Advanced Battery Sizing and Energy Management Approach for Fast EV Charging Station Microgrids
Abstract
This thesis presents a methodology for optimal Battery Energy Storage System (BESS) sizing in DC Fast Charging stations (DCFCs) using a Mixed-Integer Linear Programming (MILP) framework. The approach minimizes electricity costs, reduces grid dependence, and enhances system resilience, while accounting for battery aging and the value of unserved energy. The methodology is evaluated by comparing its performance with a commercial software. A novel energy management approach is developed based on managed charging and load curtailment and Model Predictive Control (MPC). This energy management algorithm utilizes MILP formulation with a rolling horizon to maximize profit and address challenges such as limitation in the power exchanged with the grid, improving system reliability, and maintaining power balance of the fast-charging station. Compared with a rule-based energy management method as a baseline, it demonstrates superior performance in cost reduction and efficiency. A tailored load estimation technique is also introduced to improve load prediction accuracy to use in both battery sizing and energy management approach.