Dynamic safety assessment of natural gas stations using Bayesian network

Area of Science:

• Engineering
• Risk Management
• Public Safety

Background:

• Natural gas pipelines are crucial for energy transport but pose risks in urban areas.
• Existing risk analysis methods are insufficient for natural gas stations.
• Urban development increases the threat to public safety and infrastructure from gas transportation.

Purpose of the Study:

• To develop a dynamic and comprehensive quantitative risk analysis (DCQRA) approach.
• To model accident scenarios and assess risks specifically for natural gas stations.
• To provide a robust methodology for enhancing safety in gas transportation infrastructure.

Main Methods:

• Failure Mode and Effects Analysis (FMEA) for hazard identification.
• Bow-tie diagrams for modeling worst-case accident scenarios.
• Bayesian networks for quantitative risk assessment.

Main Results:

• The failure of the regulator system is identified as the most critical accident scenario.
• Human error is the most significant contributing factor to these failures.
• The DCQRA approach effectively models complex risks in natural gas stations.

Conclusions:

• Prioritizing risk management efforts on human error and regulator system failures is crucial.
• The developed DCQRA methodology enhances the safety of natural gas stations.
• Implementing findings can significantly reduce accident probabilities and associated risks.