Think you know which way the wind blows?
Then the folks at Bonneville Power Administration want to hear from you.
In a Pacific Northwest twist on the X-Prize, two research teams have been invited to compete, to see which can best predict wind changes as much as 36 hours in advance.
“Wind power is a great energy source, but we could make even better use of it if we could anticipate big changes,” said John Pease, who’s overseeing the contest for Bonneville. The “friendly competition,” he said, will put some of the world’s best brainpower to work building prediction models.
It’s an international challenge: AWS Truewind of New York vs. Energy and Meteo Systems of Germany. The team with the most accurate predictions will be in line for a BPA contract to develop a full-scale wind forecast model for the region.
Wind, energy analysts say, has both tremendous potential and terrific pitfalls. The most obvious downside – that it doesn’t always blow when you want it to – is complicated by the fact that electricity on the grid cannot be stored. Instead, the power being used must roughly equal the power being produced.
Better forecasting, Pease said, will help grid managers (think air traffic controllers for electrons) smooth the balance of anticipated supply and demand.
If successful, the predictive model would be the first created specifically to foresee sharp changes – called “ramps” – in wind energy. And if successful, it will be key to blending the many power sources expected to make up tomorrow’s grid.
The teams will begin this month, Pease said, projecting the blow at four Oregon and Washington wind farms.
Phil Barbour, an Oregon-based research meteorologist, called it “an exciting project,” with potential to support a growing wind-power industry around the globe. “Winds are often very localized and difficult to predict,” he said. “It’s even harder to predict these specific ramp events. This is a huge challenge for the competing teams.”
Wind power on BPA’s regional grid, Pease said, can vary over one hour’s time by as much as 1,000 megawatts – the equivalent of a big nuclear plant. That means BPA must keep backup energy reserves on tap, and must charge the wind producers for those reserves, which in turn increases the cost of wind power.
A solid predictive model, Pease said, could reduce the need for reserves, and so lower energy costs for consumers.
Michael Jamison, The Missoulian – http://www.missoulian.com/news/state-and-regional/article_3436f418-895b-11de-91c3-001cc4c002e0.html