The TBM breakthrough on the 2.4km long, 3m i.d. Dartford Cable Tunnel under the River Thames, in February this year, marked the successful return to form of British tunnelling contractor – Amec.
Having some of the UK’s biggest successes in the 1990’s under it’s belt, including the Jubilee Line C102, the contractor was one of the high profile exceptions when tunnel works were awarded on the Channel Tunnel Rail Link Section 2 in 2001.
Amec is keen to re-affirm its position on the larger diameter tunnelling market and technically speaking, the recently completed Dartford EPBM drive and associated shafts should stand it in good stead.
The project certainly presented some challenges, with its bore through highly fissured water-bearing chalk under 3.5 bar pressure without compressed air for interventions, the innovative shaft lining design required and, most notably, dealing with a seriously damaged cutterhead on the homeward stretch.
The project
The new tunnel will allow the relocation of two 275kV electric circuits, previously housed within the adjacent Dartford Road Tunnel. The new cable tunnel option was chosen in 2000 by the client, National Grid, following extensive studies by consultant, Mott MacDonald, as the least risky and cheapest way to solve increasing maintenance headaches created by the existing system within the road tunnel. The tunnel will also provide sufficient scope for future expansion.
In March 2001 National Grid employed UK consultant Babtie to develop the outline design and by early 2002 five contractors had returned tenders for the design and build contract that included the tunnel, a reception and a launch shaft, surface civil works, overhead line works and the removal of the redundant circuits.
Although Amec was soon announced preferred bidder, the US$29.3M Target Cost I-Chem Green Book contract, with a 50/50 pain gain risk share with the client, was not officially awarded until almost a year later in January 2003, due to a wrangle over land acquisition.
After completing the final design in early 2003, the Amec team mobilised on site and began construction of the Littlebrook launch shaft on the southern bank of the Thames in March the same year.
Shaft construction
Construction of the 40m deep, 9m i.d. launch and reception shafts have presented their own challenges. Both have been sunk using the wet caisson method. Amec designed the innovative lining required in conjunction with CV Buchan, an Amec subsidiary, who also cast and supplied the segments.
The 250mm thick, 1m wide, precast segments have no internal bolts with only external bolts used in pockets, with steel flange pocket plates between segments. To increase the weight, to help prevent later flotation, a 250mm thick reinforced concrete ring, called a wrap, was cast in 2m pours on the outside of the segments, increasing the thickness to 500mm.
The conditions at the Littlebrook shaft consist of 0.5m-3.5m of made ground, 14m of alluvium, approximately 1m of gravel, and then chalk. Harold Walls, Amec’s project manager explained: “The problem in the soft silts is the caisson weight. But in the chalk, it is shearing the chalk with the cutting edge. You have to build rigidity and friction into it before and whilst it is sinking, to slow it down. If you didn’t take steps to control this, the caisson would disappear, there’s just not the bearing capacity in the alluvium.”
Initial excavation in the fill allowed the steel cutting edge to be established on an un-reinforced concrete ring and the concrete guide collar to be cast. To promote internal friction within the caisson, Amec built four panel rings immediately above the caisson cutting edge (these panel rings would also provide a shear connection for the final concrete plug) and filled the caisson interior with crushed concrete hardcore. The panel ring’s main function at this stage was to key in the hardcore to the segmental lining thus increasing the bearing area of the caisson.
A mini excavator was then used within the caisson to sequentially remove sectors of the cutting edge support concrete and backfill again with the crushed concrete. The caisson was pushed while filling with hardcore continued to the top of the last panel ring.
The caisson was then pushed down, always maintaining a minimum of 1m of caisson above the guide collar before the next two rings were built and the wrap poured. The push was undertaken using jacks in 2m stages with no excavation taking place. Excavation was only carried out when the jacking system reached its maximum thrust and then only sufficient material was removed to allow the next 2m of construction to be undertaken. This sequence was maintained until the caisson reached the gravel layer.
By keeping the caisson full of material, maximum internal friction was maintained (as no friction could be achieved outside the caisson due to the surrounding 50mm of bentonite slurry) thus preventing the caisson sinking uncontrollably through the soft soils.
At 15m depth, when reaching the gravel overlying the chalk, the problem was over and normal wet caisson shaft sinking operations continued to the full depth.
Once sunk, it was the intention to install tension piles at the formation level and cast the shaft base followed by the grouting up of the 50mm bentonite annulus.
The chalk turned out to be stronger and less jointed than expected, which affected progress accordingly.
Partial trial pump downs were then undertaken to help estimate water inflow rates. It was also necessary to ensure, via piezometer monitoring, that there was no water connection between the alluviums and underlying chalk as this could have led to extensive surface settlement. Once this was established, and that the water recovery rate into the shaft could be managed, a full pump down was undertaken.
The shaft was then progressed to formation by mini excavator, loading into skips, and then grouted. An under ream of the shaft rings was carried out with any water diverted it into a gravel layer. The shaft base was then concreted and the remaining water pumped out by a 155mm diameter pump, installed in steel pipes cast into the base.
The launch shaft was complete in July 2003 ready to install and start the TBM drive.
TBM launch
Being launched at a depth of nearly 35m provided a tunnel alignment entirely within the Lewis Chalk beds with possible flint contents in the face of 10%-20%. To bore the 3m i.d. tunnel and erect the segmental lining Amec refurbished its 3.35m diameter Lovat EPBM, previously used on the Fylde project in Lancashire 10 years ago. The shield was upsized to 3.5m and the cutterhead dressed with rippers and scrapers. Cutter discs weren’t deemed necessary as the rippers were thought capable of dealing with any flint nodules.
Tunnel boring began on the 28 July from the Littlebrook shaft, initially using the TBM’s jacking rams to push the machine off a thrust block, built against the shaft back wall, forcing it through a Bullflex seal, soft eye and out into the chalk.
To allow the 65m long TBM to be installed in sections, temporary precast tunnel half segments were erected within the shaft to act as an extension to the thrust block until sufficient space was available to lower and install each subsequent section of the machine.
Once fully underway the TBM made good progress. Although heavily fissured with water present at up to 3.5 bar, the chalk proved a good tunnelling medium that, as anticipated, required no foam or polymers during excavation or spoil removal. The chalk also provided a stable, self supporting face, not requiring full EPB mode, with the pressure dissipated along the TBM’s 9m long, 800mm diameter screw.
Walls explained, “we didn’t consider using any additives, bearing in mind that we were shaving off about 25mm-30mm per rotation. With the water that’s already in there, the chalk goes into a paste, a bit like toothpaste. Spoil in that condition tends to self generate into a plug, and that proved to be the case. It flowed out of the back of the machine beautifully.”
As the machine progressed, each CV Buchan designed and supplied lining ring was transported to the back of the TBM via rail guided skips. The six 1.2m long trapezoidal tapered segments were erected in the machine tailskin using the Lovat’s hydraulic full circle erector, with the key plate in the crown. Grouting was carried out concurrent with excavation. To ensure a watertight tunnel, the segments were supplied with pre-fitted VIP Heinke EPDM gaskets, which have proved extremely successful.
In the third week of August 2003, having bored 630m, and just before descending under the River Thames, Amec was required to inspect the cutterhead, as specified in the contract documents. With no compressed air available, it was vital to ensure the stoppage area was secure from high-pressure water inflows.
Walls talked T&TI through the process, “You have a slight problem in that you can’t turn the water off in the ground! If the (water) flows were too high in the surrounding chalk, you would have to drive further until you reach less fissured ground. We had ports in the tailskin to inject polymers if we’d needed to kill any flows, but in the event the flow was acceptable where we were, so we pulled the screw back and accessed the head.”
The cutterhead proved to be in good condition. The crew replaced one or two tools, although it wasn’t thought totally necessary, it was just considered convenient to take the opportunity while the head was accessible. Boring resumed and the journey under the river proved a smooth one, with rates of up to 28m per shift achieved. It was only in mid December after the TBM had travelled 1.7km, with 700m left to go, that things started to go awry.
Down tools
“We were across the river and our production started to drop off,” explained Walls. “This is usually a sign that tools are wearing, but usually the torque also reduces. Our torque was high, which told us we had good tools, and we could demonstrate that we were cutting the necessary annulus.”
Walls continued, “we thought it was harder chalk or more flints. We were trying to monitor the spoil coming through but the flints were being smashed down to small pieces, and when you wrap them in chalk they’re very hard to see.”
The crew lived with it for a few shifts, believing it to be hard ground, but soon enough it became evident that the head would have to be entered and inspected.
Strangely, on the night shift before the entry was planned, production shot up to 18 rings – 21.6m, almost double what it had been. Now sure that it had been a section of hard ground and believing the inspection unnecessary, the team was disappointed to see production drop to 6-7 rings on the next day shift.
Left with no choice, a safe area was secured, the TBM stopped and the head entered.
“We found that two tools had been lost about 300mm in from the periphery allowing wear on the head and into the actual body of the machine. We weren’t getting a total excavation overlap, one tool missing would have been ok, but not two. The tool loss was in a critical spot, just where the picks commenced angling outwards around the edge of the cutterhead,” said Walls.
In a foot width the forward-pointing tools were lost. The first remaining pick left to address the chalk wasn’t presenting a perfect tungsten carbide face. This not only led to wear into the cutterhead but also into the sides of the outer picks. Some were actually weakened to the point where they were pointing backwards.
Being a week and a half before Christmas, Amec decided to use the upcoming holiday to fix the machine, limiting delays to the schedule.
To begin remedial operations, the team generated 0.15m of space between the face and the cutterhead by driving forward on the steering rams and then pulling back. A working pocket was then excavated in front of the machine to allow immediate repairs to begin.
“We went in through the ports and hand excavated a space big enough for a man to stand and work comfortably in, a couple of meters square, and half a meter deep. The pocket’s roof was pinned from the top of the machine for support, while the muck was thrown back through the head,” said Walls.
The repairs involved re-building the lost toolboxes and replacing the tools, and welding some additional rippers in the affected area.
As the rest of the head was “in perfect condition”, only the 300mm-400mm damaged width of the whole cutterhead needed trymay plates welded into it to bring it flush with the rest of the head.
The welders worked, from inside the pocket, on a 1m -1.5m damaged section each time. Once each section was complete the head was turned and the next length welded until the complete 360º had been fixed. During the repairs, on each shift two welders worked on rotation and were protected from fumes by a comprehensive temporary ventilation system. The machine was ready for work by the 5 January.
The final stretch
Now fully back up to speed on the final 700m the TBM hit some good regular advance rates of 30m + per day. However, an unfortunate byproduct of the combination of high advance rates and a particularly wet winter was the spoil disposal logistics. The land reclamation site in Tilbury, used for the spoil disposal, was long and spread out as opposed to deep, which is preferable for dumping a material that is initially semi-liquefied.
The spoil wasn’t drying out properly due to the high winter rainfall. Although it was a clean inert material, its moisture content was remaining extremely high. Due to available space the lorries were getting bogged down in the tip, and a limit was put on the spoil tipping. This not only caused huge frustration in that tunnelling was going so well, but actually cost Amec five shifts.
Even so, the team completed the final 700m in just under a month, breaking through into the Thurrock reception shaft on the 4 February. The tunnelling operations have come in just over three weeks late, an impressive feat when considering both the cutterhead incident and the problems with spoil logistics.
“Presented with a challenge like the Dartford Cable Tunnel, it is essential to be able to draw from past experience. The benefit of knowledge we’ve gained from the continual use of our Lovats throughout the UK has contributed immensely to the project’s success,” said Amec Tunnelling construction director Peter South. “Dealing with chalk under continual water pressures of 3.5 bar and still achieving good outputs from a 40m deep pit, dealing with face inspections without the need for compressed air and effecting repairs to a damaged head over the Christmas shutdown period, with little effect on programme speaks volumes for the dedication of the team involved,” he said.
The Dartford Cable Tunnel project is now well on schedule for becoming operational in early 2005.
Related Files
Plan map of the Dartford Cable Tunnel alignment under the River Thames
