Wake-induced variations in noise levels and amplitude modulation for two interacting wind turbines

This study uses numerical simulations to demonstrate that turbine-turbine interactions significantly alter noise levels and amplitude modulation, with downstream alignment causing substantial increases in sound pressure and modulation due to wake-induced flow focusing, while side-by-side and staggered arrangements generally reduce modulation through spatial averaging and rotor dynamics.

The Gist

Imagine you are standing in a vast, quiet field. Suddenly, a giant wind turbine starts spinning nearby. It makes a "whoosh-whoosh" sound, like a giant fan. Now, imagine there are two of them.

This paper is like a detective story about what happens to that "whoosh" sound when two turbines are standing close to each other. The researchers wanted to know: Does the wind from the first turbine change how the sound of the second turbine reaches your ears?

Here is the story of their findings, broken down into simple concepts.

1. The Invisible River (The Wind Wake)

First, you have to understand that wind turbines don't just spin; they mess up the air around them.

• The Analogy: Think of a wind turbine like a boat moving through water. The boat leaves a choppy, slower-moving trail of water behind it called a "wake."
• The Reality: When a turbine spins, it creates a similar "wake" in the air. This wake is a tunnel of slower wind and swirling turbulence trailing behind the turbine.

2. The Three Ways to Stand

The researchers tested three different ways to arrange two turbines:

• The Train (One behind the other): Turbine A is in front, and Turbine B is directly behind it in the wind's path.
• The Twins (Side-by-side): They stand next to each other, like twins holding hands.
• The Zig-Zag (Staggered): One is slightly ahead and to the side, like a zig-zag pattern.

3. What Happens to the Sound?

Scenario A: The Train (One Behind the Other)

This is where things get wild.

• The Sound Trap: When Turbine B sits in the "wake" of Turbine A, the air is moving slower and is more turbulent. But here's the magic trick: The shape of this wind tunnel acts like a magnifying glass or a funnel.
• The Result: The sound from Turbine A gets squeezed and focused into a tight beam that shoots straight downwind. If you are standing behind them, the noise gets louder (by several decibels) and the "whoosh-whoosh" rhythm (called Amplitude Modulation) becomes much more intense and annoying.
• The Catch: Turbine B itself is actually quieter because it's sitting in the slow wind of Turbine A's wake. It's like a singer trying to sing while standing in a strong headwind; they can't push as hard. So, the loud noise you hear is mostly the first turbine, amplified by the wind tunnel created by the second one.

Scenario B & C: The Twins and The Zig-Zag

• The Result: When the turbines are side-by-side or staggered, the "wind tunnels" don't line up perfectly to focus the sound.
• The Smoothing Effect: Imagine two people clapping their hands. If they clap at the exact same time, it's a loud CRACK. If they clap at slightly different times, the sounds blend together and become a softer clap-clap.
• The Finding: Because the two turbines are slightly out of sync or in different wind spots, their sounds mix together. This actually reduces the "whoosh-whoosh" rhythm. The noise becomes more steady and less annoying, though the overall volume is slightly higher than a single turbine (but not by much).

4. The "Beating" Heartbeat

The researchers discovered something fascinating about the rhythm of the noise, called Amplitude Modulation (AM). This is the "swish-swish" sound that drives people crazy.

• The Perfect Sync: If both turbines spin at the exact same speed and start at the exact same angle, their "swish" sounds can either cancel each other out (silence) or boost each other (loud noise), depending on their position. It's like two pendulums swinging.
• The "Beating" Effect: In the real world, one turbine might spin at 10.2 RPM and the other at 10.3 RPM. This tiny difference creates a beat.
  • The Analogy: Think of two guitar strings tuned slightly differently. When you play them together, you hear a "wah-wah-wah" pulsing sound.
  • The Result: The noise from the turbines doesn't just stay loud or quiet; it pulses on and off in a slow, rhythmic beat. Sometimes the "swish" is very strong, and a few seconds later, it almost disappears. This makes the noise unpredictable and harder to get used to.

Why Does This Matter?

Wind energy is great for the planet, but the noise can be a problem for neighbors.

• For Planners: If you build a wind farm, you can't just place turbines randomly. Putting them directly behind each other in the wind direction might create a "noise laser" that is much louder than expected.
• For Neighbors: The study shows that the rhythm of the noise (the modulation) is just as important as the volume. Small changes in how the turbines spin can make the noise feel much more annoying or much more bearable.

The Bottom Line

Two wind turbines together are not just "two times the noise." They interact with the wind in complex ways.

• Line up? You get a focused, louder, more rhythmic noise beam.
• Side-by-side? The noise blends out, becoming steadier and slightly less annoying.
• Different speeds? You get a pulsing "heartbeat" sound that comes and goes.

By understanding these invisible wind tunnels and sound interactions, engineers can design wind farms that generate clean energy without turning the neighborhood into a noisy place.

Technical Summary

1. Problem Statement

Wind turbine noise is a primary barrier to public acceptance and a driver for economic curtailment (reduced power generation) to meet noise regulations. Current engineering models for noise prediction often rely on simplified assumptions, such as point sources and uniform wind profiles, which fail to capture complex real-world interactions.

• The Gap: While the effects of single turbine wakes on sound propagation are studied, the systematic impact of multiple interacting wakes (specifically from pairs of turbines) on both Sound Pressure Levels (SPL) and Amplitude Modulation (AM) remains poorly understood.
• The Objective: To investigate how the relative positioning of two wind turbines (aligned, side-by-side, or staggered) alters noise generation and propagation, with a specific focus on the perceptual metric of Amplitude Modulation (the "swishing" sound).

2. Methodology

The study employs a high-fidelity numerical framework combining three distinct models:

• Atmospheric Flow (LES): Large-Eddy Simulations (LES) were used to model the atmospheric boundary layer (ABL) under stable conditions (worst-case scenario for noise propagation). An actuator disk model represented the turbines. The simulation utilized a concurrent precursor method to generate realistic turbulent inflow.
• Source Model: Based on Amiet's theory, the model calculates Turbulent Inflow Noise (TIN) and Trailing Edge Noise (TEN). It uses LES data (mean velocity and turbulent dissipation) from the rotor plane to determine source characteristics. The model accounts for the specific rotational speed ($\Omega$) of each turbine, which varies based on the local wind speed in the wake.
• Propagation Model: A Parabolic Equation (PE) method (Ostashev et al., 2020) was used to simulate sound propagation through the moving atmosphere. This approach accounts for wind refraction without relying on effective sound speed approximations.
• Configurations: Four scenarios were simulated with turbines spaced 4 rotor diameters (480 m) apart:
  1. Case B: Single isolated turbine (Reference).
  2. Case C: Column (aligned downstream).
  3. Case L: Line (side-by-side).
  4. Case S: Staggered (diagonal).

3. Key Contributions

• Systematic Wake Interaction Analysis: The study isolates the specific acoustic effects of wake superposition on noise propagation, moving beyond single-turbine studies.
• AM Sensitivity to Rotor Dynamics: It highlights that Amplitude Modulation is highly sensitive to the relative angular offset and rotational speed differences between turbines, introducing "beating" effects when speeds differ slightly.
• High-Fidelity Coupling: The work demonstrates a robust chain linking LES flow fields, physics-based source models, and wave-based propagation solvers to predict complex multi-turbine acoustic fields.

4. Key Results

A. Flow and Source Characteristics

• Wake Superposition: In the Column (C) configuration, the downstream turbine (C2) sits in the wake of the upstream turbine (C1). This causes a significant velocity deficit and increased turbulence at C2's rotor, reducing its rotational speed and overall sound power level (OASWL) by >7 dBA compared to C1.
• Side-by-Side/Staggered: In Line (L) and Staggered (S) configurations, wake interactions are minimal. However, small spanwise wind speed variations cause slight differences in rotor speeds (e.g., 0.9% difference in Case S), leading to distinct temporal behaviors.

B. Sound Pressure Level (SPL) Propagation

• Column (C) - Strong Focusing: When turbines are aligned, the second wake acts as a converging lens. This enhances wake focusing, increasing SPL and AM downwind by several decibels (up to 4–5 dBA in specific zones) compared to a single turbine. The focusing zone moves closer to the turbines.
• Side-by-Side (L) & Staggered (S): The increase in average SPL is limited (< 2 dBA) in oblique directions.
• Spatial Averaging: In L and S configurations, the superposition of signals from two turbines tends to smooth out fluctuations, generally reducing the overall AM compared to an isolated turbine.

C. Amplitude Modulation (AM) and Temporal Effects

• Angular Offset (Identical Speeds): In the Line (L) case (identical speeds), AM is strongly dependent on the initial angular offset ($\Delta\beta$) between rotors.
  • If rotors are in phase ($\Delta\beta = 0^\circ$), AM is constructive.
  • If out of phase ($\Delta\beta = 60^\circ$), AM can drop to near zero due to incoherent summation, even though the average SPL remains unchanged.
• Beating Effects (Different Speeds): In the Staggered (S) case, slight differences in rotor speeds create a beating phenomenon.
  • The signal exhibits an envelope frequency ($f_e$) corresponding to the difference in blade-passing frequencies.
  • This results in intermittent AM, where the "swishing" sound fluctuates between being perceptible and imperceptible over long periods (e.g., a 400s period observed in simulations).
• Wake Influence on AM: In the Column (C) case, the strong focusing of the second wake significantly amplifies AM downwind (reaching ~10 dBA), whereas in L and S cases, AM is reduced due to spatial averaging.

5. Significance and Implications

• Environmental Impact: The findings suggest that turbine siting strategies must account for wake-induced focusing, which can significantly increase noise levels and AM in specific downwind zones, potentially violating regulations even if individual turbine noise is low.
• Perception and Regulation: The study underscores that Amplitude Modulation is not a static value; it is highly dynamic and sensitive to rotor synchronization and speed differences. The "beating" effect implies that noise annoyance may be intermittent, complicating current regulatory frameworks that often rely on average metrics.
• Optimization: Accurate prediction of these interaction effects is crucial for developing better wind farm layouts that maximize energy production while minimizing noise impact, moving beyond simplified point-source models.

Conclusion

The paper concludes that while side-by-side and staggered arrangements generally mitigate noise through spatial averaging, aligned (column) arrangements create a "worst-case" scenario where wake focusing drastically increases both noise levels and amplitude modulation. Furthermore, the temporal variability of noise (AM) is governed by complex rotor dynamics, suggesting that future noise assessment models must incorporate time-dependent rotor interactions and wake effects to accurately predict human perception and environmental impact.