Reducing the carbon footprint of house heating through model predictive control – A simulation study in Danish conditions

Abstract Around the world, electricity systems are transitioning towards renewable energy to meet humanity's climate change mitigation targets. However, in a pre-transition system, the carbon intensity of power exhibits strong variations over time, which calls for load shifting to times when its impact is lower. In this work, the case of heating in single-family houses is studied, using Model Predictive Control (MPC) to optimise multi-zone operation. Low inertia heating is used, and simulations are made upon three different insulation level using historical grid and climate data from Denmark. The results show that energy and CO 2 optimisation are relevant objectives for predictive control for lowering the carbon footprint of heating, while SPOT price optimisation is comparatively undesirable. However, benefits of energy optimisation were questioned, as a well-tuned PID control might have had similar performance. Nevertheless, gains from CO 2 optimisation in recent houses highlight the importance of considering the average carbon intensity of energy used, in addition to the amount of energy itself, when aiming to reduce the carbon footprint.