Model predictive control of District Heating substations for flexible heating of buildings

Abstract

The potential in cost and energy savings by replacing a feed forward weather compensated control (WCC) controlled radiator system with a linear MPC controller is investigated in a Modelica-Python setup. It is shown that if the MPC is optimized for minimum energy consumption it can reduce the energy consumption by up to 12 %. It is also demonstrated how variable price signal can influence the heat demand profile, and thereby shift energy consumption away from peak hours. By introducing a peak load tariff, it is also possible to reduce the rapid changes and large peaks often caused by optimization-based controllers