Infrastructure outages after disasters can cause contrasting impacts throughout a community due to differing vulnerabilities. The identification of critical social impact nodes for use in infrastructure risk mitigation planning and restoration will allow for overall disaster impact reduction and lessening of impact disparity across a community. To achieve criticality identification, this paper presents a probabilistic methodology to quantify the service outage impact of every node. Modelling of impact due to outages is accomplished by fixating outage occurrence in existing impact models and randomly sampling other input variables. The impact variability due to each node’s outage is captured through conditional impact probability density functions (CIPDFs) that are first developed with Monte Carlo simulation at base nodes. Characterization of upstream CIPDFs is achieved through convolution of downstream nodes’ CIPDFs. Mean impact curves/surfaces are developed from CIPDF and nodal vulnerability information for all components to further characterize the impact expected at a certain hazard level. Criticality identification using CIPDFs and mean impact curves/surfaces is presented for both pre-event and post-event situations. The methodology is applied for electric power network criticality identification for household dislocation impact in Galveston, TX, USA in a hurricane event.
Read full abstract