The paper that follows is the 2010 Sir Harold Harding Memorial Lecture. This lecture is the latest in the series of lectures given by eminent speakers on tunnelling and related subjects at the British Tunnelling Society.
Innovations in Technology, 1950 to Today
There have been extensive innovations in technology from 1950 to today. They have had to keep up with the development and sophistication of these machines:
¦ For surveying techniques the use of GPS, total stations, gyro-theodolites, TBM guidance below ground with Z instruments or VMT, together with laser beams. All these methods are universal today.
¦ Welding techniques have developed considerably and have been one of the factors in allowing the growth in size of TBMs above 15m diameter.
¦ Finite element design has also been a factor in design development to enable the growth in size and the handling of earth pressures up to 8 bar.
¦ PLC control for the required sophistication in the TBM control cabin.
¦ The use of conditioners, especially surfactant foams, has been a giant step forward for EPBMs.
¦ Main bearings and bearing sealing systems have developed in order to cater for the very high thrusts that are used in large diameter closed face machines and to resist earth pressures up to 8 bar. However this remains the most vulnerable system on closed face TBMs.
¦ The ability to detect wear on the cutterhead, the cutting tools and within the screw conveyor, has been getting increasing attention. Wear of pick cutters can now be detected remotely. However the technology is not yet in place to do the same for disc cutters.
TBM Design Development
There are various subjects which are described as design development.
¦ The articulation of the machines is important in order to be able to steer correctly. It helps in keeping the TBM on line and also in negotiating tight curves. It also allows the cutterhead to be withdrawn from the face without drawing back the whole TBM and consequently damaging the tailseals. There is either passive or active articulation. It’s important to use the correct one, to suit the alignment of the tunnel.
¦ Thrust and torque have been mentioned already. Closed Face TBMs must never be thrust or torque limited. Thrust has to be able to overcome a) the earth or slurry pressure, b) the force needed to bury the cutting tools, c) the skin friction of the shield, d) the friction between the tailseals and the permanent lining and e) the drag of the trailing sledges. It also has to assume that a percentage of thrust will be taken up by steering corrections. Torque is more difficult to calculate. However there is a useful empirical formula for both STMs and EPBMs which is based on the formula aD3 where D is the TBM diameter and a is the factor. For STMs a is 0.75 to 2.0 and for EPBMs a is 2.0 to 3.0. This assumes for EPBMs the use of foam conditioners which reduce torque by as much as 50%.
¦ Cutterhead drive units. Manufactures are turning more and more to the use of variable frequency drive (VFD) for driving the cutterhead. It is more efficient than hydraulic drive, produces less heat and produces less vibration and noise.
¦ The development of cutters has been considerable, the resistance to wear has also been improved considerably. For disc cutters, which are normally 17” it is now quite common to use 19” cutters. There are two major advantages when using 19” cutters. Firstly, they ill take a thrust of 30 tonne, whereas 17” cutters are limited to 25 tonnes. Secondly, they are more resistant to both wear and mechanical damage.
¦ Wear is a major subject and must always be considered in relation to the ground conditions. This is referring to wear of the cutter tools and the cutterhead for both types of closed face TBMs. In addition for STMs it is vital to control the wear in the slurry pumps and pipelines and in EPBMs it is vital to control the wear in the screw conveyor. Control of wear is aided considerably in STMs by the bentonitic slurry and in EPBMs by the foam, bentonite and polymer conditioners. Mixshields excavating in rock have a particular problem in controlling the wear in the slurry pipes and pumps.
¦ Measurement of excavated quantity is vital. On EPBMs the latest and most accurate method is the use of belt weighers, preferably mounted on a dedicated, constant tension belt. In the good old, bad old days we just counted muck skips which was extremely inaccurate. The skips often included a lot of water. Today the best method is to use two belt weighers. The belt weigher is suspended between rollers with pressure cells below. Laser profilers are also used, which are supposed to give greater accuracy. Belt weighers will probably measure the excavated spoil to an accuracy of +/-10%. However their main usefulness is that they will give a comparative reading from ring to ring. So they would usually be used, say, on a 5 ring rolling average, so that it can be quickly seen if there has been a serious problem of over excavation. On STMs the use of magnetic flow meters and infra red density meters on the inbound and outbound slurry lines is the standard method. It should be noted that on both types of closed face TBMs the spoil measurement has to take into account the varying ground density, the varying moisture content and the amount of conditioners being injected into the excavation chamber.
¦ The sophistication of today’s tail seals is enabling the use of very high earth or slurry pressures up to 8 bar. Tailseals have become standard and use multiple wire brush seals injected constantly with a fibrous grease whenever the TBM is moving. This type of tailseal was developed by the Japanese but based on an original NRDC patent dated 13th Dec 1971, which arose out of the New Cross Experiment, see Figure 19. These tail seals are now designed to be capable of resisting up to 13 bar of pressure.
¦ Main bearing and seal systems. This is the heart of any TBM and has gone through considerable development over the last 4 decades. Main bearings are now universally triple roller bearings made with great precision by specialist companies. They are normally designed with a d10 life of 10,000 hours. Seal systems are in some respects more sophisticated mechanisms than the bearings. They are also extremely vulnerable when working against high earth or slurry pressures and any failure in the system will expose the main bearing to contamination, which, if it happens, will result in rapid bearing failure. Figure 20 shows the main bearing and seals on the Hallandsaas Mixshield. An important consideration in the sealing system design is the labyrinth at the front of the sealing array. This is injected with a fibrous grease which extrudes out into the excavation chamber and protects the lip seals, which have a normal capacity of resisting 4 bar per seal.
¦ The erection of the pre-cast tunnel lining. It is becoming universal to use vacuum erectors and vacuum handling for transfer cranes. This has virtually become a necessity due to the very large weight of segments involved in large diameter TBMs, or tunnels which have to resist high ground or hydrostatic pressures resulting in very thick linings.
¦ The ability to steer TBMs accurately is vital. Most manufacturers use articulation of the body, whereas some manufacturers use an articulated cutterhead, which has the ability to move axially and also radially in any direction. Articulated TBMs can be articulated passively or actively, or sometimes a combination of both. Active articulation is a positive articulation between the front and central body of the TBM. This will allow, as well as giving the ability to articulate, the ability for the front body to move axially in relation to the central body by as much as 100mm. Passive articulation is normally the articulation of the tailshield and controlled by passively connected hydraulic rams, which results in the tailshield passively following the central body and in essence shortens the TBM length in terms of steering capability.
¦ The ability to control the earth or slurry pressure is essential. In EPBMs it is achieved by a combination of the correct design and use of the screw conveyor and the use of spoil conditioners, especially the surfactant foams. The length of the screw has to be designed to suit the maximum earth pressure. There is an empirical formula which gives the length of screw to resist pressure. It is 0.2 bar per flight and assumes each flight takes up 0.6m of screw length.
¦ The ability to deal with boulders is most important. The cutterhead on one of the Storebaelt EPBMs with both picks and disc cutters. The disc cutters break up the boulders and also protect the pick cutters, which are required for excavating the granular soils. On STMs it is also necessary to have a crusher to break down the boulder fragments, that pass through the cutterhead to a size suitable for pumping. This is normally one third of the pipe diameter.
¦ There has been marked growth in TBM size over the last decade and a half and is referred to below in the description of some major projects. This has been quite remarkable. The first really large ones were for Trans Tokyo Bay Highway Tunnel, at 14.14m, where there were 8 STMs. Then on the 4th Elbe tunnel at 14.2m. It was on the 3rd Elbe tunnel in the mid 1970’s that the modern type of EPDM gaskets for pre-cast linings were originally developed. This revolutionised the water tightness of tunnels. Before that, all tunnellers would say that you can’t make a tunnel perfectly watertight. Nowadays we aim to make tunnels totally watertight. Then the Groene Hart tunnel in Holland used a 14.87m diameter TBM which was quite a breakthrough for the Dutch. Then further down is shown the current world record in Chong Ming in China. A mixshield at 15.42m diameter. So it can be seen that the growth in size of closed face TBMs has been considerable over the last 16 years. There are also TBMs of over 16m and 19m now on the drawing board. Soft ground machines would never have reached this size without the development of closed face TBMs.
¦ The growth in maximum working pressure has also been considerable. On the French side of the Channel Tunnel it was up to 10 bar. On the 4th Elbe tunnel it was 4.5 bar, on Storebaelt it was 7.5 bar and on the Hallandsaas project, the machine has been designed for 13 bar. Currently on Lake Mead in the USA they are proposing to use a machine beneath the lake at 13 bar.
Major Projects using Closed face TBMs
The following projects are a selection of projects carried out during the last quarter of a century. Some of these projects could not have been carried out before the introduction of closed face TBM technology.
¦ The Channel Tunnel, French side used EPBMs for the marine running tunnels. The sub-aqueous tunnels were driven between 1987 and 1990. The Kawasaki Robbins EPBM is 8.7m diameter and was used on one of the Marine Running Tunnels. Hydrostatic pressures were up to 10 bar.
¦ The modern St Clair River tunnel was constructed in the mid 1990s using a Lovat EPBM of 9.5m diameter. This has historic significance as it replaced the original tunnel, which was constructed in 1880 and had become obsolete due to its inability to allow the use of double stacked container trains. There was very low cover of 5m below the St Clair River. This was equivalent to only half a diameter, which is extremely low for a sub-aqueous tunnel. The machine was provided by the owner using the OPP.
¦ The 14.2m Mixshield for the 4th Elbe tunnel constructed in the late 1990s. This project used 4.5 bar pressure in the excavation chamber. The machine had twin cutterheads rotating in opposite directions. This assisted the clearance of spoil from the centre and also reduces the effective torque in any one direction.
¦ The Storebaelt Railway tunnel in Denmark was constructed between 1989 and 1995 using four 8.75m diameter Howden EPBMs, as shown in Figure 21. Face pressures were up to 7.5 bar. The loose glacial tills combined with high hydrostatic pressures made the tunnelling extremely challenging. The most severe problem was wear to the cutter tools and cutterhead and the difficulty of entering the excavation chamber under very high pressures of compressed air. It was on that project I first became a consultant and also a gentleman after 35 years as a contractor – with apologies to all contractors!
¦ The Trans Tokyo Bay Highway tunnels constructed in the early 1990s used the first of the really large STMs at 14.14m diameter. This was a very exciting project at the time and was given a great deal of international attention. Eight STMs were supplied for the project by Kawasaki, Mitsubishi, Hitachi and IHI. Figure 22 shows one of these large TBMs.
¦ The Copenhagen Metro consisted of 7km of twin tunnel and was constructed in the early 2000s using two 5.76m diameter NFM EPBMs, shown on Figure 23. Closed face machines were used for keeping water out rather than for supporting the face. There’s a problem when tunnelling in Copenhagen because a number of the old buildings are set on timber piles. If the water table is lowered, then within 6 weeks the piles will start to deteriorate. This project is another example where the cutterhead open area was too small at about 23%. This was problematic when the machine was launched in the local glacial tills. The machines had extreme difficulty in excavating because the spoil was jamming within the cutterhead. Some modifications were made, in particular burning off the grizzly bars.
¦ The Madrid Calle 30 Highway tunnels were constructed between 2004 and 2008 using two 15.2m diameter EPBMs. The manufacturers were Herren knecht and Kawasaki. Herrenknecht used their previous technology of contra rotating cutterheads, only used before with STMs. See Figure 24. The centre cutterhead had one screw to take the spoil from that part of the face. The outer cutterhead had two screws which removed the remainder of the spoil. They were very sophisticated, as with a number of the Herrenknecht machines. But in fact the Kawasaki machine, which had a standard single cutterhead with a single screw, managed to achieve the same rate of progress. The other interesting piece of comparative data is that the Herrenknecht cutterheads had 31.6% open area and were driven hydraulically and the Mitsubishi machine had 43%, which is the Japanese philosophy for EPBMs, and used a VFD cutterhead drive system.
¦ The Channel Tunnel Rail Link (CTRL) used eight 8.11m diameter closed face TBMs to excavate 20km of twin tunnels under the river Thames and below the suburbs of East London during the period from 2002 to 2004. When comparing the Kawasaki EPBM, the Wirth EPBM, the Lovat EPBM and the Herrenknecht STM it can be seen that the Kawasaki TBM cutterhead has a large open area of 40%, whereas the Wirth machine was below 30%, which did cause a few problems in the dense dewatered sands at the start. This resulted in the need for modifications. Overall the tunnelling on this project was a major success.
¦ The Hallandsaas Railway Tunnel project in Sweden started in 1992 and will probably complete in 2015. The 10.5m Herrenknecht Mixshield can operate as a rock cutting machine in open mode or as an STM in closed mode. The fractured gneiss and amphybolite geology has hydrostatic pressures of up to 13 bar at tunnel level and is extremely challenging. At the time of writing the TBM has passed through what is known geologically as the Mollenback Zone, a very fractured horst, which has been cryogenically frozen in advance. The machine is designed for 13 bar and will complete the first drive in September 2010. The TBM will then be dismantled and used for the second drive. It is of interest that the TBM was delivered from the manufacturer with 17in cutters, but at a mid adit, half way along the drive, a new cutterhead with 19in cutters was fitted. There has often been argument as to whether 17in or 19in cutters are better. However the Hallandsaas 19in cutters have shown a large improvement in the need to repair cutters and to enter the excavation chamber. The use of cutters before and after the change has improved from 200 m3/cutter to 325m3/cutter.
¦ In San Diego, California, an outfall tunnel was constructed using a 3.98m diameter Mitsubishi EPBM. Because of the high sub sea hydrostatic pressure of 7 bar they used a 47m long screw conveyor, which may be a world record. They also used 2,000 tonnes of thrust, which is comparatively large for a 3.98m diameter machine.
¦ The Smart tunnel in Kuala Lumpur was constructed using two 13.21m diameter Herrenknecht Mixshield. The concept for this tunnel is very interesting. There are two road decks at the top and a water flood control area below the bottom deck. If there are high floods the whole tunnel can be closed to traffic and flooded.
¦ The Groene Hart tunnel in Holland was constructed in the early 2000s using a 14.87m diameter NFM STM, see Figure 25. The Dutch are great maritime engineers, so for some of their early metro works they used maritime solutions, such as a line of compressed air caissons through Amsterdam. They turned to tunnelling methods fairly recently. But have now taken it up seriously. They have also constructed the Westerschelde tunnel using Mixshields, which was again a very difficult tunnel to construct.
¦ The Rio Subterraneo water tunnel was constructed in Buenos Aires, Argentina during the period 1997 to 2000, using two 4.37m diameter Herrenknecht EPBMs. These machines were supplied by the owner using the OPP. The machines had difficulty steering when the contractor used open mode at the start of driving. It was concluded that this was due to the lack of the backward force of the earth pressure acting at the machine centre creating a turning moment in combination with the shove rams. Once the contractor turned to the use of closed mode the machines operated extremely well and broke some weekly records.
¦ The Chong Ming Highway tunnels in China are currently the largest use of closed face TBMs. The tunnels were constructed in the period 2005 to 2008 using two Herrenknecht Mixshields. In Figure 26 can be seen a picture of the Chong Ming lining which illustrates very succintly what an extraordinarily large tunnel it is.
¦ The Sheppard line in Toronto was constructed during the period 1996 to 1999 using two 5.9m diameter Lovat EPBMs. They were supplied by the metro authority using the OPP. Currently the metro authority is designing a new line, the Spadina line, which may also use the OPP.
¦ Crossrail in London is just starting and it will be interesting to see what type of machines are going to be used for that project.
¦ In Seattle, Washington State USA, they are replacing the Alaskan Way Viaduct with a tunnel under the city. It is currently proposed to use a 16.2m diameter closed face TBM. So who knows where these size increases are going to stop.
Conclusion
I strongly recommend that if a closed face machine is being used, it should be used in closed mode at all times, unless it has been designed as a mixshield. Use an accurate means of measuring the excavated volume, as a means of controlling settlement. Make the TBM all singing, all dancing within the factory and do not be left trying to modify it below ground. Always use tail skin grouting and make sure that grouting is taking place whenever the TBM is moving forward. Make sure that settlement monitoring is in real-time and is not catching up when it is too late. Always carry out a thorough ground investigation, it saves everyone money, especially the owner. Make the correct choice of tunnel boring machine type at the start. Use an adequate open area of cutter head. Double check the design of the main bearing seals, as these are extremely difficult or impossible to change below ground. And finally always be prepared for boulders of the unexpected kind.
The safety of the work force is considerably improved, tunnels can be constructed where none could be constructed before and settlement control has improved beyond recognition. Finally closed face machines give considerable assistance with the achievement of the five important criteria necessary for any form of construction namely: cost, schedule, quality, safety and the protection of the environment.
We have come a very long way in the 185 years since the Brunels started the Thames tunnel in 1825 and in the 46 years since John Bartlett brought out his patent for the Bentonite Tunnelling Machine in 1964. None of us involved at the start in the 1960s and 1970s envisaged that these TBMs could be used for the size and complexity of projects that I have described in this paper.
The future of this particular tunnelling art is in your hands. Make sure that it is done safely, to a high quality and with absolutely minimum effect on the ground. If you are currently about 25 years old who knows what you might be building in another 50 years. Just to speculate on this I find it really quite exciting. I hope that you all agree.
So ladies and gentlemen, the future is yours. So I would say keep your eyes on the horizon and this art will continue to evolve over time.
Figure 19: Wirebrush Tail Seal Figure 20: Bearing Sealing System Figure 21: Storebaelt EPBM – 8.75m Figure 22: Trans Tokyo Bay TBM – 14.14m Figure 23: Copenhagen Metro – 5.76m Figure 24: Madrid EPBM, Herrenknecht – 15.2m Figure 25: Groene Hart Rail Tunnel – 14.87m Figure 26: Chong Ming Tunnel Lining
