Welcome to the PSAM 18 Abstract Status page.
Abstract NA251Full Paper + Presentation
Why Are Japanese Nuclear Power Plants “So Safe”? — Insights from Safety Improvement Assessments
Following the accident at Tokyo Electric Power Company’s Fukushima Daiichi Nuclear Power Station, the Japanese regulatory authority required licensees to perform Probabilistic Risk Assessment (PRA) every five years under the Periodic Safety Assessment of Continuous Improvement System introduced in December 2013.
For pressurized water reactors that have resumed operation, Level 1 and Level 2 PRAs have been conducted for at-power internal, seismic, and tsunami events, together with a Level 1 PRA for shutdown internal events. The reported core damage frequency (CDF) generally falls between 1E-6 and 1E-5 per reactor-year, while the containment failure frequency (CFF) generally falls between 1E-7 and 1E-6 per reactor-year. These values are substantially lower than those reported for comparable plants in other countries. Such extremely low values are difficult to attribute solely to differences in plant design, especially among plants of similar type and generation, suggesting that the gap is more likely attributable to assumptions, modeling choices, scope definitions, and input data adopted in the PRA.
This study therefore investigates why the reported risk metrics of Japanese light water reactors are markedly lower than those of comparable foreign plants, based on publicly available information from licensees’ PRA reports. In particular, it aims to clarify, at the level of PRA modeling practice, how the “1E-06 culture” pointed out by Apostolakis [1] is manifested in actual assessments and embedded in model assumptions and quantification approaches. Based on these findings, we identify key features of Japanese PRA practice and discuss the issues that need to be addressed to further advance risk-informed applications in Japan.
References
[1] George Apostolakis, “The Evolution of the Use of PRA and Risk-Informed Decision Making in Japan,” Proceedings of Probabilistic Safety Assessment and Management 17 & Asian Symposium on Risk Assessment and Management 2024 (PSAM17&ASRAM2024), PSAM17&ASRAM2024-PL01, Sendai, Japan, October, 2024.
← Check another abstractFor pressurized water reactors that have resumed operation, Level 1 and Level 2 PRAs have been conducted for at-power internal, seismic, and tsunami events, together with a Level 1 PRA for shutdown internal events. The reported core damage frequency (CDF) generally falls between 1E-6 and 1E-5 per reactor-year, while the containment failure frequency (CFF) generally falls between 1E-7 and 1E-6 per reactor-year. These values are substantially lower than those reported for comparable plants in other countries. Such extremely low values are difficult to attribute solely to differences in plant design, especially among plants of similar type and generation, suggesting that the gap is more likely attributable to assumptions, modeling choices, scope definitions, and input data adopted in the PRA.
This study therefore investigates why the reported risk metrics of Japanese light water reactors are markedly lower than those of comparable foreign plants, based on publicly available information from licensees’ PRA reports. In particular, it aims to clarify, at the level of PRA modeling practice, how the “1E-06 culture” pointed out by Apostolakis [1] is manifested in actual assessments and embedded in model assumptions and quantification approaches. Based on these findings, we identify key features of Japanese PRA practice and discuss the issues that need to be addressed to further advance risk-informed applications in Japan.
References
[1] George Apostolakis, “The Evolution of the Use of PRA and Risk-Informed Decision Making in Japan,” Proceedings of Probabilistic Safety Assessment and Management 17 & Asian Symposium on Risk Assessment and Management 2024 (PSAM17&ASRAM2024), PSAM17&ASRAM2024-PL01, Sendai, Japan, October, 2024.