Assessment on electric vehicle dispatch-based demand-side energy storage requirement reduction via spectral decomposition

As renewable integration increases, power systems experience great net load variability, requiring coordinated solutions from both supply and demand sides. This study employs spectral decomposition to analyze net power load patterns, quantify demand-side energy storage requirements, and evaluate the potential of optimized electric vehicle (EV) charging in reducing these storage demands. The results demonstrate that spectral decomposition effectively eliminates the impact of long-term net load and isolates representative weekly fluctuation patterns, resulting in a more robust evaluation. In the case study of Beijing, the annual coefficient of variation, which reflects net load variability, decreases from 3.726 to 3.333. Compared to the filtered net load, the median weekly relative energy storage power demand reduces from 2.0 to 1.8, with the variation range changing from 0.7 to 3.5 to 1.0–2.8. The median relative capacity demand declines from 13.7 h to 12.0 h. The analysis of EV charging optimization strategies reveals that unidirectional charging already provides most of the storage reduction, while bidirectional charging offers further improvement. In a fully electrified vehicle scenario, the optimal configuration of rated power and capacity for EV under the unidirectional charging strategy is 7 kW and 40 kWh, while it is 10 kW and 40 kWh under a bidirectional charging strategy. With unidirectional control, the median storage power requirement already drops markedly from 1.8 to 0.5, and the capacity demand from 12.0 h to 3.5 h. Bidirectional control can further reduce both metrics to 0.0, but at higher cost.