Seismic Risk Assessment of Natural Gas Networks considering Cascading Effects

Earthquake disasters may not only damage buildings but significantly influence the standard of living due to the impacts on critical infrastructure such as lifeline systems. This study focuses on the seismic risk and resilience of the natural gas network and the cascading effects. We analyze the risk of failure in the mid/low-pressure pipelines under a major earthquake scenario and evaluate the impacts on systemic service levels considering the secondary disasters of earthquakes. In this study, repair rate, R.R., is applied to evaluate the failure of natural gas pipeline. With the R.R. of pipelines and ground motion parameters, e.g., PGA and PGD, the failure probability of the pipeline can be derived by Poisson distribution. By overlay analysis with seismic parameters from Taiwan Earthquake Loss Estimation System (TELES) and the GIS data of the natural gas network, the number of damaged pipelines, the number of affected users, and the closure probability of valves can be estimated through Monte Carlo simulation. The service level and resilience of the system can be further assessed. In addition to the impacts on the natural gas system, the leaking gas can also cause a potential risk of worsening post-earthquake fire. On the other hand, the repairment of damaged pipelines may affect the surrounding traffic. These should be considered during the restoration process. This study proposes a risk assessment approach for the natural gas pipeline subjected to earthquakes considering not only the physical damage of the pipeline but the closure of valves, the risk of worsening post-earthquake fire, and the sequential influence on the traffic. The proposed framework was applied to the natural gas system in Tainan, Taiwan, as the case study. This study assesses the earthquake hazard risk of natural gas pipelines from the perspective of system functionality and community resilience. Decision-makers can plan appropriate disaster mitigation strategies based on the analysis results.