Text: Hanne Kokkegård - DTU Compute and Energy Cluster Denmark
For three years, the partners in the Weather2X project will work on improving short-term forecasts for energy production based on the renewable energy sources of sun and wind, which vary over the day and seasons.
If the prediction for energy production could be improved by 10 percent, it would make it possible to adapt the power grid to handle large amounts of renewable energy better. Today wind turbines sometimes must be stopped because the power grid is overloaded. Besides, it will make it easier to plan energy production from other energy sources when the sun is not shining, and the wind is not blowing.
At DTU Compute, the researchers in the section Dynamical Systems will help fine-tune the weather forecasts locally to be able to understand the correlation between the weather forecast and the so-called power output - how much electricity the wind turbine send into the grid.
Correlate weather forecasts and data from power production
Today, wind farms in the landscape are established in a way that the wind turbines produce as much electricity as possible around the year. Nevertheless, very local conditions for the individual turbine come into play and cause the forecast for the total energy production in the wind farm to deviate from reality.
With the help of statistical analyses, the researchers will correlate weather forecasts and high-resolution data from power production on both the individual wind turbine and the total wind farm and from that get a more precise understanding of sources of uncertainties. Based on that, they will develop models that are capable of considering the uncertainty that local physical conditions create.
“Forecasts sometimes miss the mark, and they will of course also do so in the future. But it is a question of whether we can miss less and be more precise a larger part of the time, for example on an hourly basis by being able to say more about the individual wind turbines based on data from the wind farm,” says Associated Professor Jan Kloppenborg Møller.
As an additional benefit, knowledge from Weather2X could ultimately create a more reliable basis for energy price settings when you better know the total energy production. That could benefit the green transition since customers are more interested in acting flexible and moving part of their consumption into hours of the day when the price is low and the grid is full of renewable energy.
Balancing the system
The DTU spinout company ENFOR currently provides forecasts for approximately 15 percent of all wind and solar energy plants in the world. In the company they see great potential for Weather2X:
“We expect the new collaboration to save time and money for our key customers, who are both system operators and those who trade electricity. These include asset owners like Ørsted or Vattenfall – or electricity traders. For all of them, inaccurate energy forecasts can cause challenges in the electricity system they buy and supply power to, creating unnecessary costs and making the integration of renewable energy more expensive," says CEO of ENFOR Mikkel Westenholz.
In Ørsted, they believe the project will lead to significant improvements, especially in the short term.
"I also hope we will have even better energy forecasts by the end of the innovation project. For example, this past weekend (some weeks ago), we stopped our power plants and delivered all electricity, heat, and system services from wind and electric boilers. This requires accurate forecasts, and we expect to improve this discipline through Weather2X," says Jan Hvidberg, Lead Analyst at Ørsted.
Measurements at turbine height
DMI has been supplying selected weather data to the energy sector for many years – a product they have long wanted to improve.
Normally, meteorological observations of wind are made at 10 meters height and temperature at 2 meters height. Wind turbines' blades spin at heights from approximately 50-250 meters. Up there, the wind blows more than near the ground, and the difference depends on the weather.
"We lack the right type of measurements to compare with, so we can fine-tune our models. DMI has virtually no observations from these heights, either of the wind or the energy the turbines produce, making it impossible to effectively verify our forecasts," says Henrik Vedel.
As a partner in Weather2X DMI now gains access to the energy sector's observations, including measurement equipment on the turbines.
"That's one of the main reasons DMI is part of the innovation project. Here we expect, together with our excellent partners, to better predict the wind where the turbines operate. The same goes for solar radiation forecasts throughout the Danish area. We expect that closer collaboration with the major players in the renewable energy sector will lead to improvements that benefit both the sector and DMI's models," says Henrik Vedel.
