Development of the Bow-Tie Bayesian Network Model for Assessing Pipeline Failures

Recently, oil and gas pipeline leaks in mountainous areas have occurred occasionally. A dynamic risk assessment system has been established to better understand and reveal the risk evolution of such accidents to monitor and predict the likelihood of pipeline leakage incidents. Using the bow-tie model, risk factors are identified, and accident consequences for buried aviation fuel pipelines in mountainous areas are analyzed. The GeNIe software is then used to transform the bow-tie model into a Bayesian network to analyze dynamic risks. The node variables in the model are examined before probability and sensitivity analysis, allowing for the identification of the key factors leading to pipeline leakage incidents. The research results indicate that the high-risk basic events, ranked from highest to lowest risk, are untimely handling, uneven ditch bottom, unintentional damage by personnel, mechanical damage during pipeline installation, failure to strictly enforce procedures, illegal construction, and untimely detection. Therefore, to prevent leakage accidents of aviation fuel pipelines in mountainous areas, it is essential to establish an efficient monitoring and response mechanism, leverage advanced technology to monitor pipeline conditions in real time, and ensure early detection and timely resolution of issues. Additionally, enhancing personnel safety awareness, strengthening pipeline quality reviews, and conducting regular inspections and maintenance of line markers (such as mileage piles, corner piles, signposts, and warning signs) are critical. Lastly, strict quality control and construction environment management should be implemented to optimize pipeline installation and maintenance processes, effectively reducing the risks associated with mechanical damage and terrain-related issues.