Real-Time Optimal Valve Operation and Booster Disinfection for Water Quality in Water Distribution Systems

Abstract

Historically, a water distribution system’s (WDS) hydraulic performance has been the primary operational concern. Over the past two decades, however, more attention has been paid to water quality behavior in WDS and today, water quality level is an equally important issue for many water utilities. In most cases, maintaining disinfectant levels is usually of interest to avoid the bacteria regrowth and to protect against the potential cross-contamination events. However, disinfectants, such as chlorine, decay over time and produce potentially harmful disinfectant by-products when they react with organic material in the water. Therefore, maintaining a minimum chlorine residual requirement throughout the WDS is a complex but important task. When online booster disinfection is combined with source disinfection, it has been shown that the total chlorine dosage can be reduced while maintaining minimum chlorine residuals across the system. Here, optimal valve operation has been combined with booster disinfection to improve the system water quality. Valves can be operated to alter the flow distribution in the network; prevent the isolation of water; and direct disinfectant laden water to locations where it is needed. A real-time optimal valve operation and booster disinfection problem is formulated as a single objective optimization model. The objective is to minimize chlorine injection mass at sources or to minimize excessive chlorine concentrations at withdrawal points while maintaining minimum chlorine concentrations and pressures throughout the system. The problem is solved using a genetic algorithm (GA). The application to a medium-sized WDS shows that optimal operation of existing valves combined with booster disinfection can improve water quality while requiring lower chlorine doses and resulting in little significant pressure reduction. Also, real-time operations can adapt to the temporal and spatial variations of system demands.