Dynamic and interactive approach to level 2 PRA using continuous Markov process with Monte Carlo Method

Jang, S.*; Yamaguchi, Akira*; Takata, Takashi  

The current approach to Level 2 probabilistic risk assessment (PRA) using the conventional event-tree (ET)/fault-tree (FT) methodology requires pre-specifications of event order occurrence and component failure probabilities which may vary significantly in the presence of uncertainties. In the present study, a new methodology is proposed to quantify the level 2 PRA in which the accident progression scenarios are dynamic and interactive with the instantaneous plant state and related phenomena. The accident progression is treated as a continuous Markov process and the transition probabilities are evaluated based on the computation of plant system thermal-hydraulic dynamics. A Monte Carlo method is used to obtain the resultant probability of the radioactive material release scenarios. The methodology is applied to the protected loss of heat sink accident scenario of the level 2 PRA of a generation IV fast reactor.