Researchers from Penn State University developed a model that identifies optimal wind farm location and predicts energy output 24 hours in advance.

To develop the model, the team used data from the Analog Ensemble, which is a database of weather prediction data from the U.S. National Center for Atmosphere Research. The team focused on errors in wind farm electricity production predictions and determined that regions with high average wind speeds aligned with more degrees of forecast uncertainty. According to researchers, this points to the difficulty of making wind speed predictions at locations with high average wind speed.

Consequently, the researchers suggest that wind farm builders select locations with low average wind speeds, yet with more consistent and reliable winds.

Although the team realizes that this approach does not offer a definitive answer about predicted wind, they suggest that the model creates a probability curve for wind production that companies can use to inform their decision making.

Typically, identifying the ideal location for a wind farm entails factors such as average wind speed consistency and appropriate terrain. The Penn State researchers suggest that key to selecting wind farm location is wind predictability. Predicting the amount of wind energy generated is crucial, particularly 24 hours in advance.

"Electricity suppliers need to know how much power is available a day ahead," said Guido Cervone, professor of geography, and meteorology and atmospheric science. "They need to have reliable sources because they can't have a blackout. They also do not want to buy more electricity on the spot market because same-day purchases are more expensive."

The research appears in the journal Renewable Energy.