Fuzzy PSA evaluation method for passive residual heat removal system

Highlights: ► We build the fuzzy PSA evaluation model of passive heat removal system. ► We modified the traditional PSA method, especially introducing the clustering to derive expert weight coefficient. ► We give more conservative estimation for top event failure with almost all the states of events are concerned. - Abstract: Fuzzy quantification evaluation is one of the most effective ways to solve the factors' uncertainty in the nuclear power station PSA. Take AP1000 passive residual heat removal system for example, a modified fuzzy PSA calculation model is established, in which two important conceptions membership and weight are introduced. The top events probability is finally calculated by the matrix quantification way. The results show that, fuzzy PSA method is suitable for the conservative estimation of failure probability, in which the relativity between different factors and the uncertainty of their own property are both considered. Meanwhile, it can provide a comprehensive evaluation of event failure and make an overall description of the real situation, especially for conservative estimation. It gives a new calculation method for nuclear PSA research.