Wake steering makes wind turbines slightly louder, and a ring of 24 microphones just caught it for the first time
Wind farm engineers have spent years learning to deliberately point turbines away from the wind — a technique called wake steering — to coax more total power from a farm. But almost no one had asked what that intentional misalignment does to the noise a turbine makes.
To find out, researchers arranged 24 microphones in a precise circle around a single 2.2 MW commercial turbine in northern France and recorded its acoustic footprint over four days. What the instruments captured was subtle — but potentially consequential for how wind farms are permitted and regulated.
The clever trick that boosts wind farm output
Wake steering works by pointing upstream turbines slightly away from the wind. Their turbulent wakes deflect sideways, missing the turbines directly downwind. Those downstream machines then operate in cleaner air and produce more power. The upstream turbine gives up a little; the downstream ones gain considerably more. When the math works out, the whole farm comes out ahead.
Paired statistical tests confirmed this increase across multiple wind speed bins, appearing consistently regardless of whether the offset was positive or negative.
Over the past decade, wake steering has moved from research papers into real commercial deployments. It’s one of the more promising tools in wind farm flow control — a field focused on coordinating turbines toward collective goals beyond individual machine performance.
Noise was almost an afterthought. An expert survey on wind farm flow control ranked acoustic management second to last among priorities. That calculus is shifting — particularly in France, where noise regulations are among the strictest in Europe. Most newly developed French wind projects now operate under acoustic curtailment plans, which already trim average annual energy output by about 2.8%. When noise rules cost that much, the acoustic side effects of any new control strategy stop being theoretical.
A circle of microphones around a single turbine
The experiment was designed to answer the noise question directly. Researchers placed 24 sound level in a ring around a single commercial turbine — a 2.2 MW, 360-foot-diameter machine operated by ENGIE Green in northern France — each meter (3.2 feet) spaced 15 degrees apart at roughly 442 feet out. To keep measurements clean, the two nearest neighboring turbines were shut down for every test round, and the closest operating turbine was more than a kilometer (0.62 miles) away.
The misaligned turbine was stepped through yaw offsets from −20° to +20° in 10-degree increments, holding each position for roughly 40 minutes. Standard nacelle anemometers become unreliable when a turbine is deliberately pointed away from the wind, so the team deployed both a nacelle-mounted and a ground-based lidar for wind speed and direction references. The four-day campaign in March 2024 yielded more than 21 hours of usable acoustic data across wind speeds ranging from 6 to 13.5 m/s.
What the data revealed: noise is not a simple circle
The first surprise came before any yaw misalignment was applied. Even with the turbine pointed straight into the wind, its noise pattern was far from the uniform ring that most regulatory models assume.
Sound levels varied by up to 4 dB(A) depending on direction. Quiet zones appeared near the rotor plane, running 3.5 to 4 dB(A) below the louder directions — and the direction of maximum noise emission wasn’t straight downwind. It was consistently off-axis, contradicting most semi-empirical tools used in permitting.
Statistical analysis confirmed two persistent asymmetries. The downstroke side of the rotor was on average 0.6 dB(A) louder than the upstroke side. The downwind side was 0.4 dB(A) louder than the upwind side. Both differences were statistically significant, with p-values far below 0.05. Research suggests blade camber alone can cause deviations of up to 1 dB(A) between pressure and suction sides — a subtlety standard models miss entirely.
Wake steering adds roughly half a decibel — and it matters
When yaw misalignment was applied, the turbine’s overall sound power level rose by approximately 0.6 dB(A). Paired statistical tests confirmed this increase across multiple wind speed bins, appearing consistently regardless of whether the offset was positive or negative. The noise pattern also grew more complex, rotating with the turbine’s axis and developing additional lobes and dips.
Spectrally, the increase was most pronounced at low frequencies (below 1,000 Hz) and high frequencies (above 5,000 Hz) — suggesting yaw misalignment primarily affects turbulent-inflow noise rather than trailing-edge noise. A prior NREL study reported noise reductions under yaw misalignment. The researchers here argue the discrepancy likely reflects different turbine control strategies: the NREL turbine reduced rotor speed when yawed, which would lower noise. The French turbine showed no such rotor slowdown — so no corresponding noise drop.
Why 0.6 decibels can rewrite a curtailment plan
Half a decibel sounds trivial. Perceptually, it nearly is — 1 dB is roughly the threshold of human hearing. In regulatory contexts, though, the arithmetic is less forgiving.
Current acoustic curtailment plans assume turbines radiate noise equally in all directions. A variation of just ±0.5 dB in source power level can force meaningful changes to curtailment schedules, with knock-on effects on annual energy production of around 1%. In a regulatory environment where farms already lose 2.8% of output to acoustic restrictions, that margin represents real money over a project’s lifetime.
The directivity findings compound the problem. Plans built around omnidirectional sources may be simultaneously too restrictive in some directions and too lenient in others. What’s needed is a multi-objective approach that jointly optimizes power output, structural loads, and noise. The EU Horizon Europe–funded TWAIN project will use this dataset to develop more advanced turbine noise models — giving operators tools to design smarter curtailment strategies that don’t leave energy on the table unnecessarily.
More information is available here: Duc, T. and Finez, A.: Experimental investigation of the effect of wake steering on the noise emission of a commercial wind turbine, Wind Energ. Sci., 11, 3249–3272, https://doi.org/10.5194/wes-11-3249-2026, 2026.
Daniel García is an Editor-in-Chief with strong expertise in structural work and engineering principles. He combines this technical foundation with deep knowledge of energy, spatial design, and emerging technologies, bringing a forward-thinking and analytical approach to editorial leadership.
