Validation of an offshore occupational accident frequency prediction model—A practical demonstration using case studies

A model has been developed to predict the frequency and associated costs of occupational accidents in the offshore oil and gas industry. Model inputs include: (i) direct factors, such as quality of personal protective equipment; (ii) corporate factors, such as training program effectiveness; and (iii) external factors, such as royalty regime. Three applications of the model are described, two for projects in eastern Canada and one for the Gulf of Mexico drilling sector. Expert opinion is used to provide the required model input associated with the regions' safety programs. Published accident data are used to calibrate the model and validate results. The model is shown to predict actual results well, especially considering the subjective nature of the activity. The model's versatility is demonstrated through its application to different types of accident statistics and regions, and its use in generating performance measures for operators. © 2006 American Institute of Chemical Engineers Process Saf Prog, 2006