分布式光伏接入下建筑不平衡功率与灵活性需求的多时间尺度分析

Abstract Building imbalance power results from mismatches between building load and distributed renewable generation, requiring flexible resources to maintain stable grid interaction. This paper presents a frequency-domain approach to additively decompose imbalance power into seasonal, weekly, and daily components and evaluate the corresponding flexibility requirements. By applying Parseval’s theorem, the variability indicator of imbalance power is also disaggregated across these distinct timescales. In a case study on a Beijing shopping mall with 40% distributed photovoltaic (PV) penetration, the daily-scale component accounts for 85% of total variability, whereas the smaller seasonal component demands significantly higher flexible energy capacity. Furthermore, after multi-timescale decomposition, a dispatch optimization model shows that the variability reduction curve follows a power-law relationship with available flexible energy at each timescale, allowing accurate estimation from a limited number of optimization results. Moreover, distributed PV influences imbalance power differently depending on building type and regional climate. In northern China, malls with substantial midday load benefit from well-sized PV installations at the daily scale, but their winter consumption peaks intensify seasonal imbalances with increasing PV penetration. In contrast, in central and southern regions, more favorable load–generation alignment alleviates seasonal-scale imbalances. Additionally, south-facing facade PV enhances winter generation, offering a potential solution to mitigate seasonal imbalances in northern cold climates. These findings emphasize the necessity of a timescale-specific analysis for designing building flexible resources and provide a valuable basis for the integration of flexible resources and distributed PV in future low-carbon power systems.