Probabilistic safety assessment (PSA) on a specific reactor are often implemented without considering the applicability of generic reliability data, and doubt about such assessments is aroused because of the lack of plant-specific reliability data. The applicability of generic reliability data is analyzed in present paper, in order to remove the doubt in a way. Several sets of reliability data composing from different sources are researched. The following analysis evaluate a fault tree for a typical example of a reactor, using several sets of reliability data and show the differences in the results. Additionally, a comparison is made with a procedure of analysis using reliability data ranges. The results show that the probabilistic safety analysis on a specific reactor using reliability data which come from different sources is feasible. The differences are slight for most components, only a few key components should be separated in the first place and concentrated more attention on them. The superiority of plant-specific data should be advocated. In the mean time, the lack of data should not be a barrier for PSA on a specific reactor. And the analysis facing a lack of data is advised to be encouraged as an approach to improve the safety of specific reactors.

This content is only available via PDF.
You do not currently have access to this content.