MPC, GEOTABS, real implementation
Hybrid GEOTABS buildings are buildings equipped with a ground source heat pump (GSHP), thermally activated building systems (TABS), a conditioned ventilation and, optionally, additional heating and cooling systems such as variable-air-volume boxes or radiators. GEOTABS can be very energy efficient but their controllability is limited and thermal comfort is therefore not always guaranteed. Hybrid GEOTABS systems have the potential to eliminate these problems, provided the different system components interact properly. In this paper, we investigate the performance of hybrid GEOTABS systems for an office building, a retirement home, a school and a block of flats when controlled by a current practice rule-based-controller (RBC). The study is based on detailed simulation models inspired by four existing buildings. The RBCs performance is then compared with the performance achieved by Model Predictive Controllers (MPC) which optimize both the heat flow rates to the TABS and to the supplementary systems, and the supply ventilation temperature. The study shows that while thermal comfort cannot be guaranteed by RBC for all buildings, a very high thermal comfort is achieved when controlled by MPC. This means that hybrid GEOTABS systems can technically be successfully used for a wide variety of building (and occupancy) types when appropriate control is implemented. Moreover, the investigated MPCs can save between 30% and 50% of the energy cost of which 6% to 11% is obtained by optimizing both the TABS thermal powers and the ventilation supply temperature simultaneously instead of only the TABS powers. Finally, the optimal control behaviors obtained by MPCs showed clear time patterns which can be used to improve current practice RBCs. However, it is shown that simple heating/cooling curves, as used today, cannot reach the MPC performance.