This page contains an article, paper, news item or other source of evidence referred to in The Case Against Windfarms
ETSU-R-97
Why it is Wrong
Dick Bowdler
July 2005
ETSU-R-97
Why it is Wrong
1.1
ETSU-R-97 is used throughout the
1.2
The conclusions of ETSU-R-97 are so badly argued as to be laughable in parts
(the daytime standard is based on the principle that it does not matter if
people cannot get to sleep on their patio so long as they can get to sleep in
their bedrooms). It is the only standard where the permissible night time level
is higher than the permissible day time level.
1.3
ETSU-R-97 bears no resemblance to standards used for other industrial
developments. Other renewable energy developments have to meet much stricter
standards. Each time the Noise Working Group that drew up the document decide
that a particular standard is appropriate, they follow it up by saying (without
putting forward any evidence whatsoever) that such a standard would restrict
development of wind farms and so find reasons to relax it further.
2.1
It seems common sense that the impact of a new noise on existing residences is
related in some way to the background noise. For example if the background
noise level at present is 45dBA then a level of 35dB from a new industrial
source would probably be inaudible. If the background noise level at present is
20dB then an industrial noise of 35dB will clearly be heard and would be very
likely to produce complaints.
2.2
Indeed it is normal to set a noise limit relative to the pre-existing
background noise when a new industrial noise is to be introduced into a
residential area. Typical planning conditions imposed by rural local
authorities (and sometimes urban ones) require that the new noise be no more
than 5dB above the pre-existing background. This is based on the procedure set
out in British Standard 4142.
2.3
In fact BS4142 does not purport to be a method of assessing nuisance or
amenity. It was first published in 1967 and has since been revised twice though
the general principles
1
remain the same. It is simply a method of assessing the
likelihood of complaints. Its origin is obscure and it has been the subject of
endless criticism for a whole variety of reasons. But the fact is that it
works. It has been and is still regularly used to assess noise impact and I do
not know of one case where it has been suggested that BS4142 gave an anomalous
result. Furthermore it was endorsed by DEFRA in September 1998, the department
of government concerned with the environment at that time. They submitted their
Noise and Nuisance Policy under Health Effect Based Noise Assessment Methods to
the EU. This said that BS4142:1997 provides a technical means of assessing
whether or not 'complaints are likely'. The result of an assessment carried out
to BS4142 would normally be relevant to the deliberations of any court
considering whether or not a nuisance exists.
2.4 BS4142 is not
normally used to assess wind farms. This is done using the document ETSU-R-97
"The Assessment and Rating of Noise from Wind Farms".
2.5 ETSU-R-97 was
written by a Noise Working Group (NWG) of developers, noise consultants,
environmental health officers and others set up in 1995 by the Department of
Trade and Industry through ETSU (the Energy Technology Support Unit). The DTI's
mission is prosperity for all by working to create the best environment for
business success in the
2.6 The status of
ETSU-R-97 is perfectly clear. The preface says The
aim of the Working Group was to provide information and advice to developers
and planners on the environmental assessment of noise from wind turbines. While
the DTI facilitated the establishment of this Noise Working Group this report
is not a report of Government and should not be thought of in any way as
replacing the advice contained within relevant Government guidance. The report
represents the consensus view of the group of experts listed below who between
them have a breadth and depth of experience in assessing and controlling the
environmental impact of noise from wind farms. This consensus view has been
arrived at through negotiation and compromise and in recognition of the value
of achieving a common approach to the assessment of noise from wind turbines.
2.7 The first
paragraph of the executive summary says This document describes a framework
for the measurement of wind farm noise and gives indicative noise levels
thought to offer a reasonable degree of protection to wind farm neighbours,
without placing unreasonable restrictions on wind farm development or adding
unduly to the costs and administrative burdens on wind farm developers or local
authorities.
2.8 It is thus,
by its own admission, not a method of assessing impact. What is more the
compromise reached by the NWG is so lacking in basis, so full of unfounded
assertions and so badly thought out and argued that it comes up with standards
for wind farm noise that are quite unlike any other noise standards. I need to
explain in some detail why this is the case so that my point can be fully
understood.
2
3.1 I have
explained why the assessment method in ETSU-R-97 is not a measure of impact. I
need to describe how the assessment method was developed by the NWG in order to
explain how it relates to normal methods of measuring impact. The NWG starts by
pointing out that the planning advice relating to noise says that the
likelihood of complaints can be assessed, where the Standard is appropriate,
using guidance in BS 4142: 1990. In examining whether BS4142 is appropriate for
assessing wind turbine noise the NWG suggests that there are three reasons why
it might not be. These are:
Wind
farms are likely to be developed in largely rural areas and not in the areas to
which the standard is principally addressed, namely mixed residential and
industrial areas;
the
scope of BS 4142 specifically precludes situations where background noise
levels are below 30dB(A);
BS 4142
recommends that noise measurements should not be taken in extreme weather
conditions such as high wind speed greater than 5 metres per second average
".
3.2 In answer to the first point they say Although the standard is intended for use in mixed
residential and industrial areas as suggested by its title, there are no
obvious reasons which prevent its application in more rural areas and indeed
Members of the Noise Working Group have used it in such areas. So BS4142 is
not rejected for this reason.
3.3 To the second point they say, after some debate, The
question that arises is: if one intends to apply the principles of BS 4142 to
the protection of external amenity, and the instrumentation is available to
accurately measure noise levels below 30dB(A), should
a margin above background approach be pursued in low noise environments or can
an absolute level be justified in such circumstances? They leave the
question to be dealt with later. I should point out that since ETSU-R-97 was
published BS4142 has been revised so that low noise levels are only excluded
when both the background is less than 30dB and the turbine noise is less than
35dB.
3.4 Whatever the NWGs answer to the third reason, and
it is not very clear what that answer is, it is obvious that they accept that
there is no reason to reject BS4142 at higher wind speeds because ETSU itself
says that background noise should be measured at all wind speeds up to 12m/s.
3.5 In summary, thus far the NWG seem to find no good
reason to reject BS4142 except that it leaves open the possibility of whether
to adopt a limiting absolute level to be dealt with later.
3.6 At this point it is necessary for me to explain LA90
and LAeq. Noise levels can be stated in different ways. For
example if a noise is fluctuating we could talk about the minimum or the
maximum or the average. BS4142, in accordance with international practice, uses
the
3
measure LAeq to describe the specific noise - that
is the noise to be assessed. This is effectively an average. It is actually a
logarithmic average but that is of no real significance here. Again in
accordance with common practice BS4142 uses LA90 to define
background noise. This is the level exceeded for 90% of the time, so in a ten
minute period the noise level is more than the LA90 for an aggregate
of 9 minutes. So the LA90 is usually close to the minimum noise
level.
3.7 On the
question of turbine noise the NWG put forward the suggestion that LA90 should
be used to measure turbine noise. This is because the measure will eliminate
other extraneous noise. For example, if a site is affected by an occasional
passing car, the LAeq may be determined by the car whilst the LA90
may not. I have no objection to the principle of measuring turbine noise
by the use of LA90. This is a method I often use where the
difference between the LAeq and the LA90 is known and
constant. However, it would be much better to measure as LA90 and
then add back 2dB (the difference between the two) to get the LAeq value
so that the units remain consistent with BS4142 and other normal practice.
ETSU-R-97 carries on describing turbine noise as an LA90 which
simply leads to confusion. BS7445 (Also ISO1996) Description and Measurement
of Environmental Noise makes it clear that environmental noise is to be
described as LAeq.
3.8 On Page 59
ETSU-R-97 says It is proposed that the background noise levels upon which
limits are based, and the noise limits themselves, are based upon typical
rather than extreme values at any given wind speed. An approach based upon
extreme values would be difficult to implement as the difference in
measurements between turbine noise and background would depend upon the length
of time one is prepared to take data. A more sensible approach is to base
limits upon typical or average levels, but to appreciate that both turbine and
background noise levels can vary over several dB for the same nominal
conditions. What they are saying is that, having measured background noise
levels over a period of several weeks we should take the background noise level
at each wind speed as the average of all the background noise levels at that
wind speed. This is completely inconsistent with normal practice and suggesting
it is "sensible" is merely an unfounded assertion. In using BS4142 in the field
we are generally required by local authorities to measure at the quietest part
of the period in question. It is not acceptable, where traffic noise
predominates, to take an average of the LA90 values over, for
example, a whole night time period. The local authority will require the
background noise in the middle of the night when it is quietest. For example
A letter
from Renfrew Council in 2004 in connection with a planning application says
that the impact of noise on nearby dwellings should be assessed by BS4142 and
that the background noise level for the most sensitive period that the
source could operate should be used for this assessment.
At the
Portree Co-Op development it was agreed that In accordance with BS4142 the
background noise should be measured as LA90 and the noise from the
development as LAeq. Measurements of LA90 over any
specific period should be carried out in wind speeds less than 5m/s and during
a representative part of the period including the quietest part of the period.
The measurements should be made in intervals of between 5 and 15 minutes. The
average and standard deviation of all the measurements should be calculated and
the
4
background noise taken as the average less one standard deviation. So the level required is more or less the quietest
part of a quiet night.
3.9 In the case of background noise dominated by wind
it has been my practice to take the average and the standard deviation of a
group of 10 minute measurements and to define the period LA90 as the
average less one standard deviation. Typically this is about 4dB less than the
average. Statistically 15% of the time the background noise is below this
level. Unless there is a large variation between day and night time background noise I will normally use the whole 24 hour data rather than
separate day and night.
3.10 Returning to ETSU-R-97 on page 60, continuing
discussion on background noise the NWG say, Noise from the wind farm will be
limited to 5dB(A) above background for both day- and night-time. When comparing
the proposed margin with the complaints criteria suggested by BS 4142 it is
important to bear in mind that the LA90 descriptor is also being proposed for
the turbine noise. The Leq levels can be expected to be about 1.5-2.5dB
greater. An addition of 1.5-2.5dB places the margin at the upper end of the
range which can be considered to be of marginal significance ie around 5dB. What
they appear to be saying is that, because turbine noise is measured as LA90,
the margin above background noise that is proposed is actually 7dB in normal
BS4142 terms rather than the 5dB normally adopted by local authorities. There
is nothing in BS4142 that suggests that 7dB is at the upper end of the range
which can be considered to be of marginal significance. This phrase is
simply an invention of the NWG.
3.11 Further down page 60 it says that On balance
it is considered that a margin of 5dB(A) (by which
it means 7dB in BS4142 terms) will offer a reasonable degree of protection
to both the internal and external environment without unduly restricting the
development of wind energy which itself has other environmental benefits. There
is no foundation whatsoever for this assertion. No evidence is brought forward
or referred to.
3.12 So the position in the argument so far is this.
The NWG has decided, without any foundation, that the 5dB "marginal significance"
in BS4142 could be 7dB. It has decided, against all normal practice, that the
background noise level for assessment purposes ought to be the average of
background levels in any particular condition rather than the lowest level. In
wind controlled background noise the average is likely to be at least 4dB more
than a realistic background level. So the NWG consider that 11dB over
background is appropriate for wind farms as against normal practice for
industrial noise of 5dB over background noise. Of course I have to bear in mind
that ETSU-R-97 does not purport to offer a method of assessment of impact. So
the NWG is proposing that, for wind farms, a level of noise that is likely to
give rise to complaints is appropriate because of the particular public benefits
of wind farms. I cannot agree with this. As I exemplify elsewhere other
projects of public benefit have to meet the stricter standard of 5dB above
background.
3.13 Not content with establishing a margin above
background noise far greater than normal, the NWG, at the bottom of page 60,
continues Applying the margin above background approach to some of the very
quiet areas in the UK would imply setting noise limits down to say 25-30dB(A)
based upon background levels perhaps as low as 20-25dB(A). This is
5
true in principle
but in practice turbines generate less noise at low wind speeds and, at cut in,
turbine noise might have to be limited in some areas to as little as 25dB. By
the time wind speed was up to 6m/s the background noise level would be at least
25dB probably more like 30dB and so this would require turbine noise to be
restricted to less than 30-35dB rather than 25-30dB. Limits of this level
would prove very restrictive on the development of wind energy. This is
simply a broad assertion. No evidence whatsoever has been adduced to
demonstrate this.
3.14 Some measure
of loss of amenity needs to be applied in low background noise levels and it is
normal practice in rural
Co-Op
Retail Store, Portree in 2002. Noise of plant from the development should
not exceed the background noise level by more than 5dBA or, if the noise is
tonal, should not exceed the background noise at all at any noise sensitive
property. The background noise at Home Farm Road was measured at 28dB on a
calm night and this was agreed as the background noise.
New factory for Vestas at Machrihanish in 2001. At this new factory (ironically the factory that
makes wind turbines) Argyll and Bute Council require that: The rated noise
level from the development shall not exceed the predetermined ambient noise
level (the L90(A)) at the nearest noise sensitive
properties at the former RAF housing, by more than 5dB(A). All measurements are
to be taken in accordance with BS4142: 1997 with the measurement periods being
1 hour for the period 0800-2200 hours and 5 minutes for the period 2200-0800
hours. The night time background noise was agreed at 27dB which was the
lowest hourly level reached during a windless night.
Earlier measurements when there was sea noise and the background was 32dB were
not accepted by the council.
In 2004,
SEPA, at Roslin in
3.15 On page 61 the NWG say During
the night one can reasonably expect most people to be indoors and it will not
be necessary to control noise to levels below those required to ensure that the
restorative process of sleep is not disturbed. A night-time absolute lower
limit is therefore appropriate based upon sleep disturbance criteria. What
this says is that a turbine noise level inside peoples
houses of just less than the World Health Organisation say is necessary to get
back to sleep if you wake up in the night is satisfactory. It seems to me this
must be the very upper limit of acceptability, not one that is well balanced.
Since then, the WHO has revised its guidance 5dB lower. So the ETSU night
standard is now higher than WHO say you need to get back to sleep.
3.16 When they come to day time, on Page 62 of
ETSU-R-97, it says It is also the opinion of the Noise Working Group that
there is no need to restrict noise levels below a lower absolute limit of LA90,10min = 33dB(A); if an environment is quiet enough so as
not to disturb the process of falling asleep or sleep itself then it ought to
be quiet enough for the
6
peaceful enjoyment
of one's patio or garden. This is a
bizarre statement. It seems that the 33dBA is the 35dB sleep restoration level
set out by the World Health Organisation for inside bedrooms at night. They
seem to be saying that there is no need for noise levels during the day to be
any lower than is necessary to allow you to go to sleep on your patio on a
sunny afternoon.
3.17 Having
suggested that 33dB would be satisfactory because people could get to sleep on
their patio - they now say that This level
would however be a damaging constraint on the development of wind power in the
3.18 Then, on
page 63 there is another leap of credibility: There is no evidence for or
against the assertion that wind farm noise with no audible tones is acceptable
up to and including LA90,10min levels of 40dB(A) even when background noise
levels are 30dB or less. This is just nonsense. There most certainly is
evidence against this assertion. The 40dB is actually 42dB in BS4142 units.
This is at least 12dB above background noise level of "30dB or less" and BS4142
says there are likely to be complaints at turbine levels of plus 10dB.
Furthermore there is no argument that BS4142 is not applicable. Even BS
4142:1990 (which was current when ETSU-R-97 was written) might easily be
applicable here. If the wind speed is 5m/s, the background noise 30dB and the
turbine noise 42dB(LAeq) then there is no reason not
to use BS4142, it does not exclude itself in these circumstances. This noise
level is also 12dB more than (twice as loud as) the WHO considers necessary for
you to be able to get to sleep.
3.19 They
summarise this For periods during the day
the Noise Working Group has adopted the approach that external noise limits
should lie somewhere between that required to avoid sleep disturbance even if
the occupant is outside of the property and the higher level that would still
prevent sleep disturbance inside the property. In other words the lowest
turbine noise level that they would adopt, during the day, would be high enough
to prevent you getting to sleep on your patio. The highest level they adopt
during the day would not quite stop you getting back to sleep in your bedroom.
Presumably the principle is that, if it is too noisy to sleep outside on your
patio you can be assured you will be able to get to sleep indoors.
7
4.1 ETSU-R-97 is
so poor technically that its conclusions have to be queried. It is put together
through a series of unfounded assertions and there has been no research drawn
on to justify them.
4.2 Even if one
were minded to accept the suggestion that you should use very low background
noise levels and that there ought to be a level below which it would be
appropriate to use an absolute noise level, the levels proposed by the NWG are
not acceptable. The night time level is 45dB(LAeq)
and the day time level is 37 to 42dB(LAeq). Most wind farm sites are
in rural areas where background noise levels can easily be 20 to 25dBA when
turbines are operating and so the margin above background could be up to 20dB
or more.
8