Probabilistic Fragility of Interdependent Urban Systems Subjected to Seismic Hazards
Doctor of Philosophy
Urban service networks have come under increased pressure due to expansion of urban population, decrease of capital investment, growing interdependence, and man-made and natural hazards. This thesis introduces a simulation-based methodology for the estimation of the fragility of urban networks subjected to earthquake perturbation. The proposed Interdependent Fragility Assessment (IFA) algorithm abstracts the steps required for perturbation-induced damage propagation within and between networks through internal and interdependent links, respectively. Damage propagation uncertainty is accounted by considering conditional probabilities of failure for components and interdependent strengths measuring the likelihood of intersystemic failure propagation. The IFA algorithm is used in four applications. The first application subjected two simplified models of real interdependent urban power and water networks to selected seismic scenarios. Test results showed that interdependence presence worsens systemic fragility, but that the features of interdependence effects were jointly influenced by local fragility properties and interdependence strengths. A second application examined the role of cascading failures caused by component overloading in systemic fragility. The results showed that cascading failures worsen interdependence fragility, and that mitigation actions improving local component capacity have limited effect on controlling interdependent-induced fragility. Two additional conceptual mitigation measures, component fragility reduction ( CFR ) and interdependence redundancy enhancement ( IRE ), were explored. CFR , decreases component seismic fragilities while IRE adds interdependence links to dependent nodes. Test results showed that CFR outperforms IRE ; however, their combination achieved comparable fragility reductions. This outcome highlights the potential of synergistic mitigation policies in controlling interdependent systemic fragility. Finally, the IFA methodology was adapted to use a probabilistic seismic description for the estimation of unconditional systemic fragilities. The hazard description was obtained following an existing approach that uses importance sampling for the generation of intensity maps. The value of the hybrid methodology rests on its capacity to generate unconditional fragility estimates for direct use in risk assessment. Topics for future work include the development of more sophisticated models of cascading failure, the analysis of optimal mitigation actions using mitigation cost-structures and life-cycle costs, the extension of the IFA methodology for perturbation such as hurricanes and flooding, and interdependent fragility studies of theoretical network models.