The Impact of Probabilistic Modeling in Life-Cycle Management of Nuclear Piping Systems

Abstract

Flow accelerated corrosion (FAC) is a serious form of degradation in primary heat transport piping system (PHTS) of the nuclear reactor. Pipes transporting hot coolant from the reactor to steam generators are particularly vulnerable to FAC degradation, such as tight radius pipe bends with high flow velocity. FAC is a life limiting factor, as excessive degradation can result in the loss of structural integrity of the pipe. To prevent this, engineering codes and regulations have specified minimum wall thickness requirements to ensure fitness for service of the piping system. Nuclear utilities have implemented periodic wall thickness inspection programs and carried out replacement of pipes prior to reaching an unsafe state. To optimize the life-cycle management of PHTS, accurate prediction of time of replacement or “end of life” of pipe sections is important. Since FAC is a time-dependent process of uncertain nature, this paper presents two probabilistic models for predicting the end of life. This paper illustrates that the modeling assumptions have a significant impact on the predicted number of replacements and life-cycle management of the nuclear piping system. A practical case study is presented using wall thickness inspection data collected from Canadian nuclear plants.