Influence evaluation of sampling methods of the nondestructive examination on failure probability of piping based on probabilistic fracture mechanics analysis

摘要

Non-destructive examinations (NDEs) have an important role in assurance of the structural integrity of nuclear components including pipe lines. NDEs are performed for welds in piping in accordance with the rules, which prescribe the requirements of NDE such as interval and extent of examination. In general, there are two kinds of sampling method for selecting welds to be examined in each interval considering the specified extent of examination. The first method is the fixed location sampling method, in which welds for NDEs are same in each interval, while the second method is the random location sampling method, in which welds for NDEs are selected from those not examined in the last interval. The selection of the sampling method is important in assuring the structural integrity of piping. Probabilistic fracture mechanics (PFM) analysis, which is one of rational structural integrity assessment methods, can quantitatively calculate failure probability of welds in piping by considering aging degradation mechanisms, such as stress corrosion cracking and fatigue, as well as crack detections and repair of cracked welds through NDE. In some countries, especially the United States, PFM approach has been applied to NDE, such as risk-informed in-service inspection. From such backgrounds, further application of PFM approach is increasingly expected. In this study, to investigate the applicability of PFM approach on NDE, we focused on the difference in sampling method in NDE for piping and calculated the failure probability of a typical nuclear piping considering NDEs based on the two sampling methods through PFM analysis. From the results, the quantitative influence of the difference in sampling method on failure probability was clarified.