Skeabrae Wind Energy Project

Skeabrae Document OSE/3526 March 2013 Skeabrae Wind Energy Project Noise and Shadow Impact Revision 3: V29 225kW wind turbine Richard Gauld IEng MInstMC BSc(Hons) Dip. DesInn Dip. GeoSci 1

Skeabrae Wind Energy Project Environmental Health Studies: Noise and Shadow Flicker Assessment Revision 3: V29 225kW Report OSE/3526 February 2014 Project Developer: Skeabrae Renewables Ltd Garson Farm Sandwick Orkney KW16 3JD Project Design: Richard Gauld IEng MInstMC BSc(Hons) Dip. GeoSci Dip. DesInn Orkney Sustainable Energy Ltd 6 North End Road Stromness Orkney KW16 3AG Telephone 01856 850054 Facsimile 01856 851239 Email info@orkneywind.co.uk 2

1 Summary 1.1 A wind energy development has been designed for RAF Skeabrae, the redundant military aerodrome located west of Dounby in the West Mainland of Orkney; three wind turbines are proposed. 1.2 The project has been revised to reduce impacts, with three V29 225kW turbines now located on the central runway; a revised noise assessment has been completed. Wind turbines can create two types of environmental health nuisance; disturbance from noise and the potential for shadow flicker affects. This study considers the impact that both these possible nuisances could have upon neighbours and concludes that planning guidelines have been achieved. 1.3 Wind turbine noise is produced by air travelling over the turbine blades along with mechanical noise from the gearbox and generator. The Skeabrae project has been revised to have smaller generators, with each turbine now producing 225kW; noise levels have been reduced. Sound impact from the wind turbines is not expected to be significant and a detailed assessment to ISO 9613-2 has been carried out using the noise calculation module of the WINDFARM software package; background noise assessment has not been necessary. 1.4 Shadow flicker can theoretically occur under certain combinations of turbine position, time of day and time of year, when the sun may pass behind the rotor and cast a shadow over neighbouring properties. Planning guidance recommends 10 rotor diameter separation between wind turbines and dwellings to avoid this nuisance, achieved by a large margin at the Skeabrae site; 10 rotor diameters is 290m, with the turbines located more than 530m from the surrounding properties. 1.5 The V29 225kW wind turbine is a Vestas design and is manufactured to the original specifications by RBB in India, and a copy of the measured noise characteristics of this machine has been appended to the report. 3

2 Potential disturbance from noise 2.1 Sound is generated by wind turbines when they operate and generate power, and occurs over the operational windspeed range. Below the cut-in wind speed there is insufficient strength in the wind to turn the blades and in storm conditions the turbine is slowed down as a safety measure. The wind speed range for the Vestas V29 is from 4 to 25 ms -1, with the turbine switching off during storms, and the principal sources of noise are from the blades rotating in the air and nacelle noise from the gearbox and generator. 2.2 Noise levels are normally expressed in decibels. Noise in the environment is measured using the db(a) scale which includes a correction for the response of the human ear to noises with different frequency content. A 3dB change in noise level is perceptible and a 10dB change in noise level is heard as a doubling or halving of the perceived level. 2.3 The sound power level of the Vestas V29 turbine is 98 db(a), with this value decreasing with distance from the turbines, from atmospheric absorption and from various other attenuation factors. Table 1 shows the predicted impact at the neighbours to the Skeabrae project along with other environmental sound levels. Source/Activity Indicative sound level db (A) Threshold of hearing 0 Rural night-time background 20-40 Skeabrae wind turbines 30-33 Quiet bedroom 35 Busy road at 5km 35-45 Car at 65 km/h at 100m 55 Busy general office 60 Conversation 60 Truck at 50kmh at 100m 65 City traffic 90 Pneumatic drill at 7m 95 Jet aircraft at 250m 105 Threshold of pain 140 Table 1 Indicative noise levels in the environment 4

2.4 The Planning Advice Note on Renewable Energy Technologies, PAN 45, has now been superseded by PAN 1/2011, Planning and Noise [1], and by a series of online advice sheets on the different renewable technologies. PAN 1/2011 identifies two sources of noise from wind turbines; mechanical noise and aerodynamic noise. It states that good acoustical design and siting of turbines is essential to minimise the potential to generate noise. The report The Assessment and Rating of Noise from Wind Farms, ETSU-R-97 [2], describes a framework for the measurement of wind farm noise and gives indicative noise levels thought to offer a degree of protection to wind farm neighbours, without placing unreasonable restrictions on wind farm development or adding unduly to the costs and administrative burdens on wind farm developers or planning authorities. 2.5 These planning guidelines suggest that noise limits should be applied to locations where a quiet environment is desirable. Where the wind farm is near to neighbours, noise limits can be set relative to background noise and should reflect the variation in both turbine source noise and background noise with wind speed. Separate noise limits can apply for day-time and for nighttime as the protection of the external amenity becomes less important in the evening and the emphasis should be on preventing sleep disturbance. 2.6 The recommended day-time noise limits are 35-40 db(a) or 5dB(A) above the prevailing background, whichever is the greater. For night-time periods the recommended noise limit is 43 db(a) or 5dB(A) above the prevailing background, again whichever is the greater. The 43 db(a) lower limit is based on a sleep disturbance criteria of 35 db(a) with an allowance of 10dB(A) for attenuation through an open window and 2dB(A) subtracted to account for the use of L A90 rather the L Aeq. Quiet day-time periods are defined as evenings from 1800-2300 plus Saturday afternoons from 1300-1800 and Sundays from 0700-1800. Night-time is defined as 2300-0700. 2.7 Where predicted noise levels are low at the nearest residential properties, due to large distances to the wind turbines or if a small scale development is planned, a simplified noise assessment is permitted by the planning guidelines, such that noise is restricted to an L A90 level of 35 db(a) for wind speeds up to 10 ms -1. This recognizes that extensive background noise measurements are not necessary for small or low density schemes with good neighbour separation. 5

2.8 The Skeabrae project has been revised to have three V29 wind turbines. The nearest turbine is located more than 530m from any neighbour, and as none of the properties are orientated towards the site, it is felt that the simplified 35dB (LA90 10 min) limit is appropriate. Garson Farm is located over 700m from the nearest turbine and noise levels are calculated to be well within the 45dB(A) limit appropriate for those with a financial involvement. 2.9 Noise predictions have been carried out using International Standard ISO 9613-2, Acoustics Attenuation of Sound During Propagation Outdoors [3]. using the ReSoft WINDFARM ISO 9613 software module. The propagation model provides for the prediction of sound pressure levels based on either short-term downwind (ie. worst case) conditions or long term overall averages. Only the down-wind condition has been considered in this assessment, with the wind blowing from the proposed turbine locations towards the nearby houses. When the wind is blowing in the opposite direction noise levels are significantly lower. 2.10 The ISO propagation model calculates the predicted sound pressure level by taking the source sound power level for each turbine in separate octave bands and subtracting a number of attenuation factors according to the following: Predicted Octave Band Noise Level = Lw + D A geo - A atm A gr - A bar - A misc Where: LW = Source Sound Power Level D = Directivity Factor (down-wind is worst-case) A geo = Geometrical Divergence A atm = Atmospheric Absorption A gr = Ground Effects A bar = Barrier Attenuation A misc = Miscellaneous Geometrical divergence has the largest impact on noise reduction, with lesser effects from the other attenuation factors. Noise predictions have been based on measured sound power levels for the proposed turbines as shown in Table 2. The noise spectra have been taken from an assessment out on sample turbines, Table 3, normalised to the warranted sound power level at each integer wind speed; a copy of the turbine noise assessment has been appended to the report. 6

Wind Speed at 10m (m/s) V29 Sound Power Level (dba) 7 97.2 8 97.8 9 98.4 10 99.0 11 99.6 12 100.2 Table 2 Predicted sound power levels from the V29 Octave Band Centre Frequency (Hz) 63 125 250 500 1k 2k 4k 8k V29 225 76.4 83.7 88.5 93.3 92.3 89.6 81.6 71.0 Table 3 Standard octave band spectra from the V29 2.11 Geometric Divergence calculation: Ageo = 20 x log(d) + 11 where d = distance from each turbine Atmospheric Absorption attenuation calculation: Aatm = d x α where d = distance from turbines and α = atmospheric absorption coefficient in db/m Published values of α have been taken from ISO9613 Part 1 corresponding to a temperature of 10ºC and a relative humidity of 70% which give relatively low levels of atmospheric attenuation, and subsequently worst-case noise predictions, Table 4. These atmospheric conditions are in line with current good practice Octave Band Centre Frequency (Hz) Atmospheric Absorption Coefficient (db/m) 63 125 250 500 1k 2k 4k 8k 0.00012 0.00041 0.00104 0.00193 0.00366 0.00966 0.03280 0.11700 Table 4 Atmospheric absorption coefficients 7

2.12 Six properties have been chosen to represent the impact upon different parts of the community, and include the nearest neighbours; Figure 1. The open nature of the site, the extensive farmland and long distances to neighbours are important considerations [4]. Figure 1 Neighbours 2.13 Ground Effect is the interference of sound reflected by the ground interfering with the sound propagating directly from source to receiver. The prediction of ground effects depends on the source height, receiver height, propagation height between the source and receiver and the ground conditions. The ground conditions are described according to a variable G that varies between 0 for hard ground (includes paving, water, ice, concrete & any sites with low porosity) and 1 for soft ground (includes ground covered by grass, trees or other vegetation). The assessment has been carried out using the tower height of 32m, a receiver height of 4m and an assumed ground factor G = 0.5. This ground factor corresponds to a mixture between hard and soft ground and represents current good practice detailed in ISO 9613-2. There has not been any attenuation of the sound levels for Barrier or Miscellaneous Effects. 2.14 The WINDFARM noise module determines the sound levels at neighbours using Ordnance Survey digital map data at a scale of 1:50000. The programme measures distances from surrounding properties to each of the turbines, calculates the line-of-sight distances in each case, allowing for hub height and relative elevation, subtracts the various attenuation factors and then calculates the cumulative sound from all turbines. The WINDFARM noise module calculates total sound levels from all three turbines in db (Aeq), with 2dB then deducted to give theoretical noise levels in db (LA90 10min), Table 5. Figure 2 is a map showing the sound dispersal pattern around the three turbines in db (Aeq);. 8

2.15 The process used to determine the sound levels was thus to identify the neighbours, measure distances from these properties to each of the turbines, and then to subtract the various attenuation factors. Table 5 shows the range of cumulative sound levels, calculated from the above formulae, with the corrected db (LA90) assessment obtained by deducting 2dB. House Status Distances to each turbine (m) Ground conditions Total Sound db(aeq)) Theoretical noise level db(la90 10min) H1 Landowner 705, 840, 940 Pasture + road 31.98 30 H2 Neighbour 760, 780, 810 Pasture + road 32.42 32 H3 Neighbour 540, 660, 770 Pasture + road 34.58 33 H4 Neighbour 580, 720, 860 Pasture + road 33.76 32 H5 Neighbour 670, 690, 740 Pasture +road 33.67 32 H6 Neighbour 530, 580, 660 Pasture + road 35.42 33 Table 5 Predicted sound levels 2.16 The assessment reveals that the project achieves planning guidelines, with predicted noise at neighbours quieter than the recommended level of 35 db (LA90). The project also has sound levels well below the 43 db(a) night-time disturbance figure, and is also below the 45dB(A) limit recommended for project owners. 2.17 Measurements of low frequency and infra-sonic noise around modern wind farms have shown that levels of such noise are below accepted thresholds of perception even on the wind farm itself [5]. There is no evidence that there are any effects of low frequency sounds at levels below perceptibility. The World Health Organisation has also stated that there is no reliable evidence that infrasound below the thresholds of hearing produce physiological or psychological effects [6]. 9

35 db LA90 Figure 2; Sound dispersal pattern, showing the 35 dbla90 noise limit 10

3 Potential disturbance from shadow flicker 3.1 Under certain combinations of geographical position, time of day and time of year, the sun may pass behind the rotor of a wind turbine and cast a shadow over neighbouring properties. The Scottish Government has provided online planning advice regarding onshore wind turbines at http://www.scotland.gov.uk/topics/built-environment/planning/national-planning- Policy/themes/renewables/Onshore. This guidance indicates that shadow flicker occurs within buildings where a blade flicker appears within a narrow window opening, and that the seasonal duration of this effect can be calculated from the geometry of the machine and the latitude of the potential site. The online planning advice indicates that provided a good separation is provided between the turbines and nearby dwellings (as a general rule 10 rotor diameters), shadow flicker should not be a problem. For the Skeabrae project 10 rotor diameters is 290m. 3.2 The Skeabrae project has been designed in conjunction with the software programme WindFarm. This programme uses Ordnance Survey Panorama gridded topographic data at a scale of 1:50,000 to create a virtual landscape for assistance in designing layouts. The programme calculates of the risk of shadow flicker impact, based on the relative geometry of the turbine and any nearby properties. At Skeabrae the nearest properties are 530m or more from the turbines, and accordingly there is little likelihood of shadow flicker effects occurring. 3.3 It has been suggested that shadow flicker from wind turbines can have an influence upon individuals susceptible to photosensitive epilepsy. Epilepsy Scotland indicate that television is the most commonly reported trigger of seizures in photosensitive people, and that flickering sunlight through trees or strobe lighting are also triggers. The most common flash rates which are reported to produce seizures are between 12 and 24 Hz [7]. 3.4 The flash rate which triggers epilepsy in photosensitive people is at a much higher frequency that the shadows produced by rotating wind turbines. To achieve a rate of 12 to 24 Hz from a three bladed wind turbine would require a rotational speed of 240 to 480 rpm, far greater than any grid-connected wind turbine speed, and approaching the rotational speed of helicopter blades. The shadow flicker from wind turbines of the scale proposed for Skeabrae has a maximum frequency 1.5 Hz. 11

References and bibliography 1 Planning Advice Note PAN 1/2011. Planning and Noise. The Scottish Government, March 2011 2 ETSU-R-97. The Assessment and Rating of Noise from Windfarms. DTI Noise Working Group 3 ISO 9613-2. Acoustics Attenuation of Sound During Propagation Outdoors. 4 ETSU W/13/00385/REP. A critical appraisal of wind farm noise propagation. ETSU for the DTI. 5 ETSU W/13/00392/REP. Low frequency noise and vibrations at a modern windfarm. ETSU for the DTI. 6 Community Noise. Archives of the Centre of Sensory Research Vol.2(1), eds. Bergland and Lindvall. World Health Organisation. 7 Epilepsy Scotland. Factsheet http://www.epilepsyscotland.org.uk/images/pdf/photosensitve.pdf 12

Skeabrae Environmental Health Impacts March 2013 Skeabrae Wind Energy Project Noise and Shadow Impact Appendix A V29 225kW Turbine Noise Assessment 13