Optimizing operational control of a compression chiller system with photovoltaics and an electrical battery
Cooling currently accounts for ~20% of the world’s total energy consumption in buildings. This is projected to increase by 2050, when two-thirds of the global households will likely have air conditioners. As a result, the pressure on the electricity grids will also increase (especially during peak load hours). Reducing electricity peaks in cooling systems is, therefore, crucial and could be achieved using storage technologies to shift the demand. To this aim, we are investigating a compression-chiller-based cooling systems with photovoltaics (PV) coupled with an electrical battery. In addition to reducing the peak load, such a system would also benefit from storing surplus PV or grid electricity at lower tariffs. An electrical battery could also shift the non-cooling electricity load of the building. As the economics and energy efficiency of this system is tied to the combination of electricity prices, and weather and cooling demands, it has many opportunities for control optimization. To show the impact of this optimization, this paper aims to compare results using a rule-based control to results using model-predictive control. The system will be modelled within a detailed dynamic simulation environment for an office building in Chur, Switzerland. The model-predictive controller will interact with the dynamic simulations using control signals from a MILP-based mathematical optimizer. The optimized control aims to minimize the total operation cost, i.e. the cost of electricity consumption, assuming dynamic electricity prices. In the initial results, the application of model-predictive control demonstrated a significant reduction in operation costs compared to the rule-based control, highlighting its potential to enhance economic performance in PV-battery cooling systems.