Improving remediation decisions at hazardous waste sites with risk-based geostatistical analysis

The soil-related cancer risk of a hazardous waste site is usually bounded by selecting an upper confidence limit on the mean concentration value for each carcinogen and using these levels to define exposures, which are then combined with cancer potency factors to define risk. We suggest it is more appropriate to define cancer risk for each sample taken at the site. This sample-specific risk can be used in a geostatistical procedure to produce maps of block-specific risk at the site. When combined with models of site utilization, a quantity, which we call ''riskiness,'' can be developed to define the risk contribution of the k(th) site location to a particular exposure scenario. We demonstrate the derivation of block-specific risk and show how bath this quantity and riskiness can be displayed graphically to aid communication with stakeholders and decision making regarding actual remediation strategies.