Commentary for ERO …

Commentary for ERO (Environmental Registry of Ontario) number 013-3800
Amendments to the Renewable Energy Approvals Regulation
(Ontario Regulation 359/09)

November 2, 2018

This comment deals with the noise impact from wind turbine projects.

The Ministry comments on this proposal, posted on the ERO webpage, include:

“Renewable energy approvals and the environment
Renewable energy projects (as with all construction and human activity) may impact the environment. Projects could impact:
natural heritage (e.g. (for example)animals, birds, insects, trees, and vegetation)
noise
groundwater
surface water.
We require project developers to reduce these effects through the renewable energy approvals process so that the benefits of the energy outweigh the impacts to the environment.”

Project developers are required to demonstrate compliance to noise levels through a predictive, pre-approval noise assessment report, i.e. a computer model, and a post-construction noise compliance audit, i.e. a field audit.

There are shortcomings with both the predictive noise assessment methodology and with the current version of the 2017 “Compliance Protocol for Wind Turbine Noise” (Compliance Protocol) for field noise audits.

1) The methodology used to prepare the pre-construction noise impact assessment report that models the noise at the various “receptors”, (the ISO 9613-2), has several limitations that result in understated noise at those locations.

2) The methodology used to measure and confirm the turbine’s actual noise emission level, (IEC 61400-11 and the identical CAN/CSA 61400-11); and the immission noise at the receptors (the Compliance Protocol), have a number of conditions/requirements that result in some of the highest noise events being discarded from the analysis.

1) PRE-APPROVAL Modelled Noise, the Noise Impact Assessment REPORT

1.1 The noise modelling is a desktop computer calculation, (CadnaA), that predicts the noise at receptors based on the turbine’s noise “emission” provided by the manufacturers, and the setback distances to turbines around each receptor.

The CadnaA calculations are based on the ISO 9613-2 protocol that was published in 1996, and have several acknowledged limitations. Some of the more problematic limitations are:
* The ISO document states that the protocol is accurate for noise sources with a height up to 30 metres, and for modelling noise up to distance of 1,000 metres. Most industrial wind turbines deployed in Ontario have a rotor height of 80 to 135 metres.
* The ISO document states that the results have a standard deviation of 3 dBA. That means that for an average modelled noise of 40 dBA, 95% of the data (2 sigma) will fall between 34 and 46 dBA, yet all noise assessment reports are based on the average calculated noise with no consideration for the uncertainty in the calculation.
* The ISO document does nor require adjustments for a turbulent wind profile.

The ISO protocol has other deficiencies in addition to the uncertainty.

At a presentation[1] delivered by Erik Sloth, an engineer employed by the Danish turbine manufacturer, Vestas, the noise modelled with the ISO 9613-2 protocol was shown to understate the measured noise by up to 6 dBA at mid-tone frequencies. One slide from Mr. Sloth’s presentation, comparing results from ISO 9613-2 to another noise, Nord 2000 and field measurements, show the ISO 9613-2 can underestimate the noise between 200 Hz and 800 Hz by as much as 6 dBA at those frequencies; and that the Nord 2000 noise model is much closer to the measured noise.

Reference [1] Sloth, Erik; Problems related to the use of the existing noise measurement standards when predicting noise from wind turbines and wind farms.
https://waubrafoundation.org.au/wp-content/uploads/2014/01/AUSWEA-2004c…

1.2 The turbine manufacturers do not provide a noise specification for a turbulent wind profile, and turbulence will increase the turbine’s noise emission. This was acknowledged in Mr. Erik Sloth’s presentation referenced above.

The methodology used by turbine manufacturers to measure turbine emission noise is contained in the document IEC 61400-11 3rd Edition, and in the corresponding CAN/CSA 61400-11 3rd Edition, adopted without modification for use in Ontario. The IEC and CAN/CSA
documents contain an “informative” Appendix C that comments on turbulence, but corrections or adjustments for turbulence are not a requirement.

When the wind direction lines up parallel to the layout between the upwind and downwind turbines, the downwind turbines will driven by turbulent inflow from the upwind turbine blades when there is insufficient separation between the turbines. 

The most common recommendation for the minimum separation distance between turbines to minimize turbulent inflow is 7 times the rotor diameter; and some studies[2] have suggested that a separation distance of 15 times the rotor diameter is more appropriate.

Reference [2]. Johan Meyers and Charles Meneveau, Optimal turbine spacing in fully developed wind-farm boundary layers,
https://people.mech.kuleuven.be/~jmeyers/resources/downloads/WE2012_15_…

The authors of this study state:
“For optimal wind turbine spacing, the figure of merit that has been used here is the total power extracted for a given geostrophic wind velocity. Depending on the ratio of land-surface costs and turbine costs, different optimal spacings have been obtained. For realistic cost ratios, we find that the optimal average turbine spacing may be considerably higher (∼ 15D) then conventionally used in current wind-farm implementations (∼ 7D).”

If the proximity of the turbines to each other is going to create a wind profile that is turbulent. The manufacturer of the turbines should be required to provide a set of worst-case guaranteed sound power levels under the condition of the proposed site layout, including any effect of the hilly terrain. That sound power data should be used for the noise calculations.

Appendix F6 in the 2017 Compliance Protocol lists “estimates or measurements” of turbulence as only “optional” information for the turbine emission audit (E- audit) reports.

2) Measured noise from Turbines (EMISSION) or at receptors (IMMISSION) for compliance verification or complaint response

2.1 There is no acknowledgement or adjustment for cyclical or amplitude modulated noise (blade “swish” or “swoosh”), and turbulent inflow noise.

Prior to the 2008 edition of the Noise Guidelines for Wind Farms, there was a requirement to add a penalty of +5 dBA to any noise source that exhibited “cyclic” or amplitude modulated noise.[3] This requirement was removed in the 2008 guidelines, and is still absent from the 2017 Compliance Protocol for Wind Turbine Noise.

Reference [3] Ministry document NPC-104

The 2008 Noise Guidelines for Wind Farms includes comments on“swishing” noises (Section 6.4.8, second paragraph

“No special adjustments are necessary to address the variation in wind turbine sound level (swishing sound) due to the blade rotation See Section 4. The temporal characteristic is not dissimilar to the other sounds to which no adjustments are applied. It should be noted that the adjustments for special quality of sound described in Publication NPC -104 (Reference 1) were not designed to apply to sounds exhibiting such temporal characteristic.”

The comment that amplitude modulated noise from wind turbines is “temporal”, or infrequent, is not consistent with public comments made by several Ministry and industry experts.

In a 2010 memorandum[4], obtained through a freedom of information request, Mr. Cameron Hall, one of the Ministry’s senior environmental officers, stated:

“Most of the complainants who have contacted the Ministry about sound contamination from the Melancthon Ecopower Centre WTGs identify the characteristic "blade swoosh" or "swishing" sound contamination discharged into the natural environment from the WTGs as a quality of the WTG sound contamination which they find offensive. Provincial Officers have confirmed the "blade swoosh" quality of the sound contamination discharged into the natural environment from the WTGs throughout the Melancthon Ecopower Centre wind plant.”
(capitalized font is my emphasis)

and

“It should be noted that the more recent 2008 NPC Guidelines Interpretation differs from the 2004 NPC Guidelines Interpretation by stating no adjustment should apply to the cyclic variation quality "swishing sound" of the noise contamination discharged from the WTGs. The 2008 NPC Guidelines Interpretation suggests the blade swish noise is temporal. This conclusion is not supported by our field observations, or the findings in the Ministry 2007 Acoustic Consulting Report.”
(capitalized font is my emphasis)

Reference [4] NOTE; This memorandum has been published in various media outlets and is in the public domain. The following is one source.
https://nationalpost.com/news/canada/ontarios-wind-turbines-too-noisy-m…

The comment in the 2007 Acoustic Consulting Report[5] referred to by Mr. Cameron was made by the author, Mr. Ramini Ramakrishnan, who stated:

“Due to the nature of the amplitude modulation phenomenon, the swishing or thumping exists all the time.”

Reference [5] Ramakrishnan, Ramani; WIND TURBINE FACILITIES NOISE ISSUES, Aiolos Report Number: 4071/2180/AR155 Rev3, DECEMBER 2007
https://grandbend.northlandpower.ca/site/northland_power___grand_bend_w…

Mr. Brian Howe, the president of the engineering firm HGC Engineering, stated in his Witness Statement for the Cedar Point ERT that:

“Research has shown that the dominant audible characteristic is the amplitude modulation of broad-band sound (the “swoosh”)”

2.2 The wind direction filter for noise compliance testing is limited to downwind conditions at the test site and on-axis +/- 45 degrees.

A study by S. Oerlemans J.G. Schepers [6] found that amplitude modulated noise is greatest at 90° and 270° off axis to the downwind direction at the test sites. However, the current Compliance Protocol require the downwind direction to be off-axis by no more than +/- 45 degrees to the test equipment location.

Reference [6]. S. Oerlemans J.G. Schepers, Prediction of wind turbine noise and validation against experiment, based on an article published in the International Journal of Aeroacoustics, Vol. 8, No. 6, 2009, by MULTI-SCIENCE PUBLISHING CO. LTD., UK.
https://core.ac.uk/download/pdf/53034072.pdf

2.3 There is no consideration or adjustment for turbulent inflow into the turbine rotors during noise compliance audits or noise complaint audits

The comments in Section 1.2 above on turbulent inflow were in reference to the turbine noise emission specification; but it also applies to additional noise immission at receptors by operational turbines during periods of turbulent inflow.

A study by C. Doolan [7] found that turbulence generated noise is greatest at between 90 degrees to 270 degrees off-axis to the downwind direction at the test site. Since the Compliance Protocol requires the data sets to be limited to data with downwind direction no more than +/- 45 degrees off-axis at the test site, this data filter will discard any noise data sets where turbulent inflow noise is present.

Reference [7]. C. Doolan, “Concepts for the control of wind turbine noise”, Proceedings of Acoustics 2011, Gold Coast, Australia, 2-4 November 2011
https://www.acoustics.asn.au/journal/Vol40No1.pdf

This unusual directivity pattern to noise emissions is due to turbulence in the air stream ahead of the leading edge of the blade interacting with the blade surface. This noise also has a characteristic dipole directivity pattern but unlike trailing edge noise, the frequency generated is around 18Hz. Low frequencies, such as 18 Hz, are audible but very weakly attenuated by the atmosphere and can be heard at great distances.

3. Turbine emission audits allow a 0.5 dBA leeway above stated noise specifications.

The 2017 Compliance Protocol for Wind Turbine Noise allows project operators a leeway of +0.5 dBA on their turbine emission results which almost completely negates the purpose of including the turbine emission uncertainty in the noise analysis.

Summary Comments

The current modelling calculations used to predict noise levels, and compliance understate the predicted noise at receptors, resulting in turbine-receptor setbacks that are inadequate to ensure compliance to noise limits.

The 2017 Compliance Protocol for Wind Turbine Noise requires noise measurements to be filtered to include only data sets where the wind direction is downwind and on-axis +/- 45 degrees. This will exclude most amplitude modulated, and turbulence generated, noise events, and thus does not measure the worst case noise levels.

Recommendations:

1) The ISO 9613-2 protocol used to predict the noise impact if a wind turbine project should be replaced with a more accurate noise model, such as the Nord 2000.

2) The turbine manufacturers already provide a measurement “uncertainty” in the turbine’s noise emission levels. The manufacturers should also be required to demonstrate that this uncertainty includes conditions of turbulent inflow.

3) The leeway of +0.5 dBA in the turbine emission audit should be removed.

4) The 2017 Compliance Guidelines for Wind Turbine Noise should be amended so that noise is measured not only downwind and off-axis limited to +/- 45 degrees, but also at 90 degrees and 120 degrees off-axis to account for amplitude modulated, as well as turbulent inflow noise for all noise audits.

5) The Ministry should acknowledge that all turbines generate amplitude modulated noise, and lower the noise compliance level by 5 dBA to an AVERAGE daytime noise level, Leq, of 35 dBA, with a MAXIMUM, Lmax, of 40 dBA (including amplitude modulated and turbulence noise) The MAXIMUM night time noise level, Lmax, should be set 35 dBA (again including amplitude modulated and turbulence noise) to avoid sleep disturbance.

6) Any noise audits conducted in response to complaints should not be restricted by wind direction or other operational parameters. Noise complaint audits should ensure that the data gathered include the same wind and other environmental conditions (within the limits of the testing equipment) during the times noted by the complainants.