Post-event restoration simulation of water distribution systems: a generally applicable approach
Date
2023
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
University of Delaware
Abstract
This dissertation describes a new tool for modeling the post-event restoration of a damaged water supply system. Called the REstoration of Water after Event Tool (REWET), it can be used to estimate the impact of a disruptive event (e.g., earthquake, landslide, wildfire) on a water supply system’s functioning by simulating how the restoration process unfolds. Simulation results provide the duration and spatial distribution of water service disruption, which allows for assessing impacts that households, businesses, and others ultimately experience. REWET’s strengths are: (1) both the hydraulic operations of the system and the restoration process are represented in detail, (2) any system or disruptive event can be modeled supporting varying levels of complexity, and deterministic or probabilistic analysis, and (3) its code is open-source and free to use. The tool uses Pressure-Demand Analysis for its hydraulic simulations and allows detailed discrete event simulation representation of the restoration process. The dissertation proposes a general restoration plan framework developed by scrutinizing restoration models and their rationales to encompass a range of restoration models and anything in between or similar to the reviewed restoration models. REWET’s restoration model is created using the general restoration plan framework; thus, it applies to a wide variety of restoration strategies. Case study applications of REWET for a complex, real-life system—Los Angeles—and two small, simple networks—Net3 and Anytown—illustrate the tool’s functionality, flexibility, and key features. Furthermore, a probabilistic case study application for Los Angeles demonstrates REWET’s scalability and tractability. ☐ Additionally, this dissertation enhances the understanding of the effect of restoration and restoration planning. Utilizing REWET, a parametric study was conducted to investigate: (1) the effect of restoration in reducing the impacts of a water supply interruption, (2) how to design the restoration process to minimize the impacts of a water supply interruption, and (3) the performance of different restoration strategies under various circumstances (type of system, type of event). The Los Angeles and Anytown networks were subjected to damage scenarios reflecting various damage conditions (intensity and locations). The study provides insights such as the necessity of designing restoration strategies based on the network structure, restoration possibilities, and damage conditions. Additionally, the effect of restoration components, such as the discovery of damages and rerouting, has been studied under various circumstances.
Description
Keywords
Discrete event simulation, Post-event restoration, REWET, Water networks, Restoration models