A novel resilience modeling method for community system considering natural gas leakage evolution

Abstract

With rising natural gas demand, the issue of emergency management of gas leaks in communities is becoming more prominent. Resilience engineering is a successful approach to improving the system's ability in dealing with emergencies. In contrast to natural disasters such as floods and tornadoes, the consequences of a gas leak are closely related to the accident's evolutionary path. However, few reported works have taken the evolution of disasters into account in the assessment of resilience. Conventional methods fail to quantify the public safety performance of a disturbance caused by a natural gas leak. This work presents a novel dynamic approach to assessing resilience that incorporates the evolution of an incident and its interaction with emergency measures. A network structuring model of accident evolution is developed by the Functional Resonance Analysis Method (FRAM) to analyze the potential accident propagation. The explosion consequence of gas leakage escalation is simulated based on Computational Fluid Dynamics (CFD), and the personnel injury criterion is adopted to quantify the temporal and spatial variation characteristics of the system degradation. Then, the dynamic Bayesian network (DBN) is then used to track the interactions between accidents and emergency measures. Taking a real accident case (Shiyan underground gas explosion) as an example, the case study indicates that the proposed method can identify and prioritize emergency measures, as well as provide strong support for decision-making and arrangements in emergency management.