Human Reliability Analysis for the Petroleum Industry: Lessons Learned from Applying SPAR-H

摘要

Human error has shown to be a contributor to a number of major accidents in the petroleum industry. However, Quantitative Risk Analyses (QRA) have only to a limited extent taken into account the contribution of human performance to major accident risk. Human Reliability Analysis (HRA) has the potential to overcome this by systematically analysing human performance for safety critical tasks. A joint effort between industry and academia is underway in Norway to adapt SPAR-H, an HRA technique from the nuclear industry, to a petroleum setting (PetroHRA). This paper discusses some of the lessons learned so far in directly applying the technique to a petroleum case study. A case study was performed where the operator’s task was to manually activate the platform’s depressurisation system upon detection of a hydrocarbon leakage. The factors influencing the performance (Performance Shaping Factors; PSF) of the operator were analysed indicating the potential contributors to operator failure. The PSFs were: time, stress/stressors, complexity, experience/training, procedures, HMI/ergonomics, fitness for duty, and work processes. A number of issues were identified in when directly applying of SPAR-H. These were challenges in deciding on the multiplier of the PSFs, the potential for overlap between the PSFs, the industry specificity of the HMI/Ergonomics PSF description and the method’s tendency to inflated Human Error Probabilities (HEP). A first step to improve the definitions and guidance material for a petroleum-specific SPAR-H was taken in a separate literature study performed by two of the authors of this paper (Rasmussen, Standal, & Laumann, in submission). The lessons learned presented in this paper provided a basis for further development of SPAR-H and its guidance material. It was further shown that directly applying and integrating SPAR-H in QRA in a petroleum context is possible and relatively straightforward. In addition, the qualitative outcomes provided a structured and meaningful understanding of human performance previously not available to QRA. Ultimately, this effort contributes to a further integration of HRA and QRA and therewith providing a more complete picture of the factors contributing to major accident risk.