| Popis: |
This paper addresses the progress with piping reliability analysis methods and techniques in the context of probabilistic safety assessment (PSA) to support risk-informed decisions. Piping reliability analysis has been a topic of discussion and concern within the nuclear safety community for a very long time. In part, this concern has been related to the capabilities and limitations of available methods and techniques, as well as with the requirements for how to best perform quantitative analysis in support of safety cases. The team of analysts responsible for the seminal Reactor Safety Study (WASH-1400) performed a limited evaluation of nuclear power plant piping reliability based on service experience from the then approximately 150 U.S. commercial nuclear reactor operating years [1,2]. This evaluation was aimed at estimation of loss-of-coolant-accident (LOCA) frequencies for input to the two PSA models of WASH-1400. After the publication of WASH-1400 in 1975 many other R&D projects have explored the roles of structural reliability models and statistical evaluation models in providing acceptable input to PSA. Furthermore, during the past 15 years efforts have been directed towards establishment of comprehensive pipe failure event databases as a foundation for exploratory research to better understand the capabilities and limitations of today’s piping reliability analysis frameworks. Against a historical overview of past efforts, this paper addresses the question how to best utilize service experience data for quantitative piping reliability analysis. Significant progress has been made to develop pipe failure databases, as well as analysis methods and techniques to explore and analyze the body of service experience with piping from today’s (February 2007) over 12,500 commercial reactor operating years. Insights from recent pipe failure database applications are utilized to reach some conclusions about the capability of statistical analysis approaches to piping reliability analysis, including Bayesian modeling.Copyright © 2007 by ASME |
