Executive summary
1. It is generally accepted that the Scottish population has ready access to clean and reliable sources of water. However, there is a sizeable minority of the population whose water comes from a non-statutory water supplier. These supplies may be at a greater risk of microbial and chemical contamination as treatment is often minimal or non-existent. There are an estimated 31 000 to 38 000 private water supplies in Scotland (Reid et al., 1998; DWI, 1993) and although the majority of these supplies serve only one household (Reid et al., 1999), there are many others which serve businesses such as hotels, campsites and residential schools. Contaminated water can carry water-borne infections such as cryptosporidiosis and E.coli 0157.
2. The frequency of statutory monitoring of these supplies varies with the type and volume of use with those water supplies serving premises used for food production or preparation and accommodation monitored more regularly than shared domestic supplies. There is no statutory requirement to monitor sources supplying a single household. Infrequent or non-existent monitoring implies that there is considerable potential for any microbial contamination to go undetected. This has potential consequences for public health particularly where the contamination is sporadic and associated with other environmental conditions such as rainfall.
3. In response to the lack of correlation between indicator species and the presence of potentially harmful organisms and to the failure of the monitoring schemes to successfully ensure a consistent supply of uncontaminated water, the Water Services Unit of the Scottish Executive, along with the Department of Health, commissioned the development of a Microbiological Risk Assessment. This procedure was designed to systematically assess the risk of contamination of any private water supply (Lamb et al., 1998). Through the use of this mechanism, both the likelihood that a supply would be contaminated and the factors that were contributing to this risk (hazards) could be identified. Efforts could then be made to reduce or manage the risk by eliminating as many hazards as possible. The primary purpose of this research was to validate the Microbiological Risk Assessment developed by Lamb et al., (1998) for use in future Private Water Supply Regulation.
4. A total of 33 sites were selected for monitoring throughout 2002. The proportion of source types reflected the proportion of supplies in each of the Private Water Supply Regulation categories that occur nationally and included source on a range of land uses such as (arable agriculture, grazed pastures, woodland and moorland). The type of supplies within the network included wells tapping shallow groundwater, springs, surface flow to a reservoir or holding tank and field drain collection. The monitoring sites were in both the Dee and Don Catchments and ranged from Tarves in the north to Auchenblae in the south and as far west as Braemar.
5. No source sampled was free of coliform contamination for each of the 12 samples taken throughout 2002.
6. Results showed that there was a greater degree of contamination in sources surrounded by agricultural land compared with moorland or woodland. This may be reflecting a greater population of bacteria associated with the more fertile land. Although there was some indication that the presence of grazing animals increased the potential for contamination by faecal coliforms, the relationship was not significant for all statistical tests.
7. Rainfall appears to be an important driver of contamination either by inducing overland flow (observed) or by enabling the transport of bacteria to the groundwater. However, it was found that the amount of rainfall was less important than the timing of heavy or prolonged rainfall events. This antecedent rainfall would reduce any soil moisture deficit and induce near saturated conditions in the soil making transport of bacteria in larger pores more likely. The water sources associated with soils that were naturally wet were found to be at greater risk of contamination. In these soils near saturated conditions would be reached in less time than freely drained soils.
8. Other studies have been shown to support the results from this work. The complex nature of Private Water Supplies may mean that a validation of the Microbiological Risk Assessment is not possible without a considerable increase in the number of sources sampled. Intuitively, the individual components of the Microbiological Risk Assessment would seem to encompass the main factors likely to affect the quality of water supplies. Therefore the main value of the Microbiological Risk Assessment, as currently constructed, may be as a "checklist" of potential action points to secure the integrity of the supply as far as possible.
Acknowledgements: the authors would like to thank all the owners of the Private Water Supplies that were monitored for their patience and understanding throughout the year. Without their help and support we could not have undertaken this research. We would also like to thank all those individuals who took their own monthly water samples and Calum MacKenzie and Joe May (Aberdeenshire Council Environmental Health Department) for their considerable help in introducing us to the owners and managers of Category 2 supplies. Our thanks also go to Macaulay staff who helped with the sampling (Malcolm Coull) and shared the burden of reading the incubated trays (in particular Yvonne Cook and Claire Abel). Finally, We would like to thank Elaine Wilson for allowing us to quote results of the comparison between Colilert18™ and the more traditional methods to determine microbiological water quality from her unfinished PhD thesis.