Temporal resolution of EV connection data for forecast of V2G frequency services

Jemima Graham, Fei Teng, Evelyn Heylen

Abstract: This paper investigates the temporal resolution of electric vehicle (EV) connection data required to forecast aggregate frequency response (FR) capacity of vehicle-to-grid (V2G) for FR market participation. This is necessary as there is a mismatch between the temporal resolutions of EV event datasets and the temporal resolutions needed in ancillary services markets. Here, we demonstrate for the first time that 1 Hz event data provides limited added value for aggregate V2G connection forecasting as the fluctuations found in aggregate EV connection time series and distributions of aggregate EV connection time series can be well-represented by minutely data. An additional investigation into the impact of further downsampling of minutely aggregate EV connection time series indicates that filtering the time series prior to downsampling may reduce information loss.