Statutory Order on Noise from Wind Turbines

Transcription

1 Statutory Order on Noise from Wind Turbines The following shall be laid down pursuant to section 7(1), nos. 1, 2 and 8, section 7a(1), section 92 and section 110(3) of the Environmental Protection Act, cf. Consolidation Act no. of [xxx yy] June 2018, and section 33(1) and (4), section 48(1) and (2), and section 61(1) of the Marine Environment Protection Act, cf. Consolidation Act no of 4 September 2017: Part 1 Scope and definitions 1. This Statutory Order shall apply to the installation, modification and operation of wind turbines. 2. For the purpose of this Statutory Order the following definitions apply: 1) Small wind turbines: Single wind turbines with a rotor area of 200 m 2 or below and with a total height of 25 metres or below, including domestic wind turbines. 2) Prototype wind turbines: The first, non-series-produced wind turbine of a new type. 3) Series 0 wind turbines: The first, small production series of a new type of wind turbine. 4) Test wind turbines: Series 0 wind turbines or wind turbines which are prototype-certified or modified for experimental use pursuant to the Statutory Order on Technical Certification for the Construction, Manufacture, Installation, Maintenance and Service of Wind Turbines. 5) Wind farm: A cluster of three or more wind turbines. 6) Noise impact area around test wind turbines: The largest extent of the area around test wind turbines where total noise level from wind turbines is higher than 37 db(a) at 6 m/s and 39 db(a) at 8 m/s, determined in accordance with the guidelines set out in Annexes 1 and 2. 7) Noise-sensitive land use: Areas that are actually used for or designated in district plans or town planning regulations for residential, institutional, holiday home, camping or allotment purposes or areas designated in district plans or town planning regulations for noise-sensitive recreational activities. 8) Low-frequency noise: Noise in the frequency range from 10 to 160 Hz. Low-frequency noise is characterised by the A- weighted level of noise in one-third octave bands from 10 up to and including 160 Hz, calculated indoors using the method set out in Annex 1. Part 2 Requirements for wind turbines 3. Anyone who owns a wind turbine is responsible for installing, operating and maintaining it in such a way as to ensure compliance with this Statutory Order. 4.-(1) The total noise impact from wind turbines may not exceed the following limit values: 1) At the most noise-exposed point in outdoor living areas no more than 15 metres from dwellings in the open countryside: (a) 44 db(a) at a wind speed of 8 m/s. (b) 42 db(a) at a wind speed of 6 m/s. 2) At the most noise-exposed point in areas with noise-sensitive land use: (a) 39 db(a) at a wind speed of 8 m/s. (b) 37 db(a) at a wind speed of 6 m/s. (2) The total low-frequency noise from wind turbines may not exceed 20 db at a wind speed of 8 and 6 m/s indoors in dwellings in open countryside or indoors in areas with noise-sensitive land use respectively. (3) The limit values laid down in subsections 1 and 2 do not apply to the wind turbine owner's dwelling. (4) The limit values in subsections (1) and (2) shall not apply to a temporary place of residence for refugees, just as such a place of residence shall otherwise not be included in the assessment of noise from wind turbines if the municipal council has notified a dispensation pursuant to section 5u(1), or authorisation pursuant to section 5u(1)-(3), of the Planning Act allowing for the establishment of a place of residence in an area exposed to high levels of noise. 5. The noise impact, cf. section 4, subsections 1 and 2, is determined in accordance with the guidelines in Annexes 1 and 2 and is stated as the equivalent, corrected, A-weighted noise level at a height of 1.5 metres at wind speeds corrected to a height of 10 metres at 6 and 8 m/s respectively at a roughness length of 0.05 metres. 6.-(1) Measurements of the source strength of wind turbines and of tone levels in the noise are carried out in accordance with the instructions in Annexes 1 and 2 as Environmental Measurements - External Noise', cf. Statutory Order on Quality Requirements for Environmental Measurements.

2 (2) Measurements of wind turbines fitted with several generators must use the noise emission from the wind turbine operating with the noisiest generator as the basis for the noise measurement. Part 3 Test wind turbines 7.-(1) Dwellings etc. that are constructed or converted in existing buildings within a noise impact area around test wind turbines after the date of publication of proposals for the district plan designating the area for the installation of test wind turbines are not included in the assessment of noise from test wind turbines, cf. subsection (2). (2) Dwellings etc. that are constructed or converted in existing buildings within a noise impact area around offshore test wind turbines after the date of publication of the environmental impact assessment (EIA) are not included in the noise assessment for these wind turbines. (3) When wind turbines are to be installed or modified outside an area designated for test wind turbines onshore or outside an offshore area, where the Danish Energy Agency has granted permission for the installation of test wind turbines pursuant to section 25 of the Danish Promotion of Sustainable Energy Act, the total noise impact from the test wind turbines on which the noise impact area of the wind turbines is based must form the basis of the assessment of whether the noise levels are in compliance with the limits set out in section 4, subsections (1) and (2). The same applies to inspection of these wind turbines. Part 4 Notifications etc. 8.-(1) Anyone who wishes to install or modify a wind turbine in a way that may lead to increased noise impact from said wind turbine must submit a notification of this to the municipal council. This does not, however, apply to the installation or modification of offshore wind turbines. (2) Such notification must include documentation demonstrating that the wind turbines can comply with the noise limits set out in section 4. (3) Documentation must take the form of: 1) A report of measurement of noise emission from one or more specimens of the wind turbine type concerned, cf. section 6 2) Maps of the area in which the applicant wishes to install the wind turbine(s) concerned. These maps must feature a scale and north arrow as well as accurately indicate the location of the wind turbine(s) concerned, the location of existing wind turbines and of neighbouring dwellings and the distance to these and to other noise-sensitive land use. The calculation of the noise impact found at the points referred to in section 4 in accordance with the guidelines in Annexes 1 and 2. (4) For prototype wind turbines such measurements and calculations in accordance with subsection (3), no. 1, must be provided to demonstrate the likelihood of the turbine complying with the noise limits. 9.-(1) Notification will be regarded as having been submitted when the municipal council has received all the information specified in section 8(3). Notification may take place when the required planning basis, land zone permissions, if required, and EIA permit for the wind turbine have been obtained, cf. Statutory Order on the Assessment of the Environmental Impact of Certain Public and Private Facilities. (2) If the municipal council has not within four weeks from the date specified in subsection (1) made any objections, the wind turbine may be installed or modified unless this is prevented by other legislation. (3) Building and construction work may not commence before the expiry of this four-week time limit unless the municipal council prior to this announces that it will not object to the notification. (4) In areas which, according to municipal or district planning, have been reserved for the installation of several wind turbines or designated as wind farms, and where notification of individual wind turbines takes place consecutively, the municipal council may, to ensure that the total noise impact from all the wind turbines in the mentioned areas complies with the noise limits set out in section 4, set more restrictive requirements for the noise contributed by the individual wind turbine than the noise limits set out in section (1) When a wind turbine is put into operation after installation or modification, the inspection authority must be informed of this, cf. section 11. (2) If a wind turbine that has been notified, or a wind turbine the modification of which has been notified, has not been put into operation within two years of the expiry of the time limit in section 9(2), a new notification containing the information specified in section 8(3) must be submitted to the municipal council.

3 Part 5 Inspection and orders on noise measurements 11.-(1) The municipal council supervises compliance with this Statutory Order, cf. however subsection (2). (2) The Danish Environmental Protection Agency supervises compliance with this Statutory Order for offshore wind turbines. 12.-(1) The municipal council may order the owner of a wind turbine to carry out noise measurements and calculations at his own expense, cf. sections 5 and 6, 1) when a wind turbine that has been notified, or a wind turbine the modification of which has been notified, is put into operation, 2) in connection with an ordinary inspection pursuant to the Environmental Protection Act, but not more often than once a year, or 3) in connection with the processing of neighbours' complaints about noise when the municipal council considers this to be necessary. (2) The municipal council may in conjunction with its supervision of small wind turbines decide that noise measurements should not be carried out as Environmental Measurement - External Noise, cf. section 6(1). (3) For offshore wind turbines, the Danish Environmental Protection Agency may order the owner of a wind turbine to carry out noise measurements and calculations at his own expense, cf. sections (5) and (6), 1) when a wind turbine, or a modification to a wind turbine, is put into operation, 2) in connection with an ordinary inspection pursuant to the Marine Environment Act, but not more often than once a year, or 3) in connection with the processing of neighbours' complaints about noise when the Danish Environmental Protection Agency considers this to be necessary. Part 6 Complaints and penalties 13.-(1) With the exception of decisions pursuant to section 12(1) and (2), and all decisions relating to municipally-owned or municipally-operated wind turbines, decisions taken by the municipal council cannot be appealed to another administrative authority. (2) With the exception of decisions pursuant to section 12(3), decisions by the Danish Environmental Protection Agency cannot be appealed to another administrative authority. 14.-(1) Unless a greater penalty is prescribed in accordance with other legislation, a fine will be imposed on anyone who 1) installs or modifies a wind turbine in a way that may lead to increased noise impact from said turbine without notification and appropriate documentation, cf. section 8, 2) commences building and construction work or installs a wind turbine irrespective of objections from the municipal council, cf. section 9(2) or (4), 3) commences building and construction work in contravention of section 9(3), 4) puts a wind turbine into operation in contravention of section 10, or 5) fails to comply with an order pursuant to section 12. (2) The penalty may increase to imprisonment of up to two years, if the violation is committed intentionally or by gross negligence and if through this violation 1) the environment has been adversely affected or exposed to danger, or 2) a financial advantage is achieved or strived towards by the person concerned or others, including reductions in expenses. (3) Criminal liability can be imposed on companies etc. (legal entities) in accordance with the rules in Part 5 of the Danish Criminal Code. Part 7 Entry into force etc. 15.-(1) This Statutory Order shall enter into force on [15 October 2018]. (2) Statutory Order no of 21 December 2015 on Noise from Wind Turbines is repealed. (3) Section 7 applies to test wind turbines where the EIA report for offshore wind turbines or proposals for district plans designating the area for the installation of test wind turbines have been published on 1 January 2012 or later. (4) If a wind turbine, or a modification to a wind turbine, has been notified prior to [15 October 2018], but has not been put into operation within two years of the expiry of the municipal council's objection period, a new notification containing the

4 information specified in section 8(3) of this Statutory Order must be submitted to the municipal council. This does not, however, apply to offshore wind turbines. (5) Statutory Order no. 304 of 14 May 1991 on Noise from Wind Turbines still applies for wind turbines notified or put into operation prior to 1 January 2007, cf. however subsection (9). (6) Statutory Order no of 14 December 2006 on Noise from Wind Turbines still applies for wind turbines notified or put into operation prior to 1 January 2012, cf. however subsection (9). (7) Statutory Order no of 21 December 2015 on Noise from Wind Turbines still applies for wind turbines notified or put into operation prior to [15 October 2018], cf. however subsection (9). (8) Statutory Order no of 21 December 2015 on Noise from Wind Turbines still applies for offshore wind turbines notified a construction licence pursuant to the Promotion of Renewable Energy Act or put into operation prior to [15 October 2018], cf. however subsection (9). (9) Notwithstanding subsections (5)-(8), the rules in this Statutory Order shall apply in the event of modifications to a wind turbine, provided the modifications entail an increased noise impact from the wind turbine as assessed according to the guidelines set out in Annexes 1 and 2 to this Statutory Order, cf. however subsection (4). Section General rules for measurement of noise emission from wind turbines Measurement position for onshore wind turbines Annex 1 The noise emission from a wind turbine (sound power level L WA in 1/3 or 1/1 octave bands) is measured at different levels of the electrical power produced by the turbine at a point on the leeward side of the tower. Measurements must be taken at a distance R from the base of the turbine, which must not deviate more than ± 20% from the distance R 0 (see figure 1). The deviation from R 0 must not exceed ± 30 metres. The distance R 0 is the hub height of the wind turbine (h) plus the radius of the rotor (½ d). For a vertical axis wind turbine the distance R 0 is instead the height to the middle of the rotor plus the diameter of the element rotating on the vertical axis; this is not illustrated.

5 Figure 1. During the measurement the microphone must be positioned so that the direction from the tower of the turbine to the microphone does not deviate more than ±15 from the wind direction. Based on the sound measurements, A-weighted reference spectra are determined at wind speeds of 6 and 8 m/s respectively. One half of a windscreen is affixed to the microphone, which is positioned directly on a reflective board on the ground in order to eliminate wind noise in the microphone as much as possible. The board must be no smaller than one metre in any direction. The wind may produce too high a level of background noise when measuring noise emissions at low frequencies. This may be counteracted by using a secondary windscreen which is hemispherical, between 40 and 50 cm in diameter or larger and located concentrically above the microphone and the primary windscreen. If a secondary windscreen is used, the attenuation of sound caused by the windscreen must be corrected for. Correction must be made in the 1/3 octave bands. Measurement of sound spectra and wind speed The noise from the wind turbine is measured as a number of A-weighted sound spectra per 1/3 octave in a frequency area including, as a minimum, the 1/3 octave bands from 20 to 10,000 Hz. A number of sound spectra must be measured in periods of either 10 or 60 seconds. In the same period, the average electrical power produced by the turbine is registered as well as the wind speed measured with the built-in anemometer in the wind turbine at hub height and with an anemometer erected at a height of at least 10 metres near the wind turbine in a position where neither the wind turbine nor other objects in the area are deemed to affect the wind measurement. When the average electrical power produced by the wind turbine is less than 0.95 times its nominal power, the wind speed v h at hub height can be calculated on the basis of the power curve of the wind turbine. The wind speed v ref at a height of 10 metres under reference conditions may then be determined by using equation At higher electrical output the wind speed v h at hub height is determined instead by using the built-in anemometer in the wind turbine at hub height, and the wind speed v ref is determined by using equation If in exceptional cases the power curve of the wind turbine is not known, or when the wind turbine has been parked due to background noise being measured, the wind speed is determined by using the anemometer erected at a height of at least 10 metres. The wind speed v ref is then determined by using equation

6 Based on all measurements while the wind turbine is in operation with an average power of less than 0.95 times nominal power, the correlation between, on the one hand, the wind speed that is read from the turbine's power curve and converted to V ref and, on the other hand, the wind speed that is measured by the wind turbine's built-in anemometer at hub height and converted to V ref and the wind speed that is measured by the anemometer erected at a height of at least 10 metres and converted to V ref. For each measurement period of 10 or 60 seconds, the ratio between the wind speed based on the power curve of the wind turbine and the two separately measured wind speeds is determined, and finally the average value of all ratios for each of the two anemometers is determined. These averaged ratios must be multiplied with the readings from the built-in anemometer in the wind turbine at hub height when the average power produced is more than 0.95 times nominal power and with the reading on the anemometer erected at a height of at least 10 metres when the wind turbine has been parked and background noise is measured. This constitutes an in situ calibration of the built-in anemometer in the wind turbine and the erected anemometer. Because the power curve of the wind turbine is used to establish the correlation between the power produced and the wind speed, a copy of the power curve must form part of the report on source strength measurements. For an averaging period of 10 seconds a minimum of 30 spectra must be measured at average electrical power equivalent to the wind speed v ref at a height of 10 metres under reference conditions falling within a range of 5.5 m/s v ref 6.5 m/s and a minimum of 30 spectra where v ref similarly falls within a range of 7.5 m/s v ref 8.5 m/s. For an averaging period of 60 seconds proportionally fewer spectra may be measured in each of the two ranges. Of the above, measured spectra must include a minimum of 12 spectra (for a period of 10 seconds) that fall within the following four ranges for v ref: 5.5 m/s v ref < 6.0 m/s 6.0 m/s v ref 6.5 m/s 7.5 m/s v ref < 8.0 m/s 8.0 m/s v ref 8.5 m/s The A-weighted reference spectrum at 6 and 8 m/s respectively for each 1/3 octave bands (or 1/1 octave bands) is then determined as the average energy value of the measured sound pressure spectra for v ref within the specified ranges at around 6 and 8 m/s respectively. Equation Correction of wind speed measured at hub height to a height of 10 metres. Translation of formula above: hvor: = where: vindmøllens navhøjde (I meter) = the hub height of the wind turbine (in metres) referenceruheden 0,05 meter (fast værdi) = reference roughness 0.05 metres (fixed value) referencehøjden 10 meter (fast værdi) = reference height 10 metres (fixed value)

7 If the wind speed is measured at height z, the correlation between v ref and v z is shown in equation Equation Correction of wind speed measured by erected anemometer at height z to 10 metres. Translation of formula above: højden af den opstillede vindmåler (I meter)= height of anemometer erected (in metres) er rugheden af terrænet på det aktuelle målested. Terrænets ruhed z 0 estimeres ud fra tabel 1.1 = is the roughness of the terrain at the current point of measurement. The roughness of the terrain z 0 is estimated on the basis of table 1.1. Type of terrain Roughness z 0 [metres] Water, snow, sand Open, flat land, bare soil, mown grass 0.01 Farmland with vegetation 0.05 Residential area; small town; area with dense, tall vegetation 0.3 Table 1.1: Roughness for different types of terrain Correction for background noise and determination of sound power level With the turbine parked, the background noise is measured as an equivalent number of sound spectra and within the same wind speed ranges as indicated above. The wind speed is measured using an anemometer erected at a height of at least 10 metres, and the wind speed v ref is calculated using equation The average energy value of the measured background noise spectra is determined at 6 and 8 m/s respectively and used to correct the wind turbine's reference spectrum by correcting the sound pressure levels L A, ref in each 1/3 octave band (or 1/1 octave band) in the reference spectrum by use of equation If the sound pressure level in the reference spectrum is not at least 3 db higher than the sound pressure level of the background noise, the correction for background noise must be limited to 3 db. The total level L Aeq of the averaged background noise must be at least 6 db lower than the total level L Aeq of the wind turbine noise. If this is not the case, a new measurement must be carried out when the background noise is lower. However, for inspection of noise impact, measurements may be used where the different between total noise and background noise is less than 6 db provided the calculated noise level after correction for a background noise of -1.3 db is no higher than the limit values.

8 Equation Correction for background noise Translation of formula above: Det korrigerede referencelydtrykniveau I 1/3-oktavbånd (eller 1/1-oktavbånd) = the corrected reference sound pressure level in 1/3 octave band (or 1/1 octave band) Den midlede baggrundsstøjs lydtrykniveau I 1/3-oktavbånd (eller 1/1-otavbånd) = the averaged background noise sound pressure level in 1/3 octave band (or 1/1 octave band) The wind turbine's sound power level L WA, ref in 1/3 octave bands (or 1/1 octave bands) can then be calculated using equation Equation The sound power level of the wind turbine Translation of formula above: 6 db er en korrektion på grund af måling tæt ved en reflekterende plade på jorden = 6 db is a correction due to measurement close to a reflective board on the ground Den aktuelle måleafstand mellem mikrofonen og vindmøllens fod = the current measuring distance between the microphone and the base of the wind turbine Notwithstanding the above, measurements carried out in accordance with the IEC standard may also be used as a basis for the determination of L WA,ref Determination of sound pressure level L pa At a point, e.g. by the nearest neighbour, the wind turbine's sound pressure level in 1/3 octave bands (or 1/1 octave bands) at a height of 1.5 metres can be determined using equation L pa = L WA,ref 10 log(l 2 + h 2 ) 11dB + ΔL g ΔL a + ΔL m, where l = the distance from the base of the wind turbine to the calculation point 11dB = correction for distance 10 x log 4 π ΔL g = correction for terrain (1.5dB for onshore wind turbines and 3dB for offshore wind turbines) ΔL a = air absorption, (α a l 2 + h 2 ), where the α a attenuation coefficient is given in tables 1.2 and 1.3.) ΔL m = correction for multiple reflections (0 db for onshore wind turbines, for offshore wind turbines: see below) Equation Calculation of sound pressure level in 1/3 octave bands (or 1/1 octave bands) Calculations are carried out for 1/3 octave bands 50 Hz to 10,000 Hz or 1/1 octave bands 63 Hz to 8,000 Hz. A safety margin may be added so that higher numbers for the source strength L WA,ref are used for the calculations than those shown in the measurement report.

9 Octave band centre frequency in Hz ,000 2,000 4,000 8,000 αa in db/km Table 1.2: Air absorption coefficients per 1/1 octave bands at a relative air humidity of 80% and an air temperature of 10º C 1/3 octave bands centre frequency in Hz α a in db/km /3 octave bands centre frequency in Hz ,000 1,250 1,600 2,000 αa in db/km /3 octave bands centre frequency in Hz 2,500 3,150 4,000 5,000 6,300 8,000 10,000 αa in db/km Table 1.3: Air absorption coefficients per 1/3 octave bands at a relative air humidity of 80% and an air temperature of 10º C Correction for multiple reflections for offshore wind turbines The threshold distance, l 0, for multiple reflections is found for v ref = 6 m/s and v ref = 8 m/s using equation l 0 = 2000 h 30 6 v ref Equation Threshold distance for multiple reflections The rated distance, l, between the wind turbine and the calculation point is determined using equation l = l l 0 Equation Rated distance between wind turbine and calculation point Correction for multiple reflections, ΔL m, is subsequently calculated using equation for l 1 10 log l for 1 < l < ΔL m = N log l + 4 for l 5 { 10 log l + (N 10) log 5 for l > 5 where N is a frequency-dependent scaling factor given in table Equation Calculation of correction for multiple reflections 1/3 octave centre frequency in Hz N Table Scaling factor N for determination of correction for multiple reflections The total A-weighted sound pressure level L pa, tot at the point is then found by adding the sound pressure levels L pa, i in each 1/3 octave band (or 1/1 octave band), cf. equation

10 Equation Total sound pressure level The uncertainty of the calculated sound pressure level L pa, tot by use of this method is ±2 db Determination of tones and noise exposure level L r In order to determine the noise exposure level L r (noise impact) at a given point, the content of clearly audible tones in the noise is measured. If a frequency analysis, performed pursuant to Annex 2, of the wind turbine noise measured close to the turbine, as described in the procedures for measurement of the A-weighted sound power level, shows that no penalty should be given for audible tones near the turbine, there will be no tones in the noise near dwellings and a separate assessment hereof would be unnecessary. If the measurement near the wind turbine indicates a penalty, measurements are performed at the dwelling(s) where the noise exposure level, with or without a possible penalty for tones, is greatest. The noise measurement must be carried out at a representative point close to the dwelling(s), 1.5 metres above ground and selected in such a way that the wind noise has as little effect as possible on the measurement results. 1 There must be a tailwind ± 45 from the wind turbine towards the measurement position. In this context, there are no requirements for temperature gradient or cloud cover. The wind speed 10 metres above ground level, determined on the basis of a power curve for the wind turbine that may cause tone levels in the noise at neighbours, must be between 5 and 9 m/s. For small wind turbines where there is no power curve available, the wind speed is determined by measurement 10 metres above ground level. The measurement should be performed for at least one hour during operating conditions where the tone is most audible. Objective analyses of the tone level are performed according to the guidelines in section 3 and such that there are at least 5 spectra (averaging period 60 seconds) in each of the wind-speed ranges 5.5 m/s to 7 m/s (at least one spectrum at a wind speed below 6.0 m/s) and 7 m/s to 8.5 m/s (at least one spectrum at a wind speed above 8.0 m/s). The penalty K T, in db, is determined for each of the measured spectra as indicated in Annex 2, section 2.4, equation The noise exposure level, L r, for 6 m/s and 8 m/s, respectively, is determined on the basis of the highest K T value for each of the associated wind-speed ranges as indicated in equation L r = L pa,tot + K T Equation Determination of clearly audible tones When processing a notification, the tone content may be determined on the basis of a measurement on the tailwind side of an equivalent turbine at a distance corresponding to the distance to the neighbouring point Determination of low-frequency noise from wind turbines The level of low-frequency noise, e.g. in the nearest dwelling, is determined by use of equation L palf = L WA,ref 10 log(l 2 + h 2 ) 11dB + ΔL glf ΔL σ ΔL a + ΔL m, 1 See guidelines on this in the conclusion in Report no. 28 from the Danish Environmental Protection Agency's Reference Laboratory for Noise Measurements (RL) 17/15 on the assessment of the tone levels in wind turbine noise at neighbours (Danish title: Vurdering af toneindhold i vindmøllestøj hos naboer ).

11 where l = the distance from the base of the wind turbine to the calculation point 11 db = correction for distance 10 x log 4 π ΔL glf = correction for terrain at low frequencies (table 1.4) ΔL σ = sound insulation at low frequencies (table 1.4) ΔL a = air absorption, (α a l 2 + h 2 ), where the attenuation coefficient α a is given in table 1.4. ΔL m = correction for multiple reflections (0dB for onshore wind turbines, for offshore wind turbines, see equation 1.2.4) Equation Calculation of low-frequency noise from wind turbines in 1/3 octave bands 1/3 octave bands centre frequency in Hz LgLF: ground correction, onshore wind turbine (db) LgLF: ground correction, offshore wind turbine (db) Lσ: sound insulation (level difference) (db) Lσ: sound insulation, areas with holiday homes (level difference) (db) αa in db/km /3 octave bands centre frequency in Hz LgLF: ground correction, onshore wind turbine (db) LgLF: ground correction, offshore wind turbine (db) Lσ: sound insulation (level difference) (db) Lσ: sound insulation, areas with holiday homes (level difference) (db) αa in db/km Table 1.4: Terrain correction for low-frequency noise for wind turbines location onshore and offshore, respectively, sound insulation (level difference) and air absorption coefficients per 1/3 octave bands at a relative air humidity of 80% and an air temperature of 10º C Ground correction for offshore wind turbines is valid for calculation of low-frequency noise in a building close to the coast. If noise is to be calculated in a building which, seen in the direction of the wind turbines, is more than 200 metres inland, the terrain correction for onshore wind turbines is used instead. For buildings located between 0 and 200 metres from the coast linear interpolation between the two values for terrain correction is made. The total A-weighted sound pressure level LpALF, tot in the dwelling is then found by adding the sound pressure levels L palf, i in each 1/3 octave band, cf. equation Equation Total sound pressure level The uncertainty of the calculated sound pressure level L pa, tot by use of this method is ±2 db. Section 2 Special rules 2.1 Determination of noise from wind farms In this Statutory Order a wind farm means a cluster of three or more identical wind turbines irrespective of whether these are erected onshore or offshore.

12 The sound power level L WA, ref in 1/3 octave bands or 1/1 octave bands is determined by measurements of at least three randomly selected wind turbines of the same type. For the other wind farm turbines the average energy value of the three (or more) measured sound power levels is used. The erected anemometer must, if it is located on the tailwind side of one of the other wind turbines, have a distance to this wind turbine of at least ten times the rotor diameter (d) of the turbine. See Figure 1. The sound pressure level in 1/3 octave bands or 1/1 octave bands at a point is found by adding the noise contribution from each wind turbine calculated using equation 1.2.1, as shown in equation 2.1. Equation 2.1. Total sound pressure level from several wind turbines The same formula is used when the contribution from a new wind turbine needs to be added to the sound pressure level produced by the existing wind turbines around the dwelling concerned. The total A-weighted sound pressure level L pa, tot at the point is then found by using equation Correction for multiple reflections for offshore wind farms Optionally, the calculation of ΔL m for the individual wind turbines in an offshore wind farm can be calculated by determining, for all of the wind turbines in the farm, the threshold distance, l 0, for multiple reflections for 6 m/s and 8 m/s, respectively, for a wind direction that is equal to the direction from the wind turbine to the calculation point. For all wind turbines in the wind farm, the threshold distance is thus determined using equation A more detailed calculation of ΔL m may be performed by determining, for each wind turbine, the threshold distance, l 0, on the basis of the tailwind component in the direction from the wind turbine to the calculation point relative to a single, fixed wind direction, which is the direction going from the turbine closest to the calculation point to the calculation point. The tailwind component, v ref, k, i, for the i th wind turbine in the farm is determined using equation 2.2: V ref,k,i = v ref cos θ i Where θ i is the angle between the direction from the calculation point to the nearest wind turbine and the direction from the calculation point to the i th wind turbine. Equation 2.2. Calculation of tailwind component Cos θ i can be found using equation 2.3: Cos θ i = l l i 2 l m 2 2 l 1 l i where l 1 = the distance from the calculation point to the nearest wind turbine in the farm l i = the distance from the calculation point to the i th wind turbine in the farm l m = the distance between the i th wind turbine and the wind turbine nearest the calculation point. Equation 2.3. Calculation of cos θ i The threshold distance l 0,i for the i th wind turbine in the wind farm is then found by inserting V ref,k,i in equation

13 2.2 Measurement of noise emission from offshore wind turbines A. Microphone mounted on a reflective board on the ship Compared to measurements for onshore wind turbines, the measurement method is changed so that the reflective board on to which the microphone is placed is affixed to the roof of the pilot house on the measuring vessel or on a correspondingly large surface with an unobstructed view to the wind turbine from the location of the microphone. The roof or board must be no smaller than four metres in any direction. Otherwise the instructions set out in Part 1, section 1, apply. B. The microphone mounted off the side of the ship If the microphone cannot be mounted as specified under Heading A, the microphone must be positioned 3-5 metres above sea level, free from reflective surfaces, etc. and 1-2 metres away from the side of the measurement vessel with an unobstructed view to the wind turbine. A primary windscreen must be affixed to the microphone, and the microphone axis must point in the direction of the hub of the wind turbine. The noise from the wind turbine is measured as A-weighted spectra for a number of periods in accordance with the guidelines set out in Part 1, section 1 for onshore wind turbines, with simultaneous recording of the power produced by the wind turbine, the wind speed at hub height measured by the built-in anemometer in the wind turbine and the wind speed vz at a height of at least 10 metres above sea level with the anemometer placed on the same vessel as the microphone. Due to the low roughness value of the surface of the sea, v z = v ref. If the background noise is too high, its effect may be reduced by increasing the height of the microphone to five metres and reducing the measuring distance. The turbine's sound power level L WA, ref in 1/3 octave bands (or 1/1 octave bands) is then determined as indicated in 2.4. Equation 2.4. Sound power level for offshore wind turbines The instructions above for the measurement of the noise exposure level of wind farms also apply to the measurement of the noise impact of offshore wind farms. 2.3 Measurement of noise emission from small wind turbines For small wind turbines, including domestic turbines, the source strength is determined in accordance with the principles in the method specified in 1.1, with the option of the following deviation: During measurement the microphone must be positioned in such a way that the direction from the tower of the turbine to the microphone does not deviate more than ±45 from the wind direction. Small wind turbines usually do not provide the option of direct reading of the power produced in short time ranges, and the correlation between the noise emission of the wind turbine and the wind speed is, therefore, based on measurements by an anemometer erected at a height of at least 10 metres near the wind turbine in a position where neither the wind turbine nor other objects in the area are deemed to affect the wind measurement. 2

14 Annex 2 Objective method for determining the audibility of tones in noise Section 1 The 2 method contains procedures for stationary and variable tones, for narrow-band noise as well as for low-frequency tones, and the result is a graduated penalty of between 0 and 6 db 3. Section 2Objective method 2.1 General The method involves three steps: 1. Narrow-band frequency analysis (FFT analysis). 2. Determination of averaged sound pressure levels for tone(s) and masking noise within a critical band around the tone(s). 3. Calculation of the audibility of the tones, L ta and the penalty, K T Frequency analysis Each A-weighted narrow-band spectrum is determined by linear averaging for one minute (long-term averaging). The effective analysis bandwidth should be less than 5% of the width of the critical band with tonal components. Table 1 below shows the width of the critical bands. The frequency analyser must be calibrated in db re 20 Pa, and it is recommended to use Hanning-weighting as the window function. Notes: 1) With the recommended Hanning time window, the effective analysis bandwidth is 1.5 times the frequency resolution. The frequency resolution is the distance between the spectral lines. 2) With an effective analysis bandwidth of 5% of a critical band, audible tones will normally show as local maxima of at least 8 db above the surrounding noise in the averaged spectrum. 3) In rare cases, with a complex of many closely spaced tones, a higher resolution (smaller bandwidth) may be necessary to determine the noise level of the masking noise correctly. 4) If the frequency of audible tones within the averaging period varies by more than 10% of the width of the relevant critical band, it may be necessary to subdivide the long-term averaging into a number of averaging periods of a shorter duration. 2 The method is the same as the one described in Briefing (in Danish: Orientering) no. 31 of 2001 from the Reference Laboratory on a proposal for a revised, objective method for determination of the audibility of tones in noise (Danish title: "Forslag til revideret objektiv metode til bestemmelse af tydeligheden af toner i støj"), and which is reproduced in Danish Standard DS/ISO , 2nd version, on acoustics: measurements, description and assessment of external noise, part 2 noise levels (Danish title: Akustik Måling, beskrivelse og vurdering af ekstern støj Del 2: Bestemmelse af eksterne støjniveauer ) Annex C (informative) and in the British Standard BS 4142:2014: Methods for rating and assessing industrial and commercial sound Annex D (normative) Objective method for assessing the audibility of tones in sound: Reference method. 3 The audibility of tones that were on the criterion curve for the fixed penalty of 5 db to the measured noise level in the previously used method, will give rise to a penalty of 2.5 db according to this new method. Very audible tones will be added 6 db.

15 2.3 Determination of sound pressure levels Tonal sound pressure levels, L pt The tones are identified by visual inspection of the narrow-band spectrum and the tonal sound pressure level is determined on the basis of the spectrum. All local maxima with a 3 db bandwidth smaller than 10% of the bandwidth of the relevant, critical band are regarded as a tone. The levels, L pti, of all tones, no. i, within the same critical band must be energy summed to arrive at the total tone level in this band, L pt: L pt L pti log 10 Equation The total tone level in a critical band Notes: 1) If a 'tone' is actually narrow-band noise, or if the frequency of the tone varies, the tone will appear as a number of lines in the averaged spectrum. In such cases, L pti is the energy sum of all lines with levels within 6 db of the local maximum, corrected for influence from the window function used (for the Hanning window, this is the energy sum minus 1.8 db). 2) In cases where tones occur at low frequencies, it is advisable to examine whether the total tone level is above the hearing threshold (See DS/EN ISO 387-7). If the total tone level in a critical band is under the hearing threshold, this critical band should be disregarded in the determination of the audibility of tones Bandwidth and centre frequency for critical bands The bandwidths of critical bands are shown in table 2.1. Centre frequency, fc [Hz] Above 500 Bandwidth [Hz] % of f c Table 2.1 The width of critical bands. The critical bands must be placed with a centre frequency, f c, equal to the tone frequency. If several tones can fall within the same critical band, the critical band must be placed symmetrically around the most significant tones in such a way that the difference between the total tone level, L pt, and the masking noise level, l pn, (see section 2.3.3) is as large as possible. In order to determine the centre frequency of the critical band, f c, only tones within ten db from the tone with the highest level are considered significant. Notes: The centre frequency of the critical band, f c, may vary continuously over the relevant frequency area. The lowest critical band is 0 Hz to 100 Hz The sound pressure level of the masked noise in a critical band, L pn

16 The average noise level in a critical band, L pn, avg, can be found by visual averaging of the levels of the spectrum's "noise lines" within an area of between 0.5 and 1 critical band around the centre frequency, f c. The "noise lines" are found by disregarding all maxima in the spectrum that are due to tones and their possible sidebands in the area in question. The total sound pressure level of the masking noise in a critical band, L pn, is calculated on the basis of the average noise level in the same critical band, L pn, avg, as indicated in equation 2.3.2: L pn = L pn, avg + 10 log (Bandwidth of critical band / effective analysis bandwidth) Equation The total sound pressure level of the masked noise in a critical band 2.4 Calculation of the audibility of the tones, L ta, and the penalty, K T The audibility of the tones, L ta, is expressed in db above the masking threshold, MT, see figure 2. The penalty, K T, is the value which is to be added to the L Aeq for a time period in order to arrive at the noise exposure level - corrected for tones - in said time period. On the basis of the difference between the tone level and the noise level in a critical band, L pt - L pn, both L ta and K T can be determined using figure 2. A given centre frequency, f c, of the critical band and a given difference in levels L pt-l pn determine a point in figure 2. L ta is determined as the difference between L pt-l pn and the masking threshold, as indicated in the figure. K T is read by interpolating between the lines marked by different values of K T in the figure. Alternatively, L ta may be calculated using equation 2.4.1, and K T can be calculated using equation Lpt - Lpn [db] k T =6dB k T =5dB k T =4dB k=3db k T =2dB k T =1dB k T =0dB -2 Maskerings tærskel, MT , Centerfrekvens af kritisk bånd, f c [Hz] Figure 2 Masking threshold, MT, and curves for determination of the penalty, K T. L pt is the total sound pressure level of the tones in a critical band, and L pn is the total sound pressure level of the masking noise in the same band. Translation of figure above: Centerfrekvens af kritisk bånd = Centre frequency for critical bands

17 Maskeringstærskel = Masking threshold The audibility of the tones in a critical band can be calculated using equation 2.4.1: L ta = L pt L pn log (1 + ( f c / 502 ) 2,5 ) db re MT, where L pt is the total sound pressure level of the tones in the critical band, L pn is the total sound pressure level of the masked noise in the critical band, and f c is the centre frequency in Hz of the critical band. Equation Audibility of tones in a critical band The penalty, K T, in db may be determined on the basis of equation 2.4.2: 10 db < L ta: K T = 6 db 4 db L ta 10 db: K T = L ta -4 db L ta < 4 db: K T = 0 db Note: K T is not limited to integer values. Equation The penalty K T If there are several tones (or groups of tones) within the same spectrum in different critical bands, there must be separate calculations for each of these bands. The critical band that contains the most audible tone(s), i.e. the band that gives the highest L ta value, is crucial for L ta and for the penalty, K T, based on this spectrum. The tonal analysis must be documented as follows: Section 3: Documentation 3.1. The frequency analysis Measurement period, number of spectra/averaging periods and effective analysis bandwidth. Time window (e.g. Hanning), time weighting (Lin) and frequency weighting (A). A typical spectrum with indication of the location of the crucial critical band and the average noise level in said band. 3.2 The calculations in the crucial critical band State whether the results have been achieved by visual averaging or by automatic calculation. The cut-off frequencies for the critical band and the area for visual averaging or linear regression (see section 4.4).

18 Lydtrykniveau pr. analysebåndbredde, db re 20 Pa While this translation was carried out by a professional translation agency, the text Frequencies and individual tone levels and the total tone level, L pti and L pt in db re 20 Pa. The level of the masking noise in the critical band, L pn in db re 20 Pa. The audibility of tone(s), L ta, in db above the audiometric threshold. The size of the penalty, K T, in db. Tones in other critical bands that may trigger a penalty, must as a minimum be indicated by their frequency. Section 4 Detailed definitions of tone and noise levels For the purpose of computer implementation of the method, in particular, this section provides more detailed definitions of tones and noise. 4.1 General Tone når 3 db båndbredde < 10% af kr. b. Toneenergi Støjpause start Støjpause slut Hverken tone eller støj Støjniveau Lineær regression Kritisk bånd (Kr.b.) af kr.b. Frekvens, Hz Figure 3

19 Definitions of tones (red lines), noise (blue lines) and noise breaks (neither tones nor noise, green lines). is the tone seek criterion and is usually set to 1 db. Translation of formula above: Lydtrykniveau pr. Analysebåndbredde, db re = Sound pressure level per analysis bandwidth Toneenergi = Tone energy Støjpause start = Start of noise break Støjpause slut = End of noise break Hverken tone eller støj = Neither tone nor noise Støjniveau Lineær regression = Noise level Linear regression Kritisk bånd = Critical band Frekvens = Frequency Note: It is important that software implementations allow for making visual inspections of the results. As a minimum the following is required: A spectrum with the indication of the lines that have been defined as tones and highlighting the surrounding critical band and associated regression lines for the noise. Moreover, it would be helpful if the following is indicated: lines characterised as noise, noise breaks and tones. For example, this could be done using different colours for spectral lines as in figure Noise breaks Noise breaks are local maxima in the spectrum, with the possibility of the occurrence of tones. The noise breaks are defined and found on the basis of the following principle: The beginning of a noise break is found on the positive (left) flank of a local maximum as the line, s, where the following conditions are met: L s - L s-1 db and L s-1 - L s-2 < db, where L s is the level of line no. s and L s-1 is the level of line no. s 1, etc. is the tone seek criterion, which is usually set to 1 db. For normal and regular spectra, the tone seek criterion = 1 db works without problems. For irregular spectra (e.g. spectra measured over short averaging periods as mentioned in section 4.3), values up to 3 or 4 db may provide more correct results. It is recommended that this parameter be user-defined in software implementations of the method. The end of the noise break is indicated on the negative (right) flank of a local maximum as the line, e, where the following conditions are met: L e L e+1 db and L e+1 L e+2 < db. All lines from and with s and up to and including e are defined as belonging to a preliminary noise break.