New Zealand Windfarms Limited v Palmerston North City Council
[2013] NZHC 1504
•20 June 2013
IN THE HIGH COURT OF NEW ZEALAND WELLINGTON REGISTRY
CIV-2012-485-1503 [2013] NZHC 1504
BETWEEN NEW ZEALAND WINDFARMS LIMITED
Appellant
ANDPALMERSTON NORTH CITY COUNCIL
Respondent
Hearing: 4 February 2013
Counsel: J B M Smith and M J Slyfield for Appellant
J W Maassen and N Jessen for Respondent
Judgment: 20 June 2013
JUDGMENT OF WILLIAMS J
In accordance with r 11.5, I direct the Registrar to endorse this judgment with the delivery time of 4.30pm on the 20th June 2013.
Introduction
[1] The appellant was granted consent by the respondent to build a substantial wind farm of up to 97 wind turbines. By the time the project was two-thirds built, it was found that the information filed with the appellant’s application for resource consent had significantly under-predicted the actual noise generation characteristics of its wind turbines and their noise impact on surrounding residents. This fact underpins the current controversy.
[2] The resource consent as finally granted contained:
NEW ZEALAND WINDFARMS LIMITED v PALMERSTON NORTH CITY COUNCIL [2013] NZHC 1504 [20 June 2013]
(a) a general condition requiring the wind farm to be constructed and operated in accordance with the information contained in the appellant’s resource consent application; and
(b)specific conditions setting standards in relation to allowable noise levels received at sensitive residential locations nearby, and monitoring conditions.
[3] There is no proof that the specific noise levels in the consent conditions were, or are being breached. Monitoring is ongoing to determine that question. The appellant accepts however, that noise generated by the wind farm is greater than was predicted in the application and that residents are also affected to a greater degree than predicted.
[4] The primary issue in the appeal is whether the appellant is bound by both its own predictions about the sound levels generated by its turbines and the specific noise standards contained in consent conditions or whether it is bound only by the specific standards.
[5] The Environment Court found that the appellant was bound by both. In essence the appellant challenges that finding. It argues that only the specific noise conditions bite. The respondent cross-appeals against an aspect of the reasoning of the majority of the Environment Court. I will explain the details of the cross-appeal during the course of this judgment.
Facts
[6] New Zealand Windfarms Limited (NZWL) operates Te Rere Hau wind farm in the exposed hill country east of Palmerston North pursuant to a consent granted in
2005. Sixty-five of the 97 consented wind turbine generators have been erected and are operational.
[7] The turbines are Windflow 500 turbines. They are a new design built especially for the appellant and are only in use in Te Rere Hau wind farm. Prior to Te Rere Hau wind farm, only one prototype of the Windflow 500 existed. It was
situated in Banks Peninsula. This means the turbine had no real operational track record before Te Rere Hau.
Conditions 1, 4 and 5
[8] The consent contained condition 1 – the standard form condition seen in most resource consents in New Zealand requiring the project to conform with application documentation. In this case the condition required the wind farm to be constructed and operated generally in accordance with said documentation. The material part of condition 1 was in these terms:
1.The proposed Te Rere Hau Wind Farm be constructed and operated generally in accordance with all the information, site plans and drawings accompanying the application or submitted as additional information.
[9] The application included an Assessment of Environment Effects (AEE) as required by s 88 RMA, and other additional information. Attached to the AEE, was a Noise Impact Assessment Report (NIAR) prepared by NZWL’s acoustic engineer, Mr Malcolm Hunt. The NIAR predicted that the Sound Power Level (SPL) from a single Windflow 500 wind turbine would be 100.7dBA measured at the turbine itself. This figure was derived from tests Mr Hunt carried out on the Banks Peninsula prototype of the Windflow 500.
[10] SPL needs to be understood. It is the level of sound ‘energy’ created at source by a turbine. It measures the power of the sound the turbine creates rather than its audible noise. The ear cannot hear sound power although it can certainly feel its effect. It is different to Sound Pressure Level which measures the audible noise received by the ear. Sound Pressure Level (A-weighted or adjusted to the human audible range) is the metric used in standard noise control consent conditions.
[11] The NIAR also confirmed that the turbines would produce no Special Audible Characteristics (SACs). This prediction was also based on testing Mr Hunt carried out on the Banks Peninsula prototype. SACs are audible tones or impulses – sounds that are acoustically significant not because they are loud, but because they are annoying to the human ear. They include high pitched whines, hums, buzzes, oscillating sounds and so forth. A sound that contains SACs may arouse an adverse
community response at a lower level than a sound containing no SACs. To account for the adverse subjective response caused by SACs, predicted or measured sound pressure levels are given a 5dBA loading when SACs are present in the acoustic characteristics of the noise in question.
[12] So, according to Mr Hunt’s predictions in the consent application, the SPL at source would be a little over 100dBA and there would be no SAC loading. On the face of it, these predictions were information accompanying the application and were therefore caught by condition 1.
[13] Condition 4 of the consent set the noise limit measured, as I have said, as a Sound Pressure Level, at either 40dBA or 5dBA above background noise (whichever is the higher) measured at the notional boundary1 of any receiving dwelling. Condition 5 set out the monitoring processes by which to measure compliance with condition 4.
[14] As I have said, the NIAR predicted that the turbines would produce
100.7dBA of sound power at source. That prediction was wrong. Indeed the Environment Court described it as “wildly incorrect”. The production versions of the Windflow 500 turbines did not perform anywhere near as well acoustically as the Banks Peninsula prototype had.
[15] In 2009, a number of complaints were received from Te Rere Hau wind farm’s neighbours about noise effects (the tally is now 800). NZWL investigated. Four representative turbines were tested. Their SPLs at an 8m/s wind speed were
103.3dBA, 104.9dBA, 105.6dBA and 106.4dBA respectively.2 In addition, the
turbines were in fact generating SACs audible at 50m from source, contrary to Mr Hunt’s predictions, although there remains an unresolved question around whether the SACs are in fact audible at points further afield.
[16] In practical terms, an increase in SPL will lead to an equivalent increase in
Sound Pressure Level at a common receiving point all other factors being constant.
1 A line drawn 20m out from the dwelling in the direction of the noise generator.
2 Palmerston North City Council v New Zealand Windfarms Ltd [2012] NZEnvC 133 at [46].
Factors which can distort this otherwise arithmetic relationship include distance, terrain, shielding, wind direction and ground cover.
[17] I was provided by counsel with a kind of “dummies’ guide” to understanding decibels. It was provided to counsel by Marshall Day Acoustics, a consulting acoustic engineering firm. It contained the following explanatory ‘rules of thumb’:
1. A “just noticeable change” in sound pressure level is 3 decibels.
2. A “subjective doubling of loudness” is approximately a 10 decibel
increase.
3. Doubling the number of noise making devices generally results in a
3 decibel increase.
4. Moving twice as far from a noise source generally results in a
6 decibel decrease.
[18] It can be seen that, in combination, these SPL increases were potentially significant, especially when combined with a possible 5dBA loading for SACs.
Applicable standards
[19] NZS6808 is the New Zealand standard in relation to noise from wind turbines. The standard was promulgated by a technical committee comprising representatives from the wind farm industry, government and local government. Its purpose is to aid wind farm development and local authority planning procedures by:
(a) providing a method for the measurement and assessment of sound from turbines; and
(b)establishing recommended limits of acceptability for sound received at residential locations.
[20] According to cl 4.3.2 of the standard, the following mathematical equation is the means by which noise levels at any given distance from source are predicted:
LR = Lw – 10 Log (2πR2) – ΔLa ……………………………(Eq.1)
where
LR = the sound pressure level from a single WTG at 1.2 m to 1.5 m above local ground level in dBA at distance R.
Lw = the sound power level of the WTG in dBA. Measured according to IEA (International Energy Agency) procedures relating to WTG sound measurement or IEC DIS 1400-11.
R = the distance between the source and the receiver in metres.
ΔLa = a R.
a = attenuation of sound due to air absorption, dBA/m for broad band sound which is typically 0.005 dBA (refer ISO 9613-1). This value is dependent upon the spectral character of the sound and the atmospheric conditions.
[21] It will be seen that the SPL of the turbines (Lw) is the starting point for this calculation. Note also that no allowance is given for other variables such as terrain, shielding ground cover, wind direction and so forth. Rather, the standard requires a straight sound power by distance calculation. In this way, the standard intentionally over-predicts noise levels in order to err on the conservative side to protect surrounding acoustic amenity.
[22] The standard also provides a “guide to acceptable limits” at receiving residential locations. The aim of the recommended limit is to achieve indoor noise levels of 30-35dBA L95 at any affected dwelling. The limit is therefore set at 40dBA L95 outside the residence. L95 is the noise level at the particular location that is exceeded 95 per cent of the time. It is a measure of background noise rather than maximum noise levels. It is designed specifically, I understand, for measuring levels of continuous industrial type noise of the kind in question here. The recommended acceptable limit is set out as follows:
4.4.2 Acceptable limit
As a guide to the limits of acceptability, the sound level from the WTG (or windfarm) should not exceed, at any residential site, and at any of the nominated windspeeds, the background sound level (L95) by more than
5dBA, or a level of 40dBA L95, whichever is the greater.
[23] So if the background sound level (without the wind farm) is less than 35dBA, then the applicable limit with the wind farm is 40dBA, but if the background level is (say) 43dBA, then the upper limit is 48dBA, i.e. 43dBA plus 5dBA.
[24] NZS6808 is not entirely New Zealand generated. Rather it draws on extensive international experience in the measurement and regulation of wind farm noise, particularly, as I understand it, in Europe. The New Zealand standard is based to a large extent on the international standard IEC61400-11. IEC is the International Electrotechnical Commission. The Commission is the world's leading organisation for the preparation and publication of international standards for electrotechnology. The Commission utilises experts from industry, commerce, government, test and research labs, academia and consumer groups in order to produce acoustic standards having international application. The Commission has 60 full member countries and
22 associate member countries. New Zealand is a full member.
[25] I understand that wind farm consent decisions in New Zealand have, for the most part, relied on the NZS6808 levels and methods although I also understand there have been exceptions.
[26] The standard requires careful pre-consent monitoring of sound levels at local homes where sound levels are predicted to exceed 35dBA after the wind farm is established.3
[27] When preparing the appellant’s consent application and in order to identify the residences that could be affected, Mr Hunt modelled L95 background noise contours around the wind farm at 45dBA (closest to the wind farm) 40dBA, 35dBA and 30dBA (furthest away). He then plotted the homes within the 30dBA contour. Because he underestimated the sound power generated by the production versions of the Windflow 500 turbines, he used an incorrect SPL input. The modelling therefore wrongly predicted that only three residences had background noise levels of 30dBA or more, making it possible that at those sites, the wind farm might add enough extra noise to breach NZS6808. It seems that modelling in fact made three key errors:
(a) first, it underestimated the SPL generated by the turbines;
3 NZS6808 cl 4.5.1.
(b)second, it included the attenuating effect of topographical screening when the standard said it should not and in any event, it overestimated the reduction; and
(c) third, it excluded SACs.
[28] The effect of the mistakes NZWL made in calculating topography, SPLs and SACs was among other things that the actual 30dBA contour was considerably larger than NZWL had predicted incorporating many more homes. The Environment Court described matters this way:
The conclusion contained in the NIAR that only three local residential locations would be affected by receipt of sounds at levels of 30dBA or more has proven to be wildly incorrect. The acoustic experts agreed that this is due to a combination of the increase in the sound power level generated by the turbines installed (on average about 5 decibels higher than stated in the NIAR) and an overestimation of topographical screening (in the order of 5 to
7 decibels). At the request of the Court, NZWL produced amended noise contour maps showing the noise contours generated using a sound power
level of 105.7dBA, with and without topographical screening. Taking the
most conservative scenario of the amended sound power level and no topographical screening approximately 30 residences are shown within the
30dBA contour line and 16 of these are also within the 40dBA contour line.
NZS6808:1998 specifies the equation to be used for the calculation of the outdoor sound level with distance from the source. The Standard notes that this equation does not take into account attenuation due to screening effects where there is no line of sight between the turbine and the receiver locations. The acoustic absorption and reflection effects due to vegetation and ground cover are also ignored. (my emphasis)
[29] This meant that Te Rere Hau wind farm was acoustically far more problematic than the respondent and surrounding land owners had been led to believe during the consent hearings. Far from being a foregone conclusion, compliance with condition 4 and NZS6808 from which it was derived was, and remains, a matter of controversy.
[30] It is not yet known if the condition 4 upper limit of 40dBA or background and
5dBA is being breached. Initial calculations by Mr Halstead, the current acoustic engineer for NZWL, suggested that some down wind conditions (i.e. wind blowing from an SSE direction) did produce breaches of that standard at one property, but
subsequent corrections by NZWL suggested that may have been wrong. Monitoring continues.
[31] In light of the increasing clamour over acoustic problems, the respondent applied in October 2011 to the Environment Court for a series of declarations, the most relevant of which was a declaration that the wind farm was being operated in breach of the scope of consent condition – condition 1. The respondent did not apply for a declaration that conditions 4 and 5 had been breached. Nor did it issue abatement notices. The question of whether the 40dBA or background plus 5dBA levels have been breached remains at large.
The Environment Court decision
[32] The majority of the Environment Court (comprising Commissioners Beaumont and Bunting) held that the wind farm is being operated in breach of condition 1 because, in combination:
(a) the turbines have a SPL 5dBA higher on average than was predicted in the NIAR;
(b)the turbines are generating SACs, contrary to statements contained in the NIAR; and
(c) the noise from the wind farm being received at local residential locations exceeds the levels predicted in the NIAR.
[33] The presiding officer, Judge Dwyer, disagreed with that reasoning. He decided that (a) and (b) were sufficient to constitute a breach of condition 1 whether or not the levels produced by (a) and (b) created situation (c), i.e. higher than predicted levels at local residential locations. His view was that the exceedences at (a) and (b) on their own meant that the wind farm was not operating generally in accordance with the terms of the application, and condition 1 was therefore breached. I note in passing that the respondent sides with the learned Judge on this question. Its cross-appeal challenges the majority’s reasoning here.
[34] The Environment Court granted the application and issued a declaration in these terms:
Condition 1 of the resource consent is being and has been breached by the respondent in that Te Rere Hau (TRH) wind farm has been operated in such a way that the noise effects at local residential locations are considerably greater than those predicted in the application.
[35] The Environment Court’s three key reasons can be summarised as follows. First, the rule that a specific condition prevails over a general condition when they are in conflict did not apply here. There is, the court said, no conflict between general condition 1 and specific conditions 4 and 5. All three conditions co-exist as separate parameters of the resource consent. It is useful to set out the court’s
reasoning in full on this point:4
The information contained in the NIAR as to sound power levels and SACs relates to noise generated at source. Conditions 4 and 5 address the noise to be received by neighbouring properties. That is a different matter to the noise generated by turbines at source.
…
We appreciate there is a direct relationship between the noise generated at source and the noise ultimately received by neighbouring properties. The information contained in the NIAR as to the sound power level of the Windflow 500 was what led NZWL to the prediction that only three neighbouring properties would receive noise at levels higher than 30dBA from TRH. The increase in sound power levels of the TRH turbines (as compared to the sound power levels stated in the AEE) has led directly to TRH having noise effects over a substantially wider area and on more residential locations than was predicted in the AEE. In our view there is no inconsistency or conflict between Condition 1 and Conditions 4 and 5 at all. Condition 1 (and the NIAR) identifies the means (restricted sound power output and absence of SACs) by which NZWL predicted it would be able to meet the noise requirement now contained in Conditions 4 and 5.
[36] The court’s second reason was that the AEE is “the bedrock upon which resource consent applications are founded”. The “need for accuracy and integrity in the application documents is self evident in the court’s view.”5 In particular, the predictions contained in the NIAR were presented as the product of scientifically based calculations that could be relied upon by the consent authority in determining
the effects of the proposal. NZWL used the results of those calculations to make the
4 At [92] and [93].
specific prediction that only three local residential locations would receive noise in excess of 30dBA. It would, the court said, have “very wide” implications for the administration of RMA consents if NZWL were not held accountable for that prediction.6
[37] The court’s third reason (this time by majority) was that NZWL could change the operating protocol of the wind farm to compensate for the increased SPL of the turbines and achieve the level of sound received by local residential properties that was predicted in its application. The Commissioners were critical of NZWL’s failure to do this.
[38] The Environment Court also issued a declaration that the Council was quite separately entitled to conduct a review of the noise conditions applicable to the wind farm under s 128(1)(c). Three requirements must be established for s 128(1)(c) to apply. First, the information made available to the consent authority must contain inaccuracies. NZWL accepted that the NIAR contained inaccurate statements in relation to SPL and sound propagation. The Environment Court found that these inaccurate statements lead to further errors, namely:
(a) miscalculation of noise contours;
(b)the incorrect conclusion that only three local residents were likely to be affected by noise from the wind farm at levels of 30dBA or more;
(c) the incorrect conclusion that there needed to be only one monitoring point for assessment of wind farm noise compliance; and
(d)failure to establish a detailed regime to manage and monitor the effect of SACs.
[39] Second, the court considered that those inaccuracies materially influenced the decision on the application. The Environment Court found that the Commissioner at first instance accepted Mr Hunt’s evidence that only three residences would be
adversely affected by noise from the wind farm and the very limited noise effects of the wind farm generally. The Environment Court found if the information before the Commissioner had accurately predicted noise generation, he would have specified additional locations for monitoring and created a condition to address the measurement, assessment and criteria for any SACs penalty. The Commissioner may also have ordered post-installation and pre-operational testing to verify modelling results to account for the operational differences between the prototype turbine and constructed turbines.
[40] Third, the court found that the effects of the exercise of the consent were such that it is necessary to apply more appropriate monitoring conditions. Further, the acknowledged presence of SACs created the need for an enforceable and viable SAC condition. The court also said that other consequential changes to conditions might flow from the changes identified.
The appellant’s argument
[41] The appellant argued that the predictions contained in the NIAR relating to SPL, the presence of SACs and the noise received at residential properties were components of the process of setting noise restrictions and should not be treated as additional noise limits. They said conditions 4 and 5 were intended to be the only means of controlling noise and the Environment Court’s interpretation restricts noise to levels below what is authorised by those conditions, creating a conflict. The appellant submitted that this conflict should be addressed by the principle that specific conditions prevail over general conditions.
[42] In support of this, the appellant argued first, that the Commissioner intended for conditions 4 and 5 to set noise limits and control the noise produced by the turbines. In his decision, the Commissioner found that excessive or intrusive noise could be an adverse effect that was more than minor but the “the level of noise is predicted to comply with the NZS6808:1998 recommended noise levels”. He accepted evidence from the Council and from NZWL that “given the imposition of appropriate consent conditions the overall result will be one where adverse effects on the environment can be adequately avoided, remedied or mitigated.” From this, the
appellant said that it was not the consent Commissioner’s intention for predictions contained in the NIAR to be treated as additional parameters for noise control. His focus was on the effects of noise received and he formed the view that the standard would adequately address those adverse effects.
[43] Second, the appellant said that the NIAR, taken as a whole, focussed on whether the predicted sound levels were within the standard guidelines and recommended that the noise limits prescribed by the standard should be adopted. The appellant pointed to excerpts from the NIAR, that stated the standard stipulates the limits of acceptability for sound received from turbines at residential locations and the standard should be adopted “to provide a satisfactory level of protection” against potential adverse noise effects. The predictions are not mentioned as an additional means of controlling noise.
[44] Third, the appellant said s 128(1)(c) of the Act is the appropriate solution to the situation. Section 128(1)(c) allows the consent authority to review the conditions of the resource consent and remedy the effect of the inaccurate information contained in the resource consent application. This process would allow all potentially affected neighbours to participate in a further process to evaluate the actual noise effects of the wind farm and amend the conditions accordingly.
The respondent’s argument
[45] The respondent’s submissions understandably shadowed the Environment Court’s reasoning to some extent. The respondent argued that condition 1 is compatible with conditions 4 and 5 because they control different things. SPL and SACs relate to noise emitted at source while conditions 4 and 5 address noise effects received by neighbouring properties.
[46] In addition, condition 1 and conditions 4 and 5 perform different functions, the respondent argued. Condition 1 controls the scope of the application for which consent is sought. Its purpose is to ensure that the applicant does ‘generally’ what it says it is going to do. Even without condition 1, the respondent argued, the appellant would not be permitted to go outside the scope of the application. Conditions 4 and
5 on the other hand, control specific noise effects on others. Therefore, the
respondent argued, the Environment Court was correct to conclude that condition 1 should be treated as containing an additional and distinct restriction on noise levels alongside conditions 4 and 5.
[47] The respondent argued that condition 1 must be applied according to its plain terms when the result of doing so would be absurd.7 The reasoning is as follows, according to the respondent: condition 1 requires the applicant to operate the wind farm generally in accordance with any information accompanying the applications; SPL and SAC predictions were contained in NIAR which was information accompanying the application; the applicant must therefore operate the wind farm generally in accordance with those predictions, failure to do so being a breach of condition 1.
[48] The respondent reiterated the Environment Court’s reasoning that an applicant should be bound by the information contained in its resource consent application. The respondent also pointed to the practical effect of finding that the predictions contained in the NIAR are subservient to conditions 4 and 5. The respondent argued that such a finding would allow NZWL to operate the wind farm with noise effects that are twice as loud as predicted and a noise shadow that is significantly larger than what was predicted. To allow this result, argued the respondent, would undermine the integrity of the RMA process and saddle local residents with an unpredicted problem not of their making. There is, the respondent argued, an issue of basic justice here.
Analysis
[49] The starting point is that scope conditions such as condition 1 in this case are enforceable both because they lawfully and expressly bind the appellant to the parameter of its own application and because they simply express in words the general principle that no consent may grant more than what is asked for.8 A slightly
different way of putting that proposition is that the applicant may not produce
7 Jewett Investments Limited v Body Corporate 204096 [2011] NZCA 232.
8 See Gillies Waiheke Limited v Auckland City Council [2004] NZRMA 385 at [22] (CA).
environmental effects that are materially greater than or different to the effects described in the application as its outer limits.9
[50] There can be no doubt that the SPL and SAC predictions in the NIAR included as part of the AEE are information accompanying the application and so fit the description of material covered by condition 1. Nor is there any doubt that Sound Power Level and Sound Pressure Level measure different things and so are not, in their nature, inherently in conflict. It is possible, in theory at least, for the Council to have imposed separate controls on noise generated at source (SPL and SACs) and noise received at residential homes (sound pressure). The Environment Court was correct in my view to conclude that condition 1 identifies the means and
conditions 4 and 5 describe the end.10
[51] There is therefore no need to have recourse to the maxim generalia specialibus non derogant to resolve a potential conflict between a condition relating to noise generated and a condition relating to noise received. Put simply, there is no conflict.
[52] The respondent understandably therefore argues that the plain words of condition 1 must apply with full force: the SPL and SAC predictions are information accompanying the consent application and so the wind farm must be operated generally in accordance with those predictions in order to comply. Those plain words should only be set to one side if the result produced is absurd. Here, the respondent says, there is no absurdity. On the contrary, there is justice and fairness in holding NZWL to its prediction in terms of effects on innocent neighbours.
[53] In my view, the outcome in this appeal turns on the answers to two questions. First, what was the intended acoustic scope of the application? And second, what
was (or were) the intended limit (or limits) on noise in the consent decision. In each
9 See generally the discussion of Wild J in Aitkens v Napier City Council (2008) 15 ERNZ 84 and especially the test articulated at [20]: whether the change under consideration at hearing stage made the application significantly different in a relevant way. Although the question in that case was slightly different to that applicable in this case, the principle is nonetheless a useful way of thinking about scope under the RMA.
10 At [93].
case, true meaning is to be assessed objectively through the lens of an objective observer reading the documents in question and aware of all relevant context.11
[54] In this sense, the present case is similar to that in Effort Shipping Co Limited v Linden Management SA12 in which the House of Lords was faced with construing two apparently conflicting provisions in the Hague Shipping Rules. Their Lordships found that a careful reading of the rules demonstrated that in the case of dangerous goods, one provision was clearly intended by the drafters to take effect irrespective of the other.13 A similarly purposive or intention based approach yields the answer in this case.
The application
[55] An objective reading of the NIAR suggests to me that the SPL and SAC predictions were never matters of scope. When, in the NIAR, Mr Hunt set out the outer limits of noise emissions from the wind farm, he had in mind the recommended limits under NZS6808 cl 4.4.2. Three examples will suffice.14
NZ6808:1998 recommends wind farm noise limits be set at a level of 40dBA or 5dB above the background, whichever the greater. These levels are given in terms of an L95 level. (my emphasis)
Thus, while reduced terrain screening effects may be a possibility (yet to be proven), even under these worse (sic) case conditions the proposed wind farm sound levels will remain within the guidelines set out in NZS6808:1998. (my emphasis)
This assessment finds that, based on the ambient sound level data under various wind speeds together with predicted sound levels from the wind farm under various wind speeds, it is considered that the proposed wind farm installation of 104 Windflow 500 WTG will comply at all but one (Hargreaves) rural residential site with the limits set out in NZS6808:1998
Acoustics – The Assessment & Measurement of Sound From Wind Turbine
Generators. (my emphasis)
[56] By contrast, the SPL and SAC references were clearly designed to feed into the cl 4.4.2 parameter. They were not intended to be a parameter in their own right.
That is hardly surprising given their purpose. Returning to the equation at 4.3.2 of
11 See Gillies Waiheke Limited v Auckland City Council, above n 7, at [23].
12 Effort Shipping Co Limited v Linden Management SA [1998] 2 WLR 206.
13 At 212 at B per Lord Lloyd of Berwick.
14 NIAR at pp 9, 31 and 33.
NZS6808, the parameter is LR.15 LW (the SPL value), is a component from which that parameter is derived. LR was intended to provide the acoustic scope of this application.
The consent decision
[57] The consent decision of Commissioner Aburn is to similar effect in my view. The Commissioner reviewed Mr Hunt’s evidence including his SPL and SAC predictions and his predicted sound pressure levels at the monitoring sites. These did not determine scope. Rather they provided the Commissioner with comfort that the scope applied for would not be exceeded. They were the means to that end. He said this:
I consider that it is appropriate to adopt NZS6808:1998 as the basis for measuring noise from Te Rere Hau wind farm. NZS6808:1998 deals specifically with the measurement of sound from WTGs in the presence of wind. The standards state that other acoustic standards should not be used as these require the assessment of sound levels in the absence of wind, a situation that does nor [not?] apply for operating WTGs. (my emphasis)
[58] Once again effects are described in the following terms:
Based on the evidence to the hearing, I accept that there will be some adverse effects that are (potentially) going to be more than minor. These effects could include:
…
(b) excessive or intrusive noise: the evidence is that for existing rural- residential properties the level of noise is predicted to comply with the NZSD 6808:1998 recommended noise levels; or where there is the potential for those levels to be exceeded (eg Hargreaves/Flint property) a written approval has been provided such that s.104(3)(b) applies. In relation to future (potential) developments and activities on adjoining properties (eg rural residential development on the County Heights Trust and Alucard properties, and the proposed horsetrekking and associated lodges proposal of the Beales) I accept that under a full discretionary activity consent application it will be necessary to consider any adverse effects, including reverse sensitivity effects. I also accept the ‘outcome’ could be some restrictions and/or mitigation measures that may not have been necessary if the wind farm had not been established.
[59] As the appellant noted,16 the Commissioner found that the “the level of noise is predicted to comply with the NZS6808:1998 recommended noise levels”. He accepted evidence from the Council and from NZWL that “given the imposition of appropriate consent conditions the overall result will be one where adverse effects on the environment can be adequately avoided, remedied or mitigated.”
[60] It is on that basis that conditions 4 and 5 were adopted to set and then monitor the accepted acoustic limits of this project.
[61] It is true, as the respondent says, that Mr Hunt’s predictions about turbine noise generation – SPLs and SACs – were contained in documentation accompanying the application, but it is not enough that they simply exist in the documentation. In determining the meaning of condition 1, the appellant’s purpose in including those predictions will be very important. Particularly in this case because the same subject matter – noise – is dealt with in different ways in the same document.
[62] In its simplest terms, I do not consider that the NIAR’s SPL and SACs predictions were intended to go to scope. They related to how the predicted noise levels would be achieved, not what the levels should be. So operated “generally in accordance with the information accompanying the application” is to be read as affirming the scope of the application as the outer limit of consent. The term “operated” must mean operated within those limits because they were explicitly requested by the appellant and set by the consent commissioner as the allowable limits of operation. Here, conditions 1 and 4 are consistent. Properly interpreted they both say the wind farm must be operated so as to produce noise effects at the notional boundaries of local residents at no greater than 40dBA L95 or 5dBA above background noise, whichever is higher. That is what NZWL asked for. That must therefore be the scope of the application. Mr Hunt also said that these sound levels will be very achievable for NZWL because of the positive noise generation characteristics of the Windflow 500. He was of course completely wrong, but that does not change what was actually asked for.
[63] In my view, the approach favoured by the Environment Court and the respondent requires the objective observer to decontextualise the material about noise generation characteristics of the Windflow 500 turbines. That is to read them without reference to their purpose and without taking account of the L95 levels also referred to extensively in the application and consent.
[64] Thus, as with Effort Shipping, if a purposive approach is taken to construction of the documents in question, and context is properly understood, apparently conflicting provisions are able to be reconciled.
The relevance of s 128
[65] In this case NZWL has made an error. It relied on noise readings taken from a single prototype of the wind turbines the company was to use, but those readings did not transfer to production models. The error was a grave one from the point of view of the acoustic amenity enjoyed by surrounding neighbours, but I heard no suggestion that NZWL or its acoustic engineer had acted in bad faith in making these predictions. It was a mistake that needs to be addressed with all expedition. Section 128(1)(c) was designed specifically to address these sorts of issues. It provides:
A consent authority may, in accordance with section 129, serve notice on a consent holder of its intention to review the conditions of a resource consent–
…
(c) if the information made available to the consent authority by the applicant for the consent for the purposes of the application contained inaccuracies which materially influenced the decision made on the application and the effects of the exercise of the consent are such that it is necessary to apply more appropriate conditions.
[66] The Environment Court declared that this provision was available as a correction mechanism in this case. The court accepted that the application contained inaccuracies that materially influenced the decision and (implicitly) that it may be necessary to apply more appropriate conditions.
[67] The provisions in RMA covering public notification, submissions and hearings in respect of resource consents all apply with respect to a review under s 128. There is therefore a very public correction process and a reconsideration of the appropriate consent conditions.
[68] The appellant accepts that s 128 is available. It has not challenged that aspect of the Environment Court’s decision. On the contrary, the appellant accepts that such review is inevitable and must be engaged with.
[69] For his part, Mr Maassen expressed concern on behalf of the respondent that there will be no real prospect of revisiting the applicability of the NZS6808 standards in cl 4.4.2 in a s 128 review. I do not share that concern. It seems to me that in a s 128 review, potentially at least, the whole question of noise levels and compliance therewith is back on the table for fresh debate. Indeed according to s 132(3) a consent can be cancelled if adverse effects from its exercise are “significant”. I (of course) express no view as to whether that provision has application here, but it does demonstrate that a s 128 review is no mere tinkering exercise.
[70] In addition, local affected residents (whose numbers will now have multiplied with reconfigured noise contours) are able to participate with more complete knowledge of the noise impacts of the project. It will be open to them to argue that the noise level set in condition 4 (and by implication the NZS6808 standard) is too high in this environment. It is also to be remembered that monitoring continues and it is still not known whether Te Rere Hau wind farm is currently in breach of conditions 4 and 5.
[71] The differences then between a s 128 approach and the breach of condition approach advocated by the respondent is where the onus is cast. Under s 128, the respondent acts, but the appellant retains its consent with current conditions in the meantime. Under a breach of condition approach, the respondent may issue an abatement notice but the onus will be cast on the appellant to apply for a fresh consent or amended conditions and in the meantime to cease breaching the
condition. The two processes might end up at the same point, but for one party or the other, time is lost.
[72] The important point, it seems to me, is that if innocent neighbours have come to be unfairly affected by much greater noise impacts than had been predicted, then the s 128 process is available to redress that unfairness. They are certainly not left stranded by my finding with respect to the meaning of condition 1.
Disposition
[73] The appellant’s appeal is allowed. Declaration 1.9 is set aside. The respondent’s cross-appeal is overtaken by that result.
[74] Costs are reserved.
Williams J
1
1
0