An integral approach for calculating uncertainties in consequences from nuclear reactor accidents

This paper presents a Personal Computer-based model that uses an integral approach for calculation of early off-site consequences resulting from nuclear power plant accidents. The computing time requirements for a typical calculation on a mainframe computer using this model are two orders of magnitude lower than those of CRAC2 and MACCS codes, thus providing a valuable tool for sensitivity and uncertainty studies. The model predicts time-integrated air concentration of each radionuclide at any location from release as a function of time-integrated source strength using the Gaussian plume model. The concentration can be calculated at the centerline of a Gaussian profile or, optionally, as an average over the cross-section based on a top-hat distribution. The solution procedure involves direct analytic integration of air concentration equations over time and position. This is different from the differential approach currently used in CRAC2 and MACCS codes. The present model uses simplified meteorology. Dispersion parameters are calculated from exponential fits to the Pasquill-Gifford curves for six atmospheric stability classes designated A to F, and from an approximation as indicated in Reg. guide 1.145 for the seventh stability class G (extremely stable). 8 refs., 2 figs