The Assessment of Containment Functional Failure Frequency in the Revised Level 2 PRA Standard in Japan

Abstract

In this paper, we introduce the overview of the requirements and the complementary information on the evaluation of containment functional failure frequency (CFF) in the revised version of “A Standard for Procedures of Probabilistic Risk Assessment of Nuclear Power Plants during Power Operation (Level 2 PRA) “[1] in Japan, which was developed and revised at the Level 2 PRA Subcommittee under the Atomic Energy Society of Japan (AESJ). Although the Level 2 PRA standard includes the evaluation of CFF and radiological source terms, we explain only the evaluation of CFF in this paper. In the evaluation of CFF, the physical response analysis and the probabilistic analysis are included as follows. The accident progression analysis is performed for each of the plant damage states, considering the operation status of mitigation systems, thermal-hydraulic behavior and core damage progression, and occurrences of some key events such as reactor pressure vessel failure. The containment event tree (CET) is developed classifying the accident progress in tree diagram. In the CET, some headings are arranged sequentially considering the accident progression. The headings correspond to the phenomena occurrence and the systems operation status, and a branch probability is assigned at each branch of heading. The branch probabilities of the phenomena are evaluated by either the Risk Oriented Accident Analysis Methodology (ROAAM) or the Decomposition Event Tree (DET) analysis considering the containment threats. The branch probabilities on the phenomena are set as the probability distributions, because the phenomena and the analysis have uncertainties. The branch probabilities on the systems operation are evaluated using the fault tree analysis and human error analysis. The containment functional failure modes are assigned at the end state of the CET considering the type of load against containment integrity. For the evaluation of the non-energetic load, the integral codes such as MELCOR [2], THALES-2 [3], and MAAP4 [4] etc. are used. On the other hand, various mechanistic codes are used for the evaluation of energetic phenomena such as steam explosion. The containment functional failure is judged by comparing the ultimate strength or the fragility of containment structure and the generated loads. After all, the CFF can be obtained by summing the frequency of containment functional failure mode. In the Level 2 PRA standard in Japan, the requirements in each evaluation process above are described. In addition, the technical background and the examples as the complementary information on each requirement are described in the Annex of the standard to help the application of the standard. In this revision, the body is revised to clarify the requirements on the quantification of the CET. The Annex is revised to incorporate the up-to-date information on severe accident research and severe accident management (SAM) measures. The updated information includes the melt stratification (OECD/MASCA project [5]), the steam explosion (SERENA project [6] and PULiMS/SES experiments [7]), the ex-vessel debris coolability (OECD/MCCI project [8]), debris jet breakup, the melt spreading, the coolability of the particulate bed, and the containment vessel (CV) fragility evaluation. Some future challenges are extracted from the lessons learned from the Fukushima Daiichi accident, such as development of the Level 2 PRA for the external hazard as earthquake and tsunami, quantification of impact on the containment integrity of hydrogen detonation in the adjacent buildings, and human error evaluation in the external hazard.