Health Effects of Industrial Wind: The Debate Intensifies (update with Steven Cooper)
By Sherri Lange
July 30, 2020
“Exposing the fact that the cost-benefit analyses for wind farms are wrong, the power output modelling is wrong, the acoustic modelling is wrong, and the acoustic dose response data is wrong could create some interesting discussion on the back of Planet of the Humans.” (Cooper, June 20, 2020)
Master Resource has followed the work of acoustician Steven Cooper for some time. In a February 2018 interview, Sensing but Not Hearing, Mr. Cooper explained how all-body hearing mattered more than acoustic isolation and reporting.
“On discussing the residents’ observations (with the residents) for the first two weeks,” he stated, “I found the use of describing the impacts in terms of Noise, Vibration, and Sensation was accepted by the residents as a better concept.”
In a November 2019 update, (Sensing but Not Hearing, Latest), Mr. Cooper asked: “I question how an authority can propose noise criteria with no fundamental basis identified as to how a stated core objective can be satisfied.”
Today, we speak again with Steven Cooper about new research explaining why acoustic measures by wind developers are not supported with any meaningful measures or predictions from models. This interview builds upon and updates the previous two in regard to his published and presented ground-breaking work.
Question: Can you update us on your conferences, research, and presentations?
Cooper: My original concept was to determine and reproduce an accurate sound signal of wind turbine noise to then permit multi-disciplinary investigation of what the noise was doing to people and what parts of the body were reacting to the noise.
That concept has ventured into multiple investigations over the years that have been progressively presented to the wind turbine working group of the Acoustical Society of America, and then by invitation has extended into Europe.
My work has questioned/examined what our general acoustic analysis provides and the errors in such analysis; threshold of hearing and threshold of sensation; the challenges in creating that sound; subjective assessment of mono vs. stereo; and the infrasound sound of wind turbines versus the pulsation of the wind turbine sound that is occurring at an infrasound rate.
I have also done a case study that showed sensitive people can identify the presence of an inaudible wind turbine noise, putting the Kelley Mod-1 work (in the 1908s) with the work on fluctuation from Zwicker and Fastl, clarifying the incorrect use of Amplitude Modulation in relation to wind turbine noise, and with the assistance of a psychologist in a single case study, showing the response of the inaudible pulsating wind turbine noise was centered in the frontal lobes of the brain.
As a result of the International Congress of Acoustics (ICA) and the European Acoustics Association (EAA) TC Noise meetings in Germany last year, there was a request for a special Edition in Acoustics on Wind Turbine Noise. Recently a Technical Note (by Cooper and Chan) on acoustic compliance of wind farms was peer reviewed and published in Acoustics.
Having entered the fascinating world of psychoacoustics I have been able to re-analyse the Cape Bridgwater data in terms of the power output hypothesis and the modulation index and have found an amazing correlation with the complaints/observations. I propose to submit two more Technical Notes to be published in the Special Edition in Acoustics if I can make the deadline.
Question: What was the response of your colleagues to your materials presented?
Cooper: So far, I have received a great deal of encouragement to continue the work. In Germany last year at the ICA conference I had a number of senior acousticians (who collectively have published a significant number of papers on wind turbine noise) seek private discussions so they could discuss in detail the three papers that I presented (a lot of material packed into a short presentation).
In other presentations there were acknowledgments that the presenter had used the wrong terminology in the relation to wind turbine amplitude modulation (when it should be modulation of the amplitude). Those public acknowledgments to other acousticians indicate a shift in the acoustic terminology for wind turbine pulsations.
The recognition of my work in Europe, appointment to the EAA technical committee on Noise and a request (to my wife) by a number of esteemed professors to continue my research and not retire was astounding. In relation to the recent Technical Note published in Acoustics I received an email from Dr Paul Schomer congratulating me on the paper and my identification of issues with the noise dose response/setting wind turbine limits.
To receive that email from Paul I consider to be the highest honour I could obtain in acoustics.
Question: You have indicated in your notes to us that the WHO (World Health Organization) has incorrectly set 2018 criteria for wind turbines, despite being critical of the limited number of noise studies. Can you explain this?
Cooper: In many socio-acoustic studies the input of acousticians is a small component and does not involve validation of predicted levels versus real levels. If the researchers didn’t understand that the wind turbines do not operate at full rated power all the time, and they weren’t acousticians who would be aware of that very fact (whereas the community is aware the wind is not constant and neither is the noise), then it follows (as night follows day) that what turbine noise the residents had been exposed to for some time would be lower than the predicted levels.
Therefore, the noise scale on the dose response curve should be shifted to the right. This shift leads to a lower noise level for the general limit derived for 10% of the community highly annoyed.
Going to the wind turbine noise curves in the WHO 2018 Guidelines they are not all using the same acoustic parameter. In the recent Technical Note, we corrected the curves on the assumption of the same noise all the time. When that data is reviewed then, using the normal 10% highly affected level, the WHO recommended limit is too high.
Health Canada relied upon predicted noise levels for their analysis – again no validation exercise to identify the actual noise versus the predicted levels. The study report does not identify any exercise to ascertain how the turbine noise/power of the turbines would have varied over the period prior to and during the survey.
The Health Canada study did not know the actual wind turbine noise during the survey or how it varied during the survey. But they provided a dose response curve. Now if the two wind farms were subject to only an average power output of say 30% then it follows that the average noise would be significantly lower than the predicted levels. Therefore, in reality the Health Canada recommended noise levels for wind turbine noise that should have been lower.
Question: So, you are saying that the noise criteria to protect the community from adverse impacts from WHO references, Guidelines, and Health Canada’s recommendations, are not correct?
Cooper: Yes. Neither the WHO nor Health Canada identified actual wind turbine noise levels to protect the community from adverse impacts – because they used predicted levels without any validation of what the actual noise levels were (when compared to the predicted levels).
On a purely statistical basis the number of complaints in relation to wind turbines says there is something wrong with the suggested criteria. Environmental Authorities and wind farm operators conduct compliance testing and say it meets the relevant guideline or permit condition.
But the residents are still affected, complain of sleep disturbance and adverse impacts. So therefore, for the “complying wind farm” there presents a conflict with what the guidelines say will not create adverse impacts or sleep disturbance and reality.
Part of the problem is that various people on the various independent expert committees, the bulk of them, are not acousticians, or people who would have any understanding of the how (and when) wind farms operate.
Not like the work of Dr Kelley on the MOD-1 turbine where the report clearly stated that some of the residents were disturbed (when the predictions said that would not occur). More importantly the report of Dr Kelley and his team identifies the complaints were confirmed and not imagined.
The important thing to comprehend is that for the UK and Australian criteria they are based upon the WHO Community Noise 1995 document that provided a limit to protect against sleep disturbance. But the sleep disturbance criterion was based on traffic noise not wind turbine noise. So, the concept of achieving wind industry noise limits that “will automatically ensure there are no sleep or health or noise impacts,” is clearly based on a false premise.
The WHO 2018 Guideline clearly stated there was insufficient data to provide any advice as to sleep disturbance noise limit. Why? Because there are no studies on actual wind turbine noise, and the six questions I posed in our first interview have still not been answered.
Question: On examination of the WHO Guidelines, 2018, as discussed above, it seems obvious that the “conditional” recommendation (page 97) that policy makers reduce noise from turbines to below 45dB L den, is useless and invalid. So we can safely say all development of wind without adequate noise protections is a house of cards. Facile, unregulated, and improperly guided by any meaningful principles around noise measurement, impacts of impulsive sound.
Cooper: In the 1990s, the UK developed criteria for wind turbine installations (ETSU-R-97). The document identifies the limits are to protect against sleep disturbance, but they used road traffic noise as the basis of the sleep criteria.
So, if the Authorities used traffic noise sleep disturbance criteria for wind turbine noise then from the start there is a major problem.
ETSU-R-97 identified the operation of wind farms with a maximum of 450kW. But the document did not identify what the actual noise was from the turbines.
That means that where there were no complaints the wind turbine noise level could not be measurable or significantly below the background level or the nominated sleep criteria. No material in ETSU-R-97 (UK Working Group publication regarding Assessment and Rating of Noise Related to Wind Farms) to explain the sleep criteria versus the actual wind turbine noise levels.
It would have been appropriate to call the criteria preliminary or precautionary or subject to further revisions.
But we have in New Zealand and Australia wind farm noise limits based on ETSU-R-97 and applied to turbines somewhat larger target as than 450kW (being the largest turbine identified in ETSU-R-97).
But the basis of the noise limits in New Zealand and Australia are not supported by any wind farm dose response curves for the size of modern-day wind turbines or any studies into sleep disturbance as a result of modern-day wind farms.
But to be fair to the WHO they provided their criteria with a qualification of limited dose response data, no reliable sleep data and indicating a precautionary approach.
If you follow the above material, then the criteria may not be useless in that it is lower than some of the criteria used around the world.
On the basis of the WHO suggested limits it is appropriate for Environmental and Planning Authorities to undertake due diligence in reviewing their criteria/policies. A guideline from the WHO is not a regulation or a policy directive that becomes conditions of consent on a permit.
Courts can be presented hundreds of residents who provide evidence of being impacted by wind turbine noise. But the judge can say that he/she is bound to consider the application with respect to the policy/guideline of the state authority responsible for noise – not the complaints or evidence of disturbance.
In response to your question you need to understand that the precautionary wind turbine criteria issued by the WHO in 2018 relies upon predicted wind turbine noise levels. There is no material in the four social surveys referred to in the WHO Guidelines (in relation to wind turbine studies) that validated the actual noise levels versus the predicted wind turbine levels.
Furthermore, the assumed wind turbine noise levels (used by WHO) were not validated to exclude the wind noise or the vegetation noise. These additional noise components as a result of the presence of wind need to be addressed in deriving a noise contribution.
Hence my recent Technical Note raises points for consideration if such due diligence reviews were to be undertaken.
In light of the number of complaints around the world in relation to wind turbine noise, one must not follow the path previously used of just undertaking literature reviews of what other authorities have done (e.g. ETSU-R-97) – without actually analysing that material and determining its relevance to wind turbine noise.
Reviewing the development of various acoustic Standards used around the world will find (when going back to the original sources) a number of assumptions or empirical concepts used to formulate such procedures/assessment methodologies.
I found such examples in room acoustics (for echoes in rooms) in my Master’s research and identifying an error in the FAA’s INM program regarding lateral attenuation for aircraft noise (acknowledged by the US Aircraft Standards committee in 2003) as part of a PhD research study. It is this sort of questioning (raised by my supervisor) that has caused me to question the basis of the wind turbine noise criteria and the number of complaints that question the criteria.
The above situation has occurred in Australia, Authorities saying the criteria is to protect against sleep disturbance as a result of wind turbines, when there is no such data to support that position.
Question: So what you are saying is that if the authorities were to undertake a due diligence review of their wind farm noise policies in light of the WHO identifying very limited data to provide anything but a precautionary approach, then there are important technical issues to resolve in the acoustic compliance process.
Cooper: Yes, there are a number of technical issues to be sorted out in deriving the actual noise emission for a wind farm to truly present the noise contribution for comparison with the appropriate criteria. It is those matters that I raised in the recent Technical Note.
I raised a number of matters in relation to separating the ambient noise from the overall noise that was measured. But more importantly I think Case Study D presents issues as to misrepresentation of the actual operating conditions and an assumed power output of a wind farm that becomes critical in any claims of acoustic compliance with permit conditions.
For example, Case study D in the Technical Note expands the issue by questioning the material to establish that the various wind farms used in the four dose response studies referenced in the WHO 2018 guideline. On going to the various papers for the four studies there is no identification of the actual power output versus the dose responses based upon predicted levels for maximum power.
Question: So if the wind farms were operating at a lower power output and therefore gave rise to lower noise levels, then the community reaction determined in those social surveys would be incorrect because the responses would be occurring for a lower noise level that suggested in the prediction method.
Cooper: Correct. You now have the concept that the basis of noise criteria for wind farms may not be correct by using predicted noise levels.
But case study D goes a step further because it highlights the fact that the wind across a wind farm is less than the predicted wind because of the interference in flow of wind by reason of the number of the turbines being present (and the layout of the wind farm). Upwind turbines result in disturbed air for turbines downwind. The presence of wake reduces the ability of the wind farm to obtain the predicted power outputs, and the effect of wake vortices extend outside the wind farm and can give rise to increases in noise and annoyance.
Question: Does this mean in terms of noise modelling one cannot consider each turbine as operating on its own?
Cooper: Yes. There is a fair bit of theoretical material on separation/spacing distances between turbines that show significant reductions in wind speed (and directions) across a wind farm when compared to the assumptions for the pre-wind farm situation.
Question: You then add in the use of meteorology masts outside the wind farm to provide a hub height wind speed that does not represent the average wind across the wind farm.
Cooper: Yes. But in the real world you can only show that material when you have access to the SCADA material for the subject wind farm that provides all the operating parameters for each turbine.
Question: In case study D you show for Cape Bridgewater how the angles of the turbines vary across the wind farm, and you reference other figures in the Cape Bridgewater study to show that point. But you then identify the use of wake-free wind data in compliance testing (using the regression line method) misrepresents the true acoustic compliance results.
Cooper: Yes. By using wind data for a wake free situation (by meteorology masts outside the wind farm) the wind will be higher than across the wind farm. This shifts the regression line data to the right and makes compliance easier. But the compliance report fails to identify the inaccuracy of the regression line re the wake free wind data versus the actual wind on the wind farm.
But more importantly the shifted data is not quantified in terms of the power output of the wind farm. The compliance test uses synchronised wind and noise data that is averaged over a period of time. There is no identification of the wind direction nor identification of the averaged power output.
If the wind farm during testing is only operating at say 25% of the rated power output then how can that compliance test be compared with the permit limits – particularly as there is no stated qualification (in the compliance report) as to the operation of the wind farm during the testing.
Question: The question of incompetence or gross negligence arises ...
Cooper: There is a severe level of incompetence on the part of Environmental Agencies responsible for protecting the community from harm. Their “acoustic experts” did not do the correct due diligence in relation to wind turbine noise.
So, you get the few really independent acousticians around the world (who have worked at the “coal face”) when confronting the response from the community in relation to wind farms say, hold on a second, there is something not right.
I don’t mean to isolate out some acousticians here, but to some extent it is noticeable and inevitable that I would be able to “Test Drive” some of the work and study being done internationally.
Some of the best, Rand, James, Ambrose, Swinbanks (US), Stigwood (UK), Metalka (Canada), Thorne (NZ), Cooper, Huson, Hansen and Hansen (AUS) have continued to work on the problem for years (except for Hansen and Hansen, all are unfunded), to slowly try and get to the bottom of the mess – but this group is getting on in years, and the younger acousticians (without the experience of solving problems) go with the money and do not question if the industry “facts” are correct.
Sometimes, quietly following a presentation, an interested professional or acoustician will approach me and ask for clarification on amplitude modulation or limitations on digital frequency analysis and so on. So that coaching continues for me as an integral part of my work.
Question: You mentioned that disturbed air flow or “dirty” air downstream is often ignored and contaminates the real discussion and real facts. Can you explain?
Cooper: Yes. This is a very important fact: Using wake-free wind data for compliance testing is simply incorrect.
The issue of dirty air downwind of turbines and disturbed inflow air has been around for a while, but the material in the Technical Note in Acoustics from an electrical viewpoint has sent me off in a new direction. Enhancement of the noise due to wake vortex kilometers downstream of the turbine just opened up a whole pile of boxes that fit in with one of the concepts I proposed from the Cape Bridgewater Study.
Putting the wake free noise curve data with the power losses then led me to assess the actual power output when wind farm operators undertook the compliance testing – it is so low that the compliance test needs to have a qualification (for example) “ these results relate to the average power output of the wind farm at 23% of the rated power.”
Such a statement automatically says, “hold on, how does that data demonstrate acoustic compliance – surely at high wind from power output levels, then the noise must be greater.” Are such “wind farm compliance tests” misleading the Environmental Authorities overseeing compliance with permits? Or if the Environmental Agency undertakes “compliance testing” are they misleading the community?
But another question then arises from the above: “If we use the actual average wind speed across the wind farm, the results shift to the left and reveal high noise levels for the actual wind the wind farm experiences”.
Question: Well that puts the cat amongst the pigeons. Let me see if I have this right:
- The testing results may include wind noise and wind vegetation noise.
- So, this means the actual wind turbine noise could be lower.
- The wind turbine dose response curves contained in the WHO 2018 guideline were predicted noise levels – not actual noise levels at the residential receivers.
- The wind turbine dose response curves in the WHO 2018 guidelines did not involve a validation exercise to ascertain what noise was being generated at the time of the survey or for the period before the survey.
- If the predicted levels used in the dose-response studies were for the maximum output of a wind farm then it is possible that the community responses have been placed in the wrong noise level category.
- Due to the presence of turbines on a wind farm the wind at individual turbines can be lower than outside the wind farms. This means that it may not be possible for a wind farm to obtain its rated capacity.
- The wake from upstream turbines can affect the operation of a turbine and cause disturbed air, reduce the output of the downstream turbines and because of the disturbed air gives rise to greater noise level and/or increased modulation.
- The use of graphical results to show the difference between the wind farm operating versus off does not identify the operating parameters of the wind farm.
Cooper: Yes, that is a pretty good summary. We already know from the wind turbine dose response curves in the WHO Guidelines that people are more sensitive to wind turbines than for road traffic noise. And if the assumed wind turbine dose response curves are based upon predicted levels that are too high then the assessment criteria are wrong.
This has relevance when you consider the vast amount of money on the Canadian Wind Farm Health Study (Health Canada Wind Turbine Noise and Health Study) used predicted levels and failed to validate the operations of the wind farms (and the resultant noise) with respect to the predicted levels.
Furthermore if we don’t know the actual wind turbine noise levels that people experience from wind farms that gives rise to sleep disturbance then how does one undertake sleep studies to determine the wind turbine noise contribution that will not give rise to sleep disturbance?
Lange: Thank you so much Steven, for the clarity and perspective. What continues to astonish many of us, is that massive outlays of turbines continue to be developed, with no standard acoustic or physical safe “setback,” and as you note, built on false premises and faulty predictions. This is a broken clock that does not even tell the correct time twice a day. It reminds us of the Waubra Foundation’s Definitive Document on your work. And, we must again wonder, does your work bring out calls for legal correction. Rhetorical.
Appendix: Waubra Foundation on Cooper’s Work
“Since measurements and predictive noise models for wind turbines being expressed in dBA exclude accurate measurement of infrasound and low frequency noise, it follows that dBA is useless as a proxy for predicting damage to neighbours, or for setting Standards to protect them from harm. Even before Steven Cooper’s investigation, the wind turbine noise Standards were known to be dangerously inadequate.
“Responsible authorities should have altered the Standards to include sound as a whole and infrasound in particular, especially after Dr Neil Kelley’s work establishing direct causation from infrasound and low frequency noise resurfaced in mid 2013. Steven Cooper’s work at Cape Bridgewater (our note: and his subsequent diligence and reporting on irregularities, bad data, non sensical standards of all kinds,) reinforces the need for urgent revision of existing Australian (and OTHER) standards and regulations, and to develop a standard for ‘sensation’.
“These current Standards are now known to be dangerous, clearly do not protect people, and must not ever be used again.”
Dr Laurie continues on the value of the Cooper studies:
If this Cape Bridgewater research, commissioned by a wind developer, conducted by an ethical independent acoustician with the cooperation of both the wind developer and the affected residents, is not acted upon immediately to prevent further harm, the public authorities and politicians who choose not to act are then in a position of knowingly allowing the serious damage to physical and mental health from impulsive infrasound and low frequency noise from wind turbines to continue.
Given that the most serious and common complaint around the world from neighbours to industrial wind turbines and other sources of impulsive infrasound and low frequency noise is repeatedly disturbed and interrupted sleep, (resulting in prolonged and chronic sleep deprivation, which itself is acknowledged as a method of torture by the UN Committee against Torture xxviii), the individual public officials are risking future charges against them ...
Immediate action is required from public officials at every level of government who are responsible for the current situation. This is not only to prevent further serious damage to human health, but also to reduce officials’ personal risk of future successful prosecution against them by severely impacted rural residents, for torture or acquiescence to torture, or for ignoring ongoing cruel, inhuman and degrading treatment, about which officials have been repeatedly personally advised.
[Originally published at Master Resource]