基于最优清洗调度的印度大型光伏电站非均匀积灰损失评估及性能提升策略

Abstract Soiling significantly impacts the efficiency of photovoltaic (PV) systems, especially in regions with heavy dust deposition like India. The issue is exacerbated by spatially non-uniform soiling in utility-scale PV plants, where certain areas of the plant experience higher losses than others, complicating maintenance efforts. In this study, we analysed string-level SCADA data from a 50 M W p utility-scale PV plant in South India divided into several zones to create detailed soiling maps. Using these maps, we developed both string-optimized and zone-optimized cleaning methodologies. The string-optimized approach utilized four specific cleaning thresholds to help determine the most profitable cleaning areas in each zone, while the zone-optimized approach aimed to streamline cleaning processes, enhance PV plant performance, and resource efficiency. Additionally, unlike previous studies, this analysis accounted for DC cabling losses, further refining the evaluation of soiling impact. The results in-terms of cleaning profit generated were compared with the same made by actual logged cleaning. Additionally, we performed a sensitivity analysis by varying solar PV electricity tariffs and cleaning costs to evaluate the economic viability of different cleaning strategies. The analysis indicated that the 85% cleaning threshold is the most economical, particularly as PV electricity prices continue to decline. Our findings suggest that structured cleaning schedules based on soiling data can significantly improve PV plant performance and profitability. This approach can be replicated in similar PV plants to support India’s growing PV sector, ultimately helping the country become a global leader in solar energy.