Adequacy of Data Sets and Impacts on Risk Assessment. Gordon S. Randall, Julie T. Wroble, and Thomas H. Angus, Ecology and Environment, Inc., 999 Third Ave., Suite 1500, Seattle, WA 98104
Ecology and Environment, Inc., conducted human health and ecological risk assessments for a hazardous waste site with a limited number of known contaminants and source areas. A large number of samples were collected for the expected contaminants in order to derive accurate estimates of area-wide average concentrations in environmental media and to derive meaningful risk estimates for the site. However, a very limited number of samples were analyzed for other potential contaminants, including dioxin, in order to limit analytical costs and avoid sampling in areas expected to be clean. The expected site contaminants were present at high concentrations only in very localized areas and were shown not to be of significant concern to human health or the environment for the site as a whole. The other chemicals tested for only in suspected "hot spots" were found at moderately high concentrations, and as a result risk estimates for the site were skewed towards these chemicals, even though they are not likely to be present over much of the site. Similar problems related to data have been encountered when conducting risk assessments in the past, especially related to dioxin, which is frequently tested in only a limited number of samples due to the extreme cost of dioxin analytical methods. In this paper, the problems associated with conducting a risk assessment using limited or biased data and the meaning of resulting risk estimates are discussed.