The hybrid causal logic methodology for risk assessment: Quantification algorithm

The foundation of scenario modeling in traditional probabilistic risk analysis (PRA) methods, dating back to the seminal WASH-100 study, is built on the integration of event trees (ET), or event sequence diagrams (ESD), with fault tree (FT) models. This ET-FT binary logic framework provides a structured approach to analyzing risk scenarios, with causal depth represented by the "basic events" defining each scenario. Introduced in 2005, the hybrid causal logic (HCL) methodology extended the ET-FT model to incorporate Bayesian Belief Networks (BBN), capturing the logical and inferential dependencies within a single, cohesive model. HCL was developed to address limitations in PRA, especially to account for “soft causal factors†where causation cannot be fully or deterministically established, such as human and organizational failures. This methodology has found applications across sectors, including nuclear, aviation, petrochemical, maritime, space, and healthcare. While the HCL solution algorithm, covered by a U.S. patent and integrated into platforms like Trilith and PSIM, has been utilized extensively, its algorithmic details have not been previously disclosed.