Energy Demand Optimisation in PV- and EV-Integrated Buildings: A Supervisory Controller Based on MPC

This paper presents the development, implementation, and validation of a Model Predictive Control (MPC) system for managing peak demand and maximising the use of photovoltaic (PV) power when charging electric vehicles (EV) in a building microgrid environment. The proposed MPC-based smart charging controller optimises EV charging by adjusting charging rates based on user's battery energy level requirements, PV availability, building electricity demand, and electricity pricing. The controller seeks to balance the reduction of peak demand while meeting the needs of the EV user. The system is first tested in a simulation environment and then integrated into an office building in the Netherlands, and is currently operational. During the initial phase of testing, a total of 89 charging sessions were recorded during office hours. An average peak power reduction of 8.4 kW, equivalent to 17%, was observed, with the maximum reduction reaching 25.0 kW (42%), and the minimum being 0 kW. Throughout this phase, self-consumption was sustained at 100%, a consequence of the building's limited PV capacity. Self-sufficiency fluctuated between 22.5% and 0.2%, with an average of 6.4\%. In the subsequent testing phase, the contract electricity demand constraints (kW) were intensified. There, 20 charging sessions occurred during office hours in a period of 10 days. The analysis of key performance indicators revealed an average peak power reduction of 7.8 kW (19.5%), with a maximum peak reduction of 18.0 kW (45%) and a minimum reduction of 0.9 kW (2.3%). During the same period, self-consumption remained at 100%. However, the enhancement in PV power availability led to a self-sufficiency range of 45.1% to 17.3%, averaging at 29.5%.