(A) method for risk-informed safety significance categorization using AHP and BBN

Abstract

A risk-informed safety significance categorization (RISSC) is used in risk-informed regulation and application (RIRA) to allocate the structures, systems, or components (SSCs) into two or more groups, according to their safety significance using both probabilistic and deterministic insights. In this work, both the probabilistic and the deterministic insights are based on level 1 and 2 internal event probabilistic safety assessments (PSAs) and normal operating conditions. We use both the importance analyses of PSAs and the qualitative assessments based on the deterministic insights as decision factors for the decision problem RISSC. In this work, we proposed a method for risk-informed safety significance categorization using the analytic hierarchy process (AHP) and Bayesian belief networks (BBN). This method introduces early participatory of an expert panel by using the AHP and the BBN to improve the overall decision-making process and perform integrated evaluation quantitatively. The early participatory of the expert panel is applied to perform the RISSC in an early stage on the basis of the expert’s knowledge and experience; the AHP to structure the decision problem RISSC with the decision factors and to obtain the relative importance of the decision factors, that is, the weighting values of the decision factors; the BBN to evaluate the probabilities that a component falls in Hi (high safety significant category) in the respective decision factors as safety significances in the respective decision factors. The weighting values of the decision factors and the probabilities that a component falls in Hi in the respective decision factors are integrated into the total probability. The total probability is used as the final safety significance for the decision-making in the RISSC: if the total probability is greater than 0.6, the safety significance category of the component is High Safety Significance Category (HSSC); if the total probability is smaller than 0.6 ...