In December, four acoustic consulting firms collaborated to study wind turbine noise at three Brown County, Wisconsin homes that had been abandoned by their owners after the nearby Shirley Wind Project began operations.  The study, organized by regional environmental group Clean Wisconsin and paid for by the state Public Regulatory Commission, will help inform the PRC’s consideration of a proposed new wind farm in the area.  

Two things stand out about this new study.  The first is the choice to bring together several acousticians who have previously been widely cited by opposite sides of the turbine siting debate. The study team included one firm  (Hessler and Associates) commonly hired to do sound assessments for wind developers, another (Rand Acoustics) that has become widely championed by concerned citizens groups because of its much more cautionary assessment of turbine noise, and a third (Schomer and Associates) whose work has often been in the middle ground, with particular papers being seized on by each side in the siting debate; the fourth firm (Channel Islands Acoustics) has worked much less on wind farm issues than the other three.  This diverse group of acousticians produced a 13-page consensus report (edited to 12 pages in the final version submitted to the PSC), along with an appendix report from each team, all of which focus on different aspects of the study that they found most compelling. 

The second virtue of this study is that it clearly documented, for the first time, specific sources of infrasound (sound at frequencies below 20Hz) and low-frequency noise (audible sound above about 20Hz) from turbines that are consistently measurable inside homes. The data they collected clearly showed peaks in the sound spectrum that correspond to the “blade passing frequency” (BPF) of just under 1Hz, or one pulse per second, and several harmonics of the BPF up to about 5Hz.  These pulses showed up both inside and outside the closest home, 1280 ft from the nearest turbine.  In addition, they measured a more modest infrasound and low-frequency peak at 15-30Hz, which reflects the natural resonance and flexibility of typical home construction; this peak may have been triggered by turbine sound or by wind or other outdoor sound sources. One of the acousticians, Rand, notes in his appendix a possibly corresponding pulse of outdoor sound in the 9-14Hz range that can be associated with inflow turbulence hitting turbines.  Still, the infrasound that was measured in this study, as in most other similar measurements of wind turbine noise, is at lower dB levels than what is typically considered perceptible by humans. (Ed. note: two emerging yet still limited bodies of work suggest that turbine infrasound may have rapid peaks that approach standard perceptual thresholds, and that our ears may respond physiologically to sounds at lower levels than are perceived; nothing in this Wisconsin study address these questions, though later analysis of the data may contribute to the study of short-term peaks.)

Since the study took place in homes that were abandoned by homeowners who all complained of debilitating health effects, including sleeplessness, nausea, and depression, part of the goal of the study was to see whether they could identify any possible acoustic triggers for these negative responses.  The authors collectively noted that “the issue is complex and relatively new” and concluded that this work “was extremely helpful and a good start to uncover the cause of such severe adverse impact reported at this site.”  

The consensus report, signed by all members of the team, introduces a new hypothesis, based on a US Navy study that found that vibrations can trigger nausea in pilots when in the frequency range of up to 0.5-0.9 Hz, with the peak “nauseogenicity” occurring at 0.2 Hz.  Of particular concern is that as turbine blades get longer, the BPF is being reduced; only the recent generation of turbines has dropped below 1Hz (thus perhaps helping to explain the recent surge of health complaints among a subset of turbine neighbors), and planned larger blades will drop close to that 0.2Hz range of maximum inducement of nausea.  While stressing that this is, as yet, a very preliminary supposition, especially since it involves a study based on physically vibrating the body, while turbine infrasound is a vibration of the air around a body, the authors still agreed that:

The four investigating firms are of the opinion that enough evidence and hypotheses have been given herein to classify LFN and infrasound as a serious issue, possibly affecting the future of the industry. It should be addressed beyond the present practice of showing that wind turbine levels are magnitudes below the threshold of hearing at low frequencies.

In particular, the research team agreed that a further literature search for studies related to vibration-induced nausea should take place (Paul Schomer is working on this), and that a “threshold of perception” test should be conducted, to see what proportion of residents are able to perceive the faint signals in either audible or infrasonic ranges.  Only one of the five acousticians, Rand, could detect sound at all residences; he also reported headache and/or nausea (it is also noted that he is the only one among the five researchers who suffers from motion sickness).

As often happens, the reaction to this study ranged from “this changes everything” to “this is nothing new,” with some saying it proves negative effects and others that it proves wind energy is safe.  For a run-down of the reactions, a brief look at each of the four appendices, and links to download the study, click on through… 

As we begin to look at the wild range of responses, bear in mind that the consensus report makes no claims of proving anything, one way or the other.  It gathers the data, some of which is new and interesting, introduces a possible but far from fully investigated mechanism for one physical response (nausea), and urges continued study.  It may be notable that the Hesslers agreed that the data and hypothesis warranted further study, in that they’ve tended to insist that there is no indication that infrasound levels below human audibility should be a concern, and need not be measured or assessed for new wind projects. (It should be stressed, though, that the study of infrasound here was suggested by David Hessler; still, their Appendix report and public statements since the report was released affirms their contention that what was measured here does not change their current understanding of the risks.)  Many of the news reports did manage to hit a realistic middle ground of recognizing that this study opens some new pathways for study, but did not in itself show any clear correlation between turbine sounds and health responses. The headline that best captured both the innovative aspects of this study and its limited scope was probably “Study suggests wind turbines’ low-frequency noise could cause health woes”  (note suggests, rather than proves; and could, not does or does not).

But two responses stand out for their extreme hyperbole on each side of the spectrum.  The first came from Wisconsin state Representative Andre Jacque, who quickly hit the papers with his contention that the study found “dangerous levels” of infrasound and was proof that wind turbines can cause health issues; he said that “these results compel (the PSC) to act immediately to keep this nightmare from spreading,” and called for an immediate suspension of permitting for new wind projects.

On the other end of the spectrum, representatives of one national and one regional wind industry trade association issued an op-ed claiming that “Science proves wind energy is safe for Wisconsin,” which takes the complementary tack of lumping this study in with others that, similarly, compile relevant information and find valid and sometimes urgent cause for further study, but not any solid links between turbine sound and health effects (indeed, in each case, the study recommended would help clarify the largely uninvestigated questions of cause and effect).  Unfortunately, this piece also indulges in a kind of spin that has been too common in recent years, dismissing the work, and personal experiences, of Rand because he is predominantly hired by groups concerned about wind development near their homes, while putting forth a reassuring quote from David Hessler, without similarly noting that his firm is predominantly hired by wind developers.  

This semi-slander of the “other” experts and easy championing of “our” experts is an ongoing problem on both sides of the issue, and is one of the reasons that I am so enthusiastic about these diverse acousticians working together on this study.  In my mind, neither of these long-time community noise specialists should be dismissed due to their differing opinions about what sound level is an appropriate target to minimize negative responses among neighbors; they each offer valuable perspectives, and they each explain their reasoning clearly.  

On that note, while I encourage anyone really involved in these issues to download and read the entire 50-page document that includes the consensus report and all four appendices, here’s a teaser of what each of the four consultants focuses on in his individual report:

• Channel Islands Acoustics (Bruce Walker) presents a small sampling of the vast amounts of sound data collected, with a focus on how well correlated the inside and outside data was at each location (which is one indication of whether the measured-but-inaudible infrasound inside is indeed related to the turbine sound).  He also notes an important detail: units doing continuous monitoring near turbines closest to two of the homes showed “significant variability in near-field sound levels” that suggest “turbine noise emissions may have decreased shortly before the team started and increased shortly after the team stopped measuring on some days.”  This suggests that at least some of the data gathered likely doesn’t reflect typical peak sound levels (the team focused on one home per day/night).
• Hessler and Associates (George and David Hessler) present graphs showing the difference between indoor and outdoor measurements at each location, and highlight some ambiguities that need to be addressed: uncertainty about whether the house-resonance infrasound is driven by turbine sound or something else, and their ongoing questions about whether what is measured as infrasound is merely wind noise on mics, even when wind screens are used.  While also stressing that the infrasound found here is well below perceptual thresholds, and less severe by some measures than poorly-sited industrial gas turbines, the Hesslers affirm at the end of their report that “enough was learned by these investigators, all with quite different past experiences, that it can be mutually agreed that infrasound from wind turbines is an important issue that needs to be resolved in a more conclusive manner by appropriate study.”
• Rand Acoustics (Rob Rand) stresses a recurring theme of his, that “the human being is the best reporting instrument,” encouraging study of correlations between ill effects and turbine operations or proximity.  He compiles notes on his own physiological responses, as well as those of the (former) residents of the study homes, many of whom were on site during the study, and presents a summary of their experiences,  including high annoyance with the noise inside their homes, and finding partial relief when retreating to their basements and more complete relief when getting beyond about 3 1/2 miles from turbines. He also presents some results of his air-pressure data, showing clear pulses in the closest house, and little discernable effect at a more distant house when the turbines were turning slowly.  Finally, he suggests that blade flexing, occurring at the rotation rate of 0.2-0.25Hz, is another possible driver of some nauseogenic (nasea-inducing) pressure variations, and presents charts showing the evolution of turbine technology toward longer blades with slower rotation rates.
• Schomer and Associates (Paul Schomer) also focuses in on the coherence data, presenting information that complements that of Walker.  He also stresses that this study, including conversations with the former residents, was his first understanding that health issues can be completely separate from noise annoyance (i.e., that health effect may be triggered by inaudible or barely audible sound).  From this foundation, his appendix includes the most thorough summary of the Navy motion sickness/nauseogenicity research and its possible implications for the wind industry as turbines get larger.  In particular, he notes that it may be necessary to lower noise limits by 6dB for 2MW turbines, as compared to smaller 1MW turbines, if the Navy research is confirmed to be applicable to wind turbine infrasound.

To return to the consensus report for a couple of final points:

The operator of the Shirley Wind Project, Duke Energy, refused to cooperate with the study by turning individual turbines off at some times, so the researchers could measure the differences in sound levels (both inside and out) when turbines were on or off.  It’s also not clear whether Duke will be making wind or turbine power data availalble for further analysis of the data.  The consensus report stressed that “the cooperation of the wind farm operator is absolutely essential” for obtaining the complete and necessary data.

Finally, there is some controversy and contention around the fact that while the participating acousticians all signed on to a 13-page consensus report, one section was deleted before Clean Wisconsin submitted it to the PSC.  This section contains recommended noise limits for the proposed Highlands Wind project; according to Clean Wisconsin, they checked with the PSC and were directed that the study should only be looking at the Shirley project.  Presumably, each acoustician can submit their own recommendations on Highlands during the hearings on that project.

Still, most citizens won’t be digging through the dozens of submissions to the Highlands record and read through hundreds of pages of individual submissions and transcribed oral testimony and questions for each witness.  So, the recommendations that were agreed to by the five are interesting to see.  In the omitted section, the Hesslers and Schomer offered slightly different recommendations, both lower than the current 45dB limit; they note that “in this specific case, it seems justified to the two firms to be conservative (one more than the other) to avoid a duplicate project to Shirley (ed note: i.e., involving several abandoned homes and about fifty formal health effects complaints submitted to the Brown County Board of Health) at Highlands because there is no technical reason to believe the community response would be different.”

The recommendations offered are based on work that George Hessler and Paul Schomer have done independently in which, using entirely different approaches, they’ve both zeroed in on a design goal for wind projects of keeping noise below 39 or 40dB, measured as a long-term (two week) average.  In this case, the Hesslers recommend a limit of 39.5dB.  Schomer urges an expedited perception study to determine if some individuals can indeed perceive current wind turbine operations at the more distant houses in the current study; if that cannot happen, then Schomer somewhat reluctantly encourages a limit of 33.5dB, reflecting the 6dB reduction that may be necessary to assure the infrasound/pressure nauseogenic effects of the proposed 2.5MW turbines at Highlands are similar to those of the the 1.5MW turbines at Shirley.  He qualifies this by noting that at the one project currently using the same 2.5MW turbines, with a town 1700 feet away, there have been no noise issues reported to the wind turbine company.

To download the report: A Cooperative Measurement Survey and Analysis of Low Frequency and Infrasound at the Shirley Wind Farm in Brown County, Wisconsin.  Report Number 122412-1
Final official version as submitted by Clean Wisconsin, Document ID 178773 (omits Highland recommendations, includes cover letter explaining why.  See this earlier submission for more complete response to claims by Forest Voice that Clean Wisconsin had “edited” the report as signed.)
Initial version as signed by the acousticians, including Highland recommendations, Document ID 178263.  Clean Wisconsin has filed a motion to remove this version from the record. It appears to be widely available at other sites online; most media and advocacy sites that are linking to copies on their sites have this one, the first version posted at the PRC site.

To stay up to date on all the Highlands testimony:  All documents related to the proposed Highlands project, including submissions from various acousticians (including those here and others), public interest groups (including Clean Wisconsin supporting the project, and Forest Voice opposing the project), and individual citizens, go to http://psc.wi.gov/ and do a search for Docket ID 2535-CE-100 (separate this into the three entry boxes; no need to select a Document Type).  You’ll quickly see that there is a huge amount of material, including rebuttals to earlier submissions and motions from Clean Wisconsin to strike some of the submissions of others from the record.