Optimization of Overhaul Interval of Motor-Operated Valve in Safety Systems for BWRs Estimated by Bayes Analysis

Abstract

  When the maintenance program for components in a nuclear power plant is drawn up, it is important to evaluate and optimize the overhaul interval of components based on operation and maintenance experiences. Motor-operated valve (MOV) data in safety systems for BWRs are selected from the operation and maintenance database for Japanese nuclear power plants called NUClear Information Archives (NUCIA). The time dependence of the failure occurrence rate of MOV is analyzed by Baysian inference using a log-linear and regression model of nonhomogeneous Poisson process. Then, its optimal overhaul interval of MOV is evaluated based on the time-dependent failure occurrence rate as determined by Bayes analysis. This Bayes analysis is found to be effective for estimating the optimal decision for the overhaul interval of MOV in an actual nuclear plant. This analysis will also contribute to optimizing the overhaul interval of any components in a nuclear power plant by using the data in NUCIA.