This is a recording of 20 seconds of wind turbine noise (provided by Rick James) in the vestibule of a house located 1500' upwind of a wind turbine. The sound was recorded with a microphone capable of detecting sound frequencies as low as 1 Hz. Note that most recording systems (camcorders and camera systems used by TV stations) are NOT capable of detecting such low frequencies and do not represent the sound accurately. When this waveform is played by the computer you can hear the typical periodic swish of the blades rotating.
Below is the same sound recording after the infrasound has been removed (by high-pass filtering at 20 Hz). So this wave shows the audible component of the noise that is present in the waveform above. The two waveforms look very different because the upper one contains a large infrasound component that you cannot hear. Even though the above and below waveforms look very different, they sound exactly the same when played by a computer sound system. This is because you only hear the audible portion of the waveform and not the infrasound part. The waveform below represents the sound that most recording systems (such as camcorders) detect.
Below is the same sound recording in which now the audible sound has been removed (by low-pass filtering at 20 Hz). This shows the infrasound that dominated the initial recording. Even though the amplitude of this waveform is large, you hear absolutely nothing when it is played back by the computer. This is because the computer speaker system is not capable of generating such low frequencies, and even if it did, you probably would not be able to hear them at the level generated. This does NOT mean that your ears would not detect this sound. The ears are quite capable of responding to such sounds but infrasounds appear to be canceled out of conscious hearing. This is probably so that you are not distracted from environmental sounds by internal low frequency sounds such as from your respiration and heartbeat.
Also note that the biggest infrasound peaks (at 2 – 4 sec and 15 – 17 sec in waveforms 1 and 3) do not correspond to when the turbine sounds loudest (at 5 – 11 sec in waveform 2). It is therefore impossible to judge the infrasound level from the sound that is heard or from A-weighted measurements. Quantifying the sound with A-weighted measurements is analogous to having mustard and ketchup on a hamburger and trying to estimate the amount of mustard present by measuring the amount of ketchup. In order to know how much infrasound is present you have to measure the infrasound itself and not how loud the acoustic component of the sound is with A-weighted levels.
Below are shown wind turbine noise presented as a spectrum (Van den Berg, 2004 (left) and Jung and Cheung 2008 (right)). Both of these spectra show that most of the sound energy generated by wind turbines is in the infrasound range.
The level of infrasound a turbine makes depends on many factors. Unfortunately the factors that affect infrasound production are not widely presented in published literature but the wind turbine operators are certainly aware of them. Generated infrasound depends on the size of the turbine, so small, personal turbines and the turbines in use up to about a decade ago do not present a problem. In general, the biggest problems come from the bigger turbines, but even this will depend on the manufacturer and blade configuration. Infrasound generated depends on the loading of the turbine (i.e. how much power it is generating) and will be lower when the turbine is “freewheeling”, i.e. when the blades are turning under low loads. Infrasound generation apparently depends on “inflow turbulence”, i.e. how much the wind speed varies across the diameter of the rotor. For flat, open land, where the wind is not turbulent, windspeed over the rotor may be fairly uniform and infrasound production may be lower. On the other hand, when turbines are built on ridgelines or in woodland, where airflow is turbulent and is not uniform over the rotor, then infrasound generation may be greater. Similarly, when wind turbines are grouped and the wind direction is such that the turbulent air downwind of one turbine hits a second turbine then infrasound generation by the second turbine will be greater. These location differences may account for why some communities have more problems with wind turbines than others. It also account for why some measurement studies find modest infrasound levels (around 70 dB SPL) while others, such as those above, find much higher levels. It is likely that infrasound generation is highly variable. But that isn't much consolation if you live near a turbine that generates high infrasound levels while the engineers are measuring similar turbines that have smoother airflow and which generate lower infrasound levels.
The wind industry makes highly misleading claims about the infrasound content of wind turbine noise. “Renewable UK”, the website of the British Wind Energy Association (www.bwea.com, 10/28/2010) use this quotation from one of their consultants.
“I can state quite categorically that there is no significant infrasound from current designs of wind turbines”
The critical word in this statement is the word “significant”. Although most naïve readers reading the statement would conclude that wind turbines do not generate infrasound, nothing could be further from the truth. As shown above, these devices do generate high levels of infrasound. The term “significant” is used to cover the fact that although turbines DO generate infrasound, it is assumed by the writer to have no consequence as humans cannot hear the infrasound at the levels produced. This is based on the invalid assumption that you must “hear” infrasound in order for it to affect you. This assumption is totally incorrect as discussed on other pages. Hearing is not the only way that infrasound can affect you. The infrasound is transduced by the ear at levels well below those that are heard and can affect the body in a number of ways without being heard.
Similarly, the reported quotation from Tom Gray of AWEA “Wind turbine noise is as loud as your refrigerator heard from the living room.” shows either a complete lack of understanding of the nature of wind turbine noise or an intentional misrepresentation of the true situation. Wind turbine noise is not comparable to the noise from a refrigerator. Refrigerators do not generate infrasound levels of over 90 dB at 1-2 Hz. I predict that IF THEY DID, NO ONE WOULD TOLERATE THEM IN THEIR HOUSES!!
Return to Wind Turbine page
Page created 12/28/2010 by Alec N Salt, PhD, Department of Otolaryngology, Washington University Medial School, St Louis, MO, 63129, USA