Optimizing cleaning schedules for spatially distributed photovoltaic installations with site-specific variations

The rapid growth of photovoltaic (PV) installations has led to the deployment of dispersed PV fleets across diverse locations. Effective operation and maintenance (O&M), particularly proactive cleaning, is critical to mitigate soiling, which significantly impacts system performance. Standard cleaning strategies tailored to individual sites are challenging to scale across multi-system fleets due to site-specific variations, resulting in high costs and operational complexities. This paper introduces a mixed-integer linear programming model to optimize cleaning schedules for dispersed PV systems. The model accounts for site-specific factors such as soiling losses, cleaning costs, and generation capacities, dynamically adjusting cleaning frequencies to enhance fleet-wide performance and profitability. Results highlight the benefits of tailored scheduling, with some sites requiring more frequent cleaning while all contribute to overall profitability, validating the fleet-wide approach. Computational experiments show substantial profit gains, with dynamic scheduling achieving a 15–34% profit increase compared to interval-based scheduling and a 2–5% improvement over threshold-based approaches. By aligning resource use with site-specific conditions, this approach supports renewable energy goals by stabilizing production and minimizing environmental impact. The framework provides a robust, cost-effective solution for O&M contractors to optimize PV operations and promote sustainability, with potential for real-time enhancements to adapt to evolving conditions.