Agent‐Based Modeling to Simulate Contamination Events and Evaluate Threat Management Strategies in Water Distribution Systems

Abstract

In the event of contamination of a water distribution system, decisions must be made to mitigate the impact of the contamination and to protect public health. Making threat management decisions while a contaminant spreads through the network is a dynamic and interactive process. Response actions taken by the utility managers and water consumption choices made by the consumers will affect the hydraulics, and thus the spread of the contaminant plume, in the network. A modeling framework that allows the simulation of a contamination event under the effects of actions taken by utility managers and consumers will be a useful tool for the analysis of alternative threat mitigation and management strategies. This article presents a multiagent modeling framework that combines agent-based, mechanistic, and dynamic methods. Agents select actions based on a set of rules that represent an individual's autonomy, goal-based desires, and reaction to the environment and the actions of other agents. Consumer behaviors including ingestion, mobility, reduction of water demands, and word-of-mouth communication are simulated. Management strategies are evaluated, including opening hydrants to flush the contaminant and broadcasts. As actions taken by consumer agents and utility operators affect demands and flows in the system, the mechanistic model is updated. Management strategies are evaluated based on the exposure of the population to the contaminant. The framework is designed to consider the typical issues involved in water distribution threat management and provides valuable analysis of threat containment strategies for water distribution system contamination events.

Assessing the value of cooperation and information exchange in large water resources systems by agent‐based optimization

“…For example, the four agents in Figure 1a are modeled as organized in two coalitions and thus solve two optimization problems with two objectives each. The explicit formulation and solution of an optimization problem by each agent further develop the agent-based optimization approach introduced by Yang et al [2009] and replace the a priori defined rules adopted in most of agentbased water resources applications [e.g., van Oel et al, 2010;Zechman, 2011;Le et al, 2012, and references therein].…”

Using Niched Co-Evolution Strategies to Address Non-Uniqueness in Characterizing Sources of Contamination in a Water Distribution System

“…Another area of future work for NCES is incorporating the social aspects of a contamination event with the technical system through a socio-technical system analysis approach. Research is already underway in this field (Zechman 2011). Using NCES to generate alternatives instead of finding one outcome can provide an immense amount of insight for socio-technical problems.…”

On Event Detection and Localization in Acyclic Flow Networks

“…To protect consumers in a network, water security measures should include an on-line, real-time contaminant warning system of sensors to quickly identify any degradation in water quality. Efficient placement of sensors is needed to collect information for responding to a threat by identifying the location and timing of the contaminant intrusion [6] and developing strategies for opening hydrants to flush a contaminant [7]. An extensive set of studies have been conducted to develop and apply optimization-based methodologies for placing sensors in a water distribution network [8] [9].…”