The worst nightmare for prospective tunnel engineers on the recently approved Thames Tideway tunnel scheme is the thought of a catastrophic tunnel boring machine breakdown. It is a major risk, especially in a heavily fissured section of the chalk that the tunnel line runs through for the eastern half its 32km length.
The $3.9bn plus project is for a gravity interceptor sewer to link up London's Victorian outfalls and provide storage capacity for storm flows. The large diameter tunnel will run deep to allow a steady 1.2% gradient and will cut beneath the tangle of London's multitudinous rail, tube, power and other tunnels, including secret defence networks.
The start point in west London will be 40m deep, with the tunnel descending to 80m at the eastern end. In saturated ground this translates to an eight bar outside pressure, higher than any current project worldwide ? except perhaps the troubled Hallandsas railway tunnels in Sweden which sink to 100m below ground.
Deep chalk layers beneath the Thames contain flints of almost diamond hardness that are expected to take a heavy toll on the TBMs. But accessing the face or even the pressure chamber behind the face on a slurry or earth pressure balance (EPB) machine would be exceptionally difficult.
Tools can be replaced from behind the face 'but you have to get into the chamber', says an expert at one of the major TBM manufacturers.
'There are machines that allow access at normal pressure from inside the radial arms driving the face, but that is only for very large machines.' Diameter of the Thames tunnel will be 7.2m which, while large, is not big enough to allow for this technology.
Unexpected breakdown is possible, particularly in the sections running through chalk (see diagram). 'There is flint in layers in all chalk and it is very difficult to avoid, ' says David Gutteridge, tunnelling specialist at consultant Mott MacDonald.
Flint is second only to diamond in its hardness and will score steel. It causes high wear on face and cutters, and also on the pumps and pipelines that would dispose of material in a slurry supported TBM.
'It was a challenge faced by the recently completed Thames tunnel for the Channel Tunnel Rail Link (CTRL), ' says Bill Grose, head of tunnelling at consultant Arup. 'But in the end that is a very successful project.'
But that crossing is further east than the new tunnel, which will finish at the Beckton processing works in east London. A second project in chalk, the extension of the Docklands Light Railway (DLR) to Woolwich, has not been so lucky (see page 18).
Unlike the CTRL, which used slurry machines, the shallower DLR is being driven with earth pressure balance (EPB) shields and because these fill the chamber with spoil, it tends to grind the mechanism more.
'We were expecting to make about 450m [between maintenance stops] with an EPB, ' says Peter South tunnelling director at Amec, the contractor building the DLR extension. 'But we only made 400m and had to repair parts with welding.
South says that not only is int wear significant but also that the pressures will cause heavier wear because of the additional exing and loadings on bearings and moving parts.
To carry out this repair work means getting at the face.
Highly specialised operatives trained to work in pressurised environments are an option but would almost certainly be ruled out in the UK by health and safety considerations. They would anyway have to work virtually blind and be limited to short working periods and long decompression. Divers went into a machine on the Netherlands' Westerschelde project, but it is rarely done.
But the other option, forming maintenance zones by grouting to form a watertight block as done on the DLR project, would also be difcult warns South.
'Usuallyyou would do that from the surface but the line of this tunnel is underneath the River Thames and therefore you would need a jack-up barge.' That in turn creates river navigation issues, he says.
Grose is optimistic however.
Machines have improved greatly, he believes, and as long as a sufciently high spec TBM is chosen it should be possible to avoid breakdown in difcult spots.
Most TBM makers now supply hardened and armoured tools and wear resistant face plates for machines, though this would add to the cost.
The worst difficulties could be avoided by planned stops for the machine in better ground, says client Thames Water's project coordinator John Greenwood. 'If there is better ground, the ow rate inwards would be limited and it would be easier to inspect and repair the machine face.' Some more limited grouting would be needed.
He thinks that the worst of the flint would be on the transition between the chalk and the layers above, the result of historic erosion, meaning the machine would have its biggest problems over a 500m or so length.
An extra problem might be that the water in the chalk constitutes a major supply aquifer for London and its surroundings, and must not be contaminated. That means any ground treatment will be constrained. It also means the tunnel, which will carry sewage, must be well sealed.
To achieve planned maintenance stops implies good knowledge of the ground and extensive ground investigation is to be done, says Greenwood.
A variety of sources, including Thames Water's own extraction well boreholes, have already given the client substantial information.
This will be combined with British Geological Survey data and shared core information from the Crossrail project and the CTRL. 'And we expect to spend quite a few millions on detailed site investigation, ' says Greenwood.
Maintenance needs will affect the choice of machine type.
South says he favours EPB which 'leaves a good product in the ground' but the slurry type machine might handle int better, as solid material can drop through the 'thin' support slurry, which, in chalk, is likely to be simply water rather than bentonite.
It is likely that a slurry machine would need maintenance and tool changes every 1,200m or so whereas an EPB will need to stop every 500m or less.
But slurry would create problems for spoil extraction because chalk tends to break down into a mush and is difcult to handle at the separation plant, says South.