Data-Driven Prognostics for Major Piping in Nuclear Power Plants

Abstract

As operation period of Nuclear Power Plants (NPPs) is getting longer, necessity of reflecting ageing effect is increasing. Especially, when it comes to the piping in NPPs such as reactor coolant system piping or steam generator
tubes, it is vulnerable to stress corrosion crack (SCC) or wear due to the fluid with high temperature, high pressure and radiation. Accidents related to such cases have been reported. Since ruptures of the piping can result in severe accidents, it is important to predict and prevent them in advance. Current NPPs ageing management is performed with the physical model based on generic experimental data, which cannot properly consider each NPPs’ different operation environment or history. Prognostics using plant specific data can compensate this limit of ageing management using the physical model. Recently, as usable data of NPPs is increasing with the development of instrumentation technology, applicability of prognostics for NPPs has been increased. Therefore, this paper suggests some prognostics methods such as GPM (General Path Model), MCMC (Markov Chain Monte Carlo) and Particle filter that can consider ageing degradation for the major piping in NPPs. It is expected that prognostics results can be used in Probabilistic Safety Assessment (PSA) considering current or future ageing degradation.