PhD Final Defense – Roberto Guidotti
Regional Risk and Resilience Analysis of Interdependent Critical Infrastructure
Advisor: Professor Paolo Gardoni
Date: Friday May 18th, 9:00 am
Location: Newmark – Quade Conference Room
Critical infrastructure (such as water and wastewater, electric power, transportation, and telecommunication systems) constitute the backbone of modern society. They provide essential goods and services to communities, supporting the population’s well-being. Hazard events of past years revealed that the vulnerability of communities is related to the vulnerability of their critical infrastructure. Critical infrastructure are exposed to low-probability high-consequence hazard events. Reductions and interruptions in their functionality may result in considerable impacts on society. A prompt recovery of the critical infrastructure leads to a prompt recovery of the economic vitality and the general well-being of the impacted communities. Interdependencies among infrastructure and between infrastructure and social systems may increase the vulnerability of communities and exacerbate the impact of hazard events, often resulting in widespread disruption and slower recovery.
This dissertation proposes a novel probabilistic methodology to quantify the reliability and resilience of interdependent critical infrastructure. Infrastructures are modeled mathematically as networks, extending to civil engineering applications well-established tools of graph theory. Topology and flow-based approaches are used to translate the physical damage of the single components into loss or reduction of network functionality and to develop network capacity and demand models. Network capacity and demand models are integrated in a time-varying network reliability and resilience analysis to assess the network response in the immediate aftermath of a hazard event and at different times during recovery process, as the network components are repaired. To capture the role of interdependencies and propagate the loss or reduction of functionality across all dependent networks, this dissertation presents a novel multi-layer heterogeneous network model. In the proposed model, the heterogeneity comes from having different components in each infrastructure (i.e., generation, distribution and transmission components). The different classes of interdependency (e.g., physical, geographical, social, etc.) are modeled as different layers.
This dissertation applies the proposed methodology to a series of example testbeds, including isolated and interdependent, virtual and real critical infrastructure. Results shed light into the role of interdependencies among infrastructure and between infrastructure and social systems in the recovery of communities. Results of this dissertation aim to benefit civil engineers to develop costeffective mitigations measures and practices in infrastructure design, construction and planning, as well as emergency managers, planners and the community to be better prepared for future hazard events.