Assessment of Energy flexibility based on different control methods on thermally activated system: A case study on zero-energy office building

Thermally Activated Building Systems (TABS) integrated with high-performance building envelopes offer significant potential for energy flexibility through demand response. However, such thermal inertia poses control challenges for cooling control. Model Predictive Control (MPC) has emerged as a promising solution, but the impact of different modeling mechanisms on control performance in the context of flexibility needs to be investigated. This study compares two adaptive MPC strategies based on grey-box and black-box models against conventional rule-based control (RBC) for a thermally activated ceiling cooling system in a zero-energy office building. The control framework are developed to minimize energy cost within the constraint of thermal comfort through setpoint regulation. A comprehensive assessment examines model accuracy, cost savings, and energy flexibility metrics to determine the performance of each control approach. Results show that grey-box MPC achieves superior performance with 38.0% cost reduction and 20.0% energy savings compared to RBC, while significantly improving energy flexibility with 33.3% higher flexibility factor and 141.7% increase in self-sufficiency. The black-box MPC demonstrates moderate improvements with 21.6% cost reduction and 14.1% energy savings.