MPC based Community Battery System to minimize the Energy Cost of a Residential Community

Community Battery Energy Storage is highly beneficial for residential microgrids and mitigates significant household investments. This paper considers two townhouses with grid-connected photovoltaic (PV) systems and a community battery owned by them. This is in furtherance to a previous work that employed Mixed Integer Linear Programming (MILP). A Model Predictive Control (MPC) methodology is used here, which optimizes by looking ahead, considering the uncertainty in PV and demand. This study uses the synchronous averages of PV and demand for 2020 and 2021 as predictions. The study explores the utilization of MPC in a community battery scenario to maximize the energy cost outcome to motivate investments. Two prediction scenarios are simulated with varying control horizons to characterize the range of likely monthly bills. The electricity bill without a community battery is ${\$}$227.35. The savings obtained using synchronous averaged and perfect predictions are ${\$}$11.47 and ${\$}$67.91 respectively.