Use of model predictive control to enhance the flexibility of thermal energy storage cooling systems

This paper investigates the application of a model predictive controller (MPC) to both a traditional and a novel chilled water thermal energy storage system over for an Austin, Texas, climate. In the novel system, the thermal storage discharges during peak electricity times to meet building cooling load and to supply reduced temperature water for heat rejection in the chiller's condenser. Chiller efficiency improves as the condenser water temperature decreases, shifting more electrical usage to off-peak hours, but may increase overall electrical usage. The MPC is designed to optimize the discharge and recharge of the thermal storage in order to minimize operation costs or energy consumption over a 24-hour prediction horizon. The ability of MPC to level the electrical load profile is also considered. The way in which demand charges are considered in the objective function can greatly influence the system's electrical load profile.