<![CDATA[The oil and gas processing industry is critical for meeting global energy demand, yet it remains vulnerable to high-consequence accidents such as fires and explosions. This study presents a probabilistic risk assessment of a vapour cloud explosion and fire case at the BP-Husky Toledo Refinery using accident investigation report data. The objective is to obtain a more comprehensive representation of accident risk by integrating the performance of physical safety barriers and human reliability within a single analytical framework. Physical escalation pathways and consequence scenarios are modelled using event tree analysis, while human error likelihood is evaluated using the success likelihood index method based on identified error modes and performance shaping factors. Both components are then integrated using a Bayesian network to quantify consequence probabilities and examine the effect of adding a human error barrier and proposed barrier improvements. The results show that the consequence distribution is dominated by near-miss outcomes (approximately 57%) with smaller probabilities for flash fire (about 14%), vapour cloud explosion (about 13%), and rupture (about 16%). Incorporating the human error barrier produces only marginal changes in these probabilities. Nevertheless, the integrated approach improves clarity in linking specific human error mechanisms and barrier performance to overall risk and supports structured evaluation of barrier improvements for accident prevention.]]>
Copyrights © 2025
