Grid impact study of frequency regulation with EVs

Abstract: The ongoing grid paradigm shift has resulted in synchronous generators currently being phased out. As the generators have been providing ancillary services, such as frequency control regulation, this means that new technologies to perform ancillary services need to be developed and evaluated. Electric vehicles (EVs) is an energy source that today is being underutilized. Therefore, research has been investigating the possibility to provide ancillary services trough vehicle-to-grid technology (V2G) and especially utilizing EVs as frequency containment reserves (FCR). 1–6. However, real world data to evaluate this technology is lacking today. The study in this thesis utilizes real world data from one of the world’s first commercial EV fleets providing FCR-N (Frequency Containment Reserve - Normal Operation) 7, with the aim to evaluate grid impacts from the V2G technology. Initially, the grid impacts in the studied system is analyzed. Thereafter, a simulation model is created in Matlab Simulink. Through varying the cable length (0.02 km to 1.6 km) and the installed power of the EV chargers (100 kW to 800 kW) different scenarios are created and simulated, using the model. Thereby, it is evaluated if and when problems will occur for the grid operation in the studied system. Additionally, it is investigated if grid impacts in terms of voltage can be minimized through reactive power compensation. The main focus of the analysis concerns voltage issues but load profiles are also analyzed. For the studied system no voltage limit violations or thermal limit issues were found. However, the FCR-N provision changes the shape of the load profile for the building, to which the EV fleet belongs. For the reactive power compensation it was found that the suggested minimum power factor (PF) for battery plants suggested by the transmission system operator (TSO) 8 was enough to improve voltage profiles up to 0.2 km, however for longer cable lengths the suggested PF was not enough. Instead a proportional Q(P) controller is suggested which improves voltage profiles up to 0.4 km as well as for lower installed power levels for 0.8 km. For 0.8 km cable length and longer it is not possible to minimize grid impacts through reactive power compensation. Instead grid upgrades would be required if the V2G technology is adapted in systems requiring cable lengths and installed power levels of these sizes. In the load profile analysis it was seen that the peak load in the morning increased severely as a result of upscaling the installed power in the EV fleet. Additionally, the afternoon peak load increased in magnitude and time for the higher installed power levels. Strategies to minimize the peak loads such as smart scheduling are discussed but not tested for. It is concluded that the V2G technology is a possible strategy to perform FCR-N in an underutilized grid. However, if utilized to such extent that grid updates are required, the cost of grid updates needs to be weighted against the gains from FCR-N provision and cost of EV batteries for further evaluation.