Experience gained in the petroleum activities have showed that major accident risk is inherent in daily activities. Risk influence methodology is perceived as a good candidate to model the activity-related risk, as a key input to operational planning decisions. The paper reviews and summarizes 11 risk influence frameworks that integrate organizational and human factors in a structured way. The intention was to evaluate how these frameworks and identified risk influencing factors (RIFs) can be used for activity-related risk analysis. The main conclusion is that it is not necessary to model explicitly RIFs for activity consequence risk -the effect that performing an activity will have on the risk level after the activity has been completed. Operational management RIFs, direct organizational RIFs, personal risk influencing factors, task characteristics RIFs, technical system RIFs, and environmental RIFs are relevant for activity performance risk -the risk associated with performing the action. Operational management RIFs influence planning, which is important to identification of interactions while estimating period risk -the risk for a plant or facility over a period.

No abstract

In the petroleum industry, the study and impact of human reliability on safe and efficient operations is receiving heightened awareness. According to recent studies, human errors contributed to an estimated 60 to 90 percent of all accidents across many industries (petroleum, nuclear and aviation industries, etc.). Human errors can result in two possible outcomes of concern: safety risks and the cost of failure. In operations where the consequences of accidents may be unbearable, the emphasis on safety drives the study and research of human reliability. This effort has led to a number of human reliability analysis (HRA) methods, and they provide useful references for managing human errors. This paper examines how the petroleum industry could utilize and benefit from the available HRA methods by applying them to daily operations. Several HRA methods are reviewed in this paper from three aspects: human error identification, human error prediction of given tasks, and factors that contribute to human errors. A good data collection system can substantially support the HRA study qualitatively and quantitatively. This paper also introduces a practice for human error data collection that was implemented in the nuclear industry—human event repository analysis (HERA). Suggestions and remarks are provided about developing similar human error data collection systems tailored for the petroleum industry. The expected benefits of HRA include a better understanding of the causes and influencing factors of human errors in specific operation conditions, and insights into error-reduction measures for reducing accident risk and the cost of product failure.

No abstract

No abstract