Unlocking peak shaving: How EV driver heterogeneity shapes V2G potential

Abstract

As electric vehicles (EVs) and variable renewable energy sources rapidly expand, vehicle-to-grid (V2G) services have emerged as a promising strategy to enhance grid flexibility. However, their effectiveness critically depends on EV drivers’ willingness to participate, which is shaped by behavioral heterogeneity and operational constraints. While previous studies have explored participation preferences, they have largely overlooked time-specific availability and its implications for system-level flexibility. This study addresses this gap by integrating a discrete choice experiment with latent class modeling to analyze both user preferences and the peak shaving potential of V2G. To capture temporal availability more accurately, “weekday connection time” is introduced as a novel contract attribute, enabling realistic estimates of time-specific charging and discharging flexibility. The analysis identifies three distinct driver segments, each characterized by unique preferences for monetary incentives, minimum connection days, charger accessibility, weekday connection frequency, and state-of-charge guarantees. Scenario-based simulations incorporating these heterogeneous profiles indicate that tailored V2G program designs could reduce peak net load by up to 22.9 % by 2030. These findings underscore the importance of differentiated policy instruments and aggregator strategies that reflect user diversity. The study provides a behaviorally grounded framework for designing inclusive and effective V2G programs that contribute to a more flexible and sustainable power system.