SOLARFARMER MODEL VALIDATION

SolarFarmer is DNV’s solar PV design and energy assessment model and software. It combines detailed 3D modelling of sites with robust calculation engines to deliver transparent, bankable results for developers, investors, and lenders.

The SolarFarmer model has been extensively validated to ensure its results can be trusted for project design and financing. The validation work benchmarks the model against measured data and independent datasets, providing confidence in the accuracy of its results.

DNV has conducted validation studies on 6 different operating projects using this framework across the United States, totalling 1.19 GW - View DNV new release here.

For full details of the validation studies, including published papers and presentations, see the Introduction to Validation page in the SolarFarmer Technical Documentation.

Metrics for Model Accuracy

The SolarFarmer model validation applies established statistical metrics to compare modelled results against measured data. These include:

• Mean Bias Error (MBE): quantifies average over- or under-prediction
• Root Mean Square Error (RMSE): measures the overall magnitude of deviations
• Relative MBE (rMBE) and Relative RMSE (rRMSE): normalised against mean measured output

These metrics are standard within the solar industry and provide a consistent, quantitative basis for assessing model accuracy.

Learn more: Metrics for Model Accuracy.

‍

Detailed Results Examples

The validation includes worked examples showing how SolarFarmer results compare with measured site data. These examples illustrate the statistical and graphical analyses used, including:

• Scatter plots of measured versus modelled output
• Histograms of relative Mean Bias Error (rMBE)
• Seasonal and diurnal error distributions (e.g. rMBE by time of day)
• Box plots showing error spread and symmetry
• Correlation and auto-correlation plots to test for systematic bias

These examples demonstrate how the SolarFarmer model performs under real project conditions and highlight the transparency of the validation process.

See: Detailed Results Examples.

Results Summary

The validation results are summarised across multiple validation sites to show the overall accuracy and consistency of the SolarFarmer model. The summary highlights:

• Mean Bias Error (MBE) and relative MBE (rMBE) values across sites
• Root Mean Square Error (RMSE) and relative RMSE (rRMSE) ranges
• Consistency of accuracy across different datasets and conditions
• Evidence that errors are small, stable, and well-characterised

This summary provides a clear, aggregated view of model performance, reinforcing confidence in SolarFarmer’s reliability for project assessment and financing.

View the summary: Results Summary.