[1O01-07] Advanced PRA Methodology

[1O07] Research on regional risk analysis method with Continuous Markov Monte Carlo and Bayesian network

(2)Preliminary development of regional risk analysis model

*Sunghyon Jang¹, Akira Yamaguchi¹, Masayuki Hijiya², Kagetomo Miyahara², Kensuke Toyoshima², Satoshi Shinzaki² (1. The University of Tokyo, 2. Nuclear Engineering, Ltd)

Keywords:Continuous Markov and Monte Carlo method, Bayesian network, multi site risk, regional risk

A new approach for multi-site and regional risk assessment has been developed by combining Bayesian network, Markov process, and Monte Carlo method. In this method, Bayesian network was used to evaluate state transition probabilities of components in a multi-unit system with consideration of its operation. The status of the accident progression was estimated by Markov process. A large number of accident scenarios, which are unique to each other, were generated by Monte Carlo method.

In this report, preliminary analysis of the regional risk was carried out for a multi-unit site which consists of a twin-unit system and a support center.