Van den Berg, Dr. Fritz, Expert Review, NHMRC Draft Information Paper

Dr G.F. (Frits) van den Berg
Amsterdam, The Netherlands

Comments on the NHMRC Draft Information Paper: Evidence on Wind Farms and Human Health
dr G.P. (Frits) van den Berg, March 2014

General comments on the Draft information paper (‘the paper’).

  1. There is no clear definition or description of health effects. This may create confusion. In the WHO Guidelines for Community Noise and the Night Noise Guidelines annoyance is mentioned as one of the health effects without restrictions. Apparently in the paper ‘direct’ health effects do not include noise annoyance or sleep disturbance (e.g. section 7.1: “There is no (….) evidence that wind farms directly cause adverse health effects”, but “There is (…) evidence that proximity to wind farms is associated with annoyance and (…) sleep disturbance and poorer quality of life.”). Thus the paper implies annoyance and sleep disturbance are ‘indirect’ health effects. Later (p.14) they are ‘health-related outcomes’ as opposed (?) to ‘health outcomes’. I suggest to be clear about this and describe what health effects are considered and why/how they are classified as direct (somatic?) or other, or as health or health-related effects.
  2. I agree largely with the critical appraisal in section 3.4 and the prudence in interpreting research results. However, the text is too academic and suggests a partiality with respect to overinterpretation by lay (?) readers. It is academic and my objection is that the critical observations apply to most if not all environmental health effect studies, not just to wind turbine studies. Low participation rates are not uncommon in surveys, and there is no generally accepted minimum participation rate. If selection bias occurs, this would affect many other environmental health effect studies. One of the ways to try to take this into account is a non-response analysis, as was done in some studies. Of course masking may not be effective, but again this problem may affect other environmental health effect studies as well: if people close to an airport find a question on airport noise in their questionnaire, they may think that the presence of the airport is a reason for the survey. The same may be true for information bias: residents near an airport perhaps recall more health details than those further away. Confounding is a very important issue to consider, but confounding factors may have been identified and can in principle be anything the imagination produces. Secondly, the text suggests partiality in that most examples caution for positive results. But it could also be otherwise: why would recall be better at short distances from a source?; confounding factors could also reduce an association (e.g. over time those being most annoyed having moved away);
  3. Visual impact is not addressed here, except for shadow flicker. A literature review should reveal that it is a major issue in wind farm planning (e.g. M. Wolsink, Planning of renewables schemes: Deliberative and fair decision-making on landscape issues instead of reproachful accusations of non-cooperation Energy Policy 35 (2007), 2692–2704; P. Devine-Wright, Beyond NIMBYism: towards an Integrated Framework for Understanding Public Perceptions of Wind Energy, Wind Energy 8 (2005), 125–139). The visual impact of wind turbines on landscape may be associated to visual annoyance as noise is to noise annoyance. Visual impact has been investigated in several studies (such as references 8 and 10 in the paper), but not with regards to its effect as a separate stressor. Although it is plausible that such visual annoyance could lead to chronic stress, I know of no evidence supporting this. However, as this issue is important in the public debate on wind farms, it seems logical to include it as a health (related) topic, even though there may be no publication supporting a ‘direct’ link with health. Results may not show up in the systematic scientific review of studies, but it could be identified as a gap in the knowledge of the impact of wind farms. In a later stage NHMRC could recommend to be careful in planning wind farms. A guide to this could be the “What is good planning; position statement” from the Planning Institute of Australia.
  4. I strongly support the use of ‘parallel evidence’ or additional knowledge from exposure to other sources to study the effects of wind turbines. It is highly relevant when it has been shown that a specific acoustical or other characteristic increases or reduces the effect on residents. The use of parallel evidence acknowledges that we can build on available knowledge and it implies that we need not address all issues again specifically for wind turbines.

Specific comments: 

  1. End of page 10: “NHMRC could not conclude that exposure to wind farm noise causes annoyance, sleep disturbance or poorer quality of life”. The same is true for the effects of noise from road traffic or aircraft. Although strictly and academically true, in normal life we accept that annoyance from passing cars or aircraft is the result of those passages. Even though we acknowledge that the effect is not the same for all people: some or not bothered, others may be very sensitive, some are deaf. That is also the case for somatic effects, e.g. from allergens or viruses.
  2. The term ‘louder’ in the text “a low-frequency sound (lower than 100 Hertz) needs to be at a higher level (“louder”) to be heard than a mid-range frequency (e.g. 1000 Hertz)” (p. 11, 2nd par.) is incorrect. Louder refers to hearing, not to level. Suggestion: a low- frequency sound (lower than 100 Hertz) needs to be at a higher level (more physical sound, more ‘volume’) to be heard as loud as a mid-range frequency sound (e.g. 1000 Hertz).
  3. I suggest to drop “if sleep is disturbed” from the last sentence but one on p. 11. This may apply to subjective (self-reported) sleep disturbance, but not (always) to objective sleep disturbance.
  4. There is original German research on the effects of shadow flicker (Pohl J, Faul F, Mausfeld R. Belästigung durch periodischen Schattenwurf von Windenergieanlagen. Kiel, Germany: Institut für Psychologie der Christian-Albrechts-Universität; 1999. If language is the problem, it is mentioned in Parsons Brinckerhoff’s Update of UK Shadow Flicker Evidence Base (Newcastle upon Tyne, UK: Parsons Brinckerhoff; 2011) that also mentions some case studies.
  5. Reference 22 at the end of p. 12 is about German, not ‘Australian suburban homes’.
  6. I agree with “effects on sleep are likely to be modest” (section 7.2.2) from a statistical point of view. However, it may not apply to individuals: a vulnerable group are those that sleep with an open window and live close to and with a bedroom facing a wind farm.
  7. In ‘Physical health’ (p. 14) reference is made to the WINDFARM perception report (ref. 10), where the better (peer-reviewed) reference is E Pedersen, F van den Berg, R Bakker, J Bouma: Response to noise from modern wind farms in The Netherlands; Journal of the Acoustical Society of America, volume 126 nr 2 (2009), p. 634-643. There is no mention here of the Polish and Japanese wind turbine noise studies, that largely confirm the studies that are reported. Although as yet only reported in conference papers, should these not have come up in the literature search?
  8. In ‘Mental health’ (p. 14) reference is again made to the WINDFARM perception report (ref. 10), where the better (peer-reviewed) reference is ref. 14 with respect to psychological distress. I think it is relevant that in this study distress in noisy areas (with main road) was significantly related to annoyance, not to sound level. In quiet areas there were no significant relations.
  9. In the last sentence of the paragraph on ‘Annoyance’ on p. 14 the word ‘contributed’ suggests these factors all increase annoyance, but they can also reduce the prevalence of annoyance (as was the case for the ‘healthy farmers’). I suggest to use ‘influence the prevalence of annoyance’.
  10. In the paragraph on ‘Sleep’ (p. 14/15) there is no mention of ref. 14 where it was found that self-reported sleep disturbance was significantly related to noise annoyance but not to noise level.
  11. It is hard to believe that there is no background evidence on shadow flicker or other visual effects. Apart from the visual intrusion (see point 4 above) there must be psychological evidence that moving (rotating) objects attract more attention than static objects and severe light modulation (flicker) has more impact than a constant light level. As parallel evidence this would imply a possibly higher risk on annoyance.
  12. I agree with the recommendations for further research (p.20). However, as many concerns do not arise from scientific considerations but from public concerns, I recommend to involve representatives from the public in the definition and interpretation of new studies. This is to ensure that factors thought to be important from a public point of view are included in the studies.

Download the review as a PDF→