Analysis on the effect of V2G aggregation on distribution network based on traffic simulator

Abstract

Transition to electrical vehicles (EVs) is promoted in many countries to reduce CO2 emissions. EVs are also expected to work as one of the distributed energy resources (DERs), which can contribute to grid balancing. Businesses on EV aggregation are appearing, and some demonstration projects have proceeded. In such projects, the potential for grid balancing of EV aggregation has been evaluated, however, the effect of the EV aggregation on distribution networks has not been sufficiently discussed. The bundled EVs are expected to be charged or discharged at the same time, causing large power flow and an excessive voltage variation in distribution networks. Power flow analysis is required for discussing the voltage variation. However, the amount of data on charging and discharging of EVs is insufficient. The purpose of this research is to develop a method of simulating charging and discharging patterns of EVs controlled by aggregators. The difficulty of this research is in how to consider the connection of each EV to chargers, which is the critical restriction in scheduling the charging and discharging of EVs. This problem was solved by generating time-series data on connections of each EV to chargers with a traffic simulator, which had already been developed in our institution. This report proposed to schedule charging and discharging of EVs based on the data given by the traffic simulator for simulating charging and discharging patterns. This method was applied to simulated load curves made by the aggregator that operates a distribution network similarly to a virtual power plant (VPP). The aggregator scheduled charging and discharging of 256 EVs such that the cost of purchasing electrical power and the wheeling charges are minimized. The voltage variation was evaluated based on this load curve in the VPP.