'There is always one thing that stands out as a bigger risk than anything else on a project, ' says RLE technical director Mike Glover. 'On this one it was the Thames tunnel. The reason was its location across a bend in the river, which still has large ships on it. Ships don't go round bends in a straight line - they broadside their way around. That made it difcult for us to use a barge and gather good borehole data through the river bed.
'What information we had indicated int in chalk, but we didn't have enough data to be certain how much. The information also showed us that the chalk was heavily ssured, with large drainage routes going into the chalk to depth.' Increasing the depth of the tunnel was impossible: 'We already had it at the maximum gradient, 1:40, and couldn't push it down further [to find more intact chalk] because we were constrained by the QEII bridge to the north and Galley Hill Road to the south, ' says Glover.
'The Thames tunnel 3km long] is relatively short by our standards.' But unable to quantify the flint content, RLE was unsure how much damage would be inicted on the tunnel boring machine (TBM), and therefore how great the risk of breakdown would be. But it did know that if a TBM failed while working under up to 4 bar of water pressure the project would be in trouble.
'We needed the tunnel complete to get on with systemwide [railway infrastructure] work.
We'd decided our rail head should be on the south of the river because on the north the route is mostly in tunnel or on viaduct so it's difficult to gain access.' Delay to the tunnel spelled delay to the rest of the project.
'We had to ask what would happen if the machine didn't make it or got stuck. To mitigate the risk, despite having a very high end design cost, we decided to invest in a second, standby TBM. As it turned out, we probably could have made it with just one machine, but because we had it we used it, ' says Glover.
As with the London tunnels, the risk mitigation strategy lay in specification of the TBMs.
'These were the only slurry machines used on the job and were equipped with rock crushers to break up the flints, ' says Alan Myers, who was RLE's project manager for the job.
'It was a fantastically successful crossing. We did our first drive in seven months and the second in five.' Joint venture contractor Hochtief/Murphy drove their Herrenknecht Mixshield TBMs through alluvium, peat with clay base and gravels before getting to the chalk. All were water bearing. As feared, the chalk had 'big ints in it', Myers says.
But the TBM worked better than hoped. 'The advantage of the chalk slurry was that it balanced water pressure and gave a lot of lubrication to the head, which was equipped with disc cutters to chop the ints up before they were pulverised in the crusher. The slurry meant there was very little wear at all.
'The critical issue wasn't driving the machine, but dealing with the slurry, ' Myers says.
Spoil was disposed of as fill at a disused Kent quarry, earmarked for development.
'We had to turn the slurry into an engineering ll material and that was a real challenge. To demonstrate it was sufciently dry, the slurry left site via a conveyor belt - if it wasn't up to scratch you couldn't get it up the incline of the belt.'
During the first drive it became clear that the array of hydrocyclones and centrifuges used to dewater the slurry was not large enough to process the volumes being produced.
'It was holding us up. During our first drive we planned to achieve 90m per week average.
We met our expectations, but put out a peak of 130m per week. We knew we could produce higher average progress and made changes to the management of the slurry system, ' Myers says.
Additional drying muscle - more cyclones and centrifuges - enabled Hochtief/Murphy to clip two months off the second drive, averaging 130m per week and peaking at 190m.
The Thames tunnel ontract employed what were, at the time, cutting edge technologies.
A 'soft eye' - an area reinforced with glass fibre rebar rather than steel - was installed in the end wall of the TBM reception chamber to enable an easier break through. 'We put a cap on the outer side of the wall and pressurised it so that the slurry system remained effective. That saved a lot of ground treatment around the portal, ' Myers says.
Compressed air was used to replace picks, requiring operatives to go through decompression using oxygen after the work was complete.
'That also was a novel feature for the UK.'
And the back up TBM was ready to go for the second drive as soon as the first drive was complete. Myers says grouting was needed to stop up ssures around the locations of the tunnel's central cross passage, which had to be excavated by hand, but nowhere else.
Despite anxieties about the bored tunnel, Myers says the real technical difficulty encountered was at the portals, 'which consumed about two-thirds of the [£140M] budget'. Approaches to the portals are in 15m deep retained cutting, excavated in marshy ground, 'with the water table at zero', Glover says.
The combined length of the north and south approaches was 1.2km. Diaphragm walls were toed into underlying chalk by about 5m and the individual D-wall panels were 7m wide - thought to be unprecedented in the UK - to reduce leakage.
Dewatering was used to enable excavation within the diaphragm walls, with an inevitable impact on the surrounding water table.
'There is an array of highly sensitive oil and gas tanks close to our north portal, ' says Myers. 'Their owners were demanding huge compensation if our work disrupted their activities. There was no room for differential settlement so as well as dewatering we introduced groundwater recharge to control water levels around the tank farm.'