Reliability Analysis of Highway Evacuation Network Post-Earthquake Disaster

In countries with a high population density, evacuating large amounts of people in a short time post-disaster is a critical challenge for emergency response agencies. This paper analyzes the reliability of a highway evacuation network post earthquake disaster. The capacity of a road segment after disasters is modeled as a discrete probability function. A modified stochastic flow algorithm based on k-shortest path algorithm is proposed to calculate the possibility of transporting a given the number of people in a pre-defined time window on a potentially damaged highway network. Northeast Sichuan is used as case study site and real world data are collected to develop multiple scenarios in order to test the proposed methodology. The results show that the reliability of an evacuation network is highly sensitive to the tightness of an evacuation time window and the damage probability of a critical road segment. The proposed methodology illustrates prominent potential of application in regional emergency planning and risk assessment.