PVRADAR logo

Predict & mitigate soiling during project design

Built for utility-scale teams who need defensible soiling assumptions

Soiling is one of the most underestimated drivers of yield uncertainty, not only in dry and arid regions, but also in humid climates, where organic or "sticky" soiling tends be persistent. For utility-scale PV projects, small changes in soiling assumptions can shift project value, change competitiveness in tenders, and trigger difficult discussions with lenders or EPC providers.

This solution is designed for large project developers and asset owners who require high accuracy, defensible assumptions, and expert support when assessing soiling risks during development and operation.

The challenge: Estimating accurate site-specific soiling conditions

Soiling losses are highly location-dependent and often deviate significantly from generic default values.

If soiling is assumed too low, projects face a real risk of underperformance. If soiling is assumed too high, project value is underestimated.

The core issue is that soiling is not a single number. It is a dynamic process influenced by rainfall patterns, particulate conditions, and cleaning behavior, which requires a modeling approach grounded in real data.

Our solution: Best-in-class soiling modeling, ready to use

PVRADAR provides a ready-to-use Soiling & Cleaning Web App powered by a proven physical soiling model.

The app combines more than 20 years of historic environmental data from multiple sources to calculate daily or monthly soiling loss factors that can be used directly in yield estimation.

For operational assets, users can compare and optimize different cleaning technologies and schedules based on long-term costs and recovered energy.

For PlantPredict users, PVRADAR is available directly within the PlantPredict app.

soiling_app_step3.webp

How it works: From project location to daily soiling losses to optimal cleaning strategy

1. Go from your project file to a soiling estimate in minutes

Start from a PVSYST or PlantPredict project and obtain a location-specific soiling loss estimate in under five minutes.

2. Model daily or monthly soiling loss factors

Calculate daily-resolution or monthly soiling loss factors using PVRADAR’s physical soiling model and more than 20 years of historic data.

3. Optimize cleaning strategy and technology

Compare cleaning technologies and optimize cleaning schedules to minimize total cost of ownership while maximizing recovered energy.

4. Generate a stakeholder-ready report

Create a comprehensive PDF report documenting data sources, assumptions, and results for internal review and external stakeholders.

soiling_app_step5.webp

What you get: Accurate soiling assumptions you can defend

  • Location-specific soiling loss factors at daily or monthly resolution for yield estimation
  • Reduced uncertainty and improved project value in dry and arid regions
  • Cleaning strategy optimization for operational assets, including cost-benefit trade-offs
  • A transparent modeling approach grounded in real data rather than defaults
  • A stakeholder-ready report for lenders, clients, and internal decision-makers

Trusted by energy leaders!

card image
TotalEnergies conducted a blind validation of the PVRADAR soiling model across 9 U.S. PV sites. Using only location and mounting configuration inputs, PVRADAR achieved the highest accuracy (median RMSE = 1.9%), outperforming pvlib’s HSU (2.3%) and Kimber (3.1%), as presented at the 2025 PVPMC Workshop in Cyprus.
card image
Green Energy Park and PVRADAR partnered to assess soiling risks for a large-scale PV project in Morocco. Using local measurements and advanced modeling, cleaning strategies were optimized based on real conditions. The result: a data-driven plan that reduces soiling-related losses and cuts cleaning costs by up to 40%, supporting efficient procurement before construction begins.
card image
PlantPredict users can now easily import 12 average monthly soiling loss factors for any location in the USA as inputs for the yield estimation. This P50 estimate considers typical dust conditions for the area, while excluding localized influences from human activity and birds.

Licenses

PVRADAR provides both project and yearly licenses for the Soiling & Cleaning Web App. Discounts apply for PlantPredict license holders.

See it in action.

We would be happy to give you a private demo and make sure the app fully satisfies your needs.