Reliability of quantitative risk analysis through an industrial case study.

Purpose: The purpose of this paper is to analyze the reliability of the quantitative risk model used for planning inspection and maintenance activities. The objective is to critically discuss the factors that contribute to the probability and consequence of failure calculations. Design/methodology/approach: The case study conducted using one of the most widely deployed risk models in the oil and gas industry where a full assessment was performed on an offshore gas producing platform. Findings: The generic failure frequencies used as the basis for calculating the probability of failure are set at a value representative of the refining and petrochemical industry's failure data. This failure database does not cover offshore. The critical discussion indicated the lack of basis of the coefficient of variances, prior probabilities and conditional probabilities. Moreover, the risk model does not address the distribution of thickness measurements, corrosion rates and inspection effectiveness, whereas only overall deterministic values are used; this requires judgment to determine these values. Probabilities of ignition, probabilities of delayed ignition and other probabilities in Level 1 event tree are found selected based on expert judgment for each of the reference fluids and release types (i.e. continuous or instantaneous). These probabilities are constant and independent of the release rate or mass and lack of constructed model. Defining the release type is critical in the consequence of the failure methodology, whereas the calculated consequences differ greatly depending on the type of release, i.e. continuous or instantaneous. The assessment results show that both criteria of defining the type of release, i.e. continuous or instantaneous, do not affect the calculations of flammable consequences when the auto-ignition likely is zero at the storage temperature. While, the difference in the resulted toxic consequence was more than 31 times between the two criteria of defining the type of release. Research limitations/implications: There is a need to revamp this quantitative risk model to minimize the subjectivity in the risk calculation and to address the unique design features of offshore platforms. Originality/value: This case study critically discuss the risk model being widely applied in the O&G industry and demonstrates to the end-users the subjectivity in the risk results. Hence, be vigilant when establishing the risk tolerance/target for the purpose of inspection and maintenance planning. [ABSTRACT FROM AUTHOR]