Probabilistic Safety Assessment (PSA) have been steadily evolving to acquire insights and reflect lessons-learned into accident scenarios with the best realistic modelling of a system.
The reliability analysis method such as PSA should be able to take into account time dependency in quantifying the probability of functional success or failure of the systems. Practically it is difficult to deal with the time dependency when branching success/failures of a large-sized system, so in a complex system such as Nuclear Power Plants (NPPs), it is widely accepted to take some conservative results from the deterministic method (physical model), and then to treat the time dependency somewhat loosely when probabilistically checking the functional performance of the system to be. However, when the conservatism of the dualistic deterministic and probabilistic methods is combined, it is not always possible to obtain conservative results, or excessively high conservatism may mislead in decision-making.
This study develops a reliability analysis tool named DICE (Dynamic Integrated Consequence Evaluation), and it basically helps the deterministic method and the probabilistic method operate collectively according to time. We have completed the conceptual design for the scheduler, the branching rules, and the branch quantification models based on fault trees. This paper will introduce the development processes and discuss its applicability for checking the coverability of operational procedures.