American Nuclear Society 19th International Conference on Probabilistic Safety Assessment and Analysis (PSA 2025)

(Chicago, Illinois)

Paper 7: Human Error Dependency Treatment in Human Reliability Analysis: Research Gaps and Opportunities

By: Tingting Cheng, Ali Mosleh

Human error dependency has long challenged Human Reliability Analysis (HRA), as it skews estimates of human error probabilities (HEPs). Early methods simplify the dependency treatment by applying multipliers, while advanced approaches added cognitive bases but also complexity. Root causes often stem from shared organizational or design factors, yet these remain underexplored. This paper (1) reviews gaps in dependency treatment across error identification, criteria, PIFs, and computational demands, and (2) proposes using Bayesian Belief Networks to model causal mechanisms, offering a pathway to improve modern HRA dependency treatment. PUBLICATION