Date of Award
12-1-2025
Degree Name
Master of Science
Department
Civil Engineering
First Advisor
Shin, Sangmin
Abstract
Existing centralized water supply systems are critically threatened by the joint effects of climate and socioeconomic changes, extreme weather events, and physical degradation due to aging, and their changeability. The uncertain and changing drivers pose challenges for water supply systems in providing sustainable water services during disruptions. The energy field dealing related issues has endorsed the microgrid approach with a decentralized energy supply with dispersed local energy sources. The application of energy microgrids has demonstrated their resiliency for sustainable energy supply. For improved resiliency of a water distribution system, this study investigated the application of a microgrid approach and the detection of subtle anomalies in its operation. This thesis explores the setup of a lab scale water distribution model addressing the following queries: 1) 1. Can the microgrid approach enhance water distribution system resilience more effectively than the traditional centralized and decentralized water distribution systems? 2) Can the subtle anomalies in microgrid operation be detected to enhance resilience? To acknowledge the questions and test the corresponding hypotheses, this study built a lab-scale water distribution model, which can demonstrate centralized, decentralized and water microgrid systems under various disruption scenarios – e.g., power off due to extreme weather conditions, pipe leaks/bursts due to aging or freezing. Here, water microgrid systems consider the interactive operations between central and local water systems, while decentralized systems lack operative interactions. The performance of the system was evaluated by functionality-based resilience for each disruption scenario. The quantitative resilience analysis reveals that the microgrid configuration showed reduced degradation of water supply performance resulting on improved resilience during disruptions. Further, a widely used machine learning model for anomaly detection, Autoencoder was applied to detect the minor changes in water use in the lab scale model. Results highlight the ability of autoencoder on detection of subtle anomalies. However, the noise on the datasets plays a vital role in its performance. The findings suggest engineering insights into the upgrade of current centralized water systems with a microgrid approach to maximize sustainable water supply service under uncertain and extreme disruptions.
Access
This thesis is only available for download to the SIUC community. Current SIUC affiliates may also access this paper off campus by searching Dissertations & Theses @ Southern Illinois University Carbondale from ProQuest. Others should contact the interlibrary loan department of your local library or contact ProQuest's Dissertation Express service.