Examination of Witnesses (Questions 2900
- 2919)
2900. There was a concern, was there not, about
a train going at over 100 miles an hour, well over 100 miles an
hour, being on a floating track slab?
(Mr Methold) Indeed. I think
at that time it was unproven technology for high-speed rail. There
are safety issues associated with using squishy material on the
track. Can I have the next slide, please? Very interestingly,
a second undertaking was given by the Select Committee. This is
one I would like to draw the Committee's attention to now. This
undertaking specifies the number of properties that were expected
to experience the low and medium impacts. This undertaking constrained
the project such that it was not allowed to introduce any new
impacts in the low category. In other words, importance has been
placed on the low impact category of 35-39. CTRL recognised that
this was an adverse impact and that the project should be constrained
to not introducing any additional low impacts. Again, I would
suggest that this is more onerous than the situation that Crossrail
is putting forward to us at the moment.
2901. Mr Binley: Can I clarify for own
benefit, because in a number of instances we have two and sometimes
three different tunnels operating one above the other in levels,
and at times trains will be going through those tunnels at the
same time. Does that affect your calculation in any way? Nobody
has said anything about that. It may not but it may.
(Mr Methold) It could affect
the overall noise level. If two trains are passing at the same
time it is reasonable to say that the noise level will be additive,
particularly if you are using the slow time constant, because,
as we know, that means that the needle has slowed down on our
meter, so it means that there is more opportunity for the two
noise levels to occur at the same time. With a fast time response,
which is much more of an intermittent flickering noise level,
it could be argued that it is very unlikely you would get the
two fast events at the same time with a passing train. It would
only really be an issue where you have the same distance between
the property and the two tunnels. It may be unlikely that you
would find many situations where that would be the case, but clearly
if you were straight down the middle of the two—
2902. Mr Clarkson: Just to pursue the
hon Member's question a little bit further, has there been any
material in the environmental assessment that you have seen that
deals with that point?
(Mr Methold) No. Can I have
the next slide, please? I would now like to explore the origins
of the Crossrail Design Aim.
2903. LBC21.
(Mr Methold) Indeed. We
have three studies that underpin the design aim. Two were undertaken
by London Underground, and one by Crossrail. This is the study
that Mr Thornley-Taylor alluded to last week—the Crossrail
Social Survey, as I shall refer to it. Interestingly, the two
studies undertaken by London Underground are based around considering
complaints. I want to draw a distinction between the annoyance
and complaint. Many more people can be annoyed by the noise than
would complain, for all sorts of reasons which I shall come on
to a little bit later on. We think it is inappropriate that a
design aim should be based upon reported formal complaints. Mr
Thornley-Taylor mentioned this particular study which was looking
at complaints received in the 1960s and 1970s, and I think it
followed the opening of the Victoria line extension at that time.
It was reported in 1980, and I was heartened to hear Mr Thornley-Taylor
say that he felt that that information was long-in-the-tooth at
the time. It means that I do not have to dwell on it too long.
All I will do is read out a passage from the front cover of that
particular report. "The report is based entirely on complaints
and that these indicate only that a threshold level of tolerance
has been exceeded. No information is available to indicate the
degree of nuisance or annoyance caused by the railway noise below
the level at which a complaint is stimulated. The criterion developed
in this report does not, therefore, represent an acceptable limit
of noise but only a threshold point on a scale of annoyance. Further
environmental and practical factors will need consideration in
order to determine an acceptable and reasonable limit for underground
railway noise." This is a fairly straight talking summary
of that particular study. Mr Thornley-Taylor also alluded to all
of the measurements being over 40 in the survey. Inspection of
those complaint histories actually indicates that we have levels
around 40 and below 40 at certain points within buildings, for
your information. The next slide is LBC22. I would like to talk
a little bit about the Jubilee line extension. I put "circa1991"
but I am not entirely sure when the environmental statement did
appear for this one. We draw on the text I am quoting under item
a) from the Promoter's technical reports for Crossrail. The environmental
statement for the JLE identified 35 dB(A) as a desirable level.
This can be sourced directly to the APTA guidelines. They went
on to talk about 40 dB(A) as a potential complaint threshold,
based upon the study we have just talked about.
2904. Mrs James: We have very carefully
had explained to us the difference between the levels, and the
level to the power of 10 is actually a doubling. What does that
35-40 actually represent? Is it one-and-a-half times?
(Mr Methold) It is a trebling
of acoustic energy.
2905. Mrs James: Thank you.
2906. Mr Clarkson: Let us get this clear:
a trebling of acoustic energy between 35 and 40?
(Mr Methold) That is correct.
2907. And loudness?
(Mr Methold) Loudness, it
would be somewhere between zero and 10. We know that doubling
of loudness is 10. We can suggest that might be half as loud as
maybe a doubling of loudness.
2908. If I can lead on the arithmetic, and you
will correct me if I am wrong, I think to answer the hon member's
question, is that a 50 per cent increase?
(Mr Methold) It is an over
3 dB increase, so it is a noticeable difference. That is probably
the best way to describe it. The studies that indicate that 10
dB is a double of loudness is based on laboratory experiments
where somebody is asked to push a button in a booth when they
felt that a noise level sounded twice as loud. The JLE project
went for the design aim of 40 dB based upon a complaint threshold.
Item c) here, we put unchallenged here; Mr Thornley-Taylor made
the point that this was an unchallenged design aim at the time.
2909. Is the question mark deliberate?
(Mr Methold) It is deliberate.
It is interesting to read the transcripts of the Channel Tunnel
Rail Link Bill where the specialists advising the local authorities
and the specialists advising the Promoter were the same people
involved in the Jubilee line extension. What we can see when we
read the transcripts is that the specialists advising the local
authorities most certainly did want to challenge this design aim
and were engaged in discussions with the Promoter right up to
the point at which the Committee was due to take hearings. The
specialist advising the local authorities was under the impression,
or believed, that he had secured an informal assurance that 35
dB would be a desirable level and would feature in an undertaking.
We can see when we read the transcripts, however, that that did
not transpire, and that I think it was regarded as a misunderstanding
between the parties. This is an unfortunate misunderstanding because
it marks a point in history at which 40 decibels has been argued
to be unchallenged through the Select Committee process. Can I
have the next slide, please? Crossrail Social Survey. I talked
about the London Underground study being related to complaints.
This is actually an attempt to try and relate it to annoyance,
a more appropriate indicator for design aim purposes.
2910. First, please, before you go any further,
is this table, LBC23, a lift?
(Mr Methold) It is. It is
a lift straight from the main social survey report. We have two
reports that comprise the social survey. There is the actual social
survey report done by a specialist organisation. We then have
an interpretive report undertaken by the Promoter's expert witness
on that information. This is from the main social survey report.
The first thing to notice—and I will explain a little about
the table, first. We have categories of noise level on the left-hand
side of this table: less than 35.9, 36 and 39.9, and then 40.
We presume this to mean 40 plus. This is a direct lift from the
report itself. The second, third and fourth columns are then documenting
the number of respondents who said they were either "very
annoyed" "quite annoyed" or "not annoyed"
at certain noise levels. These are grouped in the table. The first
thing to notice about this is that we only have 26 respondents
on the survey where information is useable. The second thing to
notice from this table is rather peculiar patterns. We have three
respondents "very annoyed"—all of them below 40,
we should point out. Nobody is "very annoyed" above
40, and yet we have "quite annoyed" respondents above
40. These are very unusual patterns and really start to indicate
very early on that we had an unusual data set because it is so
small. This is a very small sample of information.
2911. Kelvin Hopkins: Should it be assumed
that the people who were "very annoyed" at 39.9 were
also "very annoyed" at 40?
(Mr Methold) It cannot be,
not according to this survey. Can I have the next slide, please?
This is a graph taken from the interpretive report. It is documenting
the responses of those 26. Along the left-hand side we have "not
at all annoyed", "not very annoyed", "quite
annoyed" and "very annoyed" on the scale. Along
the bottom we have the measured noise level. The crosses, obviously,
represent the measured noise level and the particular noise annoyance
response.
2912. Mr Clarkson: Just so we get it
absolutely clear: this document is produced by whom?
(Mr Methold) This is produced
by Mr Thornley-Taylor on behalf of Crossrail.
2913. Should the base data be the same as your
previous exhibit?
(Mr Methold) It should be.
I am hoping it is the same data. We can see that there are only
four respondents here; quite annoyed and very annoyed, in total.
It may mean that some of the crosses are on top of each other.
2914. Kelvin Hopkins: Was any attempt
made to establish whether or not some of these people were hearing
impaired? That is quite significant. If they all had a hearing
impairment they would have a different reaction.
(Mr Methold) That is perfectly
true and I do not have the answer to that. It is not documented
whether we have hearing impaired—
2915. Mr Clarkson: I think in the survey,
as I recall reading some of the material, some people refused
to turn off their televisions in one survey, so they could not
get a proper survey. Is that the sort of problem you face?
(Mr Methold) Indeed, I believe
that over 600 questionnaires were sent out for this survey and
they had responses from 277 but only managed to get noise and
vibration measurements within 27 sites. This is an incredibly
small sample for this type of exercise but the whole point of
this is to try and identify what we call a dose response curve,
a way that we can predict the likely annoyance based upon a set
noise level. The diagonal line you see through this data set is
an attempt to try and develop this dose response curve, so it
is a best fit line. Statistically it has been generated through
a mathematical process.
2916. Mr Clarkson: I have no doubt that
some in the room will not want to be too interested in the detail
of this but there is one area that I ask you to explain to the
Committee and that is the top right which is a statistical—well,
you tell us. What is it?
(Mr Methold) This is an
indication of the statistical confidence of that diagonal line
through that data. It comes out of a mathematical process. An
R squared at 1.0 indicates that the prediction model of the diagonal
line will give you a 100 per cent right answer every time. What
we have got here is something that is saying that less than six
per cent of the annoyance can be attributed to the noise level
LAmax.S. This is a very important conclusion. Statistically this
is an incredibly weak relationship we have here. It is almost
obvious to see this. We have slightly below 40 and we have a whole
gamut of different responses across the board, which is a very
clear indication that we have a sample that cannot be used as
a scientific basis for this type of exercise. There is one final
point and it is a very important point that I want to make about
this survey and that is to do with measurement position. We know,
reading through the social survey reports, that all of the measurements
undertaken for this survey were conducted in the corner of a room.
I will draw you back to the Design Aims that we have seen and
the assessment criteria that we have seen for Crossrail which
refer to the middle or near the centre of the room. We can expect
up to three decibels difference between a measurement taken near
a wall and near the centre of the room. We can expect up to an
average of six decibels difference between a measurement made
in the corner of a room compared to the centre of a room.
2917. Mr Clarkson: Explain.
(Mr Methold) What this means
is that this social survey is based upon noise levels at the upper
end, so it does not relate directly to the assessment criteria
that Crossrail are putting forward. In other words, if Mr Thornley-Taylor
is using this to defend 40 decibels what it actually means is
that at the centre of the room this could be round about 35 decibels.
2918. Mr Clarkson: I would like you to
flesh that point out about the corner and the wall a bit more
for the Committee.
(Mr Methold) You would expect
more reflection near the corners of the room. If you are measuring
the middle of the room certain other aspects kick into the acoustic
requirements, such as carpets absorbing some of the reflections,
which you may not get when you measure near a corner. You also
have, if you like, three sound sources from each of the walls
in a corner position radiating at the microphone in close proximity.
2919. Mr Clarkson: What was the historical
approach to that sort of measurement?
(Mr Methold) Historically
people did tend to try and measure in a corner because it was
easier to get consistent measurements but recent publications
and in particular the Association of Noise Consultants have published
a book, which I believe is called Guidelines for the Assessment
of Measurement of Groundborne Noise and Vibration, of which
Mr Thornley-Taylor is an author. This is quite clearly stipulating
the preferred location for assessment of groundborne noise as
being near the centre of a room. This is the point we are making
here. We are not comparing like for like. Can I have the next
slide please? These are extracts from Mr Thornley-Taylor 's interpretive
report on this social survey study and I would like to read them
out: "Because of the small size of the sample of interviewees
for whom both noise and social surveys were achieved, the confidence
limits of the results are very low." The second extract reads,
"Because of the small number of results, the survey does
not serve to establish, with the degree of confidence which attaches
to the surveys of noise from above-ground transportation, the
relationship between noise from underground trains and its effect
on people. It nevertheless adds to the existing small database
on the subject although any interpretation placed on the results
must be treated with caution." These are very serious warning
signals for the use of this information and I do not believe they
should be the basis for defending the project's position that
40 dB is an appropriate design level. Can I have the next slide
please?
| 
