I echnoIog~’ - IRC · 2014-03-07 · (,‘I’Ct ~-1~ ¼C 1~,3 ©International Development...
Transcript of I echnoIog~’ - IRC · 2014-03-07 · (,‘I’Ct ~-1~ ¼C 1~,3 ©International Development...
I
232.2—720
(IJ iS~.((’z.~L
I!
I
echnoIog~’~iesearchand Evaluation in AsiI
(,‘I’Ct
~-1~ ¼C1~,3
©International DevelopmentResearchCentre1982PostalAddress Box 8500, Ottawa, CanadaK1G 3H9HeadOffice 60 QueenStreet,Ottawa, Canada
Sharp,DGraham,M.
IDRC-204eVillage handpumptechnology researchand evaluationin Asia
Ottawa, Ont, IDRC, 1982 72 p. ill
!Pumps/, Ihand tools!, /appropriate technology!, /rurali, !watersupply!, /developing countries! — !project evaluation!, ftesting/,!technical aspects!, !market/, !economic aspects!, /case studies!,statisticaldata.
UDC 621.651(1-22) ISBN. 0-88936-360-9
Microfiche edition available
11 existe également une edition francarse de cette publication
The International DevelopmentResearchCentre is a publiccorporation createdby the Parliamentof Canadain 1970 tosupportresearchdesignedto adaptscienceandtechnologyto theneedsof developingcountries The Centre’s activity is concen-tratedin five sectors;agriculture, food and nutrition sciences,health sciences, information sciences, social sciences,andcommunicationsIDRC is financedsolely by theParliamentofCanada,its policies, however,aresetby aninternationalBoardof GovernorsTheCentre’sheadquartersarein Ottawa,CanadaRegional offices arelocatedin Africa, Asia, Latin America, andthe Middle East
La edicijin espanola de esta publicaciän también se encuentra disponible
720 232.2&2v1
Village Handpump Technology
Researchand Evaluation in Asia
- 2~
IDRC-204e
Editors: Donald Sharpand Michael Graham
Résumé
Depuis six ans le CRDI appuiefinancièrementdes recherchessur la mise au point depompesplus efficacespour l’approvisionnementen eau potabledes regions rurales Lesavantagesde nouveauxmatériauxetmodêlesdepompeont etéétudiés,plus particulièrementl’emplol de matièresplastiques L’Université de Waterlooa collaboréa Ia productiond’unassemblagede cylindre et clapet de piedsimple qui constitueraitle premierélémentd’unepompea main pour puits de surfacesusceptibled’être fabriqueedanslespaysendéveloppe-ment avecles ressourcesdisponiblessurplace Soumisea desessaisen laboratoire,Ia pompea ensuiteété testéedansdiversesconditionsenvironnementalesdansquatrepaysasiatiquesetdeuxpaysafricainspourdeterminersoncoüt defabrication,safiabilité et sadurabilité, safacilité d’entretienpar les villageois et son efficacité technique Cette publication passeenrevue les résultatsde rechercheprésentésa l’atelier tenua l’Université de Malaya, KualaLumpur (Malaisie) du 16au19 ao~t1982,auterme desprojetsréalisésenAsie Ellecontientégalementuneevaluationtechniqueetéconomiqueglobaledesquatreprojetsetuneévalua-tion des recherchesa faire et des priorités a leur donner Les futurs travauxporterontprobablementsur Ia possibilitéde lancerune productiona grandeéchelledepompesa mainet sur les difficultés que présenteraitIa réalisationd’une telle entreprise
Resumen
En los ültimos seis años el CllD ha apoyado investigacionestendientesa desarrollarsistemas más efectivos de bombeode agua para el area rural Se han estudiadolasimplicacionesde los nuevosmaterialesy diseñosde bombas,enespecialel usodematerialesplásticos En colaboración con Ia Universidad de Waterloo, se desarrollóun conjunto
económicodepiston y valvula-pedalcomobaseparsunabombamanualdepozospandosquepudieraserfabricadaen lospalsesendesarrolloconrecursoslocales DespuésdeserensayadaenIaboratorio,Ia bombafue sometidaa pruebabajodiferentescondicionesambientalesencuatropalsesde Asia y dos de Africa con el objeto de determinarcostosde fabricación,confiabilidad y durabilidad,capacidadde mantenimientoa nivel rural y desempeñotécnicoEstelibro ofreceuna reseñade los resultadosde lasinvestigacionespresentadosduranteunseminario realizadoen Ia Universidadde Malaya,Kuala Lumpur,Malasia,del 16 al 19 deagosto de 1982 a Ia culminación de los proyectos asiáticos Se incluyen ademáslasevaluaciones técnicas y económicas generalesde los cuatro proyectos, asI como unaestimaciónde las futuras necesidadesy pnondadesde Ia investigación,entrelas cualessecontarinprobablementeel potencialdeproduccióna granescalay los problemasinvolucradosen Ia implantacióndel sistema
2
Contents
Preface 5
Acknowledgments 6
Introduction 7
Sri Lanka PathiranaDharmadasa,Upali Wickramasinghe,andDouglasChandrasiri 11
Thailand Pichai Nimityongskul and Pisidhi Karasudhi 21
Philippines Antonio Bravo 33
Malaysia Goh Sing Yau 39
Overview of Technical Performance Gob Sing Yau 53
Economic Analysis andPotential Markets Tan Bock Thiam 57
Conclusions 67
Participants 71
3
Not all the health assistants in the world can get rid ofdysenteryand cholera if water supplies are contaminated
Barbara Ward 1976 The Home of ManW W Norton & Company Inc , New York, NY,USA Page 229
4
Preface
Many factorsareinvolved in efforts to provide safedrinking water for all duringthis the International Water Supply and Sanitation Decade One of the keys,however, is the developmentand use of a reliable handpumpthat can be locallyproduced,installed, and maintained at a reasonablecost
The International Development ResearchCentre (IDRC) has invested aboutCA$730 000 in a network of water-supplyprojectsin Asia andAfrica over the last6 yearsto help developmoreeffective pump systemsfor rural water supplies Thispublication reviews the resultsof the Asian segmentof thenetwork and identifiesfuture researchpriorities, specifically the need to investigate large-scalemanu-facturing of the polyvinyl chloride (PVC) pump that hasbeen developedand theessentialsocial and public-health factors that must be part of any implementationprogram
It should be pointed out that the technology developedand tested by theseIDRC-supported research projects is applicable to rural situations all over theworld, not just to thosefew countriesin Asia wherefield testingwas carried outThe developmentof a handpumputilizing inexpensivePVC components,whichcan be manufacturedlocally and simply enough to be maintained at the villagelevel, is a giant step forward in the struggle to provide adequate,clean, watersupplies to rural populations
The technology has beentried, tested,and proven But the question remainshow can the desire to utilize it and maintain it be best transferredto thosewhoneed it most? It is our hope that this volume will stimulateefforts to implementthis technologyandfoster new researchinitiatives in all countrieswhereprovisionof potable water is still a major problem
The paperspresentedin this publication are summariesof the full reports ofeach country project More specific details may be obtained by writing to theHealth SciencesDivision of IDRC to obtain microfichecopiesof thecompletereports
ElizabethCharlebois,DirectorHealth SciencesDivisionInternational Development Research Centre
5
Acknowledgments
Over the past6 years,manyresearchers,engineers,technicians,consultants,support staff, andothershavecontributedto thedevelopmentof IDRC’s conceptof the village level operatedandmaintained(VLOM) handpump.The list is toonumerousto acknowledgeeachpersonby name It goeswithout saying,however,that it is thededicatedeffortsof thesepeoplethatmadethis publicationpossible
Thanks are also due to the University of Malaya, where the end-of-projectseminar-workshopwas held,andto Dr Goh Sing Yau, local coordinator,andhiscolleagues,Dr TeeTiamTing, Dr Tan BockThiam, Mr ChongKah Lin, andMr TeoBeng Hoe, for their hardwork in ensuringthesuccessof themeetingMr LeeKamWing actedas IDRC coordinatoranda specialword of thanksis dueto Ai Ling Gob,Health SciencesDivision, IDRC, Singapore
Also credit should be given to Tim Journey,who carriedout the earlydesignwork for handpumpsutilizing plastic componentsunderthe sponsorshipof theWorld Bank and was later hired by IDRC to continuethe effort.
It mustbe pointed out that, although thepump describedin this publication isoftenreferredto as the IDRC—Waterloodesign, it is really nothing more than anupdatedversion of a wooden pump used in Europe about six centuriesago.Elementsof thedesignareclearlyillustratedin a 16th centuryplateappearingin abook on mining translatedby Herbert Clark Hooverand Lou Henry Hoover in1950
It is interestingthatscientistscontinually reinvent thewheelor, in this case,thepump.
6
Introduction
The precise links betweenimproved water supply and health benefits aredifficult to document However, all peopleappreciatethesignificanceof a clean,adequatewatersupply Nevertheless,an increasedsupply of safewatermust beaccompaniedby certain behavioural changesthat affect personalhygieneandsanitation practicesbefore enteric diseasescan be significantly reduced Thesechangesarecomplexandarenot likely to occurspontaneouslyThetargetpopula-tion must be suppliedwith readily understoodinformation aboutthebenefitsofchange andconvincedto adoptnew behaviouralpatternsandacceptnew tech-nologies Furthermore,consumeracceptanceof waterandsanitationtechnologydependson devicesthat canhold up to abuse,function for long.periods,andcanbe purchasedand maintainedby the villagers themselves
The selection,development,anduseof reliablehandpumpsthat can be locallyproducedandinstalledandmaintainedat a reasonablepriceis amajorsteptowardproviding reliable, safe drinking-water supplies to rural communities Due tomany technical and economicfactors, such as the complexity of engine-drivenpumps andthehigh cost of fuel, manualpumpswill continueto be usedin mostpartsof theworld, not only for potablewaterbut alsofor domesticuse,livestock,and irrigation
For thepastdecade,seniorofficials of nationalwaterauthoritiesin developingcountries,along with personnelfrom internationaland bilateral agencies,haveobservedthatoneof themostimportantproblemsin ruralwater-supplyprogramsis the high failure rate of conventionalmanual pumps Failuresoccur mainlybecausepumps werenot designedfor the level of stressandabuseencounteredfrom large usergroups within rural communities Furthermore, the materialsfrom which theyaremade,mainly castiron andsteel,arenot only expensive,butalsonot readily available locally. Consequently,manydevelopingcountrieshavebeenrelying on importedpumpsandpartssuppliedby internationalandbilateraldonors This has implications in terms of costs,maintenancerequirements,andproblemsof procurementof spareparts
For the past6 years,the InternationalDevelopmentResearchCentre(IDRC)has been supporting researchin the developmentof more effective pumpingsystemsfor rural watersupplies Theapproachtakenhasbeento examinesystem-atically the implications of new materialsand improvedpump designs.In view ofthe wide-spreadintroduction of plastics technology that has taken place indevelopingcountriesin the last decade,particular attentionwas focusedon thepolymerresins,specifically polyvinyl chloride (PVC) piping, which is widely avail-able throughout Africa and Asia In many respects,plastics technologyis todevelopingcountrieswhatcastiron wasto industrializedcountriesyearsagoandthe vast potentialof plasticshasyet to be tapped
The IDRC-sponsoreddesign work centred on developing a simple, low-costpiston andfoot-valveassemblyfor a manual,shallow-wellpump.This stageof theresearch,in collaboration with the University of Waterloo, was completedinMay 1977 The piston and foot-valve assemblydevelopedat the University ofWaterloowas testedat theConsumer’sAssociationTesting Facility in EnglandThis testingprogramwasinitiated by theOverseasDevelopmentMinistry in theUnited Kingdom to analyze the characteristicsof 10 commercially produced
7
manual pumps that were manufacturedin industrializedcountries The projectestablishedthe reliability andefficiency of theWaterloodesigncomparedwith theexisting technology The Waterloopump differs from othersin that it hasbeendesignedspecifically for fabrication in developing countries, utilizing existinglocally available resources
In 1978, after the laboratory testing, researchprojectswere set up in twocountriesin Africa andfour in Asia to field testthepump undervariousenviron-mentalconditionsandlevelsof technicalsophisticationwith differentusergroupsThe countriesinvolvedin this phasewereMalaysia, thePhilippines,Sri Lanka,andThailand in Asia, and Ethiopia and Malawi in Africa
The Waterloo hanu’pump has brought clean water to rural families in Malaysia
8
Theprimary objectivesof thesestudieswereto assesstheWaterloopumpdesignin variousfield conditionsfor characteristicssuchascapacityfor local manufacture,cost of manufacture,reliability anddurability, maintenancecapabilityat thevillagelevel, andtechnicalperformanceThe basicpiston andfoot-valvedesignproducedby theUniversity of Waterloowasusedby all theprojectswith somelocal modifi-cations The above-groundcomponentswere locally designedand produced ineachcountry
In thePhilippines,the Institute for Small-ScaleIndustriesat theUniversity ofthePhilippinescarriedout the researchin collaborationwith theNationalInstituteof ScienceandTechnology,theDepartmentof Local GovernmentandCommunityDevelopment,Departmentof Health, and the Local Utilities and Water WorksAgency In Thailand, the Asian Institute of Technologyconductedthe researchin cooperationwith theDepartmentof Health, theDepartmentof Public Works,the Office of AcceleratedDevelopment,and the National Economicand SocialDevelopmentBoard In Malaysia, theFacultyof Engineeringat theUniversity ofMalayaconductedtheresearchin collaborationwith theEnvironmentalEngineeringDivision of the Ministry of Health In Sri Lanka, the Lanka Jathika SarvodayaShramadanaSangamaya(the SarvodayaMovement), which is involved in grass-roots community-developmentwork, carried out the research
The researchincludedan economicanalysisof costeffectivenesscomparedwithother handpumpsbeingusedin the region It also involvedassessingthepotentialfor rural water-supplydevelopment,making projectionson thepercentageof ruralhouseholdsthat could be servedby pipedwater, andattemptingto determinethefuture marketdemandfor handpumpsin the region
In August 1980, themid-projectmeetingfor thefour Asian projectswasheld atthe University of Malaya in Kuala Lumpur to review the projects’ progressandestablishcommonmonitoring andmeasurementtechniquesA uniquemethodforaccuratelydeterminingpump usagewith a mechanicalcountingdevice,developedat theUniversity of Malaya, wasalso incorporatedinto the field-testingprogramThis devicemade it possibleto correlatemeasurementsof wearwith thedistancethe piston traveledor theamount the pump was used
The activities of the four projectsin Asia havenow beencompletedand theresultsare encouraging Two workshopswere thereforesponsoredby IDRC incollaborationwith the Faculty of Engineeringof the University of Malaya from16 to 19 August 1982 in Kuala Lumpur
For the first 2 days,theproject leadersfrom the four Asian countriesreviewedanddiscussedtheir resultsandassessedtheoverall technicalandeconomicimplica-tions of their findings. During thelast 2 days,a disseminationseminarwasheld topresent the results to interestedgovernmentaland nongovernmentalagenciesfrom the regionandto observersfrom variousinternationalagenciesandprivateconcerns The statusof handpumptechnologyin theregionwasreviewedandnewresearchpriorities were identified
The PVC pump demonstratedduring the field trials that it holdsconsiderablepotentialfor useat thevillage level It can be madelocally at reasonablecostandiseasilyrepairedwith locally fabricatedparts However, it must be realized that,aswith any technology, there are limitations If one is looking for a “magic,”maintenance-freepump, then this technology is not theanswer The resultsoffield trials indicatethat, although the pump is durable,therearelimitations thatmust be understoodand respectedor malfunctionswill occur Also, failure willoccur if the well is improperlydevelopedMore importantly, the outcomeof thisresearchhasclearly demonstratedthat inexpensiveplastics can be usedin hand-pump manufacture,making it possibleto producepumps andspareparts locallyand to incorporatedesignsthataresimpleto understandandeasyto maintainat anaffordablecost
This volume deal primarily with handpump technology, but it must berememberedthat thepump is morethanjust a convenientmeansof drawingwater
9
from thewell It is an essentialelementin public-healthefforts becausethe onlysafewayto provide adequatesanitaryprotectionfrom surfacecontaminationis tosealthewell andinstall a pump Unless this andother public-healthmeasuresaretaken to protect thewell, water-relateddiseaseswill continue to take their toll
In thecoming years,limited resourceswill haveseriousconsequencesupon theprovisionof safe,adequate,watersuppliesfor ruralpopulations If this problemisto beaddressed,governmentsandwaterauthoritiesmust focus their resourcesondeveloping low-cost technologies that are easily understood, operated, andmaintainedat thevillage level By publishing this volume, we hopethat theresultsof this researchwill stimulatethe implementationof suchappropriatetechnologyand at the sametime foster new researchinitiatives
10
Sri LankaPathiranaDharmadasa,
Upali Wickramasinghe,andDouglasChandrasiri
The majorityof theruralpeoplein Sri Lankaobtain waterfor daily usefrom rivers,canals,lakes, irrigation tanks, and uncoveredwellsThe waterfrom suchsourcesis often unsuit-able for drinking and most other domesticpurposesBecausefew peopleboil thewaterbefore drinking it, this results in manydiseases a fact that village people do notunderstand
The Sarvodaya Movement is playing amajor role in setting up health-educationprograms and in providing facilities forimproving the healthof the rural massesinSri Lanka Onecomponentof this programisthecovered-wellsprogram(Fig 1) The mainemphasisof this programis the introductionof low-cost handpumpsmade from locally
available materialsas a meansof providingclean drinking water for household useDuring this project, threenew designsweredevelopedfor the above-groundcomponentsof the Waterloo pump developed withfunding from theInternationalDevelopmentResearchCentre (IDRC) As well, severalmodificationswere made to the piston andcheckvalve, which wasusedin placeof a footvalve, to make thepump easierto manufac-ture with local resources The goal of thisdesign work was to develop a pump thatincorporatedthe following features the useof low-cost materialsavailable in Sri Lanka,easy maintenance and repair without theneedfor highly skilled labour, andtheuseofpolyvinyl chloride (PVC) plastic to eliminatecorrosion problems
Organizationof theProject
A preliminary survey was undertakeninJanuary1979 in severalvillagesin thedistrictsof Galle,Matara,andHambantotato investi-gate: the economic situation in the villages;the existing social conditions, the irrigationfacilities, and the attitudesof the villagerstowardhandpumpsBasedon thefindings ofthis survey, it was initially decidedto install60 pumps in six villages, but later, due to
Fig. 1. View of well showing drainage channel to removespilled water and stonelayer to asoist drainage
11
geographicalandpolitical reasons,thenumberwas reducedto four pumps in eachof fivevillages Akurala Village, Talawa Village,Hingurudugoda Village, and GinimellagahaVillage in the Galle District and YatiyanaVillage in the Matara District. In addition,one pump was installed at the SarvodayaCentre office for demonstrationpurposes
The projectwas divided into threephasesconstructionof the wells, installation of thepumps, and inspection and field testingDuring thestudy,the importanceof coveringthewells andthe healthproblemscausedbyusing water from uncovered wells wasemphasizedto thevillagers All constructionwork wascarriedout by theSarvodayaRuralTechnicalServiceThe pumpswereassembledand installed by the EngineeringSectionofthe SarvodayaMovement, and the surveywork was handedover to a team selectedforthe purpose
The preliminary surveying was completedby July 1979 andtheconstructionwork com-pleted by February 1980 During February-August 1980, the pumps were installed and
the monitoring work was started Pumppistons,pistonrings, fulcrum shafts,journals,check valves, check-valve bolts, and pumpheadswere producedat the SarvodayaMainCentre Parts that were easierto make wereproducedat the Village Centres
Locationand Constructionof Wells
Careful consideration was given to theplacementof wells The wellswerelocatedatleast 30 m from the nearestlatrine or othersourceof contaminationandin well drainedareasdevoid of surface water even duringheavy rains (1 m = 3 28 ft) The wells wereconstructedby digging a pit andpositioning aprecastconcrete ring in the hole A secondring was then addedanddigging continueduntil the water table was reached Theseconcreterings, therefore,formed the wallsofthe well This technique was so successfulthat the SarvodayaRural Technical Servicecommittee decided to construct all the wellsin the sameway Moulds for the rings were
Table 1 Summaryof material used, fabricationequipmentrequired,cosusedin different pump designs
t, andquality of thecomponents
Pumpelement Material Tools and equipment Cost (Rs)b
anddesign used’ required Material Labour Quality
FrameLi Angle iron Welder, hacksaw 145 90 SatisfactoryL2’ Concrete Mason’s tools, mould 80 60 PoorL3 Angle iron, Drill, hacksaw,welder, 210 150 Very good
GI sheetmetal sheet-metaltoolsVi CI pipe, Hacksaw, welder, drill 30 50 Good
MS plate
HandleLi Cl pipe, MS, Drill, hacksaw,lathe, 80 100 Good
brassbushings welderL2 Wood Carpenter’stools 120 80 PoorL3 CI pipe Hacksaw, welder, 100
blacksmith’stools100 Very good
Vi Wood, bolts, Carpenter’stools,hacksaw, 12 60 Satisfactorywashers dnll, files
Piston and check valve1st Wood, leather Lathe, drill, leather cutter 10 50 Not durable, low
volumetricefficiency2nd PVC Lathe,drill, solventcement, 175
blowtorch90 Leaked,broke easily
3rd PVC, wood Lathe, drill, solventcement 100 50 Good
‘Abbreviations Cl, galvanizediron, MS, mild steel, PVC, polyvinyl chloride= US$1
‘Becauseof problemswith this design,it waseliminatedfrom the field testing Wells originally having thesepumpswerefitted with L3 pumpsinstead
12
1
—Iinch
I • I I
cm
A
2
5
4
6
7
6
8
9
10
8
B
Fig. 2. Details of (A) piston and (B) check valve used in LI and L3 pumps (1) “1” bolt, (2) brass nut, (3) PVC cap,(4) brass spring, (5) brass bushing,(6) PVC washer, (7) leather plale valve, (8) leathercup, (9) wood, and(10) PVC pipe
madeout of 1/8-inch (3-mm) mild-steelplatesand angle iron These moulds were sent tothe villages where the concrete cylinderswere then made In each village, theRuralTechnicalSectiontrained six personsto makethe rings with the aid of the village masons
In all cases,the well wassealed,a drainage
channelwasprovidedto removespilled water,and a stonelayer was laid to assistdrainage(see Fig 1) Accessto the well wasprovidedby a manhole in the cover A week afterconstruction was completed,the walls werewashedand waterwasdrawn from thewellusing a gasoline-enginepump
2
5
2
13
Description ofPumps
Thethreepumpsthat weredesignedduringthis projectsharedessentiallythesamebelow-groundcomponentsbut differedsignificantlyin their above-groundconfiguration Themodifications made to both the below-ground andabove-groundpartsaredescribedin detail below and summarizedin Table 1
Below-groundcomponents
A PVC check valve basedon the originalWaterloo design was used in the Vi-typepump, in the other pumps, however, thepiston andcheckvalve werechangedto thoseshown in Fig 2 In themodified pistonsandcheckvalves,a pieceof woodwith eight holesdrilled in it is inserted in a section of PVCpipe 5 cm long with an outside diameterof4 5 cm (1 cm = 0 39 in ) In this new design,the piston seal is obtainedby using locallymade leathercups Thesewere usedinsteadof rings madeof polyethylene,a material thatis not manufacturedin Sri Lanka,is difficultto obtain,andis expensiveLeather,however,is available all over the island The PVCwashers were made locally by flatteningheated PVC pipe and then forming thewasherson a lathe The platevalve is madefrom flattened PVC pipeon which a pieceofleather is glued to createa sealand preventleaking This change was required becausesevereleakagewas observedwhenPVC and
brass were used together The use of aleather valve completelysolvedthis problem
A piece of galvanizedsteel wire, used tomakethe spring,washeld in placeby a brassnut and washer A threadedbrass“I” bolt12 5 cm long holds the parts of the pistontogether A 3 5-cm long brass bushing isplacedoverthe“I” bolt to facilitate movementof theplate valve This bushingalsoensurestheproperspacingof thecomponentsof thepistonandcheckvalveandcorrecttensionofthe spring when the partsare assembled
The connecting-rodguides in the pumpcylinder aremade of PVC pipe cut into sec-tions and solvent-weldedto the PVC rod
Threadedcouplings are used to join thesectionsof thepistonrod together(seeFig 3)Initially, the rods werejoinedusingabolt, butthebolts broke becauseof stressat the jointThe use of PVC couplings has been verysuccessful
Above-groundcomponents
The L3 pump
For field-testing,a total of six of thesepumpswere installed Four at Talawa and oneeachat Cinimallagahaand at theSarvodayaMainCentre Thesepumps were used to replacetheL2 pumpsthat hadbeeninstalledoriginallyin thesewells The frameof the L2 was madeout of concreteand useda woodenhandleTheoriginalL3 pumpsalso useda handlethatpivoted on a bearing fixed in a concretepedestal Although this arrangementprovedmore successfulthanthe L2 pump, it alsohadto be changedbecauseof excessive,andrapid,wear of the wood at the pivot point Thismeantthat thehandlecouldbeloweredto thepoint where it would makecontactwith theoutlet spout, resulting in a constantbangingby the handle during the operation of thepump that resulted in the joint betweenthepump head and the spout cracking Thisproblem was remedied by designing a newhandle that preventedoverlowering and atthe same time made the pump easier tooperate(Fig. 4)
Frame Angle iron is used to make thesquareframefor this pump Four 9-mm boltsarewelded to theuppersection of the frameto hold the pump head and four 12 5-mmholes are drilled in the lower section forattachmentto the well (1 mm = 0 039 in)Two bracketsmadeof flat iron are used to
P
:1 1
3-4-
inch 201
cm 5
A B
Fig. 3. (A) ThreadedPVC piston rod couplingsand (B) PVCpiston rod guidesglued to piston rod to reducevibration duringoperation of pump (1) connectingrod, (2) threadedcoupling,(3) riser pipe, and (4) rod guides
14
11
-4 3
5
67
8—
-9
—10
hold the riserpipe Oneis weldedto thelowersectionof the frame, theother is fixed to thestationarybracketwith 6-mm bolts (see Fig5)
Bearings Wood (satinor palu) is usedforboth setsof bearings Onesetis fitted to thepiston rod Theothersetis usedfor thepivotpoint in the handleand is designedso thatthe bearing can be reversed once wear isexcessive,thus prolonging the life of thebearing(Fig 5)
Spout The spout is made of 25-mmgalvanizedpipe. The pipe is filled with sandto preventcrimping andbentusingheaton aspecially designed jig made by the projectstaff The other end is threadedto fix it tothe pump head.
Connectingrod A 19-mm PVC pipeis usedfor the piston rod The sectionsof the rodare joined using threaded PVC couplingsand the rod is connectedto the piston usinga threadedbrasscoupling
Frame cover A galvanized sheet-metal(22 gauge)coveris attachedto theframewithnuts andbolts The uppersection,which can
be removed,is taperedto protectthepumpingmechanismfrom rain, which is important forprolonging the life of the woodenbearingsThe removablecoveralsoallows easyaccessfor inspection and maintenance
Handle Originally, the handle wasmounted on one side of the pump frameHowever, this design causedexcessivewearon the woodenbearing and thus the designwas changed The handle is now made of12 5-mm galvanizedpipe and two piecesofflat iron (seeFig 5) Thishandlewasdesignedto increasethe life of the woodenbearing,makethepumpeasierto use,andat thesametime limit the length of the stroke Whenthehandleis loweredtoo far, theuser’shandbangson the iron spout This simple designfeaturenot only protectsthe internalpartsofthe pump from damagebut also allows thepump to be locked with a chain andpadlockif desired
The LI pump
Six of the Li pumps were installed inBaddegama,three at Ginimallagaha,andthree at Hmgurudugoda The above-ground
3
0
10
20-
30
40
50 -
cm inch
4
5
67
9
10
9
A B C
10
Fig. 4. (A)LI pump, (B) L3 pump, and (C) VI pump (1) handle, (2) frame, (3) spout, (4) piston rod, (5) riser pipe, (6) pistonrod guide, (7) piston rod joint, (8) riser pipe joint, (9) piston,(10)checkvalve, (II) woodenbearingandcounterbox(LI and VI),(12) main bearing (L3), (13) big-end bearing (L3), and (14) counter base(L3)
15
componentsof thesepumps(seeTable 1 andFig 4) worked well and did not requiremodification Originally, a 7 5-cm cylinderwas used for the riser pipe anda 5-cm pipewas usedfor theabove-groundportion of thepipe However, this meantthat the 7 5-cmpiston could not be removedfor repairswith-out cutting the cylinder To make repairseasierand to save time and labour, a 5-cmpiston is now used along with a 5-cm pipefor the entire length of the riser pipe andcylinder
The VI pump
Nine of theVi pumps (Fig 4) have beeninstalled four at Akurala; four at Yatiyana,andoneat HingurudugodaThe wells at thesesites are 3-4 m deep The above-groundcomponentsof this pumpweremodifiedbasedon our experience,the results of the fieldtesting,andthesuggestionsof theusers Wefound that the original metal handlemadethis pump difficult to operate,thereforeweswitched to a woodenhandle We also dis-coveredthat thepistoncouldbe pulledout of
the cylinder during use so we installed asimple metalbracket to preventover-raising(see Fig 6) The cost of this pump is low,aboutUS$75 for thecompletepump to drawwater from 3 m However, becausethere isno lever mechanismin the design, the usermust lift the full weight of thewater in thecolumn, thus making extensiveuse tiring
Frame The pumpframeis madeof 5 i-cmgalvanizedpipe A galvanizedplate is usedtofix the frameto the well and a 2 5-cm socketis welded onto the frame for attachmentofthe spout
Spout The spout, madeof 2 5-cm pipe,is screwedto thepump frame Theother endof the spout is bent, usingheat, to an angleof 90°
Handle Hard woodsuchas satin andjakis usedto make thehandle, which is held inplace by a bolt The pump is very easytooperate,but usagefor a long period can betiring In a few instances,thewoodenhandlebrokewhen thepump wasin operation,how-ever, it is interestingto note that thepeople
inch10 20
Ol , ii ii Iii?10 30 50 B
~~top~
C (side)
A
wooden bearingsFig. 5. Details of L3 pump (A) frameand woodenbearings, (B) big-endbearing,and(C)final handledesignshowingpositionof
16
Fig. 6. (A) Stopand counter box attached to VI pumpo, (B)detailo of counter inotallation, and (C) counter aosembly fromVI pump
in the village can easilymake a new handlethemselves
Main pipe PVC pipe, 5 cmin diameter,isused for the riser pipe Although theoutsideof thepipeis smooth,the inside is roughandnot completelyroundbecauseof local manu-facturing problems Becauseof theseproblems,it was impossibleto obtainanadequatesealinthepiston andcheckvalve usingPVC rings.Instead,leather cups that could be made inthe villages were designed to solve thisproblem These cups were made using asimple pressanda locally fabricateddie Theleather was treatedwith tallow and held inthe pressfor about30 minutesto acquiretherequiredshape These cups havebeenusedfor more than 1 year and continue to givegood results Problemswerealsoencounteredwhen thelocally availablecouplingswereusedto join the 5-cm riser pipe. Thesecouplingsproduced poor quality joints and, becauseoftheir configuration,left a ridge in thejoinedpipes This ridge createda problem becausethe leather piston cups stuck at the jointsmaking it impossible to remove the pistonexceptby cutting the riserpipe To overcomethis problem,thepipeswerejoinedby makingbell joints The end of the PVC pipe wasdippedin hot coconutoil to softentheplasticand then it was forced over a locally mademetal form to increaseits diameterenoughto fit tightly over the normal end of theadjoining pipe The pipescouldthen bejoinedwith solvent cementThis systemhasworkedwell andallows thepiston to beremovedeasilyfor maintenanceand repair
TechnicalAssessment
During the course of the field testing, anumberof designmodificationsusinglocallyavailable materialsandexpertisewere intro-duced to solveproblems Extensivemeasure-ments were also made of the wear of thepiston rings, piston platevalve, check valve,journals,fulcrum shaft,andconnectingrod toassessthe durability of these pumps underfield conditions The technicalperformanceofthepumpswasalsomeasuredwith theaidofa speciallymountedcounter(seeFig 6) Thiscountermeasuredthe length of eachstrokeof thepiston andthus gavea readingon theamount of useof the pump The volumetricefficiency of the pump wascalculatedfromvolumetncefficiency = (actualdischargex 100)1
A
inch 12
B cm 30
C
17
(cylinder area x length of standardstroke)The mechanicalefficiency of the pumpswasalsocalculated
During the initial stagesof the testing,manypumpsmalfunctionedbecausethejointsin thepiston rod broke Thegalvanizedboltsused to join the sectionsbroke after about3 monthsof usedue to vibrationof thepistonrod during operationof thepump In addition,theholesdrilled in thepiston rod at thejointscausedweaknessand, occasionally,breakageof the rod The rods are now joined withthreaded PVC couplings that are readilyavailable locally The couplings are glued totheendsof the rodsandtherods arescrewedtogether This method hasbeen very satis-factory As well, guides made of sectionsofPVC pipewereattachedto thepistonrod (seeFig 3) Theseguides reducedthe amountofvibration of the piston rod andthus reducedsomeof the fatigue problems
In the Li pump, a 7 5-cm diameterpipewasusedas thepumping cylinder to increasetheoutputof thepump,however,this causedproblems because the piston could not beraisedabove thecylinder section as the riserpipe wasof smallerdiameter To removethepiston for repairor maintenance,thecylinderhad to be cut Therefore, in all pumps, wenow use 5-cm pipe throughout the entirelength of the riser pipe and cylinder Oneproblem still remainswith the riser pipe thejoints in thepipeoccasionallyseparateduringpumping becauseof thevibration of thepipeHowever, this problem was only observedwhen the riserpipeis morethan5 m long Forshallowerwells, this hasnot beena problem
Becausethe riser pipe wasopenat theend,the original check valve with polyethylenerings frequently droppedinto thewell againbecauseof vibration during operationof thepump The useof leather cups in the valveseemsto havesolved this problem For addi-tional security,theendof riserpipe is heatedandcrimped to prevent thecheckvalve fromfalling into thewell A screenplacedovertheendof the riserpipe wasalsoused to correctthis problem but crimping the end of thepipe was found to be more practical
Local fabrication of the Waterloo pistonand checkvalve causeconsiderabledifficultybecausesolid PVC stockis not availablein SriLanka Severalattemptsweremadeto impro-vise with locally availablematerials Initially,we tried making a “solid” rod or cylinder bygluing progressivelysmallerPVC pipesinside
one another Problems were encounteredwith this designbecause,whengrooveswerecut for thepiston rings, theendsof the rodtendedto breakoff It was alsovery difficultto drill holesalongthelength of this“built-up”pipe Next, we tried to fabricate the pistonand check valve from wood and still usepolyethylene rings as a seal Although theconstruction of the valves was easier, theywerenot successfulbecausethepiston ringsstuckin thegroovesanddid not sealproperlyagainstthewall of the riserpipe Also, a poorsealresultedbecauseof the roughinsidesur-faceof the PVC pipesavailablein Sri LankaThis rough surface also quickly wore thepolyethylene rings, resulting in burrs onthe edgeof the rings, which contributed totheir sticking in the grooves
The problemsrelatedto thesevalveswereeventually solved by using a design thatcombineda hollow PVC pipe with a woodencore and employed leather cup seals Thisdesign (see Fig 2) completely solved theleakageproblem Not only wastheproblemofleakageremedied but the wear of the riserpipe was also lessened When polyethylenerings wereusedduringfield testing, the riserpipe wore by 0 35 mm after 90 daysof useThis is believed to be due to silt particlesbecomingembeddedin the rings and actinglike sandpaper against the cylinder wallSimilar tests with leather cups producedmuch less wear This design makes localfabrication and repair possible and it hasproven its reliability under field conditionsfor over 1 year
During the field testing of these pumps,two types of checkvalves were tried In theVi pumps,a checkvalve using arubberplatevalve was screwed into the bottom of theriser pipe However,a retrievablecheckvalveof the samedesignas the PVC-woodpistonwasusedfor theLi andL3 pumps A rubberplate valve was usedin theVi pump and itworked very well (Fig 7) But, becauseit isscrewed to the riser pipe, its use is onlypractical in shallowwells where it is easytoremove the entire length of riser pipe toservice the valve The PVC-wood valve(Fig 2) used in the Li andL3 pumpscan beextracted without removing the riser pipeBecause it incorporatesleather cups and aPVC—leather plate valve, the villagers caneasily repair worn parts themselvesRubbersuitable for the plate valve used in the Vipump, on the other hand, is not so easily
is
Table 2 Overview of performancedata, measurementsof wear, and maintenanceand repair required
Pump typeandnumber
Numberoiusers
Volumetricefficiency
(%)
Waterhead(m)
Avg wateroutput(L/day)
Percentagewear Downtime
(days)
MaintenanceandrepairsPistonrings Cylinder Journals Parts(US$) Time(hours)
‘0
VI pumpsBA 01BA 02BA 03BA 04BH 15MY 17MY 18MY 19MY 20
L3 pumpsBT 05BT 06BT 07
BT 08BC 12MC 21
Ll pumpsBC 09BC 10
BC 11BC 13BC 14BH 16
en = 3 28 ceet,
735048564417214334
212762293250
465258423376
1 L = 0 22 gallon
80 34 900 12 75 11 7 1770 32565 1 5 403 22 9 12 6 11 75 3 5091 14 115 15 10 22 4 11.25 20074 1 4 285 12 5 5 21 5 11 10 2 0062 2 9 821 13 8 21 3 3 50 2 0085 24 292 38 25 20 3 500 20069 29 726 11 5 23 - - —
73 1 8 289 13 5 22 4 4 75 1 5055 1 2 49 16 2 8 11 3 1 25 025
77 3 0 513 14 1 7 8 6 14 00 4 5084 2 3 400 15 5 2 7 6 18 75 6 0075 5 4 536 14 6 8 8 8 6 10 00 4 7569 6 0 476 12 5 8 12 6 8 50 5 7586 7 9 340 11 5 7 9 9 7 12 00 6 2580 59 100 - - - 1 - 100
99 35 490 7 92 13 11 1900 65087 5 3 875 5 5 3 1 10 20 00 7 0089 66 832 13 5 2 11 1500 45088 28 930 12 5 7 1 6 6 14 50 3 5092 25 705 6 55 14 8 1585 52593 3 3 972 4 8 10 5 2 7 2 50 3 50
1 23 4 5 6 7 5
00 000
0 0 0
inch
0 0 0
4
o~J
ii I
cm 12
Fig. 7. Details of check valve uoed in VI pumpo (I) PVC cover, (2) rubber nng, (3) brass nut and washer, (4) rubber checkvalve, (5) PVC plate, (6) PVC screen, (7) brass bolt, and (8) brass spring
obtainedand is more expensiveUsing the mechanicalcounting device, it
was possible to determine the amount ofusageof each pump Estimates were thenmadeon the averagewater output in litresper day, thedistancethepiston hadtraveled,andtheamountof weareachcomponenthadundergonein relationship to usage Table 2provides a summary of performance datacollected from March i9Bi to April i982The resultsof this monitoring indicatedthattheLi pump experiencedthe leastamountofwear and maintainedthe highestvolumetricefficiency This was probably due to theleather cup seals Becauseit wasnot possibleto purchasehigh quality PVC piping, i e,mostpipeswereout-of-round andwereroughon the interior surfaces,the initial wear onthe polyethylenerings was substantialuntilabrasion shapedthe rings to the interiorconfigurationsof the cylinder Wearon thejournals (bearings)varied greatly accordingto the materialused For example,thebrassbearings used in the Li pump wore a maxi-mum of 2%, whereaswear on the woodenbearingsusedin the L3 pump was as muchas 23% It wasfound that, althoughwoodenbearingsare more subjectto wear,this weardoesnot hindertheoperationof thepump,incontrast,as little as i mm of wear on brassbearingsmakesthepumpdifficult to operate
It wasinterestingto note that,of the threepumps tested, type Li was subjectedto thelargestusergroups,theaveragebeingSi per-sons Type L3 experiencedthe least amountof usagewith an averageof only 37 personsper group An averageof 42 persons pergroup usedpump Vi Even with the largernumbersof peopleusing the Li pump, the
breakdownrate wasnot any higher than theL3 model The Vi model, or direct actionpump, had the lowest breakdownrate
To conclude,although theoriginal conceptof the Waterloo designwas to utilize poly-ethylenerings to create a sealbetweenthepiston and cylinder and the foot valve andcylinder, this was not successfuldue to thepoor quality of the PVC piping andthe factthat polyethylene is not easy to obtain
Because leather is easily accessible andinexpensive,this was the logical alternativeUsing a metaldie, it wasfairly easyto shapethe leather into the desired form Thistechniqueis simpleenoughto be masteredbya village worker Although the Li angle-ironpumpsweredurable,wearon thebrassbear-ings made their operationdifficult It was,therefore,more practical andless expensiveto use the wooden bearings,as in the L3model,even though they requiredmore fre-quentmaintenance
We feel that the pump that has beendevelopedas a resultof this IDRC-supportedproject is durable, can be producedat areasonablecost,andcan bemaintainedat thevillage level with very little specialtraining
Acknowledgmento TheWaterloopump researchprogram was carried out by the Lanka JathikaSarvodayaShramadanaMovement(Inc) Sri Lankaand sponsoredby IDRC We wish to expressoursincere thanksto the SarvodayaCoordinatorsintheGalle District, Mr DannyDissanayaka,andtheMatara District, Mr P Hewavitharana,Mr KarlWherle, Mr ThomasZimmerman,and Mr Guna-palaGanegamaof the RuralTechnicalSectionUnit,Sarvodaya,Moratuwa,andto all theworkersin theSarvodayaGramodayaCentresandthepeoplewhohelpedus to make this task a success
ofl
20
Thailand
Pichai NimityongskulandPisidhi
Karasudhi
The NationalEconomicandSocialDevelop-ment Board of Thailand reports that morethan 80% of the people in Thailand live inrural villages and that only 40% of thesepeople have accessto a safe water supply.Although people in the urban areas havea relatively good quality water supply, thesupply in rural communities is far fromadequateIn rural areas,potable waterandwaterfor otherdomesticpurposesis obtainedfrom various sources rainwatercatchment,deepor shallowwells, reservoirs,ponds,andstreams Of thesesources,water from deepor shallow wells is the safest in terms ofprotection from waterbornediseases
Approximately90% of theexistingwells inThailand usehandpumpsandover 5 millionpeopledependon thesehandpumpsto obtaintheir water for consumption and otherdomestic uses.As a result, handpumpsarean integratedpart of the life of the ruralpeopleandtheoperationandmaintenanceofthese pumps posesa challengingtask It isestimatedthat the cost of repair and main-tenanceof the 7000 handpumpsinstalled bythe Departmentof Mineral Resourcesaloneis over US$500000 annually Furthermore,NIDA (1978) reported that, based on arandom sample,roughly 5000 of the 19000handpumpsinstalledin Thailandby differentgovernmentagencieswere out of operationon any given day
Purposeand Scopeof theProject
Themain objectiveof this studywasto test,undervariousfield conditions,thehandpumpdevelopedat theUniversity of Waterlooandsubsequentlyto modify and optimize thehandpump design to suit local conditions.Specifically, the aim of this study was to.
(1) conduct a review of the handpumpscurrently usedby the five main governmentagenciesresponsible for rural water supplyin Thailand; (2) carry out laboratory tests onvarious handpump types, including theWaterloo pump, to compare their perfor-manceandenduranceundervarious conditions;(3) install and conduct field tests of theWaterloo handpumpconfiguration, and(4) basedon the field and laboratory testresults, adapt and improve the Waterloodesignand field test themodified handpumpunder village conditions
This project was sponsoredby the Inter-national Development ResearchCentre(IDRC) andcarriedout with thecooperationof the following Thai governmentagenciesDepartment of Mineral Resources (DMR),Ministry of Industry, Departmentof Health,Ministry of Public Health, Department ofPublic Works, Ministry of Interior, Office ofAccelerated Rural Development (ARD),Ministry of Interior, and AgriculturalTechnology Office, Ministry of Agricultureand Cooperatives In addition, the NationalEconomic and Social DevelopmentBoard ofthe Office of the Prime Minister servedasthe coordinator for theseagencies
Reviewof ExistingHandpumpsinThailand
Historically, the existing handpumpsinThailand originated from Europe andNorthAmerica and were designedfor use by asingle family in the developedcountries Indeveloping countries, the pump was sharedby manypeopleliving in theruralcommunityand, becauseof the increasedusage,it brokequite often and, in most cases,couldnot berepairedby the villagers. According to theARD Office (i980), severaldifferent typesofhandpumpshave beeninstalled in Thailandby different governmentagenciesHowever,the different handpumpsin Thailand canbe broadly classified into two groups theDMR handpumpsandthe ARD handpumps
The prototypesof the handpumpsfromthe DMR are the Demster,RedJacket,andother handpumpsdonated by, or procuredfrom, the United States of America Thistype of handpumphas athree-pinleverwith acrossheadand a cylinder that generallyhasa plunger with two leather cup sealsand apoppetvalve at theplunger The lower valve
2i
consistsof a spring-activatedpoppetand thepumphasa 3-inch (7 5-cm)cylinder, 7/16-inch(11-mm) pumping rod, and 1 25-inch (3-cm)drop pipe (riser pipe) The inlet-valve lining,piston cup, top gasket,andcylinder gasketofthesepumpsaremadeof leather,whereasthespout gasket is madeof rubber There are42 different componentsand most of themare madeof brass Cold-drawn steelis usedfor the cylinder reducing coupling, bottomfulcrum pin, and piston rod
The handpumpsprovided by the PublicWork Departmentare also included in thisgroupbecausetheyaresimilar to DMR hand-pumps except that the drop pipe is slightlylarger (1 5-inch, 3 8-cm diameter)
The ARD handpump, which is usuallycalledtheKorat handpump,hasbeenadoptedby theDepartmentof Health, theARDOffice,and the Local Administration DepartmentThe pump mechanismconsistsof a rack andpinion (gear type) Leather is used for thepiston seal, gasket, andbushing seal All ofthepackingnut caps,valves,locks,andhandlegear bushingare made of brass The pumphas a 3-inch (7 5-cm) cylinder and 1 25-inch(3-cm) drop pipe and is composedof 32components
Thereis a slight differencein thehandpumpsuppliedby ARD It is essentiallythesameasthe Korat handpumpexcept that it has a3-inch (7 5-cm) cylinder made of polyvinylchloride (PVC) and 1.25-inch (3-cm) droppipe The pumps supplied by the LocalAdministration Department are normally
used for shallow wellsThe advantagesand disadvantagesof the
DMR andARD handpumpsaresummarizedin Table 1.
In the Waterloo handpump,basically, allthe componentsare made of rigid PVC andpolyethylene,which arelow cost commoditypolymers Theseplastics havebeenselectedfor easeof manufactureandlow costaswellas for their efficiency The below-groundcomponentsconsistof a PVC well casingthatalso functions as the pump cylinder for theplastic piston The plastic foot valve isconstructedof componentsthat arecommonto thepiston(Fig 1) The flow of waterin thepiston and foot valve is regulatedby simplerubberplatevalves Polyethylenepiston ringsprovide adequatehydraulicsealson thepistonwith much less frictional resistance thanleather cupsor rings It hasbeenfound thatwear is concentratedon the rings, whichcanbe replaced easily, rather than on the wellcasing The foot valve can be removedeasilyfor inspectionor repairandservesasareservepiston whenever necessary.Both 2-inch(5-cm) and 3-inch (7 5-cm) diameterpistonswere used in this study
The pistons and foot valves for theWaterloo pumps were made at the AsianInstitute of Technology (AlT) using locallyavailable PVC rods The piston rings andfoot-valve adapterswere made from poly-ethylene supplied by IDRC Basedon theresultsof the laboratoryandfield tests, thepiston andfoot valve were later modified
Table 1 Advantagesanddisadvantagesof existing handpumpsin Thailand
DMR
AdvantagesThe air chamberof the above-groundcomponent
helpsmaintaina continuousflow of waterThe handle is madeof a single pieceof cast iron
having appropriateshapeand size, henceit iseasyto operateand is durable
DisadvantagesDirect contact between the axle and the bushing
causesrapid wearIf the axle breaks,thehandlecanmove horizontally
andmay damageother partsThecontactareabetweenthe axle andstuffing box
is relatively loose This can result in eccentricmovementof the pumpingrodandreducedpumpefficiency
Theplungercouldslam againsttheupperandlowerparts of the cylinder if not properly installed
The cast-iron fulcrum link is not strong enoughand, underrepeateduse, tends to break
ARD
AdvantagesVertical movementof thepumping rod is by using
a rack and pinionThereare fewermoving partssubjectto wearand
tearThe shock-absorbingspring prevents the plunger
from slamming against the upper and lowerpartsof the cylinder
DisadvantagesThe flow of water is not continuousdue to the
absenceof an air chamberThehandleandpinion areseparatepartsandfailure
often occursat this Joint
22
Fig. 1. (A) Piston and (B) foot-valveassemblyof Waterloo handpump (1) polyethylenering, (2) piston, (3) Pvcplate valve,(4) brass valve guide, (5) flat washer, (6) nut, (7) bolt, (8) bolt with eye, and (9) polyethylenefoot-valveadapter
LaboratoryTesting
Laboratory tests were carried out in theStructural EngineeringLaboratory,AlT, forall three handpumps The DMR and ARDhandpumpsweresuppliedby variousgovern-ment agenciesand theWaterloohandpumpswere made at AlT The parametersthatwerevaried were strokelengthof thepiston,orifice/pistonarearatio,andrateof pumping
A steelplatform, 4 m high, waserectedanda mechanizedrocker arm was installed onthis platform to drive 12 setsof handpumps
Table2 Testing program for Waterloo handpump
Piston stroke
Orifice/piston Speedof piston length (inch)’
arearatio (%) (strokes/mm) 3 4 5 6 8 10
125 20 XXXXXX170 30 XXXXXX222 40
5060
XXXXXXXXXXXXXXXXXX
Note For all threeorifice/pistonarearatios, testsweremadefor each piston speedand stroke combination EachX representsa single test that was carriedout twice
‘1 inch = 254 cm
23
A B
simultaneously A head simulation systemwas also fabricated to test the performanceof thehandpumpsunderdifferentwaterheads
A detailedtesting programfor theWaterloohandpumpwas undertaken(Table 2) In thistable, the orifice/piston area ratios of 12 5,17 0, and22 2% representan openingin thepiston area of eight holes each having adiameter of 3/8 inch (9.5 mm), 7/16 inch(ii mm), and1/2 inch (12.5 mm), respectivelyFor each series of tests, the dischargefor10 strokesof thepistonwas measuredusing abucket and a graduatedcylinder The testwas performedtwice and the averagevaluewasrecorded The volumetricefficiency wasdefined as actual divided by theoreticaldischarge times 100 where the theoretical
discharge is equal to the stroke lengthmultiplied by the cross-section of thecylinder
In addition to the testson theperformanceof handpumps,severalother tests were alsoconducted
The mechanical properties — tensile,compressive,and fatigue strengths— of thePVC material were determined A 300-kNuniversal testing machine was used for thetensionandcompressiontests For the fatiguetest,a servo-pulsatorhaving a capacityof 15 t
was usedLeakage was testedto check the perfor-
mance of the foot valve, which was themodified foot valve that was installed inthe field
100 —
90-
C-)z
UU-U-
U80 -
F-
70 —
0
SD
io•
STROKE RATE (strokes/minute)
Fig. 2. Influenceof stroke lengthson volumetricefficiencyof Waterloopumpshaving 12 5% orifice/pistonarea ratio (Waterhead, 15 feet(4 57 m), piston diameter, 3 inches(7 5 cm), valvegap, 0 25 inch (0 6 cm), piston foot-valveclearance,12 inches(30 cm))
Stroke length (inch)
304•
6A
20 30 40 50 60
24
The tensilestrengthof thePVC pipe jointusedin final designof thePVC systemwasdetermined for 3/4-inch (2-cm) and 3-inch(7 5-cm) diameterpipes The joint wasmadeusing a PVC coupling
Resultsoflaboratory testing
Performanceof handpumps
The influence of piston stroke length onthevolumetricefficiency of Waterloopumpshaving orifice/pistonarearatiosof 12 5, 17 0,
and 22 2% was plotted against the speedofthe piston for all pumps An exampleof theresult is given in Fig 2 The piston diameter,the gap of the foot valve, the piston footvalve, andthewaterheadclearancewerekeptconstantthroughout these tests The volu-metricefficiencyof thepumpincreasedas thepiston stroke length as well as the speedofthepistonincreasedTheinfluenceof orifice/piston arearatioson thevolumetricefficiencyof Waterloopumps having piston strokesof4, 6, and 8 inches (10, 15, and20 cm) werealso determined The results for a 6-inch(15-cm)strokearegiven in Fig 3 Theresultsindicatedthat, for a piston strokeof 4 inches(10 cm), theorifice/pistonarearatio hadlittleeffect on the volumetric efficiency of thepump However, for longer piston strokes,the volumetric efficiency of the Waterloopump increasedas the orifice/piston area
ratio increasedThe performanceof DMR andARD hand-
pumps under laboratory conditions wasalso plotted At low speedsof piston move-ment, the volumetric efficiency of ARDhandpumpswas lower than that of DMRhandpumps At higher speedsof pistonmovement, however, the volumetric effi-ciency of ARD handpumpsimproved signi-ficantly At very high piston speeds, thevolumetricefficiencyof theDMR handpumpsexceeded100% This can be explainedby thefact that,when thepump is operatedat veryhigh speed.extrawater from the riser pipeispumpedup
The volumetricefficienciesof the existinghandpumpsand the Waterloo handpumpshaving piston strokesof 4, 6, and10 inches(10, 15, and 25 cm) were comparedFigure 4depictsthecomparisonfor a strokeof 6 inches(15 cm). At low pistonspeeds,thevolumetricefficiency of the Waterloo handpumpwasbetweenthat of the DMR and ARD hand-pumps At higher piston speeds,however,the volumetric efficiency of the Waterloohandpumpwas thelowest of the three For apiston stroke of 10 inches (25 cm), the
I I I performance of all three handpumps was20 30 40 50 60 similar
STROKERATE (strokes/minute)
Fig. 3. Influence of orifice/piston area ratio on volumetricefficiency of Waterloo pumpshaving 6-inch (15-cm) strokelength (Waler head, 15 feet (4 57 m), piston diameter, 3inches (7 5 cm), valvegap, 0 25 inch (0 6 cm))
>,UzLuUU-U-LuU
HLu
‘-I0
100
90
80
70
ARD PumpWaterloo Pump
DMR Pump
I I
20 30 40 50 60
STROKE RATE (strokes/minute)
Fig. 4. Comparison of volumetric efficienciesfor differentpumpshaving 6-inch (15-cm) stroke length
>,UzLuULU
U
I—
100
90
80
70
Orifice/Piston
Area Ratio (%)
125170
222
Other tests
The tensile and compressivestrengthsofPVC material were determinedto be 5370and 10130 lb/mnch~(37025 and 69844
25
kPascal), respectively In the fatigue test,the mechanical properties of PVC wereconsiderablyinfluencedby the temperaturerise that was inducedby repeatedloadingsDuring field operation, however, the PVCpiston is not subjectedto continuousrepeatedloading as simulated by the servo-pulsator
The leakage test was carried out bycompletely filling the riser pipe with waterandallowing it to leakthroughthe foot valvefor 1 day Test results indicated that therewas no water leakageby the modified footvalve for a waterheadof 5 00 m (16 4 feet)Both rubberandplasticsealswere testedandthe same resultswere obtained
The variation of strengthwith curing timefor joints in 3-inch (7 5-cm) and 3/4-inch(2-cm) PVC pipeswasalso testedThe tensilestrength of 3-inch (7 5-cm) joints reachedover 120 kg/cm2 after curing for 15 minutes,which can be regardedas the initial settingtime of the solvent cement
During laboratory tests, it is impossible tosimulate the actual field conditions to whichthe handpumpswill be exposed,especiallyusing the head simulation system Otherrelevant factors such as the quality of thewater in the wells, the rate of pumping orspeed of the piston, the direction of theforceexertedon thepumprod, andthecondi-tion of the above-groundcomponentsmustalso be taken into account Therefore, theWaterloo piston and foot valves weremonitored under field conditions
Field Testing
The wells that were selectedfor fieldtesting of the Waterloo pump were useddaily by the villagers in the communityNormally, the selectedwell was equippedwith either a DMR or ARD handpumpandonly thebelow-groundcomponents(cylinder,piston, and foot valve) were replacedby theWaterloo components If the replacedcom-ponentsdid not perform as well as the oldones,or thepump did not function normallydue to the replacement,the villagershad tofind a new sourceof waterand complainedIt was, therefore,important for theworkingteam to ensure that the componentsthatwere installedin thewell functionedproperlyHence, the following modifications weremade to thePVC components,basedon thelaboratory tests, before the pumps wereinstalled in the field
Modifications
PVC cylinder
Thecylindersuggestedby theUniversityofWaterloo was ordinary PVC pipe that couldbe obtainedin the local market When thePVC cylinder wasinstalledfor testingin thelaboratory, failure occurred at the jointbetweenthe steelcap andthePVC cylinderOne solutionwasto increasethethicknessofthe PVC cylinder, but it was found that thenew cylinder was too costly. The solutionthat was adoptedincorporateda steelcasingwith a PVC lining Figure 5 showsthedetailsof the modified 3-inch (7 5-cm) diametercylinder togetherwith its steelcap Thesteelcasingstrengthenstheupperandlower jointsand thePVC lining is requiredto reducethewear of the polyethylenepiston rings
PVC piston
Thedetailsof theoriginalpistonareshownin Fig 1 However, problemswere encoun-teredwith breakageof the top and bottomribs of thepiston,which aretheweakestpartsof the piston Failure wasdue mainly to theribs hitting thewall of thecylinderrepeatedlywhen thepump rod movedeccentrically Toprevent this, the thicknessof the top andbottom ribs wasdoubledfrom 1/4 inch (6mm)to 1/2 inch (12 mm) The detailsof the modi-fied piston are shown in Fig 6
Foot valve
The foot-valve assemblyof the Waterloohandpumpis shownin Fig 1 This foot valvewas originally attachedto the cylinder wallwith a polyethyleneadapterthatsealedtightlyagainst the cylinder wall However, becausethe inner surfaceof the PVC cylinder wasnot perfectlyroundandsmooth,waterleakageoccurredat the foot valve Another problemwasthat sandparticlesoften becametrappedbetweenthe PVC platevalve and theupperface of the piston, resulting in additionalwaterleakage The first attempt to solvethisproblem was to replace the polyethyleneadapter with drilled stainless-steelplates,placeacompressionspring on top of theplatevalve, andglue a thin pieceof rubberto thePVC plate
One problemthat remainedafter the firstmodification was that sandparticlestrappedinside the grooved portion of the pistoncausedwear To solve this, the idea of using
26
-
Ii~1~
s_
A
0 H
/
~- L-‘I
B
inch(,,I
Fig. 5. Modified 3-inch (7 5-cm) diametercylinder (A) andsteel cap (B) (1) steelcap, (2) rubber 0-ring, (3) steelcasing,(4) Pvc lining, and (5) threadedsection (Total height ofcylinder is 27 inches(68 6 cm))
thefoot valveas a sparepistonwaseliminatedIn the secondmodification, the contactareabetweenthe PVC plate valve and thestain-less steel plate was kept to a minimum bymeansof two small ridges This foot valveperformed satisfactorily in the field Toreducetheproductioncost,however,a thirdmodification was made by eliminating thecompressionspring and changing the PVCplatevalve to a 1/4-inch (6-mm) solid rubberplate The moving distance of the rubber
Fig. 6. (A) Top and (B) side views of modified piston ofWaterlooha ndpump, (1) 7/16-inch (111-cm)holes,and (2)~vc piston ring
platevalve was keptat 1/4 inch (6 mm) Thethird modificationof the foot valve is shownin Fig 7 In total, 54 wells in three regionsused this type of foot valve
Site selectionand installationof PVCpumps
The sites of the field testswere in central,northeastern,and northern Thailand The
jis-
5,; /
~// /
~L4~
1
2
3
4
)
5
A
2
cm
4
10
inch
—, I I I I
cm 8
3
27
cm
Fig. 7. Third modification of PVC foot valve (1) threadedbolt, (2) stainless-steel plate, (3) PVC lining and steelcasing, (4) 7/16-inch (i 11-cm) holes on 1 5-inch (3 8-cm)pitch diameter, (s) brass guide, (6) rubber plate valve, (7)steel cap, and (8) 0-ring
three regions that were selected werelocated in the following areas (1) Saraburiand Lopburi provinces, central Thailand,(2) Khon Kaen province, northeasternThailand, (3) Lamphun and Chiang Maiprovince, northern Thailand
In each region, 18 wells having differentcharacteristics were equipped with themodified PVC cylinders, pistons, and footvalves One important factor that wasconsideredin well selectionwasto chooseawell with a high frequencyof usage,becausemany wells are rarely used in someareasAnother factor that was consideredwastheexistenceof an accessroadto thewell If thewell was too far and difficult to reach,themonitoring would havebecomeverydifficult,and sometimesimpossible,during the rainyseason
Monitoring
BeforethePVC componentswereinstalled,initial measurementswere made on threeaspects— above- and below-ground com-ponentsand the static water table
The configuration of the above-groundcomponentof the handpumpwas notedandall the relevantdimensionsof thecomponentswere recorded
For the below-ground components, thethicknessof theupperandlowerpolyethylenerings on the piston was measuredusing amicrometer For eachring, themeasurementswere taken at three different positions
spaced 1200 apart Thediameterof thepiston2 was notedandtheinner diameterof thePVC3 cylinder was measured with an internal
vernier caliper The cylinder and piston5 assemblywere inspectedat roughly 3-month6 intervals.
The staticwater level is thedistancefromthe surfaceof water (watertable) in thewellto the outlet spout of thepump This staticwater level wasmeasuredusinganelectronicsound probe A wire lead from this deviceislowered slowly into the well through a holedrilled in the pump casing When the wiretouches thesurfaceof thewater, a soundisproducedelectronicallyThestaticwaterlevelequalsthelength of thewire plus thedistancefrom thehole to thespoutof thepump Thestatic water level was observedmonthly
After installation of the PVC components,the performanceof the pump wastestedbymeasuringthe actual volume of water dis-chargedper 10 strokesof piston movementThesedatawerealso collectedon a monthlybasis
In addition to thesetests, samplesof thewaterfrom thewells in the threeregionsweretestedin thelaboratoryfor waterquality Theanalyseswere conducted by the Environ-mental EngineeringDivision using standardmethods for 10 parameters pH, turbidity,colour, hardness,calcium, chloride, nitrate,manganese,iron, and fecal coliform Thewaterquality index was calculatedbasedontwo differentapproachestheDelphi Approachusing option 1 and the WHO AlternativeApproach The test resultsindicatedthat, forsomewells, thewaterquality indexwaspoorandthe water in theselocationswas recom-mendedfor generaldomesticusesonly andnot for consumptionNormally, thevillagerspossessingthese wells obtain their drinkingwaterfrom open pondsthatcollectrainwaterduring the rainy season
Resultsoffield testing
The fluctuationsof staticwater levels forwells locatedin SaraburiandLopburi, KhonKaen and Chiang Mai, and Lamphunwithrespect to calendar time were plotted (see,for example,Fig 8) The static water levelwasratherhigh duringOctoberto Decemberand low during February to March, whichcorrespond to the rainy and dry seasonsrespectively
The volumetric efficiencies, volume of
7
8
inch 401 ‘I
10
28
waterper 10-cmstroke,changesin thicknessof upperandlower piston rings,andchangesin the inner diameter of cylindersof all 54wells locatedin thethreeregionswereplottedagainstcalendartime (see,for example,Fig 9)
Records of thedamagedpartsof both theabove- and below-ground componentsofPVC pumpsfrom installationuntil theendofthe field test were carefully tabulatedTable 3 presentsa summaryof thefrequencyof thedifferent causesof pump failure
A1T~PVCHandpump
Design
Basedon thefield monitoringandlaboratorytestresults,a final design,calledtheAlT—PVChandpump, was developed It incorporatedthe following features
A new type of fulcrum link, hingedat bothends by meansof ball bearings,was intro-
13
11
9
7
5
I_I I I I I l~l I I I I ~l I I~l
8
4
2
CM 2’
F-I
-I
u.iI-’
ii)
0
CM 4
CM 10
CM 8
3
CM 7
MA M ,l J A S 0 N D I F M A MMONTH
FIg. 8. Observed static water levels of wells located in Chiang Mai (CM3—9) and Lamphun (CMI, 2, and 10)
29
duced in this design This fulcrum linkreplacesthe type usedin theDMR handpump,which failed frequentlyeitherat thejoints orat the link itself Details of the above-ground componentsare shown in Fig 10
200 [_Volumetric efficiency
0
~ 2 — Volume of water per10-cm stroke
~ lii -___ _______i ‘,—Ip.__._• • .__.___•___•.._1,__•_._._.__sS -S--S
~ oIl I I I I I I I I I I I I
7 0 - Thicknessof uppera and lower • rings
65~
6F-
90r Cylinder diameter
~ 80u 1980 1981~ 70~ I I I I II I I I I
MAMJ JASONDJ FMAMj j
MONTH
Fig. 9. Performanceof 3-inch (7 5-cm) cylinder handpumpin Saraburi and Lopburi regions
Table 3 Statistical recordsMarch 1980-July
of pump failures,1981
Types of failure Number of failures
Above-groundcomponentsDMR pumpsBreakageof fulcrum link 13Breakageof handle 9Loosening of stuffing box 3Breakage of fulcrum pin 2Wear on flat bar 2Breakage of flat-bar bushing 2
ARD pumpsDamage of sector gear 6Loosening of handle 3
Breakage of spring 3Breakageof spout 1
Below-groundcomponentsLeakage of foot valve 24Failure at piston-rod joints 20
Failure at riser pipe 2
The riser pipe is made of standard3-inch(7 5-cm) diameter PVC pipe 4 m (13 1 feet)long This pipe is also used as the cylinderThe modified piston and foot valve (Figs 6and7) discussedearlier wereadoptedin thisnew design The only part not made of PVCis the 7/16-inch(11-cm)pump rod Initially, astandard hollow 3/4-inch (2-cm) diameterPVC pipe was used for thepump rod but thiswas not strong or durableenough to with-stand the pumping force Even the steelpump rod was found to be insufficient and,as a result, a spacer is introduced at the
6
H 11
Fig. 10. Above-ground componentsof AlT—PVC hand-pump (1) round steel bar, (2) bar guide, (3) handle, (4)
frame, (5) handle-supportjoint with ball bearings, (6) PVCfaucet, (7) steelcasing, (8) coupling, (9) PVC pipe, (10) steelrod, and (11) woodenspacer
30
10
11
12
Fig. 11. Below-ground componentsof AlT—PVC hand-pump (1) steel rod, (2) coupling, (3) brass guide, (4) PVCpiston valve, (5) PVC piston, (6) PVC piston ring, (7) PVCpipe, (8) stainleso-steelplate, (9) rubber foot-valve plate,
(10) PVCfoot valve,(11) solventcementglue, and (12) steelwell casing
joint of thepump rodto minimize vibrationsDetails of the below-groundcomponentsareshown in Fig 11
On a trial basis, three AlT—PVC hand-
pumpswere installedin wells locatedin three
Table 4 Specificationsof thandpumps that were
he three AlT—PVCfield tested
Nol No2 No3
Installation date 22/8/80 15/11/80 19/11/80Water level (m) 11 25 9 15 12 00Volumetric efficiency(%)
76 78 73E 98 91 86
Avg piston-ringthickness (mm)
UpperI 659 671 943E 628 655 721%change 4 40 2 38 2.96
LowerI 653 649 742E 631 640 723%change 3 37 1 39 2 56
Avg inner diam ofcylinder (mm)
I 8001 7995 7986E 8051 8030 8027% change 0 64 0 43 0 42
Table 5 Cost comparisonof above- and below-ground componentsof threetypesof handpumps
used in Thailand (December1980)
Type of Costhandpump (BahtY Remarks
~23 Baht = US$1‘Excluding riserpipes ForAlT—PVC handpump,the cost
of riser pipe is approximately110 Bahtlm (33 Baht/foot)
different regionsand their performancewasmonitored The identification and initialdimensionsof thebelow-groundcomponentsand the performance of these all-PVChandpumpsare summarizedin Table 4
Costanalysis
The cost of an AlT-PVC handpumpwasestimatedand comparedwith the DMR andARD handpumps(Table 5) Two separatecomponents are considered in the costanalysis, the above- and below-ground com-ponents For the below-groundcomponents,
1
2
34
5
6
7
6 ‘10mm = 039 inch, 1 m = 328 feetb1 at installation, and E, at end of field test
Above-ground
DMRARDAlT-PVC
8 Below-ground
5DMR~AlT-PVC
2100 Massproduction2500 Massproduction2300 Single order
700 Massproduction700 Massproduction800 Single order
31
only the costsof the piston andcylinder aregiven becausethe length of the riser pipedependson well depth.
The costs of DMR and ARD handpumpsare basedon mass-production,whereasthecost of the AlT-PVC system is basedon asingleorder Thus the costof the AlT-PVChandpumpcould be reducedfurther if thepump were mass produced The cost of theAlT-PVC handpumpwas3100Baht, whereasthe DMR and ARD handpumpswere 2800and 3200 Baht, respectively
Conclusion
The performanceof the Waterloo hand-pump was studied in detail underboth fieldand laboratory conditions Existing hand-pumps in Thailand were also reviewedandtheir performanceundersimulatedconditionswas assessedAfter careful observationinthe laboratory,thebelow-groundcomponentsof the Waterloo pump were modified andinstalled in 54 wells and monitored for15 months Unlike the plastic handpumpsthat are available commerciallyand intendedfor usein shallowwells (2—5 m), theWaterloohandpumpwasfound to be applicableup to adepthof 20 m For deeperwells, a smallerdiameterpiston should be employed
Basedon thefield monitonngandlaboratorytest results, a modified version, called theAlT-PVC handpump, was designed On atrial basis, three of thesehandpumpswereinstalled in the field and their performance
was observedover 6 months Thesehand-pumps performed satisfactorily and wereappreciatedby the villagers One advantageof this proposedhandpumpis theuseof PVCriser pipe that simultaneously acts as thecylinder of the pumping unit
The field tests indicatedthat theWaterloohandpumpsrequiredsome maintenanceandrepair The weakest part of the pumpingsystemwasthe foot valve,whichoftenleakedin wells that containedsandparticlesin thewater However, the resultscannotbe fullyevaluatedbecause the frequency of pumpusagewas not included in the monitoringprocess
SourceDocuments
ARD (Accelerated Rural Development Office)1980 Improvementof handpumpdesignin Thai-land Bangkok,Thailand Ministry of the Interior,ARD ReportpublishedunderthesponsorshipofUnited Nations Children’s Fund, EquipmentControl Division
Kingham, J A 1979 Progressreport on testingofIDRC prototypepumps.London,U K Consumers’Association
NIDA (National Institute of DevelopmentAdmin-istration) 1978 Evaluation on national ruralwatersupplyprogram Bangkok,Thailand NIDAResearchreport under the joint sponsorshipofthe National Economic and Social DevelopmentBoard and the United NationsChildren’s Fund(ln Thai)
Rudin,A andPlumtree,A 1978 Design for plastichand pump and well Waterloo, Ont, CanadaUniversity of Waterloo Report No 3 (Project609-01-02)
32
Philippines
AntonioBravo
This reportoutlines the field testingof theWaterloohandpumpin the Philippines fromJanuary1980 to May 1982
The project sought to attain the followingfour objectives:(1) to field test theWaterloohandpumpin the rural areas,(2) to evaluatethe technical viability of the handpumpbyassessingits operationalperformancein termsof volumetricefficiency,mechanicalefficiency,andeaseof usage,(3) to determineproblemsrelatedto theoperationanduseof thepumpand provide solutions for the improvementin design and reliability of the pump, and(4) to assessprospectsfor eventualadoptionof theWaterloopumpin thenationalprogramof handpumpdevelopment
In accordancewith specificationsprovidedby the International DevelopmentResearchCentre (IDRC), prototypesof the Waterloohandpumpwere fabricated and installed inspecific areas in the Philippines Thirtyhandpumpswere installedin Laguna,Cavite,Pangasinan,and Nueva Ecija. The siteschosenwere typical ruralcommunitieswherewater supply is a problem and where aminimum of 10-20users reside
Periodicvisits weremadeto thepumpsitesto monitor the performanceof the pumpsEmploying a monitoring format agreeduponby the representativesof the countriesinvolved in the network project during ameeting held in Malaysia in the latter partof 1980, various parameterswere observedandmeasured
PumpDesignand Construction
Below-groundcomponents
The project team used a 3-inch (7 5-cm)diameterpolyvinyl chloride (PVC) pipe forthe well casing, which also functions as the
pump cylinder Pipes, 10 feet (3 m) long,were joined using PVC solvent cementandPVC couplings.The nominalpipe dimensionswere 88 7 mm OD x 79 26 mmID Thepistonwas machinedfrom solid PVC stock Twopolyethylenerings eachwith a diameterof3 inches (76 5 mm) and a thicknessof 5/16inch (7 93 mm) were used The first (upper)piston ring was 5/9 inch (14 2 mm) widewhereas the lower one was 0 25 inch(6 35 mm) Eight 0.25-inch (6 35-mm)holeswere drilled in the piston (Fig 1A)
The project team decided to use an irre-coverablefoot valve Both piston and footvalve were basically similar except that thefoot valve had larger holes than the piston(0 5 inch, 12 7 mm) (Fig. 1B) The foot valvewasredesignedtherefore,and,insteadof twopolyethylenerings, a rubbergasket(ID 2 34inch and OD 3 5 inch, 59 5 and 88.9 mm)was used
The foot valve was further modified byincorporatinga brassspring The valve seatwasalsochamferedto preventsandparticlesfrom wedgingbetweenthe flapper valveandthe base (Fig 2)
PVC pipes(OD 1 06 inch andID 0 81 inch,26 8 and20 5 mm) wereusedas piston rods.Thesewere 10 feet (3 m) long and werejoined usingPVC couplingsandPVC solventcement A PVC screenfilter (about 10 feet,3 m, long)wasincorporatedin severalpumpsduring installation This PVC screen andsandtrap was improvised by cutting slotsdiagonally along a sectionof PVC pipe withahacksaw Theseslots were0 25 inch (6 mm)apart and 1 inch (2 5 cm) long across thepipe The endof thefilter wascutandbenttogive it a conical shape
Above-groundcomponents
The above-groundcomponentsconsistedof a woodenhandleassembly,metalyoke, aconcrete pedestal,and a concreteplatformreinforced by 0 5-inch (1 25-cm) steel rein-forcement bars. The wooden handle wasoriginally connectedto the piston rod by ametal flange This flange wasconnectedbya bolt to a steel block into which a brassfitting connected to the piston rod wasscrewed A problem in the design of thislinkagecausedthepistonrod to buckleduringoperationThis wasremediedby usinga solidiron rod for the upperportion of the pistonrod (Fig 3) The pedestalfor the pump was
33
inch~ I3
I1I~MIR~~~’o,M
_ ~“4— 2---+~
A
1~~~~~
BFig. 1. (A) Piston and (B) original foot valve (1) hex nut, (2) plain washer,(3) plastic v,alvecap, (4) valve-capsupguide,(5) plastic piston ring, (6) diagonal cut, (7) threadedbolt with 0-ring welded in place, and (8) rubber seal
o_ +
Fig. 2. Modifiedfoot valve (1) nut, (2) brass valvecap andslip guide.(3) brass spring, (4) flappervalve. (5) PVC valve,and(6) threadedbolt
I I I I I
cm 8
3—
*
5
I
34
Fig. 3. Modified versionof above-groundcomponents (1) nut, (2) lock washer,(3) brass sleevebearing, (4) woodenhandle,(5) metal sleeve bearing, (6) bolt, (7) washer,(8) handle support pipe, and (9) solid iron rod
madeof concrete(Fig 4) andtheplatformonwhich thepump wasmountedwasa 5 x 10
foot (1 5 x 3 m) concreteslababout6 inches(15 cm) thick
The piston was attachedto the pump rodby a brassconnector,oneend of which wasscrewed to the bolt on the piston while theother endwas inserted into thePVC pumprod and locked with a pin
Monitoring Techniques
The following parametersweremeasuredat leastonceeachmonth andthe resultsaresummarizedin Table 1 and Fig 5
Volumetric efficiency was determinedbythe following method The theoreticalnumber(nth) of strokes needed to fill a 20-L canwas calculatedfrom the formula
nth strokes= v/[ir(D2 - d2)/4l[L/1000l
where~v = volume of container;D = insidediameterof well casing,d = outsidediameterof piston rod, L = strokelength
1 3___~
2
8
7
<U,1
2
3
4
Fig. 4. Detail5 of concrete pump pedestal, (1) handlesupportpipe, (2) solid iron bar, (3) concrete,and (4) 3-inch(7 5-cm) galvanizedpipe
inch 8ol .. ,,
cm 20
35
Table 1 Summaryof performancedata.b
Pumpnumber Type
Dateinstalled
Numberof
users
Volumetricefficiency
(%)
Waterhead(m)
Avg wateroutput(L/day)
Periodmonitored
PA4 lift 5/2/81 12 60 1 93 715 l!3/81-1515/82PA5 lift 1/3/81 15 64 2 63 800 1/4181—15/5/82NE1 lift 30/9/81 8 56 1 41 1000 30/10l81—15/5/82NE2 lift 12/10/81 150 63 0 99 2100 15/11/81-15/5/82~No observationsmade on down-time or on maintenanceand repairs
m = 328 feet, 1 L = 0 22 gallon
~75
70
~ ~ NE~
—NE1
~‘50 I I I I I I I I I Ill I I I I
1—
251__ I I I I I I~l I I I
~ 20~~ NE2
— NE1
~PA~
— PA4
5 — 1981 1982
I I I I I I I I I l~I I I I l.............M A M J J A S 0 N D I F M A M
MONTHFig. 5. Volumetricefficiency,averagewaterhead, andmonthlymeanof the averagedaily deliveryof thefourpumpsduringfield-testing
36
At a standardfrequencyof about40 strokesper minute, the same2o-L canwas filled tocapacity while counting the actual number(n~~)of strokes.Thevolumetricefficiencywasthen computedfrom the formula:
Volumetric efficiency (%) (niiJn.~)x 100
The water head (distance between thewater level in the well andthe pump spout)was measuredby using a calibrated stringweightedat one endwith a float
Sinceno counterwas installedto measurepump usage, the determination of totalvolume of water pumped was estimatedMonitors at thevillage recordedthenumberof 20-L cans (provided at the pump site forthis purpose)that thevillage peoplepumpedperday This wasusedas the basisto estimatethe amount of water pumpedper month.
The thickness of the upper and lowerpiston rings at three points was measuredmonthly. The percentagerate of wear wasthen determinedfrom the pastand presentmeasurements
HighlightsofFindings
The projectwasableto install 30 Waterloopumps in several areas.JalaJala Islands inLaguna(10), Cavite (9), Pangasinan(7), andNueva Ecija (4) Of these30 pumps, only 4(13%)wereconsideredfunctional/operationaland appropriate for monitoring purposes.The other 26 pumps were eventually aban-donedas they were besetwith one or moreof the following seven problems. (1) highleakagerate, (2) usersfound thepump to bevery difficult, strenuous,andinconvenienttousebecauseit took quite a long time to primeand draw water (This situation is describedwith humourby oneuserwith the followingstatement“Our sweatcomesout first beforewe pump out water.”); (3) defective footvalve; (4) caving in of the well, (5) peopleeventually abandoningtheuseof thepumpsbecausethey startedto draw up brackish ormuddy water, (6) piston stuckandcouldnotbe extracted,and(7) foot valveslippedout ofthe casingand droppeddown the well
Given thealternativeof being ableto useother existing pumps (Jet-maticandpitcherpumps),thevillage userseventuallyswitchedto theseother pumps The originalWaterloo-designed handpump that was fabricated
according to specificationsdid not functionas efficiently as expected when testedinitially. The following specificobservationswere made
The foot valve did not hold watersatisfac-torily because.(1) the original gasketor theseal (polyethylene ring) was not effective,(2) sandparticles became trapped betweenthe plastic flapper valve and the valve seatand causedpoor seating of the valve, and(3) therewasexcessiveclearancebetweenthevalveguideandtheflapper valve Thepiston’sperformancewas similar to that of the footvalve.
Filter
The PVC filter didnot functionadequatelyThis may havebeenbecausethescreenslotsweretoo big andallowedfine sandandsilt topass Another reasonwas that some wellswere not adequatelydeveloped
To try to overcome these problems,changeswere madein thedesignof the footvalve Thesechangesincludedincorporatingabrass spring, using a rubber gasket, andchamfering the valve seat to preventsandparticles from wedging betweenthe flappervalve andthevalve seat Thesemodificationsimproved the performance of the pumpsslightly However, it is not possibleto stateconclusively whether the improved perfor-mancewasduesolely to themodificationsorwasdue to the fact that thewells hadlowerpumping heads:the four operationalpumpshada waterheadof lessthan 10 feet (3 m)
Even whentheMalaysianfoot-valvedesignwas tried, technicalproblemswith the footvalves continued to be a major constraintOne interesting finding that has alwayspuzzledthe team is theseepagerate encoun-tered in all the wells Although a strongerspring was installed for the flapper valve, ahigh seepagerate persisted This meantthat50-70 strokes of the pump handle wererequired to raise the water column so thatwater could be obtained; however, after ashort period without pumping, the level ofthewater in thecolumn fell backto the levelof thegroundwater. Thus, priming hadto berepeatedFor areaswith a high water table,like Aliaga, Nueva Ecija, and Pangasinan,the tedious processof raising the watercolumn was not much of a problem. Fordeepwells, however, this is a very seriousproblem
37
Fig. 6. Design of mount for mechanical counter H) housing, (2) counter, (3) counter wheel and rubber ring, (4) counter support,(5) counter adjusting nut and spring, (a) guide assembly, (7) rocker, (8) rocker pin, and (9) pump handle pin
One possible explanation of the seepageproblemcould be sandintrusion interferingwith the foot-valveassemblyBecauseof thestrong suctionpower of thepiston assembly,it is surmised that the water entering thefoot valve causes turbulence in the waterreservoir; this disturbs the sandparticlesinthe wall of the well. These sandparticlesthusbecomesuspendedin thewaterandaredrawn into the riser pipe When pumpingstops,the sandparticles settle, causingthefoot valve to malfunction Sand was alwaysfound in the foot valve assemblywhenit wasextractedfor inspection Another factor thatwasconsideredwasthequality of workman-ship on the foot valve However, laboratoryseepagetests did not indicate this as theproblem This high seepagerate problemhasyet to be resolved
Mechanicalcounter
Mounts were designedfor a counter tomeasurepiston travel The initial designwasbasedon a rubber-lined flywheel contactingthepistonrod However,this designimpairedthe free motion of the pump handle, thus
limiting the travel of the piston rod andmaking it impossible to attain the desiredstroke length This problemwas due to thepeculiarity of design of the above-groundcomponentsThe designwasabandonedanda new mount was tried This mount wasinstalled in parallel with the pump handle(Fig 6) Although initial observationssuggestthat the design was feasible, the problemswith high seepagerates prevented furtherexperimentationwith the mount
Summary
Although major problemswereencounteredin this researchproject, the conceptof usingplastic materials as pump componentshasnot beentotally dismissedin the Philippines.The project team is optimistic that, givenfurther opportunitiesto investigatedesignssuitable to local conditions, an appropriatehandpumpcouldbe developedfor usein thePhilippines Therefore, technically improvedplastic handpump designs could yet provebeneficial in thedevelopmentof rural water-supply systems
inch SO~i’
cm
38
Malaysia
Goh SingYau
Thepopulationof Malaysiais approximately13 million and about 70% of the populationlive in ruralareasMore thanhalf of theruralhouseholdsare not servedwith pipedwatersupplies.In the late1960s,theEnvironmentalEngineenngUnit (now renamedtheEngineer-ing Services Division) of the Ministry ofHealth initiated a program to improve therural water supply. Part of the programinvolved the drilling of 2000 new tubewellseveryyearandfitting them with handpumpseach to serve four or five households
At present,all handpumpsinstalledby theMinistry of Healthmust be imported Becauseof the relatively high water table in mostlowland areasin Malaysia, mosthandpumpsinstalledby theMinistry of Healthareof thesuction type, such as the Dragon, Fuji, andGibson handpumpsA limited numberof lifthandpumpssuch as the Dempsterand theIndia Mark II have beeninstalled in deeperwells in hilly terrain The Ministry of Healthhas found that the suction handpumpsoftendo not last muchlongerthan 1 year The lifthandpumps, which cost much more thansuction handpumps,aremore robust in theirconstructionandlast longer However,sparepartsfor boththesuction andlift handpumps,especiallythecastmetalcomponents,arenotreadily available locally This hadled to thepracticeof cannibalizingpartsto keepsomeofthe handpumpsin operationwhile othersareabandonedfor lack of spareparts
This joint projectbetweentheMinistry ofHealth and the Departmentof MechanicalEngineeringof theUniversity of Malayawasinitiated in an attemptto overcomesomeofthesedifficulties. The main objectiveof theprojectwasto developa relativelyinexpensivehandpump that could be produced locallyfrom locally available materials The hand-pump was to be a simpledesignthatcould bemaintainedby usersat the village level
Although handpumpshave existed for a
long time, many recent studies have beenpromptedby the recognition that they stillhave an important role to play in providingsafe drinking water to the majority of therural peoplein developingcountries In 1978,the International Development ResearchCentre (IDRC) encouragedthe developmentof a PVC plastichandpumpfor usein develop-ing countries Initial studies wereconductedat the University of Waterloo and furthertests were carried out by the Consumers’Associationin theUnitedKingdom Themajoradvantagesof the polyvinyl chloride (PVC)plastic handpumpover traditionalcast-metalhandpumps include (1) simple fabricationproceduresbecausePVC partscanbesolvent-weldedtogether,and(2) maintenanceby usersat thevillage level is feasiblebecausePVC isrelativelylight andremovalof thehandpumpassemblyfrom the well for inspection andmaintenanceis easier
PresentStudy
The resultsreportedhere covertheperiodJanuary1979-June1982 Thefirst phaseof theproject lasted approximately 1 year andinvolved an analytical study and parallelexperimentalinvestigationto determinethecritical parametersfor an optimumdesignforthe plastic reciprocatinghandpump Duringthesecondphaseof theproject, 17 handpumpswere fabricatedandfield testedin two ruralareasfor 8.5 months
For thepurposeof this report,detailsof thelaboratory investigation have been omitted(see Goh 1980)
The reciprocatingpiston handpumpcon-sidered here consistsessentiallyof a drawpipe with two identical valves The bottomvalve (foot valve) is in a fixed positionat thebottom of thedraw pipe andis immersedinwater The upper valve (piston valve) isattached to a piston rod that moves thepiston valve in a reciprocatingmotion a shortdistanceabove the foot valve The cycle ofoperationis illustrated in Fig 1
TheoreticalAnalysis
Froman analysisof theforcesactingonthepiston rod for eachstroke of theoperatingcycle, a corresponding force—displacement
39
diagram can be constructed (Fig. 2). Thenormal rectangular shape of the force—displacement diagram is distorted by thefollowing (1) the valve delays x12 and x34correspondingto strokes 1 to 2 and 3 to 4in Fig 1, (2) the pressureresistanceforces~pA0 during strokes 2 to 3 and 4 to 1, and(3) the frictional forcesFuandF0
Thework input duringthecycleof operationof the pump is given by theareaenclosedbythe force—displacementdiagram For a generalcase, theareaof the force—displacementdia-gram is given by.
~FdL = A~PghTLO(1 ‘- x/L0) + FTLO
+ A~pdL
wherethe termson theright-handside (RHS)representthe work input required: to lift thewater (first term); to overcome frictionalforces (second term), and to overcomepressureresistanceforces (third term).
The volumetric efficiency (qvot) may bedefined as actualvolume of waterdeliveredper cycle divided by the volume displacedduring the suction stroke.
= [(L0 — x12 — x3~)A~— VL]IL~A~
therefore
ii~~i= 1 — xiL0 — VL/(L�,A~) [2]
where the terms on the RHS representtheeffect of valve delay (second term), andleakagepast the piston (third term).
The volumetric efficiency is therefore ameasureof wastageof possible volumetricoutputcapacityValve delaysandleakagepastthepiston decreasethevolumetricefficiency.
The mechanicalefficiency (qm~u) may bedefined as the work done by lifting waterdivided by thework input or
~7mech= 17~0jpgL0A1,hT/~FdL
Substituting equations Ii] and 121 into this
equation,we haveequation [31~1 — VL/ILOAP (1 — x/L~)]
qmerh = __________________________________1 + lF1lA~+ ~pdL)/LIllpghT(1 - x/L,,)]
For thelimiting casewhenleakage,friction,and pressureresistancelossesarenegligible(VL = FT = L~p = 0), then
tlmech = 100% Notethat, for the limiting case,x/L neednot bezero In otherwords, if leakage,friction, andpressureresistancelossesarenegligible, themechanicalefficiency is independentof thevalve delay. The mechanical efficiency is
(1) (2) —3’
Fig. 1. Cycle of operation of reciprocating piston hcindpump
(3) —-3. (4)
40
U
0
ii-
1
Fig. 2. Force—displacementdiagram
therefore a measure of the wastage ofmechanicaleffort as a resultof leakage,fric-tion, and pressureresistancelosses.
Equations [2] and[3] requirea knowledgeof the unknownsx (the total valve delay),VL (theleakagepastthepiston), ~p (thepres-sure resistancelosses),andFT (thetotal fric-tional force actingon thepiston) OtherthanFT, which must be determinedfrom experi-ments, theother unknownsmay be derived
Delay in closureof thevalve may be signi-ficant if the valve flap is light or the valvegapis large, or both The magnitudeof thevalve delay may be determinedfrom a con-siderationof the relativemotion of thevalveflap (which is falling under its own weight)and the valve seaton the piston (which ismoving in a vertical reciprocatingmotion). Itcan be shownthat thevalve delay is a func-tion of theweightof thevalveflap, theheightof the valve gap, and the leakagepast thepiston
In thepresentdesign,thePVC piston fittedwith two polyethylenepiston rings slidesin acylinder slightly larger than thepiston Thewater-sealingactionis performedby thepistonrings, whicharein “contact”with thecylinderwall througha thin lubricatingfilm of water
Froma considerationof thewaterflow in thenarrow annular passagebetweenthe outersurfaceof thepiston rings andtheinnerwallof the cylinder, it can be shown that theleakagepast thepiston rings is given by
VL = Cld~hTIN [4]
whereC1 is a constantfor a particularsetofrings
The pressureresistanceforces occuras aresultof pressuredropsacrossthepistonandfoot valve and may be expressedas
L~p=O5ki.pUo2 15]
wherekT is thepressureloss coefficientthatcan be determinedfrom the geometryandvelocityof waterflow throughtheorificesinthe piston and foot valve
The aboveanalysishasignoredtheeffectsof accelerationandretardationof the wateron the piston as well as the oscillationsinducedin thepiston rod asa resultof impactloadingon suddenclosureof thepistonvalveIt can,however,beshownthat theareaof theforce—displacementdiagramis notaffectedbytheaccelerationandretardationforcesas thework doneby oneis canceledout by thework
+ W~2+ B!, + BR)
DISPLACEMENT
41
absorbedby the other Similarly, it may be to the top endof thepump rod The displace-shownthat, becausetheoscillations induced ment was measuredby a 25-cm potentio-in the piston rod arelightly damped,energy metric displacementtransducerconnectedtodissipation is small Hence, this simple the pin at the slide The outputsignalsfromanalysisappearsadequateas can be demon- the strain gaugeandthedisplacementtrans-stratedby comparingthepredictionswith the ducerwere fed via a dynamicstrain-metertoexperimentaldata a storageoscilloscopeA Polaroidinstant-film
camera was used to record the strain—
ExperimentalInvestigationdisplacementtraceon theoscilloscopescreen
It is important to ensurethat the appliedforce andrecordedstrainrelationshipis linear
The experimentalarrangementfor testingthe handpumps consisted essentiallyof ahandpumpassemblylifting water to a maxi-mum of 9 m from a central constant-levelreservoir Theassemblycouldbeconvertedtolift water from 6 or 3 m when required Thewater at the outlet of the handpumpwasreturned to the central reservoir through areturn pipe The handpumpwasdriven by a1-horsepowerDC motorvia a reductiongearandchaindriveassembly.The rotary motionof the flywheel wasconvertedto areciprocat-ing vertical motion by a pin and slideMounting holesat variousdistancesfrom thecentre of the flywheel were provided tochange stroke length when required The
This can be checkedby a calibrationtest Forthe linear case, the area of the force—displacementloop is equalto the work inputper cycle. The area may be obtained bymechanicalintegrationusing a planimeter
Figures 3 and 4 show a sample of thepredictionsof volumetricandmechanicaleffi-ciencies computed from the mathematicalmodel andthe experimentalresultsobtainedfront theparallelexperimentalinvestigationsFurther comparisonsof the predictionsandthe experimentaldataover a rangeof para-meters showed that the predictions are inremarkableagreementwith theexperimentalresultsconsideringthe simple model usedinthe presentanalysis(Goh 1980)
speedof stroke could be varied by changing The laboratory investigationshowedthatthespeedof theDC motor This wasachieved the leakage past the piston, the frictionby varying the input voltageto thearmature betweentherings andpump cylinder,andthecoil while keepinga full 240V acrossthefield pressuredrops across the piston and footcoil of the DC motor valve havea very pronouncedeffect on pump
The strain in thepump rod wasmeasured performance as characterizedby the volu-by four straingaugeson a proof-ringattached metric and mechanicalefficiencies
7!~rus~80
>..Uz04Uu~LUU
F-’LU
-I0>
100 -
80
60
40 -
20 -
0
100
U2:
“C 40 -
U2:4:x 20 -U
Stroke length
A 15-inch(38-cm)V 30-inch(76-cm)
0 50-inch(127-cm)
Stroke length
a 15-inch(38-cm)V 30-inch(76-cm)
0 50-inch(127-cm)
I I I I
20 30 40 50 60 70 80 20 30 40 5060 70 80
STROKERATE (strokes/minute)Fig. 3. Variation of volumetric efficiency with speed ofstroke application for three stroke lengths (theoreticallines and experimental data points) valve gap, 0 113inch (2 87 mm), valve weight, 0 029 lb(farce) (13 15 g),orifice/piston ratio, 16 40/s, and water head, 230 inches(3 66 m)
STROKE RATE (strokes/minute)
Fig. 4. Variation of mechanical efficiency with speedof stroke application for three stroke lengths (theoreticallines and experimental data points) valve gap. 0 173inch (2 87 mm), valve weight, 0 029 lb(force) (13 15 g).orifice/piston area ratio, 16 4%, water head, 230 inches(3 66 m)
42
OptimumDesign Field Investigation
It is difficult to specifyanoptimum designbecause,for practical reasons,it must be acompromise between a number of factorssuch as simplicity, easeof manufacture,andcosts as well as high volumetric and mechan-ical efficiencies However, the followinggeneral comments may be made on theoptimum design.
The ratio of orifice areasin thepiston andfoot valve should be sufficiently large topreventhigh pressuredropsat the desiredspeedof operation For the presentconfigu-ration, a valuein excessof 15% is satisfactory
The piston speed,which is a product ofstrokelength andspeedof strokeapplication,should be sufficiently high to ensure lowleakagepastthepiston rings A boy operatinga handpumpis observedto be ableto achievean averagepiston speedof the orderof 300
inches (762 cm) per minute, i e, a 5-inch(12 7-cm) stroke at 30 strokesper minute ora 3-inch (7 o2-cm) stroke at 50 strokesperminute
The valve weight must be sufficient toensureminimal closuredelayat theparticularvalve gap Too small a valve gap is likely toincreasethe pressuredrop acrossthepistonandfoot valveresulting in decreasedmechan-ical efficiency A valve flap weight between0 45 and0 96 ounce (13 6 and 27 2 g) operatingwith a valve gap of 0 25 inch (0 64 cm) givessatisfactoryperformance
Increasedwaterheadincreasesthe leakagepast the piston rings resulting in decreasedvolumetric efficiency However, if the pumpis operatedat a sufficiently high pistonspeed,the effect is much reduced
Becausethe sealingactionat thepiston isperformedby thepiston rings, small dimen-sional variations of clearancebetween thepiston and draw pipe have no effect on theperformance characteristics No significantdeterioration in performancewas observedfor up to 0 1575 inch (0 4 cm) difference indiameterof piston andpump cylinder
Becausethe use of a conicalentry in thepiston andfoot valve hasonly a small effecton themechanicalefficiencyat normalopera-ting speeds, the simplicity of the suddencontraction entry may be retained to savecost of manufacture However, the holesshould, preferably, be slightly chamferedatthe entry and outlet
In the design of pumps,there is a certain“maximum suction depth” belowwhich it isno longer possible to draw water by suctionBelow this depth, water must be raised either
by lifting or someother method This distinc-tion is important as a suction handpumpisgenerally simpler and less expensivethan alift pump
In most lowland areas in Malaysia, thewater table is relatively high andwatermaycommonly be found at depthsbelowgroundthat are less than the maximum suction depth
of about 26 feet (8 m) In hilly areasandinsomeexceptional situations in lowland areas,the water table sometimes falls below the
maximum suction depth There is, therefore,a requirement in Malaysia for two variationsto the basic design of the handpump, i e, thesuction handpumpand the lift handpump
Commonfeaturesof thesuctionand lift handpumps
Figures 5 and 6 show the main features ofthe present design of the suction and lifthandpumps The common features of thesetwo handpumpsare (1) a mild steelstand,(2) a leverage system consisting of timber
linkages, galvanizediron joints, and galva-nized iron/oil-impregnatedtimber bearings,(3) a PVC pump cylinder, (4) a PVC pistonwith two polyethylenerings,and(5) a remov-ablePVC foot valve with a rubber seal
The mild steel stand provides a firmsupport for the pumping cylinder andleverage system Timber linkages are usedbecausethey are readily availableand easyto replace Oil-impregnatedtimber bearingshave been tested in the laboratory (Stern-berg 1978) and found to have outstandingperformance characteristics The use of aPVC piston with two polyethylene ringssliding in a PVC pump cylinder substantiallyreducesthe friction without sacrificing highvolumetric efficiency (Goh 1980) The basicPVC componentfor thepistonandfootvalveis identical This allows for cost savings byusing one injection mould for both items
Field test
A total of 12 suction and 5 lift pumps werefabricated and installed for field testing in
43
Fig. 5. (A) Lift and (B) suction pumps (7) wooden parts,(5) concrete,(6) piston, (7) foot valve, and (8) casing pipe
two ruralareasin Malaysia The objectivesofthe field-testing program were to evaluatethe technicalperformanceunderfield condi-tionsandtheeconomicfeasibility of wide-scaleadoption for rural use in Malaysia.
Measurementtechniquesfor use in the field
A number of laboratory measurementtechniques for the determination of thetechnical performance of the reciprocatingpiston pumpcannotbeusedin thefield eitherbecausesophisticatedelectronicequipmentisneededor becauseof physical constraintsatthehandpumpsite Severalsimple field mea-surement techniques and apparatusweredeveloped for use in the field evaluationprogram
(2) galvanized parts, (3) PVC parts, (4) mild-steel stand,
Depth of water table In this method, thedifferencein theelectricalresistanceof waterandair is employedto detectthewater levelin thewell The arrangementconsistsessen-tially of a sufficient length of twin-core wireand a perspexconical probe head The twinwires arethreadedinto thecentralhole of theprobe headand the endsof the wires entertwo lateral holesandemergeflush with thesurfaceon either sideof theprobehead Theprobe headis machinedwith a total conicalangleof 200 andthesurfacepolishedto facili-tate drip-dry actionwhen theprobe is liftedout of the water A multitest metercapableof measuringin the rangeof up to 300kohmsis connectedto theother endof thetwin-corewire to measurethe resistanceacrossthetwo
A B
44
Fig. 6. (A) Piston assemblyand (B) recoverablefoot valverings (polyethylene), (4) six equally spaced holes, (5) bolt,pin, (9) recovery pin, ~nd (10) double-lip rubber seal
terminals When theterminalsareexposedtoair, the multitest meter readsopen circuitWhen the terminalsare immersedin water,the meter readsin the order of 200 kohmsDiscretechangesfrom infinity to 200 kohmscan be obtainedon the multitest meter forchangesin the levelof waterof less than0 24inch (6 mm) The depthof the water tablefrom thegroundmay bedeterminedfrom thelength of the wire
45
(1) lock nuts, (2) valve flap (natural rubber), (3) piston(6) valve stop, (7) PVC plastic, (8) hole for connecting
Total usageof handpump For a comparativestudy of wear and physical deteriorationofmajor handpumpcomponents,a measureofthe total usageof thehandpumpin the fieldover the periodunderconsiderationis neededFigure 7 shows an arrangementwhere aVeeder-Roottotalizing counter (which doesnot registerreversedrotation)wasadaptedtomeasurethe cumulative travel of the pistonrodduring theusefulstroke Themountingis
A B
LJ
Fig. 7. Mounting for Veeder-Root(V-R) counter on handpumps(all dimensions in mm, 70 mm = 0 3937 inch)
(1) piston rod, (2) sliding fit to allow for lateral movementof counter, (3) two 0-rings (73 mm ID x 2 mm diam(0 51 x 0 08 inch) cross-section), (4) V-R flange-casecounter (#74-6125-001), and (5) mounting screw to top plateof handpump
‘////A’///Y ‘y//
‘V// ‘/A’//l
L I
~38 38
46
made from a 0 5-inch (i 27-cm) thick plateAfter drilling the central hole (slightly largerthan the piston rod), the plate is cut intotwo similar halves A third half-plate isrequiredfor mounting the assemblyso thatthemovablehalf-pieceis clearof thetop-plateof the handpump The Veeder-Root counteris mountedon themovablehalf-plate,whichis spring-loadedto ensureconstantcontactoftheroller with thepiston rod Thefixed half-plate should be mountedclear of the pistonrod (taking into considerationany swing orinclination of the piston rod duringtheappliedstroke)
Work input The method used in thelaboratory to determine work input is notsuitable for use in the field becauseof thedifficulty of transporting several pieces ofsensitiveelectronic equipmentand ensuringthat theyremainin good operatingconditionat the field site Moreover, the methodalsorequiresa highly trained technicianto carryout themeasurementsIt is, therefore,desir-able to have a simpler, lighter, and morerobust instrument that can be readily carriedto the field andoperatedby a technicianwithminimal training
The mechanical instrument used tomeasurework input at the field site consistsof essentially (1) two steel bars, held togetherat both ends by adjustable clamps, that deflectrelative to each other when a load is applied tothemidpointsof thebars, (2) adial gaugethatmagnifies the deflection of the steel bars,(3) an indicatordrum that rotatesthetracingpaper as the instrument is displaced from afixed position, and (4) apen connectedto thedial gauge by a string that is held taut by arecoiling spring During theoperationof thepump, therelativedeflectionof thebars,afterbeingmagnifiedby the dial gauge,causesthepen to moveperpendicularto the rotationofthe drum The resultantdiagram tracedonthe paper on the drum gives a force—displacement loop that is equivalent to thatobtained by the laboratorymethod
The operating range of the mechanicalinstrumentwith respectto theapplied forcemay be changedby changing the distancebetween the adjustableclamps so that theeffective deflectinglengthof thesteelbarsiseither reducedor increased
Field monitoringprocedures
Four separatesetsof forms wereused for
monitoring pumps in the field. They were(1) Form KI wasdesignedfor useby Ministryof Healthpersonnelto recordcounterreadingsat 2-week intervals, a second copy of thisform, which was updatedregularly,waskeptat the University, (2) Form A (Well and PumpSpecifications)was used to recordbasicinforma-tion on thewell andpump,andwascompletedbefore thestart of themonitoring period; (3)Form B (Site Visit Data Sheet)wasthebasicformusedfor field monitoringandwasdividedintofour sections— (a) mechanicalperformance(completed at eachmonthly visit), (b) main-tenanceoperations(recordedwhennecessary),(c) wearmeasurements(measuredbimonthly),and (d) user feedback (whenever possibleduring monthly visits), and(4) Form C (Failuresand Repairs)was completedwhen apumpfailedand included details of repairs and othermeasures taken
Although all 17 handpumps were installedandusedby villagersfor approximately2 years,it was only during the last 8 5-month periodthat they were fully monitored
Fieldperformance
To facilitatecomparisonamonghandpumps,thedatafor all 17 weresummarized(Table1)Maintenanceand repairs carried out on thepumps were also summarized(Table 2)
Performance results were computed fromfield data for all 17 handpumps Figure 8 givesan exampleof the resultsof thedatafor onepump To isolate theeffect of waterheadonthe volumetricefficiency, plots of (1 —
hT againstcalendartime weremadefor eachhandpump(see,for example,Fig 9) Therewasconsiderableexperimentalscatterfor smallvaluesof (1 — iivo~)/h-r,which is equivalenttohigh valuesof ~ andlow valuesof hT Forlargervaluesof (1 — q~0~)/hT,theexperimentalpoints showed a more consistent trend Thestraight lines throughtheexperimentalpointswere plotted using the method of leastsquares For most cases, the expression(1 — ,701)/hTdecreasedwith calendartimeindi-cating that the sealing efficiency improveswith time
Variations of water head with calendartime showed a similar patternof variation ineach district, indicating the dry and wetseasonsin the respectiveareas
The averagevolume of waterdeliveredperday was computedfrom the counterreadings(which were the sum of the piston travel)
47
Table 1 Summary of performancedata
Broken PVC spout replacedLeaking foot valve changedtonew Linard rubbervalve flap
NoneWoodcovercrackedbutno actiontakenFulcrumtoo looseandretightenedPiston bolt brokenduringdismantling and replaced
PVC piston rod crackedat brasspin hole, relocatenew hole
Maintenance
Pumpnumber
EfficiencyNumber WaterDate of users head
Typeb installed (persons) i~,,,i t~,,,,a (m)
and repairsAverage Downwateroutput time Parts Time Period
(L/day) (days) (US$) (hours) monitored
PK oolS S 19/11/80 50 95 0 - 2 6 381 0 0 9 40 0 50 6/7/81-16/3/82PK 002S S 18111/80 40 98 0 - 2 4 58 9 0 0 26 0 50 6/7)81-16)3/82PK 003S S 19/11/80 40 95 0 - 2 8 93 8 0 - - 7/7/81—16/3/82PK oo4L L 2112/80 30 890 — 2 5 502 7 0 — - 7/7/81—17/3182PK 005L L 2/12/80 50 98 0 - 2 7 186 2 0 8 45 1 50 7/7/81—17/3/82PK 006S2 S 3/2/81 40 84 0 - 3 5 2176 60 — 2.50 7/7/81—17/3/82NS oolS S 6/7/80 25’ 83 5 59 6 3 7 487 5 3 - 0 50 25/6/81- 7/4/82NS 002S S 6/7/80 25” 93 7 59 1 3 7 133 5 0 - - 25)6)81-12/4/82NS 003L L 24/7/80 25’ 92 0 - 9.2 152 2 4 0 50 3 00 29/7/81—19/4/82NS 004S S 24/7/80 25’ 792 64 0 70 852 0 - — 29/7/81—14/4/82NS oosS S 28/7/80 25’ 84 4 62 2 6 9 59 9 0 — - 11/8/81—19/4/82NS 0065 S 28/7/80 31 81 7 65 0 4 9 351 3 0 9 40 0 17 17/6/81-12/4/82NS OO7PS PS 18/9/80 40 75 0 — 4 9 524 0 1 10 50 2 75 17/6/81— 9/4/82NS oo8PL PL 25/9/80 50’ 90 0 - 5 4 826.8 0 7 80 3 00 22/7/81— 9/4/82NS oo9L L 2/10/80 15’ 70 5 65 2 76 2426 10 ft75 21 75 21/10/81—16/4/82NS doS S 9/10/80 50’ 95 0 70 4 4 6 7182 0 - — 15/6/81- 7/4/82NS 0111’S PS 16/10/80 35’ 790 - 5 8 506 0 4 12 00 7,50 20/7/81-14/4/82
m 3 28“S. suction,
feet, 1 L = 0 22 gallonL, lift, PS, pressure-suction,PL, pressure-lift
‘The numberof householdsrather than numberof userswas givenAn averageof five personsto a householdis usedas an estimate
Table 2 Summaryof maintenanceandrepairs for pumpsat Perak(PK) and Negri Sembilan (NS).
Cost of TimePump Date Maintenance Down time spareparts required
number installed date (days) Description (US$) (hours)
PK ooiS 19/11)80 27/1/81 0PK oozS 18/11/80 27/1/81 0
PK 003S 19/11/80 0PK 004L 2/12/80 27/1)81 0PK 005L 2/12/80 10/9/81 0
9 40
0 26
8 45
0 500 50
0 50
1 00
PK 00652 3/2/81 7/7/81 60 Manyproblemsdueto well having 2 00insufficient water, lift pumpreplacedwith suction pump
10/9/81 0 Fulcrum arm/basetoo tight and 0 50reset
10/9/81 0 Wood coverpiececrackedbut not -
replaced
NS OO1S 6/7/80 30/9/81 0 Pistonand foot valve cleanedof 0 50iron stains
3/2/81 3 Pnming inlet not closedproperly - -
NS 002S 6/7/80 — 0 None - -
NS 003L 24/7/80 28/9/81 4 Pistonrod broke at socketjoint 0 50 3 00NS 004S 24/7/80 - 0 None - -
NS 005S 28/7)80 - 0 None — -
NS 0065 28/7/80 25)11/81 0 Spout missing 9 40 0 17NSOO7PS 18/9/80 20/2/81 1 Broken brasspiston-connecting 6 00 2 00
bolt replacedwith mild steel(chromed)part
,~ 23/4/81 0 Foot-valve brassbolt broken 4 50 0 50‘0 17/6/81 0 Wooden fulcrum arm/baseloose - 0 25
andretightenedNSOO8PL 25/9/80 20/2/81 0 Replacedbrassconnectingbolt 6 00 1 00
with mild steelpart (no sign ofdamage)
19/3/81 0 Brasshingejoint badly worn, 1 75hingeeliminatedfrom the design
30/9/81 0 Wood fulcrum armcracked 1 80 0 25NS 009L 2/10/80 19/12/80 0 Leaking foot valve replacedwith 0 25 6 75
Linard rubbervalve flap9/6/81 10 PVC piston rod broken 0 50 15 00
19/8/81 0 Nuts of woodfulcrumloosedueto - -
wearof thewoodandretightenedNS abS 9/10/80 - 0 None - -
NS 01bPS 16/10/80 21/1/81 1 Broken brassconnectingbolt 6 00 2 00replaced
13/2/81 3 Same part brokenagainand 6 00 3 00replaced
20/2/81 0 Replacedabove part with mild 2 50steel part
— —
800
andtheaveragevolumetricefficiency for theperiodbetweenmonitoringvisits Theaveragevolume of water delivered per day variedfrom handpumpto handpumpas well as forthesamehandpumpat differenttimes of theyear It is interestingto note that theaveragevolumeof waterdeliveredperdaywashigherwhen the water head was greater (whichcoincidedwith the dry seasonfor theparti-cular district)
The total (accumulated)work output wascomputed as the product of the averagevolume of water delivered per day and the
averagewater head for the period betweenmonitoring visits The total work output fordifferent handpumps varied considerablyeither becausesomehandpumpswere usedmore thanothersor becauseof adifferenceinwater head One would expect the generalwear andtear of thehandpumpsto be moredependenton the total work output ratherthan calendartime
Wear
Although wear measurementswere madeon a numberof components,significantwear
‘-IL)
~uZ<L)
5
4
3
2
i--n
‘—‘———I.—....
n___l
700
600
aE—’
500 ~
1.00
>0400
‘U
300 < -~
-.~
200
100
0
-®
—l
A S 0 N D J F M AMONTH
Fig. 8. Performance characteristics of handpumpPK QOIS in field test (7) volumetric efficiency, (2) water head,(3) water delivered per day, (4) cumulative(total) work output (m~= m
3 (volume) x m (lift)), (5) percentagewear oftop wooden bushing (00), (6) percentage wear of bottom wooden bushing (900), (7) percentagewear of top pistonring, and (8) percentagewear of bottom piston ring
50
was found only in: (1) the top woodenbush-ing for the piston rod, (2) the piston rings,(3) thehandpumpcylinder, and(4) the brasspins used in the pin-joints of the piston rodThe wear of the brass pins, however, was notmeasured. Wear measurements in the field
PKOO1So 1. PKOO2Sv 2
PKOO3SA 3PKOO4L. 4
6E
I-
—--.4. —
~ 3
2 ~ ~__- 2
1 - _~~-_-—---- —-.
1981 19820 I i~I I I
J A SONDJ FMAM
MONTH
Fig. 9. Variation of (1 — 17V01)/hT with calendar time for
four pumps
30- 2
25 —
~2OliD
zU~ 15 -
I--.
10
05 —
0
are not always taken under ideal conditionsand some of the measurements,especiallyofcomponents with very small wear, aredefinitely in error as indicated by negativewear readings However, when the wearbecomes large, consistent trends may beobserved
Wear in the bore of the lop wooden bushingWear in the bore of the wooden bushing sup-porting the piston rod was calculatedas apercentage from (measured diameter —
~ 6-
cv~ 51
50 100 150 200 250~ 1CUMULATIVEWORK OUTPUT(m
4)
Fig. 10. Variation of wear with cumulative(total) workoutput [or handpumpPK OOIS (I) top wooden bushing(00), (2) bottom wooden bushing (90°),(3) top piston ring,and (4) bottom piston ring
0
0
= 60°
= 120°
0-
0-.
S
I,
I& = 240°
PUMPCYLINDER
& = 1800
I I I I I I I I10 20 30 40 50 60
AXIAL DISTANCE (mm)
Fig. 11. Variation of wall thicknessof pump cylinder with axial distancefor pump NS 0105
51
original diameter) x 100!(original diameter)The angularposition of thehole wasdefinedas 8 = 0°for thediameterin theplaneof thelever and 8 = 90°for the diameterperpen-dicular to it Wear in theboreof thewoodenbushing was plotted against calendar timeand against total work output (Fig 10) Asexpected,the wear was more dependentonthe total work output thanon calendartimeThe magnitudeof the wearalsovaried fromhandpumpto handpumpdependingnot onlyon the total work output but also on thedampness of the wooden bushing and theslack in the lever system It was also notsurprising that the wear in the plane of thelever (8 = 0°)wasgreaterthan that perpen-dicular to it (8 =
Wear of piston rings Wear of the highdensity polyethylenepiston rings was calcu-latedas a percentagefrom (original thickness— measuredthickness)x 100/(original thick-ness) Wear of the piston rings was alsoplottedagainstcalendartime andagainsttotalwork output. Again, the wear is more afunction of total work output than calendartime The average total wear of the pistonrings over the B 5-month period was of theorderof 4%.
Wear of the pumping section of the handpumpcylinder At the end of the field test, thePVC pump cylinder was replaced by a newcylinder The original pump cylinder wascutinto two halvesalong its cylindrical axis andthewall thicknessmeasuredin thelaboratoryFigure 11 shows thevariation of wall thick-ness of the pumping section of the PVCcylinder of a handpumpthathadbeenin usein the field for approximately2 years Wearwasverysignificantandhadextended,in thiscase,to abouttwo-thirds of theoriginal wallthicknessof the handpumpcylinder In thepresent design, high density polyethylenepiston rings wereusedandthe aboveresultsshowthat thePVC cylinder wearsmorethanthe polyethylene piston rings This showsthat the original choice of materials is notsatisfactory Because it is easier and moreeconomical to replace the piston rings thanthepumping cylinder, it is more desirable tohave a combination where the piston ringswear while the pumping cylinder is moreresistentto wear.
A member of thepresentproject team,in acurrentexpenmentalinvestigation,hasshown
that low densitypolyethylenewearsabout10times faster than high density polyethylenewhen rubbingagainstPVC material in clearwater However, until further tests arecarriedout to determinewhetherpistonringsmadefrom low density polyethyleneor someother material can reducewear in the PVCcylinder significantly in a field environment,atemporary solution to the cylinder wearproblem is to raise the piston to a new,unworn, section of the pumping cylinderafter every2 years of use
Conclusions
After approximately2 yearsof use in thefield, except for severe wear of the PVCpumping cylinder, thehandpumpof thepres-ent design appears to have withstood thewear and tear of everydayuseandrequiredonly minor maintenanceandrepair Routinemaintenanceis requiredafter every 2 yearseither to replace the worn section of thecylinder or to raise the piston to an unwornsection of the cylinder
Because the major components of thepresenthandpumpare made from plastics,the use of injection moulding techniquesoffers great promisefor costreductionwhenthehandpumpis producedin largenumbers
Acknowledgment This projectwas fundedby agrant from the InternationalDevelopmentResearchCentre This is gratefully acknowledged Theauthor would like to thank the local Ministry ofHealth personnelin Ipoh, Seremban,and KualaPilah for their cooperationin getting the hand-pumps installedandmonitoredduringthe field testLast, but not least,our appreciationgoes to thevillage usersof our presenthandpumpsfor theirpatienceand tolerancedunng the field tests
References
Goh,S Y 1980 Theperformancecharacteristicsofa reciprocating piston water lift pump Ottawa,Ont,CanadaInternationalDevelopmentResearchCentre Interim progressreport,Waterpumpingtechnology— Global project
Sternberg,Y 1978 Testing of wood beanngsforhandpurnpsWashington,DC, USA InternationalBank for Reconstructionand DevelopmentResearchWorking PaperSeries,P U ReportNoRES 13
52
Overviewof TechnicalPerformance
Goh SingYau
After the Consumers’Associationlabora-tory tests,a proposeddesignof thepumpingelementwith a piston anda foot valve wasacceptedby the International DevelopmentResearchCentre (IDRC) for testing in theIDRC Asian network of handpumpprojects
Developmentof theIDRCDesign
The piston assemblyconsistsessentiallyofa polyvinyl chloride (PVC) piston with twopolyethylenerings (see Fig 1A) Laboratorytests were carried out in Thailand andMalaysia to determinethe critical valuesof.(1) orifice/piston and orifice/foot valve arearatios, (2) valve weight, (3) valve gap,(4) stroke length, and (5) the stroke rateneededto obtain optimum performanceascharacterized by high volumetric andmechanical efficiencies The results of thelaboratory tests were incorporatedinto themodified versions of the piston (see fig 6 ofthe Malaysian paper, page45)
Initial laboratoryandfield testsof thefootvalve (see Fig 1) in Thailand, Sri Lanka, andthePhilippinesshowedthat theoriginalpoly-ethylenecupsealsdid notprovidean effectivewater-tight seal and leakagewas excessive.To overcomethis problem, a nonremovablefoot valve (see fig 7 of the Thailand paper,page28) was used, with the foot valve solvent-welded or bolted to thepumpcylinder Later,a double-lippedrubbersealwasdevelopedbytheMalaysiangroupto replacethepolyethy-lenecupsealfor use in aremovablefootvalve(see fig 6 of theMalaysian paper,page45)
The original IDRC design recommendedaPVC or polyethylenevalve flap The useofthese valve flaps causedexcessive leakagethat was particularly noticeablein the footvalve Theinitial modificationcarriedout by
Thailandwas to usea rubberdiscwith a brassbacking-plate glued on to it To preventaccumulation of sand at the valve seat,elevatedlips were cuton thevalve seat Theseal could be further improvedwith a springto pressthe valve flap onto the seat How-ever, laboratory tests in Malaysia indicatedthat the spring-loadedvalve flap increasedthe work input substantially and hencedecreasedthe mechanicalefficiency
It was discoveredafter severalmonths ofuse that the glued-on brass backing-platedetacheditself from the rubberdisc A furthermodification was subsequentlysuccessfullyintroduced in Thailand by replacing theprevious valve flap design with a singlerubberdisk 0 25 inch (0 64 cm) thick withoutthe backing plate or the spring This modi-fication wasalso found to be successfulin theMalaysian project
In Sri Lanka,wheretherewasdifficulty inobtaining polyethylenerods,leathercupsealswere usedin placeof polyethylenerings andcup seals.
Developmentof theAbove-GroundComponents
Different configurations of the above-ground componentswere used in the fourcountries, theseare summarizedin Table 1
The concrete pedestal adopted by thePhilippines appears to be a simple, cheapalternativeto the traditional pump stand
Timber leverhandles(usedin Malaysiaandthe Philippines) and timber/galvanized-ironbearings(used in model L3 in Sri Lankaandin Malaysia)havebeenfound to be practicalanddurable When cheaptimber is available,timber componentsappearto be theobviouschoice The use of timber componentsalsosimplifies maintenanceand repairsby hand-pump usersat the village level.
Field Tests
A mid-project review meetingattendedbyinvestigatorsfrom each of the participatingcountrieswasheldin Kuala Lumpurfrom 26-28 August 1980 to exchangeexperiencesgainedin laboratoryandfield testsandalsotoproposea common field-monitoring proce-
53
0 1250251:
0 50
it
0 50
025
Fig. 1. Designof original (A) piston and (B) foot-valveadaptortestedin the Asianhandpumpnetwork(all dimensionsin inches,1 inch r 25 4 mm) (1) 7/16-inch (111-cm)diameterbolt, 5 5 inches(13.97 cm)long, (2)0 5-inch (1 27-cm)holesat 1.53-inch (3 89-cm)pitch diameter
dure for the four projects Two importantinstrumentsrequiredfor thefield-monitoringprogram were subsequentlydeveloped tomeasuretheaveragewaterdeliveredper dayover the monitored period (a totahzingcounter) and the work input of the pump(a mechanicalplotter) Thesetwo instrumentsare describedin detail by Goh (1980)
The detailedfield monitoring procedure,asproposed in the mid-project meeting, wascarriedout only in Sri LankaandMalaysiaBythe time the instrumentationrequiredfor thefield monitonngwasdeveloped,theThailand
programhadcome to anendandtheprojectin the Philippines encounteredproblemsinthe foot valve and the well
The handpumpsin Sri LankaandMalaysiawere in the field for more than 2 yearsHow-ever,becauseof thedelay in thedevelopmentof the field monitoring instrumentation, thefield monitoring program was carried outonly for the last 13 monthsin Sri Lankaandthe last 8 5 months in Malaysia The resultsof these monitoring studies are reportedinthe country papers
The volumetric efficiency of the hand-
A B
54
Table 1 Summaryof above-groundsystems
Country Leverage Water headandmodel Pedestal system Handle Top bushing Pistonrod (m)
(iiUi
Sri LankaLiL3ViThailand
DMR (modified)
ARD (modified)
AlT-PVC
PhilippinesLift
MalaysiaSuction/lift
Angle-iron frame Brassbushing 25-mmCl pipe Cl reducer 25-mm PVC pipe 2-7Metal cage Timber bearings 25-mmCI pipe Timber 25-mm PVC pipe 2-8CI pipe Direct lift Timber Timber 25-mm PVC pipe 1-4
Cast-ironpipe Cast-iron lever armwith metalbushing
Cast-ironpipe Stuffing box li-mm diam steelrod
2—21
Cast-ironpipe Cast-iron rack andpinion gear
Cast-ironpipe Metalpackingnut
il-mm diam steelrod
2—10
Mild steelpipe Cast-metalleverarmwith ball bearings
50-mmCI pipe Stuffing box li-mm diam steelrod
2-7
Concrete Concretepedestalwith brass bushing
Timber CI reducer Mild steel rod 2—8
Mild steelpipe Timber linkagesandbearings
Timber Timber Top 1 m 25-mmCI pipe, remainder25-mm PVC pipe
2-u
10 mm = 039 inch, 1 m = 3 28 feet
pumpsafter approximately2 yearsof usewasstill relatively high showing that the pistonring sealswerestill effective The mechanicalefficiency measuredat the piston rodwas oftheorderof 60% after about2 yearsof useinthe field This comparesveryfavourablywiththe resultsof mechanicalefficiency measure-ments on new handpumpsfor 23-foot (7-rn)water headsat the Consumers’Associationlaboratories(UNDP/World Bank 1982)
Although wear measurementswere madeof most moving parts, wear was generallyslight except for the top wooden or metalbushingandthepumpcylinder After 2 yearsof usewith an equivalentof 100000gallons(450000 L) of wateroutput, thecylinder wallof onepump hadworn by approximately70%.The wear in the correspondingpiston ringswas only about 1% For the presentdesign,where high density polyethylenerings wereusedwith PVC cylinders, it is recommendedthat thepump cylinder be inspectedfor wearevery 2 yearsor after an output of 100000gallons (450000L) of water
ConcludingRemarks
The detailed results of the projects inMalaysia, the Philippines, Sri Lanka, and
Thailand are given in the end-of-projectreports However, overall, the followinggeneralremarksmay be made.
The PVC handpump has been appliedsuccessfullyfor usein wellsup to a maximumwaterhead of 33 feet (10 m)
The resultsof the field testsindicatethat.(a) a concretepedestal,(b) timber handles,bearings,and bushings,and (c) PVC pumprods andcylindersarepracticalalternativestothe traditionalcastor weldedmetaldesignsofhandpumpsused for water headsof up to33 feet (10 m)
There is insufficient field monitoring dataon the PVC handpump for water headsgreaterthan33 feet (10 m) to beableto makea positivestatementon its usefor suchdepths
References
Coh, S Y i980 Theperformancecharactensticsofa reciprocatingpiston water lift pump Ottawa,Ont, CanadaInternationalDevelopmentResearchCentre Interim progressreport,Waterpumpingtechnology— Global project
UNDP/World Bank 1982 Laboratory tests onhand-operatedwaterpumpsfor usein developingcountries Washington,DC, USA UNDP/WorldBank Rural water supply handpumpsprogram,ReportNo 1
56
EconomicAnalysisandPotentialMarkets
TanBockThiam
Of thepeoplein developingcountries,75%do not have accessto adequatesuppliesofclean water (IRC 1982) The useof hand-pumps in areas where there are adequatesuppliesof groundwateris thesimplest,leastcostly methodof providing the rural popula-tion with cleanwater As increasedattentionis paid by governmentsand internationalorganizationsto theprovision of safesourcesof drinking water to the peoplein the ruralareas,the demandfor handpumpsin nearlyall developingcountrieswill increase
Thepotentialmarketin thesecountrieshasbeenestimatedat 20 million for this decade(Modern Asia 1982) However,althoughthereis an obvious demandfor handpumps,theinstallation rate of thesepumpsappearsto behinderedby the lack of a pump that can bemaintainedeasily at thevillage level Such apump, which is ideally suited for the ruralareas,is describedas a village-leveloperationand maintenance(VLOM) pump
The averagemaintenancecost for hand-pumps in EastAfrica is $4001 per pump peryear (World Water ic&z) In someinstances,themaintenancecostaccountsfor 85% of theamortized cost of installing a rural-watersupply Thus,althoughtheuseof handpumpsoffers a low-costalternativeto theprovisionof clean water in rural areas, the highincidenceof pump breakdownand theprob-lem of providing adequatemaintenancedeterthewider useof thesepumps in rural areas
The objectivesof this studywere threefold:first, to conduct a financial and economicanalysisof thecosteffectivenessof thepoly-vinyl chloride (PVC) handpumpsbeingtestedby the InternationalDevelopmentResearch
lAll costsare in US$ The exchangeratesusedhereare US$100 = $2 30 Malaysian,P8 50 Philippines,
523 00 Thailand, and R20.00 Sri Lanka
Centre (IDRC) network, as comparedwithother handpumpsin use in these countries;second, to analyze the sourcesof watersupply for rural areasand make projectionsregardingthepercentageof ruralhouseholdsto be servedby pipedwaterby theyear 1990,and, third, to undertakea preliminaryassess-ment of the market for handpumpsfor thenext10 yearsandassessthepotentialmarketfor the handpumpbeing testedby the net-work in thesefour countries.
Someof the information on thecostof thehandpumpswas obtained from the projectinterim reports Additional information wassecured using a questionnaire and fromdiscussionswith theproject leadersin eachofthefour countries Informationon thesecondandthird objectivesof this studywasgatheredfrom interviews with the project leaders,government officials, and other interestedindividuals from various organizations
One of the major limitations of this studywasthelimited time spentin Sri Lanka,Thai-land, and the Philippines Therefore,it waspossibleto obtain only a preliminary assess-ment of theoverall situation regardingruralwater supply in general, and the role ofhandpumpsin particular. Nevertheless,withthe active cooperation of each country’sproject team and information obtained frominterviews and published sources, it waspossibleto arrive at a fair assessmentof thesituation prevailing in thesefour countries
The cost information available for theIDRC-PVC pumps was for experimentalorindividually fabricated units. Thus, it wasnotmeaningful to use the data for comparisonwith other pumps produced on a large scale.It wasonly possiblein onecountry(Malaysia)to obtain theestimatedcost of thePVC pumpif it were to be producedin largequantitiesThus, a detailed financial analysiswas onlycarriedout for Malaysiabecauseof thelackofdata from the other countries
Sri Lanka
CountryAnalysis
Theproject teaminstalledabout21 pumps,primarily in the southernpart of the islandEachpump serveda cluster of four to fivehouseholdsor about30 people All the pumpsthat were installed are still functioning andthey are beingmaintainedby theSarvodaya
57
movement,which has severalworkshopsintheareaThesepumpswereinstalledin hand-dug, 4 9-foot (1 5-rn) diameter wells Thewells were lined with a concrete casing andcovered with a removable concrete cover,thus, if the pump breaks down, the villagerscan continueto obtain waterwith a bucket
Two factorsappearto influencethe instal-lation of the pumps in this manner First,becausethepumpsarestill experimentalandmay fail, this type of well offers someassuranceof aconstantwatersupply Second,becausevillagers are not familiar with themethodof drilling bore-holewells, therewasno local expertise available for trying thismethod of installation
The construction of open wells increasesthe installation cost andplacesa limit on thedepththat canbeobtainedThe averagedepthof 100 wells examinedwasabout 16—23 feet(5—7 m) andthestaticwater level was10—16feet (3—5 m) below the surface
The Sarvodayamovementplans to installan additional 250 handpumps in variousvillages The cost of installation will probablybe borne by foreign-basedaid agencies
Economicanalysis
Financial andtechnicalinformation regard-ing the threehandpumpmodelsareshowninTable 1 Thesepumps differ in their above-ground components, although the below-ground components are essentially the same
Installationchargesarenearly 300% of thecostof thepump,whereasyearlymaintenancechargesareabout23%of the total pumpcost
Water supplysituation
Of the 12 7 million people in Sri Lanka in1950, 73% lived in rural areas There were1 420000 rural households,each containingan averageof sevenpersons,andonly 2% ofthe rural populationwascurrently servedbytreated piped water
The useof handpumpsis a relatively newphenomenon,only about 2500 have beeninstalled in Sri Lanka Of these,only about1500 (60%) are still functioning If it isassumedthat one handpumpservesabout50families, then only 75000 families or about5% of the rural populationis currently beingserved
The United Nations Children’s Fund(UNICEF) is currently involved in a programto install a sizeable number of deep-wellhandpumpsin thenorthernhalf of theisland
Table 1 Information on threehandpumpmodels
testedin Sri Lanka
Model
Vi Li L3ItemInstallation ($) 283 269 241Pump ($) 60 100 120Yearly maintenance ($) 17 25 21Expected life (years) 7 9 9Persons served/pump 20 40 25Avg water use/person/day(L) 25 66 35
Water pumpingrate (L/minute) 8 8 5
Note Installationcostestimatedfor anaveragedepthof5 m (16 4 feet) Expectedlife isbasedon the projectleader’sestimates
~i L = 0 22 gallon
To date, 600 holeshave beendug and 300pumps installed Thesewells, which averagemore than 49 feet (15 m) deep, are situatedabout 0 5 mile (800 m) apart and each oneserves about 50 households The mostcommonpump is the India Mark II, whichcostsapproximately$400 per pump (excludingtax) UNICEF plans to have a two-tierprogramfor themaintenanceof thesepumpsa village pump mechanicto attendto minorrepairs and a regional pump inspectorresponsiblefor major repairs The truck-mounteddrilling rig required for this pro-gram wasimportedfrom Europeat a costofabout $250 000 A rig can drill up to about80 holes per year at an estimated cost of$1000-1500per well
Potential markets
The potentialmarketfor handpumpsin SriLanka is sizeable Theonly limitation appearsto be that 75%of thecountryis relativelydry,however, 70% of the population is concen-trated in the wet zone If 50% of the ruralpopulation were served by handpumps,theestimatednumber of additional handpumpsrequired in Sri Lanka would be approximately14000, 15000, and17000 in 1982, 1985, and1990, respectively These estimates arederived from the assumptionsthat the ruralpopulation is growing at an annualrate of2 5% and that the numberof families servedper pump remains at 50 If theobjectivewereto provide one handpumpper 10 families,theseestimateswould haveto beincreasedby500%
The government apparently places highpriority on handpumpdevelopmentfor therural areas of the country However, its
58
program is being curtailed by the lack of the30% counterpart financing that is a pre-requisite for obtaining “soft” loansfrom theWorld Bank (World Water 1982)
Only a limited numberof handpumpsexistat present, whether locally manufactured orimported The importedpumps come eitherfrom India or Bangladesh and, for shallowwells, cost about$100-250each In view oftheir limited usageandthe lack of experiencein their construction,there are no dataoneither their cost or reliability
Provided that funds are not a constraint,thepotentialdemandfor theIDRC-sponsoredPVC pumps is approximately20000 pumpsper yearbetweennow and1990 This is basedon theassumptionthat at leasthalf the ruralpopulationcanbe servedby shallowwellsandthat eachpump servesfive households
Thailand
After somelaboratory testing, a modifiedversion of the below-groundcomponentofthe IDRC-sponsoredPVC pump wasinstalledin 54 selectedwells in central, northeastern,and northern Thailand These wells werebeing used daily by villagers, and had beenfittedwith a Departmentof Mineral Resources(DMR) (Demsteror RedJacket)or anAcceler-atedRuralDevelopment(ARD) (modifiedKorat)pump The project teamretainedtheabove-ground componentsof these54 pumps andonly altered the below-groundcomponentsThus thepump that wastestedwasmadeupof a combination of plastic and cast-ironcomponents
Thesepumpswerefield testedfor 15 monthsand the results were generally satisfactoryThe majorproblemencounteredinitially wasthat of foot-valve leakage,especiallyin sandyareaswhere sandparticles lodged under theplate valve andcausedthe leakage.However,after some modifications to the design, thisproblem appearsto have beensolved
The Asian Institute of Technology(AlT)group also designeda new PVC pump thathad both above- and below-ground com-ponents of PVC, but only three of thesepumps were testedin the field The projectteam considersthat the IDRC-PVC hand-pump can be used for wells to a maximumdepthof 65 feet (20 m)
The drilled depth of the selected wellsrangedfrom 52 to 112 feet (16—34m) How-ever, the water level rangedfrom 4 9 to 52
feet (1 5—16 m) Thewaterlevel in mostwells
was less than 16 feet (5 m) in depth
Economicanalysis
The cost of the AlT-designedPVC hand-pump is comparableto the costof theDMRand ARD handpumps Excluding thecost ofthe riser pipe,which varieswith thedepthofthe well, thecostof theAlT-PVC handpumpis $135 as comparedwith $122 for theDMRand$139 for theARD handpumpsTheAlT-PVC handpumpcost is for individually fabri-catedmodels,whereasthecost for theDMRand ARD pumps is for mass production.Hence, it would be possible to reducetheAlT-PVC handpumpcost substantiallyif itwere producedon a large scale
Water supplysituation
Nearly 30% of Thailand’s population of38 million peoplelive in rural areasand60%of these rural inhabitants, or 2 6 millionhouseholds, do not have access to clean waterAlthough there is no definite figure on thenumber of handpumpsinstalledin Thailand,it hasbeenestimatedthat over 5 million peoplein the country depend on handpumpsforwater If it is assumedthat 200 peopleareserved by one pump on average, then anestimated 25000 handpumpsare currentlybeingused in Thailand Of this total, 19000handpumpshave been installed by variousgovernmentalagencies The major problemfacedby governmentagenciesis that, on anyparticularday, 25% of thesehandpumpswillbe out of service The averagecost of main-tenancefor eachof thesehandpumpsis $71per year
The main governmentagenciesinvolved inthe rural handpumpprogramare.Departmentof Mineral Resourcesin the Ministry ofIndustry, Department of Health in theMinistry of Public Health, andDepartmentofPublic Works and the Office of AcceleratedRural Development in the Ministry of theInterior
Eachof theseagencieshasadoptedits ownhandpumpdesignandthepumps areproducedlocally, i e, by local manufacturing firms whoare awardedgovernmentcontracts for suchpumps. Their design is essentiallysimilar tomodelsimportedfrom westerncountries,butthey incorporatesomeminor modifications
Theseagenciesarealso involved in drillingwells andinstalling handpumpsin the ruralareas To date, they have installed about
59
19000 handpumps,two-thirds of which arefor deep wells, more than 65 feet (20 m).About 10% of thesewells have beenfittedwith motorized pumps. In the last 2 years, theannualtarget hasbeen increased from 1000to at least 2000 wells per year
The Departmentof Mineral Resourcesusesa modified Demster pump that they areinstalling on 6-inch (15-cm) tube wells drilledto a depth of about 115 feet (35 m) This costsabout $3300 andincludesexpensesfor boththe drilling and thehandpump,which costs$150 The maintenanceand repair cost is$125 per pump per year Eachpump servesabout 250 peopleandprovideswaterthrough-out the year, whereas shallow wells oftenrundry TheDepartmentplansto increaseitspaceto drill 4500 wells annuallyto reachthetargetof a total of 20000 deepwells by 1990.
The AcceleratedRuralDevelopmentOfficeutilizes a modified Korat pump that costsabout $139 It differs from the original Koratpump in that it has a 3-inch (7 5-cm) cylindermadeof PVC anda 1 25-inch(3-cm)droppipeThesepumps are normally usedfor shallowwells, lessthan 50 feet (15 m) deep
The Department of Health’s Division ofRural Water Supply is the main agencyinvolved in the installation of pumps in ruralareasIt installs handpumpsfor both deepandshallowwells To date, it has installedabout600 deep-wellpumpsto drawwater from anaverage depth of 115 feet (35m) ThemodifiedKorat handpump that they use costs $150andits annualmaintenancecost is $130 Eachwell servesa groupof 40-50 householdsThedrilling andinstallation cost is approximately$1500
The Division of Rural Water Supply hasalso establishedabout1000 shallow wells incentral Thailand The pump usedis called theA-pump or Lucky pump and costs about $30-40 The averagedepthof thesewells is 33 feet(10 m) The averagedigging andinstallationcost is about $250, whereas the annual main-tenancecostis only $8 perpump TheDivisionplansto install between 800 and1000 of thesepumps per year for the next few years Lastyear, the Division, in cooperation withChulalongkornUniversity, wasinvolved in aproject to develop and test a PVChandpumpThis pump was manufacturedlocally and hasbeen installed in 100 shallow and 20 deepwells Thecostof thepumpis $50for shallowwells and $100 for deepwells The perfor-manceof thesepumpsis now beingmonitored.
Potential market
Thereare an estimated2 6 million house-holds in Thailandthat do not haveaccesstocleanwater for domesticconsumptionBasedon the assumption that the rural population isgrowing at a rate of 2 5% per year, by 1990,66000handpumpswould berequiredif hand-pumps were to be provided to this populationat a rate of one pump per 50 families If thetarget is to provideevery10 householdswitha pump, the potential demand would be330000 Thesefiguresdo not includereplace-ment pumps
There is generally a preferencefor deepwells in Thailand because this assures thatwater is available throughout the year Forsome parts of the country, particularly in thedrier areas, water from shallow wells is avail-able for only about 8 months of the yearAssuming that 50% of the handpumps thatwill be installed will be for shallow wells, thepotentialmarket for an IDRC-PVC pumpisabout 4500-18 000 units per year betweennow and1990 dependingupon thenumberoffamilies to be servedper pump (see Table 7)
Philippines
Thirty pumpsbasedon theWaterloodesignwere fabricated by a local contractor andinstalled in various locations on Luzon IslandNo laboratory testing wasconductedand thepump that was installed was essentially similarto the Waterloo type The above-groundcomponent,however, used a concretestandthat was relatively economical
Theaveragecasingdepthwas39 feet, range20—59 feet (12 m, range 6-18 m), and theaveragewaterdepthwas 13 feet, range3—26feet (4 m, range 1-8 m) Of the 30 pumpsinstalled,only 4 arestill functioning Theresthave been either abandonedor dismantledbecauseof foot-valve leakageor thecollapseof the wells
Economicanalysis
Thecostof theIDRC-PVC pumpwas$267Materials required for its fabrication consti-tuted 92% of this cost, while the fabricationexpenseitself represented8% of total costWell-drilling and pump-installationchargeswere relatively inexpensive,only $353 for anaveragedepthof 39 feet (12 m)
Onefactor leadingto thelower installationcost was the use of PVC pipe as the casing
60
cylinder’ PVC is one-third the cost of gal-vanizedpipe Another reasonfor the lowercost was the availability of local contractorsexperiencedin well drilling andinstallation ofhandpumps. Also, a small motorized rigcosting $4000 was used in well drillingoperations
Water supplysituation
In 1980, only 43%or 21 2 million peopleoutof a total population of 49 4 million wereservedby piped water supplies Of the ruralpopulation of 34 1 million people, 33%obtainedwater from public supply systems,whereasthe rest dependedon handpumps,openwells, rainwatercisterns, and streams
According to UNICEF statistics,thereare23 572 public artesian wells serving about4 million people Only 16000 of them areoperationalandtheir averagedrilled depthis197 feet (60 m) (World Water 1982) ThePhilippines has a large reserveof groundwateranda high averageannualprecipitation,89 inches (2260 mm)
The use of handpumps has a long history inthe Philippines In some areas, the majority ofhouseholds have installed their own hand-pumps Estimates of the numbers of ruralhouseholds of a total of 5 68 million obtainingwater by various means in 1980 are frompublic water supply, 1 87 million, from artesianwells, 0 39 million, and from privately ownedhandpumps,0,30 million
Theseestimatesarebasedon two assump-tions that an averagehouseholdincludessixpersons, and that there are 30000 privatelyowned shallow wells each serving 10 house-holds Basedon these estimates,over threemillion households are still without cleanwater for domesticpurposes
To overcome this problem, the Philippinesgovernmenthaslauncheda 20-yearprogramto provide safe and accessiblewater to allhouseholdsThe two main agenciesinvolvedin this rural water supply program are theRural Waterworks DevelopmentCorporation(RWDC) and the Ministry of Public Works(MPW)
Their programis centredon the formationof beneficiary committees into self-reliantwater-supply associations or cooperativesTheseassociationsare requiredto contributeto thecapitalcostandundertaketheoperationandmaintenanceof the water-supplysystemThe governmentagenciesprovide the tech-nical and institutional assistanceas well as
contribute toward the bulk of the cost,including 10% of the operatingand main-tenanceexpenses
Threelevelsof serviceshavebeenproposed,dependingon thepopulationsizeof thearea,the sourceof watersupply, thedevelopmentcost, and the community’s ability to pay Thegovernmentwill provide 90% of the capitalcostfor Level 1 andloanscovering90%of thecapital costare available for Levels 2 and3
The main emphasisfor Level 1 is to developpoint sources such as artesian wells andprotected springs Each shallow well isdesigned to serve a cluster of 15 households,whereas each deep well will cater to 50 house-holds The average installation cost perhousehold is $12 33 (or $i&5 per well) forshallowwells and$98 82 (or $4941per well)for deepwells The annualmaintenancecostis estimatedat $o 82 per household
Level 2 is essentiallythe same as Level 1but includesa system of communalfaucets.Its overall designis for aclusterof 100 house-holds, and the cost per householdis $71,excluding the cost of sourcedevelopment
Level 3 provides for individual houseconnections and the overall design is forurbanhouseholds The capitalcostper house-hold is $247
The Level 1 program will focus almostexclusively on the construction and rehabili-tation of shallow anddeepwells The averagedepth of shallow wells is 30 feet (9 m), whereasthe deepwells average200 feet (60 m) By1990, theRWDC andtheMPW plan to installa total of 169 000 shallow wells and 87500deepwellsthroughoutthecountry In addition,a total of 26000existingwellswill berehabili-tated
Potential markets
An indication of thepotential market for alow-cost efficient PVC pump can be obtainedby examiningtheMPW andtheRWDC targetsfor well construction(Table 2)
The cost of financing the program isestimatedat $368 million. It is expectedthatpartof this amountwill beobtainedfrom loanor aid programs of various internationalagencies. Failure to secure adequatefundsmay delay the implementation of thisprogram
The rural population in the Philippines isexpectedto rise to 43 1 million peopleor 7 2million households by 1990 Assuming thatall theproposedpumpsareinstalledby 1990,
61
Table 2 Well constructionand rehabilitationtargets
andforthe
the Ministry of Public Works (MPW)Rural Waterworks DevelopmentCorporation(RWDC)
YearShallow Deep Rehabili-
wells wells tation
1980 10000 1691 12741981 13000 5000 20001982 45000 13000 25001983 40000 14000 25001984 41000 15000 250019851986—1990
16000 15000 2500
4000 23800 12500
Source Governmentof Philippines 1980 Integratedwater supply program, 1980—2000, and RWDC — ThePhilippines rural water supply program
the country will have approximately 200000shallow wells and 110000 deep wells If all theshallow wells and 70% of the deepwells arelocatedin the rural areas,eachshallowwellwill be usedby about 17 householdsandeachdeepwell will cater to about50 householdsThus, for a target of one shallow well foreveryfive households,therewill bea needforan additional 480000 shallow wells by 1990
Malaysia
After extensive laboratory investigations,nine suction, two pressure-suction, onepressure-lift, and five lift pumps werefabricatedandinstalledin two rural areas—
Kuala Pilah andIpoh Thesepumpshavebeenfield tested for over 1 year, and no majorproblemshavebeenencounteredso far Allof them are still functioning, and from pre-liminary observations, it appears that thevillagersaresatisfiedwith theirperformanceThe majordistinctive featuresof thesepumpsinclude a PVC cylinder fitted with a slidingPVC piston anda stationary,but removable,PVC foot valve, and a leverage system con-sisting of timber linkagesandgalvanizediron/oil-impregnatedtimber bearings
These PVC handpumpswere installed inexisting wells thathadhandpumpsthat werecondemned from further use. The wellsaveraged30 feet (9 m) deepand the waterdepthwas usually 10—23 feet (3—7 m) belowthe ground Eachhandpumpserveda clusterof four to five households The pressure-suction and pressure-lift pumps, however,hadindividual householdconnectionsto eachof the four or five householdsThus, everyuser could pump the water directly into awater tank installed at their own house
Economicanalysis
A financial analysis was conducted tocomparethe cost of the PVC pump (bothexperimentalcost and projectedproductioncost)with thecost of thetwo existingpumps(Gibsonand Fuji)
The resultsof this analysisshowedthatthepresentworth andequivalentcostof thePVCpump were lower than the existing pumps(Table 3) This differencewas especiallysigni-ficant when the cost of thePVC productionmodel was comparedwith the cost of theexisting handpumps Using the present-worth and annual-equivalent-costconcepts,the cost of the PVC productionmodel was
Table 3 Summaryof various measuresof pumpcost (US$)
Type ofpump
Capitalcost
Presentworth
Annualequivalentcost
Gibson 31 145.11 79.74Fuji 61 155 45 28 40IDRCExperimental 134 145 72 27 33Production 74 85 72 15 99
Note The analysis for present worth and annualequivalent cost is basedon the format prepared byI Majumdar and discussedat the mid-projectmeeting inAugust 1980 (Goh 1980)
Table 4 Basic information on handpumps(1980constantdollars)
IDRC models’Experi- Produc-
Item Gibson Fuji mental tionInstallation cost
5
Material 85 85 85 85Labour 49 49 49 49
Transportcosta 1 1 1 1Pump and coupling 30 60 133 73Installed capitalcost 31Annual repair cost’~ 9
61
9134
274
2Period of operation(years) 0 5—2 5 3—5 7-9 7-9
EconomicliFe(years) 2 4 8 8Rateof discount(%) 10 10 10 10Salvagevalueatendofyear 0 0 0 0
‘A breakdown of the major cost componentsof bothmodelsis given in Table 5
bF a 30-foot (9-rn) well, the costsare auger,$18, Labour(14 days at $3 50 perday), $49, cement,$10, and casingand pipe, $57 These costsare the samefor all pumps~Assuming a distanceof 62 miles (100km) from port and
that the pumpsare transportedin bulk The total fixedcosts,including installation, are Gibson,$165,Fuji, $195,IDRC experimental,$268, and IDRC production,$208dRepaircost is for partsonly Assumedthat labourispro-
vided by either the user or the government
62
about50% of the costof the existingpumpsThe basic data available in the Malaysian
case are shown in Table 4 Several pointsshould be noted
Only the IDRC-University of Malaya(UM) experimental pump could beaccurately assessedin terms of cost(Table 5) This is likely to appearhigh inview of the limited number of pumpsproduced and the research elementinvolved in thedesignandconstructionThe mass-productioncost of the IDRC-UM model can only be estimated atpresent and is based on quotationsobtainedfrom local plastic manufacturersregarding bulk orders of the differentpump components.By experimentingwith different manufacturingprocesses,it may be possible to find ways toreducetheseprices
• Data on other pumps (GibsonandFuji)are obtained from the Ministry ofHealthrecordsandfrom field interviews
• The life recorded in the field for non-IDRC pumpsis extremelyshort, rangingfrom 6 months to 5 years
• All 12 suction pumpsinstalledin thefieldin the past2 yearsare still in operationHence, thereareno figuresavailableonthe actual lifespan of thesepumps An
Table5 Major costcompand production version
onents($)of e
of IDRC-UM
xperirnentalhandpump
Component ExperimentalProduction
Piston and foot valve 59 00 8 40
Spout 4 10 4 10Piston cylinder 20 20 8 70
Drop pipe and piston-rod assembly 9 80 9 80
Metal stand 17 40 17 40Leverageassembly 16 40 16 40Bolts and washers 6 10 8 20
Total cost 133 00 73 00
estimate of 7—9 years is used for thepurposeof this analysis
• The nonusablepumpsarekept for spareparts Thus there is no definite salvagevalue for the pumps Therefore, thesalvagevalueis assumedto bezeroat theend of the economiclife of the pump
• Cost data obtainedfrom 1978 to 1981showed no significant increase in costdue to inflation These cost figures areassumed to be in 1980 constantdollars
Water supply5ltuation
Only 43% of rural householdsin Malaysiaareservedwith pipedwater (Table 6) How-ever, this figure will rise to 58% by 1985 ifcurrentplansundertheFourthMalaysiaPlanareimplemented In terms of thenumberofrural households,this meansthat a total of994000 households in 1980 and 833000householdsin 1985 will still have to rely ontraditional sources for their daily waterrequirements
The Ministry of Health has,since the late1960s, supplied haridpumps to a limitednumber of rural households At present,atotal of about5500 wells servingabout22000households have been constructed Thisfigure is, however, only about 50% of thetarget set in the Third Malaysia Plan Oneofthe main reasonsfor this shortfall is thedifficulty in obtaininghandpumpsAt present,all the handpumpsused by theMinistry ofHealth have to be imported these includepumpssuch as the Dragon,Fuji, Gibson, andSGP These pumps are relatively cheapbutthe experienceof the Ministry is that theyrarely last for more than 1 year Also, thereis a lack of sparepartswheneverthepumpsbreakdown
The program under the Fourth MalaysiaPlan is to increasethenumberof handpumpsinstalled by 12382 to serve about 60000
Table 6 Number (‘000) and percentageof rural householdsservedby piped water’
ruralNo ofhouseholds
No servedwith piped water
% served withpiped water
1980 1985 1990 1980 1985 1990 1980 1985 1990
PeninsularMalaysiaSabahSarawak
1485102164
1656 1801159 175189 208
698 1043 138818 62 10641 66 91
47 6318 3925 35
776144
Total 1751 2004 2184 757 1171 1585 43 58 73
Source Governmentof Malaysia 1981 Fourth Malaysia Planlt is assumedthat rural householdswill constitute60%of populationin 1990ascomparedto 38% in 1985. and that the
numberof rural householdsthat w,ii be servedwith pipedwater for 1985—1990 will be thesameas that for 1980—1985,i e, 414 000 households
63
householdsby the end of 1985 However,becauseof the delays in obtaining pumpsfrom overseas,there is somedoubt that thistarget canbe achieved The governmentofMalaysiaplansto investlargesumsof moneyto supply pipedwaterto over70%of theruralpopulationby the year 1990 However, evenwith this ambitious program,about600000householdswill continueto dependon othersources of water supply in 1990
Potential markets
The Ministry of Healthis themain govern-ment agencyinvolvedin installing handpumpsin rural areas The entire costof drilling thewell andinstalling the handpump(includingthe cost of thepump itself) is borne by thegovernment To date, the governmenthasalsoprovideda free repair andmaintenanceservicefor thebulk of thehandpumpsinstalledThis policy may be discontinuedin thenearfutureanduserswill be requiredto undertakemaintenance,repair, and rehabilitation oftheir own handpumpsThecurrentpracticeisto supply one well with a handpumpto aclusterof four to six householdsThesewellsare, on the whole, shallow, with an averagedepthof 16—50 feet (5—15 m)
The governmenttarget is to install about2500 handpumpsper year to cater to about12500 householdsHowever, the numberofrural households without accessto pipedwateris currently about1 million This figureis expectedto fall to about600000by 1990 Ifit is assumedthatatotal of 400000 householdswill still haveto dependon waterfrom hand-pumps in 1990, there will be a needfor anadditional 80000 handpumpsto beestablishedthroughout the country between now andtheyear1990 This implies apotentialmarketof approximately10000handpumpsper yearfor the rest of this decade
Main Findingsand Conclusion
Costof handpumps
The cost of the IDRC-PVC handpumprangedfrom $93 to $267 in thefour countriesThis cost wasderivedfrom experimentalcostfigures andit is likely that it could be reducedto one-thirdor evento one-halfof thepresentcost if the pump were to be producedon alarge commercial scale by mass-production
techniques It is significant that the experi-mental pump costs only 17% more thancomparableexisting handpumps
The cost factor does not constitute themajor constraint to the IDRC-PVC hand-pump gaining wider acceptancein the fourcountriesIn ThailandandthePhilippines,forexample,it wasfoundthat technicalproblemsremain to be solved before thepump can beregardedas being sufficiently reliable forregular use This, however, illustrates theimportance of field testing to thoroughlyunderstandthe technology.The situation inSri Lanka and Malaysia is more favourableand the pumps tested in these countriescould, with some modifications, form thebasis for commercial productionand wide-spread utilization It is interesting to notethat, in both of these countries, technicalproblemswere initially encountered,however,thesewere solved during the course of thefield testing The Thailand and Philippinesprojects did not go through this sameextensivetesting process.
In the Philippines,it is unrealistic to do acost-effectivenesscomparisonbetween theIDRC-PVC pumps and the other existinghandpumpsbecausethe IDRC-PVC pumpsstill possesssomeseriousdesign faults thatrequireremediesIn thecaseof Sri Lanka,theIDRC-PVC handpumpsarefunctioningsatis-factorily, but there is an absenceof othercomparablepumps Hence,in this case,dataare not available for comparisonAlthoughThailand now appears to have solved theproblemof leaky foot valves,the field-testingprogram was incomplete at the time theproject terminated Only in Malaysia, was itpossible to comparethe IDRC-PVC pumpwith other handpumps
In this comparison,the IDRC-PVC hand-pumps emergedquite favourably comparedwith other handpumpsUsing the estimatedcost of the IDRC-PVC handpump(as if it
were produced on a large scale), it can beshown that thepump is only abouthalf thecost of theexisting handpumpsAlthough, inthe Malaysian case,conversionfactors wereavailable, an economic analysis was notundertakenbecausethere is no governmenttax on imported pumps but the tax onimported PVC ranged from 10 to 25% Inview of this, the results of the economicanalysiswill not differ significantly from thatof the financial analysis
In all thesefour countries, thedemandforhandpumpsis high, especiallyfor thosesuit-
64
Table 7 Summary of important handpump statistics in the four countries
Item
Cost of IDRC pumpInstallation chargesPump plus installation costIDRCExisting pumps
Rural householdswithout cleanwater (million)
Existing handpumps(1982 figures)
Potential market annually(1982-1990)1 shallowwell/20 households1 shallow well/5 households
able for drawing water from both deepandshallowwells The IDRC handpumpis bettersuited for wells of up to 50 feet (15 m) indepthor for shallowwells It is estimatedthatthereis acombinedannualmarketfor 29 500-118 000handpumpsfor shallowwells in thesefour countries The biggest marketis in thePhilippines (Table 7)
Watersuppib’situation
Both ThailandandthePhilippines havehadlong-termexperiencein theuseof handpumpsThus their rural water projectsplace greatimportance on well drilling and handpumpinstallation In both these countries, thehandpump projects are heavily funded byinternationalaid or loan agencies
On theother hand, theuseof handpumpsappearsto be a recent developmentin SriLanka andMalaysia However, the progressof the rural water program in Sri Lanka,including the handpumpprogram, is some-what constrainedby the lack of governmentfunds In contrast, the Malaysian programtends to be directed more toward supplyingrural householdswith piped water, ratherthan toward the substantialincreasein thenumber of handpumps However, this isextremelyexpensiveand may changein thefuture
Shallowand deepwells
The demand is for handpumpsthat candraw water from both shallow and deep
wells The distinctionbetweena shallowanda deepwell is ambiguous,but it is generallyacceptedthat a shallowwell is lessthan50—66feet (15—20m) in depth this is thedrill depthThe depthof thewatertable,however,variesconsiderablybetweenwells Thus,evenin thecaseof a deepwell of 100 feet (30 m), thewater level may be only 16-32 feet (5—10 m)below the surface In this case, a pumpdesignedfor shallow wells can still be used
Water tables fluctuate dependingon thepatternof rainfall prevailing in eachregionThus, someshallowwellsmayrundryduringperiods of drought In the case of deeperwells, thewatertablemay fall considerablysothat it is no longer possibleto draw waterfrom the well with a pump that is designedfor shallow depths
All the pumps testedin the IDRC programare essentiallymeantfor shallow wells, thatis, for wells with a drilled depthof 50-66feet(15—20 m) Thus, in assessingthe potentialmarketsfor the IDRC-PVC handpump,it ismoreappropriateto consideronly themarketfor shallowwells insteadof for both deepandshallow wells Future work on the IDRC-PVC pumpmayextendtherangeof thepumpand, in this event,its potentialmarket will beexpanded
Acknowledgments Thesupportandassistanceofthe International DevelopmentResearchCentre(IDRC) are gratefully acknowledgedThe projectteams in the four countries cooperatedwillingly insupplying all the basic information used in thepreparation of this study The project leaders,PathiranaDharmadasaof Sri Lanka,Pisidhi Kara-
Averageor
Malaysia Philippines Sri Lanka Thailand Total’
134134
267165
93
476135225
157275
268180
532450
569651
360356
432409
099 312 132 260 803
6000 55000 2500 25000 88500
20008000
1800072000
5000
20000
450018000
29500118000
‘All cost figures are averagesin US$~‘Thesameinstallationchargesareassumedfor the IDRC and existingpumps Thiscostis for a well with a depthof 9 m
(29 5 feet)
65
sudhiof Thailand,Antonio Bravoof thePhilippines,andGoh Sing ‘Yau of Malaysia, gaveguidanceandinsightson the overallhandpumpsituationin theirrespective countries Their assistancein theseareasis acknowledgedThanksare alsodue to thevarious officials in the four countrieswho con-sentedto being interviewed regardingtherelevantprogramof their institutions
References
Goh, S Y 1980 Theperformancecharacteristicsof
a reciprocating piston water lift pump Ottawa,Ont, CanadaInternationalDevelopmentResearchCentre Interim progressreport,Waterpumpingtechnology— Global project
IRC (International ReferenceCentre for Com-munity Water Supply) 1977 Handpumps TheHague,NetherlandsIRC TechnicalPaperNo 10
Modern Asia 1982 Wanted 20 million pumpsforthe Third World Washington,DC, USA Apnlissue, p 29
World Water 1982 1981—1990 decadeLiverpool,U K ThomasTelfield Ltd pp 11-15, 47—50
66
Conclusions
In an era of rapid developmentand populationexpansion,the challengetoimprove environmentalhealth confrontseverynation. In the Third World, thisproblemis mademoreacuteby limited resourcesProvisionof adequate,safewatersupplies to rural populations by 1990, an official target of the InternationalDrinking WaterSupplyandSanitationDecade,meansthatanestimated20 millionor more new handpumpsmay be neededby theyear 2000if thegoal of bringingpotable water to themillions of rural inhabitantsof theWorld is to be achieved.Thesepumpsmustbeableto withstandtheuseandabuseof themanywho dependupon their proper functioning for their daily water supply
Whena handpumpbreaksdownandremainsout of service,theeconomicloss isconsiderable The replacementparts,and the possibility of vandalismand dis-appearanceof parts if thepumpis out of operationfor morethana few days,resultin considerablecostandloss of financial investment,not to mentionthehardshipandinconvenienceto thosewho haveto walk long distancesto obtainwater. Onesolution to this problemis to focuseffortson thedevelopmentof locally fabricatedhandpumpsthatareinexpensiveto manufactureandcan be easilyrepairedat thevillage level with a minimum amount of expertise.
TheWaterloodesign,developedin 1976,doesjust this. The pistonandfoot valveareproducedfrom polyvinyl chloride (PVC), amaterial that is readilyavailableinmostdevelopingcountries.Their designis suchthat thepiston andfoot valve areinterchangeable,i e, the piston can be used as a foot valve andvice versa.
This greatlyreducesthenumberof sparepartsneededfor repairor replacementpurposes.Another advantageof theWaterloodesign is that it incorporatespoly-ethylenepiston rings,similar in designto thosein anautomobileengine.Thesecanbereplacedeasilywhen worn Finally, thedesigntakesadvantageof aPVC pipeasthe riserpipe andthe cylinder section,the place where the piston slides up anddown, is the riserpipeitself If this sectionbecomesworn, thepistoncansimply bemovedto a new position in the riserpipe A smallerdiameterPVC pipeis usedforthepiston rod The above-groundcomponentsareof local design,utilizing locallyavailablematerials Thesedesignsvary from thedirect-actiontype demonstratedby the Sri Lankan project to more complex steel lever-actionarrangementsdemonstratedby theThai project Inexpensiveconcretepedestalswereusedby thePhilippine project, a conceptthat deservesfurther investigation
The technology developedand testedwith support from the InternationalDevelopmentResearchCentre (IDRC) through theseprojectsclearly indicatesthatno universaldesignwill function adequatelyunderall conditionswith all usergroups.Moreover, it was not the intent of this researchto find sucha designAlthough thebasicpnncipleof thepumpremainedthesamewith all four projects,there were individual variations and modifications The resultsof this researchhave brought to light the fact that this technology,or any other handpumptechnology,must first be testedunder local conditionsandmodified accordingtothe needsand opinions of the user group, environmentalconditions,availablematerials, and level of expertiseof thoseexpectedto adopt it andmaintain itWithout this testing, the technologycannotbe expectedto meettheneedsof thetargetgroupand will most probablyfail
67
Oncesuccessfullyfield-tested,theremust be a concertedeffort to sensitizeandeducateall usersnot only on how thepump works but also on its limitations Ahandpump is a system with several components If one of those componentsmalfunctions,theentiresystembreaksdownandwaterno longercanbe deliveredTherefore, a thorough understandingof suchfacts as how this pump functions,what can go wrong, and what componentswear out the fastest, is essentialforadequatemaintenanceThis understandingis alsoessentialfor preventivemain-tenance,anaspectof handpumptechnologythatunfortunatelyhasbeenneglected
However, this workshop on village-level handpump technology should not beconsideredthecompletionof a network of researchprojects It is thebeginningof anew phasethatwill now seekto bring this conceptto thepeoplewhoneedit most,the rural poor The conceptof using inexpensiveplastics as pump componentshasbeen successfullydemonstrated.However, large-scalecommercialproduction bymeansof injection moulding hasyet to be investigatedThe feasibility of small-scaleproductionthroughcottageindustriesmustalso be fully examined Researchmustcontinueon theuseof new materialsandmodified designsandvariousoptionsforlow-cost above-groundcomponentsmust also be tried andtested More impor-tantly, however,if this technologyis to be appliedat thevillage level,efforts mustbe focusedon obtaining feedbackfrom the “endusers,” thevillagersthemselvesSociologicalandeconomicstudiesmust be carriedout in all thoselocationswherethe pump is to be installed,and a scientific- approachmust be used to developcommonmethodologiesfor thesestudies In addition,training programsfor ruralpeoplemustbe implementedthat aresupportedby thedevelopmentandtestingoflearning materialssuitablefor use by village-level workers Finally, appropriateinfrastructuresmust be establishedor strengthenedandmanagementtechniquesmust be developedthat arenot only gearedtoward institutionalizing the conceptof village-level operation and maintenance(VLOM) but also are capableof pro-viding follow-up through continuous monitoring services and educationalprogramsfor users
Mass-productiontechniquesshould substantiallyreducethe cost of this PVCpump; however,this hasyet to be tested Dr Goh Sing Yau of the University of
The Waterloopiston (left) snd foot valve (right) are inoulded from solid PVC, Exceptfor the piston rings andthe rubber soot on the foot volvo, they ore interchangeable
68
Malayaproposesto addressthis questionaswell as someof theothersmentionedabove His idea, which is intended to bridge the gap between developmentalresearchandcommercialization,is to investigatevariousmanufacturingprocessesin detail by developinga small-scalefabrication unit. The project would alsoseekto.
• thoroughly understandthe manufacturing processesand the actual costsinvolved in producing eachcomponent,
• developthe necessaryexpertiserequired to consultwith manufacturingunitson production procedures,
• conduct cost assessmentsof various manufacturing options, for example,subcontractingversusmanufactureat point of assembly,
• establishquality control guidelinesand standards,• field-test (utilizing Ministry of Health personnel)mass-producedmodels of
the pump and evaluatetheir technicalperformance,• develop appropriate manuals for transferring the technologies to other
interestedgroups,and• support complementaryprojects by providing prototypes, training, and
researchon solving any problems that may occurIt is expectedthat this project would ultimately result in theestablishmentof a
researchandtraining centrethatcould bethe focalpoint of a network of projectsaimed at investigating such concepts as village-level maintenanceschemes,community financingschemes,communityacceptancestrategies(socialmarketing),and the various options for manufactureand assembly
The discussionsduringthis workshoprevealedthat this PVC pumpmay be theanswer for many thousandsof rural communities for many years to comeHowever, it is only oneof manytechnicalchoices,all of which havetheir placeinthe long list of options In somecommunitiesandcountries,the PVC pump mayserve as an interim technology,until somethingbetter comesalong In othercommunities, due to varying social, economic, and environmental conditions, it
may not be acceptableat all In still other communities,a more sophisticatedlevelof technologymay be more suitable
For themanymillions of theworld’s ruralpopulationwho do not haveanoption,this technologyis a beginning,acontribution to the targetof cleanwaterfor all by1990 But, the future of the Waterloo designnow dependsupon the interest ofresearchersin investigatingtheproblemsof implementation In this ageof limitedresources,it is becoming increasingly clear that the future of handpumptech-nology lies with thevillagersthemselvesStill, onequestionremains how canthistechnologyandthedesireto maintain it bestbe transferredto thosewho need it
most?
ResearchNeeds
The following researchpriorities were identified during the workshop On thebehaviouralor “software” side, researchshould be conductedon
• developmentof methodologiesdesignedto promotecommunity acceptance,• developmentand implementationof various maintenanceschemes,• developmentand testing of community financing and self-help schemes,• developmentof instructional packagesdesignedfor village-leveluse,• investigation of water-use behavioural patterns and the developmentof
health educationprogramsdesignedto change those behaviouralpatternswhere necessary;and
• developmentof trainingprogramson themanagementof waterresourcesandthedevelopmentof informationsystemsdesignedto monitor thoseresources
On the technologicalor “hardware” side, there is a needfor researchon• developmentof appropriate,inexpensive,well-drilling technologies,• developmentof simplified, inexpensive,water-explorationequipment,
69
• adaptationandtestingof theWaterloodesignfor usein wellsgreaterthan165feet (50 m) deep,
• developmentand institutionalization of a classification system for watersourcesthat incorporatesnot only water quality criteria but also sanitaryprotection,construction,and stateof repair;
• investigationof the useof new materialsandother typesof plasticssuchasacrylonitrile/butadine/styrene(ABS);
• studieson the performanceof various above-groundconfigurations;and• studieson how waterbecomescontaminatedfrom thewell to thehomeand
how behaviourcan be changed(with the help of technologicalinterventions)to reducethe risk of contamination.
70
Participants
ZAINUDDIN ARSHAD. Ministry of Health, 3rd Floor, Block E, Jalan Dungun, KualaLumpur, Malaysia
ANTONIO BRAVO, Institute for Small-ScaleIndustries, University of the Philippines,EnriqueT VirataHall, Emiliojacinto Street,UP Campus,Diliman, QuezonCity,Philippines.
CHAN BOON TEIK, Director, Water Supply Division, Department of Public Works,Ministry of Public Works and Utilities, JalanDato Onn, Kuala Lumpur, Malaysia
CHEE KIM MENG, PlantManager,JohnsonandJohnsonSdnBdh, JalanTandang,PetalingJaya, Selangor,Malaysia
CHETFAN KARNKAEW, Chief, Rural Water Supply Division, Department of Health,Ministry of Public Health, DevavesmPalace,SamsenRoad,Bangkok2, Thailand
CHO JOY LEONG, FederalLand DevelopmentAuthority (FELDA), JalanGurney,KualaLumpur, Malaysia.
CHONG KAH LIN, Departmentof MechanicalEngineering, Faculty of Engineering,University of Malaya, Pantai Valley, Kuala Lumpur 22-11, Malaysia
PATHIRANA DHARMADASA, LankaJathikaSarvodayaShramadanaSangamaya(Inc), 77deSoysa Road,Moratuwa, Sri Lanka
ERNESTO D GARILAO, ExecutiveDirector, Philippines Business for Social Progress,4th Floor, Yutiro Building, 270 DasmarinasBinondo, Manila, Philippines
GOH SING YAU, Departmentof MechanicalEngineering,Faculty of Engineering,Univer-sity of Malaya, PantaiValley, Kuala Lumpur 22-11, Malaysia
MICHAEL GRAHAM, Regional Liaison Officer, CommunicationsDivision, InternationalDevelopmentResearchCentre,Tanglin P 0 Box 101, Singapore9124
ZAINAL HAIl HASHIM, RISDA, Risda Building, Jalan Ampang, Kuala Lumpur, MalaysiaRUS ISMAIL, State RISDA Office, Jalan Pengkalan Rama, Malacca, MalaysiaTIM JOURNEY, The World Bank, 222 New EskatonRoad,Dhaka,BangladeshLEE KAM WING, Program Officer, Health SciencesDivision, International Development
ResearchCentre, Tanglin P 0 Box 101, Singapore9124LEE KWOK MENG, FederalLand ConsolidationandRehabilitationAuthority (FELCRA),
P 0 Box 2254, Kuala Lumpur, MalaysiaLUM WENG KEE, SeniorPublic HealthEngineer,Ministry of Health, 3rd Floor, Block E,
Jalan Dungun, Kuala Lumpur, MalaysiaPETERLUTTIK, ProgrammeOfficer, UNDP,P0 Box 2544,Kuala Lumpur11-04,MalaysiaUZIR A MALIK, Department of Agriculture andResourceEconomics,Universiti Kebang-
saanMalaysia, Bangi, Selangor,MalaysiaZULAZMI MAMOY, Proyek Pedesaan,Universitas Indonesia,SalembaRaya 4, Jakarta,
IndonesiaCHARLES NAKAU, Appropriate TechnologyDevelopment Institute, P 0 Box 793, Lae,
PapuaNew GuineaPICHAI NIMITYONGSKUL, Division of Structural Engineeringand Construction,Asian
Institute of Technology,P0 Box 2754, Bangkok, ThailandPRASERT SAISITHI, Director, Institute of Food Researchand Product Development,
KasetsartUniversity, P 0 Box 4-170,Bangkok4, ThailandKAMAROL ZAMAN ABDOL RAHMAN, Department of Agricultural and Resource
Economics,Universiti KebangsaanMalaysia, Bangi, Selangor,MalaysiaSAW KIM HOCK, RISDA, Risda Building, Jalan Ampang, Kuala Lumpur, Malaysia
DONALD SHARP, ProgramOfficer, HealthSciencesDivision, InternationalDevelopmentResearchCentre, P0 Box 8500, Ottawa, Ontario,CanadaK1G 3H9
C J A STAMBO, Chief Engineer,Ground Water, National WaterSupply and DrainageBoard, Galle Road,Ratmalana,Colombo, Sri Lanka
WILLY SUWITO, Engineer,Directorateof SanitaryEngineering,4th Floor, Main Building,
Jalan Pattimura 20, KebayoranBaru, Jakarta,Indonesia
71
TAN BOCK THIAM, Facultyof EconomicsandAdministration,Universityof Malaya,PantaiValley, Kuala Lumpur 22-11, Malaysia
TEE TIAM TING, Departmentof MechanicalEngineering,Faculty of Engineering,Univer-sity of Malaya, Pantai Valley, Kuala Lumpur 22-11, Malaysia
TEO BENG HOE, Departmentof MechanicalEngineering,Facultyof Engineering,Univer-sity of Malaya, Pantai Valley, Kuala Lumpur 22-11, Malaysia
CECILIA C VERZOSA, PIACT/PATH, MCC P 0 Box 189, Makati 3117, Metro Manila,Philippines
DONALD WAUGH, Office of the Comptroller General and Treasurer,InternationalDevelopmentResearchCentre,P 0 Box 8500, Ottawa, Ontario, CanadaK1G 3H9
WIMUT KASEMSUP, Head, Water Resource Unit, Community Based Appropriate Tech-nology and DevelopmentServices (CBATDS), Population and Community DevelopmentAssociation, 8 Sukhumvit 12, Bangkok 11, Thailand
K M YAO, Water Quality ManagementAdvisor, PEPASJWHO,P.O Box 2550, KualaLumpur, Malaysia
CESAR E YNIGUEZ, Manager,EngineeringDepartment,Rural WaterworksDevelopmentCorporation, Vibal Building, 865-E de los SantosAvenue,Dilirrian, QuezonCity, Philippines
AHMAD ZAINI BIN MAT YUSOP, Ministry of NationalandRural Development,BangunanBank Raayat, 1st Floor, Jalan Tangsi, Kuala Lumpur, Malaysia
72