CU Research Could Improve Wind Power With Better Forecasting

Jan 22, 2015

As use of wind energy grows in Colorado, industry and energy grid officials are looking for more sophisticated forecasts to tell them when exactly to use the ephemeral energy source. Current weather models are helpful, but researchers know that mountains and valleys can throw off predictions — requiring more complex weather forecasting.

Enter a team from the University of Colorado, which were recently granted $2.5 million from the U.S. Department of Energy to improve weather prediction for the wind energy industry.

Julie Lundquist, project leader and assistant professor in the Department of Atmospheric and Oceanic Sciences at the University of Colorado at Boulder, said a three-year project is being launched to evaluate current wind prediction models. The research will be based in the Columbia River Gorge region in Washington and Oregon.

"The weather changes in that region have a huge impact on the wind energy production in that area," said Lundquist. "When there are dramatic changes in the wind — either speed up or slow down, we call these wind ramps — whenever there are wind ramps in the region that can have an impact on the operations of the power grid."

Behind the scenes, grid operators have to balance the supply and demand of wind energy. That can be challenging when there are large changes in the wind-generated electricity, said Lundquist.

"If we can improve forecast models that can help with the grid integration challenge," said Lundquist.

Over the next three years, researchers will collect weather observations at the same time that a computer forecast is running. Lundquist said they'll then apply weather data collected to update the forecast models.

"Once we do that we can say 'Oh, this is a physical process that needs to be represented better.' Or 'This needs to be done differently.' So that we can for future forecasts do a better job of representing the physics," she said.    

The grant will be led by Vaisala, a Finland-based company with offices in Colorado, which specializes in environmental and industrial measurements.

Ultimately, Lundquist said the research collected in the Columbia River Gorge region will inform prediction models in Colorado, which ranks among the top 10 states for wind generation in the United States. The American Wind Energy Association ranks Iowa, South Dakota and Kansas among the top three wind-producing states.

"By more accurately observing these phenomena, we can then better represent them in forecast models and predict the ramps," said Lundquist. "So a power grid operator would be prepared for a rapid increase or rapid decrease in wind speed and have another source of power available to compensate."