Assessing Pipeline Failure Hotspots and Vulnerabilities: A Spatial and Probabilistic Risk Analysis

Due to low carbon emissions and widespread availability, natural gas is a crucial source of energy. Pipelines are the most cost-effective and reliable means of transportation, but they also pose significant risks due to various external and internal factors leading to pipeline failure. To mitigate these risks, it is crucial to isolate the failure hotspot, as incidents often recur in high-risk areas. This study, utilizing historical data from the Pipeline and Hazardous Materials Safety Administration (PHMSA), focuses on identifying failure hotspots using a three-step process. The first step involves using ArcGIS Pro software to identify incident hotspots through the Getis-Ord Gi* statistical technique. In the second stage, the study evaluates critical nodes and network vulnerabilities to various external and internal factors. In the final stage, the Bayesian network is utilized to check the failure probabilities of the most vulnerable nodes, followed by sensitivity analysis through diagnosis inference. The results of this study have significant inferences for pipeline management and safety strategies. By identifying key network vulnerabilities and failure hotspots, the study underscores the need for future research to integrate real-time data and additional response variables to further enhance predictive capabilities and prevent pipeline incidents.