Quantitative uncertainty analysis for the human fish ingestion pathway : select radionuclide and chemical contaminants in the Clinch River/Watts Bar Reservoir System

The goal of this study was to perform a quantitative uncertainty analysis on the excess health risk associated with the ingestion of select radionuclide and chemical contaminants as the result of human consumption of contaminated fish. A quantitative uncertainty analysis is an efficient method of ranking contaminants and pathways and of determining the most sensitive parameters to the overall estimate of uncertainty in the model result; thus, providing a powerful tool in decision making processes. Subjective probability distributions were determined for each parameter in the risk assessment equations. These distributions were defined by using subjective judgment after extensively reviewing the literature and in some cases, after eliciting expert opinion. The Monte Carlo approach used to propagate the subjective probability distributions for the model parameters was Latin Hypercube Sampling. Subjective confidence intervals were obtained for the potential detriment from the human ingestion of fish contaminated with Co-60, Sr-90, Cs-137, Pu-239, methyl-mercury (MeHg), chlordane, Aroclor-1254, and Aroclor-1260. The results indicate that the potential risk from the chemical contamination in the fish of the Clinch River/Watts Bar Reservoir system is greater than the potential risk from radionuclides. The primary contaminants for further study are Aroclor-1260, Aroclor-1254, MeHg, and chlordane. The results of the sensitivity analysis indicate that the overall estimate of uncertainty in the total risk estimates would be most efficiently reduced by taking local surveys to obtain better subjective probability distributions for parameters such as the number of fish-meals per day and the exposure duration instead of taking more fish samples.