This paper presents a comprehensive simulation suite for a residential heating and cooling system that combines a vapor-compression system (VCS) with a water-based thermal energy storage (TES) using a three-fluid heat exchanger (TriCoil™). The VCS configuration includes the TriCoil™ as the indoor coil, a fin-and-tube heat exchanger as the outdoor coil, a variable-speed compressor, and an isenthalpic expansion device. To minimize computational cost for annual simulations, an artificial neural network (ANN) was employed to develop a data-driven model for the TriCoil™. The suite integrates models of the VCS, TES, and control systems, factoring in utility rates, building load profiles, and weather data. Validation against experimental data demonstrated accurate predictions, with TES temperature profiles having a root mean square error of 0.3 K and VCS heating capacity predictions showing mean absolute errors of 4.3% for the R410A side, 5.0% for the water side, 4.9% for the air side, and 4.3% for the air-sensible side.Simulations for a cooling season in Stillwater, OK, demonstrated significant electricity cost reductions by shifting cooling loads from peak to off-peak hours. The proposed system minimizes expenses using time-of-use rates and optimized load scheduling, showcasing substantial benefits even with a small TES unit.
Read full abstract