Modelling sound from wind turbines and drones to combat noise pollution

Recent years have seen an explo­sion in the num­ber of drones used for leis­ure or pro­fes­sion­al pur­poses, and the pro­lif­er­a­tion of wind farms on land and at sea. As such, there are more and more sources of aero­dy­nam­ic noise. And these are likely to dis­turb the pub­lic along with anim­al biod­iversity. Reg­u­la­tions are becom­ing stricter and, gen­er­ally at a European level, noise lim­its are being imposed that must not be exceeded. But noise levels are not easy to define, and the per­cep­tion of cer­tain sounds is some­times irrational.

In 2017, research­ers asked around 40 people to listen to 103 sounds made by drones fly­ing at dif­fer­ent alti­tudes and the noises made by dif­fer­ent four-wheeled vehicles. The listen­ers were then asked to rate them from “not at all annoy­ing” to “very annoy­ing”. The res­ult was that, even at the same volume, the sound of drones is much more annoy­ing to the human ear than that of cars or lor­ries¹. The authors of the exper­i­ment sug­ges­ted that this was prob­ably because we are used to the noise of these vehicles. The same applies to the noise gen­er­ated by wind turbines.

Com­pared with many oth­er envir­on­ment­al noise sources, wheth­er man-made or nat­ur­al, the noise levels gen­er­ated by a wind farm are mod­er­ate. Accord­ing to a study by Céréma², the sound rarely exceeds 40 dBA out­side a resident’s home. In com­par­is­on, traffic passing in front of a resident’s home can emit more than 70 dBA, which is a noise level that by law must not be exceeded. Wind farms loc­ated in pre­dom­in­antly rur­al envir­on­ments gen­er­ate a rel­at­ively low level of loc­al back­ground noise, but this can con­trib­ute to the per­cep­tion of noise. In addi­tion, wind noise con­tains low fre­quen­cies (20–200 Hz, aud­ible) or infra­sound fre­quen­cies (below 20 Hz, gen­er­ally con­sidered inaud­ible) which can travel fur­ther than high­er fre­quency sounds.

“While the annoy­ance caused by noise increases with the level of noise expos­ure, its devel­op­ment does not gen­er­ally fol­low a simple law and depends on each noise source and its char­ac­ter­ist­ics (con­tinu­ous noise vs. impulse noise, low-pitched vs. high-pitched noise, etc.),” explains Olivi­er Doaré, pro­fess­or at the Mech­an­ics Unit (UME) at ENSTA Par­is. His labor­at­ory is con­duct­ing stud­ies into the mod­el­ling of dif­fer­ent noise sources, includ­ing one on wind tur­bines. “For these devices, the noise level in decibels is not sig­ni­fic­ant, because the noise is not uni­form. It is char­ac­ter­ised by tem­por­al fluc­tu­ations in amp­litude, poten­tially linked to the weath­er, the shape of the blades, etc.”

The labor­at­ory has been work­ing on mod­el­ling the sources of aero­dy­nam­ic noise from wind tur­bines and the propaga­tion of sound in the atmo­sphere. The aim is to be able to sim­u­late the res­ult­ing ambi­ent noise before a wind farm is installed. “The idea is to be able to hear the farm before it is built, as if we were cre­at­ing a sound mod­el rather than a visu­al one”, explains Ben­jamin Cot­té, a lec­turer at ENSTA (IP Par­is) and co-author of the study, along with Dav­id Mas­car­en­has, now an engin­eer at Cap­gem­ini. The research­ers hypo­thes­ised that vari­ous factors were being over­looked in the meas­ure­ment of wind tur­bine noise, such as the height and qual­ity of the ground (wheth­er it is more or less absorb­ent) on which they are installed, or the effect of wind dir­ec­tion and pro­files on the propaga­tion of sound.

By devel­op­ing a tool for syn­thes­ising wind noise, using phys­ic­al mod­el­ling rather than sampling, they can ana­lyse the sens­it­iv­ity of sounds. This can be done accord­ing to a num­ber of factors: blade geo­metry, wind or tem­per­at­ure pro­file, upstream tur­bu­lence rate. This also makes it pos­sible to pre­dict the power of the noise for each seg­ment of the wind tur­bine blade. Sound syn­theses are thus pro­duced and presen­ted for dif­fer­ent met­eor­o­lo­gic­al conditions.

These pro­cesses, which aim to recre­ate a sound envir­on­ment (known as “aur­al­isa­tion” of wind tur­bine noise), can be used in vir­tu­al real­ity applic­a­tions. The study is part of the European Vir­tu­al Real­ity Audio for Cyber Envir­on­ments (VRACE) network.

In prac­tice, the chal­lenge facing the design­ers and oper­at­ors of wind farms is to reduce the noise gen­er­ated by the machines, to avoid com­plaints from loc­al res­id­ents. Dif­fer­ent approaches are being stud­ied or have already been imple­men­ted. “It is pos­sible to influ­ence the speed of rota­tion or the angle of pitch of the blades,” explains Ben­jamin Cot­té. “Some teams are work­ing on intel­li­gent tools that auto­mat­ic­ally reg­u­late wind tur­bines to min­im­ise noise while optim­ising pro­duc­tion. One of the latest advances is the use of “combs” or “ser­ra­tions”. These are elong­ated “saw-toothed” parts, dir­ectly inspired by the wings of noc­turn­al birds of prey, which can fly in per­fect silence thanks to the spread of their wing­tip feath­ers, which allow air to pass through, thus redu­cing tur­bu­lence.” These combs are attached to the trail­ing edge of the blades. They reduce aero­dy­nam­ic noise by an aver­age of 2 to 3 decibels, by redu­cing the tur­bu­lence cre­ated by air fric­tion at the tip of the blade. 

As far as drones are con­cerned, more research is under­way. The “Aero-acous­tics of multi-propul­sion sys­tems for UAVs” (APRO) pro­ject, in part­ner­ship with the Com­puter Sci­ence and Sys­tems Engin­eer­ing Unit (U2IS), aims to improve under­stand­ing and mod­el­ling of noise gen­er­a­tion in multi-propul­sion sys­tems. The aim is to lim­it noise pol­lu­tion. Doc­tor­al stu­dent Car­oline Pas­cal is look­ing to auto­mate the ana­lys­is of acous­tic fields using robot­ic meas­ure­ments. Today, her work enables the pre­cise char­ac­ter­isa­tion of pro­peller radi­ation, and will make it pos­sible to optim­ise the tra­ject­ory, speed, and accel­er­a­tion of a drone, in order to lim­it noise pollution.

Marina Julienne

Ref­er­ences :

Mas­car­en­has, D., Cot­té, B., & Doaré, O. (2023). Propaga­tion effects in the syn­thes­is of wind tur­bine aero­dy­nam­ic noise. Acta Acustica, 7, 23.

Mas­car­en­has, D., Cot­té, B., & Doaré, O. (2022). Syn­thes­is of wind tur­bine trail­ing edge noise in free field. JASA Express Let­ters, 2(3).