PV Lighthouse – Accurate and Precise Solar Yield Modelling with SunSolve

Summary

Need

Accurate energy forecasts for photovoltaic (PV) projects are critical for the solar industry. Poor forecasts increase the risk undertaken by investors, thereby increasing the cost of financing a project and reducing the return on investment. Poor estimates inhibit the commercialisation of utility-scale PV projects.

Current energy forecasts use software that apply simple models for the optical, thermal and electrical behaviour of a PV plant. These models require a performance engineer to guestimate a series of project-specific factors that depend on the system configuration, weather and location of the system. Moreover, the models oversimplify the physical behaviour of the PV plant and omit real-world interactions between optics, temperature and electronics. As a result, the forecasted energy of a system rarely agrees closely with the actual energy produced by that system when operational.

Without improvements to PV modelling—whether by removing the guesswork, increasing the sophistication of the models, better quantifying the forecast uncertainties, or validating the forecasts against operational data—lenders will continue to attribute high risk to the yield forecasted by PV projects.

Action

PV Lighthouse seeks to increase the accuracy and precision of energy yield forecasts by implementing sophisticated physics-based models for the optical, thermal and electrical behaviour of PV systems, by better accounting for the interactions between those models, and by greatly improving the procedure to quantify the uncertainty within forecasts.

The project will be comprised of four work packages:

• Work Package 1 – Uncertainty: This package will improve the uncertainty analysis of a yield forecast and tie the other work packages together to provide SunSolve users with an easy-to-understand interface that quantifies the many sources of uncertainty and how they impact yield forecasts.
• Work Package 2 – Thermal: This package will improve the thermal models that predict how cells and modules depend on temperature, and how the module temperature depends on the system configuration and weather.
• Work Package 3 – Electrical: This package will improve the models that predict the electrical behaviour of cells, modules, strings and inverters, as well as the resulting mismatch loss that arises from their interactions. It will include an expansion for emerging technologies.
• Work Package 4 – Optical: This package will expand SunSolve’s current optical model to include physical effects faced by utility-scale developers, such as soiling, wind-stow, far-field shading, variable albedo, sloped terrain and edge brightening.

Outcome

The project aims to increase the commercial readiness of utility-scale solar projects by increasing the accuracy and precision of energy yield forecasting.

Through the integration of superior models and algorithms into SunSolve, and through a validation of SunSolve forecasts against real-world operational data, the project aims to reduce forecast uncertainty by at least 25%. In so doing, the project will demonstrate the technical and commercial value of SunSolve, reduce investor risk for utility-scale PV projects, and reduce the gap between expectations and returns for investors.

Additional impact

The project will improve the simulation capabilities of Australia’s leading PV research institutes by providing access to SunSolve, as well as training seminars and white papers on yield forecasting.