Statistical analysis of the D. C. Cook preoperational environmental monitoring program

This report summarizes the major findings of an evaluation of the statistical adequacy of the preoperational environmental monitoring program for nonradioactive waste disposal at the Donald C. Cook Nuclear Power Plant. As a result of this study we found that variance components analysis methods are adequate to determine large magnitude changes in the environment. When an interaction effect between years and inner-outer factors (reference-stress) exists for the preoperation period, estimating and testing for the plant operation effect becomes difficult. This was illustrated by the benthic data analysis. It was further found that for the determination of impact hypothesis, several-factor-mixed-effects models are not needed. Simplifications, as shown by us, in the collapsed model by us, can provide the answer quite easily. Advanced methods, such as time-series analysis and biomathematical modeling, should be studied for use in the impact analysis. The limited analyses with these techniques showed promising results