Turbines on an Alberta ridge were turned deliberately off the wind and the 6 machine array they sit in gained as much as 13 percent at average wind speeds, but the yearly total from wake steering came out insignificant

A row of six machines along a ridge in southern Alberta, all facing the same way except one.

That one is visibly turned aside. Its rotor sits at an angle to the wind coming across the grass.

Nothing is broken and nobody is repairing it. The offset was commanded on purpose.

Wind projects tend to be judged on the annual figure rather than the peak, which is the same arithmetic that stalled wind capacity in one Spanish region.

The machine is now making less power than it could, and that is the intended result.

It is being asked to lose something.

On the promise that the row behind it gains more.

Why throwing away power can raise the total

A turbine does not simply take energy out of the wind. It leaves a wake behind it, a slower and more turbulent column of air.

Any machine standing in that column sees a weaker and rougher wind than the one in front did.

On a tightly spaced site those losses are large, and they are the reason a row of identical turbines does not produce identical power.

Yawing the front machine a few degrees off the wind deflects its wake sideways rather than removing it.

The front machine gives up a slice of its own output to do that, because a rotor at an angle to the flow catches less of it.

The bet is that the wake moves far enough to clear the machine behind and that the swap comes out in the plant’s favor.

What the six machines on that ridge are

The site is an operating commercial wind farm in Alberta, run by a Canadian utility rather than by a laboratory.

The array studied was six turbines of the Vestas V80 type, five rated at 1.8 megawatts and one at 2 megawatts.

These are not new machines and they are not unusually large, which is part of why the result travels.

The experiment concentrated on a narrow band of wind direction, close to 330 degrees, where the row lines up almost exactly with the flow.

In that band the waking is severe, and the potential gain from moving a wake is therefore at its largest.

Everything measured therefore describes that band and does not describe the rest of the year on that ridge.

The team picked the worst case deliberately rather than an average one.

The numbers and how long they were measured

At wind speeds near the site average the array gained between 7 and 13 percent while the offsets were applied.

At low wind speeds the gain was far larger, between 28 and 47 percent, because that is where wake losses bite hardest.

The scatter in power output also fell, by as much as 72 percent, which matters to anyone scheduling a plant rather than sizing one.

Those are real measurements from an operating farm and not a simulation, which is what gave the work its weight.

The campaign ran for about ten days, and that is the number to hold onto.

Ten days is enough to prove an effect exists and nowhere near enough to price a year of it.

The sentence the coverage kept leaving out

The same paper that reported those double digit gains also reported the annual result, and it was blunt.

The resulting gains in annual energy production at that farm were insignificant.

Both statements are true at once, and the reason is that the large gains sit in a narrow slice of wind direction and a narrow slice of wind speed.

Multiply a big number by the small fraction of hours in which it occurs and the yearly total collapses toward zero.

Wind projects tend to be judged on the annual figure rather than the peak, which is the same arithmetic that stalled wind capacity in one Spanish region.

The measured plant level results are set out in the published abstract.

A control that pays off a few hours a year is a research finding.

Where the idea stands now

A separate campaign in Colorado went at the same question with a different design and a longer record.

There only two machines were controlled, and the measured reduction in wake losses came to about 6.6 percent across the sector studied.

Translated into a whole plant that works out to roughly 1 to 2 percent of annual production, worth on the order of a million dollars a year at a large site.

That is a modest number and it is also a real one, which is the honest summary of wake steering today.

Measuring what a wake actually does still needs instruments in the water or the air, the way researchers tracked three turbines off Portugal.

The Colorado figures and the value estimate are summarized by the federal writeup.

The technique works, and what it is worth remains site by site and year by year.