Determinants of water quality, availability and use in Kurunegala, Sri Lanka.

Between January 1987 and February 1988, 4590 homes of children under five years of age were visited in three areas of Kurunegala district, Sri Lanka and data were collected on water related practices. 60% of the population used protected wells, 30% used unprotected sources and 10% used handpumps on boreholes or piped supplies. 90% of households had a source less than 1 km away. Mean water consumption was above 25 litres per capita per day and did not correlate with the distance to source. Samples of drinking water were collected and faecal coliform levels were determined in samples of stored water from 3092 households and in samples from the water sources used by 1043 of these households. The absence or presence of organisms in each sample, and the geometric mean count in samples with organisms were used as indices of contamination. Both indices changed with season and varied between areas and between types of water source. The proportion of positive source samples was uniformly high with the exception of piped supplies and handpumps. The mean count was highest for unprotected sources. There was no evidence that ground water contamination occurred in boreholes. With stored samples, boiling appeared to reduce contamination markedly. The proportion of positive stored water samples was also lower with the use of different vessels for collection and storage, with storage inside the house, and with use of a storage container other than an earthenware pot. Because surface water pollution appears to be important it is proposed that headwalls and drainage aprons be built around unprotected sources. Faecal contamination at the source may have more public health significance than contamination of stored water. In this respect public hygiene may play an important role in reducing water pollution at handpumps or protected wells.