Identification of High-Risk Zones in Natural Gas Pipeline Networks Using Hotspot Detection and Spatial Analysis

Reliable and efficient pipeline networks are crucial for supplying natural gas to end users. Pipeline networks are cost-effective but susceptible to numerous risks, notably corrosion, natural hazards, and human-induced interventions, leading to gas supply interruptions, environmental damage, and threats to property and human lives. Identifying the high-risk areas prone to recurring failures is crucial for efficient risk management. This research presents a novel spatial analysis approach to identify hotspots using historical failure incident data from the Pipeline and Hazardous Materials Safety Administration database (2011–2023). The proposed framework utilized Global Moran’s I to assess spatial autocorrelation and Getis-Ord Gi* statistics to detect localized clusters of failure events. Results show that high-risk areas are concentrated where pipeline failures exhibit strong clustering patterns. Using the buffer creation technique, example zones with high incidents in the past 12 years have been identified to perform the cause and release matrices analysis, searching for the next probable cause and type of failure and endorsing the effectiveness of the proposed framework. The identified failure hotspots lay a foundation for future research on failure propagation modeling for analyzing pipeline infrastructure resilience.