A HYBRID METHODOLOGY COMBINING FUNCTIONAL PROBABILISTIC ANALYSIS TO ASSESS THE MAINTENANCE PLAN OF A NATURAL GAS TRANSMISSION SYSTEM: A TURBO-COMPRESSOR POWER PACK CASE STUDY

摘要

A probabilistic analysis is an approach that allows to identify whether if preventive, predictive and corrective maintenance practices are efficient for keeping the overall system safety. The implementation of methodologies that evaluate the maintenance plans are a way to improve existing programs. Promigas S.A. E.S.P. is a company that must guarantee the reliability and availability of its natural gas transmission systems to its consumers. Therefore, since 2009, the company has implemented reliability centered maintenance practices to achieve it. However, considering the current scenario of the natural gas in Colombia and according to the growth plans of the company, it is necessary to guarantee that the reliability and availability indexes remain close to 100%. To achieve the aforementioned objective, it is necessary to diagnose the current maintenance plan. We propose a hybrid methodology combining functional and probabilistic analysis to assess the maintenance plan of a natural gas transmission system, specifically a turbo-compressor power pack. The proposed methodology includes a new priorization method to identify and select critical components and its critical failure modes, through a qualitative functional and quantitative characterization of the subsystems that conform the turbo-compressor power pack. The probabilistic analyses were simulated for five time periods: one, three, six, nine and thirteen years. The results allow to conclude in terms of availability that while the maintenance plan is optimal for the first-time period, from the second time period the preventive and predictive maintenance practices must be optimized increasing resources or modifying the equipment intervention frequencies.