Loading...

Probabilistic Framework for Integrated Resilience Analysis of Urban Infrastructure Systems And Application to a Virtual City

Sepehri, Sina | 2024

0 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 57057 (09)
  4. University: Sharif University of Technology
  5. Department: Civil Engineering
  6. Advisor(s): Mahsouli, Mojtaba
  7. Abstract:
  8. This thesis presents a comprehensive probabilistic framework for community seismic resilience assessment considering all major infrastructures and their dependencies in an integrated analysis. The infrastructures that are modeled in this research include the buildings portfolio, electric power system, potable water system, natural gas system, telecommunication system, transportation system, healthcare system, and businesses. In fact, this research builds upon past studies, each of which focused on assessing the resilience of individual infrastructures. Integrating resilience analysis of all infrastructures allows for modeling the effects of their dependencies on robustness, rapidity, and resilience measures, comparing these indicators amongst various infrastructure systems, and revealing the share of each infrastructure in the overall community Loss. This framework is showcased through a comprehensive case study that features a virtual city. This framework is based on integrating hazard, risk, network, recovery, and agent-based models. Hazard models include earthquake occurrence, magnitude, area and location of fault rupture, ground shaking intensity, and ground failure models. Subsequently, risk models, receive the intensity measure from hazard models and predict the status of the community including the number of injuries and human fatalities, and the ensuing social impacts, as well as the status of each infrastructure component immediately after the occurrence of the earthquake. For example, risk models for the water infrastructure include damage models of stations and pipelines, and the water station serviceability model. Thereafter, recovery models, receive the status of the community and infrastructure components from risk models and predict the time and cost required for the recovery of each infrastructure to normal conditions. For networked infrastructures, network models determine the service status of the system by receiving the status of individual components. For example, the water network model, receives the status of water stations and pipelines and determines the network serviceability at consumption nodes. Finally, agent-based models simulate the required operations for the recovery of the community to its normal state by defining organizations that are responsible for the recovery of each infrastructure as agents and modeling their resources, characteristics, behaviors, and interactions with each other and the environment. The agent-based model is executed within a discrete event simulation, where each recovery operation is subject to resource constraints of the agents through queueing, and the cost and time of operations are estimated using the aforementioned recovery models. The interdependency of infrastructures is modeled both at the level of physical dependence of components of an infrastructure to other infrastructures, and at the level of dependence of recovery operations of infrastructures with each other. For example, the performance of each water station is dependent on both the level of damage to that station and the serviceability of the power system at its location. As another example, repairing a power station that serves a hospital leads to an increase in the capacity of the hospital to treat the injured. The healthcare agent, informed of this by the power system recovery agent, schedules discrete events to accommodate and treat more injured individuals. At the end of this simulation, the cash flow of community costs and the duration of community recovery are calculated. To propagate the uncertainty, Monte Carlo sampling is performed to generate enough random samples to quantify the probability distribution of the total community cost and its recovery time. A cost-based resilience measure for the community is then computed as a function of the total community cost and its gross regional product. Additionally, resilience indices specific to each infrastructure are calculated by the area under the normalized performance curve of that infrastructure. The current framework has been implemented in the Rtx, which is software for reliability, risk, and resilience analysis equipped with a comprehensive library of probabilistic models for various hazards, infrastructures, and consequences. The capabilities of this framework are demonstrated through its application on the virtual city. Primary results are the overall resilience index of the entire community and the comparison of robustness, risk, and resilience measures for various infrastructures. Insights obtained from the proposed framework include determining the share of each infrastructure from community costs and the share of various types of costs, including direct and indirect economic, direct and indirect social, socioeconomic, and environmental costs. This framework acts as a decision-support tool for the optimal allocation of limited resources to community resilience enhancement actions
  9. Keywords:
  10. Earthquake ; Urban Infrastructure ; Probability Model ; Agent Based Simulation ; Failure Dependency ; Discrete Event Simulation ; Seismic Resilience ; Infrastructure Resilience ; Probabilistic Seismic Structures Vulnerability

 Digital Object List

 Bookmark

No TOC