Impact of limited information on estimating aggregated electric vehicle loading: A distribution system operator perspective

The major deployment of electric vehicles (EVs) brings challenges to the planning and operation of distribution grids, necessitating effective models and estimation approaches for EV charging in distribution grid management. However, distribution system operators (DSOs) may have limited or no access to relevant charging data. This paper investigates how diverse levels of charging information affect the estimation accuracy of aggregated price-sensitive EV power profiles, focusing on distribution system management. Three approaches utilizing varying levels of information are implemented to model the EV charging profiles: a full-information formulation and two reduced-information formulations that rely on aggregate energy measurements combined with either charger occupancy or plug-in statistics. The approaches are evaluated against the individually price-optimized EV schedules over a year, and the resulting cost-based optimal power profiles and associated errors are analyzed. The results indicate that while more detailed information increases the general accuracy of the estimated profiles, reduced-information models can still provide robust estimates for peak load assessment to support congestion management decisions, considering the data constraints and the risk-averse attitude of DSOs.