Builders of some of the largest wind power installations have come up against a problem: with more than 50 or 100 turbines in a facility, the effect of the upwind turbines on wind speed farther downwind becomes harder and harder to predict. According to an article in Wednesday's MIT Technology Review, wind facility developers wanting to install several hundred turbines at a go may be faced with disappointing power output unless our computer models get a lot better.
In the article, author Kevin Bullis notes that each row of turbines dramatically effects the turbulence in the wind stream downwind, with surprising results. Bullis quotes researcher Charles Meneveau of Johns Hopkins University as saying that the turbulence created by larger turbine arrays can actually propagate upward a kilometer above the turbines. (That's about 3,300 feet for you non-metrical readers; about ten times the hight of the tallest turbine assemblies.)
Complicating our modeling of the way wind interacts with turbines is the fact that wind isn't all that predictable even without the turbines there. Wind can blow steadily or change its speed and direction constantly, and predicting the interactions of such a variable entity with turbines is beyond the grasp of even the most sophisticated computer models at presewnt.
The result? As Bullis says, developers of wind projects in the smaller sizes prevalent five years ago often found their expectations of power delivery were about 10 percent too high, leading to lower earnings than projected -- and problems paying off debts. Bullis notes that models have improved somewhat since then, but their accuracy isn't keeping pace with the increase in the size of proposed wind installations. The Chokecherry and Sierra Madre project in Wyoming, for example, with construction expected to begin next year, is slated to include more than 1,000 turbines.
Bullis points out that that same turbulence in the wakes of turbines actually makes larger installations feasible, by literally pulling wind down toward the ground toward the downwind edge of a large installation. Still, it's worth noting that Johns Hopkins' Meneveau says most turbine installations are packed more tightly than a strict analysis of wind dynamics would merit. It may be that the best way to avoid unpredictable output from large clusters of wind turbines is to keep the turbines widely separated, by not cramming dozens or hundreds of them into single windy areas.