Multi-level optimization of maintenance plan for natural gas pipeline systems subject to external corrosion

Abstract Failures of natural gas pipeline systems may result in severe consequences to social security and economic loss due to the combustibility of natural gas. Pipeline maintenance should be applied throughout the lifecycle to ensure that the pipeline system is managed safely and cost effectively. In this context, this study addresses optimal maintenance planning for natural gas pipeline systems subject to external metal-loss corrosion. The corrosion growth of natural gas pipeline systems is described as a Markov process. A multi-level strategy is proposed for the maintenance optimization of pipeline systems subject to external corrosion, which includes repaired Markov states (Level 1), maintenance time (Level 2), and maintenance number during the pipeline lifetime (Level 3). The multi-level optimization maintenance model is presented by synthetizing the corrosion Markov process and the multi-level maintenance strategy. The total cost of natural gas pipeline systems subject to external corrosion can be further decreased through the proposed multi-level optimization method compared with the adoption of traditional methods. Besides, genetic algorithm (GA) is also introduced for the multi-level optimization analysis of natural gas pipeline systems. A comprehensive optimization algorithm based on GA and the Markov process that can accurately and efficiently conduct multi-level optimal maintenance planning is proposed.