Figure 1. Consumption of drinking-water by households in differentseasons.

village. A sampling unit, in this case asingle farm household, was selectedrandomly from each para to producethe study sample. Finally, within eachselected household, the family leaderwas chosen to be the interviewee. Keyindividuals included in the interviewswere the village school teacher, mem-bers of village committees, commonlabourers, and different categories offarmers and non-farmers. A total of36 farm households were interviewedfrom 12 different selected villages.

Figure I shows that during the rainyseason 35.5 per cent of householdsprefer rainwater, while during the dry

Dry seasonHand lUbewlIiI _

en CanaUdllcn watere;DPHE·UNICEFsand httered _.

waler

Sources of dnnklng walero

10%

Rainy seaSOn40% 0 Pond waler

fZl Rain waterrn River waler

The objectives of the study were:o To study the present systems of

rainwater use and storage.o To study the socio-cultural and

economic aspects of water use ofboth rainwater and other locallyavailable sources.

o To perform a comparative costanalysis between rainwater use andthe use of water from other avail-able sources.The study considered the general

condition of the water sources, storagepractices, water quality control meth-ods, and the organization involved inimproving drinking-water supply inthe area. In general, the inhabitants ofDacope upozilla use water from ponds,canals or ditches, rivers, tubewells, theDPHE-UNICEF sand-filter facilities,and rainwater. After a series of discus-sions with a sociologist, an engineerand an economist, an appropriate ques-tionnaire was prepared to collect pri-mary data. The villages in the studyarea usually consist of two or threeparas. A para is a local name for adistinct cluster of households within a

Rainwater harvesting and storagetechniques from Bangladeshby M.D. Hussain and A.T.M. ZiauddinIt seems a cruel irony that a country with somuch rainfall has to resort so often tocontaminated surface water for cooking anddrinking. There are ways to harvest the cleanrain, and to store it for use during the dry season,as many Bangledeshis are already doing. Herethey share some ideas and suggestions.

EVEN THOUGH Bangladesh has avery high rainfall from April to Octo-ber, much of the water harvested andconsumed, especially in the rural areas,is contaminated. About 85 per cent ofthe total population of the country haveno access to clean drinking-water andadequate sanitation facilities. Theproblem is very serious in the southerncoastal area (where 18 per cent of thepopulation live), as both surface andgroundwater are very saline. The prob-lem intensifies during the dry seasonand during floods. Although peopleliving in coastal and hilly Bangladeshhave been collecting and using rain-water since time immemorial, no spe-cific study on rainwater use in theregion was available. An investigationwas therefore conducted into the socio-economic aspects of, and the scope forthe development of, the rainwatercatchment facilities in the country. Acase study was conducted in twelvedifferent villages of the Dacopeupozilla of Khulna in Bangladeshduring the months of October andNovember.

Table 1: Present situation of drinking-water facilities in the villages of Dacope

Condition of hand tube wells Condition of DPHE-UNICEFsand-filtered facilities

Pond ownership

Tubewells in good working condition

Name ofvillage

Total no.of

tubewells

No. of tubewells out of

order

No. used No.unusedbecauseof salinity

Total TotalNo.

No. ingood

workingcondition

No.unused

No. ofpublicpond

No. ofprivatepond

SutarkhuliAnandanagarSatghariaKhatailKhanaBaruikhaliChunkuriBatbuniaSaheberabadSrinagarRamnagarSurirdoba

Total

84

4341813332

53

32

21391221

26

22

2

18

16

3

2

1112

11

52

2219

1122

27

2

2121

9

2

5 4

3

3

9

3

1123

7212

23Source: 1986 office of Sub-Assistant Engineer, Public Health, Government of Bangladesh, Dacope, Khulna and during inspection.

10 WATERLINES VOL.lO NO.3 JANUARY 1992

Rainwater was often harvested using corrugated iron sheets and gutters.

Method of rainwater collection:Type I - roof discharge through CI sheetType II - roof discharge through tilesType III - roof discharge through strawType IV - polythene paper under open skyType V - bedsheet/sari under open sky

(earthen water containers made bylocal authorities).

People used rainwater for drinking,cooking, bathing, and washing clothes.It was reported that vegetables, pulses,and rice cooked in rainwater tastedbetter, and it was very popular whenpreparing tea. Several users felt that itwas not healthy to drink only rain-water, because it did not containminerals, as groundwater did, but somedrank it whenever they could because

they preferred the taste.During the rainy season the villagers

harvested the water and stored it,sometimes enough for the whole year,but more often only for a few months.The roads become very muddy duringthe rainy season, so rainwater is alsoharvested to avoid them.

The method of harvesting the rain-water varied, depending on the kindof house the villager lived in (see Table2). There were three common types ofroof in the area: corrugated iron sheets,tiles, and thatch. Roof discharge was acommon method of rainwater collec-tion. Gutters made of corrugated ironsheet, bamboo, betel-nut tree and thebark of banana plant had been widelyused in the rainwater catchment sys-tems. People with thatched housesusually used other devices to collectrainwater from the open sky, such asan upturned umbrella or a polyethelenesheet, a mosquito net, or a clean sari.The water collected this way wasemptied into a kalshi, and these in tumwere emptied into motkas, largerearthen storage vessels. The capacityof storage ranged from 32 to 5376Iitres. Motkas and kalshis were the onlypopular rainwater storage facilitiesavailable in the area.

A motka was usually used by richpeople for long-term storage. Thesewere semi-permanently placed insideor outside the house. The quality of thestored water varied, depending on themotivation and education of the house-holds. Two households visited inSutarkhali village were so enthusiasticthat they stored rainwater in largemotkas for year-round consumption.They took water out with a smallpre-cleaned pot, and kept the motkaclosed all the time.

An effective technique to maintainthe quality of the stored water bytreating it, using locally available

V

v

IIIIII

IV

11

Method ofrainwatercollection

(type)

season this dropped to 2.4 per cent, andponds or tubewells were used.

A joint Department of Public Health(DPHE) and UNICEF programme inthe area had been working to providebetter quality drinking-water. This pro-gramme included the establishment ofa sand-filtered tank of 2m x 2m x 2m.A handpump pumped water from apond (the groundwater was too salty)to the tank, where it was filtered. Evenwith this filtering, during the dryseason the water became salty andpeople did not want to use it.

Table 1 shows the condition ofdrinking-water facilities in the twelvevillages in Dacope. About 49 per centof tubewells were found to be out oforder, and of those working, 41 percent were not in use because theirwater was saline. Of the DPHE-UNICEF sand-filtered facilities, 45 percent were found to be idle. Only a fewof the ponds were government-owned,most being private.

Women played a significant role inthe collection of drinking-water. It wasassumed that water collection was awoman's job only, and men would doit only under exceptional circum-stances. Sometimes the women had towalk very far to collect water in kalshis

Table 2: Rainwater collection and storage scenario in village of Dacope

Total rainwater storage capacity for

Number ofdrinking purposes only

Type and num-households ber of storage(persons) For rainy sea- Also for off-sea- facilities

son use (Iitres) son use (litres) (Type/number)

6 90 A-59 5376 {0-4C-86 4224 0-65 96 A-68 64 A-422 640 C-26 192 {B-1A-55 128 {B-1A-112 576 {C-1B-25 128 {B-1A-15 32 A-216 960 C-37 288 {B-2A-36 724 {C-2A-420 384 {B-3A-29 1280 {C-4A-27 80 A-5

Storage facilities:Type A - a Kalshi of 16 litresType B - a Motka of 120 litresType C - a Motka of 350 litresType 0 - a Motka of 750 Htres

WATERLINES VOL.10 NO.3 JANUARY 1992 11

• a a a a

A motka, a large earthen vessel often used to store water.

Casng. weII5creen, droIIn,j f~ and 1<><>Is,puorf>$.~ In!!!I31meni and .,...,..Ienance ~ ferwa,ur weIs

T•••.071 435302112Fa. 071 4364394T••••• 262549 V,"us G VIRTUS

VIMU$ lid164 New CavendtSt'l 5tLondon W,M 7FJEngIircl

WATERwellsupply

drinking-water storage facility in-cluded the labour cost of collection,and the installation and maintenancecosts. The rainwater catchment systemwas found to be slightly cheaper thanthe sand-filtered system.

The study concluded that the propercollection and storage of rainwatercould solve the drinking-water prob-lem where there is a serious shortage.

M.D. Hussain and A.T.M. Ziauddin areAssociate Professors, Department of Farm Powerand Machinery, Bangladesh AgriculturalUniversity, Mymensingh, Bangladesh.

toads, and the dead bodies of rodents,lizards, cockroaches etc. in the storedrainwater. This was caused by im-proper storage practices.

Some people did realize the benefitsof drinking pure water though, andTable 3 shows the different sourcesused in the villages.

To improve their supply, about 75per cent of the villagers wanted to buildbetter and larger rainwater storageunits for private use. But only eco-nomically solvent households withmore than five acres were able to dothis. Farmers possessing less than anacre of land preferred the building ofa drinking-water storage facility forcommunity use. Most of the usersassume that it is the responsibility ofthe government to provide pure waterfor them. Only a few households werewilling to take out a bank loan at lowinterest rates to build the storagefacility, and then only if the applicationprocess was simplified.

Cost comparisons of the differenttypes of water sources were carriedout. The cost was calculated on thebasis of a consumption rate of threelitres per person per day for a familyof seven. The collated costs of the

material, was practised by two house-holds. They stacked snail shells inseveral layers, and in between thelayers, cowdung, straw, or rice huskwere placed. Rice husk was spreadover the layers and the stack wasburned gradually. The burnt shell wascollected and added to the stored waterat least once a month. It is believedthat the calcium (lime) or its compoundin the burnt shell is the main elementwhich purifies the water.

Most households with limited stor-age capacity found it very difficult tomaintain quality, as mosquito larvaeand other insects developed within twoweeks in the stored rainwater. Certaincommon varieties of fish, known lo-cally as koi (Anabas Testudinews),sing (Heteropneutes fossils), or magur(Clarias batrachus) were introducedinto the stored water to eat the larvaeof mosquito and other similar insects,but the fish discharge their own excretain the water, which again reduces thequality of the drinking-water.

Some consumers drink the rainwaterafter simply filtering it. Few familiestreated water using known chemicalssuch as alum, lime, or water-purifyingtablets. Sometimes people found frogs,

Table 3: Percentage of households using different sources of drinking-water in Dacope, KhulnaWater source %

Pond River Canal/ditch Rainwater Hand tubewe/l DPHE-UNICEFSand filter

Villages Rainy Dry Rainy Dry Rainy Dry Rainy Dry Rainy Dry Rainy Dryseason season season season season season season season season season season season

SutarkhaW 20 40 13.47 6.6 6.6 13.33Anandanagar 25 25 25 25Satgoria 50 50Khat.all 14 28 29 29Khona 10 8 8 16 16 26 16BarolkhaW 14 14 14 29 29Chunkuri 14 14 14 29 29Batbunia 25 25 25 25Shaheberabad 11 28 17 22 22Srinagar 11 10 23 11 11 11 23Ramnagar 20 40 20 20Burirdoba 16.5 16.5 34 16.5 16.5Mean 14.75 34.08 8.32 7.80 2.66 2.66 35.50 2.24 27.58 36.00 12.60 15.72

"Note: October to April is considered to be the dry season.

12 WATERLINES VOL. 10 NO.3 JANUARY 1992