Analysis of Risk Protocol Options for Private Water Supplies

Listen

4. Comparison of risk assessment scores

Two quantitative risk assessments were compared in this project. The first was developed at a workshop hosted by the Scottish Executive in 2003. This risk assessment (which is still at a draft stage) identified all the specific activities that were hazardous to the maintenance of good microbiological water quality at the source. Each factor was given a weighting, which was decided by consensus at the workshop attended by experts from all parts of the UK and based on some previous work that used the West Wales data. Each site was assessed, the hazards scored and summed to give an overall risk of contamination. Again, due to the non-normality of the data, a non-parametric analysis of the difference in the median risk score between the Welsh and North East Scotland sites was conducted (Figure 7).

Figure 7: Non-parametric Mann-Whitney U-test examining the similarity between the risk scores from the North East Scotland and West Wales PWS using the Scottish Executive quantitative risk assessment.

Figure 7: Non-parametric Mann-Whitney U-test examining the similarity between the risk scores from the North East Scotland and West Wales PWS using the Scottish Executive quantitative risk assessment.

The second risk assessment, developed for the National Trust by Jarman (1996) is not strictly a quantitative risk assessment. It works on the principle that individual hazard groups are scored and weights applied, these are scaled such that the maximum score in any hazard group is 10. It was not intended that the scores for each hazard group be summed, rather it translates the individual hazard scores into a risk category. However, in order to compare Welsh and Scottish datasets and to assess the capability of the NT risk assessment to represent the risk of contamination, the scores were summed. Since the assessment is not strictly quantitative, the term semi-quantitative has been used to describe it.

A non-parametric analysis of the difference in the median risk score between the Welsh and North East Scotland sites is shown in Figure 8. These analyses indicate that there is no significant difference in median risk scores between the two datasets.

Figure 8: Non-parametric Mann-Whitney U-test examining the similarity between the risk scores from the North East Scotland and West Wales PWS using the National Trust semi-quantitative risk assessment.

Figure 8: Non-parametric Mann-Whitney U-test examining the similarity between the risk scores from the North East Scotland and West Wales PWS using the National Trust semi-quantitative risk assessment.

Page updated: Thursday, March 27, 2008