NHMRC Public Consultations

Skip Navigation and go to Content
Visit NHMRC website

NHMRC Draft Information Paper: Evidence on Wind Farms and Human Health submission

This submission reflects the views of
Organisation Name: 
Hatch Pty Ltd
Personal Details
Q1. Is the draft Information Paper presented and written in a manner that is easy to understand?: 

{offline submission form entered into portal by NHMRC}


Q2. Does the draft Information Paper clearly outline how the evidence was reviewed and interpreted by NHMRC?: 


Q4. Is there additional evidence on the likely level of exposure to emissions produced by wind farms at nearby residences? : 

A technical journal article was provided in Acoustics Australia Vol. 40 April (2012) No. 1 p 32- 34. The paper was a technical note and was peer reviewed, with the title “Low frequency, infrasound and amplitude modulation noise from wind turbines – some recent findings

It was a review of other papers obtained up until 2012 and considered, amongst other things, the: 

  • frequency spectra of noise emission from wind turbines at different distances, and 
  • the human sensitivity of sound perception in the low frequency region.

The information presented identified that for the wind turbines measured, the low frequency and infrasound levels at distances greater than typically 200m are below the threshold of sensitivity of humans at those frequencies. The paper also presented sound frequency spectra levels from other sources of low frequency sound which have higher values and greater exposed populations (road and rail traffic noise and surf noise) but no similar claims of annoyance.

Some of the issues claimed by those annoyed by wind farm noise relate to low frequency or infrasound not well measured by current parameters for environmental noise – statistical averaged A-weighted or C-weighted sound levels. (dBA or dBC). To further identify if there are physical phenomena that can be measured related to annoyance, some further research could be done. This area is being researched by Health Canada with a report due later in 2014. See http://www.hc-sc.gc.ca/ewhsemt/consult/_2013/wind_turbine-eoliennes/research_recherche-eng.php

The research efforts for low frequency and infrasound effects could be addressed at a number of points in the effects chain – from: 

  • the sources and their emissions, through 
  • sound propagation through the atmosphere and modeling of that, and 
  • human perception and annoyance of the sound spectra produced by wind farms and similar low frequency sources of noise.

This is commonly considered in sound science as “source – path – receiver”.

For the source aspect, sound levels at different distances from wind turbines are relatively well known and can be measured. However there may be a need for additional measurements of the spectra at low frequency and infrasound to better determine their range, emission level and variability, and perhaps suitable parameters for those measurements.

Source spectrum sound levels could also be measured at increasing distances from wind turbines and wind farms to provide comparison levels with human perception and annoyance criteria in those low frequency ranges. As a part of the comparison, similar frequency range measurements should also be included for other known sources of low frequency sound; this is considered an important component to determine if there is any temporal or frequency spectrum difference between the types of received source sounds which might relate to annoyance.

Measurement of sound in the low frequency range has been done for wind turbines but it is considered to have not been for longer periods of weeks or months as typically done for long-term environmental noise assessments. This is another area of research that is considered to be necessary to determine the variability of received emissions.

For new or proposed developments of wind farms (and other industry) in Australia, the approvals regime requires prediction of sound levels at residential receiver locations. These predictions are made using proprietary or publicly available computer noise propagation models. The accuracy of these models varies and their verification to Australian conditions and typical wind farm to receiver arrangements is considered in need of improvement. Low frequency spectrum predictions is also an area yet to receive what is considered to be an appropriate level of investigation – this is only necessary if low frequency sound is identified as being an issue from the types of studies discussed above, as the low frequency spectrum content of wind turbines is relatively lower than that from the mid frequencies of their spectra. Some work has been done with two different models for typical wind farm layouts. The range of accuracy demonstrated in those studies is below what is considered to be preferable for a reasonable assessment of impact to be made. Therefore further research into accuracy of a wider range of propagation models is recommended. This research could be done for operating wind farms and the predictions should cover the full range of meteorological conditions that occur over the long term, to compare with actual measurements.

The final part of the additional research recommended would be into the human response to wind farm sound including at low frequencies and infrasound parts of the sound frequency spectrum. There has been significant work done by others into the threshold of perception of sound at low and infrasound frequencies and some of these were referred to in the journal paper described earlier. Sound from wind farms is noted above to occur at sound levels below these thresholds at typical distances of residential receivers, yet some respondents claim annoyance and health effects at these. No evidence has yet been identified that exposure to sound at levels below perception can cause measurable health effects. Further research to assess if there are measurable health effects of specific wind farm sound spectra at the predicted and measured sound levels would demonstrate the probability or likelihood of its occurrence. This would be a difficult and expensive area to investigate. Some work would be required to identify how to test and measure – both in testing exposure levels and what health effect responses to measure.

Page reviewed: 11 February, 2015