Fuzzy Bayesian network approach for reliability and sensitivity analysis of a multistate nuclear power plant under common cause failures

Reliability assessment of Nuclear Power Plants (NPPs) is important for ensuring operational safety and efficiency. This study presents a novel integration of Fuzzy Bayesian Network (FBN) with Common Cause Failure (CCF) modelling for multistate NPP reliability assessment under uncertainty. This study uses a Fuzzy Bayesian Network (FBN) approach to assess the reliability and sensitivity of a multistate Nuclear Power Plant (NPP). It considers Common Cause Failures (CCF) by combining Bayesian Networks with fuzzy probability. This helps to model uncertainties and the relationships between system components. The reliability analysis examines three operational states: fully operational, degraded performance, and failure. It also compares system performance in cases with and without CCF. The posterior probability distribution of root nodes is computed to determine the influence of individual components on system states, while sensitivity analysis identifies critical components affecting system reliability. The results show that the probability of the fully operational state increases, while the probabilities of degraded and failure states decrease when CCF is considered. The main contribution of this work is that it develops a simple and effective framework that considers multiple system states, uncertainty, and component dependencies together by including CCF within a single FBN model. The findings indicate that incorporating CCF improves overall system reliability by reducing the likelihood of failure and degraded performance. Sensitivity analysis indicates that the Steam Generator, Coolant System, and Condenser have the highest impact on system failure, emphasizing their importance in maintenance prioritization. The proposed approach provides a systematic framework for analyzing multistate NPP system under uncertainties.