Smoking Gun: Research Proves ‘Inaudible’ Pulsing Wind Turbine Noise Direct Cause of Adverse Symptoms

The evidence proving the unnecessary damage done to wind farm neighbours by the noise generated by giant industrial wind turbines is mounting by the day: Germany’s Max Planck Institute has identified sub-audible infrasound as the cause of stress, sleep disruption and more (see our post here); and a Swedish group have shown that it’s the pulsing nature of low-frequency wind turbine noise  (‘amplitude modulation’) that is responsible for sleep problems in those forced to live with it (see our post here).

In a World first, Australia’s Administrative Appeals Tribunal (AAT) held that “noise annoyance” caused by wind turbine generated low-frequency noise and infrasound “is a plausible pathway to disease”. The AAT also slammed wind turbine noise standards as irrelevant and, therefore, totally unfit for purpose: Australian Court Finds Wind Turbine Noise Exposure a ‘Pathway to Disease’: Waubra Foundation Vindicated

The wind industry and its pet acoustic consultants maintain the fiction that wind turbine noise is the equivalent of a refrigerator at 500m, and steadfastly refuse to admit that the low-frequency noise and infrasound generated has any impact on humans, at all.

A few ‘research’ teams stacked with wind industry ‘friendlies’ have conducted bogus experiments, claiming to prove that the infrasound generated by wind turbines has no effect at all on humans.

Part of the fiction depends upon trying to ‘recreate’ the noise generated by giant industrial wind turbines in a laboratory setting. The friendlies – clearly eager to exonerate their paymasters – replay what purports to be wind turbine noise through headphones. Predictably, the subjects rarely complain about their few minutes of exposure to noise which bears no comparison to the real thing. On the friendlies very specific remit, they get to claim ‘mission accomplished’ and very fat pay-cheques.

However, as with most things in life, the answers you get depend on the questions you ask.

STT Champion, Steven Cooper has been asking the right questions about wind turbine noise for nearly 8 years. Now, thanks to his tenacity and independent, enquiring mind, a small experiment has yielded very big results.

The experiment written up in the paper he co-authoured with Chris Chan, and which Cooper delivered to the Acoustical Society of America’s conference in New Orleans last month, provides scientific evidence that sound between 30Hz and 1200 Hz (i.e. not infrasound) which is inaudible, but pulsating, i.e. at or below the threshold of hearing, can directly and reliably induce symptoms in noise sensitized individuals.

The Waubra Foundation, the charity behind the AAT’s ground-breaking decision, supported and encouraged Cooper’s research, and encouraged noise sensitized people to participate in it.

Here are the key points from Cooper and Chan’s research.

Subjective perception of wind turbine noise – The stereo approach
Proceedings of the 174th Meeting of the Acoustical Society of America
Steven Cooper and Chris Chan
7 December 2017

Abstract

The conduct of stereo measurements for both playback in high-quality headphones and in a hemi-anechoic room has been undertaken for a number of wind farms and other low-frequency noise sources as an expansion of the material previously presented at the Boston ASA meeting. The results of the additional monitoring, evaluation, and subjective analysis of this procedure are discussed and identifies the benefits of monitoring noise complaints and assessments of wind farm noise in stereo.

The laboratory mono subjective system was used to reproduce the audio wave file obtained in a dwelling. The test signal, being inaudible, was presented as a pilot double blind provocation case control study to 9 test subjects who have been identified as being sensitized to wind turbine noise and low frequency pulsating industrial noise. All test subjects could detect the operation of the inaudible test signal. The use of a stereo manikin to investigate detected inaudible “hotspots” is discussed.

…

3.0 SENSATION INVESTIGATION

In 2013 Schomer proposed the possibility that a limited number of residents subject to noise from wind turbines may be experiencing motion sickness and suggested the construction of a test facility that utilise special transducers to extend down to very low frequencies (0.05 Hz or lower). Schomer proposed to undertake sensing tests that could then lead to further medical examinations on animals to develop an understanding why the phenomenon seems to affect some residents near wind farms and establish who are affected by wind turbine infrasonic emissions in various ways.

We have previously utilised one of our reverberation test chambers (having a volume of 126 m³) with twelve 15” sub-woofers mounted on in the aperture between the reverberation chambers to investigate threshold of sensation versus threshold of hearing in the infrasound region, investigations into the ‘infrasound signature” from wind turbines. Those investigations were undertaken using pure tones or external (free-field) noise measurements of wind turbine noise.

The chamber has been used to investigate the generation of recorded wind turbine noise versus field measurements to identify the issue of pulsations across the entire spectrum and that the synthesis method that has been proposed for creating the source signal over a wide band of frequencies and a concept of synthesising a digital signal from analysed Leq FFT results but limited to just the infrasound region. Those investigations found the synthesised results did not agree with our analysis of the original external source data that has been obtained in the field. Utilising a synthesised signal from an averaged (Leq) FFT to produce a steady signal lacks the on/off transitions, transients and variations that existed in the original time record.

For the subject study the original wave files obtained at house 87 from the Cape Bridgewater study was used with a focus on the region of 30 Hz – 1250Hz. The source wave file signal obtained from measurements inside dwelling 87 at Cape Bridgewater, that have been used by several authors as a reference FFT Leq spectrum, was reproduced in the chamber utilising the sound system described above and provided the 1/3 octave band spectra shown in Figure 3. For the frequency range of interest the reproduced signal approximated the original signal as a 10 minute Leq level.

As a pilot study, 9 persons identified as sensitive to wind turbine noise or pulsating low-frequency industrial noise (test group 1) have attended our test chambers to participate in an experiment along the lines of the sensing tests in the format described by Schomer. A control group of 9 persons not previously exposed to turbine noise or pulsating low-frequency industrial noise (including 4 acousticians) participated in the same tests.

The reverberation room, with the addition of acoustic absorption treatment, satisfies the requirements of European Broadcasting Union Technical Document 3276 Listening Conditions for the Assessment of Sound Programme Material: Monophonic and Two-Channel Sound. The maximum noise level under that standard for a mono signal is set at 85 dB(A). The distribution of absorption around the perimeter of the reverberation room leads to the absence of lateral reflections from wall surfaces. As the walls of the chamber are core filled blockwork, from sound intensity and vibration measurements it was established that neither the walls, floor or ceiling of the chamber are generating structure borne noise from the speakers mounted on the baffle in the aperture.

The levels that were generated in the room approximate the 1/3 octave band levels obtained in house 87 (in the Cape Bridgewater study) over the range of 40 – 1250 Hz. The response that falls off below 16 Hz reflects the absence of any graphics or parametric equalisation, and the limitations of the A-D convertor.

Table 2 presents the measured sound levels of the generated and ambient levels in the test chamber, with the derived sound level contributions in both the Leq level and the L90 level.

By any of the general measurement parameters used for wind farm assessments, the test signal contribution is at or below the ambient level. Of relevance to researchers of wind turbine noise, the testing had the wind turbine noise contribution as an Leq level of 12 dB(A) in a background level of 23 dB(A).

For the levels that were generated. the testing was undertaken in accordance with Australian Standard AS 1269.4 Occupational Noise Management, Part 4: Auditory Assessment and the testing conducted in accordance with the ASA Ethical Principles of the Acoustical Society of America for Research Involving Human and Non-Human Animals in Research and Publishing Presentations. An observer was present in the reverberation room during the testing.

The testing was conducted as multiple blind study tests. At no point in time were any of the participants advised what signal (if any) was being applied.

After a period of between 45 seconds to 3 minutes, all the 9 people in test group 1 could sense the presence of the wind turbine signal on 100% of the occasions in which the signal was presented, even though they were unable to hear the signal. At no point in time did any of these test subjects detect any audible signal.

One test subject (from the test group 1) identified a disorientation in the room where there was a perception of a tilt in the floor of about 20°.

The control group were exposed to the same test set up. After a period of some two minutes 2 people (including one a very distinguished Australian acoustician) could identify sensation, whilst the remainder of the control group never detected any sensation.

3.1 Observed Differences in the Sound Field – Hotspots

All the test group 1 subjects were requested to move around the room and identify any hotspots where there was a perception of a greater impact.

Two general areas were identified on either side of the radiating pattern for the baffle speaker systems (see Figure 4).

The test subjects identified the sensation that they were experiencing occurred in different parts of the body.

Seven people from test group 1 noise identified sensation in the back of the neck or the back of the head, and in four subjects there was also a tingling in the legs.

All the people from test group 1 were requested to rotate 360° to identify whether there was any position at which the sensation became stronger.

In all cases except for two women (one person who has a hearing impairment), the test subjects identified that the greatest sensation occurred for an orientation where the back of the head was towards the speaker baffles but the body was turned at an angle of 45° so that the ear adjacent the baffle propagating sound field was closer to the speakers (see Figure 4).

The test subjects were then presented with audio headphones that provide an SLC 80 attenuation of 11 dB, and then a set of hearing protectors providing an SLC 80 of 26dB.

All test subjects (except the two women noted above) identified there was a difference in the perception of sensation in their head, but had difficulty expressing what that difference was. Both woman identified the test signal produced a sensation across the forehead. The headphones provided a slight difference but when using the ear muffs both participants felt the onset of nausea and the experiment was terminated. Does this result support the observation by Salt of a greater Guinea pig ear response when there was less high frequency masking?

…

Conslusions

Testing of the response of individuals to audible wind turbine noise in recent years has typically utilised a mono noise source with a large bank of speakers in a modified reverberation room or listening environment.

Other testing purporting to assess the impact of infrasound from turbines, has not actually used the infrasound signal but has used either pure tones or a synthesised signal based a result of an FFT Leq analysis of the original signal and incorrectly claimed such noise sources as being “wind farm infrasound”.

Analysis of wind farm noise using wave files of actual wind farm noise (rather than any synthesised format or digitally designed signal) has found the typical FFT acoustical analysis is incorrect in terms of the fundamental formula of BT=1 for frequency analysis. That is, a finer resolution or small B requires a large T, and therefore a low temporal resolution to make the result valid. In the infrasound region the pulses are not present long enough to satisfy BT=1.

A modification of the Infrasound Logger from Huson Associates (Mark II) incorporates a modified filter and increased sample rate to address signal droop and obtain a faithful wave file to 150Hz.

Analysis of the wave files recorded at Cape Bridgewater reveals the presence of a dynamically pulsed amplitude modulation of the signal that occurs across the entire audible frequency band. The dominant bands where such noise is audible are in the low and mid frequency region.

All our field work to date that provides FFT or 1/3 octave band measurement data in relation to wind turbine infrasound, identified levels well below the nominal threshold of hearing. The limitation of instrumentation and sampling rates to provide an accurate and valid spectrum measurement in the infrasound region has been questioned (BT=1).

The previous work by the authors that identified the analysis/signature of pulses that occur at an infrasound rate, leads investigators to view the signal in the time domain and examine/describe/review the method of modulation with dynamically pulse amplitude modulation suggested as a more accurate description.

In endeavouring to reproduce an accurate signal in the time domain we have raised the issue of much higher sampling rates than normally encountered.

There are also issues with the creation of wind turbine “infrasound” in the laboratory. The authors are of the opinion that experimental research limited to just wind turbine infrasound”, whether tones or synthesised digital signals, is a waste of research time and money.

Reproducing and analysing the wind turbine signal including the audible range is an easier and simpler task to undertake and permits the essential work of identifying what creates sleep disturbance and physical impacts from wind turbine noise. Such research should be undertaken inside dwellings (in the field) and (subject to qualification of the sound field) may be undertaken in the laboratory.

Utilising wave files and playback of such signals at inaudible levels without requiring reproduction of infrasound is an easier and simpler task to undertake. The benefits of using a stereo signal for subjective assessment is clearly a superior method and a logical approach for any serious investigation into wind turbine noise.

 

Our previous paper into the stereo effect found microphones spaced 1.9 metres apart for recording the signal and playback in a hemi anechoic space using line array speakers to be the preferred method by all test subjects for the subjective assessment of external wind turbine noise.

For utilising headphones, the recent testing has confirmed that the use of a stereo head torso (or in this exercise a cheaper version identified as a manikin) is the appropriate mechanism for undertaking further investigation into the subjective effects of wind turbines.

The application of the manikin to support the investigation of the subjective response of wind turbine affected persons in a mono generated sound field, utilising inaudible wind turbine noise, identified slight differences between the “ears” at the position identified by the test group as the hotspots (i.e. a greater perception of sensation with their backs to the sound source and one ear on an angle of 45° to the sound source).

The sensation perceived by the specific sensitive people (rather than the control group) was significantly stronger in the sound field exposed to the entire body when compared to just utilising headphones.

The results of the sensitivity testing require the expertise of other disciplines to explain the mechanisms by which the test subjects perceive the wind turbine noise to their entire body and should be of interest to other researchers.
Acoustical Society of America

Download the full paper here