The TBM is like a factory, industrialising the tunnelling process" says design engineer Werner Burger at German TBM maker Herrenknecht. His firm supplied all the TBMs used on the AlpTransit projects, four for the long Gotthard and two more on the drives for the 35km long Lötschberg, further west in the Swiss Alps.

But deep tunnels, with rock cover sometimes over 2,000m can bring additional problems he says. Pressure effects from such overburden loading can cause rock bursts and the danger of squeezing rock. Water pressures, if saturated faults are hit particularly, can be exceptionally high with a head of hundreds of metres and pressures potentially of more than 100 bar.

"That is well beyond the capacity of any TBM, even the most effectively sealed shield machines," he says. Currently pressures can be coped with up to around 15 bar and 20 bar is a realistic projection for future machines, but not much over.

"Further more the water will possibly be hot from the ambient temperatures of the rock around," he says. Even if the bore is dry, hot rock is a major issue with temperatures on the Gotthard rising to 50oC and to 30oC or more even on shorter tunnels.

First factors to be thought about are the robustness of the machine for drives that can be in the 20km range. "The mechanical side is not exceptionally difficult but the elements need to be thought about early on in the design, working in conjunction with the geologists to upgrade them as needed."

At US manufacturer Robbins, vicepresident Doug Harding says that a basic issue is simply ensuring robustness and longevity for components, "trying use higher design safety factors that allow for these long tunnels without breaking down; coping with extended main bearing life, and so on."

Greater robustness needs to be planned in conjunction with the likely maintenance regime says Burger "planning routine stops as you would for any factory system, synchronised perhaps with holiday periods." That in turn will depend on the tunnel design itself and whether there will be some chance of outside access to the machine, perhaps at points where ventilation shafts join or access adits. "It needs to be thought through quite carefully." Machines also need careful thinking to allow for more maintenance from inside the machine than shorter drives might require.

On the long Gotthard drives for example the hard rock gripper machines driving from the south portal at Bodio were given a major overhaul when they reached the caverns of the multifunction stations at Faido, after 14km of the initial bore.

"In fact they were reconfigured with a greater diameter for the cutting heads to allow for the rock conditions and potential convergence in the next even longer drives" says Berger. A thicker lining was required for those sections. Wear plates and protection were also changed.

As machines develop, the capacity to record multiple parameters of the equipment performance for different components, gives the capability for predictive diagnostics to be developed too. Software is evolving which can examine complex data patterns and help plan maintenance and allow anticipatory component replacement.

As well as the machine equipment itself, there is a greater need ancillary equipment on the machine in deep tunnels, both for rock support and for forward investigation. For machines with an "open" configuration, there is also possibly a need for shielding panels against rock bursts or simply falls, which are more likely in very deep tunnels.

Forward probing drills and a machine configuration to allow them free use ahead and at angles, is important. "The potential problems that can be met are severe, and particularly water might be discovered with extreme pressure so it is even more important to know what is ahead" says Burger. That can mean drilling but increasingly other forward detection methods are being developed using sonic or geophysical systems, though they are still evolving; where possible suppliers like Herrenknecht are keen to offer their own systems because the design is then well integrated with the machine, but third party options can also be installed.

Machine parameters can also help determine rock mass behaviour, as on the on the Lyon-Turin base tunnel project in the current 7.5km long Maddalena drive, now about halfway through. A 6.3m high performance Robbins main beam machine is being used specifically for data collection with recording of parameters from its own performance as well as other instrumentation. It is also mounted with extra tools for high advance rates and forward probes.

Drills have another function too, providing for forward ground treatment by grouting and injection of foams and other chemicals perhaps. These are increasingly used on the face too, to reduce friction effects and the very high temperatures that cutter discs can reach.

"We are also increasingly seeing the application of water spray for the cutterhead to cool the discs," says technical sales director Jean-Camille Brochard at French TBM maker NFM Technologies. Heat removal is especially important in the deepest areas where rock is hot.

As well as the mechanical side of machines there has to be better provision for workers on long drives where logistical access is limited. "So machines need more or better lunch rooms, toilets and safety chambers, and even the hand railings need to be thought about," says Robbins’ Harding. That may include thinking about associated equipment like the manriders bringing crews up to the machine, which might need better ventilation or air-conditioning he says.

Provision of workspace cooling may well be a major element around the design. It is not necessarily what the TBM provides itself says Burger, as coolers and chillers will be organised by the contractor, but there needs to be the space to accommodate the equipment and for the incoming water cooling lines that remove the heat. Internal lines may be needed on the machine to connect cooling and extra ventilation, all of which needs good forward discussion between contractor and machine maker.

Once cooling and other equipment is added "you also have to think about greater power demands for such equipment," adds Brochard, "and the implications all through the machine." The same goes for wear protection he says which may need to be heavier for the cutterhead face and edges, screw conveyors and other points, particularly for long drives. But special hard steels do not always weld easily to other components and intermediate steel is needed. This changes the weight and performance of other parts. Accessibility for changing such components needs be considered too.

NFM is currently building a large diameter machine for the Lyon-Turin project’s largest drive yet, being carried out by a joint venture led by French contractor Spie Batignolles. The tunnel is a test drive but will run 9km precisely along the main alignment of the tunnel from the St Martin de la Porte access tunnel on the northern end of the scheme. The TBM will be 9.5m in diameter.

Unlike the gripper TBMs used on the Swiss tunnels, this will be a shield machine with the capacity to erect segments behind it. The main reason for the choice is that the rock conditions expected on the drive are for softer rock, which could cause difficulties for a gripper machine which must thrust against the tunnel sides to push forwards, whereas a shield pushes off against the segment ring. The TBMs on the Gotthard did have some difficulties on small parts of the drives for just this reason, slowing down in some sections significantly just inside the Bodio portal for example.

A shield configuration is already being used at the Koralm tunnel in Austria where two Wirth machines are making drives of 17km and 16km through hard rock, but with blocky conditions. Here the main advantages are logistic because the tunnel is accessed by two 60m deep shafts, making supply of shotcrete, steel and mesh more difficult than ready made segments. But rock temperature is also a factor. "Eliminating the need for shotcrete limits the associated heat generated by hydration," says tunnel design coordinator Hanns Wagner with tunnel client ÖBB-Infrastruktur, at Austrian railways. "The overburden here gives us 30oC which is just within limits and shotcrete would add more, creating a need for cooling systems." Use of segments, with waterproof gaskets, also means that apart from a few sections in bad rock, there need be no inner lining for the tunnel, again reducing hydration heat effects.

"And finally there is additional safety for the workers against rock falls or rock bursts when there is a segmental lining." For the French project these factors are important too but the rock conditions more so. And the choice does involve some risk as there is a danger of squeezing ground, as already encountered in the Saint Martin de la Porte access tunnel where deformations of over a metre had to be dealt with.

"You must also carefully design against the segments being damaged by irresistible squeezing forces," says Lyon-Turin joint director general Alain Chabert. Fortunately the main alignment is not in the same highly heterogeneous carboniferous formation which created the access problems. "The rock is carboniferous but of a different type – Brequin Orelle – which is less squeezing we anticipate. But we will find out, that is part of the point of the investigation tunnel."

In case of problems however the NFM machine will have special features including overcutting at the face by up to 250mm. "The cutter head has two cutter discs which are extendible on hydraulic cylinders to give an extra 100mm says Brochard. "Once some space has been made the machine can have the head further extended using shims, to a maximum 250mm."

The head can also be offset from the axis to make an asymmetric overcut. Other features required on the machine are gas detectors for a wide range of gases including hydrogen sulphide, carbon monoxide and dioxide, oxygen, and sealed cabinets for the electrics to prevent sparking.

A significant consideration for segment methods is the advantage of not needing a waterproof membrane and therefore additional inner lining work. At Koralm segments are fitted with gaskets offering with a 2 bar resistance at a computer controlled factory on site equipped with Herrenknecht moulds. For Lyon-Turin’s Saint Martina de la Porte 9km probe tunnel Italian company Fama is supplying both moulds and its own gasket system with a patented method of anchoring the gasket in the segment concrete. A segment lining is also being installed on the next section of the Brenner base tunnel’s exploratory drive, again despite the risk of squeezing. A Herrenknecht machine, of approximately 7.8m external diameter will drive 15km southwards from an existing 6km long section at Innsbruck, through Innsbruck quartzphyllite.

The rock bedding orientation will be perpendicular to the drive it is expected from the initial, conventionally done, drive, allowing for enough arching effect to prevent machine jamming. But this is an investigation and will find out if such machines can do the main drives.