The kick-off meeting for the research project took place in July of this year (2023). The project is being funded by the Federal Ministry for Economic Affairs and Climate Action and is expected to receive about 1.2 million euros over the next three years. Previously mentioned partners include the University of Kassel, anemos GmbH, and other companies from the wind industry.

The consortium is researching the effects of shortening the measurement period from one year to just a few months – based on the knowledge that seasonal fluctuations in wind patterns are an important factor, as project manager Dr. Alexander Basse explains: "Here in Germany, there are usually significantly lower wind speeds in summer than in fall or winter. If we only measure a few months instead of a whole year, the measurement no longer covers the entire annual wind pattern; it is therefore no longer representative of average wind conditions." In addition, there are seasonal variations in wind direction and or wind shear (the change in wind speed with regards to altitude). "Our AI-based models are designed to recognize and learn these seasonal patterns and apply them to other locations - essentially we are applying machine learning to wind and weather."

Considering Constraints and Time Losses

In addition to wind potential, energy losses must also be calculated in order to be able to estimate exactly how much electricity a wind farm can generate. This includes, in particular, shutdowns due to so-called licensing requirements. The shutdowns and or throttling measures are required to protect people living nearby and to possibly protect endangered species. This would entail ensuring, for example, that noise emissions remain low and that shadows cast on residential buildings only occur to a limited extent. The turbines also must be temporarily shut down when bats start migrating. " Nowadays, it’s close to impossible to realize any wind farm project without such conditions - which makes it even more important to determine their impact on the electricity yield," says Lasse Blanke, managing director of anemos GmbH.

Like wind conditions, essentially all these losses are time-dependent: Bats only fly under certain meteorological conditions, and shadow flicker only occurs when the sun is shining. Predicting the extent and periods of such shutdowns as accurately as possible and calculating the corresponding losses is important and therefore a focus of our project.

Automated Models for Real-World Applications

For the purpose of real-world applications, the STRAIGHT project aims to develop models that calculate yields as automatically as possible based on information about wind and weather as well as the technical conditions of the wind turbines. "Together with our partners from the industry, we are ensuring that our results can also be used in practice and that the models are quickly put to use," says Dr. Doron Callies, a scientist at the University of Kassel. 

But it is not only the wind industry which will benefit from the results: Fraunhofer IEE is adapting the models so that they can be applied to entire regions. This is of particular relevance, as the German states will be specifically designating areas for wind energy use in the coming years. "Our models should be able to more accurately predict how much wind power can be generated on these sites. As a result, we also support the plannability of the energy turnaround in Germany," says project manager Basse.