A model predictive control-based framework for evaluating building thermal mass energy flexibility

Buildings’ thermal mass provides energy flexibility to support power grids with increasing shares of renewable generation. Since the amount of shiftable energy varies with boundary conditions, a reliable time-dependent quantification is required. This study proposes a framework to assess the time-varying energy flexibility of building thermal mass using the flexibility envelope metric. Unlike existing works, which either rely on simplified control-oriented R-C models or on detailed models combined with rule-based control, the proposed framework integrates high-fidelity building and HVAC models with a Model Predictive Control algorithm through co-simulation. This approach enables flexibility envelope calculations that jointly capture detailed thermal dynamics and predictive control capabilities. The framework is applied to two residential buildings with floor heating. The results show that simplified R-C based approaches underestimate flexibility in 71% of the evaluated conditions, with an average deviation of −20%. Finally, the storage efficiency ranges between 40%–52% and 46%–55% for the two case studies. Highlights Building’s thermal mass energy flexibility quantification. Co-simulation of the physics-based building model with Model Predictive Control (MPC). Model Predictive Control to predict and optimize the building’s thermal response. The use of co-simulation to quantify the flexibility envelope method boundaries. Heat storage process evaluation through the calculation of storage efficiency.