(A) study on the pressurized water reactor (PWR) containment response analysis methodologies for postulated severe accident

Abstract

The present study contains two major parts: one is the treatment of uncertainties involved in the current APET and the other is the importance analysis of the APET uncertainty inputs. A clear disadvantage of the expert opinion polling process approach for uncertainty analysis of the current probabilistic risk assessment (PRA) is that the sufficient robustness in the final results may not be attained against the ambiguity of the information upon which the experts base their judgement or the judgemental uncertainty arising under various imprecise and incomplete information. For the treatment of such type of uncertainty, a new approach based on fuzzy set theory is proposed. Then its potential use to the uncertainty analysis of the current PRA is proved through an analysis of accident progression event tree (APET). As a product, a formal procedure with computational algorithms suitable for application of the fuzzy set theory to the APET analysis is provided. Comparing with the uncertainty analysis results obtained by the statistical approach currently used in PRA, the present approach has several major advantages: Firstly, it greatly enhances the robustness in the final results of APET uncertainty analysis by modeling the judgemental uncertainty that arises in the probabilistic quantification of APET top events. Secondly, the modeling of APET uncertainty analysis is far more convenient because of the nonprobabilistic features of fuzzy probabilities used for uncertainty quantification of the APET top events. Thirdly, the APET model can easily be operated by means of a well defined formal propagation logic of fuzzy set theory without going through a tedious sampling procedure. Finally, the fuzzy outcomes provide at least as much information as the existing methods based on the statistical approach. Thus, the present approach can be used as a valuable alternative approach to uncertainty analysis used in the current PRA. Two importance measures for the importance analy...