Multiobjective Optimization to Explore Tradeoffs in Rainwater Harvesting Strategies for Urban Water Sustainability

Abstract

Urban water systems are designed for centralized management, where water is collected at a central location, treated, and delivered to a population of users through a pipe network. Decentralized systems may generate water and energy savings beyond conventional approaches, as they reduce the demands on the potable drinking water system and the energy required for treatment and conveyance. For example, rainwater harvesting systems that are installed at individual lots can be used to capture and reuse rainwater to irrigate lawns. This research explores the tradeoffs among infrastructure costs, energy savings, and water savings as consumers adopt rainwater harvesting within an existing centralized water supply system. The presence of rainwater harvesting within a community of individual households is a sociotechnical process, as interactions among existing water supply infrastructure, utility managers, and consumers can influence the adoption of decentralized technologies and the performance of centralized infrastructure. The urban water supply system is simulated as a complex adaptive system to analyze the water use behavior of consumers and their influence on system-level sustainability. An agent-based model is constructed to simulate households as water-consumer agents and is coupled with a system dynamics simulation of a water reservoir to capture the feedbacks that drive the household-level adoption of rainwater harvesting. An evolutionary computation approach is coupled with the agent-based modeling framework to optimize multiple objectives and explore tradeoffs among energy requirements, water savings, and the cost of rainwater harvesting systems. The framework is demonstrated for a virtual case study to develop management strategies for sizing rainwater harvesting cisterns and achieving sustainability goals for a sociotechnical water supply system.