Contaminated land risk assessment: variability in site assessment and decision making in the UK

摘要

Judgement forms an integral part of a risk-based approach to the assessment of land affected by contamination. Legislation and guidance suggest that the assessor should use a rational step-wise process to identify pollutant linkages in order to assess risk from land contamination. The present study aims to investigate the decision-making processes that are used by experienced contaminated land assessors. This study required 29 participants with a minimum of five years' relevant experience to rate the level of risk from land contamination on 27 hypothetical housing development sites. Each site was designed with specific information (variables) used as indicators of the potential for unacceptable risk. Linear regression analysis was used to identify the significance of each of the variables in determining the level of risk assessed by participants. The first of the key findings was that considerable disagreement was observed between participants, and this was correlated to cases with contradictory information. This may have also been related to the participant's perception of the available risk scale. The linear regression analysis showed that the most influential variables were chemical-test data and the presence of human-exposure pathways. These findings would suggest that experienced assessors focus on a few key aspects of the information available to assess risk from land contamination. However, analysis of the qualitative data collected in the study supported a more holistic decision-making process, in line with use of pollutant linkages described in guidance. The results suggest that when presented with limited data for development sites, assessors may rely on a few variables to rate the risk, but that a coherent picture of the interaction of all of the variables is required for a more confident assessment. The findings of the study presented here can be used to inform training and future guidance in this sector.