Enhanced Bayesian THERP: Lessons learnt from HRA benchmarking

The Enhanced Bayesian THERP (Technique for Human Reliability Analysis) method uses as its basis the time-reliability curve introduced in the Swain’s human reliability analysis (HRA) handbook. It differs from the Swain's Handbook via a transparent adjustment of the time-dependent human error probabilities by use of five performance shaping factors (PSFs): (1) support from procedures, (2) support from training, (3) feedback from process, (4) need for co-ordination and communication, (5) mental load, decision burden. In order to better know the characteristics of the Enhanced Bayesian THERP from a more international perspective, the method has been subject to evaluation within the framework of the international “HRA Methods Empirical Study Using Simulator Data”. Without knowledge of the crews’ performances, several HRA analysis teams from different countries, using different methods, performed predictive analyses of four scenarios. This paper gives an overview of the method with major findings from the benchmarking. The empirical comparison gives confidence that the time reliability curve is a feasible and cost effective method to estimate human error probabilities when the time window is well defined and relatively short. The comparison of empirical observations with predictions was found as a useful exercise to identify areas of improvements in the HRA method.