El texto, el discurso y el código.

The issues relating to annoyance due to wind turbines are presented in Section 2.8 of this work. This section presents New Zealand specific research.

Attitudinal studies by Mosley (2007), Phipps (2007), and Phipps et al. (2007) present significant evidence as to the perception of noise affecting people who live in the

26

Further investigation of appropriate methodologies to analyse sound character from distant wind turbines is desirable. This work, however, is not part of this research.

vicinity of extensive wind farms in New Zealand. The work by Mosley presents brief case studies of residences affected and unaffected by wind farm noise and/or vibration. The work by Phipps et al. presents a comprehensive attitudinal study of residents affected and unaffected by wind farm noise. My observations run parallel with the research of both Mosley and Phipps et al.

In July 2006 as an expert witness to the Environment Court (New Zealand) I presented the outcomes of my investigations on behalf of the Makara Guardians Incorporated (MGI) into potential adverse effects of wind farm noise. The case involved the ‘West Wind’ wind farm Wellington New Zealand. Professional technical advice on behalf of MGI was tendered to the Environment Court by Dr Fritz (GP) van den Berg who supported his evidence by reference to his published work (van den Berg, 2006). In accordance with standard practice the acoustics experts were called to an Experts’ Caucus meeting to discuss the technical issues and to arrive at conclusions that would help the Court make a decision. The applicants’ advisors relied on the provisions of NZS 6808:1998 - Acoustics – The Assessment and Measurement of

Sound from Wind Turbine Generators. The standard provides a nominal guideline as

to acceptable sound from a wind farm:

As a guide to the limits of acceptability, the sound level from the WTG (or wind farm) should not exceed, at any residential site, and at any of the nominated wind speeds, the background sound level (L95) by more than 5dB(A), or a level of 40 dB(A) L95,

whichever is the greater.

Under NZS 6808 the limits of acceptability are subject to an assessment for special audible characteristics. The Experts’ Caucus process is normally of one day. In this case, the caucus continued for nearly 4 days and the issues concerning wind turbine noise, measurement and assessment were vigorously debated. The Conditions (detailed in Decision W59/2007) developed during the Caucus relevant to this research relating to ‘acceptable’ levels and special audible characteristics are:

17. Wind turbine sound levels, when measured at the notional boundary of dwellings existing or holding all resource consents necessary for construction at the date of this consent, or able to be constructed as a permitted activity shall not exceed the appropriate regression curve of the A-weighted background sound level (L95) by more

When the background sound conditions are at or below 25dBA L95 determined from the

appropriate regression curve, without the interference of the wind farm, and when the mean wind speed at a representative location for the dwelling is less than 1.5m/s measured at a height of 10m AGL, then noise from the wind farm shall not exceed 35dBA L95 at the dwelling.

33. When wind farm sound within the notional boundary of a dwelling has a special audible characteristic, i.e. impulsiveness, tonality and/or an audible modulation, the measured sound level of the source shall have a maximum 5dB penalty applied by adjustment of the measured sound level by the arithmetic addition of the penalty. The total penalty for all special audible characteristics shall be no more than 5dB.

34. Sound with a special audible characteristic includes clearly audible tones. A test for the presence of tonality shall be made by comparing the levels of neighbouring one- third octave bands in the sound spectrum. An adjustment of +5dB for tonality shall be applied if the level (Leq) in any one third octave band exceeds the arithmetic mean of

the Leq levels in the two adjacent bands by more than the values given in Table 1.

Table 1: One-Third Octave Band Level Differences One-third octave band Level difference

25 - 125Hz 12dB

160 - 400Hz 8dB

500 - 10,000Hz 5dB

There might be cases where this analysis does not result in a tonal component being defined although the sound is in fact tonal. For these cases it will be necessary to undertake a narrow band analysis in order to determine if a sound is tonal using Joint Nordic Method Version 2 with the penalties in that document applied.

35. A test for modulation is if the measured peak to trough levels exceed 5dBA on a regularly varying basis or if the spectral characteristics, third octave band level, exhibit a peak to trough variation that exceeds 6dB on a regular basis in respect of the blade pass frequency.

A significant outcome of the Caucus was the acknowledgement that there needed to be specific tests for tonality and modulation. Neither Dr van den Berg nor I agreed with the acceptable criteria of Condition 6 and approval was given for a dissenting statement to the agreed matters:

We believe that the conditions here agreed upon will protect residents from severe annoyance and sleep disturbance, but not from annoyance and loss of amenity. We believe annoyance and loss of amenity will be protected when the wind turbine noise limit would be 30 dBA L95 in conditions of low wind speed at the dwellings and

modulation restricted to 3 dB.

In an affidavit to the Environment Court (Hayes, 2007, clause 34) states that:

…In contrast, in Project Westwind the selected turbines are to be warranted as free of tonal noise. On this basis, one would expect the wind turbine to be audible even if the turbine noise was 10 – 15 dB below the background noise level, something which is confirmed by the analysis undertaken by Nelson. It should also be noted that the intent of NZS6808 is not inaudibility but the prevention of severe annoyance. …

The issues of low frequency sound and modulation was addressed by Hayes in his affidavit, clauses 15-18, and confirms that modulation of the aerodynamic noise of the blade passage frequency can be audible within a dwelling and can cause complaint. The Environment Court decision requires continuous noise monitoring (Condition 30A) to measure background levels in order to ensure compliance with the conditions. Continuous monitoring of the characteristics of the sound in order to assess special audible characteristics is not required. In hindsight, this is a significant failure within the monitoring program as ‘noise levels’ are not the sole determinant of noise intrusion. The measurement and assessment of special audible characteristics is more important for noise intrusion within the home.

Subsequently, at another wind farm (Mill Creek) hearing near to the West Wind proposal, evidence was presented by James (2008) and Trevathan (2008). The prevention of health risks, Mill Creek evidence submitted by Hayes, measurement artefacts due to windscreen limitations, amplitude modulation, computer model accuracy and flaws in the New Zealand wind turbine standard are critiqued.

The decision and submissions are significant in the development of this work as they are taken into account for the low amplitude sound and intrusive noise analysis assessment methods.