Sensitivity Analysis of the Reactor Safety Study

摘要

The Reactor Safety Study (RSS) or Wash 1400 developed a methodology estimating the public risk from light water nuclear reactors. In order to give further insights into this study, a sensitivity analysis has been performed to determine the significant contributors to risk for both the PWR and BWR. The sensitivity to variation of the point values of the failure probabilities reported in the RSS was determined for the safety systems identified therein, as well as for many of the generic classes from which individual failures contributed to system failures. Increasing as well as decreasing point values were considered. An analysis of the sensitivity to increasing uncertainty in system failure probabilities was also performed. The sensitivity parameters chosen were release category probabilities, core melt probability, and the risk parameters of early fatalities, latent cancers and total property damage. The latter three are adequate for describing all public risks identified in the RSS. The results indicate reductions of public risk by less than a factor of two for factor reductions in system or generic failure probabilities as high as one hundred. There also appears to be more benefit in monitoring the most sensitive systems to verify adherence to RSS failure rates than to backfitting present reactors. The sensitivity analysis results do indicate, however, possible benefits in reducing human error rates.