Impact of the Weather Forecast Quality on a MPCdriven Heat Pump Heating System

Abstract

Electrically driven heat pumps offer in combination with thermal energy storage systems the potential to response to fluctuating renewable energy sources, e.g. photovoltaics. To fully exploit this flexibility and financial potential, smart predictive control strategies such as Model Predictive Control (MPC) are needed. For such a controller, weather forecast data are mandatory to perform the optimization. Several sources of weather forecast data are available with variable forecasting quality. In this study, the impact of the weather forecast quality on a realistic heat pump heating system is investigated in experiments and simulations. Therefore, the operation of a MPC strategy is carried out for a perfect forecast compared to two imperfect forecast scenarios over a consecutive period of 4 days on a Hardware-in-the-Loop test bench with a geothermal heat pump and a thermal energy storage system. In order to evaluate the benefits in real operation compared to rule-based controllers, a heat-controlled (HC) and a PV self-consumption optimized controller (PVC) are also operated on the test bench. In addition and as a validation process, all scenarios are simulated and compared to the measurement results. Compared to a standard rule-based HC strategy the PV self-consumption can be increased by using a PVC and MPC strategy by 6.2 % and 38.9 %, respectively. The accurately the weather forecasting quality is in general the higher the performance of the HP heating system. Thus, the PV self-consumption is reduced for high-quality and low-quality weather forecasts by 4.6 % and 11.1 %, respectively, compared to a perfect MPC. Even a MPC with low-quality weather forecast data can achieve higher system performance as a simple rule-based HC strategy. For achieving higher system performance by using a MPC instead of a rule-based control strategy like PVC the, forecasting quality has to be as accurate as possible.