24 June 2014
The study is concerned with a 14-storey office building, which requires comfort in
temperature and in humidity during the working hours only. The Heating, Ventilation and Air-
Conditioning (HVAC) system is of the forced-air type with energy recovery capabilities from
the exhaust air. The heating and refrigeration systems comprise not only standard gas
boilers and cooling machines, respectively, but also geothermal heat pumps running in both
heating and cooling modes.
From earlier evaluations of the total system over one year, the authors have observed that
the thermal inertia of the building (due to the concrete structure) plays an important role as
for the daily profile of the energy required.
In this study the benefit from shifting the energy load is evaluated, i.e., from storing into the
building some energy produced by the heat pump only, preferably at higher Coefficient Of
Performance (COP) values, and leaving the building deliver the energy when needed. Due to
the slow dynamics of the system, one must account for the future evolution of the weather
conditions and of the building occupancy (which yields important heat gains); for this reason
a Model-based Predictive Control (MPC) scheme is developed in Matlab®. The algorithm
calculates every evening with the rest of the week as a horizon the control variable values
allowing for the minimal energy cost while satisfying the comfort constraints. The strategy
applies with good results to both heating and cooling periods. In the future the authors will
first evaluate the robustness to inaccuracies in the model structure/parameters and in the
inputs, then implement the control scheme on site.