Graduation Semester and Year
Spring 2025
Document Type
Dissertation
Degree Name
Doctor of Philosophy in Civil Engineering
Department
Civil Engineering
First Advisor
Mohsen Shahandashti
Second Advisor
Kyeong Rok Ryu
Third Advisor
Nilo Tsung
Fourth Advisor
Evan Matthew Mistur
Abstract
Water distribution networks (WDNs) are critical infrastructure systems that are highly vulnerable to seismic hazards. Utility managers are often required to make proactive rehabilitation decisions under uncertainty, constrained by limited resources. However, existing seismic rehabilitation decision-making models typically ignore the influence of risk preferences, assume cooperation among conflicting objectives, and lack flexibility in addressing the diverse needs of utilities. This research aims to develop comprehensive risk-averse and multi-objective decision-making models to identify critical pipelines for proactive seismic rehabilitation in WDNs. Three distinct models are proposed to support utility decision-making across different planning contexts. First, a single-objective risk-averse simulated annealing (RSA) algorithm is formulated to identify critical pipes by maximizing the expected post-earthquake serviceability index (ESI) while controlling risk metric through a value at risk (VaR) constraint. This model enables decision-makers to select a rehabilitation strategy that aligns with their specific risk tolerance level. Second, a multi-objective nondominated sorting genetic algorithm (NSGA) is developed to generate a set of Pareto-optimal rehabilitation strategies that simultaneously maximize the ESI and minimize risk metric, quantified using both VaR and conditional value at risk (CVaR). This approach offers greater flexibility by presenting utility managers with a spectrum of trade-offs between post-earthquake serviceability and risk, rather than a single optimal rehabilitation strategy. Third, a multi-objective game theory model (MOGM) is introduced to explicitly account for the inherent conflict between minimizing rehabilitation cost (RC) and maximizing ESI. By modeling these objectives as strategic players in a game, the MOGM identifies Nash equilibrium strategies that provide balanced and simplified rehabilitation strategies that address the strategic conflict between cost-efficiency and post-earthquake serviceability enhancement. Each model integrated several methodological components, including seismic repair rate estimation using empirical fragility curves, multi-physics hydraulic and seismic damage modeling, and Monte Carlo simulations (MCS) to capture the probabilistic nature of pipeline damage. The models are applied to benchmark Modena WDN to identify critical pipes for proactive seismic rehabilitation. Each of the proposed frameworks offers a distinct advantage in addressing the complexities of seismic rehabilitation planning for WDNs. The outcomes of this research offer novel contributions to the field of decision-making by advancing optimization-based approaches for the seismic rehabilitation of WDNs. Specifically, the RSA framework equips decision-makers with a structured approach to select rehabilitation strategies under a controllable level of risk aversion. Building on this, the NSGA framework provides a set of Pareto-optimal solutions, enabling utility managers to explore a range of trade-offs between post-earthquake serviceability and risk. Furthermore, the MOGM framework introduces a strategic layer to decision-making by treating conflicting objectives, such as cost and post-earthquake serviceability, as independent players, resulting in more balanced and actionable rehabilitation strategies. Together, these models form a comprehensive, flexible, and practical decision-making toolkit for enhancing the post-earthquake serviceability of WDNs.
License
This work is licensed under a Creative Commons Attribution-NonCommercial-No Derivative Works 4.0 International License.
Recommended Citation
Sharveen, Sumaya, "Risk-averse and Multi-objective Decision-making Models for Rehabilitating Water Pipeline Networks Subjected to Earthquakes" (2025). Civil Engineering Dissertations. 511.
https://mavmatrix.uta.edu/civilengineering_dissertations/511