Random risk assessment model and risk-based maintenance decisions for natural gas pipelines

Abstract

Pipeline failures can lead to interruptions in natural gas supply, incur downtime and maintenance costs, and easily trigger major accidents. Corrosion has been identified as a major cause of pipeline failure. The consequences of pipeline corrosion vary depending on the characteristics of the pipeline laying area. However, accident consequence assessment models that can address the randomness in pipeline corrosion are limited. We evaluated the random risk of natural gas pipelines, while considering the failure probabilities and consequences, and constructed a non-periodic preventive maintenance decision model based on the corrosion state and potential risk. First, we developed a random process model for simulating the corrosion of natural gas pipelines; a random risk assessment model was established by considering the failure probability with respect to the random corrosion depth and failure consequence with respect to the random corrosion length. Second, based on the results of random risk assessment, a preventive maintenance policy was developed pertaining to the corrosion conditions and future risks. Third, a maintenance decision model, with the aim of minimising the average maintenance cost in a finite horizon, was established to determine the optimal preventive maintenance risk threshold and inspection cycle for natural gas pipelines. Finally, we conducted an experiment while considering the corrosion data of a real pipeline. The results indicated that the proposed policy can effectively control pipeline failure risk during operation and reduce maintenance costs.