering it, andwater which splashes against people whodraw it is prevented from draining back intothe source.

• Untreated surface water sources usuallycontain pathogens from human faeces;

• The quality of drinking water supplied by asource is not die most important factor inthe fight to reduce the transmission ofcommon diseases.

• Sufficient quantity of water needs to bemade conveniently available for domesticand personal hygiene but this water doesnot need to be of a very high quality. Peopleneed to understand and practice personalhygiene. Households need to understand,and practice, hygienic methods ofcollecting, storing and using drinking waterfrom a protected source to prevent the waterbecoming contaminated before it isconsumed. Communities need to adoptappropriate sanitation methods to preventthe spread of excreta related diseases.

• The three usual interventions which aredesigned to deal only with breaking thecycle of Guinea worm infection do little toimprove the bacteriological quality of thewater (Table 2.1)

Table 2.1 presents a simplified summary of therisks of transmission of Guinea worm diseaseand faecal pathogens.

Table 2.1 Advantages and disadvantages of the three main interventions for interrupting thecycle of transmission of Guinea worm disease, with particular reference to the resultingbacteriological quality of the water.

IKTEHVOTIOi

1* Pre^MAtino cortlictPrevention of infected peopleentering the water and contami-nating it with larvae.

For exanple:-

•Fencing or other physical fil-ters barriers to prevent every-body entering the water andpreventing any water which maycom in contact with infectedpeople from carrying larvae backinto the source

•Rules imposed by the comnunitybaning any infected person enter-ing the mater

2. StrainingSeparating the copepods fromwater used for drinking by pour-ing it thmugn a finely woven

i cloth or a monofilament meshi!

ii

3. ChemicalsApplication of tenephos ('Abate')in particular to the source(s) ofwater used for drinking. Chlorinecoanounds and iodine have alsobeen used.

ADVANTAGES

Cycle of transmission of Guineaworm should be interrupted

Also prevents people bringingother contaminants (eg. faecalmatter) into the water

Interrupts the cycle of transmis-sion if used on every ocassion. Afairly simple technide which canbe adopted by individuals for usein the home or when in the field.

Also kills off mosauito larvaewhich could transmit malaria

DISADVANTAGES

It only needs one infected personto bypass the barrier or breakthe rules to start the cycleagain.The whole commonty must under-stand and cooperate.

Some water collecting deviceacceptable to the community isneeded (eg. shaduf, bucket andrope, handpump)

Uninfected people need to bewilling to draw water for theinfected people.Uninfected people are likely tocontaminate the water with bacte-ria and other pathogenic organ-isms when they step into thewater

The straining will not renovepathogenic bacteria, viruses etc.but people may have a false sense iof security that the water is |safe.People need to raraflei to carrya strainer with them if they aregoing to drink water elsewhere.

The chemical needs to be readilyavailable, and to be added to thewater body at suitable frequen-cies and in the right quantitiesto ensure the copepods arekilled.Temepnos d o n not kill pathogen-

ic bacteria, viruses etc. butpeople m y feel that the water issafe because it has been treatedwith a chemical.Where there are a large numberof sources it is not practical totreat all of then.

Ifc

IIIIIIIIIIIIII1III

2.1 Water Sources. Guinea WormTransmission And Water Quality

Mankind collects water from various points onthe hydrological cycle which operates betweenwater in the atmosphere and water on the earth.It is necessary to understand the relationshipbetween the various water sources and Guineaworm disease and other diseases to ensure thatimprovements aimed at eradicating Guineaworm do not inadvertently lead to increases inother diseases.

2.1.1 Rainwater

This is one of the purest sources of wateralthough it can be contaminated as it flows overthe surface used to capture it. There are no risksof Guinea worm transmission related to its usefor potable purposes until:

it mixes with water already containing infectedcopepods orit is stored in such a way that copepods canbreed in it and the copepods can becomeinfected by Guinea worm larvae

The use of covered jars or tanks to store watercollected from roofs offers no risks of Guineaworm transmission as long as a hygienic methodis used to draw the water. The quality ofrainwater from above-ground catchments isgenerally good with regard to bacteria and otherpathogens. However the storage of suchrainwater in open tanks and ponds has risksassociated with it which are similar to those forponds used for storing rainwater collected fromground surfaces.

If rainwater flows over ground surfaces beforeentering storage there is risk of Guinea wormtransmission if the water passes through poolscontaining infected copepods or Guinea wormlarvae. Even if the water is not infected when itenters storage, if copepods breed in the pond orstorage tank, and if water is collected in a waywhich allows it to be infected with larvae, thesource provides a transmission site. It should benoted that although larvae are bound tocontaminate the stored water if people withGuinea worm lesions on their lees enter it thereis also a risk where people do not enter thewater. This risk arises if water which splashes

against a Guinea worm lesion is allowed todrain back into the source.

Even if there is no risk of Guinea wormtransmission the bacterial quality of water islikely to be poor when collected from a groundcatchment and this may lead to the spread ofother diseases.

Although rainwater collected from an elevatedsurface may be potable, people who arepreviously used to different sources of water maydislike its taste and may initially reject it as asource of drinking water.

Rainwater needs to be stored so that it isavailable when there is no rain. If there is a longdry season the volume of storage required willbe large. If the catchment surface is the roof of ahouse the storage can be very convenientlyplaced near to the home. With an above groundtank water can be collected hygienically from atap. Subsurface tanks require some liftingmechanism.

2.1.2 Surface Water

As rainwater runs over the ground it can pick upcontaminants and once it has joined larger waterbodies it can be further polluted by people usingthe water for different purposes. Usually it willcontain pathogenic organisms originating fromfaeces, so some form of treatment is alwaysadvisable to prevent the spread of disease.However, small scale methods of treatingsurface water in rural areas of developingcountries often meet with operational andmaintenance problems.

Untreated water may also contain a quantity ofsuspended matter such as tine clay particleswhich affect the colour and taste of the water,but existing users of untreated surface waterhave often grown accustomed to these and donot reject the water on these grounds.

Copepods are not usually found in fast runningstreams and rivers (Muller, 1979) so the risk ofGuinea worm transmission in such cases is low.However in slow flowing water copepods can bepresent and they can become infected by larvaefrom people with Guinea worm lesions. Even ifone community draws water in such a way thatihey do not allow Guinea worm larvae to enter

the water, if the water is flowing they are at riskfrom infection from upstream users whocontaminate the water with larvae. Where wateris stationary the situation is similar to thatexplained above for rainwater storage ponds.

2.1.3 Groundwater

As polluted surface water infiltrates throughgranular soils and sub-strata to join thegroundwater it filters out virtually all pathogenicbacteria and viruses. Groundwater is thereforeusually of potable quality but it can becontaminated during collection, for example bythe use of a dirty bucket and rope in a wellwhich can lead to the contamination of thewater. Covering the well and installing ahandpump prevents this source of contaminationbut adds another problem since the handpumpwill need proper maintenance to be sustainable.Where suction handpumps are used it isimportant that the footvalve design is good orpeople may contaminate the pump by priming itwith polluted water.

Occasionally some natural chemicals may occurin the groundwater at concentration levels whichmake the water unsuitable for humanconsumption, or which cause it to be rejectedbecause of taste, colour or smell. Sometimes thewater is 'hard' and it may be rejected by usersbecause it increases the time, and therefore fuelneeded to cook some foods and it increases theamount of soap needed to form a lather. In anumber of areas in west Africa the groundwatercan also be naturally corrosive and this createsproblems where galvanised iron or mild steelhandpump components are used because theycan corrode and fail very rapidly.

Groundwater is often found in aquifers whichextend over a large area so although sometimesefforts may be necessary to find the highestyielding site, often a well or borehole can belocated conveniently close to the community thatwill use it

Springs only occur in specific locations but theyhave Hie advantage that they can deliver waterabove ground. If the spring is on a hillside abovea community they can be served by gravitythrough a piped system.

Groundwater offers no risks of Guinea wormtransmission until it enters a pool or well wherecopepods breed and where Guinea wormsufferers can contaminate the water with Guineaworm larvae. Hence there is no risk oftransmission in groundwater from protectedsprings, boreholes/tubewells or covered handdug wells equipped with handpumps or fromdraw wells' (where water is collected in abucket which is lowered from the surface on arope) with parapet walls.

Where open wells are not properly constructedand Guinea worm larvae from users can enterdie water, there is a risk of transmission only ifcopepods live in the water. Interestingly, itseems that copepods do not usually survive inthe water found in most draw wells. Muller(1985) reports that due to the absence of enoughlight for the growth of the algae and protozoa onwhich the copepods feed "they are not usuallyable to reproduce in draw wells, provided thediameter is less than about 4 metres" but noconfirmation of this has been found in otherreferences. There is a great risk associated withstep wells' which are a type of well wherepeople descend a path or steps into anexcavation to draw water by dipping a containerinto, or even sometimes entering, the well water.

2.2 Transmission Of Diseases Other Than

2.2.1 Faecal Contamination In The Home

The risks of contamination of sources of waterby faecal pathogens has already been mentioned.Unfortunately the provision of good qualitydrinking water at a source does not necessarilymean that people using the source will drinkgood quality water because it can becomecontaminated by faecal pathogens at an numberof points before someone drinks it. For example,it may be contaminated by a diny container usedto collect it from the source or to draw it fromthe storage container, or if poorly stored in thehome it can become contaminated in other ways.

Appropriate health education is necessary beforepeople make the changes in habit which arenecessary to improve the level of domestichygiene needed to maintain the quality ofpotable water until it is consumed.


IIIIIIIIIIIIIIIIIII

Although this paper relates in the main towater supply and the transmission ofdracunculiasis it must be recognised thatimprovements in water quantity, quality andwater handling practices only go part waytowards the reduction of diseases in acommunity. Because many of the water bornediseases caused by these pathogens are alsotransmitted by other routes, such as by flies orunwashed hands, only supplying uncon-taminated drinking water does not necessarilylead to a dramatic reduction in theirprevalence. It is necessary that appropriatemethods of sanitation and good standards ofhygiene are also widely adopted if the spreadof excreta-related infections is to be reduced.

It has been noted that an increase in the quantityof water used for personal hygiene, even if it isnot of a good potable standard can lead to amore general benefit than purely addressing theijuaiiti of the drinking water supplied. One ofthe main ways of increasing the quantity ofwater used is to make it more convenientlyaccessible.

2.2 Other Water Related Diseases

In addition to the nsks of the transmission ofGuinea worm disease and faecal pathogensmentioned above there are a tew other diseaserisks associated with water sources which are ofimportance in relation to some sources of supplyused in Guinea worm endemic areas: -

2.2.1 Malaria

Mosquitoes which transmit malaria from oneperson to another breed in stationary water. Tinsis equally true for very small amounts of wateras for the large volumes in water storage tanks.If it is planned to introduce open storagereservoirs to an area then the risk of increasingthe spread of malaria should be investigated andminimised. For example, preventing the growthof weeds along the waterline can reduce theopportunities for mosquitoes to breed.

2.2.2 Schistosomiasis (Bilharzia)

Snails which form the intermediate host duringthe transmission cycle of schistosomiasis canbreed in still or slow moving water. If they

become infected by the miracidia which developfrom the ovum passed in the urine or faeces ofan infected person the snails will later releasemany cercariae which can burrow into the skinof someone entering the water, or lesscommonly enter their bodies through thestomach when they drink the water. Wherebilharzia is endemic in an area clearly there arerisks associated with some surface water sourceswhich need to be minimised.

3 COST EFFECTIVE TECHNOLOGIESFOR GROUNHWATER EXPLOITATION

SECTION SUMMARY

Groundwater exploitation has already played animportant part in the eradication of Guineaworm disease from a number of places in theworld. Groundwater has two main advantages:

• it is usually potable and free of anycopepods to be infected with Guinea wormlarvae. It does not usually contain faecalpathogens or other contaminants;

• aquifers are often extensive and in suchcases it is usually possible to locate a wellvery near to a community needing water.

The main disadvantages of exploitinggroundwater are:

• it can be difficult to excavate the holenecessary to reach the ground water. Thisproblem increases as the depth to thegroundwater increases and with thehardness of the sub-strata.

• except in the case of springs and artesianwells, some way has to be found to lift thegroundwater up to ground level. Thisproblem obviously increases with the heightof lift required, and extraction using ahandpump is increasingly difficult as thegroundwater depth increases. Fewhandpumps are acceptable to communitieswhere the lift is beyond about 45 metresbelow the surface and because of theincreased forces on the components of suchpumps they need to be very robust.

The most important observations andrecommendations to arise from the section are:

• Where springs occur full use should bemade of them. When they are properlyprotected they can supply potablegroundwater, in most cases without theneed for a pumping device. Cost effectivespring protection designs which do not needspring boxes are available.

• Appropriate groundwater locationtechniques should always be used to avoidas much as possible the cost of wasted bore-holes or wells;

• Hand-drilled boreholes have great potentialwhere the strata is suitable. They are oftenmore cost effective than hand dug wells butunlike draw wells they need sustainablehandpumps for the withdrawal of water.The boring equipment is relatively cheapand usually very portable, and thecommunity can be fully involved in its usefor borehole construction.

• In a limited category of strata the use ofwash-bored tubewells is very cost effectiveand it can lead to the very rapidconstruction of tubewells. The equipment isrelatively cheap and easily transported butafter completion the tubewell will need to beequipped with a sustainable handpump.

• In many cases low yielding strata,traditionally not considered suitable forgroundwater exploitation, can be used tosupply water to a community. This isespecially true when hand dug wells areconstructed but very low yielding aquiferscan be exploited by a tubewell if it isdesigned to serve only a small population,such as would be the case when using theBlair bucket pump.

• Where the groundwater is not excessivelydeep, open hand dug wells have greatpotential and such a method of exploitinggroundwater makes maximum use of localmaterials and community participation.Where unprotected wells are already beingused by a community the transmission ofGuinea worm larvae is a risk, and properprotection of those wells, or construction ofnew wells, is the obvious choice oftechnology. The use of a bucket andwindlass, or a 'shaduf reduces the risk ofwater pollution by users.

• Where a sustainable handpump acceptableto the community is available, hand dugwells should be covered and water should bewithdrawn using the handpump becausethis reduces the opportunities for the wellwater to become contaminated. However

t ^H


facilities for water to be withdrawn from thewell by bucket should also be provided foruse in an emergency. The simple, easilymaintained, direct action type of bandpumphas great potential for most hand dug wellsince the pumping lift usually does notexceed 15m. Where the community beingserved by the well is less than 60 the Blairbucket pump is an alternative which shouldbe considered.

Where there are more than one protectedsource in a community, such that the veryoccasional failure of one handpump on onesource would not mean people had to returnto polluted traditional sources, the 'casedwell' design or 'partially lined well' designshould be considered since these useconsiderably less materials than fully linedwells. Both types of wells are suitable forhandpumps or the Blair bucket pump. The'Modified Chicago Method1 of temporarysupport can be very cost effective in theconstruction of such wells.

Powered drilling rigs for constructingboreholes need to be chosen with care tofind a type most appropriate to the fieldconditions. Relatively simple, lightweight,cheaper rigs are now available which havethe potential for much cheaper unitborehole costs. Their manoeuvrability oftenmeans that boreholes can be drilled inremote places inaccessible to the heavierrigs.

There are a vast number of differenthandpumps available on the market but alsoa number of sources of useful advice and theresults of field tests to enable a wise choiceto be made for particular field conditions.Undoubtedly the suitability of a pump forvillage level operation and maintenance(VLOM) is very important. Of equalimportance are the standardisation in anyone country on just a few models ofhandpump, and the in-country manufactureof the pumps, or at least the commonlyneeded spares.

The use of plastics for handpumpcomponents in corrosive groundwatersoffers many advantages and is very costeffective compared to the alternative use

stainless steel. The successful use of plasticrising mains, particularly with open tophandpump cylinders has already beenproven in the field, and current researchshould shortly produce improved connectorswhich will allow the pipe sections to beeasily separated when necessary. Directaction handpumps which often have onlyplastic components below ground aresuitable for use in virtually all situationswhere the water lift does not exceed 15m.The use of stainless steel rods and plasticrising mains with deepwell pumps is alsonow well tried.

• The simple, relatively inexpensive andeasily maintainable Blair bucket pump isworthy of consideration where the numberof people relying on a source is less than 60and the lift is less than 15m. The low dailydemand from such a group may extendfurther the type of strata which can beconsidered suitable for yieldinggroundwater. If due to the convenientlocation of the borehole people are willingto spread their demand throughout the day,avoiding sustained peak use, the pumpcould be used on a very slow yieldingborehole.

• The use of a shaduf with the Blair bucket ispotentially very cost effective but to date hasnot been tested in the field.

• Where the aquifer appears to be suitable fortheir use, the potential for directly installingthe types of open top cylinder handpumpswhich have extractable foot valves shouldbe investigated. Such a method ofinstallation dispenses with the need for alarge diameter borehole casing and cantherefore be very cost effective.

• There is a wide variation in the way inwhich different projects calculate unit costs,and caution is needed when consideringcosts in isolation from full details of theirbasis.

3.1 Exploiting ^mundwater found at groundlevel

Springs: Where the groundwater table intersectsthe surface, as in the case of a spring, it is veryeasy to reach the groundwater and if the site ofthe spring is conveniently positioned in relationto a community it is usually very cost effective toprotect it to provide a potable source of water.Simple spring protection designs which do notuse costly 'spring boxes' are suitable where thespring is flowing at a sufficient rate to satisfypeak demands [Skinner (1992), Rous (1985)].Many springs can be protected using localmaterials, a short section of delivery pipe and 6-10 bags of cement (Figure 3.1.1).

Where the spring is slow flowing some form ofstorage with a supply controlled by one or moretaps may be needed to satisfy peak demand.Often if sufficient storage is provided, then evena slow yielding spring can be exploited andferrocement tanks can be a cost effectivemethod of providing storage. Although the useof cement mortar jars (see Section 4.4.2) forspring water storage has not been noted in theliterature studied there is no apparent reasonwhy one or two large cement mortar jars shouldnot be used for this purpose as long as there issufficient slope to the land below the spring toprovide for the height of the jar.

Where the spring occurs in an area where theground has only a very shallow slope there maybe insufficient height for the construction of atraditional headwall and spring outlet pipe butother methods of exploitation may be possible.For example the spring can be routed into abelow-ground tank (which could be similar to alined hand dug well) from where the water canbe drawn using a suction pump or a bucket andrope. However since the surface gradients areslack and surface water runoff is slow it is veryimportant with this type of spring that pollutedsurface water is prevented from contaminatingthe groundwater. In some places in Uganda anexcavation in clay filled with stones and coveredwith a protective layer of clay has been used forgroundwater storage with an offset SWS Rowersuction pump (see Section 5.2) to allow itsextraction (WaterAid 1992).

If a spring occurs on a hillside above acommunity then there is potential for a gravitypiped supply to tapstands which can be locatedconveniently close to the community. The costs

of such schemes are very site specific and percapita costs are rather meaningless. Where suchgravity supply schemes are possible they shouldbe investigated because there is great potentialfor health improvements due to the improvedquality and quantity of conveniently availablewater. Also the maintenance of the schemes canbe simpler than that for handpump schemes.Where a community is very scattered a largenumber of tapstands are needed (to provide asuitable density of standposts to make at leastone of them conveniently close to each user) andtheir cost and that of the associated pipeddistribution system is usually prohibitive.

Morgan (1990) shows that with some hand dugwells it is feasible to construct siphon wells' or'gravity wells' to allow piped water to becollected as shown in Figure 3.1.2.

Artesian boreholes/tubewells: Artesianconditions can occur where an aquifer isconfined between two impermeable strata andthe pietzometric head is above the top of theaquifer. If excavations (including drilling) reachsuch aquifers then the effort required to lift thewater are reduced since it will naturally rise partway up the well. In a few cases the groundwater•ressure is even sufficient to raise the water up

to the surface and no pumping is required, butsince such artesian conditions are rare they arenot further considered in this report.

3.2 Finding Sufficient Groundwater

Unfortunately groundwater is not alwaysavailable because some geological formations donot hold water or do not yield it at a sufficientrate to supply water to a community. Wheregroundwater is being extracted from an aquiferit is important that the strata is also receivingsufficient infiltrated water to replace what isremoved or the groundwater level will fall. It istherefore important that wherever possible ahydrogeologist advises any project planning toexploit groundwater.

Groundwater is found in the pores between thegrains of granular strata or in the fractures inweathered rock. The first mode of occurrenceoften exists over a large area and the exact sitingof a well may not be very critical once a suitablearea has been identified. However where the


TIIIIIIIIIIIIIIIIII•

water occurs in fractures the amount of wateravailable is very dependent on the width andpattern of fractures and in certain situations twoboreholes can be just 5m apart yet one can bedry and the other can have a reasonable yieldRecent developments in the use of hydro-fracturing of dry or low yielding boreholes infractured rock has been giving encouragingresults. In India 1012 of 1350 previouslyunsuccessful boreholes gave acceptable yieldsafter such treatment. The average cost ofhydrofracturing per borehole was US$200compared to a drilling cost of USS770 perborehole (Davis, 1992).

Cost effectiveness of groundwater exploitationdepends upon the avoidance of dry holes or lowyielding holes. In some countries a fairly reliableindication of groundwater fairly near to thesurface is known to be given from the presenceof specific types of trees which indicate thatwater is present at certain depths. Traditionalwater diviners' or 'dowsers' have sometimesproved to be reliable. Hand augering of smalldiameter exploration holes or the use of jettingprobes are also possible for prospecting forgroundwater at shallow depths. Blankwaardt(1984) notes that in Tanzania 30 - 40 surveyboreholes of 100mm diameter can be drilled forthe cost of just one hand dug well lined withconcrete rings. He observes that since handdrilled borehole are cheaper than hand dugwells, so an intensive survey using 100mm testboreholes to find a suitable site for a 230mmdiameter hand drilled borehole is usually alsovery cost effective.

There are various other techniques ahydrogeologist can use to assist in the locationof exploitable groundwater even if it is not at ashallow depth. These methods include on-sitegeophysical methods (e.g. electrical resistivityand electromagnetic methods) and remotemethods (e.g. using aerial or satellitephotographs). Some of the equipment isexpensive and has high maintenance costs somethods of investigation can appear at first to bequite costly. However, de Rooy et. al. (1986a)found that the average cost of geophysicalsurveys to locate boreholes in basement rocks mNigeria came to between 2.5% and 5% of thecost of the borehole. Jones (1987), also reportingexperiences in Nigeria, notes that the cost of

equipment to check up to 30 sites per month wasequivalent to the cost of two dry boreholes.

Avoidance of dry wells is only one aspect of theuse of geophysical equipment. It is also useful inenabling well/boreholes sites to be chosen whichwill have greater yields and specific capacitiesthan would otherwise be possible (cit.ibid).Proper interpretation of strata as drilling ofboreholes proceeds is also important since itmay allow savings to be made with regard toaspects of the design such as casing length,screen size, screen length and gravelpack/geotextile sizing.

Air flush mechanised drilling methods (with theuse of foams where appropriate) allow the yieldand strata to be more easily assessed as drillingproceeds than is possible when mud drilling isemployed and the identification of narroweraquifers in sedimentary strata is therefore madeeasier. Down the hole geophysical logging cansometimes be cost effective in giving guidancefor the accurate location of screens opposite tothin high yielding aquifers and boreholes maythen not have to be as deep as they would havebeen if such methods were not employed(cit.ibid).

Where a programme is to be based oncommunities participating in the construction ofa hand dug or hand augered well there isanother factor which needs to be considered. Ifin such a case an excavation does not find waterthen the whole programme may be affectedbecause other communities will not want to takethe nsk of labouring in vain. Properhydrological investigations can avoid such anoccurrence and reduce costs.

When considering the suitability of waterbearing strata for exploitation it is importantthat the small yields that are generally requiredfor handpump extraction are recognised. Yieldsof 0.2 - 0.3 litres/sec are usually enough forhandpumps which means that many areas notgenerally regarded as potential groundwatersources are in fact suitable for this method ofextraction (Arlosoroff et. al. 1987). Howeverallowance will need to be made in low yieldingaquifers for the likely large drawdown whichwill arise in the borehole during pumping. Anaquifer with a specific yield of only 0.05 1/sec/m

can meet the demands of one handpump if thedrawdown reaches 5m.

Unlike small diameter boreholes, hand dugwells are able to store appreciable quantities ofwater and this often means that even loweraverage yields than that mentioned above canmeet the water needs of a community becausethe aquifer can be yielding water to storagethroughout the day and night. For example acommunity of 200 people requiring 30 litreseach per day can probably be satisfied by anaverage yield of only 0.7 litres/sec if about 4 -5m3 of storage is provided (5m^ is equivalent to3.8m depth in a 1.3m diameter well). Wherelarge numbers of people are likely to use a handdug well in a slow yielding strata it can be dugwith a larger diameter to improve the areaseepage face in the aquifer and to increase thevolume of storage per unit depth. Another veryeffective method of increasing the yield, but onewhich involves the use of mechanical plant, is tobore horizontal holes radially through the wallsof the well and into the aquifer (Figure 5.4.1.4).The British Geological Survey has beeninvolved in the development of this technologywhich needs a well of at least 2 metres internaldiameter. The equipment has been field tested inSri Lanka and Zimbabwe in alluvium andweathered rock where dramatic improvementsin yield resulted [Wright & Herbert (1985),Herbert (1990), Herbert & Rastall (1991)].

Whatever method for reaching and abstractinggroundwater is used it needs to be able toconveniently provide sufficient quantities ofwater which can be collected at a rate whichmakes use of the facilities more attractive thanthe use of any other local sources of water. Oneapproach is for the hydrogeologist to starthis/her investigations from the position chosenby the users and then to work outwards from thispoint to find the nearest site capable of yieldingsufficient water

Groundwater near to existing surface watersources is considered in Section 5.4 wheresurface water infiltration systems are examined.

In Section 2 possible problems relating togroundwater quality, corrosive groundwatersand the rejection of potable groundwater bysome communities were mentioned and thesewill also need investigation. It should be noted

that groundwater held in fractured rocks, unlikethat in fine grained aquifers does not benefitfrom natural filtration as it passes through thefractures, and the risk of contamination aretherefore greater. However, the groundwatermay have reached the fractures throughsuperficial deposits. Other factors which needconsidering when promoting groundwater as asource are seasonal or annual movements of thewater table and competing demands from nearbypowered pumps.

3.3 Reaching The Groundwater

3.3.1 Reaching groundwater using largediameter excavations

Hand dug wells:Traditional useSome communities have traditionally dug holesdown to the groundwater table to extract waterfor domestic use. Where the excavation is in ariver bank the well not only exploitsgroundwater approaching the river but, underthe right conditions, it can also collectinfiltrating surface water from the river. Whereseasonal rivers occur excavations in the riverbed during the dry season are often used by localpeople to collect water flowing below the bed.These river bank wells and river bed wells areconsidered further in Section 5.4.

Usually traditional wells are of poor design andas a result the water is often polluted with faecalpathogens and in Guinea worm endemic areaswith infected copepods. It is usually possible toimprove a traditional unprotected well toexclude much of the contamination by providingan improved lining and a parapet wall andapron slab. Further improvements can resultfrom the provision of a rope and windlass, or theuse of only one dedicated bucket which is neverplaced on the ground by users. The cost of amass produced windlass used in Zimbabwe, asillustrated in Figure 3.3.1.1 is about S13. Thesesystems have been provided free with three bagsof cement as an incentive to users to protectadequately their household wells involving atotal subsidy of between $60 to S80 (Morgan &Chimbunde, 1991). Still further improvements;ire possible if the well is covered and fitted witha more hygienic water drawing device (Figure3.3.1.2) as discussed below when consideringnew wells.



1A:

Plp"

, . \ - '

S|H»iq«y» _ JIX '

Two ways o< coiiactlng and conveying lh« watar Iretn the spring aya

Ground sunace slopesalong wall to shedsurface water sideways

Stones behind headwail and" wingwalls act as a drainage layer

Sharp stones set in thetop ol wall discouragepeople tram sitting or

it

Plastic ppe insidegalvanized iron pipe

Bowl-shapedconcrete slab todirect water intopipe, coveredwith slab orstone and alayer ot clay

VU.W

\ V./7JJ-J Hard rock to resist

Wastewater

Typical headwail details

Suction pipe

Siphon well

Inverted ball valve

Water pointApron

The siphon well.

Earth backfill

Brick 2.15 mlining

Reservoir

Cross-section of a gravity well.

Improved constructionAs discussed in Section 2 pure groundwater canoften become polluted and can easily become asource of copepods infected widi Guinea wormlarvae because the design of the collection pointis poor. There are a number of good referencesto those design features of hand dug wells whichare necessary to reduce the risks of contamina-tion by users, spilt water or surface water [Watt& Wood (1976), DHV (1979), DHV(1985)] andthey will not be repeated here. The commonmediods of construction and lining are alsodescribed in these two references together widithe necessary safety precautions. Differentmethods of construction are appropriate fordifferent strata.

In firm self-supporting strata common practiceis to excavate in the dry season untilgroundwater entry causes difficulty and then,where necessary, to construct in situ liningstarting at the bottom of the excavation.Deepening into the water table then continues byexcavating inside a sinking caisson until thewell has penetrated sufficiently far into theaquifer. During this caissoning stage a pump isused to dewater the excavation.

In strata which are not self supporting it ispossible to line the well in short sections asexcavation proceeds from the surface until thegroundwater causes problems. Then wherenecessary proceeding by caissoning.Alternatively virtually all of the well can beexcavated from inside a sinking caisson. Bothmethods can also be used in firm strata.

Permanent lining and caissoning is oftenconstructed from mass or reinforced concrete orhard burnt bricks. Winter (1987) gives details ofthe use of precast ferrocement linings. All theselinings are relatively costly and their construc-tion is often time consuming. One time savingidea used for the production of precast cementmortar rings in Zimbabwe is to compact themortar between slightly tapering shutters. Theheight of these rings is only 300mm and theyare lightly reinforced widi wire. The shutterscan be removed almost immediately so thatabout 20 rings can be cast in one day using diesame shutter.

For the 'backfilled well' design (see below) nocostly permanent large diameter lining is neededbut instead a small pipe casing is installed(Figure 3.3.1.3). Another cost saving is possibleif permanent lining is only used in the section ofthe well in the aquifer and this lining is thencapped with a strong concrete cover slab fromwhich only a smaller diameter casing rises to thesurface (Figure 3.3.1.4 ). One disadvantage ofboth these methods is that a reliable pump mustbe available for withdrawing die water unlessthe pipe is at least of sufficient diameter to takea bucket (or a 'Blair' bucket pump is used).Another disadvantage is mat. unlikeconventional wells, these backfilled wells cannot be easily deepened should the groundwaterlevel ever drop below the bottom of the well.

Modified Chicago Method of temporary supportOne interesting hand dug well constructionmethod which does not use any large diameterpermanent lining is the 'Modified ChicagoMethod' of providing temporary support to theexcavauon. This mediod is promoted byConsallen, a British manufacturer of an open-top cylinder handpump. It reports that themethod has been used in Liberia and recentlymore extensively in Busoga, Uganda where todate 250 wells have been constructed with 90being completed last year (Consallen, 1992).

The temporary support system used with dieModified Chicago Method consists of verticaltimbers (about 150mm x 50mm x 4200mm)positioned around die circumference of dieexcavauon and held against die excavated faceby wooden wedges driven between the insideface of die boards and number of circular steelwailings (Figure 3.3.1.5). The wailings whichare each in die form of a 1.22 metre diameterring of welded steel channel section (about100mm x 50mm) are spaced vertically at about1 metre centres down the excavation to formcompression rings against which the wedges act.Each wailing comes in two halves which can belowered down die well and be bolted togetherand positioned against the boards at die bottomof die excavation. The spacing of the boardsaround the circumference of me excavation canbe chosen to suit the ground conditions.

Excavation near to die circumference of the wellproceeds under the end of each board in tum.Excavation periodically stops and the wedges


WlndlauBucfcat hung onWinOasa nanal*

Tin ltd

Vi-l.l

Rock lining .undw waiar

'Croa-ttction of an upgraded wtW from Blair's Upgradedwell manual for (ieldworkers.

3 m

E

Improving anexisting well by addinu ahund pump, a cover slab,an apron and drain, and apuddled clay barrieragainst seepage ot surfacewater

E

Source: After WaenerandLanoix (

L/,L_:.;\_ o . . •

0.20 m 15m0.2 m

-.1

Backfill well

T

l

holding individual vertical boards for thatparticular sector of the circumference areloosened so that each board can be moved downso that its end again touches the bottom of theexcavation. Once a full length of board has beeninstalled in the well additional boards can be fedbehind the rings from the top to follow eachdescending board. As excavation proceedsadditional wailing rings and wedges are addedat the bottom of the excavation to support theboards which have been moved down.

The advantages of the Modified ChicagoMethod cited by Consallen mclude:-

* It employs very little capital equipment.* Local manufacture of the equipment ispossible.* The equipment can be carried from site to siteas head loads.* In all types of soil conditions the workers inthe well are always fully protected.* Excavation work can safely proceed belowgroundwater level in dewatered excavations* It maximises the contribution of untrainedvillage workers allowing trained personnel tosupervise the construction of several wellssimultaneously (e.g. one technician can super-vise 6 wells).* The water yield can be tested beforecompleting the well.* There is no loss of materials in the eventthat the hole has to be abandoned.

Backfilled (Cased) WellsIf temporary supports are used to support thesides of an excavated well, once the excavationhas sufficiently penetrated the aquifer it isfeasible to install just a 110mm diameter uPVCcasing pipe from the surface, then to backfill thewell and to install a handpump in the casingpipe. The section of the excavation in theaquifer can be permanently lined with largediameter rings (Figure 3.3.1.4). An alternative islo use a slotted uPVC pipe surrounded with ageotextile or gravel pack, which in turn issurrounded with free draining material whichfills the remaining space to the edge of theexcavation (Figure 3.3.1.3). Both designs canconsiderably reduce the cost ot constructing ahand dug well but the small diameter of the pipemeans that a reliable sustainable handpump hasto be purchased because the usual bucket andrope system cannot be used to draw out the

water (other than as with the Blair bucket pump(Figure 3.3.3.6).

Consallen (1992) reports that the capital cost ofthe equipment needed to excavate a typical 15mwell excavated using the Modified ChicagoMethod, including the jet ejector pump used fordewatering is about £1000. Since these materialsare all reused the costs can be spread over alarge number of wells. The cost of four sets ofequipment allowing a technician to supervisework on four sites at one time is only £2000-£3000 if the pump, which is only needed for partof die time, is shared between die four sites. TheUK based NGO which is supervising theprogramme in Busoga has calculated Uiat, wherefree local labour is used for excavation, the totalcost of constructing each well and equipping itwidi a handpump is about £1200 including allUK and field staff overheads (Consallen, 1992).

Advantages of hand dug wells

There are significant advantages of hand dugwells worth noting.

They can be used in low yielding aquifers andwhere appropriate can be used with horizontalboring to give considerably improved yields.

It is possible to remove or break up large stonesor rocks which would be troublesome in a handuugered hole.

If the aquifer is a fractured rock then once thislevel has been reached, if pick axes are notsuitable to make progress, excavation may beable to proceed using a hand held aircompressor driven tool. If a cased well is to beproduced a simplified down the hole (DTH)hammer drill can be operated to make aborehole into the rock to save the expense of alarge diameter excavation. In certain situationscarefully controlled use of explosives down thewell may be appropriate to allow the continuedconstruction of a large diameter well in harderrock. A Zimbabwe method, using hand chiselledcharge holes, is described in non-technicallanguage in a publication for fieldworkerswritten by Laver (1987).

In a hand dug well programme in the Houet andKenedougou provinces of south-west BurkinaPaso 53 wells with an average depth of 25

26

metres were dug in 1984-85 with a unit cost ofUS$7,472 (excluding a handpump but includingUSS3388 administrative costs). It used pumps,compressors, jack hammers and sometimesdynamite to excavate into a weathered zone andthe underlying fractured granites and schistsexcavating at least 3 metres below the dryseason groundwater level. On average, one teamtook two months to complete each well with theaid of the community. (Roark et al. ,1986).Another project which operates in northernGhana, and which also uses compressors andjack hammers to construct draw wells up to 12mdeep to penetrate decomposed rock has 'all in'unit costs of about £1800 (Water Aid, 1992).

If the strata at the base of the excavation is finegrained it is possible to probe ahead of theexcavation using simple water jetting equipmentto check what other strata is to be encounteredand to confirm the groundwater level. If a casedwell is to be produced the 110mm screen can bedirectly installed in the jetted hole. This isparticularly useful where running sand isencountered since such strata makes excavationdifficult.

Hand dug wells make maximum use of localmaterials and labour and the community cantake a major role in their construction leading toa greater sense of ownership. Their participationin its construction demonstrates their strongdesire for an improved water supply which isimportant if the facilities are to be properlycared for. Communities where such a desire isnot present are less likely to attempt a hand dugwell project giving an early indication of theneed for additional health education andmobilisation without which any water supplyproject in that community would be likely tofail.

Where the well is left open as a draw wellsimple bucket and rope systems can be used fordrawing the water, and particularly with thelarger diameter wells more than one person isable to draw water at the same time.

3.3.2 Reaching groundwater using smalldiameter excavations

Various techniques of boring small diameterholes are detailed below and some specificadvantages or disadvantages are discussed in

each section. There are however some generaladvantages and disadvantages which apply to allsmall diameter excavations: -

Advantages:• Greater depths are usually possible than

with hand dug wells.• Progress is usually fast compared to hand

dug wells.• With proper completion it is relatively

easy to prevent the groundwater beingpolluted before collection.

• The design of some 'open top cylinder1

handpumps means that in some situationsno borehole casing is needed because therising main never needs to be extracted.

Disadvantages:• The level of technical support is usually

higher than for hand dug wells.• A handpump is always needed (unless a

*bucket pump' is used) and this can lead toproblems of sustainability.

• Where borehole casing, screens, pipes,handpumps etc. are not manufactured in thecountry there is a need for foreign exchangeto import them. There is not usually a needto import materials with hand dug drawwells.

• Some strata is unsuitable for boreholesbecause in slow yielding aquifers a largevolume for water storage in the well isneeded. This can only be provided by largediameter wells, or the large diameterbackfilled wells already discussed.

Machine drilled wellsThere are a wide variety of drilling rigspromoted by various manufacturers to producesmall diameter holes in the ground suitable forexploiting groundwater and this report is not asuitable place to describe all the methods andthe particular strata and other situations inwhich each is appropriate [Annex 2 of Hofkeset. al. (1983) is recommended as introductoryreading]. The term boreholes is used to describesuch holes in this report. The term tubewell isalso in common usage but this is a categorytaken to also includes wellpoints which may bedriven or jetted into a strata with no bonngactually taking place.

There is no ideal all-purpose drilling ng orsystem of drilling but there are some important

27 I

1

factors which should always be considered whenchoosing a rig for use in a developing country inorder to achieve cost effectiveness.

What degree of sophistication can be supportedin the field?

For example, although a powerful rotary rig willusually drill boreholes much faster than a simplecable-tool (percussion) rig, if when the formerbreaks down it can not be repaired for a longtime overall progress will be hampered. Theusual high running and maintenance costs andthe high level of technical support needed forthe larger rotary rigs may also mitigate againsttheir use. Because the capital cost of a powerfulrotary rig is very high it may be possible to pur-chase a number of simpler rotary or percussionrigs which despite a slower individual rate ofoutput could together give acceptable overallprogress in providing boreholes.

What size of rig is appropriate for the groundconditions?

If the rig chosen is larger than what is reallyneeded for the strata and type of borehole beingconstructed then the unit costs of the holes arelikely to be higher than necessary. It should alsobe noted that the access requirements for a largerig will also restrict the places where boreholescan be drilled because the size and weight of therig may mean it is unable to reach many sites,especially during a wet season. Smaller, moremobile, rigs on the other hand may be able toreach remote areas away from the main roadsnear to which the large rigs usually operate.

In Sudan, the convergence of resources (leadingto reduction of overheads and costs of logisticalsupport), production incentives (leading toincreased output) and community involvement(allowing the government to optimize capitalresources and minimise recurrent costdisbursements) has brought about substantialreduction in unit costs. The cost of a handpumpequipped borehole in Sudan was USS2800 in1989, less than one third of the cost of the sameprovision in 1985 (WES 1992)

In UNICEF-assisted projects in Nigeria acombination of lighter and cheaper drilling ngsand support vehicles, effective use of state of ilie;irt geophysical equipment and methods for

borehole location, import substitution ofborehole construction materials, a reduction inthe ratio of expatriate to Nigerian staff and thedepreciation of the Naira led to the unit cost of ahandpump equipped borehole dropping fromUS$28,000 in 1981 to US$3,700 in 1990 (deRooy 1992). The average annual exchange rateduring this period fell considerably from aboutUS$1.6 to US$0.1 per Naira (cit.ibid) but thereduction in unit borehole costs is nonethelessvery commendable and other projects shouldconsider making similar changes to theirdrilling programmes to reduce costs. By 1990the cost of boreholes in UNICEF-assistedprojects in Nigeria ranged from an average ofUSS85 to US$115 per metre drilled whichcompared favourably with those in Malawi andGuinea (citibid quoting Beale, 1990).

Other recent unit costs per handpump equippedborehole (drilled about 40m deep), based onprogrammes of over 65 boreholes were obtainedfor Zimbabwe (@ £4000) and Angola (@£5000). From the same source (Jackson, 1992)other probably less reliable unit costs werequoted as follows:- Senegal (@ £12,000),Namibia (@ £5,500), South Africa (@ £4,000)and Mozambique (@ £2,700).

Chauvin et. al. (1992) quote an average cost ofUSS 14,858 per completed borehole in a projectwhich completed 298 installations in the northof Zou province in Benin. This figure includesthe cost of:- unproductive wells (there was a79% success rate), installation of a handpump(India Mark II), training of handpumptechnicians, salaries and benefits of allpersonnel and the cost of all equipment andsupplies. For the purposes of comparisonChauvin quotes a figure of USS9,700 percompleted borehole reported in 1985 for theTogo Rural Water Supply and Sanitationproject.

Roark et al. (1986) give an average cost ofUS$16,116 (including US$4,405 administrationcosts) for a programme of 15 new drilled wellsand 6 rehabilitated (deepened) boreholes drilledduring 1984 and 1985 towards the end of a 7year programme in south-western Burkina Faso.Tins includes the cost of all equipmentconsumables, personnel and die operational andadministrative costs. It is also believed toinclude the cost of 3 unproductive wells but this

2S

is not clearly stated. The average depth of thewells which were drilled into both sedimentaryand crystalline rock was 60m. Roark points outthat if the cost of external technical support issubtracted, as would be the case if the work wasperformed entirely by Burkinabe technicians, thecost would reduce by 11%. He notes also thedifficulties experienced maintaining the drillingrig for which in 1985-86 alone the repair costsreached about US$20,000. In a number of yearsduring the project, breakdowns led to longperiods when the rig was unavailable.Additional constraints on the rate of boreholecompletion were caused by rain which preventedwork during 2 months of each year.

This wide range of borehole costs betweencountries shows that caution is needed ifcomparing or making use of such unit costs fornew projects. For example the basis on whicheach project calculates its unit costs (such as theamortisation of plant costs) and the fieldconditions in which each operates can be verydifferent. The density of boreholes being drilledin a project area also has a major effect on therate of completion and therefore the percentageof overhead costs charged to each hole. In-country experience of borehole drilling is theobvious the best starting point for cost estimatesfor any proposed projects but even within onecountry, operating conditions may varyconsiderably.

Figure 3.3.2.1 shows a useful map produced byUNICEF in September 1991 to indicate the typeof drill rigs most appropriate to thehydrogeology of different areas of Nigeria(Donaldson, 1992). Note that it includes areaswhere manually operated rigs can be used.Presumably manually operated ngs are alsoappropriate in localised zones within other areasof the country, but these cannot be shown onsuch a small plan. For example Figure 3.3.2.2shows areas where the wash-boring of tubewellsis possible. The production of similar maps forother countries would provide a very usefulproject planning tool.

The use of larger numbers of simple low costpowered ngs instead of large sophisticated onesis a worthwhile area for further investigationinto cost effectiveness. Simpler designs of rigsare probably available from a number of sources

but an example from just one manufacturer willsuffice to show their potential cost effectiveness.These are purely examples about which data wasreadily available (Eureka, 1992) and the authorsare not in a position to recommend them aboveany other types that may be available from othermanufacturers.

Eureka UK Ltd. produce two simple low cost topdrive rotary rigs. Their recently produced 'Port-a-Rig', costing around £8,000, is highly mobilesince it can be carried in the back of a standardpick-up. In order to make things as simple aspossible a manually operated chain and sprocketsystem is used for lifting the drill string and therotary head is shaft driven, avoiding the need forany hydraulics. Its water/mud tank which restson the ground provides stability, and usingwater or mud circulation the rig can drillboreholes of up to 200mm diameter down toabout 30m in unconsolidated strata. Use of£10,000 worth of ancillary equipment,consisting of a down the hole (DTH) hammerand a small towable compressor, allows holes ofup to 100mm diameter to be drilled in hardrock.

Eureka's trailer mounted rig has similar featuresto the Port-a-Rig' but with a higher horsepowermotor and a stronger lifting winch which allowboring to proceed to depths of 60m. The trailerbody acts a a water/mud circulation tank. Themast allows the easy handling of 6 metre lengthsof casing. The basic unit with a pump for directcirculation rotary drilling costs around £22,000with an extra £10,000 for the DTH hammer andcompressor.

One other important factor which can have adramatic effect on the rate of boreholecompletion, although not related to thetechnology, should be considered in everyproject. This is the paying of bonuses to drillingteams. Me Pherson et al. (1989) report that inSudan a bonus system shared between managersand field crews on a pro-rata basis led to adoubling or even tripling of output in somedrilling programmes reducing costsdramatically. With such a system problems ofnegligence in care of vehicles and a generallowering of quality due to poor workmanshipneed to be decisively addressed. This can bedone through close supervision, quality control

I

I

Drill Rig Zonation MapVillage Water Supply

Drill Rig Type.Manually OperatedSmaltMediumMedium LargeLarge

Map drown by UNICEF Nigeria9Seolember 1991

f mi major norm Niqtna nvm

and across the board payment cuts wherestandards are not maintained (cit.ibid).

Dodge et al. (1986) report similar dramaticimprovements in rates of borehole completionand consequent reduction in borehole costs in aUNICEF supported programme in Uganda.

Hand drilled boreholesIn a number of types of geological strata it isfeasible to drill small diameter boreholesmanually. The lighter hand boring equipment issuitable for drilling small diameter holes (up to100mm diameter) which can be used to assist inprospecting for water and in locating hand dugwells or larger diameter hand drilled boreholes.In a small range of strata 100mm diameterboreholes can be equipped with handpumps butusually a larger diameter is needed andboreholes up to 300mm diameter are possiblewith the heavy duty rigs. Such sizes allow areasonable thickness of gravel pack to beinstalled around the borehole screen. The use ofa gravel pack is not usually possible with thesmaller diameter holes but sometimes geotextilefilters or fine slotted wellpoints can be usedinstead. In some strata the use of a gravel pack,fine screen or geotextile is particularlyimportant in preventing abrasive sand enteringthe borehole and damaging the handpump. Thelarger perimeter of the gravel packed holesmakes them more suitable for slower yieldingaquifers.

A very comprehensive guide to all aspects ofhand drilled wells, covering location of suitablesites, design, to maintenance of pumps and wellsis given by Blankwaardt (1984). This reference,based on experience in a large hand drilled wellsproject in Tanzania, is strongly recommended asa source of information concerning the varioustechniques. Tanzanian experience has been thatthe construction time for a hand dug ring well is3 - 7 weeks, depending on depth and soilconditions, whereas a hand drilled tubewell isnormally constructed in only 3 - 5 days(cit.ibid.). Other useful information is given inpublications by DHV (1979 & 1985).

Figure 3.3.2.3 shows a graph of cost per metreversus depih for ring wells and tubewells in aproject in Tanzania (DHV 1985). A steepergradient to the graph beyond a depth of 6-7mcan be observed. This is due to the fact that

suction pumps can no longer be used fordewatering can no longer be used and moreexpensive deepwell and due to the cost of labourand transport of rings which increases rapidlywith depth (cit.ibid). So for an average depth of8 - 10m a ring well was found to be 2 to 2.5times as expensive as a tubewell (Blankwaardt,1984).

The "Vonder Rig' a hand auger rig developed inZimbabwe where it is now manufactured by V &M Engineering (Pvt.)Ltd. who export themthroughout the world. The rig has been usedextensively in Zimbabwe and about 250 weremanufactured and sold between 1983 and 1989.In March 1992 the typical cost for a completerig air freighted into Ghana was USS 2,500 (V& W Engineering, 1992). Similar rigs are alsomanufactured in Kenya.

The Vonder Rig (Figure 3.3.2.4) is a lighter rigthan that used by Blankwaardt but it can be usedto hand drill 170mm diameter holes throughsoils and decomposing rocks. It takes about 30minutes to unload and assemble on site and theinitial drilling rate is about 3m/hr. After 2 • 3mdepth the drilling rate reduces to 1.5m/hr in softformations and 0.5 - 1.0m/hr in harderformations. Hence a 6m hole can typically bedrilled in approximately 5 hours and a 12m holein 1.5-2 days. Several types of auger or drillingbit are available, including a hole saw which isdesigned for cutting through soft decomposingrock formations. Usually the augering takesplace in an uncased hole and a casing is addedafter completion but there is an adaptationwhich allows temporary casing to be installed asdrilling proceeds. (Morgan 1990)

Recent experience of hand drilled boreholes inNiger is reported by Naugle (1992). A set ofaugers can be locally manufactured in Niger bya skilled welder for about US$200 and some setshave already been successfully used for over 50boreholes with only minor repairs beingnecessary. The cost of the wells which are usedfor irrigation and hence do not have apron slabsis essentially the cost of the 150mm diametercasing (which in Niger costs about USS15/m),and labour (one days wages for two drillers orabout USS15). The screen necessary to excludesand consists of a section of slotted casingsheathed in geotextile filter. A well extending 4- 5m below the water table with a depth ot 10 -

M)


1

t

50.00

40.00

30.00

20,00

§5£ t 10.00t- * •W * 0O O-o 0 2

DEPTH IN M

8 10 12 14 16 18

Costs of shallow wells

Wortcinq pans ofthe Vonder Rig.A hand-ooarataddrilling ng.

12m can be installed in less than 5 hours insandy soils. Flow rates of up to 9m^ per hourwere found to be possible when using motorisedpumping from a 7m deep well. (Naugle, 1992).

There are several advantages to the use of handdrilled wells.

There is a greater cost effectiveness than handdug wells and the speed of completion is greater.The technology is well suited to full village leveloperation. Villagers can easily construct theirown boreholes. The amounts of work and localmaterials required are less than what is neededfor a band dug well and progress is faster.

The smaller rigs, like the Vonder, are verymobile since they can be easily carried to the sueby the community. It should be noted that whereconcrete is being used to line a hand dug well avehicle is often required to carry precast rings tothe site or to carry the considerable amounts ofsand, aggregate and cement needed for the on-site production of rings or in situ lining.

Operations can proceed throughout the wholeyear and no dewatering pump in needed. This isbecause the small diameter holes are relativelystable, or can be cased during drilling, to allowboring to proceed in strata well below highgroundwater levels usually without anydewatering. Such conditions can pose majorproblems with hand dug well programmeswhich are sometimes only possible during aperiod at the end of the dry season when lowergroundwater levels allow the dewatering anddeeper excavations which is necessary forsufficient penetration of the aquifer.

It is more feasible to reach some deeper, morereliable aquifers than when using hand dug welltechnology.

However there are also some disadvantages. Insome strata the smaller diameter holes (such asthe 170mm hole produced with the Vonder rig)will be unable to yield enough water for acommunity. (However the construction of asecond borehole to double the total amount ofwater available, may still be cheaper than diealternative of a hand dug well.)

The auger bits cannot penetrate hard rock, andproblems are usually experienced with verycoarse gravels or layers of pebbles.

Wash-bored (Jetted) TubewelLsWash-boring, also known as well jetting, is atechnique used to construct tubewells in strataconsisting of fine or medium sized sand. Itusually uses low pressure water (or sometimesmud) which is pumped down to the bottom endof a vertical pipe pushed into the ground so thatthe fluid washes the sand particles at the bottomof the hole into suspension. The flow of fluidthen carries the soil up the annulus between thejetting pipe and the wall of the hole to the sur-face. The pipe is raised and dropped and rotatedduring the pumping operations to improve therate of progress. Various attachments can beadded to the end of the pipe such as cutting teethor jet orifices to suit the soil conditions. If wateris in short supply recirculation pits can be usedas in direct circulation rotary drilling but thepump then needs to be tolerant to the passage ofsome solids. Mud is circulated instead of waterwhen drilling in ground where there would behigh water loses from the walls of the hole. Ifaffordable, biodegradable polymer muds areusually preferable to bentonite since the formerdo not permanently block the pores between thesand particles and hence do not adversely affectthe yield of the borehole. Figure 3.3.2.5 showsthe major components of a simple jetting rig.

The technique has been in use around the worldfor some time in particular by civil engineeringcontractors for installing 'wellpoints' fordewatering around excavations.

The use of wash-boring in developing countrieshas been under research for a number of years atthe National Institute of AgriculturalEngineering (N1AE) at Silsoe, Bedfordshire,UK. [Mentianu (1982), Carter (1985) andComish (1987)]. Two different systems havebeen promoted. The first NIAE method uses a100mm steel pipe which produces a hole ofabout 150mm diameter. When the equipment ishand held this has a practical depth limit ofabout 12m (Carter, 1985) beyond which theweight of the pipes become excessive. The largediameter of the jetting pipe allows the insertionof a permanent casing into the pipe before thepipe is withdrawn. Cheap 80mm outsidediameter (72mm inside diameter) flexible



VVZ.5"

Major components of a simple jetting rig

Figure I: The drill vwc a rimed, and its tor end is scaled Figure 2: The drill vtpe is forced down, deepening the well.i\ liter flows down to the base ot the well, loosening sin/ Tim top ot the drill pipe is open, allowing soil and water tothere. ifIln ollt

n

litswA1

' Ur.ll , , , ,*

g- tllJVJIlunulk-d with wdivrf(

I -. Well

I

perforated plastic field drainage pipe transportedas a coil has been used successfully in somewash bores as a screen. It is sheathed in 1000gauge 'layflat' polythene tubing or with a rigidnon-perforated tube to form an impermeablecasing where needed, such as for the top 3m ofthe hole to prevent the easy entrance ofpollutants. The annulus between the aquifer andthe screen is filled with coarse sand or finegravel pack.

The speed of jetting through coarse sand can beas fast as lm/minute but it is unacceptably slowin heavy clays being only about lOOmmAninute.Under reasonably favourable conditions 6m deepwells have been jetted using 2.7m^ of water. Ateam of four men is recommended for thistechnique who could jet between two and fourwells per day (Metianu, 1982).

The 1982 UK costs quoted by Metianu were:-the local fabrication of the jetting equipment(£100 materials and £400 labour); the purchaseof a 50mm, 5kw, centrifugal engine (£350);100m of layflat pump hose with quick releasecoupling (£200). So the total cost of reusableequipment suitable for a 12m deep wash borewas £1050. The cost of the casing materials forthe same depth came to only £5.

The second NIAE method uses a 50mmdiameter jetting pipe of light gauge steel. Thisallows the hand operation of the equipmentdown to about 30m depth, at which stage theweight of the drilling column becomes excessive(Carter, 1985). The small diameter jetting pipeprevents the internal installation of a permanentcasing or screen as was possible in method 1.Where necessary a temporary or permanentcasing (100 -125mm diameter) can be pushedinto the jetted hole as work proceeds, so that itfollows closely alter the jetting head. Where atemporary casing is used it allows a smallerdiameter permanent screen and casing to beinstalled inside it before the temporary casing isremoved. Coarse sand or gravel pack can becarefully poured down the annulus between thetemporary and permanent casing as the formeris slowly withdrawn. In self supporting strata nocasing may be needed during jetting and apermanent casing can installed after the jettingpipes are removed from the wash bore.

Comish (1987) reporting on the use of thesecond method in Mexico gives a materials costof USS65 for a complete 9m deep lined jettedwell with suction handpump (the latter costingonly USS34). Making allowance for plantrunning, maintenance and depreciation costs thefigure rises to about USS80 per installation.

Carter (1985) reports extensive use of jetting innorthern Nigeria where the flood plains, orfadamas, of the seasonal rivers are often severalkilometres wide (Figure 3.3.2.2). For the greaterpart of the year the rivers are confined torelatively narrow channels but they experienceseasonal flooding and hence the recharge ofshallow alluvial aquifers. A large number ofjetted wells, usually 8- 10m deep, have beeninstalled in such areas mainly for irrigationwater which due to the high water table can beextracted by powered suction pump.

SWS (1992) promotes a different jetting system(SWS Well-Jetting Technique) which directlyjets a wellscreen into the ground. The "selfjetting" modification to the screens uses twosimple valves' which allows the full force of thejet of water to exit from its lower end, ratherthan through the screen, to enable it to be jettedthrough the soil and into the aquifer. Whenjetting is completed the valves return to aposition which admits only water from the finesides of the screen and none from the un-screened jetting end. This wellscreen currentlycosts £75 in the U.K. Between 1983 and 1989more than 3,000 wells were installed in KanoState, Nigeria using this 'SWS well-jettingtechnique'. Commonly these were no more than10m deep although where necessary, it ispossible to penetrate through 30m of siltymaterials to reach a sandy aquifer. Most of thejetted wells were for irrigation water, withtypical yields of up to 5 1/sec with a 3.5 HP 2"inlet diameter powered pump. (SWS 1992)

SWS have also promoted the use of a secondjetting pipe used alongside a pipe with astandard Johnson wedge wire screen which islowered down the hole as jetting takes placethrough both pipes. Before the second jettingpipe is withdrawn coarse sand is added to thehole and this sinks through the rising water andsilt to reach the bottom of the wash-bore tolocally improve the porosity and to form astraining sand pack around the screen.

IIIIIIIIIIIIIIIIII•


•v Ilevel to >4_^ iwhich the•7^!".',.coarse i.L£""-i*sand

Jettingdown. Note that water ispumped into both the jetprobe and the well pipe.Right: Gravel packing.Coarse sand (A) sinksdown through the risingcolumn of water whilefine sand and silt (B) iswashed to the surface

(Wiseman , 1983) This is illustrated in Figure3.3.2.7.

UNICEF has also been promoting wash-boringin Nigeria but using a hand rotated bit and wateror mud pressurised by a 12HP pump. These areused in conjunction with a manually operatedtripod and 3-5 ton winch which allows longheavy strings of jetting pipes to be removed fromthe completed wash-bore. The equipment can belocally fabricated in about 3 days in a suitablyequipped workshop and is suitable for producinga 200mm diameter hole to depths even greaterthan 66 metres. Production rates of one 50mdeep wash borehole every 5 days are possibleeach requiring about 2.5m^ of water if it isrecirculated. The jetting equipment andmaterials for the subsequent completion of theborehole and handpump installation can betransported on a single axle trailer towed by aLand Cruiser or similar vehicle. The samevehicle can tow a bowser to supply water wherenecessary. One vehicle can serve up to 3 rigs. A3 kV generator, welding plant and submersiblepump are also required for each set of 3 rigs.(Donaldson et al. 1987)

Donaldson suggests that one geologist shouldbe able to manage two experienced drillers whocan each supervise three wash-bore rig teams,each consisting of one assistant driller and fourlocal labourers. The field testing in Imo State,Nigeria suggested that the technology wouldalso be suitable for riverine areas of Cross River,Rivers and Bendel States as well as the alluvialplains of the Niger, Benue and possibly oilierlarge rivers. In fact it is thought that it may beapplicable throughout 15% to 20% of Nigeria.(Donaldson et al. 1987).

Ismail (1992) reporting on recent use of thetechnology described above stated that thetypical cost of a 60m deep wash-bored tubewellis currently about US$1,500. In some instancesdepths of just over 100m have been reached.

Sludger MethodA variation of the usual jetting technique whichenjoys widespread use in the soft alluvial soils ofBangladesh, north-east India and in some partsof Nepal and Nigeria is known as 'sludging1.This method does not require a pump but relieson the reverse circulation of water and soilpanicles up a 40mm GI pipe as it is raised and

suddenly dropped into the ground using ascaffolding and lever. A man on the scaffoldingplaces his hand over the top end of the pipe toform a valve system. The valve' is closed as thepipe is raised and opened as it falls to allowwater and soil to cascade from the top of thepipe (Figure 3.3.2.6). The method is welldescribed by Smith (1988), Carter (1985) andWorld Water (December 1979). Wells may be asdeep as 50m and rates of progress can be asmuch as 10m per hour. UNICEF is reported tohave funded the sinking of nearly 12,000km ofuPVC pipe into Bangladesh's soil by this method(Carter, 1985). The cost of each well includingthe installation of the handpump was US$100 in1979.

Advantages and Disadvantages of usingWash-boring (Jetting) to construct Tubewells

The technique usually uses inexpensivelightweight equipment which is easilytransported.

Most methods can be used in running sandswhich can be difficult materials in which toexcavate hand dug wells. Since no dewateringtakes place during construction instability ofgranular aquifers is not usually a problem andthe additional positive head of the water addedto the hole helps to provide additional support tothe walls of the hole. Where necessary drillingmud can be used to provide additional support.

The disadvantages include the type of stratawhich wash-boring can penetrate. These arelimited usually to unconsolidated silts, sandsand fine gravels. The presence of thick layers ofclay leads to very slow progress and coarsegravels may not be lifted by the rising water andmay wedge against the jetting pipe.

In some strata the loss of jetting water can be amajor problem unless adequate quantities ofwater are locally available or can be transportedto the site to replace the lost water. The use ofdrilling muds can overcome this problem buttheir cost and availability often limits their use.

3.3.3 Raising Groundwater to the Surface

Although there are many successful methods ofraising groundwater to the surface for thepurposes of this research it is necessary to

I


investigate those which are cost effective forremote, scattered communities. For this reasonelectric powered pumps, whether from mains,generators or photovoltaics have been omittedfrom the discussion.

Raising water from hand dug wellsBucket and RopeThe problems of users contaminating the waterin hand dug draw wells has already beenmentioned, and to prevent Guinea Worm larvaentering the well it is important that there is noopportunity for spilt water which splashesagainst lesions on infected people to re-enter thewell. Faecal contaminants however can alsoenter on buckets and ropes which are handled byusers with dirty hands or which come intocontact with the ground. Some wind blowndebris and dust can also carry pollutants intouncovered wells and hinged access covers aresometimes used to prevent this.

Whatever system of drawing water is to beadopted at a well it has to be acceptable to theusers and most sanitary methods only allow oneuser to draw water at a time. Some communitiesstrictly enforce the use of only one communalbucket for drawing the water and this is hung inthe well or placed upturned on a post after useso that it does not come into contact with theground although flies may still bringcontaminants onto the bucket.

A windlass has the advantages that the rope cannot come into contact with the ground and thebucket and rope is only handled once by theuser. Figure 3.3.3.1 shows a simple windlasssystem and instructions to a community used inZimbabwe.

If a windlass can not be afforded, or if it isdesirable that more than one person uses thewell at the same time, then the use of one ormore rolling bars or pipes fixed across the wellnear to the edge makes water lifting easier and ithelps prevent the chaffing of ropes on the edgeof the raised wall.

In some situations a 'shaduf (a pivotedcounterweishted pole from one end of which thebucket is suspended by a rope or flexible pole) isappropriate and like the windlass this alsoprevents the rope coming into contact with thearound (Fiaure 3.3.3.2)

The 'Blair Bucket Pump' described on the nextpage can be used in a casing installed in a handdug well (Figure 3.3.3.6). An interesting designfeature of the casing is that it is closed at thebottom end but it has one 8mm diameter hole inthe side, near to the bottom of the well whichadmits water fast enough to keep up with theremoval of water by the pump. The advantage ofthis hole is that it does not allow much water toflow from the casing into the well chamberimmediately following the impact of the bucketon the water. This reduces the chances of anyconiaminants on the bucket polluting the wholewell because they are quickly removed from thecasing with the water drawn out by the bucket.(Morgan, 1990)

Some interesting ideas about a purpose madesteel cage bucket holder to allow the use of avariety of containers as buckets and increase thelife span of well bucket handles are given byCarty(1990).

HnndpiimpsAll hand dug wells can be equipped withhandpumps and undoubtedly, providing aconcrete cover slab to a hand dug well, orconstructing a 'cased well', and installing ahandpump is the best way of protecting thegroundwater from contamination. Howeverwhere wells are covered and equipped withhandpumps then the provision of an lockableaccess hole in a small raised parapet is alsorecommended (Figure 3.3.3.6). This will allowbuckets to be used in an emergency if the pumpcan not be repaired, but every effort must bemade to make sure that the type of pump andprocedures for maintenance and repair makesuch a breakdown very unlikely.

Hand dug wells are often used where thegroundwater is relatively near to the surfacemaking the use of simple handpumpsappropriate.

If the water is always within about 7m of thesurface then a suction handpump can be used.This need not be directly mounted above thewell but can be offset from it if bends are usedon the suction pipe (Figure 3.3.3.3). One of theproblems with suction pumps is the loss ofprune due to leaks from the suction check valve.This can lead to the practice of users using

You can dnnk dean water tram your upgradad wallby ksapino tne well ana tna bucket dean

"i.W

1. Keep ine buckat clean2. Hang the bucket on tna windlass3. Keep the well cover in place4. Kaap tne apron ana runoff clean5. Always usa tha tame buckat in wall6. Keep cnain wrapped arouna windlass

Figure 2. 'How lo look after your well' from Blair's Upgraded wellmanual for ficldworlcen.



contaminated water to reprime the pump therebyspoiling the quality of the water subsequentlydrawn from the well. A UNICEF sponsoredimprovement to this valve on the Bandungpump in Indonesia has led to dramaticimprovements in watertightness and valve lifewhich mean that even after extensive use theprime is retained for very long periods (Mudgal,1992). Similar improvements could no doubt bemade to other suction pumps to overcome thisweakness in their design.

In areas where the general trend is for thegroundwater level to fall, care needs to be takenwhen using suction pumps since they may soonbecome unsuitable because of the excessive lift.In this respect seasonal variations ingroundwater level are also important.

Another type of handpump particularlyappropriate to situations where the groundwaterdoes not have to be lifted more than about 12mis the direct action type of handpump. These arelight and easy to install and maintain at villagelevel and give high rates of delivery. Theoperator pulls and pushes directly on a T barconnected by a rod to the piston which is in therising main below the level of the groundwater(Figure 3.3.3.4). The rod often consists of ahollow plastic pipe which floats in the waterheld in the rising main causing a reduction inthe lifting force needed to raise the water(Figure 3.3.3.5). Two designs which have beenwidely used to date with good results are theTara and the Nira AF85. The latter is soon to bemanufactured in Ghana by a new companycalled Ghanira Ltd. (Baumann, 1992). Both usecorrosion resistant downhole components,mostly plastics, so are suitable for use in corro-sive groundwaters.

It should be noted that the widely used standardIndia Mark II handpump, which was designedas a deepwell pump, is not very suitable forwells where the total length of the 12mmdiameter operating rod string is less than 25m(Inalsa. undated). This is because the chain linkat to the top of the rod string is unable to pushthe rods down against the frictional resistancesometimes experienced between the piston sealsand the cylinder wall, and which in shallowwells can exceed the weight of the string ofstandard rods. The use of 16mm diameter rodscan over come this problem as long as the total

length of rods exceeds 15m, and the piston canthen move down under the force of gravity.Fixed link versions of the India MarkJI are alsoavailable but much simpler and more easilymaintained pumps such as those described aboveare usually more appropriate for shallow lifts.

Long rising mains suspended in large diameterwells can swing from side to side when thepump is operated and this can lead eventually todamage or failure of pump components.However the rising main should not be clampedto the wall of the well unless an open toppedcylinder with removable foot valve is used. If atraditional cylinder is used then a short sectionof casing pipe can be fixed near the bottom ofthe well so that the rising main or cylinderpasses through the casing. This casing limits thesideways movement of the rising main yetallows the main to be easily withdrawn formaintenance. An alternative is to clamp aprojecting rod to the side of the piston and thenduring installation this rod can be pushed intothe sand or gravel at the bottom of the well.

The reference giving criteria for the choice ofvarious types and makes of handpumps is'Community Water Supply:Tbe HandpumpOption' (Arlosoroff et al. 1987). IRC TechnicalPaper No. 25 (IRC, 1988) is also recommended

Raising water from horeholes/tubewells BlairBucket PumpA normal bucket and rope is clearly unsuitablefor small diameter tubewells. The BlairResearch Laboratory in Zimbabwe has for sometime been promoting the use of the 'bucketpump' mentioned earlier, which is based on asimilar principle to the rope and windlass(Figure 3.3.3.6,7 &8). This name *bucket pump1

is a little confusing since it is used by others todescribe a completely different type of pumpwhich uses a rotating drive shaft and an endlesschain of small buckets to lift water from a largediameter well. The design concept of the Blairbucket pump is quite different being similar tothat of a typical "bailer' used to remove waterand mud from boreholes sunk using the cabletool (percussion) method.

The Blair bucket has a steel windlass with oiledwooden bearings which are usually supponed bytwo purpose made steel support posts, although

• : • /

Tara

Grouno waiar lame

— Spoul

Phingw i V - i 1 - ^

J- Fool vaw

A direct action pump such as thelara doaa not havea lever handle and bearing. The pumpingmechanism is movsd to a cylinder below th* levelof the groundwater and when the handle is raisedthe plunger lifts the column of water to the surfacewhilst the loot valve opens to allow the cylinder torefill. The pump is not limited by atmospheric pres-sure, though it is most efficient at depths of up to15m. The pump rod is a hollow sealed PVC pipe,which, by using the effect of bouyancy, distributesthe force required tor pumping more equally bet-ween the up and down stroke.


I

1

Bucket Pump fitted on standard well.

Concrete head block

Well cover slab

350 mm diameter holein well slab

8 mm holedrilled in casing

Water level• Well lining(bricks or concrete I

125 mm class 6PVC casing

300 mm diameterconcrete footing

timber ones can be used to reduce cost. A chainfrom the windlass is used to suspend a long thinbucket inside the borehole casing. The originalbucket design used a 110mm diameter uPVCpipe 700mm long which gave sufficientclearance to the inside of a 125mm permanentwell casing but more robust steel buckets areusually in use; both types have a capacity ofabout 5 litres. At the bottom of the bucket thereis a simple valve which opens when the buckethits the water and closes when the windlass isoperated to lift it to the surface. At the surfacethe operator grasps the lifting loop at the top ofthe bucket, lifts it out of the borehole and placesit on a metal funnel unit attached to one of thelegs of the windlass. This unit has anattachment which pushes the valve up, causingit to open, so that the water in the bucket isdischarged into the user's container below thefunnel. Alternatively the bucket can be tipped todischarge the water through the top of thebucket but this procedure increases the amountof handling the bucket receives and hence therisk of contamination. When the bucket is not inuse it rests inside the casing with a cap coveringthe head of the casing. (Morgan 1990)

The cost of an all steel bucket pump unit and the125mm diameter uPVC casing suitable for a15m deep installation if air freighted into Ghanafrom Zimbabwe would cost about USS350 (V &W Engineering 1992). The pump could almostcertainly be locally produced more cheaply inGhana.

The time taken to draw water obviouslyincreases with depth and the windlass is moresuited to shallower depths of groundwater and itis rarely used beyond 15m lift. The smallcapacity bucket means that the windlass has tobe operated to raise and lower the bucket threeof four times to fill a normal sized container andthis can frustrate users especially with deeperinstallations. It is probably therefore onlysuitable for family or multi-family groups andMorgan (1990) suggests that it should only servebetween 30 and 60 individuals. However ifcheap methods of borehole construction are usedtogether with the relatively cheap bucket pumpit may prove to be economical to install a largenumber of them to meet the needs of a largercommunity. This would be necessary in any casein slow yielding aquifers.

Despite the need to always handle the bucket,field testing in Zimbabwe has shown that inpractice the resulting level of contamination in atubewell is not very high. Frequent testing of anumber of installations over a period of a yearshowed that the levels of Faecal E. coli in manycases were zero and that the count rarelyexceeded 15 FC/lOOml (Morgan, 1985). Thismay be in part due to the very high rate ofturnover of water in the casing and gravel packwhich quickly removes the contaminants andthe smooth surface to the casing both of whichreduce the potential for the multiplication ofbacteria compared to a draw well (Morgan,1990).

Field experience in Zimbabwe, where by 1988over 1500 units were in service shows that thechain should last for 6 years and the bucket forfour years when the pump is used by 30 peoplebut that these periods reduce to three and twoyears respectively for use by 60 people. Thechain wears more rapidly at the bucket end andthus may only require partial replacement. Theleading edge of the bucket wears the most and tosave cost this can be replaced by an new 'leadingedge/valve unit', which is riveted onto anexisting bucket. Usually all maintenance can becarried out at village level and spares are verycheap and easy to produce in-country. Becausethe pump is simple and easy to understand manyrepairs can be improvised using wire, bent nails,pieces of car inner tube etc.

Naugle (1992) reports the use of a 110mmdiameter bailer bucket but one 2m long with acapacity of 18 litre, which is used in Niger toprovide irrigation water from 140mm diametercasings. The bucket is raised using either a tallwinch frame or a simple shaduf. The secondoption is a very interesting idea which may besuitable as a cheaper, easily repaired liftingdevice for use with the Blair bucket to raisewater for domestic use. The large capacitybucket is also of interest in view of thelimitations on the rate at which water can beraised using the small Blair bucket which werediscussed above. However some aquifers will beunable to yield water fast enough for the largerbailer.

Suction Pumps. Direct Action Handpnmps nnrioermanentlv installed Jlisin" Mains


T

I

One cost effective way of installing thesehandpumps in suitable high yielding aquifers isto avoid the use of permanent casing to theborehole by installing only the suction pipe orthe rising main pipe in the tubewell. To give thegroundwater an easy flow path into the pipe itneeds to be connected to a suitable length ofscreen which may be wrapped in geotextileand/or surrounded with gravel pack or coarsesand. The SWS well jetting technique explainedearlier is one method of installing such a lengthof pipework and screen but it can also beinstalled in any self supporting borehole. In lessstable strata installation is possible usingtemporary casing which is extracted as thetubewell is carefully backfilled with coarse sandor fine gravel.

The valve and piston of a suction pump are bothabove ground so they are readily accessible formaintenance and there is therefore usually noneed to remove the suction pipe and screen fromthe ground and this pipe can therefore beinstalled in the manner describe above withoutcausing subsequent maintenance problems.

Direct action pumps usually have open-topcylinders which allow the piston to bewithdrawn through the rising main without theneed for this pipe to be removed from aborehole. If the footvalve can also be removedthrough the rising main, as for example ispossible with the Tara handpump, then the usualmaintenance procedures will not be affected bythe rising main acting as a permanentlyinstalled casing. In fact the Tara has beeninstalled in this manner in a large number ofinstallations in Bangladesh where the pump isproduced. Unfortunately most of the other directaction handpumps such as the Nira AF85, theWavin and some types of the Blair do not havesuch extraciable foot valves.

One disadvantage of the directly installed risingmain system is that the screen, cylinder andrising main cannot be removed from the ground.It should be noted that removal of the screenshould never be necessary if it is properlydesigned for the grain size of the aquifer and thewell is properly 'developed'. (Development isoften achieved by pumping water out at a highrate in stop start surges to create a naturalgraded filter in the aquifer around the screen. )

The cylinder should not usually need replacing.It could wear if it was abraded by sand particlesbut these should be held back by a properlydesigned screen. Even if the cylinder did wearthe rapid movement of the piston would meanthat the pump would usually continue to deliverwater. In fact some direct action pumps like theBlair have no seals at all between the wall ofcylinder and the piston block.

In direct action pumps where the cylinder is anunlined uPVC pipe, should such excessive weartake place that the pumps performance wasbadly affected, it would be possible to shortenthe operating rod to make the piston operate inan unworn section of the pipe. The designlength of the cylinder pipe could be made extralong to allow for such an eventuality if it wasthought likely to be a problem.

Replacement of the rising main could benecessary it became worn due to abrasion fromthe rising main. This is most unlikely to happenand could be guarded against by havingreplaceable sacrificial centralisers on the pumprod.

One possible problem with the permanentlyinstalled rising main system relates to the needto ensure that it remains as straight as possibleduring the backfilling of a tubewell. However,although the 63mm plastic pipe often used forrising mains is fairly flexible problems in thefield with bent rising mains have not beenreported in the literature read during thepreparation of this report.

Deepwell PumpsIn the past the performance of many handpumpsused with community water supply systems wasoften very poor. In recent years considerableeffort has therefore gone into designingimproved handpumps to overcome the mainproblems of the older designs of pump(Arlosoroff et al. 1987). The development ofunproved suction pumps and new direct actionpumps has already been referred to but beyond12-15 metres lift a direct action pump usuallybecomes too difficult to operate and deepwellpumps are needed.

Deepwell pumps, like the direct action pumps,have a cylinder below the ground water table,and usually have a lever handle to make it

convenient for the operator to exert enough forceto lift water from depths much greater than15m. However as the lift increases so does thedifficulty experienced in operating the pump andthe increased forces on the pump componentsincrease making it more difficult to design aneconomic pump suitable for village-levelmaintenance. Most types of deepwell handpumpcan be used up to lifts of about 25m but a muchsmaller number are suitable up to 45m and veryfew are satisfactory at lifts much beyond thisdepth.

Arlosoroff et. al. (1987) compare a large numberof different handpumps in discussing theimportant points to be considered whenchoosing between handpumps. Briefly these are:

The maximum pumping lift. Obviously the pumpneeds to be suitable for the maximum lift(including drawdown and seasonal variations).Although deepwell pumps can be used in therange of lifts up to 15m the use of the simplerdirect action pumps should be considered up tothis depth.

77ie daily output required per pump. Usually itis prudent to limit the number of people relyingon the pump to a maximum of 250 and ifpossible to have two pumps available in a villageso normal users of a broken down pump can usethe other while their pump is beinc repaired.Some designs of handpump are oniy suitable fora few families and should never be used ascommunity pumps.

The type of maintenance system to be used. TheVillage Level Operation and Maintenance(VLOM) system is showing the most promise asa sustainable maintenance system forhandpumps. This system requires a handpumpspecially designed for easy maintenance byvillagers and an effective means of distributingessential spares.

The minimum rating needed for corrosionresistance. Aggressive groundwaters are aproblem in a number of the countries whereGuinea worm is endemic

The use of expensive corrosion resistantstainless steel rising mains and rods hasfrequently been necessary to replace galvanisediron components normally used with many

designs of handpumps. Although expensive, thestainless steel pipe has the advantage that it ismuch lighter than galvanised pipe, makingcylinder removal from the borehole much easier.In 1990 the modified India Mark II, which usesstainless steel pipes and rods had a unit price ofUS$900 for a 20m setting in Ghana (Baumann,1992). The use of plastic rising mains isconsidered below.

The minimum rating needed for abrasionresistance. Boreholes should be properlydesigned to prevent the ingress of sand particleswhich can damage pump components. Wheresand is likely to enter a pump, such as whenpumps are to be installed in old poorlyconstructed boreholes, a handpump withabrasion resistant features (such as nitrile rubberpiston seals) should be chosen.

The potential for manufacture of the wholeliandpump in the country of use should beinvestigated. If this is not possible then it maystill be possible to manufacture the sparescommonly needed for its maintenance. In-country manufacture of handpumps and sparesis very important to the sustainability ofhandpumps in developing countries. Aspreviously mentioned the Nira AF85, which hasa very good track record in the field is shortly tobe produced in Ghana. Mali is reported to bemanufacturing the India Mark II. The SWSsuction pump is being assembled from a mixtureof local and imported components in Nigeria. Itis not known if any other countries in whichGuinea worm is endemic also producehandpumps which have given sustained goodperformance but the existence of localmanufacturers should always be investigated.

Cost effectiveness also depends uponstandardisation. If a country standardises on oneor just a few types of handpumps to suit thedifferent operating conditions experienced in thefield, then it makes proper stocking of sparesand training of handpump caretakers andmechanics much easier. However the choicemust be made very carefully in view of the manyfactors which need to be considered to find thebest pump. It should be noted thatstandardisation of pumps can lead to increasedpotential for in-country manufacture because thedemand for the pump, or pumps, chosen willincrease. The experience of countries such as

I

T

1

Bangladesh (with the Tara) and Pakistan (withthe Afridev) showed that rather high initialproduction costs were rapidly reduced when thepumps were produced in bulk and differentmanufacturers were encouraged to compete.

Plastic Rising MainsPlastic rising mains can be very cost effective,particularly in corrosive groundwaters where thealternative would be costly stainless steel.However there are a number of particularaspects that need to be considered if plastic is tobe used.

In particular it should be noted that a number ofproblems have been experienced with many ofthe pipe jointing systems tried to date and withthe connection between the plastic pipe and themetal pumphead. On long plastic rising mainslong term creep, leading to a change in theposition of the cylinder relative to the piston,can also be a problem with short cylinderlengths.

Glued joints to the rising main can be used withpumps which have open top cylinders withextractable foot valves because the main doesnot need to be removed from the borehole. TheUPM pump which uses a lever, rope and pulleyto raise the piston in an open top cylinder isreported to have performed well with 57mmuPVC rising main at settings as great as 60m(Bauman, 1992). However to obtain a reasonablerate of discharge at such deep settings needs upto four people to operate the handle (PM1,undated).

Where a traditional large diameter cylinder is tobe used with small diameter plastic rising mainglued joints can not be used and the pipe willneed joints which can be dismantled to allow themain to be extracted to bring the cylinder, pistonand foot valve to die surface for maintenance.Normal threading of uPVC for pipe joints hasusually failed unless pipe wall thickness exceeds10mm (Baumann, 1992), but a connector withinjection moulded rounded threads has beenproduced by Entermap in Abijan, Cote d'lvoirefor use on 50mm outside diameter uPVC pipes.This is has been field tested with the India MarkII in Mali with pumping lifts of up to 25m, andindications are that this design gives reasonableperformance (Baumann, 1992).

It should be noted that the use of plastic risingmains of smaller diameter than the cylinderusually leads to compression in die pipe on theupstroke (due the pressure acting on the reducerat the top of the cylinder and to friction at thepiston seals) and followed by tension on thedownstroke (due to the whole weight of thecontained water). This can cause a 45m risingmain to extend and contract by as much as30mm. Since this means that the cylinder alsomoves by this amount this change in lengthadversely affects the rate of discharge (cit.ibid).

Polypropylene and ABS are less 'notch sensitive'than uPVC and some pumps make use of theseplastics. The Nira AF85 direct action pump useshigh density polyethylene (HDPE) pipe. TheVergnet foot pump which operates usinghydraulic pressure to inflate a rubber diaphragmuses twin polypropylene hoses which aredelivered coiled. Consallen (1992) is unique inusing threaded ABS plastic for the rising mainsto its pumps.

In Nigeria threaded polyethylene joints mouldedto uPVC rising main pipe with an outsidediameter of 75mm and internal diameter of65mm has been tried to produce an open toppedIndia Mark III to allow die 2.5" diameter pistonto be withdrawn through the rising main when itneeds maintenance. The current situationconcerning these joints which can just fit insidea 101mm internal diameter casing is unknown.

Current research and testing by the Consumers'Research and Testing Laboratories (CRTL), UK,under World Bank/ODA sponsorship, hasidentified two good dismandeable jointingsystems suitable for 63 mm outside diameteruPVC pipes, suitable for use in borehole casingwim an internal diameter of 102 mm. Thesejoints are currently being field tested and thedesign should be available for potentialmanufacturers widiin a year. They are alsocarrying out work concerning possibleimprovements to the design and manufacture ofthe uPVC pipe material and the solvent cementsused for joining some handpump rising mains.The behaviour of pump rods, including a designmade from glass reinforced plastic (GRP), andpump rod centralisers is also being investigatedto overcome the problems of buckling which canoccur on the pump downstroke and which

sometimes can lead to wear of the rising main(Deveraux, 1992).

Afridev HandpumpOne good lever operated deepwell pump whichhas emerged from a large amount of research,design and field testing is the Afridevhandpump, the design of which is now in thepublic domain. It is in production in severalcountries including Kenya, Ethiopia andPakistan. It is a VLOM pump which has anopen top cylinder which is used with 63mmoutside diameter solvent cemented uPVC risingmain. The pump bearings are made frominjection moulded plastics which are cheap toproduce and easy for the villagers to replaceevery year. Unfortunately field experience hasshowed that under heavy use the bent stainlesssteel hook and eye rod connector design is notal .vays reliable and the mild steel ones havegiven very poor performance. A forged stainlesssteel design gives much better performance andbetter quality control during manufacture wouldprobably considerably reduce the rod failure ratewhich can be as high as 10% (Baumann, 1992).A useful feature of the Afridev pumphead is thefact that it has a flange which is suitable forbolting directly onto an India Mark II should itbe necessary to replace such a pump.

It should be noted that a number of failures ofthe Afridev in the field have been due to somemanufacturers not complying with the approvedspecification and due to poor quality controlduring manufacture (Wurzel 1992). Pumppurchasers need to check these aspects whenpurchasing fro a manufacturer.

The current price in Ghana for an Indianproduced Afridev, complete for a 30m setting, isabout US$700 (Baumann, 1992).

Experiments are been made to produce a directaction version of the Afridev, using the samecylinder and footvalve components as thedeepwell version. If successful this is an idealsituation since identical downhole componentscan be then be used for both types of pump tocover the whole range of operating depths up to45 metres. Unfortunately at the time of writingno detailed information lias been foundconcerning this development which wasreported to have taken place in Kenya andMozambique.

Using the rising main ns a permanent casingDirect installation of direct action pumpswithout using a permanent borehole casing wasdescribed above. Open top deepwell cylinderpumps with extractable footvalves can beinstalled in a similar manner if the aquifer issuitable. Experiments have been carried out inKenya using the Afridev in this way but to dateno details of their performance have beenobtained. Consallen (1992) produce aninteresting cylinder for their 'Consallen DirectlyInstalled System1 CDI. This unique designallows the piston, the footvalve or if necessarythe whole stainless steel lined cylinder to beextracted through the rising main. Thisovercomes one of the foreseeable problems ofusing rising main without a borehole casing.Other possible problems with directly installedrising mains for deep well pumps are similar tothose discussed when considering shallowpumps above.


IIIIIIII

iiiiiiii

4. COST EFFECTIVE RAINWATEREXPLOITATION

SECTION SUMMARY

Rainwater is pure, and if collected from anelevated surface such as a roof and stored in anhygienic manner it is usually suitable fordrinking without treatment and offers no risks ofthe transmission of Guinea worm disease.

• The main disadvantage of its use is thatthere is usually a need to store a largevolume of it for consumption during periodswhen there is no rain. If only water fordrinking purposes is stored and traditionalsources are used for other purposes thevolume of storage is greatly reduced.

• If rainwater is captured from the roof of theuser's house there is the advantage that thestorage can be placed conveniently close tothe point of use and if the storage is aboveground, the water can be collectedhygienically from a lap or hose.

• Where roofs to buildings are not suitable,cheap temporary catchments of wovenplastic sacks may be feasible.

• The annual rainfall figures and rainfallpattern for many of the areas of countries inWest Africa where Guinea worm is endemicare high enough to suggest that rainwatercatchment is a potential source of potablewater.

• Simple cement jars with hygienic draw-offhoses can provide cost effective storage forpotable water supply

• Where the materials are available in thecountry the use of galvanised iron sheetsand galvanised wires for the DANIDAdesign of suspended gutters is likely to becost effective.

4.1 Raised Catchment Surfaces

This section considers the cost effectiveexploitation of rainwater collected from elevatedsurfaces as a source of water for domestic use.The use of rainwater collected from ground levelsurfaces is discussed in the Section 5.6.

4.1.1 Roofs

The roof to a domestic dwelling is thecommonest forms of raised catchment. There area wide variety of roofing materials and roofshapes, some being more suitable than others forrainwater catchment:

Mud roofs and thatch roofsRoofs of mud and thatch are generally not verysuitable catchment surfaces because they usuallycontribute contaminants to the runoffthroughout the period of rainfall and theyusually adversely affect the taste and colour ofthe water. Furthermore thatch is notimpermeable and some rainwater is retained inthe roofing material reducing the amount whichcan be captured for domestic use. It is usuallydifficult to provide gutters for thatched roofs,especially for circular buildings.

However some communities have traditionallycollected water from thatch and mud roofsurfaces so their use for drinking water ispossible if the water is accepted by users, butbecause 'first flush' diversion systems will notremove much of the contamination on the roofthe bacterial quality of the water is likely to belower than that from impermeable surfaces. Theuse of charcoal and sand filters to remove colour, taste and other contaminants is theoreticallypossible (Edwards et al. 1984), but it is notknown to have ever been practised on a widescale. It should be considered impractical andunsustainable until proven otherwise.

Tiled or sheeted roofsTiled or sheeted roofs are much better thanthatch ones since nearly all the water falling onthe surface runs off. That is they have a high'runoff coefficient', which is the ratio of theamount of water collected from the roof to thatfalling on the roof. However, if an impermeableroof has very poorly installed guttering, orguttering which is too small for peak flows invery heavy storms, it will not be possible tocollect all the water so the overall runoffcoefficient should take into account theefficiency of the gutters and not just the roofingmaterial.

Where galvanised corrugated steel sheets are tooexpensive for rooting some communities are

41

now using large fibre cement tiles producedlocally by co-operatives using simply operatedvibrating tables (Parry, 1992) but during theliterature search carried out for tfiis presentreport no projects were noted where thepromotion of rainwater catchment wasaccompanied by the promotion of moreaffordable impermeable roofing materials. Thisis probably because in rural areas very fewpeople can afford new roofing materials and arainwater catchment system at the same time.However, in areas where only thatch is presentlyused, rainwater schemes using roof catchmentswill not usually be feasible until impermeableroofing materials are affordable.

4.1.2 Other Raised Surfaces

Where suitable roofs do not exist on dwellingsor public buildings in a community, sometimesroofs without buildings have been built tocapture rainwater. These shaded areas can thenbe used by communities for activities such aspublic meetings. Such catchments, 5m by 10min size, have been used in northern Togo wherethe cost was about US$900 (O'Brien, 1990a).These are combined with four 6m^ aboveground cement brick storage tanks costing aboutUS$300 each to provide drinking water for up to44 people at a unit cost of USS53 per personserved (cit. ibid) but all these figures exclude thevillage contributions of labour and localmaterials and the project-provided skilledlabour. Lee and Visscher (1990) also discussthis particular programme and they note thatalthough 60% of roofs in many of the ruralvillages are of corrugated sheets, that the projectrejected the possibly cheaper cost of usingexisting roofs, due to the increased logisticalrequirements. They also note that cheapermethods of tank construction may now beavailable. The unit cost quoted above arecompared by O'Brien to a cost of US$47 perperson served elsewhere by deepwell handpumpsystems which were not an option in thesevillages.

Smaller sizes of elevated sheeted catchmentsare sometimes designed to become the roof to anew house as in Botswana (Lee & Visscher,1990) but the provision of such catchments on alarge scale is likely to be very expensive.

In one part of Kenya, a very cheap, temporary,6M^ elevated catchment made from wovenplastic sacks which have been split open, sewntogether and stretched over a frame has beenused by villagers living in thatched houses toprovide a catchment surface (lies, 1989). Thissystem is shown in Figure 4.1.2.1. This idea isworth replicating elsewhere where there are onlythatched roofs. It would also provide additionalcatchment area to supplement the roof area of atiled or sheeted roof to increase the volume ofwater captured from each storm.

4.2 The Quality Of Stored Rainwater

As previously mentioned human faeces are thesource of many endemic diseases in developingcountries. Unfortunately birds, reptiles andanimals can carry human faecal pollution fromground level onto a raised catchment, and theycan deposit their own faeces on the roof.Windbome dust may also carry faecal pathogensonto a raised surface. Therefore the debriswashed off a roof during rainfall will not onlyaffects the appearance of the water but also itsbacterial quality.

Fortunately a number of studies have shown thatalthough bacterial contamination in water fromroofs is not uncommon, die level ofcontamination is usually very low and does notpose a major health risk (Wirojanagud et al.1989). If a 'first flush' diversion system is used,such as those shown in Figures 4.2.1,4.2.2 and4.2.3, it prevents the initial flow of heavilycontaminated water from the roof entering thestorage vessel and the quality of the stored wateris then considerably improved. If the water inthe storage vessel it JS carefully protected fromcontamination until it is consumption then itsquality usually improves further during storagebecause many of the bacteria soon die off in thetank (cit.ibid).

The use of a mosquito mesh screen, preferablyof stainless steel, at the entry to the tank is alsorecommended in order to intercept debris afterthe first flush system has ceased to operate. Anangled screen as shown in Figure 4.2.4 has theadvantage of being to some extent self cleaning.The removal of organic pollutants, whatever themethod, is advantageous because their presenceaids the survival of bacteria in the storage tank.

riiiiiiii

42

TI

iI

11

Streched woven plastic sack catchment ('sarlsa') used In Kenya(Source: Lee & Visscher 1990 after lies 1989)

warn

{^MDirty »

S K M I MPlug

* ;

t

. *

~

•no «nv n a n ) dHns. mm Iw¥ ump I U W ttwi 9m link

b: u w l g H . a M i m i M c: nnww o> ln» mugcemHimucnlHtaHin. pus CMKU em «w WIMI andinto «w anK Mean* > M amp aanMnMunpis kill prapwmg a tor n* nan

Sump • typ* ot tlrat ilu»h' dlvartar

IIIIIII

ftouiaBU ipaui

Movaabl* downplpausad to divan tha l i n t tlusn

(Source: Atler VaananivmKit et al 1984)

Rainwater Harvesting

u •%•?>••

1 roof -2 gutter - . -3 down-pipe4 collection box

5 drain valve6 overflow pipe7 rainwater storage tank

UNICEF Burundi

'ML.

IIIIII

Angtod mMh ttmnw used tolnt«rc«pt largar dabrit in captured rainwaitr

IIIIIIII

IIIIIIIIIIIIIIIIII\

Unpolluted water which enters a storage tankcan be contaminated during storage. The roof ofa covered tank needs to be of a good standard toeliminate small animals, amphibians, dust andmosquitoes. Uncovered tanks promote thegrowth of algae in the water and the breeding ofmosquitoes which can transmit Malaria.

Poorly constructed buried tanks can allowpolluted water from outside the tank tocontaminate the contents.

The way in which water will be withdrawn fromthe storage tank needs to be carefullyconsidered. If it is collected directly incontainers dipped into the water contaminationcan easily occur, and for above ground tanks dieuse of a tap fixed to a pipe passing through thewall of the tank is recommended (Figures4.4.2.2b and 4.4.2.4). Alternatively a siphonhose, with or without a tap, can be passedthrough the cover of the tank to deliver water(Figure 4.4.2.2c). Water collections systems forbelow ground tanks need to be similar to thosefor hand dug wells (see Section 3.3.3). As wasexplained in Section 2, attention also needs to bepaid to the proper storage and use of drinkingwater in a home to ensure that the good qualitywater drawn from the mam storage tank doesnot become polluted by the users beforeconsumption.

Roofs, gutters, screens and tanks all needcleaning from time to tune to avoid the build upof debris and contaminants.

From the discussion in Section 2 it should beclear diat if people only consume rainwater froma hygienic storage tank then there is no risk ofthem being infected with Guinea worm larvae,and the chance of being infected by otherpathogens is very low. However the change inhabits necessary to make this occur in practicedo not come automatically with the provision ofa rainwater catchment scheme, and appropriateand sustained health education is necessary tobring about these changes especially if the tasteof rainwater is new to users.

4.3 The Technical Feasibility Of A RainwaterCatchment And Storage System

The technical feasibility of a rainwatercatchment and storage system to fully meet agiven water demand depends on four factors:-* the area of the catchment available* the rainfall pattern over the year* the pattern of demand over the year* the volume of storage available

43.1 The area of the catchment

The product of the area of the catchment, therunoff coefficient and the plan area of a roofgives the volume of water captured from it. Forexample if a corrugated galvanised steel sheetroof is 3Qm2, and the runoff coefficient is 0.9,then if during a year 729mm of rainfall falls onthe roof a maximum volume of

V = 30X0.9X729 =captured.

19,683 litres can be

If this maximum volume is less than the totalamount of water required during the year thenthe demand can not be met unless the catchmentarea is increased.

4 3 2 The rainfall pattern over the year

It is important that reliable monthly, or evenbetter daily, rainfall records are available over anumber of past years if an economic raincatchment scheme is to be designed for an area.Often the location of the nearest meteorologicalrainwater gauge(s) will not be convenient to theproject area and initially only approximaterainfall figures can be used. However rainfallgauges should be installed in the project area assoon as possible to monitor daily rainfall so thatdesigns can be amended if the amount of rainfalling is very different from that expected.Fortunately very cheap, transparent, graduated,wedge shaped plastic rainwater gauges are nowavailable from which daily depths of rainfall canbe read by any literate person. This makes itfeasible to collect reasonably accurate recordsfrom a number of gauges in an area to getaverage figures and to compensate for missedrecords at any station.

The relationship between seasonal rainfallpatterns and die risks of being infected withdracunculiasis is discussed in Section 4.3.4 (seealso Figure 1.4).

Lee and Visscher (1990) in their study ofrainwater harvesting in four African countries(Botswana, Kenya, Mali, Tanzania and Togo)noted that in general rainwater harvesting(rooftop and ground level catchments) haspotential for wider expansion in areas where theannual rainfall is between 200mm and 1000mm.

In areas with above 1000mm of rainfall theysuggest that there is usually no need forrainwater harvesting due to the plentiful natureof surface and sub-surface sources. The lowquality of the 'plentiful' surface water sourcesshould have been noted at this point in theirdiscussion. Unless an appropriate method oftreatment is available for surface water, ifgroundwater sources are not available, thenrainwater will always have good potential forsupplying potable water. In such situations it issuggested that the higher the rainfall the greaterwill be the potential for the use of rainwatercatchment. The treatment of surface watersources is particularly difficult on a small scaleso rainwater is especially suitable for smallvillages and scattered households of rural areaswhere no groundwater can be exploited.

Lee and Visscher also quote Ongweny (1979)who suggests that for drinking water suppliessurface catchments should be widely applied inthe sub-humid and semi-arid areas, and rooftopcatchment systems in the humid and sub-humidareas.

In view of their discussions cited above. Lee andVisscher illustrate 'zones where rainwaterharvesting can be well applied' with the use amap of Africa showing the zones where rainfallis between 200 and 1000mm. The map in Figure4.3.2.1 may be more helpful. This focuses onWest Africa, where Dracunculiasis is mostendemic. A comparison of Figure 4.3.2.1 withFigure 4.3.2.2 shows that most of the land inendemic countries lies in a zone where theannual rainfall exceeds 1000mm and virtuallyall the remaining areas of these countries are inthe zone between 250mm and 1000mm. It issuggested that this is indicative of the highpotential for rainwater catchment to supply thepotable water needs of those small communitiesin these countries which cannot be economicallyor sustainably served by the alternatives ofgroundwater or treated surface water.

Surprisingly there is not much reported inliterature published in English about rainwatercatchment in these countries other than in Togo(e.g. O'Brien (1990a), Lee & Visscher (1990))and Mali (e.g. Lee & Visscher (1990)). Figure4.3.2.3 shows some representative meanmonthly and annual rainfall figures for a fewselected stations in West Africa. It can be notedthat in general in those areas north of latitude9°N there is only one rainfall peak, usually inAugust, and the wet season becomesprogressively shorter with an increase in latitude(MHCE undated).

4-3.3 The pattern of demand over the year

A family's pattern of demand for water during ayear is often assumed to be constant, whereas inpractice it is likely to vary, especially if water isbeing used for livestock and small-scaleirrigation in addition to uses in the home. Sincehigh quality water is not needed for the lattertwo purposes cheaper methods of collection andstorage of rainwater can be used for them andwater for these uses will not be consideredfurther in this section of the report. In fact thewater for livestock and small-scale irrigationcan usually come from unprotected traditionalsources.

The amount of drinking water consumed perperson varies somewhat with climatic conditionsand drinking habits. Usually it falls within therange 3 - 5 litres/person/day. When consideringwater for all domestic uses, including drinking,the minimum domestic demand is 5 - 25litres/person/day but for a good standard ofhygiene a total volume of 30 - 40 litres/per-son/day, or higher, is recommended.

If, as is usually the case, the storage of all waterused for domestic purposes is too expensive,then as an absolute minimum only water forpotable purposes needs to be stored, and this isprobably the best starting point for a newproject. If the roof is of sufficient size, additionalstorage can be added at a later stage when it canbe afforded. However, the human problem of theconveniently available stored rainwater beingused for purposes other than for drinking needsto be addressed. If people are not disciplined thestored water may run out before the end of thedry season and they will have to return todrinkina water from traditional polluted sources. I

44

i 1 - 9 9 1000 - 9999 AKEAS NOTSEARCHEDTO DATE

RAINRN4.INWESTAFR1CAMtAM MONTHLY AND AMNUAL.

RAIWFAUS FOR, SSUCTBO 3TATWMS

10° W-

ASIOIAMI95S

IIIIIIIIIIIIIIIII

IIIIIIIIIIIIIIIIIII

These will put them at risk of infection fromGuinea worm larvae and other pathogens.

The usual lack of discipline in dedicatingrainwater to potable use only is the main reasonwhy catchments serving more than one familyrarely work in practice but it has worked withsome, but not all the communities in the Togoproject cited above [Lee & Visscher (1990),O'Brien (1990a)]. However in many countriesthere has often been success in communitystorage at schools, hospitals and health centresbecause at such places use can usually be morestrictly controlled, by lockable taps if necessary.

The fact that a strict rule is imposed by a familyon the uses to which the stored water can be putduring the dry season does not necessarily meanthat these have to apply during the whole of therainy season. When, during a rainy season, atank becomes nearly full some water can bedrawn off to make space for additional rainwaterwhich would otherwise overflow. However theusers need to have advice on this aspect and aneasy method of monitoring the water level, suchas a transparent hose (which can also be used asa delivery pipe as shown in Figure 4.4.2.3),needs to be provided.

The particular implications of the typicalpatterns of rainfall in Guinea worm endemicareas on the minimum volume of storagerequired are discussed below.

4,3.4 The volume of storage

If during a typical year there is a long periodwithout rain then the volume of water stored tomeet demand throughout this period obviouslyneeds to be equal to, or greater than the volumenormally consumed by the users during that dryperiod. However the patterns of rainfall anddemand during a year often make calculations ofHie minimum volume required rather morecomplicated, but there is useful guidance on howto calculate this volume in a large number ofpublications dealing with rainwater catchment[e.g. Gould (1991), Institute for Rural Water(1982), Pacey & Cullis (1986), Edwards et al.(1984)]. It is suggested that graphical methods(mass curve technique) based on a chosendesign year' rainfall pattern are the simplestapproach (Skinner & Cotton, 1992).

The feasibility of storing drinking water onlyfor seasons when infected copepods can beingestedWhere the main purpose of a rainwatercatchment project is to avoid Guinea wormdisease the volume of storage needs only to besufficient to meet the drinking water needsduring the period in which the traditional sourceis likely to contain infected copepods.

In semiarid regions that have distinct wet anddry seasons and minimal rainfall within 3-4consecutive months, incidence of dracunculiasisoften coincides with the rainy season (Muller,1970). This means that volumes of storage maynot have to be very large since water drawn fromthe tank during the rainy season is soonreplaced.

In areas widi longer periods of rainfall,dracunculiasis can be transmitted for as many as8 months in any year but most cases occur in thelatter half of the dry season and continue intothe rainy season (Muller, 1970). This diseasepattern would require sufficient water from theend of one rainy season to be saved forconsumption from the middle of the dry seasonuntil the rainy season provided enough water toreplace the daily demand.

Figure 1.4 shows a simplified map illustratingthe peak seasons of transmission of Guineaworm disease. Note that Guinea worm is alsotransmitted just before and after these periods.Needless to say, the adoption of the idealisticpatterns of storage and consumption discussedabove is not at all straight forward and noreferences to projects promoting the adoption ofsuch practices has been found during thepreparation of this report. Also such a pattern ofconsumption will still expose people to faecalpathogens in the unprotected traditional sourceswhich they have to use when they areconserving rainwater, a situation which shouldbe avoided if at all possible.

4.4 Cost Effective Guttering And WaterStorage

If a suitable catchment surface, such as a sheetedroof, is already available then water can becollected by placing containers on the groundunder the edge of the roof to collect waterduring periods of rainfall. In fact many

communities already collect some water in thisway although it may not be used as a source ofdrinking water. However this collection systemis not suitable for collecting and storing all thewater from a storm, and the water collected inshallow containers can be easily polluted bywater splashing off the ground. More efficientand hygienic collection systems use some type ofgutter to collect virtually all the water runningoff a roof and to channel it to a suitably sizedstorage tank. Both gutters and storage vesselscan be expensive but there is potential for costsavings as detailed below.

4.4.1 Cost Effective Guttering

Supporting the gutterIn developed countries gutters are oftensupported by brackets fixed to the fascia boardwhich runs along the ends of the rafters justbelow the edge of the roof. However, in ruralareas of developing countries roof rarely havefascia boards so such a system is not usuallysuitable and other methods have to be used.Sometimes the gutter is supported on the top ofvertical poles, often using forked branches tohold the gutter. Sometimes a horizontal timberis also used to help support the gutter over largespans between poles because it becomes veryheavy when it is full of water. In other situationsgutters can be supported on wooden or metalbrackets fixed to the side of the rafter poles or tothe walls.

One idea developed by DANIDA in Kenyawhich shows great potential is the suspension ofgutters from a 'deflector plate' fixed to the edgeof the roof. This is illustrated in Figure 4.4.1. lb.Not only does this neatly solve the problem ofsupporting a gutter on a roof constructedwithout suitable rafters or a fascia board, but italso improves the amount of water captured bythe gutter during heavy rain as described below.The system is usually used with 'V shapedgutters but it could be used with any type ofgutter.

To obtain a sufficient rate of flow along a gutterit is usually necessary to install it so it has alongitudinal gradient. Tins gradient means thatthe level of the gutter becomes progressivelylower and consequently further below the edgeof the roof. With a long length of gutters therecomes a point when it can no longer be fixed to

the standard fascia board, and/or it is positionedso low that the rainwater running off the edge ofthe roof during heavy storms is likely toovershoot the gutter as shown in Figure4.4.1.1a. When a deflector plate and suspendedgutter are used this problem does not occur sincethe water hits the plate and then f.lls downvertically into the gutter. The fact that with thissystem the gutter does not have to be very closeto the edge of the roof means that it can be laidto a much steeper gradient, and therefore with agreater water carrying capacity, than is possiblewith one fitted to a fascia board.

Gutter materialsGutters can sometimes be made from locallyavailable materials such as split bamboos. Sawntimber planks nailed together to form a V or 'U'shape or variously shaped metal gutters made bytinsmiths from recycled metal sheet can also beused, but all these materials have relatively shortlives. Purpose made uPVC gutters although rotand corrosion resistant are rarely available andthey, or the alternative of a plastic pipe cut inhalf, are usually quite expensive. Semicirculargutters made from galvanised steel bent to shapeby tinsmiths, and often with bent edges to givelongitudinal stiffness, are usually more readilyavailable (at least in urban areas) but often theyare also expensive.

To reduce water losses all gutter joints should beleakproof and this can be a major problem withmany of the locally produced gutter designs.However if no suitable sealing material isavailable wastage can be considerably reducedby using a reasonable length of closely fittingoverlaps and a good longitudinal gradient.

The simplicity of the DANIDA 'V gutter systemmentioned above has much to commend it. Itcan be manufactured from flat galvanised steelsheets by people with very little training usingsimple timber forms and clamps. Furthermore,since it is installed to steeper gradients thanconventional gutters its erection does not required high level of skill. The suspension system alsohas the advantage of allowing easy adjustmentsto levels to be made during installation ormaintenance to correct any faults that develop.The manufacture, design and installation of this'V gutter system is well described by Nissen-Petersen and Lee (1990a). The same reference;dso describes a 'U' gutter system which is

IIIIIIIIIIIIIIIIIII


Watftr ovOThootng coovsnoonai gutterAt low •nd of sloping t#ngtft

Waur droctad into gutter by detector plat*which also supports wires wuatmnang me

b: Detector M M M ana(Sourc«:tMMa on

a: Convwioontt 9Jtur

Figure Advantages of using flow deflectors end suspended gutters

gutter1990s.

FontiwocK for

Sac*

a: Shaped «*cfc lilted withQfeaflUtaV fTUttaffleal J

Figure

b:Shapaa blocks lar inand ptasiMM wnn mud

c. SsOonM matat tn ant<[tgtptaatarsawmmua)

Examples of tormwork used for cement mortar |ara

attached to a fascia board on one edge and issupported by a wire fixed to the roof sheets onthe other.

In ruraJ areas of developing countries there arelikely to be only a few homeowners who canafford a complete system at the outset. Even ifsomeone could afford it, until they are convincedof its usefulness they will not be willing to investin such a system. However it should be notedthat it is not initially necessary for a houseowner to gutter the whole of his house. If he orshe is offered advice as to the best site of astorage vessel they can at first add as muchguttering as the house they can afford and thenthey at least collect some of the water falling onthe roof. At later stages, if the owner isconvinced of the advantages of a rainwatercatchment system he/she can extend theguttering, and eventually add to the storage, tomake the maximum use of conveniently avail-able potable water. The disadvantage of thisapproach is that until enough guttering andstorage is available to meet all the potable waterneeds of the family they are still at risk ofdisease from consuming water from unprotectedtraditional sources.

The water from the gutter needs to bechannelled into the storage vessel. A verticaldownpipe of adequate size is ideal for thispurpose and it may be combined with one of thefirst flush diversion systems already discussed.Downpipes can be made from galvanised steelsheet by skilled tinsmiths, but the cost andavailability of suitable pipes may mean that asteeply angled gutter running down from theroof to a screened opening into the vessel has tobe used instead. Brick 'chimney' downpipes havebeen used on some projects in Kenya.

4.4.2 Cost Effective Storage Vessels

A wide variety of materials are available for theconstruction of tanks or vessels for the storage ofcaptured rainwater either above ground or belowground. These nearly all use cement and includeclay brickwork, concrete blockwork, stonework,concrete or ferrocement. Cylindrical corrugatedgalvanised steel tanks are one of the exceptionsto this general use of cement but even these arebest provided with raised plinths which willneed cement for their construction.

An extensive literature survey has recently beencarried out to study different constructiontechniques available for tanks and jars forrainwater storage (Skinner & Cotton, 1992).Two particular methods, cement mortar jars andferrocement tanks, appear to be very costeffective in their use of cement.

Cement mortar jarsMortar jars have been very widely used inThailand where more than ten million 2m^ jarsare in use (Gould, 1992) and factory productionis common with this size of jar currently beingsold for about USS40 (Brandford, 1992). Thesejars are typically made from 1:2 cement:sandmortar with a wall thickness which does notusually exceed 100mm and which is often muchthinner. The cement mortar does not usuallyneed reiniorcement although with the larger jarswidely spaced loops of wire are advisableespecially if they are to be transported. Themortar is usually applied to a removable innerformwork (e.g. Latham, 1984, Bradford, 1992),for example as shown in Figure 4.4.2.1 and isvery well cured to avoid cracking. The mortarsurface is smoothed and waterproofed with athin coat of cement slurry.

A UNICEF design for an unreinforced jar of2m3 capacity uses 4 bass of cement (UNICEFundated). A typical 2m^ jar design fromThailand uses 2.5 bags of cement, 0.25m3 ofsand and 1.5kg of plain 1mm diameter wire(Wirojanagud, 1991). The latter design isextremely efficient in its use of cement, needingonly 1.25 bags of cement per nA This is betterthan virtually all the above ground ferrocementdesigns which for capacities of up to 15m^ havecement usage rates usually not less than 1.6 bagsper m^ and to which the costs of wire mesh andsometimes bar reinforcement also need to beadded (Skinner & Cotton. 1992).

Usually for a fixed volume of storage, a singlelarger tank would be expected to be cheaperthan a number of smaller tanks with the sametotal capacity. This is normally the case ifferrocement or other materials are used for bothsizes of tank. However unreinforced, or lightlyreinforced mortar jars are not suitable forvolumes much above 2m^ so there are no furthereconomies of scale available with diisconstruction material. However it should benoted that the cost ot a number of cement mortar


jars compared to the cost of a large ferrocementtank of equal volume usually show the jars to bethe cheaper option. An important aspect of thiscost effectiveness is that a householder can overa number of years add to his collection ofstorage jars to increase his total volume ofstorage without the usual disadvantage ofmissing out on the cheaper unit cost of a largetank.

There are some problems which occur with theuse of a number of jars rather than just one largevessel which need to be recognised but oftenthey are not insurmountable.

One problem concerns the provision of asuitable hygienic vessel filling system to suit allthe jars from one downpipe. However, if two ormore downpipes are necessary for the roof it canbe an advantage to have a number of storagevessels compared to using a single large tank.This problem is avoided if separate sections ofgutter discharge into each jar but watercollection from a number of jars at differentlocations may then become difficult as notedbelow. Where there is only one gutter outlet theuser will having to move a flexible section of thedownpipe, or a gutter chute, to each vessel inturn. Making cheap, easy to use, hygienicconnections to tanks with such a movablesystem is not very easy. Also if a heavy stormcomes when the system is unattended somewater might overflow from the jar being served.An overflow system where the overflow fromone jar enters another is one way of overcomingthis problem but the necessary pipework mayprove expensive.

A second problem relates to the hygieniccollection of water from a number of jars since itwill probably not be cost effective to supply eachone with a tap. An alternative is a smalldiameter siphon tube with a tap or a bung(Figure 4.4.2.2). One end of the tube is pushedthrough a small hole in the jar cover until it isbelow the water level inside the jar. A tap orbung is fixed to the oilier end on the outside otthe jar. When the pipe is first installed it isnecessary to suck on the tap or tube to fill thepipe with water but thereafter closing the tap orreplacing the bung alter use should maintain thesiphon ready to deliver water when the tap isnext opened. When necessary the siphon can bemoved to another jar. Where it is affordable, or

there is only one jar a flexible pipe can bepermanently fixed near the base of each jar. Thispipe can be stored in the vertical position untilwater is required at which time the pipe islowered until it delivers water (Figure 4.4.2.3).Transparent flexible pipe, although useful forshowing the water level inside the jar mayencourage the growth of algae if it is usedoutside a house.

A third problem could be the greater spacewhich a number of jars take up near to the housecompared to a single large tank.

All jars, whether used inside or outside thehome should have close fitting covers. Largejars which cannot be easily tipped over whenthey are being cleaned out should be equippedwith a washout pipe which is opened to drainthe whole jar when necessary. A single outletcan serve both purposes if provided with a shortsection of removable pipe within the tank aslong as this can be reached through the openingin the vessel (Figure 4.4.2.4).

Cement mortar jars, particularly those of smallersizes than that mentioned above also have greatpotential for use in a home to hygienically storewater from other sources than rainwater and toallow hygienic collection of the water by meansof a lap and/or flexible pipe.

Ferrocement tanksFenocemeni consists of a carefully cured cementrich mortar (e.g. 1:3) reinforced with layers ofwelded and/or woven wire mesh, sometimeswith additional plain wire or steel rod rein-forcement for added strength.

There are a wide variety of ferrocement tankdesigns around the world and many thousandsare in use in Thailand and Kenya in particular.There are also some designs for plain wirereinforced mortar tanks using mainlycircumferential wires which although strictlyspeaking should probably not be termedferrocement will be considered here togetherwith them.

Most ferrocement designs are tor above groundtanks and capacities can be as great as 150m^(Watt, 1978). Above ground tanks have thedistinct advantage of allowing a tap for hygienic

i s ;

IIIIIIII

llIIIII1111

I_V • i-V . ^ . M-

water withdrawal. A Kenyan design for a 46nFtank cost about USS1500 in 1991 (Gould, 1991).

Ferrocement can also be used very effectively fortanks which are below or partly below ground.For example there is one very cost effectivelarge capacity design (over 75m^) consisting ofa ferrocement lining to an excavated hemispherewhich is widely used in Kenya (e.g. Nissen-Petersen & Lee, 1990a). This is sometimesprovided with an arched ferrocement roof. Thesetanks are discussed further in Section 5.7.2.Such below ground storage requires the use of ahandpump or the very careful use of a rope andbucket system for the withdrawal of water if itsgood quality is to be maintained.

It is suggested that the use of ferrocement tankson a wide scale for rainwater storage in mostGuinea worm endemic areas will probably notbe as cost effective and as easy to introduce asthe use of cement mortar jars. Amongst otheraspects, their feasibility will depend on theamount of water which needs to be stored andthe cost of the reinforcing materials needed forthe ferrocement.

Other tank materialsAlthough the mortar jars and ferrocement tanksappear to offer the best potential for costeffective water storage it is recommended thatproject designers check to see if this is the casefor the circumstances in a particular country.The previously quoted reference (Skinner &Cotton, 1992) gives comparative schedules ofmaterials for a number of designs and itcontains some useful advice about the variousother important factors which need to beconsidered when choosing a particular construc-tion method.

Where suitable sands or reinforcing meshes arenot readily available for mortar jars orferrocement tanks, or in other specialcircumstances, it may be cost effective to useother construction materials. For example,barbed wire reinforced concrete is reported to bemost suitable for Machakos District, Kenya,where stone aggregates are abundant (Gould,1991 & 1992 V

Instigating a rainwater catchment projectWhere flat galvanised sheet and galvanisedwires of the right gauge are already available in

a country it is likely that suspended V guttersare the must cost effective guttering system. Asmentioned above, cement mortar jars, or maybeferrocement tanks are likely to be the cheapestmethod of storage. The costs of a rainwatercatchment and storage system using suchcomponents needs to be calculated to see if it islikely to be affordable to users and how itcompares to the cost of improving traditionalsources or providing new sources of water.

Unfortunately in many situations the use ofloans or subsidies will probably be necessary tomake the system affordable (Gould, 1991).Gould also gives examples of the successful useof revolving funds in Kenya (quoting de Vrees,1987) and loan/income generation schemes inthe Philippines (quoting Appan et al. 1989) andin Indonesia. Before rejecting the idea of asubsidy it should be noted that many other ruralwater supply systems, such as boreholesequipped with handpumps, are also usuallysubsidised, so where a groundwater extractionsystem is not feasible it is not unreasonable togive a rainwater catchment system a similar percapita subsidy if without external assistance itcannot be afforded by potential users.

It should also be noted that maintenance costs ofa rainwater catchment scheme are very low andare borne by individual households who have avested interest in caring for their system.

If there are sufficient houses with sheeted ortiled roofs in a Guinea worm endemic areawhich has a suitable rainfall pattern and the ideaappears to be feasible, then a small pilot projectto demonstrate the system can be instigated withthe participation of some interested members ofthe community. If the pilot scheme showspromise then a wider programme ofinvestigating the community's views concerningrainwater catchment can be carried out, followedby health education and the promotion ofrainwater catchment schemes. Where some formof rainwater catchment is already practised theadoption of a new project is likely to be morerapid.

The Rainwater Harvesting Information Networkat the Water and Sanitation for Health Project(WASH) is a useful source of information on allaspects of rainwater harvesting. It can be usedby project designers tor finding information

IIIIIIIIIIIIIIII


about rainwater projects in any particular area otthe world. The recently formed InternationalRainwater Catchment Systems Association(ICRCS) may also be helpful. The addresses forthese organisations are:-

Rainwater Harvesting Information Centre,c/o Dan Campbell,Suite 1001,WASH Project,1611 North Kent Street,Arlington,VA22209,U.S.A.

International Rainwater Catchment SystemsAssociation,c/o John Gould,Department of Environmental Science,University of Botswana,Post bag 00222,Gaborone,Botswana

5 COST EFFECTIVE IMPROVEMENTSTO SURFACE WATER ANDUNPROTECTED GROUNDWATERSOURCES

SECTION SUMMARY

Surface water is usually of poor bacteriologicalquality and whenever possible an interventionwhich removes the risk of faecal pathogens aswell as infected copepods should be used.

Interventions At The Source

Of the interventions which mainly address onlythe problem of copepods the most effective relateto the separation of people from the source.

• There is a need for physical barriers and theacceptance by everyone of a communityrule that no one enters the water but that itis drawn only from a dedicated point.

• There are three preferred methods ofdrawing water at this 'dedicated' point,described in order of preference. The firstcomprises a sustainable handpump andfloating intake. The handpump can eitherbe directly connected to the floating intakeor it can draw water from a well-likereservoir in one of He banks which receiveswater from the intake.

• Secondly the use of a shaduf or other bucketand rope method from a vertical abutmentwith a surface which drains away from thepond is recommended.

• A final option is the use of a hand heldbucket dipped into the water from aplatform or ramp close to the surface of thesource.

• In order to give a good level of protectionagainst copepods the three recommendedoptions can be modified by the use ofstrainer material. Either the floating intakecan be protected by a box of strainingmesh, or, adjacent to the access platform, anopen lopped strainer box can be used. Thisbox floats in the water with straining meshon its sides and base so that water drawnfrom within the box when users dip their

buckets into it is automatically strained(Neither of these two suggestions are knownto have been tried so they both need to befield tested to discover how effective andsustainable they are.)

• Where necessary and feasible chemicalinsecticide (Abate) can be used if itsapplication is properly managed

Of the interventions which provide copepod freewater of reasonable or good bacteriologicalquality, depending on the field conditions, thefollowing methods are recommended.

• Convert stepwells to protected draw wells orto wells equipped with a handpump (if asustainable handpump is available).

• Collect water which is naturally infiltratingnear to water sources or collect water fromman made infiltration systems.

Interventions After Water Has BeenCollected

• The straining of water before consumptionis recommended if other water treatmentmethods which remove most bacteria aswell as copepods are not feasible. Thisstraining is particularly appropriate wherepeople need to consume water away fromtheir usual protected source.

• Settlement, long term storage and strainingusing vessels in the home can lead to agreatly improved bacteriological quality tothe water and this should be field tested.Cement mortar jars are suggested to beideal containers and they could be sold withpurpose made strainers, and taps or pipesfor hygienic withdrawal of water.

• The use of locally available coagulantscould be promoted to lead to an even betterquality of water from the system justmentioned but this needs further study inthe field.

• It is considered doubtful whether domesticor pondside slow sand filters can beproperly sustained to produce water of goodquality but their use would lead to someimprovement in the quality of the water.


IIIIIIIIIII1IIIIIIII

• Solar disinfection does work during periodsof bright sunshine and in the absence ofother methods of improving thebacteriological quality of the water it couldbe field tested to see how it works out inpractice

• Chemical disinfection at a users home is notconsidered to be sustainable.

Other Points Regarding Surface Water

• Improvements to the ground levelcatchments areas which contribute to rainfed ponds will

• lead to a better quality of water in the pond.These improvements include fencing,interception and diversion of surface waterfrom outside the catchment, settlement andthe use of gravel filters.

• Where sustainable suction pumps are notavailable it may be possible to convert directaction and deepwell pumps to allow them tobe used as offset pumps for drawing waterfrom floating intakes or bed infiltrationsystems.

• Where man made infiltration systems are tobe used in static water bodies, or where slowsand filtration systems are to be used, it isrecommended that the feasibility of usingone or more mats of open weave filteringfabric on the surface of the filter isinvestigated. This has the potential tosimplify the cleaning process and can leadto improved lengths of filter run and ashorter 'ripening1 period alter filtercleaning.

Introduction

It was noted in Section 2 that surface waterusually contains pathogenic micro-organismsoriginating from faeces and these need to beremoved to remove the risk of disease transmis-sion. Any method used to remove suchpathogens will also effectively remove die muchlarger copepods which carry the Guinea wormlarvae. The fact that the occurrence of copepodsin surface water is also often related to the speedof flow of the water was also mentioned.

Unfortunately it is not very easy to producepotable water from polluted surface water. Anumber of possible technical systems areavailable but it is usually hard to sustain these insmall communities in a low-income country.This problem increases as the size of thecommunity decreases or as it becomes moredispersed.

Figure 1.3 in Section 1 shows the various pointsof intervention possible in the life cycle ofdracunculus medinensis and each of the threepoints of intervention are considered in thefollowing sections.

This section investigates some simple methodsof preventing the copepods in surface water orunprotected groundwater sources from becominginfected with Guinea worm larvae. In additionto these interventions at the source, it should benoted that the medical interventions ofbandaging a lesion, or using surgery to remove aGuinea worm before the expulsion of any larvae,are also effective in preventing larvae reachingcopepods (Akhter, 1992).

Methods of killing or removing infectedcopepods from water used for drinking purposesare then considered. Unfortunately most of thesustainable methods of intervention in thesethree categories do not usually lead to much of areduction in the risks associated with faecalpathogens in the water. Interventions which inaddition to removing the risk of the transmissionof dracunculiasis also remove, or considerablyreduce, the risk of the transmission of diseasescarried by faecal pathogens are thereforeinvestigated. It is hoped that such methods ofproducing potable water, although often moreexpensive than those dealing only withcopepods, will be used wherever possible to leadto a reduction in diarrhoeal diseases as well asthe elimination of dracunculiasis.

Considering ways of improving the existingsources of water used by a community is a goodstarting point in any programme since thesesources are already acceptable to users and areconsidered by them to be the most convenientsource. If a project is promoting the use of abetter quality, but less convenient source thecommunity will have to be fully convinced of theadvantages of the new source before theyabandon their existinsz one. To achieve the

necessary changes of attitude and practiceusually takes a long time. With appropriatehealth education it may be possible to promotean improved source just for drinking water andto allow water for other uses to be drawn fromexisting sources. If possible, this has theadvantage of reducing the amount of waterneeding purification which can reduce costs andmaintenance problems.

With all the methods discussed in the followingsection it should be remembered that people whomove out of their villages for any reason, andthereby away from any improved water sources,are at risk of catching diseases from anyunprotected sources they then use away fromhome. The reader is also reminded about thediscussion in Section 2.2 which stresses that theadoption of hygienic handling and storagepractices are necessary to maintain the quality ofthe water from any protected source until it isconsumed.

5.1 Barriers To Prevent Copepods BeingInfected With Guinea Worm Larvae

5.1.1 Community rules on access

One of the simplest ways of preventing Guineaworm larvae entering the water is to make suremat no people infected with Guinea worm comeinto contact with the water. However this ideawill only work if the whole communityunderstand and accepts the importance ofeveryone following the rule and care needs to betaken that visitors to the community do notbreak the rules. In addition people need to bewilling to draw water for those who aresuffering from dracunculiasis. Some reportsfrom Ghana suggest that where such improvedwater drawing practices are usedthey make a major contribution to the reductionin transmission of Guinea worm disease (Bugn,1992).

The risks of contamination of the source byGuinea worm larvae arise not just from peoplewho come to draw water but for example alsofrom people who come to bathe in the water orchildren who come to play in the water. Faecalcontamination arises not just from peopleentering the water but from the washing ofclothes, the entry of animals into the water andfrom polluted surface water that flows into the

pond. This entry should be prevented as much asis possible by raising the edges of any pond ortank to divert surface flow away. If such flow isrequired to fill the pond then other precautionsare needed and these are discussed in Section5.6.

Tliere are two approaches to restricting access toopen water sources. The first is to only letuninlected people enter the water and the secondis to provide barriers to prevent everyoneentering.

5.1.2 Access only for the uninfected

The practice of people entering a body of waterto collect water is very detrimental to thegeneral quality of the water. Although peoplewithout dracunculiasis lesions can not addlarvae to the water anyone who enters the wateris likely to add some faecal contaminants to thewater every time they collect it. Suchcontamination contains numerous pathogenswhich can cause a number of illnesses in thosewho drink the water untreated. Despite thisdrawback restricting access to only thoseuninfected with Guinea worm is a possible startpoint in the fight against dracunculiasis where acommunity accepts the idea.

Where schistosomiasis (bilharzia) is endemic inan area there is a danger of the source havinginfected snails and anyone who enters the wateris at risk of being infected with cercaria whichmay burrow into their skin (see Section 2.2.2).

5.2 Barriers to prevent anyone entering the'•ater

Preventing everybody entering the water ismuch preferable to Die restricted access justdescribed. Such barriers will prevent waterirawers from continually introducingcontaminants, although during the rainy seasoncontaminants will often also be washed from thesurrounding ground into any unprotected opensource. Various methods can be used to avoid orprevent the need for water drawers to enter openwater sources and a number of these will now beconsidered.

FencesOf value for ponds and tanks, unless fences arevery dense or made from thorns people, espe-


IIIIIIIII1IIIIIIIIII

cially children, will often find a gap throughwhich they can pass. If people choose to theywill of course also make gaps unless thecommunity metes out some sort of punishmentto offenders. However, even very simple fencesare a reminder to the community, or visitors,that they should not enter the water and whensomeone is observed inside the fence it is clearthat they are breaking the community rule.

Because a fence restricts access to the water itneeds to be used with some appropriate methodof drawing water which can be operated fromoutside the fence and a number of water liftingdevices are considered later in this section.Alternatively there needs to be a gap in the fencethrough which people gain access to a platformnear to the surface of the water so that they canreach down and immerse their container to drawthe water. Whatever method of drawing water isused it will usually need to make allowance forchanges in the water level during the year.

Fences are also useful for preventing the accessof livestock. Where livestock are traditionallyallowed to enter the same water source as thatused for drinking water the practice should bediscouraged. Such animals not only bringcontaminants into the water but they often stirup a lot of mud which makes subsequenttreatment more difficult. Where there are anumber of open sources it may be possible forthe community to agree to dedicate some of dieponds just to livestock and to use fences toprevent them from using the ponds chosen forwater for domestic purposes. Alternatively waterlifting devices can be used to draw water for thelivestock and/or they can drink any water spiltby users if it is channelled away from the pondto a suitable drinking trough. A new pond couldalso be dug with its use dedicated to livestock. Ina similar way, a pond could be dedicated tobathing and/or washing clothes.

PlatformsPlatforms near to the surface of the waier fromwhich people can reach down to draw water in ahand held container can avoid the need forpeople to enter the water. However, certain typesof fixed platform may be unsuitable if during (lieyear the water level of the source varies by alarge amount.

For sources without steep sides it may bepossible to construct a ramp, raised at least750mm above the sloping base of the pond, togive access to the water. Whatever the pondlevel, it will then be possible for someone towalk down the ramp to near where it reaches thewater and then they can draw water over theside of the ramp. Wide steps can be used in asimilar manner.

Fur sources with steeper sides it may be possibleto construct an overhanging platform, orvertically walled abutment from which users candraw water using a bucket and rope. Someadditional excavation will usually be necessarynear to the platform to make sure that water isaccessible from such a structure at low waterlevels. The advantage of the use of an abutmentor solid platform is that it can be constructed toslope away from the water source so thatcontaminated spilt water can be drained awayfrom the source.

Another option is to use a ramp spanning fromthe bank to a floating platform from which thewater can be easily collected whatever the waterlevel. A variety of materials can providebuoyancy. Empty 200 litre oil drums could beused as floats, possibly coated with a protectivelayer of ferrocement to prevent them becomingholed due to corrosion. One disadvantage ofdie use of ramps, steps and walkways down tothe water is that users' feet can depositcontaminants such as faecal pathogens on theaccess and these can be easily washed into thewater by rain. Open slatted timber walkwayswill also allow din from feet to fall directly into(lie water. Also spilt water which comes intocontact with Guinea worm lesions can drainback into the water. In this respect the abutmenttype of collection point described above is betterthat the other types of access.

Upgrading of step wellsSome communities make excavations near torivers, or in dry river beds and they go downinto these excavation to draw water. Such'stepwells' are often points of transmission ofdracunculiasis and diseases carried by faecalorganisms but there is usually potential forupgrading the collection point to remove orminimise the nsk of disease transmission. Oftenthis can be accomplished by converting thestepwell to a draw well (Figure 3.3.1.2). Good

Crosslined

-section of completed concreted ring'Upgraded well'.

WindlassHandle

^U Wood Pole. Collar

SlabWell liner.

Concrete

Bucket

Earth backfill

Concrete

Water level

Sand pack

Wall of well

Well liner

Note: There is some variation in thebackfill. This can successfully be madewith the cuttings from the well itselfor with well-washed river sand. Soilfrom the surface should not be usedfor the lower backfill. Soil can be usednearer the surface.

a: Suction pump witn angled suction pips b: Rower pump with flexible suction pipe

c: Alternative of cylinder pump or bucket orbucket and windlass in well served byinfiltration gallery connected to reservoir

Perforated plastic pipe surrounded with grigravel ana sana

Filter box surrounded bygraoea gravel and sand

Impermeable soil around well and infiltration gallery

Some methods of filtering polluted water from an open reservoir


IIIIIIIIIIIIII•

design features of converted stepwells areidentical to those for draw wells (see Section3.3.1 and 3.3.3).

Riverside excavations above flood level canoften be converted to draw wells by providing alined shaft and backfilling the excavation takingcare to provide a sanitary seal around the shaftnear to the surface to prevent contaminated spiltwater or surface water finding an easy path tothe groundwater (Figure 5.2.1).

Seasonal excavations in river beds can also beconverted to draw wells but since the site isjnder water during the wet season the structuremay be subject to loads from fast flowing waterand flood debris. Also the bed of the river mayfluidise undermining and destabilising thestructure. These factors make the construction ofpermanent river bed wells which can be usedseasonally rather difficult and it is preferable tohave wells in the riverbank, with infiltrationgalleries if necessary passing into the river bed.Such river bank wells have the advantage ofmaking water available all the year round.

If the river bed well is likely to be undamaged byflood waters then one can be constructed. If it isleft open then the debris deposited in it can bedug out by the community at the end of the wetseason, just as traditionally they would have duga stepwell every year, but these wells are likelyto be polluted when first used alter re-excavation. Also it will be difficult for thecommunity to excavate very far below the initialhigh water table without the aid of a poweredpump and it may be necessary for the well to bedeepened several times as the water level belowthe river bed falls One further problem that canoccur is that the debris below the water in thewell may limit the entry of water through theperforated sections of lining.

It is preferable to build a lined well in a riverbed only it it is provided with a strong accesscover which can be closed during the floodseason in order to prevent the entry of most ofthe silt and other debris. Such a well can only beused as a draw well seasonally when the cover isremoved.

One further possibility is to have a totally buriedwell in the nver bed from which water is

withdrawn using a pump. This option isdiscussed later.

Water lifting devicesThe use of a water lifting device will make itunnecessary for the user to come very close tothe water and if the collection point is welldesigned there is no risk of a user contaminatingthe source with Guinea worm larvae.

Methods of collecting water from open bodies ofsurface water can be similar to the rope andbucket methods used for hand dug wells whichwere discussed earlier (Section 3.3.3). Usuallydie height of lift is quite small and a simple'shaduf will be appropriate.

Suction handpumps are usually very suitable fordrawing water from surface water sources. Sincethey can be set back a little distance away fromthe edge of the water source the risks of userscontaminating the source in any way is veryunlikely. The suction pumps rising maincontains no operating rods so it can be laidhorizontally from below the pump and thendown the bank and into the water body (Figure5.2.2). A continuous flexible pipe make idealsuction pipes since, unlike stiff pipe systems,there will only be need for one joint and thisreduces the potential for air to enter the pipe atjoints and spoil the pumps performance. Flexiblepipes also have the advantage that they can beused in conjunction with a tlouting intake asmentioned below.

There are some handpumps which areconsidered acceptable for irrigation but not forgroundwater. They have usually not beenconsidered suitable for drinking water becausethere is some nsk of water being polluted as it iscollected at the pump. Because the amount ofextra pollution is likely to be small compared todie likely existing pollution in a surface watersource it is suggested that the use of such pumpsshould be allowed if there are no feasiblealternatives. They are considered useful becausethey keep users away from the source and henceprotect it from contamination.

One simple design of suction handpump whichcan be used with surface water sources is the'Rower' pump (Figure 5.2.3). This easilymaintained pump was first manufactured inBangladesh where its mam use is as an

* Tht rope-washerpump mciuttd

5.1.5

IIIIIIIIIIIIIIIIII•

I1IIIIIIIIIIII1IIII1

irrigation pump. A derivative called the SWSRower Pump is now being manufactured in theUK where it is being sold for export at a cost ofabout £135. When using a Rower pump forwater supply a water container is placed belowthe open end of the sloping 2" or 2.5" diameterplastic pipe which acts as a cylinder. The pistonis operated using a rowing action with twohands gripping the T bar handle connected tothe piston rod and this lifts the water whichflows out of the open end. Wastage of water canbe high if users are not careful since some watercan overshoot the container, but the SWS pumphas some attachments which can be used tominimise this. These also solve another problemwhich is that users can sometimes being sprayedwith water at the start of the downstroke. Theangled pump body simplifies the routing of apipe to the water source (Figure 5.2.2). The longpump stoke possible when used by adults leadsto a very fast delivery rate which is appreciatedby users. The open ended cylinder has beencriticised as being prone to contamination and tothe theft of the pump handle, rod and pistonassembly but in many situations thesedisadvantages may be outweighed by theadvantages. Some communities overcome thetheft problem by removing the handle at nightand they also cover the open cylinder pipe witha tin to prevent the entry of contaminants. Aswith all suction pumps the loss of prime can be aproblem, but the maintenance or replacement olthe foot valve on this pump is very easy.

Another pump which can usually be locallyproduced and maintained using commonlyavailable materials is the rope and washer pump(Lambert, 1989). It can be used in an inclinedposition for open water bodies (Figure 5.2.4) orin a vertical position for hand dug wells but inthe latter case other more hygienic pumps arepreferable to maintain the good quality ofgroundwater. Like the Rower pump this pump ismore commonly used for irrigation . Theinclined version draws water from near thebottom of a pond so cannot benefit from theadvantages of a floating intake which arediscussed below. The vertical version could beused in a bankside 'well' fed by a floating intakein an adjacent water body (similar to what isshown in Figure 5.2.7a). Complete rope andwasher pumps are locally manutactured and soldin Zimbabwe for less than £60 (Lambert, 1992).

A third simple pump which could also be locallyproduced in some countries is the inclined coilpump (Figure 5.2.5). Its floating drum makes itsuitable for a varying water level. It is notreported that this has ever been used to drawwater for domestic purposes. Due to the depth of.the drum in the water use of this pump does notgain the benefit of shallow floating intakeswhich are discussed below.

Direct action handpumps are less suitable thansuction pumps for water collection because theformer usually need to be mounted directlyabove the source and they are therefore moresuitable for a bank well. However, where such awell is not thought to be appropriate, and if it isconsidered that a direct action pump is the bestpump then it can still be used as long as it is ofthe type which allows the footvalve to bewithdrawn through the rising main. Thearrangement would have to be similar to thatshown in Figure 5.2.6 where one end of asuction pipe or hose is connected to a bend atthe base of the rising main and the other endenters the pond.

Deepwell handpumps are overdesigned for theshallow lifts encountered with most surfacewater sources but open topped cylinder deepwellpumps could be used in a similar manner to thatjust described for the direct action pump if nocheaper pump were available.

Floating intakesAs mentioned in Section 1, one of thecharacteristics of copepods containing theinfective third stage of the Guinea worm larvaeis that they become sluggish and sink towardsthe bottom of a water source (Muller, 1985).This indicates that unless effective pond bedfilters are used to prevent copepods being suckedinto the pipe, the intake should be positionednear to the surface of a water body rather thannear to the base. Such a position is alsoadvantageous if turbid surface water can flowsinto the pond during heavy rain because waternear the surface will be the first to become clearas the suspended solids settle down to thebottom of the pond. It is also advantageousbecause if the water is stationary some bacteriain the clear water near to the surface of areservoir will be killed off by me ultra violetcomponent of strong sunlight (Section 5.9.3).

EZ:I-

X

i iX\ .̂VJ

Lvu U

IIIIIIIIIIIIIIIIIII•

1

S.I.7

However, the extent of this effect has not beenquantified in field studies.

Where the water level in the source fluctuatesduring a year a floating intake is ideal forwithdrawing untreated water from surface watersources. Such an intake can be surrounded witha strainer to ensure the exclusion of all copepodsas discussed in a Section 5.3.2.

Numerous designs of floating intake are possible(Figure 5.2.8). One interesting design describedby Morgan (1990) is to make a closed ring ofpolyethylene pipe to act as a float and then towire to this another ring fitted to a tee piececonnected to an outlet pipe. This second ringhas a series of 10mm diameter holes drilled inthe underside of the ring to admit the water anda few 5mm holes in its upper surface to allowany air to escape when the ring is filling withwater. The outlet pipe to any floating intake canbe the suction pipe to a pump (Figure 5.2.6a) orthe gravity feed to a bank-side 'well' reservoir(Figure 5.2.7).

5.? Killing Copepods and Straining OutCopepods from Unprotected Water

Where it is not possible to prevent all copepodsin a source from becoming infected with Guineaworm larvae then they need to be eliminatedfrom drinking water to make sure that thespread of dracunculiasis is halted. There are twocommon approaches to their elimination, one isto use chemicals to kill the copepods in thewater and the other approach is to strain the livecopepods out of the source water before it isconsumed.

53.1 Killing Copepods

There are two places where the copepods can bekilled either in the source water or in the wateralter it has been collected.

Killing copepods in the source water

Using chemicalsThe commonest method of killing copepods instatic open water sources is through the use ofchemicals applied in concentrations which killthe copepods but which are harmless to humanswho may consume the treated water. Temephos(Abate) is now the chemical most widely used

for control although chlorine and iodine havesometimes been used. 'Guidelines for ChemicalControl of Copepod Populations inDracunculiasis Eradication Programmes' (CDC,1989) is the standard reference on the subjectand it recommends that only Temephos is used.It suggests that chemical control of copepodsshould be considered when the followingconditions apply.

when there are few sources, each less than5OOm3 in capacity, which are being used mymany people;

where the provision of safe drinking water fromother protected sources is not feasible;

where health education compliance is poor;

during outbreaks of dracuncuiiasis whilevillagers wait for the provision of permanentprotected sources of drinking water;

when an additional security measure is neededto prevent transmission in areas whereelimination is imminent or recently achieved.

Detailed advice on deciding which sources totreat, and when and how to apply temephos isalso given in the quoted document.

Timing is particularly critical, the firstapplication needing to be made at least onemonth before the expected appearance of lesionson infected people and applications needing tobe repeated every 4 to 6 weeks throughout thetransmission season.

In order to apply enough temephos to achievethe recommended dose of 1 mg/1 it is necessaryfor the correct volume of the pond to beestablished for each of the varying water levelsthat will be encountered during the treatmentseason. Temephos kills both the copepods andthe Guinea worm larvae (UNICEF, 1992).

The American Cyanamid Company is willing todonate Abate to approved projects which follow(lie correct procedures for requesting it throughGlobal 2000 and UNICEF but projects must bearihe shipping costs. Details of this scheme arefound in 'Strategies and Resources' (UNICEF,1992).


TIIIIIIIIIIIIII1III1

WELL SUMP

HIGH j tW/L ',

W/L <{

HIGH W/L2 GUIOE POLES

BARREL

-STONE PITCHING

LOW W/L

,_ PERFORATED PIPEjj SECTIOM

40 MM PIPE j FLEXIBLE JOINTRUBBER OR PLASTIC HOSE

Although chemical control may appear simple, ahigh level of training is needed in to ensure thesafe and effective use of the product and in thebasic survey methods needed to accuratelycalculate the volume of any water bodies to betreated. Proper storage, planning, monitoringand logistics are also necessary for a successfulprogramme so that the space between applica-tions is not long enough to allow infectedcopepods to transmit dracunculiasis. Somecommunities may not readily accept the additionof chemicals to their drinking water, particularlyif they note changes in colour, odour or taste.

T Fsing fishIf a pond contains sufficient water all the yearround to allow copepod-eating fish to thrive inthe water then the fish will obviously help tokeep down the numbers of copepods. During thepreparation of this report no references werefound to indicate how effective the use of fishcan be in controlling the numbers of copepods ina pond. It is reported as having been promotedalongside the conversion of stepwells to drawwells in the SWATCH guinea worm eradicationprogramme in India (Akhter, 1990).

Killing copepods after the collection of water

Copepods and the larvae they contain can bekilled after a user has drawn water from a sourcebut before the water is drunk. There are twomain methods of accomplishing this, one is touse chemicals and the other is to boil the water.

chemicnlsTemephos is not promoted for domestictreatment of copepod infected water. Instead, ifanything, the use of chlorine compounds oriodine are usually suggested. Use of either ofthese chemicals has the advantage of also killingoff a number of other potential pathogens but forgood results the water needs to be free ofsuspended solids and organic material.

Chlorine compounds are usually more readilyavailable than iodine, particularly in the form otbleach such as Jik' or Javel'. A concentration of2me/litre of free chlorine residual is required tokill cyclops (Muller, 198?) and application of ahigh concentration would be necessary toachieve this level of residual chlorine in watercontaining organic substances. The free residualameentraiion required just to ensure the

elimination of pathogenic bacteria is about0.3mg/litre, much less than for killing copepods.but to achieve this even in clean water requiresan application of about 2mg/litre (Caimcross &Feachem, 1983). A free residual level of2mg/litre is required to kill off amoebic cysts(cit.ibid.) which are more resistant to chlorinethan are the bacteria.

Bleaching powders are available in somecountries but are liable to loose much of theirchlorine before use unless very carefully stored.In the preparation of this report no referenceswere found to the necessary application rate ofiodine to kill copepods. To achieve disinfectionfrom other pathogens application rates of 10-15mg/litre are required. One of the factors againstits use is the fact that it is highly volatile inaqueous solution (Hofkes et al. 1983) and thereare some fears about unpleasant side effectsfrom long term use (Caimcross & Feachem,1986).

In addition to the problems of rural householdsbeing able to purchase, properly store and applythem in the correct concentrations, it is oftenfound that the taste and smell of such chemicals,particularly free chlorine, is objectionable tomost users (Muller, 1985).

BoilingBoiling effectively kills copepods and the larvaethey contain and it is also effective in killingother pathogens including bacteria, viruses,cysts and ova. Unlike the use of chemicals it iseffective in water containing suspended solids.However it is expensive in firewood, requiringabout lkg of wood to boil each litre and with theincreasing deforestation in many developingcountnes the adopuon of this method of copepodcontrol has been rather limited. Some people donot like the taste of boiled water.

5.3.2 Straining Out The Copepods

The minimum size of an infected copepod isabout 0.5mm (Larsson, 1992) so passing waterthrough a strainer with an aperture size smallerthan this is an effective way ot removing themfrom drinking water. Such methods of control;ire often referred to in the literature as'filtration' or 'the use of filters' but since suchterms can be confused with slow sand filtration,winch is quite different in that it also removes

IIIIIIIIIIIIIIIII1


other pathogens such as bacteria, it is suggestedthat the terms 'straining' and 'strainers' are usedinstead. This terminology has been adoptedthroughout this report.

The straining material can consist of a finelywoven cloth but nylon or polyestermonofilament tnesb is much better since it canbe cleaned more easily after use. The E.I.DuPont de Nemours Company and PrecisionFabrics Group have pledged the free supply andshipping of a suitable 30 denier nylonmonofilament straining mesh with a pore size of0.1mm to approved programmes which followthe correct procedures for requesting it throughGlobal 2000 (UNICEF, 1992).

The use of a locally available tightly wovencloth, or the use of a double layer of moreloosely woven cloth, is being promoted by anumber of projects but people can growimpatient with the time it takes for water to passthrough such strainers cloths especially as theyoften soon clog up and are hard to clean.

It is important that water passes through astrainer in one direction unless it can bethoroughly cleaned. This is because after oneuse live copepods, or dead copepods containinglive larvae, may be trapped on what was theupper surface of the material and if it is theninverted they may be washed into the watercontainer collecting strained water. A lastingdistincuve mark on one side of the strainer helpsensure correct use.

It is worth emphasising here that strainingthrough such a mesh or cloth does very little ifanything to improve the bacteriological qualityof the water, and although it is copepod free thewater from a surface water source will stillharbour many other pathogens after filtration.

// is clear that for strainers to be effective in thefight against dracunculiasis extensiveappropriate health education is needed tochange peoples attitudes so that everyonerealises the importance of drinking onlycopepod free water. Such chan qes in attitudesare not easily achieved.

To date straining has usually been promoted a.san intervention to be used at home or in thefields atter water has been collected from a

source but it also possible in some instances forwater to be strained as it is collected from thesource.

Straining at the sourceNothing has been noted in the literature aboutthe use of strainers at the source of infectedwater and this is one area of intervention whichwarrants further consideration. It is not reallynecessary to strain all the water taken home fordomestic use but the users need to be fully awareof the risks of dracunculiasis before they arelikely to be able to separate effectively waterdrawn through a strainer from water drawnwithout the use of a strainer.

Where people draw water from a source byimmersing their containers in the water, if adedicated water drawing point is chosen it maybe possible to provide straining before the wateris collected. This could be accomplished if anopen box' of mesh, partly submerged in thewater, is provided. Water drawn by users fromthe inside the 'box' will have been strained as itpasses into the box through the sides and base.A floating box may be necessary where thewater levels vary and the mesh on the sides andbottom of the box would always need to beenclosed in a strong grille to protect it fromdamage from the containers used to collect thewater (Figure 5.3.2.1). The vertical mesh sidesand horizontal base to the box would probablybe self cleaning since any copepods or debrisdrawn onto the mesh when water enters the boxto replace what is drawn out are likely soon tofall back into the water. Field tests of such anidea are recommended to ensure that it works inpractice.

The provision of horizontal strainer systemsnear to the source, through which people pourtheir water after drawing it is not thought to befeasible. A major problem foreseen with suchstrainers is that they would need frequentcleaning to remove debns which would collecton top of the mesh. Such a system would alsorequire two containers so one could be under themesh while water was poured onto the meshfrom the other.

A shallow sand filter box system which actsmore like a strainer than a conventional slowsand filter has been proposed and tested bySndhar et al. (1985). Although copepods were

.VI. 1


removed by this system it rapidly silted up. Evenwhen it was clean it gave only a slow rate ofdelivery of water taking about 2 minutes for 25litres.

One situation where straining at the source ismost appropriate is where the user is away froma normal source of strained or protected watersuch as when working in a remote field. If sucha person can not be persuaded to carry potablewater to the field with them it may be possible topersuade them to carry at least a piece ofstraining mesh so that they can pre-straih thewater they draw from a traditional unprotectedsource. Such a mesh could be tied over the endof any vessel used for drinking when it isimmersed to draw water. Alternatively watercollected in another container can be pouredthrough the mesh as it is held over a secondcontainer used for drinking water. Both methodswould ensure that the water in the cup or bowl iscopepod free although of course it wouldprobably contain oilier pathogens.

An interesting idea of distributing to farmersthousands of baseball caps made from strainingmaterial has recently been tried in Ghanaaccording to de Rooy (1992) but theeffectiveness of this approach has not yet beenreported. The use of a stiff cap shaped strainerheld partly in the water by one hand wouldpresumably allow water to be scooped out of thecentre by the other hand if no drinking utensil isavailable. It would also ensure that the copepodsonly came in contact with one side of thestrainer (unless elsewhere water was pouredthrough the cap from the inside!).

The use of floating intakes enclosed by a box ofstraining mesh is considered to have greatpotential for supplying copepod-free waterwhere water is to be collected by any of thefloating intake methods discussed at the end ofSection 5.2.

Straining at homeThe use of a strainer at home to strain all watercollected from an unprotected source at firstappears to be simple solution to the preventionof the transmission of Guinea worm disease.However as pointed out in 'Guidelines for HealthEducation and Community Mobilization inDracunculiasis Eradication Programs' (CDC,1991) the sinsle behaviour of.straining water is

more complex than it appears. There are severalinterrelated actions required, including:

buying or obtaining the strainer;

using the strainer the right way round for alldrinking water;

removing the strainer carefully so that copepodswill not spill into the strained water;

cleaning the strainer by backwashing it withstrained water after use;

storing the strainer in a secure place where itwill not be damaged;

inspecting the strainer before each use to be sureit has no holes or tears;

discarding a damaged strainer and replacing it.

According to local water gathering practices andavailable materials different strainer designs canmake straining easier. Options cited bySilverfineetaU 199 Dare:

using elastic (eg. strips of inner tube) around theedge so the strainer can be easily fitted over themouth of a number of different sizes ofcontainer;

sewing strainer material to the centre of a pieceof cloth to save expensive strainer material;

making a cylindrical frames of pliable bark withoilier cylindrical pieces nailed on the outside tohold the straining material in a form convenientfor use (see also Duke 1984);

making small straining devices especially forfarmers to carry with them for use in the field.

The hygienic storage of water in a home hasalready been discussed in Section 2 and the costeffectiveness of cement mortar storage jars wasnoted in Section 4. It is suggested that Guineaworm eradication projects consider theproduction of purpose made strainers to beused in conjunction with suitably sizedcement mortar water storage jars fitted withtups/hoses and dedicated for use only forstrained drinking water. As mentioned later

(M)

improvements to bacteriological quality can alsobe promoted using settlement and short termstorage in such jars.

Unfortunately some individuals may usestrainers for sieving corn and cassava starch orother uses which may damage them or preventthem being used for the intended purpose.

As previously noted to be sure of avoiding beinginfected with Guinea worm disease a personwho strains drinking water when at home mustalso consume only strained drinking water whenaway from home or must only take drinkingwater from a protected source.

g,4 Infiltration Systems

One of the ways of considerably improving thequality of surface water is to collect it after it hasinfiltrated through granular material. As itpasses through granular material the level ofsuspended solids in the water considerablyreduces due to the removal of material bysettlement, staining and adsorption. Undersuitable conditions an active layer of micro-organisms which feed on bacteria can alsodevelop leading to a considerable improvementin the bacteriological quality of the water. In facta properly designed and maintained slow sandfilter should produce an effluent free of allpathogenic bacteria, viruses and cysts.

There are two ways of ensuring that filtration ofsurface water takes place before it is collected:

by collecting it alter it has passed throughexisting naturally deposited granular strata;

by collecting it alter it has passed throughgranular material which has been especiallycarried to the site to facilitate filtration.

There is little difference between methods ofwater collection via naturally deposited strataand those methods of exploiting groundwaterwhich have already been discussed in Section 3.The main difference is that the infiltrationsystems to be considered will be fairly close todie surface water source although the furtherthey are away from the surface water source,eenerally the better will be the quality of thewater (due to the longer period of flow from the

source to the extraction point during whichbacteria will die off).

Both methods of exploiting surface water shouldalways be explored in preference to directcollection of surface water because they providewater of a greatly improved bacteriologicalquality with few if any suspended solids. Withall except the coarsest strata the process offiltration will also effectively remove anycopepods present in the surface water. However,because infected copepods tend to becomesluggish and therefore sink, they will collect onthe surface of pond-bank or pond-bed filters sothe design, construction and maintenance ofsuch systems needs to be such that there is nodanger of the infected copepods passing throughthe system.

Where infiltration systems are used in the bedsof rivers or streams care needs to be taken toensure that the collection systems will not bedamaged by scour during flood flows.

The usual flow of water in perennial rivers isuseful in that it will carry away any solidsdeposited on the bed or banks as the waterinfiltrated. This effect does not occur in pondswhere over a period of time pond-bed or pond-bank infiltration systems are likely to becomeblocked by the solids which are filtered out ofthe water.

5.4.1 Filtration Through Existing Strata

Bank-side wellsIf the soil and sub-strata next to a body ofsurface water are permeable then thegroundwater level in them will be similar to thelevel of the surface water. This means that awell or borehole constructed adjacent to such abody of water will yield either groundwaterwhich is approaching the surface water, orinfiltrated water which has passed through thestrata between the well and the surface watersource. The further away the extraction point isfrom the source the better will be the quality ofthe water.

Infiltration wells and galleriesMany surface water infiltration systems cancontinue to yield water even after none is visiblein the source because infiltrated surface water isoften still present in the .strata below the bed of

ft I



the pond or river and below the banks if they arepermeable.

Where there is a long season when there is nowater flowing in a river, if it has a deep bed ofgranular material it may be possible to drawwater throughout the year from an infiltrationgallery' connected to the bed deposits. These'infiltration galleries' in their simplest formconsist of trenches excavated into the waterbearing strata which are backfilled with suitablygraded fine aggregate and coarse sand to act as aconduit for the collected water to be transmittedto a collection point such as a bank-side 'well'from where the water can be extracted using alifting device (Figure 5.4,1.1). A perforated pipecan be laid in the material filling the trench toconsiderably improve the rate at which the watercan be collected from the surrounding strata(Figure 5.4.1.2).

The construction of such infiltration galleries inthe beds of rivers can be quite difficult, and itusually has to take place in the dry season withthe assistance of powered pumps and some formof temporary trench wall support. Where thebanks are of a material which yields waterinfiltration galleries can be constructed alonethe bank (Figure 5.4.1.3 ) which can be easierthan constructing one in the river.

With the right equipment it is possible to workfrom within a large diameter bank-side well toinstall horizontal perforated pipes into granularmaterial in river banks or under river beds tocollect the infiltrated water (Figure 5.4.1.4).Horizontally boring equipment developed by theBritish Geological Survey (BGS) for use inwells was mentioned in Section 3. Unfortunatelyit is expensive and it requires a well with aminimum internal diameter of 2m but theradials can often be installed relatively easily upto about 30 m in any direction to collectgroundwater or infiltrated water over a largearea. [Wright & Herbert (1985) and Herbert &Rastall(1991)].

The use of offset hnndpnmpsIt may also be possible to construct buried wellsin the river bed from which water can beextracted using a handpump sited on the bank.Alternatively a suction pump can be directlyconnected to a suitable screenins device buned

in the river bed and this is discussed in Section5.5.

A recent use of a buried well system in seasonalstreams in southern Ethiopia is described byBradfield (1992). In this project 400mmdiameter concrete rings, some of which hadporous walls, were stacked vertically in adewatered excavation in a sandy river bed. Apipe from the pump on the bank was routed intothe space in the centre of the rings and the top ofthe stack was sealed some distance below bedlevel with a permanent concrete lid. Theexcavation was then backfilled with the riversand which also covered the lid.

These wells yielded 3000 litres per day and thematerial and labour costs came to £285including £90 for the handpump. The pumpused was a small rotary progressing cavityMono M' pump from the UK, which Bradfieldadmits could be open to criticism since it is notstrictly speaking a VLOM (village level opera-tion and maintenance) pump. However becausethe handpump site was offset from the well itwas not thought to have been possible to useeither of the two VLOM handpumps available inthe country, that is the direct action Nira AF85and the deepwell 'Ibex' (a locally producedversion of the Afridev). These pumps were alsomore expensive than the Mono M but the longterm performance of the latter in the field is notyet clear although after a year there werereportedly no signs of wear.

Where a suction pump is not readily available,or where it is desirable to restrict the number ofdifferent types of handpumps in use in acountry, there is a way of using direct action ordeepwell handpumps as offset pumps toinfiltration systems. This is in a similar mannerto that described earlier for collecting unfilteredwater from a water body. That is by fixing abend and horizontal suction pipe below thefootvalve of such pumps as illustrated in Figure5.2.6b.

5.42 Groundwater dams

Groundwater dams are usually constructedacross the valley of a seasonal river or streamwinch passes through an area of impermeablestrata such as rock or clay. These dams obstructthe tlow of groundwater in the bed of the

cross-orofile over river

pvc pipe

River well with sand trench to the bank

RIVER

— COLLECTOR DRAIN

WELL

COVER

r- 5 . U . V . IHorizontal collector arain under river bed

IIIIIIIIIIIIIIIIIII

IIIIIIIIIIIIIIIIII

IIII11

I1

IIII

I'/II

< i

l l

X X

2 TO 3 M.

_GR0UN0WATER FORMATIONSOF LOW PERMEABIUTY

PERFORATEDLATERALS

Radial collector well

watercourse in order to provide water storagebelow the surface.

There are two main types of groundwater dam-The sub-surface dam is constructed below theexisting river bed to prevent the water containedin the river bed deposits from flowing down-stream (Figure 5.4.2.1, 5.4.2.2 and 5.4.2.3). Thesand-storage dam is extended above the originalbed level to also impound water in sedimentswhich accumulate behind the dam whenfloodwaters flow in the valley (Figure 5.4.2.4).More details of both types of dam are givenbelow.

To date groundwater dams have not been verywidely used but they have great potential forproviding good quality water throughout theyear from the nver bed deposits found in someseasonal rivers. They can provide water in somesituations where the use of just an infiltrationgallery would not be successful.

Nilsson (1988) is recommended to readers whowant to explore this technology further. Abriefer introduction to the subject is given bySmith (1990) and Skinner and Cotton (1992).Detailed guidance on the construcuon of somedesigns of clay and stone masonry sub-surfacedams is given by Nissen-Petersen & Lee(1990c).

If water is not present in the bed material of ariver throughout the dry season, in may bepossible to construct a sub-surface dam to retainwater which would otherwise flow downstreamthrough the bed deposits. The dam which isconstructed below the bed of the river can bebuilt from a variety of materials including justclay (5.4.2.2a). Water in the bed materialupstream of the dam is extracted by wells and/orinfiltration galleries etc. In a few situationswhere the ground slopes sufficiently it may alsobe possible to use a pipe to deliver water bygravity to a downstream water collection point(Figure 5.4.2.5b).

ri.imsSand-storage dams can only be sited where tliereis a seasonal flow of surface water which carriessuitable granular material which can bedeposited behind the dam to till the plannedvolume of the reservmr. 'Hie choice of site and

the design of the dam is therefore rather morecomplicated than for the sub-surface damsmentioned above which rely only on existinggranular deposits.

The dam is usually built in stages which are of aheight suitable to allow the deposition of onlythe coarser materials carried in the seasonalsurface water which flows over the dam.Because the dam is above the bed of the river itcan usually be used in conjunction with a pipedsystem which delivers water by gravity to adownstream water collection point (Figure5.4.2.5).

The use of a sand-storage dam has at least twoadvantages when compared to a conventionalsurface water dam. The first is that the naturalfiltration of any polluted surface water throughthe sand will lead to a relatively good quality ofwater being collected from the lowest layers ofsand. In the dry season this protective layer ofsand also means that there is a greatly reducedrisk of contamination of the stored water byusers because it is not exposed. The second isthat there are very low evaporation losses fromthe sand filled reservoir compared to an openreservoir, but of course a larger volume ofreservoir' is needed to compensate for thevolume of the sand.

Infiltration systems directly connected tosuction pumpsInstead of collecting water from a river-bedinfiltration system by using gravity flow to abelow-ground reservoir such as a riverbank'well' it is possible to couple a suction pumpdirectly to a pipe connected to a suitablescreening device buried in the river bed.L.ultrated water can then be pumped out of thedevice using a remote pump. Since most of thesesystems involve the construcuon of a filter ofspecially selected materials rather than the useof existing strata they are considered more fullybelow.

5.5 Infiltration Through Man-Made Filter!;

Two types of man-made filters are considered inmis section. One type is constructed belowground and is useful where the bed or banks of ariver or pond are impermeable or can onlytransmit water slowly. In such situations it ispossible to construct an infiltration system by



' ' - t ' " • " ' H " . * Cm* .'

Figure A sub-surface dam constructed from brickwork.L, 1 , \ (Source: Lee &. Visscher 1990)

Sronw Rivtr bed

Riv»r btd

230 *,';..;" Sand"; ' '

a) Clay b) Stona majonry

Figure Cross sections of clay and stone masonry subsurface dams"2_ (Source: NISSEN-PETERSEN & LEE. 1990c)

Note: In otrw designs the top ot the dam is often below bed level • So* Figure 12.4,1.

Figure Arrangement of wlngwails, soilembankments, spillway and boulders needed to

control bank erosion, and bed erosiondownstream of a groundwater dam.

(5ourc«: NISSEN-PETERSEN & LEE. 1990c)

' • • • • •• : - . .• ;• • ' • • ' ' , ' • . • ' . • ; '.':.'• • . ' • , ' • . V : • ' ' • ' • ' • ' . • * . ' 7 S " « * * \

1 ^ : . • • : • • : • • • ; :

^ / ' l . V / , - * ' .'oiiom*'"W 0.0 , 1'- I / /^/.V, .oiiom*W 0.0 , 1 - I / i l.-'

Flgure A typical sand storage dam.U . 1 . U (Source: NILSSON 198B)

„ Grama

a) Well and handpump

New bod level

Ground watar laval

Grove) drains if appropnata

Pipe to discharge

b) Gravel Alter and pipe pomt

Figure Water extraction options tor groundwater dams (Source: SMITH 1990)Option a \s suitabia tor all types ot gmundwaW dams, oooon b)is often not practical wim sub-surtaca aama bacausa a long pipe.

laid in a daao ranch, is usually needed downstream ot the aam to bring mm ascnarge pomt abovm gmuna laval.

1

excavating trenches or holes which can be filledwith selected granular material to form a filterthrough which water can gravitate or fromwhich it can be removed by a suction pump.

The other type is a slow sand filter. This isusually constructed separate from the watersource and is often partly or wholly aboveground in a tank. Unless the water source is at ahigher elevation than the filter, so that water cangravitate through the system a water liftingdevice is needed to raise the raw water to a levelfrom where it can gravitate through such filters.

5.5.1 Man-made Infiltration Systems

In many respects these are similar to theinfiltration systems discussed earlier except thatthey usually have a smaller surface area exposedto the water body because the granular materialhas to be transported to the site, or it has to besieved out of excavated material, and this makesit expensive and time consuming to constructvery large filters. This size restraint has thedrawback that such filters are more prone toblockage than when filtering can take placethrough existing strata which usually has a largearea in contact with the water. Smet andWheeler (1990) give a useful review ofinfiltration systems.

If a pond has impermeable banks a man-madeinfiltration system can be created byconstructing a well in the bank and connectingit to the pond by a wide trench filled withsuitable sand and gravel to filter the water as itapproaches the well' which acts as a reservoir(similar to what is shown in Figure 5.4.1.1).During the preparation of this report few reportsof field experiences of such a system have beenfound h'ii is believed that UNICEF Nigeria arecurrently experimenting with this type ofinfiltration system (Larsson, 1992).

An alternative method is to excavate one ormore trenches or a large hole in the bcu of ihepond and to fill it/them with sand and aggregatethrough which the water can filter. Theseexcavations are connected to the bank-side well,preferably some pipes near to the base of thegranular filling so that the filtered water has aneasy How path to the 'well' (similar to what isshown in Figures 5.4.1.2 and 5.2.7b ). Yet

.mother method is to construct the filter on thebase of the pond (Figure 5.5.1.1).

It is also possible to construct open reservoirexcavations and associated infiltration systemsbeside a river. The excavations can periodicallybe flooded with water, preferably when the riveris not carrying a high load of suspended solidsto block the infiltration system. If two suchexcavations are used it will be possible to cleandeposits out from one while the other remains inuse. Alternatively one can act as a sedimentationreservoir to reduce the suspended solids carriedin the water allowed to pass into the other whichis equipped with an infiltration system (Figures5.5.1.2 and 5.5.1.3).

Some man-made infiltration systems can bedirectly connected to a pump (Figures 5.2.2 and5.2.6c)and this type of system has been usedwith a pond bed filter in some UNICEF projectsin Nigeria and is being further investigated there(Larsson , 1992). The existing installationsconsist of a deepwell pump located on the bankof a pond, with a horizontal pipe connectedbelow the cylinder to draw filtered water from apond bed filter. The filter was made byexcavating a hole 1.8m by 1.8m in plan andthen filling it with layers of aggregate followedby coarse sand (Ismail, 1992). It was used with afilter fabric laid on the surface of the filter andthis idea is discussed further below

Another example of a man-made infiltrationsystem, sometimes called the 'Cansdale filter', ispromoted by SWS Filtration Ltd., UK. Thisconsists of an inverted glass fibre plastic boxwith a screen part way up die box to separate aclear space at the top of the box from a filling offine aggregate. The box is buried in the granulardeposits in a river-bed or lake-bed (Figure5.5.1.4). Where the bed material is not suitableto act as a natural filter, such as would be thecase with a clay bed, a hole is dug and this ismen backfilled with suitable granular materials(Figure 5.5.1.5). In both cases the filter box isbest surrounded with some fine aggregate andcoarse sand. A flexible hose from the top of theinverted box is connected to a suction pump(such as the SWS Rower pump) which is used todraw surface water which is filtered as it passesthrough the sand into the box. The filter boxunit alone is currently sold in the U.K. for £145(300mm x 300mm x 600mm for up to 6

64

,, fencing or h*dg*

collecting w*ldry iton* pitchingbctnan nun «ndlow w«t»r

•"' '^^^^^W-i^-0^^0^^^,-.

tils ar«ln« orperforated A.Cpipe (gr»v«lsurroundnl)

5.5-1,

I, ^v

har if*mt»ln l«v*l at "A"

iftpi* t loat vali-* ' suyt^^'iti

uirvd) or allow to fill to

ing w«t stason.

-zletr «ll vundardraln*

5.5.1.1-



Section B • B

Raw water line

Gravel layers '

PUn

Dank

Apron

Well

Inlet

An infiltration pallcry under a separate channel; the gallerv is builtbefore diverting me sunacc water over ti. i From Raiatjopaian and Shiftman )•

The S.W.S. filler. (From Canidalei.

Diameter of hole approx. 2*0 m

10 mm gravel

Stones or broken roCK

11111

litres/sec) or £95 (300mm x 300mm x 300mmfor up to 3 litres/secWSWS, 1992)

The whole SWS filtration system, whether itconsists of imported material or existinggranular bed material, is best 'developed' byfrequently stopping and starting suctionpumping through the system for a few hoursusing a powered pump. Pumping water in thereverse direction through the granular materialcan also help this development process whichremoves very fine material from the granularmaterials and which creates an improvedgrading of the filter (SWS, 1992). Use of theSWS filter box system in river beds in Nigeria isreported by Joy (1983).

SWS have also promoted a mini suction filterunit used inside a small upward facing box filledwith granular material covered with a syntheticfilter mat. This has been used by placing the boxon the base of irrigation canals which were asource of schistosomiasis. Water abstractedthrough the box was free from cercaria but itwas only suitable for use by up to about 10people. The filter mat was periodically replacedwhen it became blocked, and after long periodsof use the granular material in the box alsoneeded washing or replacing. (SWS, 1992)

A summary of various field tests on SWSinstallations is given in Smet and Visscher(1990).

One of the disadvantages of all suction systemsis that if a reciprocating handpump of any typeis used mere are numerous sudden changes invelocity of the water being filtered as the pistonmoves up and down the cylinder and tins candraw solids and pollutants much deeper into thebed of filtering material than would be the casewith gravity flow. This can lead to a poorerquality of filtered water and the deeperpenetration of solids makes it much moredifficult to clean die sand if it becomes clogged.However it is possible periodically to couple thesuction hose to the outlet from a powered pumpor a gravity supply from an elevated tank to gainsome advantage from 'backwashing' while diesand is being raked to resuspend depositedsolids. In a river, these solids can be earnedaway by flowing water, but in the case of a pondthe solids mav soon settle back onto the filter.

The biological treatment processes which takeplace on slow sand filters are not likely tobecom fully developed with suction infiltrationsystem^ especially when use is only periodic. Aevaluation of field installations of a number ofthe SWS filtration system (LSHTM & Gifford,1986) indicated that the bacteriological qualityof the water was sometimes not very good andsuch a system could therefore not be relied uponto produce high quality water. However it is tobe expected that such systems will be effective inremoving copepods which are much larger thanbacteria. The removal of much of the suspendedsolids and attached bacteria is feasible with suchinfiltration systems so that they can be expectedto lead to at least some improvement in thebacteriological quality. The system can of coursebe the first stage before another treatmentprocess.

5.5.2 Slow sand filters

For a long time it has been recognised that slowsand filters are especially appropriate forproducing potable water for communities indeveloping countries. They can produce water ofhigh quality from bacterially contaminatedsurface water without the need for complicatedplant and operational procedures usually foundin other types of conventional water treatmentsystems. Organisations such as the InternationalReference Centre for Community Water Supplyand Sanitation (IRC) has been active inpromoting research and field tests to findappropriate designs of filter (Visscher et al.1987) and appropriate methods of reducing theturbidity of surface waters to a level where slowsand filters can operate (Smet & Visscher,1990).

It appears that most of this research has beendirected towards serving communities generallylarger than the 250 - 500 people beingconsidered in this report. Such smallcommunities can be served by small slow sandfilters but during the preparation of this reportthere were no field reports were found whichshow that in practice such a community are ableto run and maintain them properly.

For good performance slow sand filters shouldoperate continuously and should receive a steadyHow of water and they are therefore best servedby a regulated gravity supply direct from the

source or from a elevated tank filled by a pump.This requirement often cannot be met or cannotbe afforded in small communities. If thecommunity is scattered a relatively expensivecentralised slow sand filtration system itsposition may not be sufficiently convenient toencourage universal use.

A slow sand filter needs periodic cleaning andinitially this is achieved by the simple operationof skimming off a thin layer of sand.Unfortunately for a few days alter this operationthe bacteriological quality of water from thefilter drops until the biological purificationprocess is re-established in the new top layer,although the use of a geotextile filter mat speedsup this process. During this 'ripening' processthe water passing through the filter has to be runto waste or used for non-potable purposes whiledrinking water is supplied by another filter orfrom a storage reservoir filled by the filteredwater before maintenance started.

Eventually, because the periodic removal of sandleads to an excessive reduction in the depth ofthe sand bed in the filter new sand has to beadded. During this procedure and for 3 - 7 daysafterwards while the filter 'ripens' no goodquality water can be obtained from the filter andagain the community has to rely on stored wateror another filter.

Where the source water is turbid additionalcomplications can arise from the need to provideand maintain a 'roughening' filter to reduce thelevel of suspended solids to a level which willnot quickly block die slow sand filter. "Hieturbidity of water suitable for slow sandfiltration should usually be less than 20 NTU togive a reasonable length of filter run. So formany surface water sources some form of pre-treatment (eg. sedimentation, infiltration orroughening filtration) is often needed,particularly during the rainy season.

From the above discussion, and from anunderstanding of other control proceduresnecessary with slow sand filters which have notbeen mentioned, it may be concluded thatalthough the running and maintenance of slowsand filters may not be technically verycomplicated it does demand a trained person.This person needs to be able to understand thetreatment process and to always carefully follow

die important maintenance and operationprocedures. He or she needs to be willing andable to regularly commit time and effort to thiswork and will probably need to receive somepayment from the community as an incentive.

The evidence suggest that the number ofcommunities of 500 people or less in which slowsand filtration is sustainable is very small.However, where the conditions for successappear to be present it should be consideredsince such a treatment process produces a highquality potable water.

Packaged treatment plantSmall packaged slow sand treatment plantwhich can be quickly installed in a communityhave been successfully field tested are availablefrom some manufacturers. One such unit is thePotapak which uses polyethylene tanks 1.25mdiameter and 1.2m high and simple flow controldevices connected by plastic pipework (Figure5.5.2.1). The same size of tank used to hold thefilter sand can also be used to hold gravel toform a prefilter to pretreat high turbidity waterand the tanks are designed for stacking togetherto save space during shipping. The sand andgravel for the filters is not supplied with theunits so it needs to be locally available. Twocomposite filter mats, each held in a separateplastic frame, are used on the surface of the sandfilter to trap most of die suspended solids, auseful function which is discussed further at theend of this Section. The top mat is removed andwashed clean weekly and then replaced belowthe other mat. This interesting procedure causeslittle disturbance to the useful biologicalprocesses which take place in the mats andupper layer of sand. The pipework also allows agentle backwashing procedure to be periodicallycarried out to remove accumulated silt from theupper layers of sand.(POTAPAK 1992)

Each unit can give an output of about 12m3/day.The unit cost of the Potapak for a purchase of atleast 10 slow sand filter units is about £2070FOB UK port each. The gravel filter units areabout £900 FOB UK port. The cost of a storagetank for the treated water is excluded from thesefigures. As mentioned above it is also necessaryto proved a continuous flow of water to thefilters.


RAW WATERINLET

Dimension* and contraction of package plant and flowcontroller |X

nA

fe>

J

J

Plan view of Inrge PSF wiihnin lid

Filter Chaabei-

Stor*ge|/Ctaafflberf

Pump

f -i/R«4aedPlatf ora

Storag* Ctaaaber Corer

1_Tap

Section 73-8 Section C-C ,!

Potapak are willing to consider providingvariations to their standard system to meet Lheparticular demands of purchasers (POTAPAK1992). In many situations it would be possible toconsiderably reduce these costs by constructingferrocement tanks.

It is suggested that further investigation intoreducing the cost of packaged treatment plantsis carried out. In the meantime a few of theexisting units could be installed in communitieswhich showed the most promise of operatingand maintaining the units and if such a field testwas successful it could be repeated elsewhereusing die reduced cost plant. Alternatively thepackaged plant could be replaced by a morepermanent slow sand filter unit and the originalunit could be tried out in another community.

Pond sand filterSince 1984 field testing of a small slow sandfilter for improving water from ponds has beencarried out in Bangladesh (DPHE & UNICEF1989). A report on their performance wasrecently produced (WHO, DPHE & UNICEF1991).

The pond sand filter (PSF) has been promoted intwo sizes one for up to 300 users (PSF300) andone for up to 500 users (PSF500). Each consistsof a small two chamber tank constructed aboveground (Figure 5.5.2.2). Water is lifted from ascreened floating intake in the pond byhandpump from which it is discharged througha prefilter into the water above the sand in thefilter chamber. Originally coconut fibre andbrick chips were used in this prefilter but this isnow discouraged and only brick chips arerecommended^(WHO, DPHE & UNICEF 1991).Water which passes through the sand filter iscollected at the base of the filter chamber fromwhere it flows into the adjacent storagechamber. It is collected from the chamber usinga pipe connection to a tap.

The report on the performance of 153 PSFsystems found that 18% had been abandonedand a number of the others were not being wellcared for (cit.ibid). Unfortunately only 22installations were tested for bacteria, and itseems that these tests were for numbers of 'totalcoliform' rather than the more specific faecalconform. Somewhat surprisingly in the light ii!

the aeneral evidence of the indifference of users

to the proper care of the system, in 18 of theseinstallations no coliform bacteria were identifiedper 100ml tested but the other 4 PSF units hadunspecified coliform counts greater than zero.The raw water in 55% of the ponds servingthese PSF had total coliform counts of less than10 per 100ml and in the others the counts wereall less than 25 per 100ml which again seemremarkably low.

The report also investigated the users attitudesto water quality and the use of the system. It israther surprising that there are no reports of lazypeople drawing water from the tap withoutpumping an equivalent amount of water into thefilter chamber. The rate of filtration is about 4-9litres/minute so unless users are patient thesystem requires that people draw filtered waterwhich was pumped into the filter chamber byanother user.

Amongst other things, the report came to theconclusion that the use of PSF units shouldcontinue where there is no alternative source ofpotable water but that greater beneficiary par-ticipation is necessary in construction andoperation of the units and that health educationin the community and training of the caretakersis very important.

It would seem therefore that the PSF does havepotential for treating pond water to potablestandard and since it removes bacteria it can beexpected to effectively also strain out anyinfected copepods that reach the floating intake.However the present authors have reservationsabout its widespread adoption for reliableremoval of bacteria in situations which maydiffer from those in Bangladesh.

These reservations include firstly the fact thatthe flow rate through the filter is likely to besubject to a number of changes throughout theday as it varies from zero to the maximum ratecontrolled by the difference in level between thetop water level in the filter chamber and level ofthe inlet pipe in the storage chamber. The flowrate during the night may cease all togethersince the amount of water held above the bottomwater level in the filter chamber (the same levelas the top of the inlet pipe in the storagechamber) is relatively small and this may soonpass through the filter. As previously mentioned

fi7

IIIIIIIIIIIIIIIIIII

such changes in flow are detrimental to goodslow sand filter performance.

Secondly if there is only one PSF unit it will notalways be able to supply bacteria free waterbecause immediately after the top layer of sandis scraped off the quality of the water will reduceand it will not return to a good standard until aday or two later when water passing through thefilter has caused the useful bacteriological layersto re-establish. The water passing through thefilter during this period should not be used fordrinking water. Someone therefore needs to bewilling to pump the water without gaining thebenefit of potable water.

Thirdly die design seems to omit instructions forfilling the filter chamber with water from thebottom when the sand is first installed, or whensubsequently it is replaced after washing. This isusually recommended for sand filters becauseinitial filling from the top often traps air in thepores of the sand which adversely affects theperformance of the filter. It appears that thisinitial filling could take place via the storagechamber if the vertical pipe in the storagechamber ( see Figure 5.5.2.2) was temporarilyremoved. Ideally filtered water should be usedfor this process. If raw water is used it must bestrained to remove copepods which mightotherwise survive in the storage chamber. Theuse of raw water will also introduce somebacteriological contamination into the chamberwhich may persist for some time.

Finally, as mentioned earlier, the running andmaintenance of such a sand filter needs thededication and commitment of a well trainedand highly motivated caretaker. Such a personmay be hard to find and train in many of thesmaller communities which are beingconsidered in this report.

Despite these reservations the use of PSF unitsshould be considered where no other appropriatemethods of producing potable water areavailable. Even if a PSF does not producebacterially free water the quality can be expectedto be much better than that of the surface waterfrom the source, and the filtered water shouldalways be copepod free. Furtherexperimentation, design changes and fieldtesting in Guinea worm endemic areas may

overcome some of the problems mentionedabove.

Domestic sand filtersThe use of very small slow sand filters fortreatment of water for an individual family hasbeen suggested by a number of authors. Thedesigns are often based on the use of old 210litre oil drums to contain the sand (e.g. Yaziz &Din, 1988, Cairncross & Feachem, 1986,Morgan, 1990) but in the preparation of thisreport no literature concerning their widespreaduse has been discovered.

UNICEF Kenya has in the past promoted the useof an upward flow water filler based on the useof two cement mortar jars one acting as anuntreated water reservoir and the other as a filter(Childers & Claasen, 1987). The filter jarcontained a thick layer of charcoal sandwichedbetween two layers of fine sand (Figure 5.5.2.3).An evaluation of the performance of this filterwas carried out by Singh and Chaudhuri (1992)and they came to the conclusion that thereduction in faecal coliform from its use was notsufficient for it to be used alone to treat moder-ately contaminated water, but that it should beused with another jar with a 600mm layer ofsand acting as a conventional (downflow) sandfilter.

UNICEF Nigeria is currently evaluating anumber of designs of domestic sand filter(Larsson, 1992)

Domestic sand filters will usually lead toreductions in the numbers of faecal pathogensand should always exclude copepods. However,there are reservations about the effectivenessand sustainability of household sand filters toproduce bacterially pure water. The maincriticisms relate to the rate of flow and thesustainability of operation. With most designs(such as that in Figure 5.5.2.4) the rate offiltration is likely to regularly vary from zero toa value near to the maximum rate and asexplained for the pond sand filters this is likelyto adversely affect their performance. A fewfilters designs (such as Yaziz and Din (1988))h;ive another design fault which is that the outletpipe which leaves the bottom of the filter is notraised above the level of the top of the sand.Tliis omission is a bad feature since without

Concnn cover

4S q«llon Qrum

I

IIIIIIIIIIIIIIIIII1

IIIIIIIIIIIII

i:I1IIi

such a feature the whole filter bed can beinadvertently be drained from the low level tap.

Often the depth of sand is less than the absoluteminimum of 0.5m usually recommended forslow sand filters (Visscher et al. 1987). Alsowith these small filters the ratio of the area ofsand in contact with the wall of the container tothe horizontal cross sectional area of the bed isvery high compared to any convenuonal slowsand filter. This means that in the small filters ahigher proportion of the raw water passes downthe easier flow path next to the wall (where it isnot filtered as efficiently as in the main bed ofthe filter). This detrimental effect can bereduced if the walls of the vessel are very rough.

In addition, many designs do not have a facilityfor filling the sand filter from the bottom whenthe filter is first commissioned or when the sandhas been washed and replaced. As explained forthe pond sand filter this is often detrimental tothe subsequent performance of the filter since aircan otherwise become trapped in the sand.

A final criticism is the suspicion that a memberof each household will be insufficientlymotivated to maintain the filter properly, evenwhen they have been adequately trained. Alsoafter the sand has been skimmed or replaced thefilter will not provide good quality water. Astorage vessel to contain previously filteredwater for drinking will therefore be requiredduring the period that the filter is producinglower quality water. This low quality water canof course be used for other domestic uses andneed not be wasted.

As was noted for the pond sand filters, althoughdomestic slow sand filters may not producebacterially pure water if there is no otheravailable source of drinking water they can beuseful because with a reasonable standard ofmaintenance they should improve the quality ofthe raw water from an unprotected surface watersource. An arrangement similar to that shown inFigure 5.5.2.5 or 5.5.2.6 would seem to showmost promise but as can be seen to satisfy therequirements for good slow sand filtration thelayout and operation becomes rathercomplicated. For best performance, for thereasons listed above, initial filling in both thesesvstems would need to be carried out through

the last pipe (for example by using water pouredinto it via a funnel).

More field work is required to examine thepracticalities, effectiveness and cost of suchdomestic filters and the ability of families to useand maintain them properly. If the surface watersource is acceptable for domestic uses other thandrinking and the community realise the need toonly drink filtered water then the size of filterrequired need only relate to the small volume ofwater used for drinking purposes.

The use of cement mortar jars as containers forthe filter sand would appear to be promising butthere could be problems with large vessels whenit is necessary to empty the sand when it needsto be cleaned or replaced. Flexible plastic tubingcemented into holes made in the jars wouldseem to be the best type of interconnectingpipework (see Section 4).

Use of geotextile filter matsIn recent years there has been some interestingresearch into the use of one or more layer ofthick open-weave geotextile fabrics on thesurface of conventional slow sand filters to makeit easier to clean the filters (e.g. ODA 1989,Wheeler et al. 1986, Graham, 1988). Theadvantage of the use of such fabrics is that theycan be periodically removed from the filtersurface carrying with them much of thedeposited solids which can then be washed outof the material before it is replaced. This leadsto much longer 'filter runs' because it takesmuch longer for the top layer of sand to clog.The use of fabrics also means that the upperlayers of sand are not disturbed very often. Thisis important since in these layers usefulstraining, adsorption and predation by micro-organisms leads to a dramatic, or even completeremoval of pathogenic bacteria and viruses(Visscher et al. 1987). Such fabrics are usedwith the small Potapak packaged slow sandfiltration water treatment plant mentioned in thelast section and with the much larger Oxfamslow sand filtration treatment package (OXFAMundated).

As previously mentioned the purificationprocesses which take place in bank or bedfiltration systems are not identical to those inslow sand filtration and the same level ofmicrobiological predation and straining action

Improving water quality by filtrationPivm water may Oe the only source available at times, butit is often airty ana not hvgierucolly safe A simple treatmentsystem to supply up to eignt families can be built usingtour covered ion, tour valves, plastic lutxng. sand (0.5- 3mmana 0.5mm grade) and travel, as foikws

The lubes con be titled to jarj either bv making holes inthe walls when the moflar is still soft or Dv cutting holes witha small hammer and a chisel (mode from a sharpenedscrewdriver) lubes con then bo sealed in with cementmortar.

Jar 1 is lor the storage and settlement of the m treatedwater ond should be as large as possible, VOIve A is fittedat the Dottom and is used for cleaning out the jar Valve Bis fitted about 100mm above A ana controls the flow ofwater to iar 2. Jar 1 is raised above the other jars in thesystem to proviae pressure to push water through.Jar 2 is an upwoid (tow filter which will remove much of thecoarse dirt A 250 litre |ar can treat 20 litres of water onhour Bock-wash the filler weekly by closing valve B andopening vatve C to d o n the iat Clean, but not hygienicalsafe, water can Oe diawn trom valve D.Jar 3 is a downward ftow flttst A 250 litre tor can treat 20lilies an hour.

It is cleaned Dy diainlng Ihe standing water off the topond scraping Ihe top 20mm of sand away every two or

Jar 4 is the container tor sate drinking water

• Tests in Thailana snow that the treated water can be freeof faecal bacteria tor prolonged periods of use

• information supplied bv S B Watt (UK). H Mann (UK). MrOpas (Siam Cement Co. Bangkok. Thailand). Professor RSuwanik (Siraq Hosolal. Bangkok. Thailand)

3+al7mm GRACE- CRAVEi.

THISJAR

CAN BE.AA/Y

^UiTABUT

size.

Slow sand filter system

Floating bowl in raw watertank

IIIIIIIIIIIIfII1III

•

which is useful for the removal of pathogens cannot be expected to result. However, the use offilter fabrics such as those described above couldbe useful in any man-made pond infiltrationsystems which have to deal with turbid watersince they should be effective in reducing therate at which the infiltration system will clog up,but some method of removing and cleaning thefabric will be need to be developed andultimately the sand will also need cleaning.Large areas of dirty fabric will be very hard tohandle and clean, so in such situations, smalleroverlapping pieces will probably be needed.

It is suggested that the use of fabrics with pond-bed infiltration systems, pond-bank infiltrationsystems, above ground pond sand filters andpossibly with domestic slow sand filters isworthy of further research, development andfield testing.

5.6 Ground Level Rainwater Catchments

The collection of rainwater runoff from groundlevel surfaces in considered to come into thecategory of surface water. There are threecategories of ground level catchments:

unimproved natural surfaces;improved natural surfaces;artificial surfaces.

Different aspects of these surfaces are discussedin the following sections.

Ground level surfaces are likely to becomecontaminated in a number of ways includingdeposits brought onto the catchment by people,animals (domestic and wild) and birds. Ifpossible the area which contributes runoff to areservoir should therefore be fenced off and thecommunity should be educated as to the impor-tance of avoiding entry into such areas and ofkeeping animals away from the catchment areasin order to reduce the risk of subsequentpollution of the runoff.

However pollutants can also be brought onto thesurface by wind and surface water from outsidethe fenced area. Very little can be done about thewind borne deposits but proper sanitarypractices in the surrounding area will reduce therisk of faecally contaminated dust being blownonto the catchment. With some catchments it

may be possible for the surface to be periodicallyswept with a clean brush used by someone withwell washed feet. The entry of surface waterfrom outside the fence can usually be preventedby raising the ground around the edges of thecatchment and/or using free draining ditches tointercept and carry away the polluted surfacewater (Figure 5.6.3 and 5.7.1). The use offences, ditches and bunds is also recommendedaround ponds to prevent the entry of pollutedsurface water from the land immediatelyadjacent to the pond, especially that falling inthe area where users approach it to collect water.

The quality of the water entering the storagepond or reservoir can be improved by making itnecessary for the water to first pass through asedimentation pond, silt trap and/or a gravelfilter or a coarse sand filter. These will needmaintenance since the collected silt, sand anddebris will need to be periodically removed orthey will block the filter or be carried on into thestored water. There is not much informationavailable about the performance of such methodsof improving the quality of water caught fromsurface catchments but details of somearrangements are shown in Figures 5.6.1 and5.6.2.

Although above existing ground level has manyadvantages diis is not usually possible withground-level catchments. However it may befeasible if the slope of the ground is sufficient toallow water from a catchment to gravitate to adownstream above-ground tank (Figure 5.6.3)or if a dam can be constructed (Figure 5.6.4).During the preparation of this report only a fewreferences to below ground storage were found.Some relevant details are given in Section 5.7.

The size of ground level catchments and theassociated storage are calculated in the sameway as for above ground tanks and catchmentsusing an appropriate runoff coefficient for thesurface. When calculating Hie volume of storageof uncovered or unlined reservoirs it will also benecessary to allow for the evaporation andseepage losses which can be quite high.

Some communities encourage floating plants tocover the water surface of open ponds to reduceevaporation losses but it is not clear whether mismethod is effective. Raised banks of excavatedmaterial around the reservoir or tees planted

70

Coarse mesharound intotiotank

Baffle acrosschamber

catcrvnantcollactadlnchannel

Drain

5 .G. IFigure • A s imple silt t rap

Siza o( chamber naaas to M largo anough so flowof watar into it does not causa alot of turbulence

which would pravant solids settling

Opan reservoir

Silt and dabris ara trappedb«tw««n gravel which alsoreduce* the turbulence ofttia water flowing into machamber

Inlet pipe discharges onto concreteslab to prevent erosion

Figure A gravel filter

Graval naads removing and cleaning ragularlyif it is to be effective. II it blocks watar can

ovarflow chamber and be lost

<House

Tap

^ ^

Tank

iVrrr.-f

Fence

ratfihnwnt X.

^T- //A&?R^ Surface watardiversion ditch

%.&•!> Figure Ground level catchment serving an above ground tap

Figure Various arrangements ot masonry dam wall* for rock catchments(Source: NISSEN-PETERSEN & LEE 1990b)


nearby (but at a distance where their roots donot draw water from the reservoir) can reducewind velocities an consequently evaporation.

Unimproved surfacesThe best natural land surface is a smoothsloping surface of impermeable rock or of soilswith high clay content (especially ones on whicha natural impermeable crust forms).

The proportion of runoff which can be collectedfrom a natural catchment depend on the soil andthe slope of the land. On steeper ground smalldepressions are less likely to intercept the flowbut the problems of erosion of the surfacebecomes more of a problem. Natural landcatchments, if sloping less than 1 in 10 (10%),will rarely have runoff coefficients above 0.3.

Improved natural surfacesThe runoff coefficients of natural surfaces can beimproved by a number of techniques includingthe addition of cement, lime, clays or chemicalsto the soil (Maddocks, 1975). The simplestmethod, but one best suited to mechanical plant,is to smooth and compact the soil. The use ofnone of these methods is usually feasible in thesmall communities on which this report focuses,but the less effective manual methods ofimproving the catchment are also worthwhile.These include the removal of vegetation (as longas this will not promote erosion), the smoothingof the catchment area and the filling ofdepressions. Rock catchments can beconsiderably improved by removing soil andvegetation and filling any open joints with clayor cement.

Some local technologies such as mixing soilwith clay, marl, lime or dung are used in anumber of developing counines to produce fairlyhard an impermeable surfaces for flat roofs,floors to houses or threshing floors, and thesehave obvious applications to small ground levelcatchments. Threshing floor catchments areused in Botswana (see Figure 5.6.5).

Artificial ground level catchmentsNumerous artificial ground coverings have beenused in developed countries for rainwatercatchments. Most of these are based on a layer ifimpermeable material (such as concrete orbitumen) applied to the surface, or sheets offlexible membrane (such us PVC or butvl

rubber) laid over the surface. Such surfaces areusually quite costly and are therefore rarelyappropriate in developing countries.

One interesting form of catchment which couldbe cost effective is suggested by Morgan (1990)and is illustrated in Figure 5.6.6. This uses alarge piece of plastic sheeting buried in sand toprotect it from physical damage anddeterioration due to sunlight. No details aregiven by Morgan about the long termperformance of such catchments but like thepolythene lined tanks mentioned later, the sheetmay be susceptible to attack by termites. Oneadvantage of the use of die sand on thepolythene is that it will strain out some of thecontaminants that might be carried onto thecatchment.

Another design, proposed by Kukita (undated)of UNICEF Namibia uses a polythene sheetunprotected by sand but no field report of its usewas available for examination at the time ofwriting this report. It is believed that it wasenvisaged that the sheet would only be usedduring four months of each year (presumablyjust during the rainy season) and that it wouldbe stored out of sunlight for the rest of the yearso that the useful life of the polythene could beas long as to 5 or 6 years.

5.7 Below-Ground Tanks And Reservoirs

5.7.1 Unlined Excavations And NaturalDepressions

If the soil is fairly impermeable then water canbe held in any natural depression or excavationand such ponds provide a water sources formany people in developing countries. Suchtanks are widely used in countries such asYemen and Sudan for example. Evaporation andseepage losses from such water bodies can bequite high over a long period. The evaporationlosses can be reduced by using deeper reservoirs,which will have a smaller surface area per unitvolume. The seepage losses can be reduced bysome form of lining.

5.7.2 Lined Excavations And NaturalDepressions

Traditional linings

Clay lining although labour intensive is feasibleif suitable material is readily available. Sandyclays are preferable to pure clays since they areless likely to crack when the water level fallsand they dry out. Sometimes there is alreadysufficient clay in the base and sides of the pondbut it needs to be 'puddled', that is worked whenwet and plastic to improve its impermeability.This can be done on a small scale by peopletrampling it with their feet, but driving livestockback and forth over the clay rich soil is moreeffective for larger scale projects. The additionof cattle dung to the soil on the floor of thereservoir is also reported to prevent seepage(Nissen-Petersen, 1990) and presumably alter aperiod of time it offers no risk of contaminatingthe water.

Polythene liningPolythene sheeting is not very strong and it iseasily punctured during and after installation.Also it is not very easy to make waterproof jointsbetween sheets and it rapidly deteriorates whenexposed to sunlight (although black polythene ismore resistant to ultra-violet deterioration thanclear polythene). It is inferior to the veryexpensive man-made membranes such as PVCand butyl rubber but it is likely to be the onlymembrane material affordable in the situationsconsidered in this report.

Widespread successful use of polythenemembranes is not reported in the literature but ithas been tried in Botswana. Gibberd (1969)reports the use of four layers of 0.0381mm thickpolythene sheets sandwiched between thin layersof mud and protected with sloping revetmentsmade from of layers of horizontal tubes ofpolythene sheet filled with a weak cementmortar mix and pinned together with heavygauge wires (Figure 5.7.1). Gould (19X4)reported that for various reasons only four of thetanks constructed by this method in Botswanaand Zimbabwe proved to be successful but hesuggests that if they are carefully constructedthis design could give good service. Pacey andCullis (1986) indicate that a similar techniquehas been used in West Africa.

Some designers have suggested the use of avertical cylinder of polythene in a cylindricalexcavation to form a storage tank. This wasrecommended by Kukita (undated) inconjunction with the polythene catchment

mentioned above. He proposes that the inside ofa 1.2m diameter excavation which is 5m deep isfirst lined with mud to prevent any sharp stonespiercing the polythene. He suggests that thesheets should be joined with splicing tape. Thereliable performance of this tape is critical to thesuccess of the reservoir. The reservoir is coveredwith a slab equipped with a handpump in asimilar way to a protected well. A similar ideawas noted in a French publication where thepolythene sheet was ready supplied in a longflattened cylindrical form which can be drawntogether at one end which is then tightly boundwith string and doubled back on itself beforebeing tied again to produce a waterproof end.

One problem often encountered with polythenelinings is damage from the roots of plants andtrees and from the action of termites. The formertwo can be controlled to some extent by carefulsiting and maintenance. Insecticide isrecommended by some authors to controltermites but in a study of 'hafirs', which usedvarious arrangements of polythene sheeting andlayers of mud for lining, Pontin and Woolridge(1977) found that even when insecticide wasused, termite damage was still the main cause offailure. In areas where this problem might notoccur or where it can be overcome there ispotential for further field tests of polythene linedtanks.

Un-reinforced cement mortar liningUnreinforced cement mortar lined excavationswith capacities of up to HP are reported to beused in Togo and ones of up to 8m^ to be usedin Tanzania (Lee &. Visscher, 1990). O'Brien(1990b) briefly reports the use in Togo of a12.8m3 capacity cement lined spherical pit witha dome shaped concrete cover using no metalreinforcement which used materials costingUS$240. .Since the mortar is not reinforced therisks of cracking must be high unless the soil isvery firm. Where no field experience ofsuccessful un-reinforced mortar linings isavailable it is therefore suggested thatferrocemenu although probably more expensive,is the safer option.

Ferrocement liningAbove ground ferrocement tanks were discussedin a previous section and many of the aboveground designs can also be buried. The use ofexcavated ferrocement lined hemispheres was

72



also mentioned in that section and this methodof construction has been well proven in Kenyawhere sizes up to lOOm^ have been constructed(Gould, 1992), often using barbed wire toprovide additional reinforcement to the chickenmesh [Nissen-Petersen & Lee (1990a). UNICEF,(1986)]. The timber framed roof (Figure 5.7.2)originally used with such reservoirs to reducethe rate of evaporation and to protect the storedwater was found to rot. Now a domedferrocement roof is preferred, supported ifnecessary by a central concrete pillar. WaterAidhave used a conical ferrocement roof andsupporting pillar in a design promoted in Kenya(WaterAid, 1992).

Sand Tilled reservoirsSome writers suggest the use of excavations inimpermeable ground filled with sand and graveland to store water in the pores between suchgranular material (Figure 5.7.3). Like thegroundwater dams discussed earlier this reducesthe rate of evaporation and offers someprotection from contaminants. The disadvantageof such an idea is that because the voids onlytake up a small volume of the material (rarelymore than 25% for a natural river deposit(Nilsson, 1988)) a large excavation and a largevolume of imported sand are needed. Howeverrates of evaporation from exposed water surfacescan be as much as 2000mm per year comparedto the very low rate from a filled excavation sothe actual volume of water needing to be storedis less. Water is usually extracted from suchsand filled reservoirs using a well with a suitablewater lifting device.

5.S Improving The Quality Of The Water BySettlement. Long Term Storage. SolarDisinfection Or Chemical Disinfection At TheUsers Home

5.8.1 Settlement

If it is not feasible to purify surface water bysome method of filtration improvement canresult from allowing the suspended solids tosettle and then abstracting the clearer waterfrom the top of a vessel or reservoir. This has;Uready been mentioned with respect to the useof floating intakes and sedimentation ponds.

A domestic method using the same principles,sometimes called 'the three pot system' is

promoted in some countries. It follows a threeday cycle. On the first day water collected fromthe source is put into a pot where it is leftundisturbed so that the suspended solids cansettle out. The next day the clearer water at thetop of the first pot is carefully poured into thesecond pot. The dirty water and settled solids inthe bottom of the first pot are then emptied andit is refilled again from the water source to startanother cycle. On the third day the clear waterat the top of the second pot is poured into a thirdpot which is used for drinking water.

In addition to the benefit of the clearer waterobtained from settlement this process alsoresults in the removal of those bacteria whichare attached to the settled solids. Although thiswill improve the bacteriological quality of thewater it will sail be polluted by other bacteria sostill cannot be considered to be safe. If there areinfected copepods in the water they too may beremoved by this settlement process but it shouldnot be relied upon as a removal method. Theirremoval is much more reliably carried out bystraining which can take place as the water ispoured into the drinking water storage vessel.

If water is collected from a source whichcontains species of snails which are intermediatehosts in the transmission of schistosomiasis(bilharzia) the water may be infected withcercariae which can infect man (Section 2).These become non-infective after about 48 hours(Cairncross and Feachem, 1983) so the risk oftransmission of this disease through drinkingwater is removed by any sedimentation orstorage methods which hold the water for morethan two days after collection from the source.

Some communities use locally availablecoagulants such as powdered Morinda seeds(Jahn, 1981) or alum purchased from local mar-kets to act as a coagulant to speed the rate atwhich the finer solids settle in a single pot. Suchcoagulants can of course still be used togetherwith the three pot system to improve theeffectiveness of the process. These coagulantsusually capture many of the unattached bacteriawhich settle with the floes (cit.ibid.), leading toa much higher rate of removal than by usingsettlement alone and sometimes giving aremoval rate as high as 99.9 per cent(Sutherland etal., 1990).

70*

Eiltnsion woll 65cm high of bricks or blocks (OR) •xttnaan walla* ttonts stt in mortar

• Hanapumq

3 63

(W) reintorcM witn 12 rounds ol barbtdwirt 55cm apart ana covtrto WI I nmoftdc

5cm»100fm BVWIIIIW

Stonx

f ni irwlntmn nl mortar II3) rtintorCtd

»}V2 x 342em61 OJB» with 2-3mm

Figure Cross section of hemispherical ferrocement tank(Sourca: NISSEN-PETERSEN & LEE 1990)

Tha tarrocamam is 55-60mm thick rainforcad with strands of I6g barbad wira at 200mm cantras and ona layer of chickan mash

^ ^ C ^ I R A I N W A T E R CATCHMENT

DRAIN PIPE *%V*Sb

EXCAVATIONwater prootea with anysuilaDle material anabacklilltd with cleansana or gravel

'#*? INFILTRATIOH

~p—-~——_— — —

I. —i a /

dk /In fir*

WATER WITHDRAWAL WELL

Figure S".7>3Artificial recharge using rainuater

Source: ITDG


IIIIIIIIIIIIIIIIII

Settlement alone does not produce good qualitywater and often cannot over a few days producecompletely clear water. Although settlementalone or settlement alter small scaJe coagulationand flocculation will not produce bacteriaUy freewater they give useful improvements to thequality of the water and should be promotedwhere sustainable. Both are useful for precedingsand filtration methods because the removal ofsetUeable solids will extend the period ofoperation of a filter. They are also useful prior tochlorination since they will reduce the wastageof chlorine which oxidises organic solids, someof which will be removed by settlement andmost of which could be removed by the use ofcoagulants.

If the use of natural coagulants such as Moringaseeds can be promoted then the speed andamount of clarification which takes place will beconsiderable and the resulting quality of thewater will be much improved. Furtherinvestigation into existing field experience (eg.[Dian Desa (1985), Jahn (1984) and Sutherlandet.al (1990)] of such coagulants should becarried out to see if they are worth promoung fordomestic use.

Cement mortar jars are useful containers inwhich settlement of suspended solids can takeplace in a home. Where other betterinterventions are not available it isrecommended that the use of such jars should bepromoted together with straining out ofcopepods.

5.8.2 Long Term Storage

If surface water is stored for a long period thenthere is more time for suspended solids to settleout and this can lead to a clearer water.Fortunately, because the conditions in storagevessels are not very conducive to the survival offaecal bacteria after a period of time these alsodie. Although some studies have been done intothe rate of die off in open reservoirs (WRC,1975 and Smethurst, 1979) tins study has no'found much research relating to the die off raicin smaller covered vessels. One reference (Jahn,1984) quotes some research from Tanzania(MWDEM, CIDA & SID A. 1979) whichshowed a 90 per cent die-off for most bacteria;ifter two days of storage. It is foreseen that withhighly turbid waters taste and odour problems

may arise from the growth of anaerobic bacteriaduring long term storage so it would be best toprecede it with some settlement to reduce theorganic content of the water to a minimum.

Muller (1972) found that infected copepodslived for a maximum of 50 days. They may livefor a much shorter period in the unnaturalenvironment of a covered container and thisaspect might be worthy of further study if thereare communities where straining is unlikely tobe adopted but where long term storage wouldbe acceptable.

If long term storage of surface water fordrinking purposes is feasible in a communitythen there is another way of avoiding theconsumption of infected copepods. This is tostore water collected outside the season whencopepods are infected with Guinea worm larvaeand to consume only this water during theinfective season(s). It is doubted whether inpractice this pattern of consumption could beimplemented and it would probably be mucheasier to protect against the transmission ofGuinea worm by straining, and if feasible to uselong term storage to improve the bacteriologicalquality.

If a system of domestic filtration of water whichdoes not remove all bactena is adopted thensubsequent long term storage may possibly beused to allow them sufficient time to die off.

Cement mortar jars with lids would seem to beideal storage vessels for long term domesticstorage. It is suggested that a more exhaustiveliterature, search and further investigation intothe effectiveness or practicalities of long termdomestic storage is carried out in the field.

5.8.3 Solar Disinfection of Water

In the early 1980s work by Acra at in Beirut(Acra et al., 1984) showed that exposing a bottleof clear water to intense sunlight for a few hoursresulted in the deaths of many, if not all, of thefaecal bacteria contained in the water. This is asa result of mainly the ultra-violet component ofthe sunlight. Since then numerous otherresearchers have examined this phenomena. In1988 a workshop brought many of theseresearchers together to discuss the topic and the

proceedings give a state-of-the-art review(IDRC, CRDI & CIDD, 1988).

The workshop confirmed Prof. Acra's findingsthat solar radiation exerts a germicidal effect onsmall quantities of bacterially contaminatedwater. However such treatment is not feasibleduring periods of continuous rainfall and heavyovercast conditions. It has been demonstratedthat on a clear sunny day, clear water, with abacterial coliform count of less than 1000coliforms/100ml will usually be free of livebacteria after about 3 hours exposure in a cleartransparent glass or plastic bottle (cit.ibid). At aradiation rate of 500W/m^ this is accomplishedin a longer period of about 5 hours (cit.ibid).The effect on other pathogens found in water isalso reported and research is continuing in thisfield. Literature on the effect of solardisinfection on copepods has not been noted butthey are probably unaffected by it.

The use of solar disinfection therefore certainlyhas potential for improving the bacteriologicalquality of clear water during certain seasons ofthe year but as yet it has not been extensivelyfield tested to see how its adoption by people inrural communities works out in practice but amore extensive literature search may discoversome field reports.

It would seem that where bottles are available,and clear bacterially impure water sources are inuse, that there is nothing to loose from theintroduction of this technique if people find itacceptable. Since the water will become warmduring its period of exposure it will probably beconsidered unsuitable for direct consumptionand it will need storing until it has cooled. It isbest stored in the same container in which it wasexposed so that there is no danger of itbecoming contaminated when stored in anothervessel but this will mean that a large number ofbottles will be needed. The water drawn fromthe suction infiltration systems or floatingintakes mentioned earlier in this section shouldin particular benefit from solar disinfection.

5.8.4 Domestic Chemical Disinfection

Problems with the use of chemicals byhouseholders to kill copepods was discussed inSection 5.3.1. The fact that chlorine and iodinewill also kill faecal pathogens was noted. For the

reasons given in that section the authors do notconsider that domestic chlorination of drinkingwater to make it bacteriologically pure issustainable in the type of communities beingconsidered in this report. No field reports of itsextensive use for this purpose in such situationsare know to exist.


IIIIIIIIIIII

Inin

6. CONSIDERITS1G THE COMMUNITY

This study has investigated 'Cost effectivetechnologies' for eradication of Guinea wormand for potable water supply for smallcommunities.

None of the technologiesinvestigated will be effectivewithout adequate consumerinvolvement.

This involvement by households andcommunities may be reflected in communitymanagement of projects or in customer choice oftechnologies. Any involvement will bedependent upon information, education(particularly health education) andcommunication from project sponsors and oilierinterested agencies.

The results of this study indicate thatappropriate CHOICE between the variousoptions for cost effective technology is moresignificant than any individual technology. It isnot the role of this study to investigateprogramme implementation. However, toachieve cost effective technology theimplementation process must be designed toenable consumers to choose the technologywhich they find to be most suitable. Obviouslymost consumers do not have the opportunity tounderstand all the options available. Tlierefore itis the responsibility of the sponsors to presentthe most relevant options in any particulargeological, hydrological and economic settingin a way that consumers can understand andappropriate.

To maximise community involvement andtherefore ensure health and social benefitsthe role of implemeniers needs to change from a'top-down project based approach' to a marketresearch and development with consumerchoice based approach'.

This change from a 'supply driven'approach toa 'demand led' approach will require asignificant change of thinking on behalf ofproject managers and administrators. The role

changes from that of engineering designer andprovider to market researcher and seller.

Figure 6.1 illustrates the likely affects of aproject approach where rapid take-up can beachieved but with a significant reduction inbenefits when the initial phase is completed.The community or consumer (or customer)approach will always have a longer time lag butwill be more efficient and effective in the longterm. For sponsors who desire sustainablebenefits as rapidly as possible the onlysolution is to increase their spending oninformation, education and communication(marketing and selling) dramatically toencourage a faster take-up of the consumer'schoice of Improvement.

Further information on promoting communityinvolvement and demand-led approaches toeradicate Guinea worm and domestic watersupply may be found in:

Community-Based Initiatives To EradicateGuinea Worm: A Manual for Peace CorpsVolunteers (Silverfine, Brieger and Churchill,1991);

Guidelines for Health Education andCommunity Mobilization in DracunculiasisEradication Programs (CDC, 1991);

Adding Guinea worm control components:Guidelines for Water and Sanitation Projects(Prins and Yacoob, 1988);

Infrastructure for Low-income Communities(Francevs, 1991).

h

7. CONCLUSIONS ANDRECOMMENDATIONS

There is a large range of cost-effectivetechnologies available for water supply inGuinea worm endemic areas.

For the sake of long-term consumerunderstanding and health promotion it isrecommended that wherever possible assistanceshould be aimed at providing potable drinkingwater, not just water free from infective Guineaworm larvae.

The results of the research indicate that the issueis not primarily the determination of a cost-effective technology but rather the clwice oftechnology suitable for a particular location.This is of vital importance to achieve lifecyclecost effective technologies which are necessaryto ensure sustainability.

Locations differ according to socio-economic,hydro-geological and hydrological conditions.To obtain the most suitable choice thecommunity/household/consumer need to beinvolved in an education, information andcommunication process that will enable users tomake their own choice according to theirunderstanding of improvements in health andconvenience and according to their willingnessto pay for sustaining those improvements.

The desired rate of implementation can beachieved only by innovative use of promotionand marketing campaigns not normallyassociated with 'civil engineering products'.

Cost effectiveness of implementation is likely tobe enhanced by the encouraged involvement ofprivate sector suppliers.

The starting point for upgrading to potablewater free of Guinea worm is normally toimprove existing water sources, though theprotection of potable water is difficult withsurface water sources.

Life-cycle cost effective technologychoices: Groundwater

When making choices, the exploitation ofgroundwater is normally the first priority since

it can often supply sufficient quantities of goodquality water.

Within this category every effon should be madeto cap any spring sources (and where possible topipe the water to a number of more convenientcollection points)

If there are no springs and communities arescattered but groundwater is relatively shallow,a number of hand drilled/hand dug wells withBlair type bucket pumps should be promoted.Where population density is higher the sametechnologies can be used with direct actionhandpumps.

If the ground conditions make it necessary,powered mini-rigs are applicable with deepwellopen top cylinder pumps for water collectionfrom beyond 15m depth.

Life-cycle cost effective technologychoices:Rainwater

Where groundwater cannot be exploited,rainwater catchment using roof catchments withsupported "V' gutters and cement mortar jars canbe promoted to give a good quality of potablewater. Nylon sack catchments can be used wherethere are only thatched roofs.

Life-cycle cost effective technologychoices:Surface water

Where there is no alternative but to use existingsurface water sources, strainer boxes can be usedto prevent Guinea worm as an interim measureeven where there are no pumps.

The next stage of improvement for surface wateris to use hand powered suction pumps withfloating intakes protected by strainers.

Neither of these methods remove pathogenicbacteria so they should be promoted inconjunction with household cement jars forstorage (with strainer covers as an addedprecaution). Any length of storage will begin toreduce the risk from pathogens not removed bystraining.

IIIIIIIIIIIII

I


Surface water sources should be upgraded toinfiltration galleries or river bank wells toensure potable water at the earliest opportunity.

Portable, individual strainers have a vital role toplay as back-up to the other methods ofpreventing the ingestion of Guinea worm larvaeand these can be used by people moving awayfrom protected household sources.

7N

REFERENCES

Acra A. et al., (1984), Solar Disinfection ofDrinking Water and Oral Rehydration Solution -Guidelines For Household Applications inDeveloping Countries, American University ofBeirut

Akhter M., (1990), SWACH: An IntegratedApproach to Control Guinea Worm Disease inIndia, Water Quality Bulletin, Vol. 15, No. 1,Canada Centre for Inland Waters, Burlington,Ontario, Canada, pages 23 - 28, and 63,

Akhter M., (1992), Surgical extraction ofGuinea Worm, WATERfront, Issue 1, Waterand Environmental Sanitation Section,UNICEF, New York, page 8

Appan A., Villareal C.L. and Wing L.K.,(1989), The Planning, Development,implementation of a Typical RWCS: A CaseStudy in the Province of Capiz, proceedings ofthe 4th International Conference on RainwaterCistem Systems, Manila, Philippines (quoted byGould (1991))

Arlosoroff S., Tschannerl G., Grey D., JourneyW,, FCarp A., Langenegger O., and Roche R.,(1987), Community Water Supply - TheHandpump Option, The World Bank,Washington, D C , USA

Bauman, Erich, (1992), (Handpump specialistwith 3 years recent experience in Ghana, SKAT,Switzerland), personal communication.

Beale, G. Water Management Consultants(1990), Development of the Drilling Industry inNigeria - An Overview, (unpublished), quotedbydeRooy(1992)

Blankwaardt, Bob, (1984), Hand Drilled Wells -A Manual on Siting, Design, Construction andMaintenance, Rwegarulila Water ResourcesInstitute, Dar es Salaam, Tanzania

Bradfield W.J., (1992), Extracting clean waterfrom streams in south-western Ethiopia.Waterlines, Vol. 10, No. 3, IntermediateTechnology Publications, London, U.K., pages30-32

Brandford, Bonnie, (1992), Focus on Thailand,Raindrop Rainwater Harvesting Bulletin,Volume 7, Water and Sanitation for HealthProject, Arlington, Virginia, U.S.A., pages1,3,6,7 and 10.

Brieger W., Ramakrishna J. and Adeniyi J.,(1986), Community Involvement in SocialMarketing: Guineaworm Control, InternationalQuarterly of Community Health Education, Vol.7, Issue 1, pages 19-31

Bugri, Sam, (1992), (Co-chair DORN, GhanaMinistry of Health), personal communicationCairncross, Sandy and Feachem, Richard G.,(1983), Environmental Health Engineering inthe Tropics - An Introductory Text, Wiley,Chichester, U.K.

Caimcross, Sandy and Feachem, Richard,(1986), Small Water Supplies, Bulletin No. 10,The Ross Institute of Tropical Hygiene, LondonSchool of Hygiene and Tropical Medicine,London, U.K.

Carter, Richard, (1985), GroundwaterDevelopment Using Jetted Boreholes,Waterlines Vol. 3, No. 3, IntermediateTechnology Publications, London, U.K.

Carty, Dermoi, (1990), Water-well bucketprotection in Zambia, Waterlines, Vol. 8, No. 4,Intermediate Technology Publications, London,U.K., pages 26 - 28

I


CDC (1989), Guidelines for Chemical Controlof Copepod Populations in DracunculiasisEradication Programmes, World HealthOrganisation Collaborating Center for Research,Training and Control of Dracunculiasis, Centersfor Disease Control, Atlanta, Georgia, U.S.A.

CDC (1991), Guidelines for Health Educationand Community Mobilization in DracunculiasisEradication Programs, World HealthOrganisation Collaborating Center for Research,Training and Control of Dracunculiasis, Centresfor Disease Control, Atlanta, Georgia, U.S.A.

Chauvin J., Plopper S. and Malina A., (1992),Final Evaluation of the Benin Rural WaterSupply and Sanitation Project, WASH FieldReport No. 349, Water and Sanitation for Health(WASH), Washington D.C., U.S.A.

Childers, Laurie and Claasen, Fran, (1987),UNICEF's upward-flow water filter, Waterlines,Vol. 5, No. 4, Intermediate TechnologyPublications, London, pages 29 - 31

Consallen (1992), personal communicationwith, and trade literature from Consallen GroupSales Ltd., Loughton, Essex, U.K.

Cornish, Gerald A., (1987), Well Jetting inMexico, Developing World Water, GrosvenorPress International, Hong Kong, pages 94 - 97

Davis, Colin, (1992), Hydrofracturing convertsdry wells into successful wells. Issue 1,WATERfront, UNICEF, New York U.S.A.,pages 24 & 23

de Rooy C, Kamfut M. and Sambile O. (1986a),An Empirical Resistivity Model forGroundwater Prospecting on Basement Rocks,UNICEF - Assisted Water and SanitationProgramme in Nigeria, Papers presented at thefirst annual symposium and training workshopon ground water resources in Nigeria, 23rd -25th July 1986, UNICEF, Lagos, Nigeria

de Rooy C, (1992), Personal correspondence(information given in documentation submittedto the Water, Engineering and DevelopmentCentre (WEDC) Loughborough Univers, ofTechnology, U.K., during the preparation of uPh. D thesis)

de Vrees L., (1987), Machakos DioceseRainwater Tank Programme, DevelopmentOffice, Diocese of Machakos, Kenya, (quoted byGould (1991))

Deveraux, Scott, (1992), (Researcher studyingdismantleable connectors and other aspects ofuPVC rising mains for handpumps, Consumers'Research and Testing Laboratories, Harpenden,U.K.), personal communication.

DHV, (1979), Shallow Wells (2nd edition),DHV Consulting Engineers, Amersfoort, TheNetherlands

DHV (1985), Low Cost Water Supply - ForHuman consumption, cattle watering, smallscale irrigation. Part 1: Survey and Constructionof Wells, DHV Consulting Engineers,Amersfoort, The Netherlands

Dian Desa (1985), Water Purification withMoringa seeds, Waterlines, Vol. 3, No. 4,Intermediate Technology Publications, London,U.K., pages 22 and 23

Dodge C.P., Skoda J.D., Ekvall T. and Espie A.,(1986), Drilling and pump replacementincentives in Uganda, Waterlines, Vol. 5, No. 2,pages 19-22

Donaldson L., Olsson U., Anderson B. andOlushola I., (undated but believed to be 1987),An operational modality for the tripod (manual)rig, 9 pages, diought to be an unpublishedreport, UNICEF, Lagos, Nigeria

Donaldson, Lloyd A., (1992), Rural water andhealth: the challenge to water and sanitationsector professionals, WATERfront, Issue 1,UNICEF, New York U.S.A., pages 2, 3, 11 and19

DPHE & UNICEF (1989), A Report on theDevelopment of a Pond Sand Filter, Departmentof Public Health Engineering, Government ofthe Peoples Republic of Bangladesh, andUNICEF, Dhaka, Bangladesh

Duke, Brian O.L., (1984), Filtering out theGuinea Worm, World Health, March 1984, page29

Edwards D., Keller K., and Yohalem D., (1984),Workshop Design for Rainwater RoofCatchment Systems - A training guide, WASHTechnical Report No.27, WASH, WashingtonD.C., U.S.A.

Eureka (1992), personal communication with,and trade literature from. Eureka UK Ltd,Hassocks, West Sussex, U.K.

Feachem, Richard, (1980), Bacterial standardsfor drinking water quality in developingcountries, The Lancet, August 2,

Franceys R.W.A., (1987), Handpumps •Technical brief No. 13, Waterlines, Vol. 6, No.1, Intermediate Technology Publications,London, U.K., pages 15 - 18. (also in ITP(1991))

Franceys R.W.A., (1991), Infastructure for Low-Income Communities, unpublished Ph.D. thesis,Loughborough University of Technology, U.K.

Gibberd V., (1969), Introduction of RainwaterTanks and Micro-Irrigation to Botswana,Intermediate Technology Development Group,London, U.K.

Gould, John, (1984), Rainwater catchmentsystems in Botswana: present, past and future,Waterlines, Vol. 2, No. 4, IntermediateTechnology Publications, London, U.K., pages14-19

Gould. John E., (1991), Rainwater CatchmentSystems for Household Water Supply,Environmental Sanitation Reviews, No. 32,Environmental Sanitation Information Centre,Asian institute of Technology, Bangkok,Thailand

Gould, John, (1992), Recent developments inrainwater catchment systems, Waterlines, Vol.11, No. 1, Intermediate TechnologyPublications, London, U.K.

Graham, Nigel (1988), Fabrics and Slow SandFilters, Developing World Water, GrosvenorPress International, Hong Kong, pages 172 -173

Herbert R., (1990), Dug well vs. collector wellperformance, British Geological Survey (BGS)Technical Report No. WD/90/34, BGS,Key worth. Nottinghamshire, U.K.


IIIII1III1IIIII1IIII,

Herbert R. and Rastall P., (1991), Developmentof Horizontal Drilling for Alluvial Aquifers ofHigh Permeability ODA/BGS R & D Project(91/7), Final Report on Work in Zimbabwe,British Geological Survey (BGS) TechnicalReport WD/91/50, BGS, Key worth,Nottinghamshire, U.K.

Hilton, David J., (1989), Further development ofthe inclined coil pump, Waterlines, Vol. 8, No.2, Intermediate Technology Publications,London, U.K., pages 20 - 22

Hofkes EH. (editor), Huisman L., SundaresanB.B., De Azevedo Netto J.M. and Lanoix J.N.,(1983), Technical Paper Series No. 18,International reference Centre for CommunityWater Supply and Sanitation (IRC), Wiley,Chichester, U.K.

IDRC, CRDI & CUD (1988), Solar WaterDisinfection, Proceedings of a workshop held atthe Brace Research Institute, Montreal, Que.,Canada, 15 - 17 August 1988

lies, Karen, (1989), Harvesting rainwater usinga simple catchment surface, Waterlines, Vol. 7,No. 4, Intermediate Technology Publications,London, U.K., pages 23 - 25

Inalsa (undated), India Mark II Deep Well HandPump, trade literature, inalsa, New Delhi, India

Institute for Rural water, (1982), Constructing aHousehold Cistern, Water for the WorldTechnical Note RWS.5.C.1, Agency forInternational Development, Washington D.C.,U.S.A.

IRC (1988), Handpumps - Issues and conceptsin rural water supply programmes, IRCTechnical Paper Series No. 25, InternationalWater and Sanitation Centre (IRC), The Hague,The Netherlands

IRC (1988b), Community Self-Improvements inWater Supply and Sanitation - A training andreference manual for community health workers,community development workers and othercommunity based workers. Training Series No.5, International Water and Sanitation Centre(IRC), The Hacue. The Netherlands

Ismail, Olushola (1992), personalcommunication with UNICEF Nigeria

ITP (1991), The Worth of Water - Technicalbriefs on health, water and sanitation.Intermediate Technology Publications, London,U.K.

Jackson, Brian (1992), Overseas DevelopmentAdministration (ODA) Engineering Division,Personal communication

Jahn, Samia Al Azharia, (1981), TraditionalWater Purification in Tropical DevelopingCountries - Existing Methods and PotentialApplication, German agency for TechnicalDevelopment (GTZ), Eschbom, Germany

Jahn, Samia Al Azharia (1984), Traditionalwater clarification methods - using scientificobservation to maximise efficiency, Waterlines,Vol. 2, No. 3, Intermediate TechnologyPublications, London, U.K., pages 27 and 28

Jahn, Samia Al Azharia, (1986), Proper use ofnatural coagulants for rural water supplies -Research in the Sudan and a guide for newprojects, GTZ, Eschbom, Germany

Jones Charles, (1987), Investigations for a RuralWater Supply Project in Northern Nigeria,Developing World Water, Grosvenor PressInternational, Hong Kong, pages 144-145

Joy, Derek (1983), Cleaner water from fast-flowing nvers, Waterlines, Vol.1 No.4,Intermediate Technology Publications, London,U.K., pages 26 -28

Kukita. Jun, (undated), Plastic Lined Rain-Water Catchment System, unpublished memo,UNICEF, Windhoek, Namibia

Lambert, Robert, (1989), How to Make a Ropeand Washer Pump, Intemediate TechnologyPublications, London, U.K.

Lambert R.A., (1992)the need for and design ofa pressurised discharge human-powered treadlepump. Proceedings of the Institution of CivilEngineers (ICE), Vol. 92, Issue 2, ICE,London,U.K., pages 66 - 73

Larsson, Robert W., (1992), Strategies for theEradication of Dracunculus medinensis fromAfrica, unpublished paper (presented asqualification from the M.Phil, into the Ph.Dprogramme at the Water, Engineering andDevelopment Centre, Loughborough universityof Technology, U.K.)

Latham, Brian, (1984), Building the Thai jumbowater jar using a brick form, Waterlines, Vol. 3,No. 2, Intermediate Technology Publications,London, pages 26 - 28

Laver, Sue, (1987), Well Sinking - a step by stepguide to the construction of wells using theblasting method, UNICEF, Harare, Zimbabwe

Lee M.D. and Visscher J.T., (1990), WaterHarvesting in Five African Countries,International Water and Sanitation Centre(IRC), The Hague, The Netherlands

LSHTM and Gifford (1986), EvaluaUon of SWSFiltration System, Overseas DevelopmentAdministration Project No. R3957, LondonSchool of Hygiene and Tropical Medicine,London and Gifford and Panners, Southampton,U.K.

Maddocks D., (1975), Methods for creating lowcost membranes for use in the construction ofrainwater catchment and storage systems,Intermediate Technology Publications, London,U.K.

McPherson H.J., Livingstone A., MohamedY.A. and Liao M., (1989), SustainableOperations and Maintenance of Rural WaterSupplies in Sudan, Community water Supplyand Sanitation WHO/CWS/89.19

Metianu, Andrew, (1982), A simple method ofjetting tubewells, Waterlines Vol. 1, No. I,Intermediate Technology Publications, London,U.K., pages 6-8

MHCE (undated), Land and Water Resources ofAfrica - A Teachers' Guide, Moray HouseCollege of Education, Edinburgh, U.K.

Morgan, Peter. (1990), Rural Water Suppliesand Sanitation, Macmillan, London, U.K.

Morgan, Peter and Chimbunde Ephraim,(1991), Upgrading family wells in Zimbabwe,Waterlines, Vol. 9, No. 3, IntermediateTechnology Publications, London, U.K., pages10-12

Mudgal A.K., (1992), (Technology Specialist,UNDP/World Bank Water and SanitationProgramme, New Delhi, India), personalcommunication

Muller R., (1970), Laboratory experiments onthe control of CycJops transmitting Guineaworm, Bulletin No. 42 , World HealthOrganisation (WHO), Geneva, Switzerland,pages 563 - 567 (quoted in CDC (1989))

Muller R. (1972), Maintenance of Dracunculusmedinensis in the laboratory and observationson experimental infections, Parasitology No. 64,pages 107 - 116, quoted in Muller (1985)

Muller, Ralph L.J., (1985), The Biology ofCyclops in Relation to GuineawonnTransmission and Control, in Dracunculiasis inNigeria, Conference Proceedings March 25 th -27th 1985, Federal Ministry of Health, Ikoyi,Lagos, Nigeria, pages 22 - 29

llll

l

ll

ll

IIIIII1IIIIIIIIIII

MWEDM, CIDA & SID A (1979), Rural WaterQuality in Tanzania, Ministry of WaterDevelopment Energy and Minerals, CIDA andSID A, (quoted by Jahn 1981)

Naugle, Jonathan, (1992), Affordable water forirrigation: An experience in Niger, WaterlinesVol. 11, No. 1, Intermediate TechnologyPublications, London, U.K., pages 28 - 31

Nilsson, Ake, (1988), Groundwater Dams forSmall Scale Water Supply, IntermediateTechnology Publications, London, U.K.

Nissen-Pelersen E., (1990), Harvestingrainwater in Semi-Arid Africa, Manual No. 2,Small Earth Dam Built by Animal Traction,ASAL Rainwater Harvesting Nairobi, Kenya

Nissen-Petersen E. and Lee M., (1990a),Harvesting Rainwater in Semi-Arid Africa,Manual No. 1, Water Tanks with Guttering andHandpump, ASAL Rainwater HarvestingNairobi, Kenya

Nissen-Petersen E. and Lee M.,( 1990b),Harvesting, Rainwater in Semi-Arid Africa,Manual No.5, Sub- Surface and Sand-StorageDams, ASAL Rainwater Harvesting, Nairobi,Kenya

O'Brien, Louis, (1990a), Report on RWH Projectin Togo, Raindrop Rainwater HarvestingBulletin, Vol. 3, Water and Sanitation forHealth Project (WASH), Arlington, Virginia,U.S.A. pages 1,5 and 6

O'Brien, Lois, (1990b), Local Networking inKara, Togo, Raindrop Rainwater HarvestingBulletin, Vol. 3, Water and Sanitation forHealth Project (WASH), Arlington, Virginia,U.S.A, page 7

ODA (1989), Performance of Fabric ProtectedSlow Sand Filters Treating Poor Quality RiverWater, Final Report to the ODA Project R4373,Overseas Development Administration (ODA),London, U.K.

Ongweny G.S., (1979), Rainfall and stormwaterharvesting for additional water supplies inAfrica - Report to UNEP, Nairobi, Kenya(quoted by Lee and Visscher (1990))

OXFAM (undated). OXFAM Water SupplyScheme for Refugee Camps - TreatmentPackage, OXFAM, Oxford, U.K.

Pacey A. and Cullis A., (1986), RainwaterHarvesting - The collection of rainfall andrunoff in rural areas. Intermediate TechnologyPublications, London, U.K.

Parry (1991), trade literature of J.M.P. Parry &Associates Ltd., Cradley Heath, U.K.

Philippeaux H. and Sassarao P.R.G., (1992),Rainwater Harvesting in Burundi,WATERfront, Issue 2, Water andEnvironmental Sanitation Section, UNICEF,New York, U.K.

PMI (1992), Track- literature about the UPMpump, manufactured by Pompes MaupuInternational, Orleans, France

Pontin J.M.A. and Woolridge R. (1977), AStudy of Hafir Linings in Dafur Province,Western Sudan, Report OD/10, HydraulicsResearch Station. Wallingford, U.K.

Potapak U992), i literature of, andpersonal communication with, Potapak Ltd.,Newbury, Berkshire, U.K.

Prins, Agma and Yacoob, May (1988), AddingGuinea Worm Control Components: Guidelinesfor Water and Sanitation Projects, WASHTechnical Report No. 51, Water and Sanitationfor Health (WASH), Washington, U.S.A.

Rajagopalan S. and Shiffman M.A., (1974),Guide to Simple Sanitary Measures for theControl of Enteric Diseases, World HealthOrganisation, Geneva

Roark P., Kittle B., Yango J., Nignan D., YonliR., Tindano R. and Gnoumou M., FinalEvaluation USAID/Burkina Faso Rural WaterSupply Project, WASH Field Report No. 191,Water and Sanitation for Health (WASH),Washington, D.C., USA

Rous, Tim, (1985), Protecting a shallow seepagespring, Waterlines, Vol. 4, No. 2, IntermediateTechnology Publications, London U.K., pages23 - 25

Silverfine E., Brieger W. and Churchill A.,(1991), Community-based initiatives toEradicate Guinea worm - A manual for PeaceCorps Volunteers, Vector Biology and ControlProject, US AID, U.S.A.

Singh, Vinay Pratap and Chaudhuri, Malay(1992), Performance Evaluation andOperational Modification of the UNICEFUpward Flow Water Filter, a paper presented atthe First Middle East Conference on WaterSupply and Sanitation in Rural Areas, Cairo,Egypt, 23-25 Feb., 1992

Skinner, Brian, (1992), Protecting Springs: Analternative to Spring Boxes, Technical BriefNo.34, Waterlines, Vol. 11, No. 2, Intermediate,Technology Publications. London, U.K., pages15- 19

Skinner B. and Cotton A., CommunityRainwater Catchment, (an as yet unpublishedmanual, completed 1992), International LabourOrganisation, Geneva, Switzerland

Smet J.E.M. and Visscher J.T. (compilers),(1990), Pre- Treatment Methods for CommunityWater Supply, IRC International Water andSanitation Centre, The Hague, The Netherlands

Smet J.E.M. and Wheeler D., (1990), ModularSub-Sand Abstraction Systems, in Smet andVisscher (1990), pages 39-51

Smith, Michael, (1988), Sludging - Hand Jettingof Tubewells, Developing World Water,Grosvenor Press International, Hong Kong,pages 91 and 92

Smith, Michael, (1990), Groundwater Dams,Technical Brief No. 24, Waterlines, Vol.8, No.4, Intermediate Technology Publications,London, U.K., pages 15 -18. (also ITP (1991))

Sridhar M.K.C., Kale O.O. and Adeniyi J.D.,(1985), A simple sand filter to reduce Guineaworm disease in Nigeria, Waterlines, Vol. 4,No. 2, Intermediate Technology Publications,London, U.K., pages 16 - 19

Sutherland J.P., Folkard G.K. and Grant W.D.,Natural coagulants for appropriate watertreatment: a novel approach, Waterlines, Vol. 8,

No. 4, Intermediate Technology Publications,London, U.K.

SWS (1992), personal communication with, andtrade literature from SWS Filtration Ltd,Morpeth, Northumberland, U.K.

UNICEF (undated), Cement jars for the storageof grain and water, UNICEF/Ministry ofHousing and Social Services Village TechnologyUnit, Nairobi, Kenya

UNICEF (1986), Ground catchment ferrocementwater tank, Waterlines, Vol. 5, No. 2,Intermediate Technology Publications, London,U.K., pages 28 - 30

UNICEF (1992), Strategies & Resources -Operational Strategy Framework forDracunculiasis Eradication, Version 1.1,

UNICEF, New York, U.S.A.Visscher J.T., Paramasivam R., Raman A. andHeijnen H.A., (1987), Slow Sand Filtration forCommunity Water Supply - Planning, design,construction, operation and maintenance,Technical Paper No. 24, International ReferenceCentre for Community Water Supply andsanitation, The Hague, The Netherlands

V & W Engineering (1992), Costs and tradeliterature from manufacturers V. & W.Engineering & Installations (Pvt.) Ltd., Harare,Zimbabwe; information obtained from Water Aid(1992)

Wagner E.G. and Lanoix J.N. (1959), WaterSupply for Rural Areas and Small Communities,WHO Monograph Series, No 42, World HealthOrganisation, Geneva

Water Aid (1992), personal discussions withWaterAid staff, London, U.K.

Watt, S.B. and Wood, W.E., (1976), Hand DugWells and their Construction, IntermediateTechnology Publications, London, U.K.

Watt, SB., (1978). Ferrocement water tanks andtheir construction. Intermediate TechnologyPublications, London, U.K.

WES, (1992), Low-cost, high impact approachesin Sudan, Water and Environmental Sanitation

I

1I1IIIIIIIIIIIIIII


(WES) Section, WATERfront, Issue 1. Development Centre (WEDC), LougbborougbUNICEF, New York, U.S.A., pages 18 and 19 University of Technology, U.K., pages 19 - 22Wheeler D., Skilton H., Pardon M. and LloydB., (1986), The enhancement of slow rate sandfilters by the incorporation of synthetic fabrics,ODA Research Project R3760 at the Universityof Surrey U.K., Del Agua, Guildford, U.K.

WHO, DPHE & UNICEF (1991), Report onStudy of Pond Sand Filters, WHO, DPHE andUNICEF, Dhaka. BangladeshWinter, Stephen J., (1987), A ferrocement wellfor Micronesia, Waterlines Vol. 6, No. 2,Intermediate Technology Publishing, London,U.K., pages 20 - 22

Wirojanagud W. and Hovichitr V. et al., (1989),Evaluation of Rainwater Quality: Heavy metalsand Pathogens, IDRC, Ottawa, Canada, (quotedby Gould (1991))

Wirojanagud P., (1991), (Associate professor atWater Resources and Environment Institute,Khon Kaen University, Thailand), personalcommunication

Wiseman, Robin, (1983), Kano's Jet Set, WorldWater, issue of October 1983, Liverpool, U.K.World Water (1979), Well-sinking made easy inBangladesh's ideal terrain, World Water, Issueof December 1979, Liverpool, U.K.,

WRC (1975), The effects of storage on waterquality, Papers and proceedings at a WaterResearch Symposium, 24-26 March, WaterResearch Centre (WRC), Buckinghamshire,U.K.

Wright E.P. and Herbert R., (1985), CollectorWells in Basement Aquifers, Waterlines, Vol. 4,No 2, Intermediate Technology Publications,London, U.K., pages 8-11

Wurzel, Peter, (1992), Concept versus Reality:The Atridev Handpump, WATERfront, Issue 2,Water and Environmental Sanitation Section.UNICEF, New York, U.S.A., pages 8 and 9

Yaziz, Mohammad Ismail and Din, Omar,(1988), Portable slow sand filter, paper inproceedings of 14th WEDC Conference, Waterand Urban Services in Asia and the Pacific,Kuala Lumpur; Water, Engineering and

IIIIIIIIIIIIIIIIII

APPENDIX I

Terms of Reference

II unicefIIIIIIIIIIIIIIIIl

United Nauow Children's Fund 3 United Nuiont P t i uFondi to Nations Uoiet pour I'cn&iu* N c w v&rk. New York 10017rondo de lax jsacionu U m u t pan li Ifi&ncw 212 326-1000

Tele** 175989

TERMS OF REFERENCE

RESEARCH INTO COST EFFECTIVE TECHNOLOGIES FORWA TER SUPPLY IN GUINEA WORM ENDEMIC AREAS

1. BACKGROUND

The eradication of Dracunculiasis by the year 2000 is the target set by theUnited Nations and other National and International agencies. The goal to eradicateguinea worm was initially endorsed in April 1981 by the Steering Committee of theInternational Drinking Water Supply and Sanitation Decade. In 1989, the UNICEFExecutive Board (E/ICEF/1989/CRP.2 dated 22 March 1989) supported theelimination of guinea worm as 'A disease that can be eliminated by providing safedrinking water and heaith education using already existing cost-effective measures'.The eradication of guinea worm is also one of the goals of the World Summit ofChildren declared in September 1991.

Dracunculiasis is the only disease that can be eradicated completely by theprovision of improved water supplies provided that these are used at all times. Asoutlined in the forty first World Health Assembly (A41/INF.DOC/2 dated 13 April1988) in their review of progress in the elimination of guinea worm:

'Priority should be given to improving drinking water sources in villages wheredracunculiasis is endemic because {a) those villages suffer from dracunculiasisin addition to all other problems befalling villages without safe drinking water;(b) they constitute only a small fraction of all unserved villages in endemiccountries; and (c) the obvious success achieved in such villages can help toincrease and sustain support for the water supply programme as a whole inaffected countries'

However efforts need to be made to investigate alternative low costtechnologies that can be used for isolated communities and areas with difficulthydrogeological conditions in order to reduce per capita costs which should notexceed US$30 per capita (low cost rural water supply systems at present averageUS$20 per capita in Africa). These technologies have to be easily replicated on alarge scale and, if eradication is to be achieved this decade, appropriate low costwater supply services have to be provided as quickly as possible.

Obviously in many cases a handpump-equipped borehole may be the mostsuitable technical option. However the cost escalates rapidly in isolated communitieswith populations under 200 persons and where the hydrogeological conditions areunsuitable i.e. ground water level beyond recommended 50 metres depth and/or

I2 I

aquifer is not easily accessible. UNICEF and other agencies have been activelyinvolved in investigating alternative technologies that can be used in Africa. These •have included rainwater harvesting, solar powered systems and promotion of |household water filters. However there is a need for suitable guidelines to bedeveloped outlining the introduction of suitable low cost technologies in guinea worm •endemic countries on a large scale. A project proposal needs to be formulated mrecommending suitable improvements that could be undertaken to make thesetechnologies more technically, socially and economically effective. •

I2. OVERALL PURPOSE I

To investigate and propose low cost effective methods to provide clean drinkingwater to guinea worm endemic areas. The results of the findings will be used to •develop suitable projects in a selected number of guinea worm endemic countries. I

5. SPECIFIC OBJECTIVES |

3.1 BACKGROUND PAPER

Collect relevant information from suitable sources including a literature review, •interviews with suitable persons/agencies/private companies regardingappropriate technologies, including those listed below. The consultant will Idocument which technologies have been proven to work and those that need ™to be further tested. Some technologies are listed below. However theconsultant is recommended to include others as he considers advisable. I

3.1,1 Rainwater Harvesting

UNICEF recently funded a study 'Water Harvesting in Five African •Countries'1 which investigated appropriate rainwater harvestingtechnologies in countries including Mali and Togo. Presently most Isystems have been based on rooftop catchment systems. Howeversurface catchment systems can be very suitable for provision of _sufficient drinking water (4-5 litres/person/day)when combined with •other sources for domestic purposes. According to the UNICEF report,rainwater harvesting has potential for wider expansion when the rainfallis between 200-1000 mm per annum. The study also recommends thatfor drinking water supplies, surface catchment systems should be widelyapplied in the sub-humid and semi-arid areas and rooftop catchmentsystems in the humid and sub-humid areas. These findings need to befurther investigated with respect to cost effectiveness of household

Netherlands.

I1 Lee, M.D. and Visscher J. T. (1990). Water Harvesting in Five African

Countries. IRC International Water and Sanitation Centre. The Hague, The I

I1

1IIIIIIIIIIIIIIIII

versus community systems and compared with previous and more recentfindings.

3.1.2 Solar Powered Systems

Many West African countries have been involved in installing solarpowered water supply systems. However these have mainly been forlarger communities at a per capita cost between US$40-80. Many largecompanies have been involved in their manufacture. GRUNDFOS havebeen involved in installing systems in Mali, the Gambia and Senegalusing French manufactured photovoltaic cells.

UNICEF and other agencies have much information of smallentrepreneurs who are manufacturing solar pumps at lower cost. Theseneed to be reviewed and_suitable meetings conducted to Investigate theirapplicability for guinea worm enaemic areas.

3.1.3 Slow Sand Filtration

This can be a simple, efficient and reliable technique for the treatmentof water, Its cost usually lies within the resources of communities whocan be actively involved in the construction, operation and maintenance.Various countries have tested these systems and the InternationalReference Centre (IRC), the Hague, have supported a number of researchprojects. However major problems have arisen in the operation andmaintenance of these systems which require regular maintenance inorder to periodically remove the top of the filter bed. If the water is tooturbid the system is not suitable.

It is recommended that the consultant should liaise with IRC to follow upthe findings of the IRC 1982 report2 and document successful andunsuccessful case studies analysing these for future technical research,if found necessary.

3.1.4 Infiltration Galleries

With suitable bed conditions, water can be extracted from rivers bydigging tunnels parallel or constructing wells alongside the river. Tosome extent water is filtered by passing through the bed material. Littleinformation is presently available regrading successful projects. A reviewshould be made of all relevant literature and discussions held withagencies that have been involved in supporting this technology.

Slow Sand Filtration of Community Water Supply in Developing Countries,IRC International Reference Centre for Community Water Supply andSanitation, September 1?32.

II

3.1,5 Clav Pot Filters

In many guinea worm endemic countries clay pot filters are presentlyused. However it would be useful to review the designs presently used „with respect to capacity, shape and ability to keep clean. Most are used mwith mesh filters usually provided by government.

The use of community versus household water filters needs to be further * •investigated.

r

I3.1.6 Chemical Treatment

This can be an expensive and difficult method to maintain at the _community level. However various private companies are now |manufacturing more cost effective and innovative methods to disinfectwater. It may be the most cost effective method for areas where -literacy rates are higher and all other methods of water provision are too |expensive.

3.2 DEVELOPMENT OF GUIDELINES |

3.2.1 Liaise with leading agencies involved in guinea worm eradication •including Global 2000, Vector Biology Control, USAID, Peace |Corps, CDC and others, in order to collect relevant information,exchange ideas and ensure agreement on major research findings •prior to the official meeting described in Section 3.4. |

3.2.2 Prepare a suitable background conference paper including the •results of the technical review. I

3.2.3 Work with UfMICEF and collaborative agencies to convene a •meeting at a suitable venue to be attended by all relevant agencies Iin order to share the findings of the background paper andincorporate the conclusions into the final report. •

3.3 DEVELOPMENT OF PROJECT PROPOSAL

Develop a project proposal for follow up action to field test the different •technical options on a large scale in selected countries including detailedmethodology and cost estimates. I

III1

IIIIIIIIIIIIIIIiIII

4. METHODOLOGY

Interviews will be conducted with UNICEF, Global 2000, CDCand other international agencies and private companies eitherpersonally or by telephone. A literature review will beundertaken of documentation pertaining to appropriatetechnical options including those listed in section 3.1.

The consultant will liaise frequently with the agenciesquoted in section 3.2 in order to share information. Hewill visit the International Reference Centre, The Hague,in order to collect relevant information. He will assistin making the necessary arrangements for the meeting tobe convened. Travel and DSA for this trip will beprovided by UNICEF. Consulting fee is inclusive withinthe contract. This should be conducted during the sixthweek of the consultancy in order to allow time toassimilate the conclusions into the final report.

S. FINAL PRODUCTS/OUTPUTS

5.1 A final report of approximately 25-30 pages withillustrations will be produced by the Consultant at thecompletion of the consultancy. This will be submittedwith the Project Proposal outlined in Section 3.3 to theUNICEF New York office for comments and approval.

5.2 Participate in the Technical support Team meeting forDracunculiasis, 24-28 August in Annecy, France to reviewand discuss development of guidelines with TST members,UNICEF field officers and agency partners.

Travel and DSA for this trip will be provided by UNICEF.consulting fee is inclusive within the contract.

5.3 An advanced draft should be submitted in October to serveas a background document for an interagency meeting onthis topic. Consultant should participate in thisconsultative meeting in New York with UNICEF andcollaborating agencies, October 1992.

Travel and DSA will be provided by UNICEF. Consultingfee is inclusive within the contract.

5.4 A final draft should be submitted early November 1992 toincorporate comments and suggestions from the interagencyconsultation.

6. DURATION

The consultancy will be for a duration of seven weeks.\

V. TobinWET/92/295/TEC.PRG

i research into cost effective technologies for … · research into cost effective technologies...

Documents