Development of risk importance measures for dynamic PRA based on risk triplet, 2; Trial measurement of risk importance through dynamic level 2 PRA with RAPID

Zheng, X. ; Tamaki, Hitoshi ; Shibamoto, Yasuteru ; Maruyama, Yu ; Takada, Tsuyoshi  ; Narukawa, Takafumi*; Takata, Takashi*

Traditional frequency-based risk importance measures (RIMs) have demonstrated its practicability in the nuclear regulation. The authors investigate the definitions of existing RIMs and associated applications in risk-informed nuclear regulations, for instance, the risk-informed categorization of structures, systems, and components (SSCs), risk-informed changes to technical specifications, etc. However, when evaluating mitigation effects of accident countermeasures, importance assessments involving consequence and timing has the potential of providing valuable information for decision making. By widely using numerical simulations of possible accident progressions, dynamic PRA enables a straightforward assessment of risk triplets. Recent advancements in the development of dynamic PRA tend to explicitly incorporate the dynamics of accident progression and failure events into risk assessment, and it allows a provision of more detailed risk information. The approach to appropriate estimation of risk importance within this framework has not been established, exposing a significant research challenge in the use of risk information for decision making in the nuclear industry. Possible accident sequences are sampled using RAPID by randomly branching, and risk triplets are quantified, including key quantities such as source term release amount and release timing to the environment, and the associated frequencies. Risk triplets are used to calculate the new RIMs to rank the importance of pivotal headings in the event tree model. As the exemplary results of the analysis, source term release amount and timing are largely influenced by the mode of containment failure and the termination timing of reactor coolant injection. As the conclusion, when issues such as timing or seriousness of consequence are important for judgement, dynamic PRA and the new RIMs is capable of supporting decision making by providing more detailed risk information.