Fragility analysis of nuclear containment vessels under accident internal pressure considering statistical uncertainty

Abstract

Assessing the probabilistic safety performance of nuclear containment vessels (NCVs) under accident conditions is essential for ensuring nuclear safety. This study investigates the statistical uncertainty involved in fragility analysis of NCVs subjected to internal pressure. Based on the mathematical statistical principle and reliability theory, a new method is proposed for quantifying the statistical uncertainty of the fragility curve. On this basis, a method for determining the minimum number of necessary samples to meet the desired analytical accuracy is presented. After validating the effectiveness of the proposed methods, it is applied to assess the fragility and reliability of the NCV under internal pressure. The results show that statistical uncertainty significantly affects the fragility curve and total failure probability of the NCV when the number of simulation samples is limited, and the extent of the impact is closely related to the confidence level. The proposed methods can effectively quantify the statistical uncertainty in fragility analysis and can determine the minimum number of simulation samples required for fragility analysis at a specified confidence level and analytical accuracy.