Gholam Abbas Shirali
Ahvaz Jundishapur University of Medical Sciences, Iran
Title: Application of Bayesian network in estimating human error probability: A case study in a petrochemical plant
Biography
Biography: Gholam Abbas Shirali
Abstract
Along with the improvement of equipment reliability, human error has become a great threat to the oil industry reliability and safety. Statistics show that human error is a major contributor to over 80% of accidents in chemical and petrochemical industries. Therefore, in order to ensure effective prevention of catastrophic accidents, the role of human in accident dynamics should be considered during risk assessment processes. The purpose of this study is to provide a method for estimating the instant and precise of human error probability (HEP) using cognitive reliability and error analysis method (CREAM) and Bayesian network. For this purpose, data related to dynamic context (or common performance conditions) was collected by a self-design questionnaire. Then, the gathered data was processed via MSBNx software. The results indicate that the highest HEP value is associated to the outside operators with 0.0912. In this study, factors such as unavailability of procedures/plans, multiple simultaneous goals, inadequacy of training and experience, and poor crew collaboration were identified as the common performance conditions that could effect on the HEP. Therefore, the method can be used as a useful and applicable tool to estimate the HEP value, in particularly in complex and uncertain systems such as oil industries.