Penn State Works on Improving Weather Forecasts to Help Renewable Energy Production

Researchers at Penn State University continue to work this summer on developing better models to help the solar and wind energy industries.

Researchers will be in Boulder, Colorado, until August poring over weather data compiled by the National Center for Atmospheric Research. Penn State model developers will learn by comparing their predictions to what actually occurs with factors such as cloud cover and wind speed, the university said in a press release.

These industries could plan accordingly to what the models are showing, AccuWeather Senior Meteorologist Bob Smerbeck said.

"A strong wind pattern or light wind pattern or a sunny or cloudy pattern that shows up consistently from run to run would give an industry confidence on how to generate energy or even how shift to alternate energy or buy energy if the weather will not cooperate," Smerbeck said.

The summer research will focus what's known as the "day market," or the 24 hours from a specific time, Penn State said.

"As the use of wind energy has greatly expanded over the last decade, wind forecasts have become critical tools for grid operators to predict wind output from several hours to several days in advance and plan their operations accordingly," Senior Director of Research Michael Goggin of the American Wind Energy Association said.

"This is similar to the weather models grid operators have long used to predict electricity demand based on how many people will be running air conditioners or heaters depending on the temperature and other weather conditions," Goggin said.

Grid operators use the same tools to accommodate changes in wind output that they have always used to accommodate changes in electricity demand and supply - typically trading with neighboring power systems and changing the output of power plants so that supply matches demand, Goggin said.

"Grid operators have gotten very good at using specialized weather forecasts, called wind energy forecasts, to predict wind energy production from several hours to several days in advance," Goggin said. "They are also assisted by the fact that, across the large geographic areas covered by America's interstate power systems, changes in wind energy output in one region are typically canceled out by an opposite in another region. Moreover, a decrease in wind output is often canceled out by an unrelated decrease in electricity demand or an unexpected increase in supply from another power plant."

Accurate forecasts help both power grid and wind energy operators.

"Wind forecasts make grid operators more comfortable with placing greater amounts of wind energy onto their power system," Goggin said. "Forecasts also allow grid operators to maximize the benefits of wind energy by turning off fossil power plants that will not be needed because wind energy will be providing the electricity instead."

Weather models have greatly improved since Smerbeck began as an AccuWeather meteorologist in 1984.

"Over the years, the models have become more realistic as the physics improved and they are more graphical than the standard black lines on white paper of 30 years ago," Smerbeck said. "The accuracy of a five-day forecast today is as good as a three-day forecast 20 years ago."