Abstract
Large-scale deployment of electric vehicles (EVs) to meet climate goals imposes dual pressures on power system reliability: increased electricity demand and intensified daily load variability arising from uncertain charging behavior. This study quantifies the magnitude and spatial distribution of behavior-induced load variability in China under alternative climate scenarios, using a scalable model calibrated with minute-level EV charging data. EV adoption consistent with carbon-neutrality goals is projected to raise electricity demand by 3.2% in 2050, requiring an additional annual investment of approximately 220 billion CNY (about 31 billion USD) in battery storage capacity. Accounting for behavioral uncertainty, potential load fluctuations could increase by up to 82.7%. Regional heterogeneity in charging patterns drives distinct spatial profiles of electricity variability, influencing the effectiveness of demand-side management strategies. These findings highlight the importance of targeted, behaviorally informed interventions to mitigate grid reliability risks under deep EV adoption.