SCHEDULING OF COOL STORAGE USING NONLINEAR-PROGRAMMING TECHNIQUES

Thermal energy storage for space-cooling is a load management technology that achieves load shifting by creating and storing cooling capacity during off-peak hours in anticipation of peak period loading. The charge and discharge process in a cool energy storage system is typically a 24-hour cycle. In this paper an optimal controller for this process is designed that minimizes the operating costs of the storage facility. The cool energy storage system is characterized mathematically as a state-variable system, and management of its operation is developed as a discrete-time optimal control problem. The optimal system operating strategy is defined to be a management policy that minimizes the total operating costs over a 24-hour horizon, while satisfying a set of constraints imposed by the cooling requirements on the system and the capacities of individual components of the system. The control variables are the operating levels of the ice-builder compressors and auxiliary chiller plants in the cooling system. A 24-hour load-forecast model is developed to achieve dynamic system response to the variations in the ambient weather conditions. Hourly building cooling loads are predicted over the 24-hour horizon to schedule adequate charging of storage to meet peak-period loads. The degradation of performance of the optimal control algorithm attributable to the lack of perfect foresight of weather conditions and associated building loads is analyzed via a case study, and a safety margin is calculated and presented to compensate for same.