IN THE ENVIRONMENT COURT

IN THE MATTER OF Appeal No. ENV-2007-WLG-000098 under sections 120 and 121 of the Resource Management Act 1991 (“the Act”)

BETWEEN MOTORIMU LIMITED Appellant

AND PALMERSTON NORTH CITY COUNCIL First Respondent

AND HOROWHENUA DISTRICT COUNCIL Second Respondent

______Brief of evidence by Dr Dave Bennett called by the Tararua-Aokautere Guardians Inc. ______

STATEMENT OF EVIDENCE OF DAVE BENNETT CALLED BY THE TARARUA- AOKAUTERE GUARDIANS INC.

INTRODUCTION

1. I am presently a director of several public companies, namely: Trans-Orient Petroleum Ltd, TAG Oil Ltd, and Rift Oil Plc. The two former are Canadian companies involved in oil & gas exploration in , while Rift is a British listed company involved in oil & gas exploration in Papua New Guinea.

2. In addition to these executive and directorial positions I also work as an Exploration and Energy Consultant.

3. I have a BA (Canterbury) : Natural Sciences (Physics/Maths); an MSc (Leeds): Exploration Geophysics, and a PhD (Australia National University): Geophysics.

4. In my work capacity I have acted as adviser to various NZ energy and electricity generation companies, and have played a significant role in the discovery of several of New Zealand’s existing oil and gas fields.

5. Of some relevance to this hearing is that I led the commissioning of a 1 MW power plant using associated gas from a Taranaki oil field; and that I have also chaired full day sessions of the annual NZ Electricity Conference, hence have a familiarity with the NZ electricity generation industry.

6. Also of relevance is that I have more than 30 years experience in applied seismology, which is closely allied to the field of acoustics, since it deals with propagation of pressure waves through solids and liquids, whereas acoustics concentrates on propagation through air. There is a considerable degree of overlap between the two disciplines, particularly regarding effects at the earth- air and water - air interface. I have considerable experience in conducting seismic surveys, where energy is propagated through the ground from explosive or vibratory sources. In this capacity and in the management of drilling operations I have had direct dealings with residents affected by noise

2 propagation affecting their residences in the form of audible noise and/ or vibration. In this capacity I have used NZ noise standards.

7. I am also a resident of the Makara Valley near Wellington, sited close to Meridian’s West Wind complex, now under construction. I was elected by the Makara community onto the Project West Wind Liaison Group, which has an oversight role on the activities of Meridian in this development.

8. I have read and understand the conditions of conduct for being an expert witness. I comply with all criteria for being an expert witness.

3 EVIDENCE

My evidence consists of three topics.

Electricity Supply & Demand

9. A discussion of NZ's electricity demand and the electrical energy output of Motorimu wind turbines as opposed to the existing economic alternative - combined cycle turbines. This demonstrates that Motorimu's contribution to NZ electricity supply, in either the 75 or 113 turbine configuration, is small and will not contribute to security of supply

Global Warming, GHG Savings and Kyoto Targets

10. A view on comparative Greenhouse Gas (GHG) emissions between wind and combined cycle turbine generators, which leads to the conclusion that the GHG savings of wind turbines are extremely small and can even be negative; hence that GHG is not a significant factor in considering wind farm developments,

Noise from wind turbines

11. The section on turbine acoustic noise is complementary to the submissions of Mr Bob Thorne of Noise Measurement Services Pty Ltd. I provide evidence that demonstrates that turbine noise is a global problem (including New Zealand), by reference to papers at the recent Wind Turbine Noise conference which I attended. I also cite a case study I conducted which has some relevant pointers for Motorimu, as well as relevant literature.

12. A description is given of the history and shortcomings of New Zealand Standard NZS 6808, which is commonly relied upon to protect residents from excessive noise, but can be demonstrated to have failed.

13. I also provide evidence to complement Dr Phipps submission (S274 party).

4 Electricity Supply & Demand

14. The Motorimu proposal as now sought is described as 113 turbines and an installed capacity of 96.05MW. At a of 0.35 over the 8760 hours of the year, and deducting a 10% maintenance downtime, such a complex would generate approx. 270 GWH (Gigawatt-hours) of electrical energy, as compared to NZ's present usage of approx. 44,000 GWH per annum ( on the basis of Meridian's own figures obtained under OIA, Te Apiti had an overall capacity factor of 34% for 2005 and 36.5% for 2006, so 35% for Motorimu is realistic).

15. NZ's electricity demand is growing by approx 2% per annum, so the Motorimu proposal as sought, represents only 0.6% of NZ's present demand or only 3-4 month's worth of demand increase. An individual turbine represents 0.005% (50 parts per million) of present NZ demand, or 1 day's worth of electricity demand increase.

16. NZ has a grid connected installed capacity well in excess of 8,000 MW, while the average load is approx 5,000 MW. The record NZ demand peak of 6919 MW was in July 2007. This coincided with a cold, still period when the wind contribution to meet high demand amounted to less than 30MW or 16% of installed wind capacity. Thus at a time of peak demand the contribution of windpower was very poor. High pressure systems, leading to cold still nights and still days, are frequently developed across the entire NZ landmass, let alone just the Manawatu region. Thus peak demand tends to coincide with low windpower output.

17. The recent closure of Contact's obsolete in New Plymouth has been offset by Contact's planned 300 MW ' peaking plant', to be located near its Stratford power station site. A declared intent of this new plant is to provide 'rapid start' support to . This is necessary to maintain grid wide voltage and frequency control, given the known and documented risk of widespread and rapid wind power fluctuations.

5

18. From the above, it is clear that the Motorimu proposal ( in common with other wind farms) will not contribute significantly to NZ supply, NZ growth in demand or NZ security of supply. In Germany, for example, which has the highest installed capcity of windpower in the world, the major distributor, EON, considers that in order to meet increased demand, new thermal capacity must be installed at approx. 70% of the rate of windpower installation, to ensure security of supply, because of the uncertainty associated with windpower. In the Motorimu instance, the developer has claimed that the existing Huntly facility can provide backup to Motorimu's fluctuations, but in fact the old Huntly station is now firmly in the baseload category, and cannot generally be relied on in such manner. The new 'fast start' Contact 'peaking' station or the Genesis e3p would generally be required as backup. They have much smaller GHG emissions as discussed below.

19. Points 1.1 to 1.4 demonstrate there is plenty of scope to reject all of the additional turbines without any significant impact on NZ's security of electricity supply. The permitted 75 turbines have 2/3 of the energy generation and whatever GHG savings exist, of the 113 sought, so the impact of the additional turbines is trivial.

20. The developer has represented that the113 turbine proposal cannot be reduced as that will make the project unviable. I do not accept this. The same developer originally claimed that less than 127 turbines at Motorimu was uneconomic. The larger proportion of the capital expenditure on development will be in 'variable' rather than 'fixed' costs. In other words, most of the CAPEX is proportional to the number of turbines. Each turbine requires an access road, a foundation excavation, a concrete base, a tower, a turbine, a transformer etc. Similarly regarding operating costs, each turbine has its own maintenance costs. There are some fixed project costs ( ie not proportional to the number of turbines), such as the transmission line out of the site and the control station, but these are lesser.

6

21. In other words, deleting 30% of the 113 turbines and leaving it at the 75 permitted turbines would not make the project 30% less profitable. At a rough estimate it might reduce project DNPV relative to the 113 turbine option by around 10%. Of much greater impact on project DNPV are such variables as exchange rate and future electricity prices and the assigned value and amount of any carbon credits. 22. It would generally appear that the additional locations sought by the developer are in areas of higher windflow, but also higher visibility. All locations are not equal, and the 'better' sites from a windpower point of view tend to be the less desirable sites from an impact point of view. It is disingenuous to claim that the additional sites are necessary without referring to their impact relative to other sites. Each location should stand or fall on its own merits and demerits.

Global Warming, GHG Savings and Kyoto Targets

23. The analysis presented here is approximate but sufficient to demonstrate that on this project, as indeed on all wind farms, the overall GHG savings, as compared to the available economic alternative, are negligible or even less than zero in certain circumstances. As a general proposition, no wind farm developer should be able to claim GHG benefits unless he provides an independent 'full cycle' audit for his project.

24. It is a commonly held view that wind turbines are self-evidently greenhouse gas effective, and that substantial savings of GHG's will result from their installation. This view is founded more on perception than reality.

25. World total GHG emissions per annum are of order 27 billion tonnes CO2 equivalent. New Zealand, with 0.07% of world population, emits approx 77 million tonnes per year of GHG's', ie less than 0.3% of global emissions. The greater majority of these emissions come from the transport and farming sectors; the electricity generation sector being a lesser contributor at about 10% of the total.

7

26. To estimate the GHG savings of a wind turbine plant, one must consider what the economic and technical alternative to such plant would produce per annum. This would be a combined cycle gas fired unit such as has operated for a number of years at Contact's Stratford TCC plant, or Genesis Energy's recently commissioned e3p plant at Huntly, or Genesis' proposed Rodney combined cycle project north of Auckland, or most particularly the 300 MW 'peaking plant' now proposed by Contact to provide backup to wind generation in order to ensure security of supply and voltage and frequency stability, all of which are jeopardised by wind power variability.

27. Figures from the MfE (Ministry for the Environment) website show that the combustion of natural gas produces 52kg CO2 per GJ of gas (52 tonnes per Terajoule (TJ) of gas). A Megawatt-hour equates to 3.6 GJ of electricity. Thus the likely Motorimu annual output of ~270 GigaWatt Hours equates to 972 TJ of electricity. The Contact 'peaking' station would consume ~1,500 TJ of gas in producing this same amount of electricity.

28. Thus considering only the direct fuel replacement GHG savings of Motorimu, these amount to some 78,000 tonnes CO2, or 0.1% of NZ's total GHG emissions. Direct fuel savings of an individual Motorimu turbine would contribute less than nine parts per million to New Zealand's GHG credit account, or approx. 25 parts per trillion of global GHG credit. (I note Mr Faulkner's figures are more than twice this amount - because they are biassed by low efficiency, high GHG emitter coal fired generation, as at Huntly, rather than high efficiency gas fired combined cycle plant, which is the actual present economic alternative in NZ).

29. Even this very small GHG contribution is very significantly overstating the GHG savings, since it only covers the direct fuel usage. For a proper comparison, one must perform a 'full cycle' evaluation of the GHG's involved in the construction, maintenance and removal of the wind farm, as compared to the combined cycle alternative. On all such scores, wind farms compare badly,

8 because of their distributed nature, multiple small power units, and the unpredictable conditions in which they work.

30. The installation of a wind turbine requires extensive earthworks in the form of access roads and foundation preparation. It requires extensive concrete pour for foundations and extensive construction emissions per MW-hour of generation, as towers are erected and turbines, blades and geartrains are lifted into place. (Cement manufacture is a very heavy GHG emitter: approx 0.5 tonnes CO2 per tonne of cement; or nearly double this if the CO2 emissions of the sinter heat source are considered. There is virtually no potential to alleviate these emissions. For every million tonnes of concrete poured in construction, there will be cement associated CO2 emissions of 1-2 years worth of the predicted Motorimu fuel savings on CO2 emissions).

31. During the maintenance cycle, the distributed nature and difficult access to the turbines mean that enormously greater GHG's are emitted in their maintenance than in the maintenance of a single combined cycle unit (eg Genesis e3p) at an accessible and centralised location. At Motorimu, removal of a faulty blade or gearbox, for example, will require mobilisation and use of a high capacity crane, with its associated GHG emissions. At Te Apiti, 1/4 of the turbines suffered blade or gearbox failure within two years of installation, despite a 5 year warranty, and the manufacturer () is now carrying hundreds of millions of euros of warranty liability on its books. This demonstrates the inherent unreliability of wind turbines in the NZ wind environment.

32. Due to the varying loads on the turbine and gear train and blades associated with wind variability, wind turbines are high maintenance and have a short operational life, as compared with combined cycle units. Not only are standard maintenance costs and GHG emissions of a centralised, accessible combined cycle unit (eg Genesis' e3p) far less per unit of electricity produced, but since turbines and gear trains work best at uniform rates, they last much longer. Wind turbine manufacturers typically place the life of their units at around fifteen years, whereas combined cycle units can have three times that life.

9 33. Windpower is only 'sustainable and renewable energy' providing one continues with what, compared to the economic alternative, is excessive maintenance and frequent replacement to keep in operation, involving many times the GHG emissions of the economic alternative. The only aspect of windpower that is sustainable is the wind - ie the fuel - all else is not!

34. The backup and spinning reserve capacities are also real costs of wind generation and should be counted on wind's GHG debit ledger. Standby offline rotating capacity must be available at all times to provide backup if wind output drops. This does not come for free. If backup is hydro it can be 'dry run', ie by rotating the generator as a motor, with penstock closed, but that requires power input. One cannot count one's fuel twice!

35. Estimates of these GHG debits vary from protect to project, depending on individual infrastructure and reliability factors, but are typically in the many years range of fuel associated CO2 savings (certainly not the two days quoted at last year's West Wind hearings, which would be insufficient to even power the bulldozers to complete basic earthworks). Typically, construction GHG emissions alone are estimated at between 6 months and two years, depending on the difficulty of access and the earthmoving requirements. Motorimu would lie at the high end of this range, given the site topography.

36. Whatever, the precise figure, it is too small to claim that the Motorimu proposal, either in its presently permitted 75 turbine form or in its sought 113 turbine form, or indeed any NZ wind project, will contribute significantly to New Zealand's Kyoto obligations.

37. In conclusion, the actual GHG savings associated with the entire Motorimu project, as proposed, lies between zero (or negative, depending on reliability) and 0.1% of New Zealand's total GHG emissions.

38. The Government's NZ Energy Strategy, released in October, seeks to achieve a 90% dependence on 'renewable' energy resources by 2025. At present, 'renewables', ie (hydro, wind and geothermal) account for 70% of NZ supply.

10 The identified, undeveloped geothermal resource alone could ensure the meeting of that goal. Tidal and wave projects are actively under consideration (eg Crest Energy in the Kaipara Harbour - a project five times that of Motorimu), and their energy potential in NZ is far greater than wind. NZ is also a good location for solar energy systems.

39. In the longer term, ie by 2050, nuclear fusion will become the baseload supplier - fail safe, no long lived radio-nuclides, limitless fuel supply. At present nuclear fission could provide that potential for NZ. After all, 'la belle France', the world's number one tourist destination, generates most of its power from nuclear fission, safely, economically, cleanly and in a GHG effective manner. A base- load modern technology nuclear power station north of Auckland would undoubtedly be the best means in the immediate term of reducing the huge line losses in NZ's transmission system, and meeting the Auckland region electricity market - nearly half that of NZ. However, NZ public policy has so confused the nuclear weapons and nuclear power debate as to make that practical option a distant goal.

40. At the end of the day, all New Zealanders need to recover the age old virtues of prudence and personal responsibility in energy use. I grew up in a generation where to leave food on one's plate was tantamount to a crime. New Zealanders individually need to rediscover such simple virtues, and learn to switch off lights in rooms they are not using, only put the required amount of water in their kettles and switch off their computers when not in use. And insulate their homes effectively, adopt passive solar heating and other simple solutions, which would save far more than Motorimu could ever generate. In this respect EECA has so far failed to lead NZ - hence its slavish devotion to windpower, to cover its inadequacies.

41. In all of this, windpower has a role to play, albeit relatively minor. It has its virtues and its drawbacks, but it is certainly not a solution to demand growth or GHG emissions, and should not be represented as such by developers keen to jump on the 'windrush' wagon. Certainly not as an automatic justification for imposing significant lifestyle, economic and health impacts on nearby residents.

11 The report of the Parliamentary Commissioner for the Environment ( Wind Power, People and Place, November 2006) fully recognises this, and rejects the large power complex approach which has typified the land grab mentality of NZ wind developers to date.

42. From a power supply and Kyoto obligation perspective, there is no basis for approving the additional 38 turbines over the 75 presently approved for Motorimu.

Noise from wind turbines

43. I show below the complete list of papers presented at the recent 2nd International Conference on Wind Turbine Noise, attended by several hundred delegates, as evidence that wind turbine noise is the subject of much debate and research. From attending all the papers, I can say that the problem of noise impact is widespread and that there is significant uncertainty in the estimation of wind turbine noise, due to atmospheric and terrain complexities. Impacts on nearby residents range from mild annoyance through irritation, sleep deprivation, health effects and loss of lifestyle and concomitant loss of quality of life hence of property value, all correlating with a dose relationship of sound relative to background. Several of the more pertinent papers are highlighted in yellow. A CD of these papers is provided with this submission.

12

Proceedings of Wind Turbine Noise

Held in Lyon, 20/21 September 2007

The Conference was organized by Dr Geoff Leventhall On behalf of INCE/Europe

Authors' Abstracts

Lyon - France - 20th and 21th September 2007

Second International Meeting on Wind Turbine Noise

Theoretical analysis of wind turbine used to power a stand-alone solar desalination unit in selected coastal areas of Egypt Zeinab Sayed Abdel-Rehim Long range sound transmission over the sea with application to wind turbine noise Mats Åbom and Mathieu Boué In-Home Wind Turbine Noise Is Conducive to Vibroacoustic Disease Mariana Alves-Pereira Masking of wind turbine sound by sea waves Paul Appelqvist A Variety of Wind Turbine Noise Regulations in the United States - 2007 James D. Barnes Investigating the audibility of wind turbines in the presence of vegetation noises Karl Bolin What is the real background noise? Dick Bowdler Wind Farms Noise Predictions and the Risks of Conservatism Andrew Bullmore Living with aerodynamic modulation, low frequency vibration and sleep deprivation - how wind turbines inappropriately placed can act collectively and destroy rural quietitude Julian.T.Davis Review of post-construction compliance assessment conditions in various wind farm planning permits from New Zealand and Victoria (Australia) Christophe Delaire Applicability of TLM to Wind Turbine Noise Prediction Guillaume Dutilleux Recommendations for an improved quality of the acoustics of wind farm projects Pierre Dutilleux, Joachim Gabriel Domestic Wind Turbines Gwyn Mapp Low Frequency Noise from Wind Turbines Malcolm Hayes

13 Assessment of Sound and Infrasound at the Pubnico Point Wind Farm Brian Howe Towards a review of NZ Standard NZS6808 : 1998 Acoustics - Assessment & Measurement of Sound From Wind Turbine Generators Malcolm Hunt Passive and Active Dynamic Vibration Absorbers for Gear Box Noise Reduction in Wind Turbines Illgen, A.; Drossel, W.- G.; Wittstock, V.; Neugebauer, R. Noise prediction of a new 34 MW wind farm Antonio Iannotti, Fabio Serpilli An Approach to RANS Based Prediction of Airfoil Trailing Edge Far-Field Noise M. Kamruzzaman, Th. Lutz and E. Krämer A new mechanism in vawt Prof.R.V.Sharma, Ravi Kumar Singh, Nitin Kumar Advanced methods for online vibration monitoring of wind turbines Patrick Labeyrie Aerodynamic noise from micro wind turbines : Current situation and future perspectives Damien Leclercq, Con Doolan Auralization and assessments of annoyance from wind turbines Søren V. Legarth Evaluating the potential health impacts of wind turbine noise for environmental assessments Dr. D.S. Michaud, Dr. S.E. Keith Residual loudness of wind turbine sound in the presence of ambient sound David A. Nelson Prediction of wind turbine noise and comparison to experiment Stefan Oerlemans and Gerard Schepers Uncloaking the Nature of Wind Turbines - Using the Science of Meteorology William K. G. Palmer Wind farm perception – a study on acoustic and visual impact of wind turbines on recidents in the netherlands Eja Pedersen, Jelte Bouma, Roel Bakker, Frits van den Berg SIROCCO: Silent Rotors by Acoustic Optimisation G. Schepers, T. Curvers S. rlemans K. Braun, T. Lutz, A. Herrig, W. Wuerz A. Matesanz, A. Gonzalez Calculate noise of wind-farms Sénat, Claude; Garrigues, Sébastien; Gamba, René Low frequency Noise from Large Wind Turbines Bo Søndergaard Wind Farm Noise and Regulations in the Eastern United States Hilkat Soysal, Oguz Soysal Wind profiles over complex terrain Frits van den Berg Taking into account of atmospheric conditions for a spatio-time localization of the aerodynamic sources on a moving blade by the method of acoustic imagery Dominique VINCENT, Vincent BENOIT

44. Standard NZS 6808, together with NZS 6801 & 6802, are used as the basis for measuring wind turbine noise and for assessing whether or not it is 'acceptable'. For a variety of reasons NZS6808 (1998) is not consistently satisfactory in properly protecting residents from significant adverse impact from wind turbine noise. Actual case studies from Gebbe's Pass and Te Apiti have shown that. I will discuss the Gebbe's Pass situation.

45. It is important to understand the underlying physical principle on which NZS6808 is based, in order to appreciate its drawbacks. The principle of sound screening is assumed, whereby turbine noise is masked by background noise. Thus the

14 basic criteria of NZS 6808 that enables the developer to continue to operate if he meets 40 dba sound power immission or 5 dba above background, whichever is the greater. A 5dba penalty applies if tonal or 'special audible characteristic' noise (not defined) is detected

46. Noise at a location is generated across the relevant wind speed range, and sometimes in different times of day/year and wind direction, as an average of immission power level over a ten minute interval. Many such observations are collected ( 2-3 weeks worth is typical) and the resulting values are plotted as a scatter diagram of sound power relative to wind speed. A regression line through this scatter is drawn, and this is used as the reference line for background noise and (subsequent to a wind plant commencing operation) to background plus turbine noise. An example is shown below (Fig 1):

47. The plot demonstrates the double averaging that has occurred, first over a ten minute interval for each point, and second as a regression line through the points. Clearly a lot of the time the background noise at any wind speed is above the regression line, and equally a lot of the time it is below it.

48. Now this might not matter much where there is little observational scatter. Across the flat and boring plains of North Germany and Denmark, a uniform and boring wind blows from the Baltic. In such circumstance there could be a good correlation between variation in wind induced noise at the receiver and noise power output from the turbine, and the masking effect would then work well. However, that is not the norm in NZ's hilly and gusty environment.

15

49. Figure 5, also taken from the Bowdler paper, shows that the correlation between wind speed (hence background noise level) at a sheltered site to that at a turbine site (where noise output from the turbine increases with increasing wind velocity), can be very poor. Thus the underlying principle on which the Standard is based, is not being satisfied - the noise output from the turbine can frequently be high when the background masking noise is very low.

16 50. This situation commonly prevails in New Zealand. Typically turbines are sited on the exposed ridges, while residents live in sheltered valley situations.

51. This has been demonstrated at Te Apiti by both Paul Botha of Meridian and by Malcolm Hunt Associates in their Stakeholder Review of NZS 6808, prepared for EECA and the NZ Wind Energy Association in May 2007 Ref22.0001.06.04(C) / 2006622W 001). The example shown below is based on data from observation stations near turbines at Te Apiti. The second graph shows the wide scatter of background noise with respect to turbine site wind speed, hence that much of the time the noise from the turbine is not adequately masked.

52. Much of the time the turbine noise will be 10 dba above background. As the report suggests (below Fig 6) NZS 6808 as it stands will not be appropriate at such sites.

17

53. This is by way of introduction to the case study I will quote at Gebbe's Pass / McQueen's Valley on Banks Peninsula. Windflow Technology's single 500 kW turbine is located on the ridge line at Gebbe's Pass on Banks Peninsula, with nine residences situated below it and 1500m way to SE in McQueen's Valley. the turbine is subject to a BPDC consent which requires it to maintain a tighter noise emission standard than NZS6808

54. The valley has dimensions similar to that of the Kahuterawa Valley, being long and narrow and situated between ridgelines which rise a couple of hundred metres above the valley floor. As at McQeen's Valley, the Kahuterawa Valley lies downwind of the turbine(s) on the ridge above.

55. On 26 June 2005, I visited McQueen’s Valley, to interview the residents. Ambient conditions in the valley are typically quiet and still even when wind is

18 clearly blowing across the ridge. My notes, (with names removed): Common feature of all parties (independently) were:

• Their response to the question ‘when the turbine is operating, how much of the time do you hear it?’ All independently said ‘nearly all of the time’ _ All commented on the quiet environment of the valley and the manner in which sounds can carry, ie voices from a long distance down the road.

• All noted that it could be still and quiet in the valley when windy on the ridge.

• All parties distinguished blade noise from gearbox noise.

56. Some individual comments:

Party A • _ Advised that Marshall Day had represented to the Banks Peninsula District Council that the Gebbe’s Pass turbine was very quiet, and that by comparison the Te Apiti turbines (Vestas V72) made a noise 'like a Boeing 747'. _ BPDC officials now reported as saying they now regret granting the resource consent because it has become a full time job to monitor it.

Party B • Described noise as a fading and rising Whump Whump sound • Said that when wind noise level drops suddenly the turbine noise can startle you ‘like a helicopter noise suddenly leaping out or descending above you’ • Said ‘you can’t escape the noise — its constant’ • Said (wrt Makara wind farm) ‘if they put that project in, your life will not be worth living’ • Said “far, far more irritating than we ever imagined before it started.” • Said could hear it outdoors ‘all the time’ when it was operating.

19 • Cannot hear it indoors. • Has stereo on when working in garden, to drown out the noise. Said when turbine is rebuilt and starts again, has the option of moving, since is renting, but that is more difficult for landowners. • Said BPDC was now much more considerate of their concerns than before; now they have realized (eventually) that there are genuine problems. However, BPDC has said there is nothing they can do because the resource consent has been granted, and if they keep within the noise standard WFT can continue. The resource consent is for 10 years. • Mentioned that University of Canterbury did some trials, simulating noise from the site and they could not hear 4OdBA from that experiment, but can definitely hear the turbine, especially at low wind speeds. • Mentioned a property (brick house) further down the valley at 2.5km range, without line of sight to the turbine, where noise was clearly heard, although it is not heard in an adjacent, closer house (stucco). Also mentioned noise heard by residents in Gebbe’s Pass (near Wantage) at 2.5 km range. ( these points also noted by others)

Party C • Said the noise was most irritating typically in late evening, night and early morning, when the stillest conditions prevailed. • Described noise as a whooshing sound, rising and falling in intensity

Parties D & E • Said it had affected property values, as they are trying to sell and had had several parties say when they saw the turbine that they were no longer interested because of the noise. • Also mentioned that the landowner where the turbine is situated is having difficulty selling. Also mentioned the affected homeowner near Wantage • Party D description of sound: whoo whhoo Whoo sound, rising and falling in intensity. Party E description: vrum Vrum vrum, rising and falling in intensity.

20 • Both compared it to noise of chainsaw about 500m away across valley, clearly heard. Party D said not quite as loud, Party E said it was as loud. • Party E says can hear noise inside the house, with windows closed, and on quiet nights it keeps the person awake. Also mentioned that one of the dogs howls whenever the turbine operates.

Party F • Reportedly the most affected person — health has suffered due to stress from noise. Trying to sell. Not at home so did not interview. • All parties mentioned that other landowners that I did not interview held similar views. I note that these were not, by any means oddballs — they were all rational professional people with responsible jobs, stating their experiences.

57. The purpose here is not to criticise Windflow Technology, it is to draw generic conclusions regarding what is causing the problems. Particularly at night, in the evening and early morning, wind speed at low levels drops (especially in confined valleys such as Kahuterawa but also on the plains around Linton and Tokomaru). The ratio of low level wind to high altitude wind is known generally to decrease at such times, especially if a temperature inversion occurs (common on still, clear nights), and the correlation between windspeed (hence ambient noise) at a residence as compared to wind speed ( hence poise power emission from turbine) is poor.

58. Sound can get trapped at ground level by internal refraction within an inversion gradient, hence can be amplified relative to other times. In confined valleys the sound can be further amplified in various positions by the formation of standing or slowly propagating sound waves.

59. Changes in the wind regime are not necessarily synchronous at observer and turbine. Hence the ' helicopter suddenly descending' type descriptions. Noise from the turbine increases but the background noise field at the observer does not. A wind gust on the ridge can precede a wind increase in the valley by two or three minutes, yet the increase in noise output will only take five seconds to

21 reach the residents.

60. It is the nature and continuity of the noise which causes distress, as much as it is the level. Despite that developers and regulators maintain that the noise from turbines is not distinctly different from background at affected sites, nevertheless in this low frequency range it is detectable as a varying intensity and tonally distinct presence.

61. The noise is undoubtedly dominantly trailing edge turbulence, which from studies reported on at the Lyon conference, typically peaks as a burst of noise off the trailing edge of the downgoing blade, as it is braked by slower air as it nears the ground. This explains the repetitive thumping sound as a burst of broad spectrum noise emitted at the blade passing frequency. It is typical of descriptions of wind turbine noise.

62. It is of interest to note that the turbine is considered by BPDC to stay within the limits of its consent, ie more stringent than NZS 6808. Yet the residents are significantly adversely affected. It is the repetitive nature of the sound that is disturbing. The time-dose relationship of the characteristic noise is what causes stress, despite that supposedly the turbine stays within its allowed noise immission level. The human ear picks up the special audible characteristic, which is not adequately masked by background sound.

63. Thus we can certainly say that a substantially more stringent standard than NZS 6808 does not work in this instance. The relationship of windy ridge to quiet valley, and the poor correlation between wind speeds in the two locations, mean that masking is inadequate.

64. The paper by D Nelson at the Lyon conference graphically illustrated this. A drawback of standards like NXS 6808 is that they consider noise power across the entire frequency spectrum of interest. However, turbine noise is frequently tonal, ie narrow bandwidth, and commonly modulated, ie varying in intensity on a repetitive basis, as in the example shown. The Nelson paper shows it is necessary to have adequate masking noise in the frequency band of interest,

22 and that just because background noise across the spectrum is sufficiently close to turbine noise across the spectrum, it does not mean that masking of particular tonal or special audible characteristic sound will be adequately achieved.

65. In a demonstration at the conference, Nelson played through the conference speakers, noise recorded from a field turbine, normalised to an immission of 40 dba and containing a tonal peak. The use of a 41 dba background mask of ambient noise at 41 dba was clearly inadequate in masking the characteristic sound of the turbine, which even could be heard through an ambient mask of 49 dba. Yet NZS 6808 allows sound power immissions of 5 dba above ambient as satisfactory.

66. For these and other reasons not covered here due to complexity, NZS 6808 does not work adequately in the typical NZ environment

67. We can expect that if the Motorimu proposal proceeds as sought, the residences in Kahuterawa Valley would be particularly affected, due to its generally downwind and confined location. However, Linton/Tokomaru area residences, while generally upwind, will also be significantly affected, particularly in easterly winds and also in conditions of temperature inversion. Even when they are upwind, refraction of sound in the upwind direction will be such that some residences will lie around the 'skip distance', the zone where the sound wave front impacts on the ground. Thus residences at certain ranges may be quite badly affected while others at slightly greater or lesser, will not. There is an interesting summary of this phenomenon in New Zealand Standard NZS6801.

Conclusions

68. So what is to be done?

• The residents can reasonably expect to be protected from significant adverse noise and that the Council regulatory officers will ensure an appropriate standard is met

23 • The Council wants an appropriate standard to apply.

• The developer needs to know the standard he must work to.

• Both Council and developer have a duty of care with respect to the residents.

• The noise standard, and even more stringent conditions such as those applied to Project West Wind ( not yet tested in practice) are unlikely to work satisfactorily in the topographic and atmospheric conditions around Motorimu.

69. Clearly, a precautionary approach needs to apply. But what is it?

70. At McQueen's Valley, an offset of 1500m was inadequate.

71. At Te Apiti there have been significant issues at a range of 2400m.

72. The French National Academy of Medicine in its report of March 2006 "Repercussions of wind turbine operations on human health" called for a moratorium on all turbine construction within 1500m of homes

73. Thus I consider a 1500m offset from homes is the minimum distance that could be considered, and that does not guarantee that there will not be instances of significant adverse noise impact.

74. In all other respects with regard to the Noise Conditions, I support the recommendations of Mr Bob Thorne.

75. The Noise Conditions applied to Project West Wind have yet to be tried in practice. Yet already we are experiencing the reluctance of the developer to conduct adequate pre installation noise monitoring. Their modelled 35 dba contour, within which pre installation monitoring is to be conducted, is not

24 accurate to better than 3 dba, yet Meridian have made no allowance for computational uncertainty, in order to limit their pre installation monitoring, hence leave residents who may be subsequently adversely affected without means of redress.

76. The special noise condition which applies in West Wind when background is below 25 dba is too stringent, and should be lifted to 30 dba.

David Bennett 29 February 2008

25