Application of Binary Decision Diagram Approach for Fault Tree Quantification in Seismic Probabilistic Safety Analysis of a Research Reactor

In most cases of probabilistic safety assessment model quantification, the minimal cut set (MCS) generation technique is effective and fully sufficient. But as the number of high probability events increases, e.g. due to seismic risk assessments, more accurate methods may be necessary to compensate for the overestimation of the core damage frequency resulting from using MCS methods. Furthermore, in some applications, a relevant numerical treatment of dependencies and success in sequence analysis in noncoherent fault trees may also be required to avoid overly conservative results.

To mitigate these issues, this work introduces the binary decision diagram (BDD) method for calculating the exact top event probability. BDD efficiently captures and processes complex Boolean relationships within a fault tree, allowing for more accurate system reliability evaluations. The BDD method is highlighted for its ability to handle dependencies and success branches more accurately than the MCS approach.

This study demonstrates the feasibility and effectiveness of using BDD within the seismic probabilistic safety assessment of a nuclear research reactor. The results suggest that the utilization of this method provides reasonable assurance, allowing for robust decision making regarding real-time risk status with confidence.