The growing concerns over greenhouse gas (GHG) emissions coupled with the development of electric vehicle technology have motivated the electrification of the transportation sector. Plug-in electric vehicles (PEVs) are becoming an alternative to vehicles powered by internal combustion engine, and their deployment is expected to increase in the future. Although this leads to environmental and economic benefits, power grid must be prepared for new challenges. Therefore, investigating and understanding the impacts of PEVs load demand on the distribution power system are of great importance from the perspective of power utilities. Furthermore, developing smart charge scheduling algorithms is essential to mitigate the impacts associated with PEVs charging load demand.
This thesis presents a comprehensive study of PEVs charging impacts on distribution networks. The analysis was performed on a real low voltage grid in Budapest which is modeled using DIgSILENT Power Factory software. 24-hour time series simulation is executed to investigate the key parameters of the grid such as transformer loading, feeders loading, voltage profile at the feeders’ furthest points and the system daily energy losses for increased penetration of PEVs. Uncoordinated and delayed charging scenarios have been studied and compared. Furthermore, uncoordinated charging scenario with uneven distribution of PEVs is simulated to explore the impacts on voltage unbalance factor and neutral current. The results demonstrate the adverse effects of uncoordinated charging of PEVs at high penetration levels. As an example, the distribution transformer and feeders loading exceeded the limits at 60% penetration level. In contrast, with delayed charging, all the grid components function at their normal rating at all penetration level, and no network upgrade is required.
Despite the effectiveness of delayed charging to alleviate the impacts of PEVs, at higher PEVs penetration level this scheme can result in shifting the peak load to the valley periods of the load profile. Therefore, smart charging technique is necessary to ensure the optimal scheduling of PEVs charging load. For this purpose, a smart charging algorithm was developed to manage the charging/discharging of PEVs with the objective to achieve peak shaving and valley filling of the grid load profile. The PEVs charge scheduling problem was formulated as an optimization problem and solved using particle swarm optimization technique. The algorithm was implemented and tested within MATLAB environment. The results show the effectiveness of the algorithm to achieve peak shaving and valley filling and ensure PEVs and network technical constraints.