Multi-dimensional resilience of water distribution system for water quality sensor placement

Author

Albira Acharya, Southern Illinois University CarbondaleFollow

Date of Award

12-1-2022

Degree Name

Master of Science

Department

Civil Engineering

First Advisor

Shin, Sangmin

Abstract

Water distribution system (WDS) is very critical to human health and societal welfare. Maintaining the quality of the water so that potable water gets distributed to consumers has always been a challenge in the water industry. Deterioration of water quality can happen either accidentally or deliberately and the widespread geography of the water system makes it even more vulnerable to contamination. In this respect, researchers and utilities have some response action to flush out the contaminants when they are detected. But not all networks have reliable sensors to detect the contamination and lack of guidelines for sensor deployment has made the situation even more serious. Given this context, framework for decision-making in the case of WDN against contamination is a much-needed approach. Understanding the capability of the water system to handle the contamination event could provide ample insight on how to better protect the system and how to handle if the contamination does enter the system. In this regard, this study explores the concept of resilience to define the system performance when a disruption occurs, which in this case is the intrusion of contaminants. Resilience of a system can be viewed from different perspectives, each highlighting different aspect of the system. With this insight, the objective of this research is to characterize the resilience of the water system against contamination for multiple aspects of performance or functionalities and use that concept to further elucidate the decision-making process. Hydraulic and quality simulation to emulate the contamination intrusion in WDN is performed by using EPANET-MATLAB Toolkit which has the needed package for both EPANET and EPANET-MSX. EPANET-MSX is widely used for simulating multiple intrusions in the system. The result from the MATLAB simulation gives the quality at each node which is then used to draw the performance time-series curve. Resilience is then computed for each of the performance metrics using the area under the curve method. This study makes a comparison study for multi-dimensional resilience and describes in detail the need of considering the attributes of resilience which are resistance, loss rate, recovery rate, failure duration, and recovery ability. To perceive the concept of resilience with respect to the failure scenarios, a sensitivity analysis was performed for four failure contexts namely, intrusion time, intrusion duration, intruded contaminated mass, and the number of intrusion nodes. Furthermore, a system measure is defined to aggregate different individual resilience to overcome the challenge of multi-objective decision-making. Application of both integrated and multi-dimensional resilience was conducted for optimal sensor placement in the network to maximize the resilience of the whole system. The goal of this thesis is to introduce the multi-dimensional resilience concept as a tool for decision-making based on multiple aspects of system performance by characterizing the WDS resilience and water quality sensor optimization based on different aspects of system functionality under contaminant intrusion events.