Why did the EPA perform an extensive monitoring program at Waterloo wind farm?
There are multiple wind farm sites across South Australia. The largest number of reported complaints regarding noise effects has been associated with Waterloo wind farm, although post-construction monitoring had not provided evidence for noise that did not meet the EPA Wind Farm Environment Noise Guidelines.
Given the concerns of the community in the region, the EPA decided to conduct a comprehensive noise monitoring program to find whether a scientific basis exists for the described effects.
What were residents’ concerns?
Among a range of concerns were noise, infrasound, health, effects on livestock and wildlife; and impacts on visual amenity.
How long did the study go for?
Noise equipment was first deployed in early April. Monitoring began from the middle of April to the middle of June for a total of approximately 10 weeks.
Who participated in the study?
Six houses were selected from a list of volunteers to host noise equipment based on an assessment of optimum distances and directions from the wind farm for monitoring.
However, other residents were also invited to participate in the study to complete weekly noise diaries, to provide information on noise events and other information that they wished to contribute for the project.
How close did participants live from the wind farm?
The houses hosting the equipment were between 1.3 km and 7.6 km respectively from the closest wind turbines.
The houses of other diary participants were between about 1 km and 11.5 km respectively from the closest turbines.
How big is the Waterloo Wind Farm, how many turbines are there?
The Waterloo Wind Farm is located in the Mid North region of South Australia. It is situated atop a north-south ridge stretching for 18 km. It comprises 37 Vestas V90 3-megawatt wind turbine generators with a hub height of 80 m. The entire site has a rated generation capacity of 111 Megawatts.
How does the monitoring equipment work?
The monitoring was performed using instruments that automatically acquired noise data acquisition and audio records. These noise instruments were supported by continuous weather monitors at all but 1 site. Five of the 6 stations were accessible remotely via mobile phone connections; while the remaining station needed to be serviced manually.
Did the EPA investigate noise inside and outside the houses?
Yes, the EPA deployed microphones both inside and outside the houses at 5 locations. At the 6th location, 7.6 km from the wind farm, only outside measurements were taken.
Why did the EPA do inside and outside measurements, rather than monitoring at more locations?
In order to get the best results, the EPA wanted to concentrate the limited suite of equipment to get the most comprehensive data at locations selected. This approach provided a better understanding of the relationship between outdoor noise levels and those inside houses..
Did the EPA analyse health effects?
No. The aim of the study was to investigate whether a physical, scientific basis could be shown for descriptions of noise by residents near the Waterloo Wind Farm, so the focus of available resources was specifically directed to answering that question.
However, community members were entirely free to provide information through the diary returns on the effects they felt during the study period, and any other factors that they considered important; and many have done so.
Proper evaluation of these health-based descriptions was considered to fall within the purview of health authorities, so EPA has always planned to refer the information to relevant health authorities as soon as practical.
What is the EPA’s approach for analysis of the reported effects?
The EPA utilised many data points collected from the monitoring places and the wind farm site.
These include acoustical parameters and weather parameters at the houses; and weather and operating information from the turbines themselves.
Diary entries were used to focus analyses of audio records , but other periods of possible noise from the wind farm were also examined.
The wind farm operator also arranged 6 full shutdowns, to assist in identifying wind farm contributions from background noise and other noise sources.
What weightings were used (and how do they work)?
Different weightings have been applied to the measured noise, to reflect the typical response of the human hearing system.
The most commonly used weighting is the A-weighting, which relates to audible noise.
C-weighting emphasises the importance of low frequencies, that characterise some sources such as wind farms.
G-weighting is used when measuring infrasound levels.
What is the difference between audio noise, low frequency noise and infrasound?
The human ear has been tested to be most sensitive to noise between 20Hz to 20kHz. Sounds between these frequencies are known as the Audio Noise Spectrum.
Low frequency noise exhibits significant components in the low frequency spectrum. These spectra normally consist of noise below 250Hz.
Infrasound is noise whose frequency spectrum lies mainly below the 20Hz frequency. Infrasound noise is generally not audible to a typical listener unless at very high levels.
Did the EPA explore infrasound levels from the wind farm?
Yes, the EPA deployed special instrumentation at 2 monitoring sites to measure infrasound inside and outside the houses down to a frequency of 0.25Hz.
The blade pass frequency component is also within the infrasound frequency span and was analysed. This relates to the speed at which the turbine blades rotate.
Were low frequency levels investigated at the monitoring locations?
Instruments at 2 sites were capable of acquiring data at very low frequencies including infrasound.
Monitoring equipment at three other locations had extended low frequency capacity down to at least 12.5Hz. The data was checked against relevant low frequency criteria.
Was noise measured both with the wind farm operating and when the turbines were shut down?
On 6 occasions when specific environmental conditions established by the EPA were met, the wind farm was shut down by the operators for about 50 minutes each time. The shutdowns were chosen to represent active power generation by the facility, so that any differences in noise character and levels between operational and shut down periods could be distinguished.
The 6 shutdowns gave a good spread of wind directions, so that analysis of downwind conditions (expected to provide the 'worst case' noise conditions) at most sites was possible.
Did the EPA explore some of the reported noise effects from the wind farm like thumping, rumbling and swishing?
The EPA was careful in examining all data relevant to the wind farm. Audio records and acoustic descriptors were analysed for potential presence of noise characters that were described by the residents in the diary entries.
A rumbling noise was detected at some sites under particular conditions, but only after amplification of the audio records.
Were weather conditions recorded at the time of measuring noise to determine if noise was measured during worst case weather conditions for noise transmission?
Except for 1 location, all locations were equipped with weather sensing equipments that would record weather data simultaneously to the noise recording equipment.
Was vibration measured at residents’ houses?
No, vibration testing was not a part of the scope of this project. The monitoring locations are situated at high separation distances from the wind farm.
What are the major outcomes of the study?
Comprehensive analysis of data was performed to detail noise impact from Waterloo Wind Farm. The findings are now available.
Do the results cover other wind farms?
The results found in this report are specific to the Waterloo Wind Farm and should not necessarily be considered as applicable to other wind farms.
Does the wind farm comply with the EPA’s guidelines for wind farms?
Based on this study, wind farm complies with the EPA guidelines for wind farms.
From the study outcomes, were any changes made to the operation of the wind farm, or the EPA’s guidelines?
In this study, the wind farm could be detected at levels that were compliant to the EPA wind farm guidelines. There is no indication that the guidelines need to be changed on the basis of this study.
However, it is noted that the National Health and Medical Research Council (NHMRC) is investigating evidence for health impacts from wind farms and Guidance may be reviewed in the light of its findings.
Why was monitoring undertaken during autumn/winter and not summer?
There have been some views expressed that more noise problems exist in the region during summer periods because of less wind and lower generation rates.
Planning for the study commenced as early as possible once discussions with the community had taken place in December 2012 and January 2013, but the earliest that monitoring could commence was in April.
However, the monitoring period included a range of wind speeds favourable to commercial power generation , and covered the main directions. The combination of acoustic and weather monitoring, in combination with the six shut down periods during the study, has provided a good spread of conditions under which noise emissions from the wind farm could be evaluated.
Will there be a follow-up study?
No, the EPA will not be undertaking further studies of this type at Waterloo.
Will the EPA be sharing data with other researchers?
The EPA has made its intent clear from commencement to enter into information sharing with other organisations and researchers upon request, subject to any remaining privacy issues for equipment hosts.
Did the EPA receive any funding from the wind farm operators or other sources to conduct the study?
No, although the EPA obtained diary information from the community ,and supplementary weather and operational information from the wind farm, it remained fully independent from all parties in relation to the study design, acquisition of monitoring data, and scientific analysis and reporting, throughout all stages of this study.