A multi-state Semi-Markov model for nuclear power plants piping systems subject to fatigue damage and random shocks under dynamic environments

• Published in: International journal of fatigue Vol. 168; p. 107448
• Main Authors: Liang, Qingzhu, Peng, Changhong, Li, Xiangyu
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01.03.2023
• Subjects: Dynamic environments; Fatigue degradation modeling; Random shocks; Fatigue degradation modeling; Random shocks; Dynamic environments
• ISSN: 0142-1123
• DOI: 10.1016/j.ijfatigue.2022.107448

•A multi-state Semi-Markov model for degraded piping systems in nuclear power plants was proposed.•The effects of the stochastic environment and random shocks were considered.•The analytical solutions of the time-dependent state probability of the system were derived.•The correctness of the analytical results was verified based on the benchmark case in the literature.•Two case studies were presented. Fatigue degradation is a vital damage mechanism for piping systems in nuclear power plants (NPPs). The accurate fatigue reliability evaluation of piping systems is significant for assessing the safety and reliability of an NPP. In this work, we presented a multi-state model for the fatigue degradation process of piping systems in NPPs, which accounts for the effects of random shocks and dynamic environments. The fatigue degradation of the piping was described by a Semi-Markov process, which allows accounting for generic distributions of the holding times of the system states. The arrival of the shocks was assumed to be governed by a Poisson process. The dynamic environment in which the piping is located was assumed to be governed by a Markov process. The analytical solution of the time-dependent state probability of the piping was derived. We developed a Monte Carlo (MC) algorithm for the simulation of the stochastic process describing the integrated stochastic evolution of the dynamic environment, system degradation and random shocks, which was applied to verify the correctness of the proposed model. The effectiveness of the proposed model was demonstrated by applying the model to numerical illustrations.