Your browser is no longer supported

For the best possible experience using our website we recommend you upgrade to a newer version or another browser.

Your browser appears to have cookies disabled. For the best experience of this website, please enable cookies in your browser

We'll assume we have your consent to use cookies, for example so you won't need to log in each time you visit our site.
Learn more

Water torture

Geotechnics DLR

Tunnelling under the Thames in east London was expected to be wet. The reality is far worse Amec tunnelling director Peter South tells Andrew Mylius.

This time last year Amec tunnelling director Peter South was feeling bullish as a bespoke Lovat tunnel boring machine (TBM) nosed into the ground on the first of two drives under the River Thames.

Anticipating int-laden and water-bearing chalk, the TBM had been built to a robust specification. Its picks and cutters were instrumented so that wear could be monitored every inch of the way. And should the cutter head become damaged, dry locations had been identified ahead of the machine, providing dry refuges for repair.

Referring to his diary, South says: fiWe launched our machine at the end of last April and had it buried by the beginning of May. We set up the pit bottom [to serve the TBM] and moved off properly in July. We then had rather a tortuous adventure under the river. fl This put the project two months behind schedule.

Amec is part of Woolwich Arsenal Rail Enterprises, a joint venture with Royal Bank of Scotland, which is extending London's Docklands Light Railway from its North Woolwich terminus 2.5km south under the Thames to Woolwich Arsenal (NCE 1 June 2006). Halcrow is designer. The twin bore tunnels are each 1.8km long.

Amec's fishockingfl experience on the first bore has informed its approach to the second, which got under way at the start of this month, says South: fiWe've modified the machine.

Principally, the cutters in the head have been redesigned with considerably more carbide [making them much stronger] and we've made significant changes to the grouting system.

But we've also simplified the computer logic which controls the machine, modified the water proofing capability of the segment erection system and other electrics, strengthened the gantries, improved the pumping arrangement and added an extra tail seal.

fiWe went down through alluvium, gravels and Bullhead beds [a matrix of flints, clays and silts] and into the chalk, and the torque on the cutter head started to increase. fl fiThere are flints the size of beach balls in the Bullhead beds and we were concerned there would been wear. We tried to stop the machine at lining ring 190 [285m into the drive] so that we could inspect the head, but water flows from the chalk were so great we couldn't get in. fl The TBM is an earth pressure balance machine. As the cutter head is thrust forward, eating away at the ground ahead of it, spoil falls between its spokes into a chamber behind, from where it is removed via a screw conveyor. To prevent collapse of the cut face, the screw conveyor is used to maintain pressure equal to or greater than that of the surrounding ground: full of earth, the screw behaves like a bung. Driving at slightly above earth pressure also pushes water away from the face filike stepping on wet sandfl, explains South.

In order to get men to the front of the machine to inspect the cutter head, spoil has to be emptied from the chamber.

If ground is dry then entry to the chamber is ideally done at atmospheric pressure. However, the Lovat machine was equipped with compressed air so earth pressure could be maintained if conditions were wet or the ground proved unstable.

fiYes, we had compressed air to get men in, but the chalk was so fissured we felt we'd probably blow air up through them into the river, fl says South.

Water pressure was a maximum of 3.5bar, while minimum cover to the river bed was a mere 8m. fiWithout compressed air, getting in was like trying to part the river. Water was gushing through the screw, which is 750mm diameter, and a 100mm diameter port. All we could do was lock the machine back into earth pressure mode and move on. We tried to get in again at ring 229 [343m into the drive] and had the same problem. fl South isa veteran tunneller with 30 years' experience under his belt, including the Channel Tunnel and London Underground's Jubilee Line extension, yet he describes the -rst bore of the DLR Woolwich extension as fiprobably the most dif-cult job I've worked onfl.

Morale plummeted and South chewed his nails to the quick as wear on the TBM worsened and it limped forward in search of a place in the chalk dry enough to allow access. It came to a halt a mere 15m from the -rst of the dry refuges Amec had identi-ed before tunnelling started.

fiWe eventually got in at ring 290 [at the 435m mark]. It was a case of so near yet so far. The machine couldn't go any further, fl South says.

Things were still wet. fiThere was enough water coming through the face to sink a ship, fl says South. But it was at least possible to pump intensively and keep water levels in check.

fiThe head had worn away - we lost 76.2mm of plate. The cutters were worn completely down. We couldn't have gone any further. We were very lucky really that we were able to stop where we did, fl South shrugs.

To repair the damaged cutter head South's team had to go through its spokes to hand dig a heading about 2.5m long and 2.5m high, shoring it with timber.

fiWe sat there in the chalk for three and a half weeks welding the head back up, building up its thickness, restoring its pro-le, putting in new pick boxes, replacing the cutters, putting on new chromium carbide Trimay wear plates.

fiYou need skilled welders and they have to be willing to work in the worst of conditions. There they were in a very con-ned space, with water pouring in. The head was rotated in increments so they were working on only a small section of it at a time. fl Despite the horrendous conditions, the TBM started again with a repaired head including redesigned rippers.

fiWe tried to get into the head on the south side of the river. The second planned inspection point was at an intervention shaft just next to the river, but things were bad there as well, fl South recalls.

fiAccording to our site investigation the chalk was supposed to be very closely jointed. Clearly it wasn't. fl fiThough grout had been injected into the ground to seal up -ssures ahead of the TBM's arrival, due to the presence of the river wall we were unable to make the block the length of the machine. The TBM head was driven into the grout block, which was theoretically impermeable, but the machine length was in untreated land. fl The BM as tted with a tail skin grouting system, enabling the 130mm annulus between the excavated tunnel diameter and the tunnel lining to be back-lled.

fiAt the intervention shaft we grouted tight up behind the TBM shield in case there was water following us. We injected a long chain polymer jelly grout behind the head, which soaked up the water like wallpaper paste, and eventually we got ow rates low enough to get in and change the cutter heads. But again the volume of water coming in was humungous. fl Back in earth pressure balance mode, the machine again moved forward but water began sluicing forward from the tail of the machine. This was washing out grout from the annulus.

It was discovered that the TBM's tail skin seals had fidisappearedfl, South says. To prevent grout injected into the annulus between excavated tunnel and tunnel lining from oozing forward the inside of the TBM shield was -tted with three rings of steel wire brushes. These were generously lubricated with grease.

But South thinks that as the TBM negotiated tight 300m radius bends the brushes were worn away.

fiWith water pressures of 3 to 3.5bar, the grout was just being hosed forward to the face. fl Amec solved the problem by using more polymer grout to staunch water ow before pushing the TBM forward and welding on a fourth ring of brushes just in front of the perished seals. By removing segments from the last tunnel lining ring one at a time, Amec's welders could also burn away the remains of the third brush seal and replace it, bit by bit.

fiWe used huge volumes of tail seal grease to keep the seals intact for the remainder of the drive, fl says South. For the second drive, all three of the TBM's original brushes have been restored and the extra seal retained. fiWe are positioning the shield so all four brushes are trapped within the tunnel lining at all times, fl he adds.

South continues: fiThe third inspection point was at ring 900 [only 450m from the end of the drive], which was well into the ground on the south side of the river. Water coming into the tunnel had gone from being saline to fresh - we were away from the inuence of the river - but when we stopped the machine we had to have two 225mm pumps going full bore and we had 1m of water in the tunnel invert.

fiWe were now going up hill and water was cascading down the tunnel like a river. The men at the face were changing cutters like crazy to try and get the work done so that we could shut up the face again. fl Launch pit labours Setting up for the second drive, Amec took advantage of changes to the north Woolwich site.

The launch pit for the -rst drive was cramped, forcing Amec to bury the TBM and then advance it in short steps as the rail-mounted gantries for segment erection, grouting and the muck conveyor were added behind. Initial mucking out was done by skip.

Over the past 12 months Amec has built a retained cutting to the wast of the launch pit, connecting to the rest of the DLR. This has enabled the whole TBM train to be laid out in a straight line, avoiding the need for start-stop tunnelling while kit is assembled.

Streamlining the launch process, combined with an accelerated re-t of the TBM, has meant that Amec started its second drive only a month behind its original schedule.

Progress in the -rst week was 100m. South says that the second drive should be complete in August.

Safety facts

AFR: No reportable accidents so far

Safety innovations: every individual is responsible for safety and is expected to intervene if they spot breaches or dangers. There is a total segregation of plant from people. Only one train is used at a time to service the TBM to eliminate risk of crashes. Rolling stock is -tted with speed sensors that will trigger automatic braking.

Have your say

You must sign in to make a comment

Please remember that the submission of any material is governed by our Terms and Conditions and by submitting material you confirm your agreement to these Terms and Conditions. Please note comments made online may also be published in the print edition of New Civil Engineer. Links may be included in your comments but HTML is not permitted.