Modeling wind farm noise emission and propagation: Effects of flow and layout

Colas, Jules; Emmanuelli, Ariane; Dragna, Didier; Stevens, Richard J. A. M.

doi:10.1016/j.renene.2026.126052

Publication 130 · Wind-energy acoustics

How wind-farm layout changes modeled noise

J. Colas, A. Emmanuelli, D. Dragna, and R.J.A.M. Stevens, Renewable Energy 273, 126052 (2026).

The article and Figure 14 are open access under CC BY 4.0.

Main finding

In one modeled stable-atmosphere case, a staggered 4 × 4 wind farm produced spatially averaged downwind sound levels about 3 dBA higher than the aligned layout beyond 2 km. The coupled calculation attributes the difference to both layout-dependent turbine source emission and sound propagation through the wind-farm flow.

Spatially averaged modeled overall sound pressure level and amplitude modulation versus downstream distance for aligned and staggered four-by-four wind farms, with the staggered case about 3 dBA higher beyond 2 km — **How to read the figure.** Panel (a) shows the modeled overall sound pressure level, averaged across receivers within |y| < 720 m, versus downstream distance. The gray band and dotted lines mark the wind-farm rows. Beyond 2 km, the staggered layout (blue) remains about 3 dBA above the aligned layout (black). Panel (b) shows the corresponding amplitude modulation; near 2.5 km its sampled maximum is about 2 dBA for the staggered layout and 1 dBA for the aligned layout. The curves compare one modeled stable case and are not a general layout law. Open the full-resolution figure. Figure 14, cropped. J. Colas, A. Emmanuelli, D. Dragna, and R.J.A.M. Stevens (2026), CC BY 4.0.

Why this matters

Wind-farm noise is not obtained by multiplying an isolated-turbine prediction. Layout changes the wake velocity and turbulence encountered by each rotor, which changes the modeled source emission. The combined wind-farm wake also changes atmospheric refraction and focusing between the turbines and receivers. Both parts of the source-to-receiver chain therefore have to be represented when layouts are compared.

Research context

The comparison couples actuator-disk large-eddy simulation, turbulent-inflow and trailing-edge source models, and wide-angle N × 2D parabolic-equation propagation. It covers aligned and staggered 4 × 4 farms of NREL 5 MW turbines at 4D streamwise and spanwise spacing, one stable atmospheric boundary layer, one ground model, and frequencies from 50 to 1080 Hz. A model-bookkeeping comparison assigns about 1.5 dBA of the downwind layout difference to turbine source powers and the remainder to propagation, but this is not an uncertainty-bounded causal decomposition. The calculation has no fully three-dimensional acoustic benchmark, displayed source-height or angular convergence, flow ensemble, field validation, or human-response evidence. It therefore does not establish a universal staggered-layout penalty, a worst atmospheric condition, or a permitting or annoyance outcome.

Read the article Open-access version View in the complete publication list Related research highlights