Page 35 - Volume 9, Issue 3

that the relevance of the threshold to wind turbine infra- sound is not clear. Many natural and man-made infrasonic sources exceed the threshold in the higher infrasonic region (Turnbull, Turner et al. 2012)
Infrasound from wind turbines
An early association of wind turbines and infrasound was the work of NASA in the 1980s. Investigations of the MOD-1 and similar downwind turbines revealed pressure pulses from interaction between the blades and the disturbed flow behind the tower. Downwind turbines were largely experimental mod- els and were completely replaced by the three bladed upwind turbines, which make up the current operating fleet of utility scale turbines. Analysis of pulses from the MOD-1 and similar turbines, which typically have a repetition rate of 1Hz, leads to a harmonic series based on 1Hz and to the linking of infra- sound with wind turbines (Shepherd and Hubbard 1991). However, this infrasound is of the type which might be pro- duced by a single person hand clapping, or even a ticking clock! The problems from the MOD-1 turbine were not from the fre- quency, but from the peaks of the pressure pulses in the down- wind turbine design, which caused vibration of loose building components and were also audible. Building response is the same over a wide range of pulse repetition rates, up to the point where the decay time of the this response merges with the rep- etition rate of the pulses.
There have been a number of measurements of infra- sound from modern wind turbines (Hayes 2006, Hepburn 2006, Jung and Cheung 2008, O’Neal, Hellweg et al. 2011, Ambrose and Rand 2011 (December), Turnbull, Turner et al. 2012, Walker, Hessler et al. 2012, Evans 2013). Current meas- urements are often made to alleviate concerns of those objec- tors who believe that infrasound from wind turbines is harm- ful. Measurements show similar results. At typical nearest residential distances, the one-third octave level at 10Hz is around 60dB, with a negative spectrum slope of 3 to 6dB per octave. The levels decrease with distance and may merge into background infrasound. Evans showed that the averaged infrasound levels at residences 1.8km and 2.7km from the nearest turbine of a 140, 3MW turbine wind farm were simi- lar when the turbines were on or off (Evans 2013). This con-
firms earlier work (Guldberg 2012, Howe, McCabe et al. 2012). Although the average infrasound level may not differ between wind turbines on and off, the characteristics of the sound may change when the turbines are operating. For example, the inaudible infrasound close to a wind turbine may have cyclic variations, whilst the inaudible background infrasound has random variations in level.
There is no evidence that the low levels of infrasound from wind turbines, as shown by these measurements, are harmful to humans.
A narrow band analysis of noise from the Shirley wind farm at a residence 335m from the nearest turbine is in Fig. 4 (Walker, Hessler et al. 2012). The outdoors spectrum is 38- 39dB at 10Hz, 0.05Hz band, leading to a one-third octave level of about 55dB. The slope of the outdoors spectrum is close to 20dB/decade (6dB/octave). The rise in the living room spectrum in the region of 20Hz is from building reso- nances, but is nearly 50dB below the hearing threshold at 20Hz. The residents had left their home, complaining of ill- ness (nausea) caused by wind turbines, although all the levels in Fig. 4 are below the Salt OHC threshold of Fig. 3.
One of the authors of the Shirley report suggested direct action of infrasound on the vestibular otoliths as a cause of illness, but the next section on infrasound and the ear shows that this is an unlikely explanation.
Infrasound and the ear
The pure tone hearing threshold has been measured in a chamber down to 4Hz (Watanabe and Møller 1990) and to lower frequencies using earphones (Yeowart and Evans 1974). The chamber data is shown in Fig. 5, where it is com- bined with the ISO standard threshold (ISO:226 2003). The Watanabe and Møller threshold at 4Hz is 107dB. At 12 Hz it is about 90dB. Yeowart and Evans give higher binaural thresholds: 112dB at 4Hz and 121dB at 2Hz.
The mechanism of hearing down into low frequencies is through normal excitation of the auditory cortex, as shown by fMRI investigations (Dommes, Bauknecht et al. 2009 ). Dommes, Bauknecht et al used functional Magnetic Resonance Imaging (fMRI) to investigate responses of the brain when exposed to infrasound both above and below the hearing threshold, at the following frequencies and levels:
Freq Hz 500 48 36 12 12 12
Level dB 105 100 70 120 110 90
Audible infrasound excited the auditory cortex, which is where hearing perception occurs. Inaudible infrasound did not show an excitation. This is to be expected if infrasound enters into the hearing system, and is transmitted to the brain in a similar manner to higher frequency sounds. Dommes, Bauknecht et al summarise the results of their work as:
In our study, no other cortical regions owed a comparably extensive response to the high-level stimuli as did the auditory cortex, indicating that LFT [low frequency tones] were mainly perceived via acoustic pathways instead of representing a
Fig. 3. Comparison of inner and outer hair cell thresholds. Above: Inner hair cell. Below: Outer hair cell.
34 Acoustics Today, July 2013