Loading...

Probabilistic Modeling of Networked Infrastructure in Community Resilience Analysis

Biazar, Sina | 2019

666 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 52111 (09)
  4. University: Sharif University of Technology
  5. Department: Civil Engineering
  6. Advisor(s): Mahsuli, Mojtaba; Safdarian, Amir
  7. Abstract:
  8. This thesis proposes a probabilistic modeling framework for networked infrastructure systems in community resilience analysis. In particular, the framework is presented for the electric power infrastructure. Models are proposed at two levels of refinement. The basic model only determines the connectivity of different elements in the network, whereas the refined model also quantifies the amount of the power transmitted through the network elements. In a community resilience analysis, the proposed basic model provides a rather simple tool to determine those buildings and infrastructure systems that suffer from power outage when the supplying power station loses connectivity with the network. In contrast, the proposed refined model is capable of quantifying the voltage and the current at any point within the network. Hence, in addition to the loss of connectivity, the refined model controls if the delivered power satisfies the acceptable bounds. The proposed models account for the uncertainty in the failure of the power network elements subject to an earthquake, the uncertainty in the failure of the network elements due to a high current, and the uncertainty in the repair time and repair cost of the failed elements. As a novel development, the restoration process of the electric power network, i.e., the so-called “maneuver operation” is modeled in this study using agent-based simulation. The proposed models are implemented in Rtx, a computer program for risk and resilience analysis. Rtx is based on a framework that employs a cost-based measure to quantify community resilience. The measure includes the costs incurred by the community immediately after the occurrence of the hazardous event and the costs of recovery over time, which are simulated using agent-based modeling. The Rtx framework is capable of accounting for the intra-dependencies between the elements of the power system and interdependencies between the elements of the power systems elements and those of other infrastructure systems. Hence, Rtx is capable of modeling the cascading failures triggered by an interruption in the network. For instance, the proposed framework quantifies the indirect social and economic impacts of power outage, and their effect on the community resilience measure.The proposed method is showcased by an application to a hypothetical part of Tehran, Iran. This community is composed of residential and commercial buildings, healthcare facilities, an electric power infrastructure, and a water system. The results show that the total cost incurred by the community under a seismic event of magnitude 7 would be underestimated by 20% if the power system was not treated as a networked infrastructure. This amounts to $193M in costs for a region whose size of economy is $307M, measured in terms of annual gross regional product. This emphasizes the need to properly model the dependencies in resilience analysis
  9. Keywords:
  10. Resilience Index ; Infrastructure ; Interdependency ; Probabilistic Modeling ; Interdependent Infrastructure Systems ; Network Resiliance

 Digital Object List

 Bookmark

No TOC