New algorithm to improve solar energy forecasting


GNARUM, in cooperation with the Universities of Huelva and Sevilla, has developed a new algorithm for solar energy forecasting that considerably improves deviations for hourly predictions.
This new algorithm represents a 5 point improvement in the deviated energy between forecasts and real production and an estimated saving of €500,000 for a 500MW portfolio according to the deviation costs published by REE (Spanish Grid Operator). Carlos Alberto Castaño, Head of R&D, explains that “this algorithm comes from a previous research to generate highly accurate predictions for those cases in which a daily metering update is not possible, or for those plants without specific metering data”.
The implementation of this algorithm -which has already been tested- takes place only a few months after GNARUM released a new Control Panel for its clients that includes new functionalities and tools to manage forecasts, with an all new design.
GNARUM R&D further steps are to optimize algorithms for the rest of renewable technologies and to combine different data metrics from other prediction models.
GNARUM improves its forecasting results and consolidates itself as the IT Solutions Provider for renewable energy.