[3G06-08] Dynamic PRA

[3G08] Study on the Applicability of Dynamic Level 2 PRA to Estimating Large Early Release Frequency

*Xiaoyu Zheng¹, Shogo Takahara¹, Hitoshi Tamaki¹, Tomoyuki Sugiyama¹, Yu Maruyama¹ (1. Japan Atomic Energy Agency)

Keywords:Dynamic PRA, Source term, MELCOR2.2/RAPID, Large early release frequency, Safety margin

Obtaining from conventional probabilistic risk assessment (PRA), risk information inevitably includes uncertainties originating from the variability such as order and timing of event occurrences. Dynamic PRA is a promising approach to reducing the uncertainties. It tightly combines deterministic simulations and probabilistic methods, so system dynamics characterizing dependencies and stochasticity can be explicitly modeled. In this study, using Monte-Carlo-simulation-based dynamic PRA, authors simulate time distributions of core damage and source term release for various accident sequences, and estimate the risk metrics of large early release frequency (LERF). With evaluating the time-dependency of accidental influences, the research can provide valuable risk information for determining disaster management plans.