Study on evidence-based LPSA method in nuclear power plant under abnormal operating conditions

Abstract

Abstract Living probabilistic safety assessment (LPSA) is an online risk assessment technology, which requires updating PSA models manually or automatically in time when a NPP system is abnormal or its configuration changes, so that the assessment results reflect the actual states of the plant changes in operational risk. So far, most LPSA model updates are based on the assumption that the information used for model updates is correct, and little attention has been paid to the impact of the uncertainty of the information used to update the LPSA model. This paper presents a Bayesian method based diagnostic process for evaluating the impact of personnel judgment sufficiency and the plant information authenticity on the risk assessment by LPSA under abnormal operation conditions. The proposed evidence-based method provides a solution to analysis the impact of the accuracy of plant information and the completeness of alarm response procedures by the operators on NPP risk.