Integrating fuzzy expert system and scoring system for safety performance evaluation of offshore oil and gas platforms in Malaysia

Abstract

This study proposes an integrative safety performance evaluation framework using both scoring and fuzzy expert systems for offshore oil and gas platforms in Malaysia. Fourteen safety factors are used as the basis of performance evaluation. To generate the safety score of each safety factor, weights of the indicators grouped under the safety factor are first multiplied with the compliance levels assigned to the respective indicators. The products of weights and compliance levels are then summed up, yielding the safety score. The scores of safety factors serve as inputs of the fuzzy expert system to determine the overall safety performance of an offshore oil and gas platform.The scoring and fuzzy expert systems are tested on ten offshore oil platforms in Malaysia. All the platforms tested meet the status of ‘compliant’ defined by the rules of the expert system. The crisp outputs correspond well with the total safety scores of the platforms. However, irregularities are observed in the crisp outputs when tested with various compliance scenarios, indicating that rules set by the experts have larger effect on the outputs than the total safety scores obtained. All scenarios of membership function used for the inputs and outputs do not demonstrate significant difference in the R-squared values obtained from the plots of total safety scores against crisp outputs.This study contributes to identification of the most appropriate set-up of fuzzy expert system which can be integrated with this framework to permit integrative safety performance evaluation for early warning provision and performance benchmarking. This paper presents a novel use of fuzzy expert system in safety performance measurement of offshore platforms which contributes significantly to the management and retention of experts' knowledge in decision-making pertaining to safety performance.