A new methodology to evaluate severe accident management strategies using decision tree

Abstract

Accident management strategy is defined as the operator action to prevent core damage and to terminate the progress of core damage during accident. It also contains the actions which can maintain the integrity of containment and minimize off-site release. It provides an extension of the defense in depth for core melt accidents. The emergency operating procedures of pressurized water reactor extend up to the time of inadequate core cooling, but have not been specifically developed for significant core damage. After the accident of Fukushima nuclear power plant is occurred, many question about the utility of severe accident management strategy have been raised. The effectiveness of SAMG(Severe Accident Management & Guidance) which is representative guide of severe accident management strategies also becomes significant issue. In this study, the evaluation methodology for severe accident management strategies is developed by using decision tree. Decision tree has an advantage that it can evaluate feasibility, effectiveness, adverse effect by comprehensive consideration. Cavity flooding strategy is selected for evaluation and decision tree methodology is applied at selected strategy. The quantification result of decision tree which is based on PSA(Probabilistic Safety Assessment) methodology and dose from containment at restricted area boundary which is analyzed by MACCS2 code are combined to perform risk analysis. By using the result of risk analysis, cavity flooding strategy is evaluated its effectiveness. New quantification methodology is also developed by Monte-Carlo method to reduce uncertainty which is caused by using discrete value during the quantification process. Distributed quantification result can be obtained from using developed methodology.