House of Lords - Crossrail Bill - Minutes of Evidence

Select Committee on the Crossrail Bill Minutes of Evidence

Examination of Witnesses (Questions 200 - 219)

200. MR ELVIN: Mr Thornely-Taylor, would it be worth jumping ahead because you have got two slides that illustrate this? If we come ahead to slide 48, we have got an example of resilient rail support.[75]

(Mr Thornely-Taylor) This is a form of rail support that is becoming more and more popular. I am not saying necessarily Crossrail will use it, but it is a good example of modern rail support. Instead of the rail sitting with the foot resting on a wooden block, there is an air space underneath the rail foot and the rail is supported by means of rubber blocks, one either side of what is called `the web', and those blocks are held in place by shoulders. The sheer stiffness of the rubber blocks is low, so the rail deflects under the passage of a train and there is no rigid mechanical connection whatever between the rail and the supporting concrete beneath. For most railways, for most receivers of groundborne noise, that alone is enough to bring noise levels well below the 40 figure that I mentioned previously, but there are circumstances where you need to do better than that if you are either running under a very sensitive building, like an important concert hall or a recording studio or MPs' offices. Portcullis House has the District Line running through its basement because, when the Jubilee Line Extension to Westminster Station was constructed, the District & Circle Line was completely reconstructed and the District & Circle Line station is completely new and is in fact a large bridge suspended over the huge void of the Jubilee Line Extension station. Through Westminster Station there is floating slab track where the rails are attached to the concrete block which itself is supported on rubber bearings.[76] That one was my design. I regret one thing which is that I chose to omit additional resilient support between the rail and the concrete block and, if you stand on Westminster Station when trains come in, they do rumble as they run onto the floating slab. They would not do that if there were also resilient support between the rail and the concrete as well as the bearings between the concrete and the structure, but I have failed to find anywhere in Portcullis House where you can hear trains. I hope that is still the case. There was a very exacting specification for that piece of work and, unless I am informed otherwise, it continues to be extremely successful. There are obviously engineering cost consequences to floating slab, but there are important sections of Crossrail that will have it for reasons which we will go into later in the proceedings.

201. LORD YOUNG OF NORWOOD GREEN: As one supplementary to that, and that was very helpful, but this is the Thornely-Taylor floating slab actually!
(Mr Thornely-Taylor) It is actually.

202. The one other question I wanted to ask you was about the design of the tunnel itself in terms of preventing transmission of noise and vibration. Is that a factor? Can you do anything with that?
(Mr Thornely-Taylor) It is not a very large factor. Many of the old tunnels are made of cast iron and modern tunnels are bolted concrete segments and the additional mass and rigidity of the concrete is beneficial, but most of the improvement is in the resilient support of the track.

203. LORD SNAPE: If we followed the French example, Mr Thornely-Taylor, and put rubber tyres on some of the trains, would it put your future job prospects in peril?
(Mr Thornely-Taylor) It would. I actually think my job prospects are in peril anyway because it is now quite normal for an urban underground railway to use resilient rail support, and you will have very low groundborne noise levels without trying so long as you do not mess it up. I used to spend vast amounts of time predicting noise levels from underground railways where it is no longer really necessary because it is a foregone conclusion that levels will be very low.

204. LORD SNAPE: And the French rubber tyres?
(Mr Thornely-Taylor) French rubber tyres do almost eliminate groundborne noise, but have a massive cost and engineering penalty attached to them, not least that it is difficult to run them to Reading and Shenfield and Abbey Wood.

205. MR ELVIN: I was going to ask about how you put rubber tyres on the Great Western! When you say that your job prospects are in peril, unless the resilient rail is messed up of course, there are design issues also in that it requires proper maintenance and inspection?
(Mr Thornely-Taylor) Yes. It is quite possible to mess it up and there are some examples of non-functioning resilient rail support. I will not actually give them, but it is also possible through neglect to let the rails acquire surface roughness which is so great that noise levels rise quite dramatically and it is of great importance that, along with the basic engineering design of the railway, there is a maintenance policy to grind the rails, for example, when the roughness grows. My Lord Chairman was kind enough to credit me with devising the London City Airport noise conditions, and I also wrote the noise policy for the Docklands Light Railway which has a system for detecting when it is necessary to grind the rails and when it is necessary to turn the wheels. I will be guided by Mr Elvin as to how we are doing for time and whether I should go back to the slides we have bypassed.

206. You are 20 minutes over estimate and on slide 37, but let us just see how we go.[77]

(Mr Thornely-Taylor) Slide 37 is continuing the story of the maximum noise level. I explained earlier on that there were two types of LAmax, slow and fast, the fast in the days when we were looking at a moving needle and the needle was jiggling about, and with the slow the needle was damp to make it easier to measure a single figure. For an underground train passing under a building, if you measure it in both units, you will find that they are two different, so, if you measured 40LAmax,S, it would come out at about 42LAmax,F, and that compares with the LAmax,F guideline value for bedrooms at 45, coupled with the advice that it should be lower for low-frequency noise.

207. If we just put up the actual page of the guidance, I hope that is legible at least on the screens, if not the main screen.[78]

(Mr Thornely-Taylor) This is an extract from those guidelines and I will read from paragraph (c): "Sources with low-frequency components. Disturbances may occur even though the sound pressure level during exposure is below 30dB(A). If negative effects on sleep are to be avoided, the equivalent sound pressure level should not exceed 30dB(A) indoors for continuous noise. If the noise is not continuous, sleep disturbance correlates best with LAmax", and it becomes clear later that it is LAmaxF that is being referred to, "and effects have been observed at 45 dB or less. This is particularly true if the background level is low. Noise events exceeding 45 dB(A) should therefore be limited, if possible. For sensitive people, an even lower limit would be preferred. It should be noted that it should be possible to sleep with a bedroom window slightly open", and this is irrelevant for groundborne noise because obviously it does not come in through windows, but for surface noise, "(a reduction from outside to inside of 15 dB). To prevent sleep disturbances, one should thus consider the equivalent sound pressure level and the number and level of sound events. Mitigation targeted to the first part of the night is believed to be effective for the ability to fall asleep." There are lots of passages of that kind of detail in the guidance which one can refer to, but I think that sums the current topic up for present purposes. Slide 38, if we are talking about underground trains passing during the night period, we are interested in LAeq as well as LAmax, as those guidelines were suggesting we should be.[79] A typical example for the Crossrail case would be 78 trains in the night period and they will be 200 metres long. If we assume we are going, as they will be in a lot of areas, at 80 kph, 40 LAmax,S comes out as about 24 LAeq 8 hour, whereas the eight-hour LAeq in the WHO guidance table we have been looking at is for 30, or BS 8233 says 30LAeq 8 hour is a good standard and 35 is a reasonable standard.

208. So, if it comes out at 24, that indicates that it is doing better than good?
(Mr Thornely-Taylor) It is indeed.

209. What if it said that for at least a proportion of that eight-hour LAeq it is unlikely that trains will be running, ie during the night?
(Mr Thornely-Taylor) We then have a difficulty in that we do not know what guidance to turn to because the only guidance on night-time LAeq in both the World Health Organisation document and BS 8233 is in terms of the eight-hour value, but, if you double the number of trains or compress all those trains into half the time, it would go up from 24 to 27 and, if you compress them all into quarter of the time, it would go up to 30. I think the current intention is that Crossrail trains, revenue trains, will cease at about 12.30 and start again at about 5.45 with some empty trains going to and from the depot in between, so, of the eight-hour night, about two and three-quarter hours are normal revenue service with some trains in between, so, even if you did find guidance which dealt with night in hour periods other than eight, we still would be significantly below 30. If we move on to slide 39, we see what the Environmental Statement bases its conclusions on.[80] The significant threshold in the Environmental Statement is 40 dBLAmax,S and the conclusion is that there is no significant effect from groundborne noise. It draws on a specialist technical report which includes a full set of contours of groundborne noise from tunnel portal to tunnel portal. This particular example is for Central London, and that is Soho Square there, Tottenham Court Road Station, so, as trains are travelling relatively fast between stations, the lowest contour plotted is 25, though it is quite hard to see, but the outer contour is 25 and the most we get up to is in the 30s and, as the trains slow down, the contours disappear off the map, so to speak. The essential thing is that the same track form is assumed from end to end, except for locations where, for the reasons I have alluded to, we need floating slab, but it is not going to be the case that the track form will be degraded wherever the prediction is for much less than 40, so we will not suddenly go back to rigid rail support just because we are well below 40. A track form will be used for the same kind, chosen to achieve 40 in as many places as possible, as a result of which in a lot of other places these contours show much, much lower levels than the overriding 40, design A.

210. LORD BROOKE OF ALVERTHORPE: Forgive me, just very quickly to go back to the previous slide, 38. The running time is 12:30?
(Mr Thornely-Taylor) Revenue service trains, the trains with passengers on them, I think it is true to say, taking Tottenham Court Road as an example, the last one will have gone by about half past midnight and there will not be another one until about 5.45. During that period half a dozen or so trains will be moving to and from depots.

211. LORD BROOKE OF ALVERTHORPE: We are planning ten years ahead not to change anything?
(Mr Thornely-Taylor) I am not an expert on those matters, my Lord.

212. MR ELVIN: To some extent the level of service will be predicated on the Access Option in due course.
(Mr Thornely-Taylor) We move from underground operational noise to the operation of plant on the surface. If we go to slide 40, here we bring in a new standard called "4142" which is used for all kinds of noise of an industrial or commercial nature.[81] To be specific, it applies to noise from factories, industrial premises, fixed installations or sources of an industrial nature in commercial premises. The most important application of it for Crossrail is for what is called loosely "fixed plant tunnel ventilation fans". Being the biggest example of fixed plant, the fans on all underground systems are very large. Those on Crossrail will be substantially larger than they are for underground lines because we are dealing not with an underground line, we are dealing with a main line in a tunnel with major consequences from that. There is a procedure set out in this British Standard. It is intended for predicting the likelihood of complaints and it does it by using LAeq but with a penalty of five notional decibels added to the physical level of LAeq if the noise has special characteristics. One of those characteristics is what is known as "tonal characteristics" and you almost always get tonal characteristics with sources such as fans. You compare it with a completely new index that I have not talked about yet called LA90, so the same family as the L10 I touched on in the context of highway noise, it is the level of noise exceeded for 90 per cent of the time. It is really measuring the quiet bits between noise events. If you are standing out in the street and it is not heavily trafficked, vehicles going by, aircraft going overhead and you wait for a quiet moment and you measure it, that is probably close to L90.

213. MR ELVIN: Is that what is sometimes loosely referred to as background noise level?
(Mr Thornely-Taylor) It is and it is illustrated in slide 41.[82] If you perform a statistical analysis on this varying ambient level, the L90 comes out more or less a typical low value, it is not to say occasionally it gets quieter than L90, but it is a typical low level and it is against that level that the LAeq plus five, if it is a tonal noise, is compared in the BS 4142 process. In slide 42 the standard concludes that if LAeq plus five, called the rating level, should differ from the L90 plus ten that is an indication that complaints are likely.[83] If it should differ from LA90 by plus five it is marginal as regards complaints and, going to the other extreme, if it is ten below the L90 it is a positive indication that complaints are unlikely. The reason for that is if a sound is ten dB less than a background level it is effectively inaudible. Moving on to slide 43, we come to construction noise as the first of the sequence of slides I have on mitigation, the ways in which noise is controlled and reduced.[84] Starting again with construction noise, it is controlled primarily by methods of working and modern plant. Construction plant has fallen in noise level significantly in the past couple of decades since BS 5228 was first written and extensive use of noise barriers, enclosures of noisy machinery, monitoring and management of construction noise is vital, but if all that still leaves levels above the relevant thresholds there is a noise insulation policy and the temporary re-housing policy. The line at the bottom is crucial, the Control of Pollution Act 1974 introduced a system of controlling construction noise. Section 61 of that Act provides for the voluntary application for a consent for construction works, which is devised to bring about the use of the best practical means to minimise noise, and a local authority can grant consent or grant it with conditions and, having done that, it becomes an offence to exceed or to contravene the terms of the consent. It is part of the Crossrail scheme that all construction sites will be subject to section 61 applications and, therefore, controlled in this way.

214. LORD BROOKE OF ALVERTHORPE: If it cannot be kept, can people be compensated?
(Mr Thornely-Taylor) The only form of compensation is through the payment for the installation of noise insulation or financing the cost of temporary re-housing.

215. LORD BROOKE OF ALVERTHORPE: Not for disturbance?
(Mr Thornely-Taylor) There is no simple --- There is another session on compensation, I would not presume to be a compensation expert, but I can say within the noise scheme there is no payment of money other than for physical mitigation of the kind that I have described. Slide 44 deals with vibration.[85] The vibration is not susceptible to noise barriers and enclosures, it can be controlled by methods of working, the timing of events that might cause vibration annoyance and particularly monitoring and management of any vibration that might cause harm to buildings or sensitive equipment. Interestingly, vibration is considered as noise as far as the Control of Pollution Act is concerned and the section 61 consent will deal with vibration in just the same way as it does noise. Moving to slide 45 surface railway noise will be mitigated by noise barriers, noise insulation where eligibility arises and, of course, given my comments about the importance of the rail running surface and the wheel tread condition, maintenance is also an important method of mitigating operational noise.[86] Going underground again, slide 46 [87]—

216. MR ELVIN: I think we have already anticipated—
(Mr Thornely-Taylor) We have just about done that, I think, and slide 47.[88] It is important to bear in mind that all the predictions show vibration that is perceptible not by the hearing but by the technical sense of touch, it is body movement we are talking about, disappears altogether in modern underground railways. The old fashioned railways with the direct fixed rails often produce both rumble and physical vibration, you either feel it or if there is a chair with its back against the wall and you hear the chair buzzing against the wall as the structures vibrate, that kind of vibration is universally predicted to disappear from the environment above modern underground railways as a consequence of the improvement in the track support system. We have seen the next two slides, the picture of the rail support and the floating slab, so we jump to slide 50 and we are very nearly at the end.[89] The fixed plant I was talking about, the mitigation is by the fitting of things known as "noise attenuators", loosely called "silencers" on fans. They are very large units, particularly for the enormous fans that we find on Crossrail and the acoustical engineering of plant installations is a well understood discipline and it is possible to reduce noise by design. That brings us to an end of the mitigation section. I have just got a few slides to talk about prediction methods and the uncertainty associated with them, given that we do rely on prediction in the planning of the railway. Slide 51, construction noise being an important potential impact, it is important to have an idea as to how accurate the predictions one makes will be and when the Crossrail scheme was first promoted in the early 1990s the Jubilee Line Extension was under construction and there were some major work sites, not just Westminster that I have already referred to but Canada Water and Canary Wharf.[90] The then Crossrail project financed a study of those sites to compare measured construction noise levels with predictions made using a variety of different approaches. The uncertainty was found to be confined within a limit of about 3 dB(A), which is actually a reassuring conclusion, and given that DS 5228 includes some basic noise information which is now very old, it is quite common to find construction sites that are much quieter than predicted. Slide 52 again is about prediction methods, this time surface noise.[91] There is a statutory procedure for doing that, produced to accompany the railway noise insulation regulations. Because of the importance of track quality, as we have been discussing, Defra commissioned a study to look at the influence of track conditions on the predictions you get and that recommended what it calls "a back end correction"—I see that I failed to get the "n" into "speed dependent"—it suggests that there should be a correction of two and a half dB for 100km/h for average variation in the acoustic quality of the track. Slide 53, the fixed plant I have been talking about.[92] There are standard procedures in the mechanical services industry for predicting received noise level at sensitive locations and the uncertainty associated with that process is built into the base assumptions and the prediction method that is used. Penultimately, slide 54, underground noise is predicted by simulating the running of the train in the tunnel numerically.[93] When Mr Elvin introduced me he referred to a finite difference computer package called "Findwave" which has been used in the preparation of the Environmental Statement and all predictions of groundborne noise which actually in a numerical form simulates with the passage of time in very tiny increments of small time steps the movement of a train over the rails on their resilient supports on the concrete in the base of the tunnel, including the tunnel wall, the different kinds of soil around the tunnel and the nature of the buildings above. There is uncertainty associated with it, there is a validation report in the specialist technical report, but the important consideration is that, although in detail there can be uncertainty—and slide 55 continues this topic—the prediction can come up with a peak in the noise at slightly the wrong frequency.[94] There can be uncertainties of that kind, but because we use high level indicators like dB(A) or NC curves when we are dealing with recording studios, they effectively remove the sensitivity of the results to small errors in the prediction such as peaks being at the wrong frequency, you still end up with a very similar high level index value such as a dB(A) level. We did extensive studies on the uncertainty of this prediction method, as a result of which the actual model outputs used for the Environment Statement and subsequent predictions have all had five dB(A) added to them for uncertainty so where, for example, there is a prediction of 40 dB LAmax the computer output actually said 35. That, my Lords, completes all my slides. I hope I have covered everything that is of interest, but I would be glad to expand on anything that I have not.

217. CHAIRMAN: Do any of my colleagues have any further questions?

218. BARONESS FOOKES: Could I raise very briefly something totally different. You referred to the impact on the human ear and its range of sounds; it is quite different for animals. I am wondering about the impact on domesticated animals and maybe living with human beings affected or not affected by these noises?
(Mr Thornely-Taylor) It is true to say that animals are affected in slightly different ways from humans, their hearing organs are different. Some of them can hear to higher frequencies, as we all know. In general, all the research into the effect of noise on animals shows them to be remarkably unaffected by noise unless there is a startle component to it. Sudden noises cause fright but, as you approach Gatwick Airport, you see a field of sheep grazing quietly in the field on the left as you go down the road. It is quite difficult to find any research results other than startle effects on animals. I think, in general, there is not a particular need to treat them differently from humans in that respect.

219. BARONESS FOOKES: That is all I needed to know. Thank you.

75 Crossrail Ref: P4, Mitigation-Operation-Groundborne Noise and Vibration-Resilient rail support (example) (LINEWD-RTT01-048) Back

76 Crossrail Ref: P4, Mitigation-Operation-Groundborne Noise and Vibration-Floating Slab (example) (LINEWD-RTT01-049) Back

77 Crossrail Ref: P4, Standards-Operation-underground railway (LINEWD-RTT01-037) Back

78 Crossrail Ref: P5, World Health Organization 1999, Guidelines for Community Noise , p46 (SCN-20080219-001) Back

79 Crossrail Ref: P4, Standards-Operation-underground railway (LINEWD-RTT01-038) Back

80 Crossrail Ref: P4, Crossrail contours of predicted groundborne noise LAmax,s (LINEWD-RTT01-039) Back

81 Crossrail Ref: P4, Standards-Operation-Fixed Plant (LINEWD-RTT01-040) Back

82 Crossrail Ref: P4, Standards-Operation-Fixed Plant (LINEWD-RTT01-041) Back

83 Crossrail Ref: P4, Standards-Operation-Fixed Plant (LINEWD-RTT01-042) Back

84 Crossrail Ref: P4, Mitigation-Construction Noise (LINEWD-RTT01-043) Back

85 Crossrail Ref: P4, Mitigation-Construction Vibration (LINEWD-RTT01-044) Back