Understanding spatiotemporal dynamics of V2G participation in megacities: A data-driven study

Vehicle-to-Grid (V2G) technology has the potential to enhance energy resilience and optimize electricity use, but its effectiveness depends on how participation unfolds at scale. Most existing studies rely on small-sample surveys or limited, context-specific pilot observations, overlooking the heterogeneous V2G participation shaped by mobility needs and spatiotemporal variation, which determine when, where, and how much vehicles can exchange energy and how participation scales different across diverse urban contexts. To address this gap, this study develops a data-driven agent-based modeling framework that captures heterogeneous V2G participation in megacities by integrating heterogenous mobility needs—derived from three months of trajectory data on 18,000 private battery electric vehicles (BEVs) in Shanghai—with spatial variation in land use and infrastructure. This combination reflects both seasonal dynamics and citywide heterogeneity. The model simulates daily charging and discharging decisions, with agents maximizing arbitrage profits while maintaining mobility requirements. Results show that V2G integration intensifies BEV–grid interactions at the individual level, increasing charging volumes by up to fivefold and generating average monthly net gains of 230 CNY per user. At the system level, spatiotemporal variations emerge, shaped by electricity pricing, urban–suburban context, and land-use functions. The framework also demonstrates strong adaptability and transferability, supported by submodule replacement analysis and sensitivity analysis across key parameters, providing a policy-oriented tool for anticipating user behavior and system outcomes in megacities as V2G moves beyond pilot testing. These insights support more effective infrastructure planning and regulatory strategies for a smooth and equitable transition toward large-scale V2G adoption.