Optimal double-layer preventive maintenance decision-making of natural gas pipeline networks based on gas supply reliability and vulnerability assessment

In the pipeline network, the degradation of pipelines and compressor stations may result in low-probability but high-impact events, such as natural gas supply shortages and leaks. If maintenance decisions are based solely on gas supply reliability, the low probability may downplay their severity, leading to misguided decisions. To address this, this study introduces the concept of vulnerability to quantify degradation consequences and studies the double-layer maintenance decision-making problem, incorporating both gas supply reliability and vulnerability assessments. First, to account for the continuous and discrete variations in gas transmission capacity resulting from pipeline and station degradation, various indicators are developed to model performance vulnerability. A device vulnerability assessment model is established by integrating structural vulnerability, evaluated using topological indicators, with performance vulnerability. Second, gas supply reliability and vulnerability serve as the criteria for triggering network maintenance and for grouping devices for maintenance, respectively, forming a double-layer maintenance decision model. To account for the interdependencies and constraints between network and device-layer decisions, a heuristic double-layer nested optimization algorithm is proposed. Finally, two case studies of the pipeline networks demonstrate that the proposed model can minimize the occurrence of low-probability, high-consequence events by increasing maximum supply capacity and reducing corrective maintenance costs.