記住我
Commentary on: Chaban R, Ghazy A, Georgiade E, Stumpf N, Vahl C. Negative effect of high-level infrasound on human myocardial contractility: in-vitro controlled experiment. Noise Health 2021;23:57–66.
Introduction
The experiment discussed here is not the first trial of the working group “infrasound” of the medical department of the University of Mainz to find evidence for their hypothesis that wind turbines (WT) infrasound could affect heart muscle contractility. It appears to be a follow-up of a previous study whose results were disclosed as a conference presentation with the title “Are There Harmful Effects Caused by the Silent Noise of Infrasound Produced by Windparks?” However, as described in the available abstract,[2] the team did not bother to expose the in vitro heart muscle samples to genuine airborne infrasound. Instead, they affixed one sample side directly to an “industrial vibrator,” which was then excited with frequencies of 10 and 20 Hz to produce excursions equivalent to 5% and 10% of the tissue lengths (4 and 6 mm) of the samples. So instead of exposing the myocardial tissue to dynamic pressure, which is a scalar and can be described as a slow and homogeneous variation of the quasi-hydrostatic pressure at these low frequencies, a sinusoidal translational force was applied, stretching and squeezing the tissue considerably along one axis. Little surprisingly, the authors indeed detected a loss of contractility, which however was partially recovered after stopping the mechanical interference.
Probably after realizing that this setup is in no way a valid representation of airborne sound exposure, the working group devised a new experimental design that avoids the direct mechanical connection of the muscle sample with the vibrator. We briefly recapitulate the design and analyze whether it is suitable to accurately emulate the action of airborne infrasound waves on the body or not.
Afterward, we analyze infrasound measurements conducted by a German government authority in the vicinity of WTs and compare their actual exposure to the infrasound levels applied in the experiments of Chaban et al. We also compare them with the intensities deployed in ultrasonography as a reference to sound exposure levels considered safe for muscle tissue.
Critical analysis of the setupAs explained by Chaban et al., the atrial heart muscle samples were obtained from patients who had previously undergone heart surgery. From each patient, two small strips were prepared and affixed to a test bench with tweezers. The strips were then dipped in a warm saline solution with glucose and infused with a mix of 95% oxygen and 5% carbon dioxide to keep the cells alive. One strip was kept in a quiet environment, while the other was put in a test box where it was exposed to sinusoidal sound at 16 Hz with three different sound pressure levels (SPL) (100, 110, and 120 dB) for 1 hour. Both samples were excited with electrical pulses at 75 bpm to simulate heart activity. Before and after this period, the contractility of the samples was compared with those that had been left quiet. While 100 dB did not show a noteworthy effect, the contractility force was impaired by an average of 11% at 110 dB exposure and 18% at 120 dB exposure using 18 sample pairs.
A schematic drawing of the experimental setup is given in [Figure 2] of the paper of Chaban et al. and shown here in [Figure 1].
Figure 1 Schematic drawing of the arrangement copied from the publication of Chaban et al.[1]
Figure 2 (Left) Experimental setup presented in documentary of German TV broadcaster ZDF (Zweites Deutsches Fernsehen)[3] showing the large gap in the acrylic hood above the blue ground plate (time: 24:09). (Right) Airflow at forward (green arrows) and backward (red arrows) cone excursion of the woofer, showing the acoustic short circuit occurring for low frequencies. It decreases the acoustic impedance and further increases the particle velocity, which is much higher than in the far field of a sound wave with the same sound pressureThe section “Infrasound application” explains that a “30 cm Woofer was connected to a power amplifier, to a computer, and fixed at the top of a specially-made closed chamber, where the muscle investigation system was inserted.” However, this information does not seem to be accurate with respect to a crucial detail: the chamber was, in fact, not closed. The evidence is given by the authors themselves. They presented their experimental setup in a documentary video.[3] This documentary discussed infrasound emitted by WTs and the alleged detrimental effects on the health of residents, claiming the supposed debilitation of the heart muscle to possibly be one of the effects. In the recording, it can be clearly seen that the “closed chamber” in reality is merely a hood made of acrylic glass, leaving a large gap on the lower side ([Figure 2]).
The fact that the air shoved by the loudspeaker on top of the hood easily escapes via the large gap without encountering notable resistance is corroborated by a scene in which the first author of the article, Rayan Chaban, is holding a sheet of paper in front of the gap to demonstrate the strong vibration induced by the air moving through the gap. This is an indication that his team did not realize the principal design flaw of their setup: such low-frequency experiments necessarily have to be conducted in a hermetically closed vessel; otherwise, excessive particle velocity invalidates the results. The reasons are given in Annex A. To summarize, in close vicinity to a low-frequency sound source such as the 30 cm woofer used in Chaban’s setup, the particle velocity and displacement are many times greater than in the far field of a sound wave with the same sound pressure.[4] The acrylic hood even acts as an extension of the very near field of the woofer because its constant rectangular cross-sectional area partially guides the large air volume periodically displaced by the loudspeaker on top. At the low frequency of 16 Hz used in the experiment, the gap at the foot of the hood exhibits only minor resistance to the airflow and the displaced air volume passes easily – exactly what Rayan Chaban showed with the strongly moving sheet of paper in the video.[3] Once outside the chamber, it partially recombines with the volume displaced at the other side of the loudspeaker cone, which is almost opposite in phase – a phenomenon called an acoustic short circuit, which further diminishes the acoustic impedance. The air volume under the hood is pushed on one side and pulled on the other side.
The capacitive electret microphone located under the hood and used to monitor and adjust the SPL constitutes a pressure sensor and is insensitive to the grossly exacerbated particle velocity. As a consequence, in the experiment of Chaban et al., the modulus of the sound intensity is much higher than suggested by the SPL calculated from the microphone signal. Whether this signal was correctly captured is dubious because instead of relying on laboratory-grade instrumentation, a cheap (albeit calibrated, according to Chaban) amateur electret microphone (“Superlux ECM999”) was chosen for this purpose, phantom powered by a simple preamplifier for amateur musical production whose gain can only be adjusted continuously by a potentiometer, in contrast to the fixed precision gain steps of professional measurement preamplifiers that allow for gain changes without loss of calibration.
The problem with the excessive periodic airflow under the hood is that it unavoidably induces considerable vibration in the whole setup and especially in the long arms of the tweezers of the “heart muscle investigation system.” These vibrations are transferred to the tips of the tweezers that fix the muscle samples. They are attenuated to a certain extent by the saline solution in which the samples are immersed during the experiment; however, they undoubtedly still exert a notable periodic biomechanical strain on the tissue. This external strain, which consists of periodically pushing and pulling the material at 16 Hz, does not exist in the normal situation of the heart embedded in our thorax. Owing to the large wavelength of over 20 m, the infrasound pressure acts as a homogeneous scalar with an almost equal phase on all sides of our almost incompressible body. The isentropic compressibility of water, of which our body is mostly composed, is almost 16,000 times lower than that of air. Moreover, the infrasound intensity inside the body is reduced more than 3500 times (35.5 dB) as given by the relation of the acoustic impedances between water and air. The result is the quasi-zero movement of the tissue molecules. In the experiment, the protective layer of bones and liquid that normally shields the heart against extraneous particle velocity is partially bypassed by means of the direct rigid mechanical connection from the outside world to the tissue. The periodic biomechanical irritation to which the fibers are subjected presumably entails mechanical fatigue and adverse biochemical reactions, especially at the edges of the samples that are fixed with the tips of the tweezers causing an abrupt change of elasticity.
As the periodic perturbation which the heart tissue suffers is thus a mixture of vibration, strain, and, to a lesser extent, the dynamic pressure which was meant to be the exclusive stressor in this setup, it is impossible to draw a conclusion which actually was the predominant cause for the loss of contractility force. Presumably, the sound pressure alone did not exert any impact at all.
The inadequacy of the physical conditions yielding void results could have been avoided by using a hermetically closed vessel (airtight pressure chamber), in which all experiments related to dynamic pressure at very low frequencies, including the calibration of pressure sensors themselves, are conducted.[5] It allows establishing a high frequency-independent sound pressure with only moderate loudspeaker cone excursion, suppressing the excessive particle velocity of an open setup. It is doubtful that the muscle strips would show any reaction at all under correct biophysical conditions.
Suggestion for real-world heart debilitation experimentA valid experimental design to verify the direct influence of airborne infrasound on the heart (and all other organs) could be to put volunteers in an airtight chamber[6] and let their whole body be exposed to infrasound generated by large woofers mounted in the chamber’s shell. Before and after the exposure, their maximal oxygen uptake VO2max could be determined using spirometry during an all-out effort on a bicycle ergometer. Any detrimental effect of the airborne infrasound could be easily identified by a decline in aerobic capacity in comparison to a control group. In order to allow for double-blinded conditions, the frequency should be lowered to less than 2 Hz to guarantee inaudibility, which even represents WT infrasound far better than the 16 Hz used by Chaban et al. However, the design and commissioning of a pressure chamber able to produce up to 120 dB of infrasound without revealing the presence of the signal (necessary for double-blinded conditions) by audible harmonics, rattling of loose parts or by turbulence noise at tiny leaks, is challenging.[7]
Pointless recommendation of maximum infrasound exposureChaban and his team could not detect a significant debilitation of the heart tissue at 100 dB SPL exposure even though their inadequate setup produces excessive particle velocity (and ensuing vibration of the tweezers fixing the samples) compared to a sound wave of the same SPL in the far field. Nevertheless, they claim that the maximum tolerable limit of infrasound pressure should be set to 80 dB for “chronic exposure.” While this arbitrary value is not warranted by their own results, it is also useless without specifying the frequency range and an appropriate frequency weighting for which it shall be valid. The ambient infrasound noise floor caused by natural sources starts to rise considerably below approximately 2 Hz even in only moderately windy weather conditions.[8] Besides wind gusts, ocean-generated microbaroms dominating the ambient noise spectrum between 0.1 Hz and 0.5 Hz [9],[10] and the infrasound emitted by seismic events [11] travel thousands of km overland. Looking at less-calm weather conditions, thunderstorms[12] and strong alpine downslope winds[13] can easily reach an infrasound SPL of 130 dB and more.
All this means that specifying a maximum infrasound level targeted to technical infrasound sources does not make sense without specifying frequency bounds and a suitable weighting filter, preferably the widely used G-weighting.[14],[15] Otherwise, the infrasound of the technical source to be measured simply drowns in the prevailing noise floor and the resulting overall measured SPL will basically reflect just the natural ambient noise.
Infrasound emissions of a mid-sized WT[Figure 4] provides an instructive example for machinery-generated infrasound in relation to the natural infrasound noise floor. The blue trace on the left side is a pressure signal captured for 30 minutes at high wind conditions at a 200 m distance from a mid-sized WT north of Hannover, Germany. The recording is an excerpt of raw microbarometer sample data (publicly available[16] and converted with an open-source tool[17] to standard text format) raised during a measurement campaign of the German BGR (Bundesanstalt für Geowissenschaften und Rohstoffe – Federal Institute for Geosciences and Natural Resources) during a measurement campaign in 2004.
Figure 3 Heart muscle tissue sample fixed by the tweezers of the “heart muscle investigation system,” lifted from the container with nourishing isotonic liquid after the experiment. Screenshot taken from documentary video [3]
Figure 4 (Left in blue) Half an hour of noise captured by the BGR with an MB2000 microbarometer at 100 Hz sampling rate in high wind conditions at 200 m distance of a Vestas V47 WT with its 660 kW generator active (26 rpm). (Right) Overall 256K-FFT spectrum (dark blue) and 1/3 octave SPLs (orange) of this signal. Green: Hearing threshold of the especially sensitive person (Lydolf[35]). Dashed light blue: Alec Salt´s hypothetical OHC threshold. Red: Isolated BPHs and close-in noise. (Left in red) Time signal of isolated BPHs obtained by IFFT. The pink bar at 10 Hz: G-weighed power sum of isolated BPHs (52.2 dB(G)). BPHs, blade passing harmonics; OHC, outer hair cells; SPL, sound pressure levelWTs produce gentle pressure pulses with the blade passing frequency (BPF) each time one of the three blades passes in front of the mast. However, in this recording, these tiny pulses are almost completely overpowered by the natural ambient infrasound noise floor, which exhibits a random structure with peaks up to almost 8 Pa. Signal processing is needed to make them visible.
As a first step, the 30-minutes segment was Fourier transformed to yield a high-resolution spectrum, shown logarithmically in amplitude and frequency on the right side of [Figure 4] in blue. The periodic pressure pulses of the blade passages buried in the time signal have now become apparent by a characteristic “signature” of some benign blade passing harmonics (BPH) with quickly diminishing amplitude toward higher integer multiples of the BPF. This indicates that the slew rate of the correspondent pulses is relatively slow – narrower pulses would produce a more extended overtone spectrum with less steeply declining amplitudes.
The third-octave SPL, which result from summing up the power of all frequency components within each band, are shown as orange staircases. The green line in the upper right part is the hearing threshold of a specifically infrasound-sensitive person determined by Lydolf and presented by Møller.[18] It can be considered a threshold below which the vast majority of people cannot perceive anything. The dashed light blue line is a theoretical threshold for the neuronal activity of the outer hair cells (OHC) in the cochlea as postulated by Salt and Kaltenbach.[19] The OHC is an active part of the “cochlear amplifier”[20] but does not directly contribute to the hearing sensation. All infrasound third-octave SPLs in [Figure 4] keep well below both the Lydolf threshold and Salt’s self-defined OHC activity limit.
The BPHs and their adjacent close-in noise were extracted into a separate trace (red in [Figure 4]). As already shown in other studies,[21]-[23] the energetic sum of this spectrum yields only a moderate 65 dB. In contrast, the overall SPL from a few mHz up to 50 Hz (blue curve) is 99.5 dB, a value almost entirely dominated by the rising ambient noise floor below 2 Hz. The WT itself contributes less than a 0.02 dB rise to this overall infrasound SPL. Thus, it would be indistinguishable whether it is operational or not (yet always be blamed for infringing the 80 dB limit) when only considering the unbound infrasound SPL up to 20 Hz.
In a further step, the isolated BPHs (red segments) were back-transformed to the time domain by IFFT (inverse fast Fourier transform) and plotted (in red, to scale) over the original unfiltered pressure signal (in blue). As can be seen, the amplitude of the resulting extracted pressure pulses is almost negligible relative to the overall infrasound signal. To view individual pulses, a 10-second segment of the 30-minute recording was selected for [Figure 5]. With the amplitude scale set to the same ±8 Pa as in [Figure 4], necessary to represent the broadband pressure signal unclipped, the diminutive infrasound pulses created by the blade passages are now barely visible as tiny spikes superimposed to the large pressure fluctuations (blue trace in the left graph of [Figure 5]). Only isolating their frequency components surgically as demonstrated here (or at least applying a higher order band-pass filter between 1 and 10 Hz) and then zooming the amplitude scale for the resulting residual signal by a factor of 40 (right graph in [Figure 5]) allows to view them comfortably.
Figure 5 Ten seconds extracted from one of the most energetic sections of the 30-minutes recording in Figure 4. (Left) With the same amplitude scale as in Figure 4. (Right) 40× amplitude magnification of the filtered BPH-only signal to render the tiny pressure pulses created by the blade-tower interaction visible. BPH, blade passing harmonicsTheir frequency is similar and their peak amplitude of about ±0.1 Pa is not higher than the benign atmospheric pressure fluctuations occurring at the eardrums when walking, induced by the diminutive differences of the altitude when moving the head up and down,[25] as given by the barometric formula. A change of just ±1 cm in altitude corresponds to an atmospheric pressure change of approximately ±0.12 Pa – equivalent to a peak SPL of more than 75 dB.
The overall BPH level of 65 dB and the distance at which it was measured also allow for a rough estimation of the total infrasound power emitted by the WT. In a first step, the infrasound intensity can be calculated from the infrasound pressure. It is a measure of the sound power that impinges on 1 m2 perpendicular to the wave propagation direction and is given by the following equation:
I is the intensity, p the pressure, v the particle velocity, and Z the acoustic impedance. In a second step, the total radiated power P can be obtained in a similar way as described in IEC 61400-11 (wind turbines–acoustic noise measurement techniques)[26] by integration of the intensity over a semisphere whose radius r is equal to the distance, assuming that the ground is fully reflective and that the radiation is perfectly omnidirectional (which of course is not rigorously true):
This relatively low infrasound power of less than 1W is in accordance with the moderate unweighted SPL. WTs are no strong infrasound sources. According to Møller,[18] their emitted sound power is slightly over-proportional to their nominal electric power (10 dB increase in nominal electric power corresponds to an increase of 11 dB in sound power), so even the biggest turbines with several megawatts of output power capacity should emit infrasound power only in the one-digit watt range. Moreover, as they are rotating slower, their BPH spectrum is shifted to lower frequencies, where the perception threshold is even higher.
The link between the presented WT and the experimentsInterestingly, the particular measurement presented in [Figure 4] and [Figure 5] has major relevance for the work of Chaban et al. It is based on the measurement campaign of the BGR at the WT near Hannover in 2004. More than 10 years later, they published their results after using Matlab to analyze the data.[30] However, they inadvertently committed a dramatic programming error in an equation to convert the power spectral density (the preferred type for seismology) to SPL spectra. This error, discovered and disclosed by Holzheu[21]and only reluctantly admitted by the BGR[24] after further public objections,[22],[23] led to an overestimation of 36.1 dB (factor 4096 for the sound power) of the SPL spectra, yielding a claimed infrasound SPL of 100 dB at 200 m distance of the investigated WT in the 30 minutes interval shown in [Figure 4] (filtered by the BGR with a 4th order Butterworth high-pass with 0.5 Hz cutoff to suppress the low-end noise).
The erroneous value of 100 dB was gladly adopted by organized community groups in Germany founded to combat the installation of WTs and also by C. Vahl, who is the supervisor and senior coauthor of the paper of Chaban et al. In a public talk[31] and TV and newspaper interviews,[32],[33] he disclosed that the idea to investigate the impact of WT infrasound on myocardial tissue was spurred by the desire to raise funds and by complaints about (audible) wind farm noise of a friend in northern Germany. The initial value of 100 dB in the experiment series was based on the (erroneous) publication of the BGR.[30]
On the plausibility of heart debilitation by infrasoundThe presumption that airborne infrasound generated by technical sources like WTs might debilitate the heart is special in two aspects. Firstly, bearing in mind that it is a vital organ whose failure leads to immediate death, any public insinuation that its health could be threatened by a noxious external interference is particularly alarming to most people. For the most fearful of those believing to suffer AHE under infrasound exposure, it is even able to evoke psychogenic responses with the potential to raise anxiety disorders such as cardiophobia. Secondly, the heart is an extremely rugged organ, beating over 2.5 billion times over an average lifetime (assuming an average of 60 bpm in 80 years). Each of these systolic-diastolic cycles entails considerable tissue strain and a large pressure difference in the bloodstream of about 40 mmHg at resting state (correspondent to a level of almost 163 dB peak re 20 μPa), rising to over 200 mm Hg (177 dB peak) during heavy exercise. Suggesting that heart muscle motility might be compromised by external sound pressure several thousand times weaker than the dynamic pressure produced by itself constitutes a bold thesis.
It is actually inconceivable that an in vivo heart would suffer any weakening from infrasound up to 120 dB even at prolonged exposure. Such a sensibility would have been noticed long before in everyday clinical practice. Industrial or agricultural[27] workers operating heavy machines, truck drivers,[28] and ravers at electronic dance music festivals would all suffer from acute cardiomyopathies. Likewise, swimmers would suffer hypoxemia and drown after a short time. The vertical displacement range of the upper body of some Olympic swimmers easily reaches 30 cm in their breaststroke cycle,[29] entailing a dynamic pressure amplitude in the water (as given by the hydrostatic pressure equation p = ρ g h) of over 1 kPa RMS (root mean square). This corresponds to an SPL of 154 dB, which effectively acts upon the chest – and, quite interestingly, with a frequency that corresponds well with the BPF of modern megawatt WTs. The dose (i.e., energy) that resulted in an 18% decrease of myocardial contractility (1 hour at 120 dB) in the experiments of Chaban et al. is reached after just one breaststroke cycle. No medical complications are known from the literature for any of the aforementioned conditions. It can thus be assumed that Chaban et al. identified an artificial problem that doesn’t exist in reality.
Nevertheless, after the outcome of the experiments, the senior coauthor of Chaban et al. alleged publicly[2],[31],[32] that WTs constitute “jammers for the heart,” despite the fact that the work of his research group had clearly shown that the real infrasound level of only 65 dB (measured in high wind conditions at 200 m distance) of the WT that predefined their experiments is completely innocuous for the heart health. The 65 dB are 35 dB below 100 dB, which was the lowest level applied in their experiments. No significant effect was observed at 100 dB, even though undue vibration was mechanically coupled into the muscle tissue.
Many natural sources [34] and motor-driven appliances expose us with much higher infrasound levels, as can be seen in [Figure 6]. Particularly, driving in a conventional car produces infrasound often above 100 dB,[36],[37] obviously without causing heart pathology. Enforcing a general limit of 80 dB for chronic exposure would unequivocally mean that vehicle traffic should be prohibited. Time would probably tell that the percentage of heart diseases among the population would indeed decrease considerably in this case, however by promoting healthy physical activity for the cardiovascular system, such as commuting by bike, rather than avoiding infrasound exposure.
Figure 6 Comparison of unweighted 1/3 octave levels when traveling inside a car (dark blue), a regional diesel railcar (brown), a high-speed ICE train (red), and an IC train in a railway tunnel (violet). Orange: Worst case in 48 days of E115-3.0 turbine in 500 m distance recorded with MB2005 microbarometer. Green: Hearing threshold of the especially infrasound-sensitive person (Lydolf[35]). Dashed light blue: Salt´s hypothetical OHC threshold. Blue: Hearing threshold according to Møller/Pedersen.[35] Levels in the headline are unweighted but restricted to the band 0.5 to 20 Hz. OHC, outer hair cellsOwing to its considerable wavelength, the infrasound acts perpendicularly as a gradual and almost perfectly homogeneous fluctuation of the atmospheric pressure on all sides of our body, which in turn causes the same slow periodic variation of the quasi-hydrostatic pressure inside the body. As blood, tissue, and bones are almost incompressible compared to air, very high sound intensity would be needed to put their molecules in notable motion. Moreover, owing to the huge impedance mismatch (approximately by a factor of 3500) at the air-to-water interface,[38] airborne infrasound intensity is reflected from or diffracted around the body to over 99.9%. For infrasound, which entails the same dynamic pressure outside and inside the body, the intensity is attenuated to 1/3500 of the incident airborne sound intensity, turning a couple of μW/m2 into a few nW/m2 inside the body. It is intuitively obvious to anyone with at least a modest capability of appreciation of physical magnitudes that such a low intensity is not capable to cause any harm whatsoever at the molecular, cellular, or organic level. An example of the sound intensities that our body is actually capable to endure without harm is now given.
Comparison with intensities in ultrasonographySound waves share with electromagnetic waves the property that only extreme amplitudes or very short wavelengths are capable to cause harm at the cellular or molecular level, so ultrasound is potentially more aggressive than infrasound. Indeed, ultrasound is used for cleaning, gluing, welding, and even destroying kidney stones. As deleterious effects of WT infrasound at the cellular level are asserted in some studies,[2],[31],[32] it is illustrative to compare it to the intensities deployed in an application rarely blamed for health issues: ultrasonography, considered to be the safest of all imaging techniques to examine the interior of the body.[41]-[44] The intensities of diagnostic ultrasound are much lower than those in the aforementioned technical applications; however, they are still considerable. In contrast to the airborne sound intensity, which is kept out of the body to 99.9% due to the impedance mismatch at the air-water interface, the solid piezoelectric transducer of an ultrasound imaging device, aided by water-based gel applied on the skin, effectively couples almost all acoustic power into the soft tissue where it partially dissipates, causing a local elevation of the temperature or, at higher intensities, even cavitation.
As detailed in Annex B, the typical average intensity ISPTA (spatial peak, temporal average) in the focus of the transducer during ultrasound imaging is around 100 mW/cm2, considered safe even for extended exposure time.[41],[44] This is over 300 million times (equivalent to +85 dB) the infrasound intensity of the typical 3 μW/m2 impinging on the skin of a person at a 200 m distance of the mid-sized WT at high wind presented in [Figure 4]. To put this astronomical difference in an astronomic context, the illuminance for a clear moonless night with the stars and the natural airglow as the only light sources is of the order of 0.001 lux.[39] In contrast, at noon, the unobscured sun provides illumination of up to 120.000 lux on the Earth’s surface.[40] The difference between the two intensities is around 81 dB, similar to the difference between the intensities of diagnostic ultrasound inside the body and WT infrasound on the skin.
Another interesting observation is that 100 mW/cm2 is 30 dB or 1000 times above the intensity of the highest infrasound exposure (120 dB) that Chaban et al. used in their inappropriate setup, claiming a decrease of heart muscle contractility of 18% after 1 hour. No such effect has ever been reported from sonography. This is another indication that the artificial sound exposure of isolated muscle fibres in the near field of a woofer does not quite faithfully represent the situation of the in vivo heart.
When looking at the doses, in less than 10 seconds, the region at the focus of the transducer receives more sound energy per unit area than would impinge on the skin from the WT in continuous high wind operation during over 100 years. Inside the body, this energy is further reduced by a factor of 3500, as given by the acoustic impedance relationship between water and air. This means that in less than 1 second, the cells at the focus of the transducer receive more energy than by the WT in 36,000 years.
Of course, the high intensity of the ultrasound beam is concentrated to only a small volume; however, as Vahl publicly[31] hypothesized that the damage to the heart occurs at the cell level, interfering with the myosin- and-actin-based contractions of the muscle filaments, the comparison seems pertinent.
Another important aspect is that in contrast to the continuous BPHs of a WT and the continuous 16 Hz sine wave deployed by Chaban et al., the ultrasound emitted by the imaging systems in scan mode constitutes a periodically pulsed signal, which some practitioners worried about WT infrasound, including Vahl, claim to be particularly deleterious. The duty cycle (on time in relation to the pulse repetition period) of the gated ultrasound bursts is very low to provide enough time for the transducer to listen to the reflections, whose intensities and times-of-flight are used to build up the real-time image on the screen. This means that the ISPPA (spatial peak, pulse average intensity) is even considerably higher than the ISPTA of typically 100 mW/cm2. The on/off transitions and non-linear effects (intermodulation) with a pulse repetition frequency usually up to a few kHz also generate some audible signal power, strong enough to potentially be heard by the fetus in an obstetrical check-up.[47]-[49] Though harmless, even this relatively low leakage is higher than the inaudible exposure from WTs. While claiming that these constitute “jammers for the heart”, Chaban et al. presumably submit the hearts of their patients routinely to ultrasound imaging without even caring about any sanitary implications, secured by the excellent safety record[41]-[43] of echocardiography over many decades.
Dissection of the discussion section of Chaban et al.Most theories how WT infrasound is supposed to create AHE are restricted to aural pathways. References [51]–[57] give a good overview. The work of Chaban et al. is unique in claiming organic damage from infrasound levels normally considered safe. In their discussion section, Chaban et al. even generalize infrasound as a potential health threat to the whole body.
The general remarks start with the statement that infrasound is enabled “by means of reflection; refraction and diffraction, to pass through and around different obstacles.” However, while infrasound is in fact diffracted over noise barriers, it can be shielded fairly well by stiff closed building shells, especially by those of energy-efficient airtight dwellings.[61] In this context, the authors also fail to recognize that the human body itself constitutes a solid obstacle that refracts over 99.9% of the infrasound intensity (which is what defines the physical impact on organs and tissues) around it. Instead, they do not only claim that “the human body itself does not shield against infrasound,” but even contend that the human body may emphasize infrasound “by mean of resonance, as it has been shown that the upper human torso tends to resonate between 5 and 250 Hz”. They cite[62],[63] to back this allegation. However, neither reference deals with infrasound. As appositely described in[64], “the quoting of sources that are not related to infrasound is very common, even when the debate is about infrasound”. Indeed, reference[63] exclusively deals with vertical vibration induced into the feet of standing test persons and the resulting axial resonances of the torso. These resonance modes do not occur when the body is exposed to airborne infrasound, which acts homogeneously as a scalar from all sides. This is another indication that the authors are not aware of the difference between vibration (mechanically coupled translational force) and scalar sound pressure.
Reference[62] deals with ground-based ramp-up turbine tests of supersonic combat aircraft (including afterburner use) producing up to 147 dB SPL along a line parallel to and 12.8 m from the longitudinal centerline of the aircraft. Some peak acceleration was measured on the chest of the courageous test person in the 63 to 100 Hz bands; however, the authors admit that the strong and hot wind gusts of 30 to 40 knots emanating from the nearby turbine outlet and a blast deflector may have influenced the results. The conclusion of the paper is that “Infrasound occurring at 40 Hz and below did not appear to be a problem.”
Regarding the impact of airborne infrasound on humans, Chaban et al. contend that “many individuals, who describe a feeling of a drum in their entire body, easily perceive it.” However, the vibrotactile threshold for infrasound is even 15 to 20 dB higher than the perception threshold of the ears[65] and easily surpasses 120 dB below 10 Hz. This is at least 45 dB more than worst-case infrasound exposure in the close vicinity of WTs. Furthermore, the vibrotactile threshold defines just the onset of sensation – “a feeling of a drum” would require levels well in excess of 130 dB.
In a digression to room acoustics, Chaban et al. state that “it is also common for infrasound to generate high energetic standing waves in enclosed spaces, when the space dimensions are multiples of the half wavelength.” As shown in [Figure 4], the BPHs of the mid-sized V47 turbine vanish in the noise floor above 10 Hz, where half of the wavelength equals 17 m, more than encountered for most rooms in common dwellings. The maximum rotational speed of modern megawatt WTs is only half of that of the V47, shifting the BPHs to even lower frequencies.
Reference[58] cited by Chaban et al. only deals with low-frequency noise (LFN), not with infrasound. They also confound standing waves with Helmholtz resonances, though the difference is clarified in another paper they cited.[59] It clearly states that “the frequencies of the standing-wave room resonances are above the infrasonic range for residential dwellings.” Standing waves of BPHs in dwellings are thus a nonexistent problem. In fact, simultaneous measurements inside and outside dwellings regularly demonstrate that the infrasound inside the house is considerably attenuated.[60],[61].
In an effort to show that “there is plenty of evidence regarding the damaging effect of infrasound upon the heart,” Chaban et al. reviewed experiments of other research groups. They cite a number of studies in which small mammals, mostly rats, were exposed to prolonged abnormal sound levels of up to 130 dB, for example, in the experiments of Pei et al.[66] These studies, some of which with questionable designs (including the absence of a pressure chamber), are often not only using infrasound, but also low frequency or simply broadband noise, leading to strong stress reactions by the extremely high audible noise the rodents were submitted to.
From these experiments with excessive SPL, Chaban et al. draw the invalid conclusion that “it seems reasonable to attribute some complaints about wind turbines to the infrasound radiated by them.” However, measurements in the vicinity of wind parks invariably show that the BPHs produced by the blade passage are deeply below all known perception thresholds. Neither laboratory experiments with infrasound levels up to 25 dB above WT infrasound exposure[67] nor epidemiological studies among residents[69] back their assertion.
Finally, Chaban et al. claim that their “recommendation is similar to the 85 dB(G) level recommended by the Danish Environmental Protection Agency in 1997.”[15],[68] It is not. The difference is that the 80 dB proposal of Chaban et al. does not involve any frequency weighting, while the Danish recommendation of 85 dB(G) applies the psychoacoustic G-weighting, attenuating frequency components below 10 Hz with a 12 dB/octave decrease toward lower frequencies. The G-weighting corresponds well with the increasing perception threshold toward lower frequencies.
Linear-weighted levels and G-weighted levels are not compatible and are thus not comparable. The difference between them depends on the frequency composition of the analyzed signal. The lower the frequency, the more the component will be attenuated by the G-weighting, which correctly represents the strongly decreasing sensitivity of the ear and is also valid for the hypothetical OHC activation threshold self-defined to 60 dB(G) by Salt.[19]
When summing up the BPHs in [Figure 4], which are the frequency components held responsible for AHE by Chaban et al., the unweighted result is 65 dB, well below Chaban’s proposed 80 dB limit, and the G-weighted sum of 52 dB(G) is more than 30 dB below the conservative Danish recommendation. The average perception threshold for infrasound is approximately 95 dB(G). The Danish limit of 85 dB(G) thus includes a 10 dB safety margin to also cover very sensitive persons such as the one presented by Lydolf.[35] With a G-weighted level more than 30 dB below this threshold, it can be safely deduced that even worst-case infrasound exposure emanating from the blade passage of WT has no impact on humans.
ConclusionsThe results of the paper of Chaban et al. are invalidated by a physically inappropriate setup that produces excessive air movement, not present in a regular sound wave of the same sound pressure in the far field. This strong air movement causes the tips of the tweezers to vibrate, exposing the affixed muscle tissue samples to localized mechanical irritation, which is not present in the in vivo heart. The unsuitable setup is a reminder that in interdisciplinary experiments, it is important to consider the expertise of specialists in each involved area to avoid methodological inadequacy. The outcome of the experiments that low-frequency sound of merely 120 dB causes a contractility decrease of 18% after only 1 hour of exposure is disproved by clinical and everyday evidence.
The proposed general limit of 80 dB for “chronic infrasound exposure” is not backed by the results of the research group and, without specifying frequency bounds and weighting, is useless and misleading. It would involve the prohibition of most large machinery – including those used for transport on land, water, and in the air. WTs, which initially motivated the investigation of Chaban et al. and at which their proposed infrasound SPL limit was actually aimed at, would however not be covered by this ban – even when summing up in large wind parks, their infrasound exposure is regularly well below the proposed limit and any other recommendations and thresholds of perception published in the past 50 years.
Annexure A – some basics about sound wavesSimilar to electromagnetic waves, which are composed of an electric and a magnetic component, sound waves are also composed of two components: dynamic pressure p and particle velocity v. In the far field, the particle velocity is proportional to the sound pressure and in phase with it.[70],[71] They are linked by the acoustic impedance Z, which is the product of the air sound velocity c (approximately 340 m/s) and the air density ρ (approximately 1.2 kg/m3, both c and ρ vary with meteorological conditions and altitude):
In the far field of a spherical sound wave, both sound pressure and particle velocity are inversely proportional to the distance r of the sound source (assuming the source is small relative to the wavelength):
with p0 = sound pressure at reference distance r0
If, for instance, the distance is cut to half, both quantities double. Hence, the sound intensity, which is defined as the product of sound pressure and particle velocity, quadruples. The SPL, which is defined by
increases by 6 dB. The situation changes when further approaching the sound source and entering into the near field zone, where the distance r is considerably smaller than the wavelength λ = c/f (over 21 m at the frequency of 16 Hz used in the experiments of Chaban et al.). Here, the velocity starts to exhibit a dependence that is inversely proportional to the square of the distance due to the second term inside the parenthesis:
with i = imaginary unit and wave number
At the same time, the velocity suffers a phase shift with respect to the pressure that reaches almost 90° at the surface of the sound source, where the second term prevails. The acoustic impedance and the sound intensity near the source carry an overwhelmingly strong reactive component. This means that large amounts of air are moved forward and back, however, without building up significant pressure or transferring considerable acoustic power to the emanating sound wave. In other words, a loudspeaker excited with very low frequencies, corresponding to sound wavelengths of many meters, is essentially producing periodic local wind in front of the loudspeaker cone without creating significant sound in the far field. Objects located in the near field are nevertheless affected by the strongly reactive sound field, being subject to considerable vibration, as shown by Chaban with a strongly vibrating sheet of paper in front of his experimental setup.
Annexure B – safety criterions for ultrasound diagnosticsOwing to the quest for higher resolution and repetition rate, the applied intensities in ultrasound diagnostics have risen considerably in the past decades, especially for the B-scan mode, whereas they always used to be high for special modes like color Doppler. The American FDA (Food and Drug Administration) has set an upper limit for the ISPTA (this is the average intensity at the focus) of 720 mW/cm2 for devices to be commercialized in the United States. It has become the de facto world standard. In 1992, even the permissible intensity for the examination of fetuses was raised to this limit.[48]
The only caveats concerning ultrasound exposure in the body are rising tissue temperature (hardly a problem in B-scan mode due to the low duty cycle of the repeated bursts) due to absorption and the risk of inducing cavitation.[44],[45] These risks are covered by two parameters displayed in real time on most devices, the TI (thermal index) and the MI (mechanical index). The TI is the ratio of the actual intensity to the intensity that would raise the temperature of the tissue by 1°C and the MI the relative risk to induce cavitation. While practitioners are cautioned to adhere to the “ALARA” principle (as low as reasonably achievable), a TI of 2 is generally considered safe for adults.[46] Typical ISPTA values in an examination are around 100 mW/cm2, which is considered safe for infinite application time.[41] For focussed ultrasound, even 1 W/cm2 is not considered critical as long as the exposure does not exceed 50 seconds. For the typical scan modes, the ultrasound signal is pulsed with a small duty cycle, yielding even considerably higher values for the ISPPA (spatial peak, pulse average intensity) in relation to the ISPTA.
References
留言 (0)