| The prediction of human error probabilities and their consequences in offshore platform musters, through a human error probability index (HEPI), provides a proactive quantitative approach for the inclusion of human factors in risk assessments. Due to the lack of human error databases and in particular, human error data for platform musters, an expert judgment technique, the Success Likelihood Index Methodology (SLIM) was adopted as a vehicle to predict human error probabilities for three credible muster scenarios of varying severity.; A panel of twenty-four judges active in the offshore gas industry provided data for both the rating and weighting of six performance shaping factors that influence the probability of success for eighteen muster actions which range from point of muster initiation to the final actions in the temporary safe refuge (TSR). Through the use of predicted human error probabilities and a consequence table specific to muster scenarios, the risk is estimated for each muster action. Risk mitigation measures are provided to allow a re-ranking of risk potential. Predicted human error probabilities are compared to empirical data gathered from three years of muster reports from an existing operation. An example calculation is provided, utilizing HEPI, based on an actual muster event.; Results provide a clear indication of situations that provide the highest risk and which actions are the most likely to end in failure. Through HEPI, error reduction recommendations for training, procedures, management systems and equipment are brought forward, allowing operators and designers to effectively apply human factors in a proactive approach to high risk scenarios. The framework of HEPI has universal application in the offshore industry providing a human factors tool to engineers, operators and health and safety personnel.; This thesis reviews and defines human error and the modes in which it is manifest. A review of past literature is conducted with focus placed on three expert judgment techniques that have enjoyed a degree of acceptability and application. A more detailed review of the SLIM technique is conducted and how this work addresses the concerns of past authors. A statistical review of performance shaping factor weights and ratings is conducted followed by the determination of human error probabilities for three muster scenarios. The results stemming from these calculations are applied in HEPI and are checked against empirical data. Recommendations conclude the work and provide a basis for future research. A separate data book complements this text and provides a listing of the over 15,000 pieces of raw data and associated calculations in the determination of human error probabilities. |
