Empirical grid impact of in-home electric vehicle charging differs from predictions

Abstract

Accurate assessment of the impact of electric vehicle (EV) charging on the electric grid is critical for energy policymakers to design efficient EV subsidy programs as well as to provide reliable electricity infrastructure. Despite the fact that 80 % of EV charging is conducted with residential in-home chargers, very few empirical studies have examined the load and environmental impact of residential EV charging based on actual electricity consumption data. Our paper fills this critical gap in the literature, applying a difference-in-differences approach to high frequency smart meter data of about 1600 EV homes from 2014 to 2019 in Arizona, United States. First, we find that the electricity demand during the system peak hours from 6 to 8 pm in summer can increase by 7–14 % at an average household with in-home EV charging. Second, EV households respond to electricity pricing signals by increasing their charging in lower-priced off-peak hours within the EV-specific time-of-use (TOU) pricing. Third, we find evidence of rebound effects in driving that lead to a reduction in home-electricity consumption in certain hours of the day. Lastly, we show that our empirical estimation of the grid impact due to in-home EV charging is different from that predicted by existing simulation models due to factors such as consumer behaviors. Such deviations between predicted and actual behaviors imply potential adjustment of relevant policy interventions.