Hanning, Wind Turbine Amplitude Modulation & Planning Control Study, INWG

Independent Noise Working Group

Work Package 3.2 – Excessive Amplitude Modulation, Wind Turbine Noise, Sleep and Health

Dr Christopher D Hanning BSc MRCS.MRCP MB BS.FRCA MD Honorary Consultant in Sleep Medicine, University Hospitals of Leicester

Prof Alun Evans MD Visiting Research Professor, Centre for Public Health, The Queen’s University of Belfast
Carmen Krogh BSc Retired Health Canada executive
Dr Sarah Laurie BM, BS (Flinders) Chief Executive Officer, Waubra Foundation

1 Executive summary

1.1 Excessive noise is harmful to human health, particularly through adverse effects on sleep (WHO 2011(1)). Regulation of wind turbine noise is recognised as necessary to prevent adverse effects on the human population.

1.2 UK guidance ETSU-R-97(2) (‘ETSU’) states in its executive summary “This document describes a framework for the measurement of wind farm noise and gives indicative noise levels thought (my emphasis) to offer a reasonable degree of protection to wind farm neighbours, without placing unreasonable restrictions on wind farm development or adding unduly to the costs and administrative burdens on wind farm developers or local authorities.”. It is reasonable to infer that the authors had no certainty that their recommendations were adequate nor were they solely concerned with protecting the sleep and health of wind farm neighbours and therefore moderated their recommendations accordingly.

1.3 The acoustical shortcomings of ETSU have been discussed in detail elsewhere (Bowdler 2005(3)and Cox, Unwin, Sherman 2012(4)). Despite the growing evidence of harm and the authors’ caveats, no substantive review of the fundamental principles of ETSU has been conducted nor has any substantive research been conducted in the UK. The Hayes McKenzie Partnership conducted a small study on behalf of the DTI5 in 2006 as a result of which they recommended reductions in night time noise levels. These were removed from the final report, only emerging after the earlier drafts were obtained using Freedom of Information requests (DTI 2006, a, b, c(5)).

1.4 The methodology and indicative noise levels of ETSU have been adopted in many other jurisdictions.

1.5 A large body of evidence, presented below, demonstrates that human sleep and health are adversely affected at wind turbine noise levels permitted by ETSU. There is particular concern for the health of children exposed to excessive wind turbine noise. The inadequate consideration of excessive amplitude modulation (EAM) is a major factor in the failure of ETSU to protect the human population.

1 WHO 2011 Burden of disease from Environmental Noise

2 ETSU-R-97 1996 The Assessment & Rating of Noise from Wind Farms

3 Bowdler D 2005 ETSU: Why it is wrong New Acoustics

4 Cox R Unwin D Sherman T 2012 Wind turbine noise impact assessment, where ETSU is silent

5 DTI 2006 The Measurement of Low Frequency Noise at Three UK Wind Farms – W/45/00656/00/00 – Hayes McKenzie Partnership Plus draft reports 2006 a, b, c.

2 Objective and Scope

2.1 Objective

To summarize the effects of Excessive Amplitude Modulation (EAM) on people living close to wind turbines including annoyance, sleep disturbance and health effects through a review of the available health related literature.


2.2 This report discusses ETSU’s ability to protect noise sensitive receptors (‘receptors’) from sleep disruption and therefore harm to their health and in this context to consider the contribution of EAM. Sections 3 and 4 discuss necessary preliminary matters relating to noise levels and setback distances and the characteristics of wind turbine noise. Section 5 describes the evidence of the inadequacy of ETSU.

2.3 Appendix A is included to explain the effects of noise on sleep. Appendix B lists the author’s qualifications. Appendix C includes the figures and tables referred to by Sections 3, 4 and 5 and Appendix D the bibliography

2.4 Source material Publications cited and other source material are noted at the foot of each page and the complete list, including website links where appropriate, is attached as Appendix D. Where several articles come to the same conclusion, only the most recent may be cited, in the interests of brevity. As far as possible, articles published in peer-reviewed journals are cited. However, it is inevitable that some of the material is available only on the internet reflecting the paucity of government sponsored research. Reviews are generally only cited if they have been published recently and are comprehensive reviews of the literature.

2.5 Work Package 3.1 details a survey of local planning authorities which demonstrates convincingly that EAM and complaints are a common feature of UK wind farms yet most research of the effects of wind turbine noise is from outside the United Kingdom.

3 Noise levels and setback distance

3.1 The predicted wind turbine noise experienced by receptors is generally calculated from manufacturer’s predictions as to turbine noise at varying wind speeds using standard formulae for the attenuation of sound with distance, allowing for wind direction, ground conditions etc. As an approximation, it is assumed that noise levels will decrease by 6dBA for a doubling of distance. Allowance is made for additional turbines to derive a predicted noise level. Comparison is then made to background noise levels with the assumption that the turbine noise is masked by background noise. As is shown below, neither assumption is justified. Contrary to logic and common sense, ETSU permits a higher external noise level of 43dBA at night even though background noise levels are generally lower at this time.

3.2 After the application of the ETSU methodology, setback distances for human habitation from modern 2.5-3MW turbines in the UK are typically in the region of 500-600m.

3.3 An Institute of Acoustics (IoA) working group (Bowdler et al 2009(6)) issued advice on allowing for wind shear where wind speeds at hub height are greater than those at ground level. In theory, this should have led to greater setbacks as receptor predicted noise levels would have been greater, especially during the night. In practice, setbacks have remained unchanged or decreased. Stigwood 2011(7) has investigated the method and concluded “… in all cases analysed there was a loss of community protection when adopting the article (IoA) method.”

3.4 Most published research has used setback distance rather than measured or calculated noise levels, not least because of the expense of measurement and the inaccuracies of calculation. ETSU mandates setbacks of around 600m. Therefore any study which shows harm at distances of 1000m or greater is unequivocal evidence that ETSU does not provide adequate protection for wind farm neighbours.

6 Bowdler D et al 2009 Prediction and assessment of wind turbine noise Acoustics Bulletin March/April:35

7 Stigwood M 2011 The effect of common wind shear adjustment methodology on the assessment of wind farms when applying ETSU. MAS Environmental 27th September 2011.

4 Characteristics of wind turbine noise

4.1 Wind turbine noise is not comparable to that from other common environmental sources such as traffic, rail and aircraft. Several studies have shown it be more annoying than these sources for comparable A weighted noise levels (van den Berg et al 2008(8) Fig 1, Pedersen and Persson Waye 2004(9) Fig 2). “Annoying” in this context is used in the psychological sense of causing a degree of stress sufficient to cause concern for health, not simply an irritation (WHO 2009(10)) (Appendix A 1.29, 1.30 and 1.31). In addition, wind turbine noise is not well masked by ambient noise, being audible 10-15dB below background noise (Nelson 2007(11), Hayes 2007(12), Bolin 2009(13), Pedersen et al 2010(14)).

4.2 Two characteristics of wind turbine noise have been advanced to explain these differences, modulation and low frequency noise (LFN) (James 2012(15), Thorne 2012(16), Large and Stigwood 2014(17)).

4.3 Wind turbine noise emissions are amplitude modulated (AM) as the turbine blades pass the tower and pass through areas of differing wind speeds. The effect may be increased if there is interaction between the emissions from nearby turbines. The result is an impulsive noise character often described as “thumping” or “rumbling” (Lenchine & Song 2014(18)). The degree of AM varies with a number of factors including wind speed and direction and blade configuration. Especially prominent modulation is deemed to be excessive amplitude modulation (EAM). ETSU makes some allowance for AM (3dB peak to trough) in the near field but makes no allowance for far field modulation nor for lower frequency noise content.

4.4 There is a large body of evidence to show that AM is more annoying than un-modulated noise (See WP2.2 for detailed consideration). Of particular note is a laboratory based study by Lee et al 2011(19). 30 subjects were exposed to recorded wind turbine noise at varying volumes and degrees of AM. They concluded that “….the equivalent sound level and the amplitude modulation of wind turbine noise both significantly contribute to noise annoyance.” Pohl et al 2014(20) asked residents near the Wilstedt windfarm in Germany to record annoying sounds. They concluded that perceived annoyance correlated with the presence of AM.

8 van den Berg GP et al 2008 WINDFARM Perception. Visual and acoustic impact of wind turbine farms on residents. FP6-2005- Science-and-Society-20. Specific Support Action Project no 044628

9 Pedersen E and Persson Waye K 2004 Perception and annoyance due to wind turbine noise – a dose-response relationship J Acoust Soc Am 116 3460–347

10 World Health Organisation 2009 Night noise guidelines for Europe, Copenhagen

11Nelson D 2007 Perceived loudness of wind turbine noise in the presence of ambient sound. Second International Meeting on Wind Turbine Noise. Lyon, France

12 Hayes M 2007 Affidavit in reply Makara Wind Farm New Zealand Environmental Court W59/2007

13 Bolin K 2009 Wind Turbine Noise and Natural Sounds-Masking, Propagation and Modeling Doctoral Thesis Royal Institute of Technology, Stockholm

14 Pedersen E, van den Berg F, Bakker R, Bouma J 2010 Can road traffic mask sound from wind turbines? Response to wind turbine sound at different levels of road traffic sound Energy Policy 38:2520-7

15 James R 2012 Wind Turbine Infra and Low-Frequency Sound: Warning Signs That Were Not Heard Bulletin of Science, Technology & Society 32 108–127

16 Thorne R 2012 Submission to the Senate Environment and Communications Legislation Committee

17 Large S and Stigwood M 2014 The noise characteristics of “compliant” wind farms that adversely affect its neighbours Presented to inter.noise 2014, Melbourne, Australia

18 Lenchine V and Song J 2014 Special noise character in noise from wind farms Presented to inter.noise 2014, Melbourne, Australia

19 Lee S et al 2011 Annoyance caused by amplitude modulation of wind turbine noise Noise Control Eng J 59:38-46

20 Pohl J et al 2014 Untersuchung der Beeinträchtigung von Anwohnern durch Geräuschemissionen von Windenergieanlagen und Ableitung übertragbarer Interventionsstrategien zur Verminderung dieser. Deutsche Bundesstiftung Umwelt

4.5 Noise character is an important factor in determining whether a subject arouses or awakens from sleep, impulsive noises being more likely to cause an arousal (Solet 2010(21)). An arousal is a brief lightening of sleep. Both arousals and awakenings fragment sleep, impairing its restorative properties. Impulsive sounds are chosen for fire alarms as being more likely to awaken sleepers (Bruck 2009(22)). Such studies provide support for the assertion that EAM is an important factor in the effects of wind turbine noise on sleep.

4.6 Wind turbine noise contains a large element of LFN, the contribution increasing with turbine size (Moller and Pedersen 2011(23)). The adverse effects of LFN on health have been recognised for decades, although they have been generally overlooked until recent years (Enbom 2013(24)). A WHO report in 1999(25) stated “It should be noted that a large proportion of low-frequency component in a noise may increase considerably the adverse effects on health.” Kelley (1985(26), 1987(27) reported adverse health effects from wind turbines over 25 years ago and identified LFN as the likely cause. James (2012(15) has documented the failure of the wind industry to acknowledge the early research. Pedersen (2004(9)) reviewed the available evidence in 2004 and concluded “… low frequency noise, at comparatively low sound pressure levels, disturbs sleep.” Schomer, an American acoustician and one of the co- operators in the survey of LFN (Clean Wisconsin 2012(28)), subsequently stated in a letter to the Public Service Commission of Wisconsin (2013(29)) “….this case as well as the literature and case studies all over the world have suggested that people are leaving their homes because they are being exposed to significant levels of pulsating ultra-low frequency sound produced by wind turbines. In addition there is no question that larger turbines produce more infrasound below 1 hertz which increases the likelihood that health problems will occur unless noise limits are dramatically reduced through the use of smaller turbines or lower noise limits are required at each house.” Schomer (2015(30))has recently proposed that the biological effects of LFN from turbines are related to susceptibility to motion sickness (Schomer 2015).

4.7 ETSU makes no allowance for LFN despite the increase in turbine size since its formulation.

21 Solet JM et al 2010 Evidence-based design meets evidence-based medicine: The sound sleep study Concord CA: The Center for Health Design

22 Bruck D et al 2009 How does the pitch and pattern of a signal affect auditory arousal? Journal of Sleep Research 18:196-203

23 Moller H and Pedersen CS 2011 Low frequency noise from large wind turbines J Acoust Soc Am 129: 3727-3744

24 Enbom H and Enbom I 2013 Infrasound from wind turbines: an overlooked health hazard Läkartidningen 110:1388

25 World Health Organization 1999 Guidelines for community noise

26 Kelley ND et al 1985 Acoustic noise associated with the MOD-1 wind turbine: its source, impact and control Solar Energy Research Institute Report SERI/TR-635-1166

27 Kelley ND 1987 A proposed metric for assessing the potential of community annoyance from wind turbine low-frequency noise emissions Solar Energy Research Institute Report SERI/T9-217-3261

28 Clean Wisconsin 2012 Report Number 122412-1 A Cooperative Measurement Survey & Analysis of Low Frequency & Infrasound at the Shirley Wind Farm in Brown County, Wisconsin.

29 Schomer P 2013 Letter to Zuelsdorff Environmental Analysis and Review Coordinator, Public Service Commission of Wisconsin

30 Schomer P et al 2015 A theory to explain some physiological effects of the infrasonic emissions at some wind farm sites J Acoust Soc Am 137:1356

 5. Wind turbine noise, sleep and health

5.1 Introduction

Several types of evidence lead to the conclusion that ETSU does not provide sufficient protection for receptors:

  • Epidemiological studies and anecdotal reports of harm following exposure to wind turbine noise;
  • Opinions from other experts as to appropriate setback distances and noise limits;
  • Studies of health related effects such as annoyance. Some of these studies have commented on the effects of sleep but have not used appropriate outcome measures;
  • Studies of health effects and sleep

5.2 Epidemiological and anecdotal studies

 There are a large number of anecdotal reports and surveys. In the interests of brevity, they will not be detailed here but are described in an online review (Hanning 2010(31)). One survey is particularly worthy of mention, WindVOiCe (Krogh 2011(32)), as the results have been published in a peer-reviewed journal. WindVOiCe is a self-reporting survey of Canadian communities affected by wind turbine noise. As of July 2010, 144 responses had been received of which 118 reported one or more health effects of which 84 (58%) reported sleep disturbance. There were no age differences between those that reported sleep disturbance (mean (range)) (51.5 yr (19-79)) and those that did not (52.2 yr (26-86)). All bar five of those reporting sleep disturbance live within 1500m of turbines adding further support to a minimum setback of at least that distance.

5.3 Anecdotal reports are commonly dismissed in industry sponsored reviews (eg Colby et al 2009(33)) as not acceptable evidence. Phillips, an epidemiologist, in a peer-reviewed article (Phillips 2011(34)), has examined these claims, reviewed the evidence and concluded:

“There is overwhelming evidence that wind turbines cause serious health problems in nearby residents, usually stress-disorder type diseases, at a nontrivial rate. The bulk of the evidence takes the form of thousands of adverse event reports. There is also a small amount of systematically-gathered data. The adverse event reports provide compelling evidence of the seriousness of the problems and of causation in this case because of their volume, the ease of observing exposure and outcome incidence, and case-crossover data. Proponents of turbines have sought to deny these problems by making a collection of contradictory claims including that the evidence does not “count”, the outcomes are not “real” diseases, the outcomes are the victims’ own fault, and that acoustical models cannot explain why there are health problems so the problems must not exist. These claims appeared to have swayed many non- expert observers, though they are easily debunked.”

31Hanning C 2010 Wind turbine noise, sleep and health

32 Krogh C et al 2011 WindVOiCe, a Self-Reporting Survey: Adverse Health Effects, Industrial Wind Turbines, and the Need for Vigilance Monitoring Bull Sci Tech Soc 31: 334

33 Colby et al 2009 Wind Turbine Sound and Health Effects; an Expert Panel Review American and Canadian Wind Energy Associations

34 Phillips C 2011 Properly interpreting the epidemiologic evidence about the health effects of industrial wind turbines on nearby residents Bull Sci Tech Soc 31:303-8

5.4 The weight of epidemiological evidence is that wind turbine noise adversely effects health at distances of at least 5km.

5.5 Expert opinion

The opinions on setback distances for 19 groups of scientists, legislators and acousticians are shown in Table I, Appendix C (Hanning 2010(31). The mean (range) setback distance recommended is 2.08km (1-3.2). Other recommendations are given in the text.

5.6 Thorne, an Australian acoustician who has investigated wind turbine noise and its health effects, concludes, in a peer-reviewed article “A sound level of LAeq 32 dB outside a residence and above an individual’s threshold of hearing inside the home are identified as markers for serious adverse health effects affecting susceptible individuals.” (Thorne 201135). Thorne 201436 also has recommended a 2km setback as a result of health studies on three Australian windfarms.

5.7 Schomer (one of the co-operators in Clean Wisconsin 2012(28)) recommended “….a criterion level of 33.5 dBA”.

5.8 Arra et al 2014(37) conducted a systematic review of the literature and noted that “All peer- reviewed studies captured in our review found an association between wind turbines and human distress.” They concluded “….we have demonstrated the presence of reasonable evidence (Level Four and Five) that an association exists between wind turbines and distress in humans. The existence of a dose-response relationship (between distance from wind turbines and distress) and the consistency of association between studies found in the scientific literature argues for the credibility of this ”

5.9 A recent exhaustive review (Schmidt and Klokker 2014(38)) concluded “Exposure to wind turbines does seem to increase the risk of annoyance and self-reported sleep disturbance in a dose-response relationship. There appears, though, to be a tolerable level of around LAeq of 35dB”. This is about 6dB less than the permitted ETSU night time level, implying a doubling of the setback (assuming a decay of noise level of 6dB per doubling of distance).

5.10 Jefferey and et al 2014(39) reviewed the literature, with special reference to LFN, and concluded “we expect that, at typical setback distances and sound pressure levels of IWTs in Ontario, a nontrivial percentage of exposed people will be adversely ”

5.11 The weight of independent expert opinion is that wind turbine noise adversely effects health at distances of at least 1.5km.

35 Thorne R 2011 The Problems With “Noise Numbers” for Wind Farm Noise Assessment Bulletin of Science Technology Society 31:262-290

36 Thorne R 2014 The Perception and Effect of Wind Farm Noise At Two Victorian Wind Farms An Objective Assessment June 2012. Reissued June 2014

37 Arra I et al 2014 Systematic review 2013: Association between wind turbines and human distress Cureus 6:e183

38Schmidt JH and Klokker M 2014 Health effects related to wind turbine noise exposure: a systematic review PLoS ONE 9:e114183.

39 Jefferey RD et al 2014 Industrial wind turbines and adverse health effects. Can J Rural Med 19:21-26

Download the complete Study, including appendices, figures and tables →