Individual household water supplies as a control measure against Schistosoma mansoni

SCHISTOSOMIASIS Individual household water supplies as a control measure against Schistosoma mansoni* A study in rural St Lucia G. 0. UNRAU1 As part of a programme to evaluate single control measures for reducing the transmission of Schistosoma mansoni, household water supplies were installed in S rural settlements in the Riche Fond Valley ofSt Lucia. About 2 000 persons who previously were dependent on rivers and streams are now receiving safe water at their homes. The systems provide useful design data on individual water requirements in rural areas. This experience suggests that future rural water systems can be designed more economically and efficiently by using consumption rates that are closer to the actual requirements and by eliminating water wastage at the taps. Schistosomiasis is one of the most widespread parasitic infections in the tropical and subtropical world and, in terms of incidence and prevalence, is probably increasing more rapidly than any other parasitic disease. In 1965, the Government of St Lucia and the Rockefeller Foundation formally agreed to establish a Research and Control Depart- ment that would investigate the efficiency and cost- benefit ratios of several methods for reducing the transmission of schistosomiasis mansoni on the island. One possible method, aimed at preventing exposure to infection, is the provision of household water supplies in rural areas in order to make it unnecessary for the population to enter natural waters harbouring the infective larval stage of the schistosome. In the period 1970-72, 5 rural settlements in the Riche Fond Valley of St Lucia were provided with household water supply systems designed to give convenient and reliable service at low cost. This paper reports on 3 years' successful operation of these systems, gives details of their design, construc- tion, and cost, and offers evidence that prevailing estimates of water consumption rates in rural areas are unrealistically high. A preliminary report on the epidemiological effectiveness of the project is given in a companion paper by Jordan et al. (see this issue, P. 9). * From the Research and Control Department, Castries, St Lucia, West Indies. Reprints may be obtained from the author at that address. 1 Staff Member, the Rockefeller Foundation. PROJECT AREA St Lucia, one of the Windward Islands in the eastern Caribbean Sea, is approximately 43 km long and 23 km wide and has an area of about 616 km2. The island is mountainous, with a forest-covered central ridge extending almost its entire length; rivers and streams drain eastward to the Atlantic Ocean and westward to the Caribbean through valleys separated by hills rising to 610 m. The island receives rain throughout most of the year, but there is considerable monthly variation; usually July and October are the wettest months, and February and March the driest. Annual rainfall averages 127 cm at the extreme ends of the island and more than 406 cm in the central mountainous region. Annual temperatures range from 20° to 340C. The Riche Fond Valley is situated on the eastern slope of the central mountain range. There are numerous streams in the area, and these supply an acceptable volume of water because of the high rainfall; rarely do the streams dry up. Rainfall records for the valley indicate a mean of 208 cm per year over the past 30 years. The 5 settlements provided with household water supplies-Grande Ravine, Thomazo, Grande Riviere, Morne Panache, and Debonnaire-lie near the foot of the mountains, with many of the houses built on the lower slopes (Fig. 1); these settlements were all previously dependent on rivers and streams. The other 6 settlements shown in Fig. 1 are equipped 3315 - 1 BULL. WORLD HEALTH ORGAN., Vol. 52, 1975 G. 0. UNRAU Fig. 1. Map of Riche Fond Valley showing locations of 5 settlements provided with household water supplies: Grande Ravine, Thomazo, Grande Riviere, Morne Panache, and Debonnaire. The 6 other settlements shown serve as the comparison area. with a public standpipe water system and serve as the comparison area. (See also the companion paper by Jordan et al., p. 9 of this issue.) COLLECTION OF BASELINE DATA A household census, conducted in 1968, showed that the population of the 5 settlements numbered about 2 000, with an average of 5.2 persons per household (Table 1); the mean age was 15 years, and the 0-4-year age group represented about 20% of the total. A detailed geographic survey was made to deter- mine the suitability of various water sources in relation to the location of the houses, and stream flows were measured to determine water volumes at different elevations. In addition, data were collected on the quantity of water used and on the distances water was carried from the rivers and streams to the houses (Table 1). The average daily consumption Table 1. Size of 5 settlements in 1968, and the distances water was carried from source to household before the installation of supply systems Distance water Settlement No. of Popula- was carried (m)houses tion range mean Grande Ravine 70 400 137-480 342 Thomazo 49 245 37-549 217 Grande Riviere 154 770 14-549 226 Morne Panache 53 275 4-494 304 Debonnaire 60 300 82-242 165 2 CONTROL OF SCHISTOSOMA MANSONI was estimated to be about 15 litres per person. The daily work load of transporting the water, over rough and sometimes steep paths, was estimated to involve 837 persons carrying a total of 34 m3 of water. In an initial stool survey, specimens were obtained from occupants of each household to determine the prevalence of schistosomiasis, hookworm, ascariasis, and infections due to Trichuria and Strongyloides spp. These surveys were repeated annually to mea- sure the incidence of new infections. *PROJECT DESIGN CRITERIA The overall criterion was that each household be furnished with its individual supply of safe, potable water, adequate in amount and available for 24 h each day. This required the design of 3 separate systems: one for Grande Ravine; one for Thomazo, Grande Riviere, and Morne Panache; and one for Debonnaire. A few houses were located either too far away or at too high an elevation to be served economically with an individual supply, and mainly for their convenience public laundries and shower units were installed at 3 selected sites; these units have since become popular with many of the other householders. In addition, 3 swimming pools were constructed, since children of the settlements regu- larly swim and play in the streams and rivers, and their exposure in terms both of length of time and of body area may be an important factor in the trans- mission of schistosomiasis mansoni. The household supply systems were designed to serve a total population twice the initial size, this being the standard for rural water systems in many countries of the Americas. It is, however, an over- estimation, because rural populations usually grow more slowly than urban populations. In fact, since the project was designed, the population served has grown at the rate of only 1.8% per year, and will therefore take 40 years to double. Since water requirements in a rural community are rather arbitrary, a daily quantity of 114 litres per head of population was allowed; given the ultimate aim of the project, a sizable margin of safety in consumption demand was considered desirable. Evidence from other studies (5) suggests that the availability of ample quantities of water may be as important as the bacteriological quality of the sup- ply, or even more so. Since the sophisticated treatment of water may not be an absolute requirement, simple types of treat- Table 2. Experimental variations in water treatment System Source characteristics Treatment Grande Ravine slow-flowing stream infiltration gallery, storage Thomazo, Grande rapid flow, intermittent sand sedimentation, Riviere, & Morne & stone stream-bed, storage Panache relatively isolated watershed Debonnaire slow-flowing stream sedimentation, storage ment were used in the interest of experimentation (Table 2). MATERIALS AND FACILITIES Distribution systems In order to provide effective service at low cost, a departure was made from the conventional design of distribution systems (corporation and curb cocks at each service, etc.). In all 3 cases the main distribu- tion line serves a group of households, each having a water tap, and the only valves in the system are located at the inlet and at major junctions. A single meter measures the delivery to the distribution line. Meters were also installed at the public laundries and showers. The use of poly(vinyl chloride) (PVC) pipes con- ferred several advantages. Firstly, since the rugged terrain required that pipes be transported manually to the construction sites, the low weight of these plastic pipes let us use the help of women from the settlements, who often carried one length each; manual transport of iron or steel pipes of equivalent size would have required far more labour. Secondly, the lower friction losses with PVC pipes permit- ted the use of cheaper, small-bore pipes. Thirdly, since pipes had to be imported in any case, selection of the lighter PVC pipes resulted not only in a saving on shipping costs but also in a considerable saving (about 50%) on the capital cost. Schedule 80 heavy-walled pipe was used, both because it can be threaded with ordinary pipe threading tools and the need for threaded adapters to install valves and taps in the line is thereby eliminated, and also because it gives the extra strength desirable in rough terrain. Water taps Provision of an adequate water supply from avail- able sources indicated a design of minimum waste, 3 4 G. 0. UNRAU 2/ o 0 4 8 12 16 litres per minute . W~~~~~~~~~~~~~~~~~~~~~~~~~~~HO 7S553 | Fig. 2. Typical quantity-pressure curve for a Fordilla water tap. and the discovery that a certain amount of pumping would be required stressed the need for low water loss. Accordingly, exclusive use was made of the spring-loaded Fordilla tap (Ford Meter Box Co., Wabash, IN, USA), which automatically shuts off the flow after a predetermined quantity of water has been discharged; to obtain more water the top button must be released and pushed down again. This tap thus provides an unlimited amount of water but cannot be left open. Borjesson & Bobeda (2) reported satisfactory results with the Fordilla tap in a test conducted in Paraguay, and Aris (1) found that it gave good control of water consumption in Guatemala. In addition to eliminating waste, the Fordilla tap allows the use of smaller-diameter pipe in the distri- bution system because fewer taps are operating simultaneously. The unit controls the volume delivered, which in the case of standard models is about 1 litre per cycle. For this project the manufacturer provided units that deliver about 7.5 litres per cycle. Since the flow can be terminated at any point in the cycle by manual release of the button, there seems to be no disadvantage in the use of this model; furthermore, the larger volume means fewer interruptions in the filling of laundry tubs and large containers. The design capacity of a Fordilla tap is about 10 litres/min at 10' Pa. A typical quantity-pressure curve (Fig. 2) shows that the delivery rate is not changed greatly by an increase in pressure.a For those who wanted shower facilities at their homes, a shower-head attachment was devised by a Ford, R. V. (1964) The hydraulic design of Fordilla pods. U.S.A.I.D. Community Water Supply Seminars, Bo- gota, Addis Ababa, and Bangkok. Fig. 3. Household shower arrangement with a single Fordilla water tap. fitting the tap outlet with a thread and adding a tee joint and a simple shut-off valve. A 2-m length of pipe, 2 elbows, and a shower-head make up the shower (Fig. 3). In operation the lower valve is closed, so that the flow is diverted through the shower; when the valve is opened the tap functions in the usual way. A foot control can be made by drilling a small hole through the push button on the tap and attaching this with a wire to a short board used as a pedal. Storage tanks Because of the difficult terrain and the lack of roads to suitable tank sites, a sectional bolted, 12- gauge galvanized steel tank was considered the most appropriate and economical storage reservoir. Where necessary, the tanks were transported man- ually in sections and assembled on site. Public laundry and shower units The public laundry is a simple structure having a concrete centre bench with 3 or 4 concrete tubs mounted along each side, a slightly sloping concrete floor, and a hip-type roof. Water is supplied to each tub by a Fordilla tap. The 6-stall public shower unit is built of concrete block walls and partitions. Each stall uses a Fordilla tap with an attached spray-head controlled by a pedal. CONTROL OF SCHISTOSOMA MANSONI Fig. 4. Infiltration gallery, Grande Ravine water system. Swimming pools A simple fill-and-draw (as opposed to recirculat- ing) pool, measuring 6 x 12 x 1 m, was designed. For pools less than 1.5 m deep the common minimum- area requirement is 0.9 m2 per person; this pool, when used by an assumed maximum of 50 children at one time, provides an area of nearly 1.5 m2 per child. Each pool was constructed below ground, of reinforced concrete flooring and concrete block walls with a coat of plaster applied to the inside. At one end the floor is raised to provide a water depth of only 0.6 m. DESCRIPTION OF THE SYSTEMS The 5 settlements were provided with household water supplies over a period of approximately 2½/2 years: Grande Ravine in 1970; Thomazo, Grande Riviere, and Morne Panache in 1971; and Debon- naire in 1972. Grande Ravine For this settlement of 70 houses situated on one side of a small hill, the only practicable source of piped water was the nearby small, shallow stream on which residents had previously relied (Fig. 1). This stream had a flow rate (1 ms/min) sufficient for the population. The site selected for the intake permitted the use of an infiltration gallery (Fig. 4). The water is collected after having permeated through many feet of sandy soil, a process that aids in the removal of potentially harmful organisms by filtration; further- more, being subsurface water, the supply does not dry up as quickly as surface supplies in times of drought. From the suction well the water is pumped by a 2- stage turbine type pump through a 5-cm diameter PVC line to a 83.6 ms storage tank. The total dynamic head is 73 m. A 5-cm main, 117 m long, serves the loop net- work, with 2.5-cm pipe used for the distribution loop and 2-cm pipe for the service lines. Experience has proved that this distribution system provides ade- quate supply and pressure to the 70 houses at all times. Public laundry and shower facilities (6 tubs and 6 shower stalls) and a swimming pool were also constructed at Grande Ravine. Impervious Material Backf ill Filter 1/4" -1/2" GRAVEL 4e?WeM;terial'i''~ 1/2" -1" GRAVEL / :~.?t 8" 0 PIPE TYPICAL CROSS SECTION WHO 75555 PLAN 5 6 G. 0. UNRAU Thomazo-Grande Riviere-Morne Panache The system for these settlements (a total of 256 houses), which lie along the major road close to the main river, is supplied by a nearby tributary. Water is collected by means of a dam, from which it flows through a sedimentation chamber into a suction well and is then pumped through a 6.5-cm line to a 159.6 m3 storage tank. The dynamic head is 103 m, which requires a 15-kW motor. From the tank, distribution is by gravity by means of 885 m of 6.5-cm PVC main to Grande Riviere and either 2.5- or 4-cm PVC pipe for the other distribution lines. Since the location of the houses at different eleva- tions requires both low- and high-pressure distribu- tion networks, pressure-reducing chambers or valves were installed where appropriate to limit working pressures to a maximum of 483 x 104 Pa. Public laundry and shower units (8 tubs and 6 shower stalls) were constructed at Morne Panache and at the opposite end of Grande Riviere. In addition, swimming pools were provided at each end of Grande Riviere; these are suitably situated for children from Thomazo and Debonnaire, respec- tively. Debonnaire This system, serving 60 houses, is the simplest. Basically it consists of intakes on 2 small streams, a sedimentation chamber, a storage tank of 83.6 m3 capacity, and a loop distribution line of 4-cm PVC pipe, with a 2.5-cm spur line at the extreme end. The system operates by gravity. Recently another source has been tapped, at a higher elevation, to supply a group of 10 houses above Morne Panache, and the surplus from this is used to augment the Debonnaire supply. Public laundry and shower units have not been provided at this location. Since the nearby small streams are not especially suitable for the washing of clothes, it was assumed that the distinct convenience of a household water supply would make residents prefer to do washing at their homes. DISCUSSION Water quality In the 3 years of the project's operation the chemicophysical properties of the water supplied were within acceptable ranges. The iron content, however, was somewhat high in the Grande Ravine and Debonnaire systems, with consequent discolora- tion of the water in the Grande Ravine swimming pool. Only the swimming pool supplies are chlo- rinated, although such treatment can be provided for the other water supplies if required. Water samples are collected routinely for bacteriological analysis by the Millipore filter technique (3); 80% of the sam- ples so far examined have been completely negative for coliform bacteria, and the samples that were positive had only a few colonies per 100 ml. Henderson a has suggested that, in tropical cli- mates, retaining the water in reservoirs for a longer time may lead to an improvement in quality owing to an accelerated rate of bacterial die-off. Although the present designs allowed for generous storage capacity, the extent to which this is responsible for the good quality of the water remains uncertain. However, increased storage time certainly reduces the viability of schistosome cercariae, and if storage can be substituted for sophisticated treating equip- ment, the maintenance of a system is simplified and made less costly. Consumption rates The average daily use of the systems, including the laundry-shower units, varies between 40 and 50 litres per person, being an increase of approximately 35 litres over the estimated consumption of carried water. The exact number of people using the laun- dry-shower facilities is difficult to determine; how- ever, records show an average daily consumption of 2.4 ms for the 6 laundry tubs and 6 showers at Grande Ravine, and daily averages of 1 ms and 1.8 ms for the 6 showers and 8 laundry tubs at Morne Panache. This consumption rate is far below the 114 litres per person per day allowed for; future designs based on 60 litres/person/day would provide adequate capacity. This figure, which is about 3-4 orders of magnitude lower than those now used in many areas, is similar to the figure given by Aris (1) as the basis of design for rural aqueducts in Guatemala. Cost analysis Table 3 presents a brief summary of the costs of installing household water supplies for the 5 settle- ments combined. The figures do not include engi- neering and supervision overheads, which would have to be calculated according to local availability and price for a given area. a Henderson, J. M. (1965) Community water supply development in underdeveloped areas. Tropical Public Health Seminar, Harvard School of Public Health, Boston, MA, USA. CONTROL OF SCHISTOSOMA MANSONI Table 3. Costs a (US $) of individual water supplies to the 5 settlements Design capacity Equipment & Installation Total Per capita (No. of materials labour hired cost cost persons) 4500 $31 002 $9 038 $40 040 $8.90 a When the costs of the 3 laundry-shower units and the 3 swim- ming pools are added to these figures, the total cost becomes$48 613 and the per capita cost about $10.80. During the period of installation (about 21/2 years) costs increased owing to several factors: a rise in labour rates; the fact that the second and third systems required more manual labour than the first because of more difficult terrain; and, associated with the latter, a reduced level of volunteer help. It had been hoped initially that much of the manual labour needed would be volunteered as part of a community effort. Although this effort never reached the level of expectation, nevertheless each household did provide the labour for digging the trenches for its service line pipe and a small portion of the distribution system pipe. Since there may be those who would argue that developing countries can afford a well-distributed public hydrant supply in rural areas but not the luxury of individual household service, it should be noted that once the essential equipment and mate- rials for an adequate public hydrant supply have been purchased and installed, the added expense of service lines and taps amounts to only a small percentage of the total cost. In the case of the 3 systems described the capital cost for individual service averaged 10% of the total expenditure. Cal- culation of the costs of the service line, fittings, and taps showed that the additional expense averaged slightly over US $12 per house, or approximately $2.40 per person. Moreover, water experts have long maintained that carried water is not good enough. According to Wolman (4): "Twenty-five years ago it was our assumption that bringing water to a tap in the central square of a village constituted progress. Today we see our goal differently. If we can't bring water into the house itself, or at the very least to an adjacent courtyard, we simply have not done the job." Preliminary data on the incidence of schisto- somiasis mansoni in the household water supply and comparison areas supports this view (see the paper by Jordan et al., p. 9 of this issue). To the direct health benefits associated with a household water supply can be added the saving of the time and labour previously expended in carrying water from rivers and streams. CONCLUSIONS Experience with these household water supply systems indicates: (1) that if water wastage can be eliminated in a design that furthermore is based on a consumption rate more realistic to actual needs for rural areas, many sources normally considered of unacceptable capacity can become adequate sup- plies; (2) that the use of PVC pipe and Fordilla taps, by eliminating waste, permits the design of economi- cal, simple distribution systems easily maintained by limited numbers of skilled personnel; and (3) that installation labour costs can be significantly reduced when local populations are willing to contribute to the project as a community effort. ACKNOWLEDGEMENTS The author wishes to express his gratitude to Dr P. Jordan, Director of the Research and Control Depart- ment, St Lucia, and to Dr P. N. Owens, Associate Director for Health Sciences, the Rockefeller Foundation, for their review of the manuscript and the help and suggestions they provided during all phases of the project; to Mr Jorge Saravia, group leader, and the 2 groups of sanitary engineering students from the Universidad del Valle, Cali, Colombia, who helped with the field surveys and with the drafting and design of the water systems; to the field technicians of the engineering section, Research and Control Department, whose devoted work made it possible to complete the water systems; and to Mr 0. F. Morris for his help with the project. 7 8 G. 0. UNRAU RESUMt L'INSTALLATION DE L EAU COURANTE DANS LES HABITATIONS COMME MOYEN DE LUTTE CONTRE SCHISTOSOMA MANSONI Des etudes ont ete entreprises A Ste-Lucie pour evaluer l'efficacit6 de diverses mesures de lutte contre la trans- mission de Schistosoma mansoni. L'une de ces mesures consiste A installer l'eau courante dans toutes les habita- tions rurales car l'organisation d'un bon service d'eau fait beaucoup pour eviter aux populations des campagnes les contacts avec des eaux infectees. Par le moyen d'enquetes geographiques, de recense- ments des habitations et d'6tudes annuelles sur des echantillons de selles, des donnees de base avaient et6 recueillies dans cinq villages de la vallee de Riche Fond pour aider A concevoir un systeme de distribution d'eau adequat et, par la suite, A evaluer les effets de l'approvi- sionnement en eau saine sur la frequence des infections helminthiques. Sur la base de ces donnees, trois systemes distincts ont ete r6alis6s pour fournir l'eau courante A chaque habitation. En outre, 3 buanderies-douches pu- bliques et 3 piscines ont ete amenagees en des points strategiques. Tout a ete fait pour que les installations soient aussi simples et aussi economiques que possible. Dans deux des systemes, le reservoir d'eau est aliment6 par pompage et dans le troisieme par gravite. Toutes les conduites installees sont en chlorure de polyvinyle et tous les postes d'eau sont pourvus d'un robinet # Fordilla> A 6coule- ment contr6le. Ce type de robinet permet de tirer autant d'eau qu'on le desire mais empeche tout gaspillage car il se ferme automatiquement si on ne le maintient pas en position ouverte. L'analyse des couts fait apparaitre que les frais d'equipement, de materiaux et de main- d'oeuvre (A 1'exclusion des frais d'etude technique et de surveillance des travaux) se sont eleves A US $8,90 par habitant pour ces trois systemes congus pour desservir au total une population de 4 500 personnes. Si l'on fait aussi entrer en ligne de compte les buanderies-douches et les piscines, le cout par habitant s'eleve A US $10,80. Pendant 3 annees, ces systemes ont foumi de l'eau potable, 24 h sur 24, A une population rurale d'environ 2 000 personnes. Pendant cette periode, la consommation d'eau par personne et par jour a augmente de 35 litres, ce qui signifie que l'hygiene et les pratiques sanitaires se sont ameliorees dans les m6nages; au credit de l'op& ration, il faut mettre aussi d'importantes economies de temps et d'energie puisque les villageois n'ont plus A aller chercher l'eau. II ressort de cette experience que, moyennant un traite- ment minimal, il est possible d'utiliser les cours d'eau locaux pour assurer aux petites collectivites rurales un service d'eau commode et de qualite acceptable. Sont egalement A retenir un certain nombre de points qui meritent d'etre pris en consideration pour l'etude de futurs reseaux d'alimentation en eau en milieu rural: 1) Bien que les installations aient ete concues pour une consommation de 114 litres d'eau par personne et par jour, la consommation journaliere effective a ete de moins de 50 litres par personne. 2) Les installations ont ete calculees pour la desserte d'une population double de la population actuelle; c'est lA peut-etre une marge de securite excessive car l'accrois- sement des populations rurales est generalement plus lent que celui des populations urbaines. Depuis la construc- tion de ces distributions d'eau, le taux d'accroissement demographique des cinq villages a ete de 1,8% par an, ce qui signifie qu'il faudra 40 ans pour que la popula- tion atteigne le double de ce qu'elle est actuellement et que, pendant cette periode, la capacite excedentaire ne sera pas utilis6e. 3) Une comparaison de cout entre l'installation de raccordements individuels pour chaque habitation et l'ins- tallation de postes d'eau publics bien repartis fait appa- raitre que l'excedent de depenses d'equipement (conduites et robinets) dans le premier cas ne represente que 10% de l'investissement total. Etant donne les nombreux avantages des raccordements individuels pour chaque habitation, cet excddent de depenses paraft tout L fait justifi6. En tablant sur des valeurs de consommation realistes et en eliminant les possibilites de gaspillage, on peut admettre une norme de 60 litres par personne et par jour pour l'approvisionnement en eau des populations des campagnes. Des installations de cette capacite sont suffi- santes pour desservir commod6ment et economiquement beaucoup de collectivites rurales. REFERENCES 1. ARIS, D. In: International Conference on Water for Peace, May, 1967. Washington, DC, US Government Printing Office, 1968, pp. 2-4. 2. BORJESSON, E. K. G. & BOBEDA, C. M. Journal of the American Water Works Association, 56: 853-862 (1964). 3. TARAS, M. J. ET AL. (EDITORS). Standard methods for examination of water and waste water, 13th ed. Washington, DC, American Public Health Associa- tion, 1971, pp. 678-687. 4. WOLMAN, A. World Health, July/August, 1964, p. 29. 5. WOLMAN, A. & BosCH, H. M. US water supply lessons applicable to developing countries. In: White, G. F., ed. Water, health, and society. Bloomington, Indiana University Press, 1969, pp. 219-233.