Minimizing carbon footprint from power generation for electric vehicle charging considering demand uncertainty and grid operation constraints

Matching electric vehicle (EV) charging with low-carbon electricity generation through charging power optimization is critical to enhancing the environmental benefits of vehicle electrification. However, existing strategies often overlook the balance with distribution grid constraints and rely on offline optimization, which requires complete daily EV charging demand data in advance. This study develops an online optimization framework integrating load flow analysis and the rolling horizon approach. Load flow analysis first determines an electrical load limit based on distribution grid constraints, including bus voltage deviation, transformer loading, and power flow transmission limits. Subsequently, the rolling horizon approach dynamically optimizes EV charging powers to align with periods of low-carbon electricity generation, accommodating future uncertain charging demand. Simulations based on real-world data from a workplace parking lot in California, USA, demonstrated that the framework could reduce daily CO2 emissions by up to 20 % compared to uncontrolled charging, showing a non-linear relationship between charging demand and emission mitigation. Statistical analysis revealed that the online optimization approach closely matched the performance of offline optimization. The emission reduction gap was less than 3.3 %, with no statistically significant differences observed in most scenarios. These results verified the efficacy of the framework under diverse operational conditions.