At a first glance there is nothing particularly different or exciting about the enlargement of a 4.4m i.d. pilot tunnel to a 7m i.d. station tunnel. Nothing, that is, until you consider that the muck is being removed through the advancing face of the drive, that the rings being dismantled are predominantly of the expanded lining type and that tunnelling was carried out in the footprint of one of the most famous structures in the world – St Stephen’s Tower (Big Ben) in the centre of London.

This paper examines the engineering and contractual requirements that led to this situation during construction of the Westminster Station platform tunnels on Contract 102 (C102) of the Jubilee Line Extension (JLE). Having outlined the main constraints, the paper will then describe the way in which the different challenges were approached and overcome. The paper concentrates on the eastbound drive on which the author worked as a shift engineer.

Ground treatment and settlement control

One of the biggest engineering challenges on Contract102 was tunnelling under highly sensitive structures and high profile buildings. These included Big Ben, the RAC club, the main line viaducts passing through Waterloo to Charring Cross, key sewers and high-pressure water lines. Added to this was the necessity of working in one of the world’s busiest cities.

A JV of AMEC and Geocisa (AG) was subcontracted to prepare programmes; proposals; designs; grout mixes; injections; and installation of the injection arrays. These were intended to counteract the anticipated settlements, calculated from BBA’s assessment of volume loss settlements, which could result from the tunnelling operations.

Plastic tubes à manchette were installed about 6m below the surface to carry out permeation grouting. This grouting would provide a safeguard for tunnelling operations from the upper water bearing gravels, the clay cover at times being less than 6m thick. It would also provide a homogeneous layer between the tunnel and the surface, thus facilitating the compensation grouting and limiting the potentially damaging differential settlement.

For the high pressure compensation grouting, AG drilled over 38km of steel tubes `a manchette at a depth of 16m in the Waterloo and Westminster areas. The compensation grouting was split into three types: pre-treatment; concurrent (the type mainly referred to in this report); and observational grouting, a long-term measure carried out following completion of the tunnels to restore levels and correct deformations.

In the case of Big Ben, it was tilt rather than vertical settlement that was of main concern, as small differential movements across the base become greatly exaggerated at the top of the tower, which is approximately 60m up. When the drive was finished, the differential settlement recorded across Big Ben’s foundation was 5mm, approximately equal to that predicted. Predictions for vertical settlement were exceeded by only 3mm. At this time, excavation was also being carried out in the Westminster Station box.

Permeation grouting was originally envisaged as a means of strengthening the ground and allowing safe construction of the eastbound station. A pipe arch roof support system was proposed by BBA as a cost-effective alternative in what was an extremely difficult location to carry out permeation grouting. The pipe arch roof support system, although previously used in other countries, had not been used in the UK.

The arch, while obviating the need for permeation grouting, was also intended to protect the drive beneath it from excessive deformation caused by the compensation grouting that would still be required. The grouting was performed to prevent settlement from volume loss as the drive progressed.

The layout of the tunnels was also designed to minimise settlement. This included crossing the eastbound tunnel above the westbound tunnel, adjacent to the Westminster Station box and Big Ben, and positioning the 9m station shield erection chamber at the west end of the drive, the end farthest from Big Ben.

Environmental Considerations

Contract 102 was located in the heart of London. The contract requirements reflected the potential impact a project of this scale could have on the local environment. One of the requirements was for muck from the running tunnel and station drives to be removed by river, not by road. The jetty was therefore located at the Jubilee Gardens site (the platform opposite to the erection chamber), and spoil had to be removed through the front of the shield towards Jubilee Gardens. Restrictions on truck movements from the Storey’s Gate compound also detracted from its use for spoil removal. The total volume of muck removed from the eastbound station drive (unbulked) was around 7200m³. Approximately half this had been removed earlier during construction of the running tunnel.

Running tunnel construction

The east- and westbound running tunnels were constructed in 4.4m i.d. bolted precast concrete (PCC) and 4.5m i.d. expanded lining. The expanded lining comprised ten 1m long segments and two wedge shaped keys, one at each knee. There were two sizes of key, one 60mm wider than the other. Different combinations could be used, depending on the ground conditions encountered. The ring, with no bolting or grouting required, was formed when the keys were shoved home (using up to 300 bar), expanding the ring tightly against the ground. During construction of the running tunnel, plates were initially held in place using build bars, which were only withdrawn once the ring had been partially expanded.

Progress on the running tunnel when using this type of lining could be over 24m/12h shift compared to the 6m or so that could be achieved using the 600mm long, 4.4m i.d. bolted lining. The latter was therefore only used at the start and end of the drives, where break-ups were required. In these locations, the structural integrity of the ring was paramount and the bolted lining provided a greater stability when segments were removed.

The Westminster platform drives

The Westminster Station drives required the enlargement of a section of the 4.5m i.d. running tunnel, which acted as a pilot tunnel, to a 7m i.d. spheroidal graphite iron (SGI) lining. The drives were approximately 165m long and horizontal, having a 1000m radius. The eastbound drive was the second to be carried out. The ground encountered was London Clay with occasional claystone. Unforeseen ground conditions were not anticipated as the running tunnels had already been built.

The pipe arch also acted as an exploratory borehole above the platform tunnels, ensuring that there were no unknown stream channels at the clay/gravel interface which might extend down to the tunnel crown. However, the drive was still open-faced and there was no room for complacency.

The platform drives were connected to the Westminster Station box by a series of hand mined adits.

The shield and construction cycle

To safeguard the contract programme, two running tunnel machines were commissioned from Wirth Howden (now Wirth UK) in Glasgow to avoid any delay in the construction of the running tunnels. In the event, the two drives were carried out sequentially. As the first drive reached completion in May 1995, the backhoe unit, forepole boxes and breasting plates were stripped from the machine. Following some alterations involving an extension to the backhoe arm and adjustments to the forepoles, these were fitted to a larger shell. As well as having a 360° rotary backhoe, five breasting plates and telescopic forepoles, the 7.5m o.d. shield had three lower face support rams and a ‘birdwing’ erector for erection of the 7m i.d. SGI rings.

Notable aspects of the machine and drive were the dismantler shield located in the pilot tunnel ahead of the main shield, the arrangement for muck removal through the face of the drive and the method adopted for the shield erection.

Shield erection

The ideal arrangement for a shield erection chamber is that the working space should be at the top, enabling lifting operations to be carried out with the greatest space available. Although this was possible for the lower westbound drive, the locations of the pipe arch roof and the high-pressure water main meant that it could not be achieved on the eastbound drive. The requirement for working space above the shield during its erection was not an option, so an alternative solution had to be found.

The 9m diameter, hand mined erection chamber was constructed as close to the gravels as possible. The shield was then fabricated in the bottom of the chamber. Sections of the shield were brought in from Jubilee Gardens, where they had been stored when the westbound drive had finished. On completion, the shield was jacked up nearly 1m.

The shield, backhoe and erector weighed 85 tonnes, with a further 80 tonnes of back-up sledges and equipment. The jacking operation was carried out using four 50 tonne jacks. These were set up independently to take account of the uneven distribution of weight through the shield. However, jacking was carried out simultaneously in order to keep the shield level. Guides with 5mm clearance were attached to the front and sides of the erection chamber.

Once in position and the relative line and level checks having been carried out, concrete was poured into the invert and the trailing back-up constructed in the pilot tunnel.

Dismantling the pilot tunnel

The drive required the dismantling of a predominantly expanded 4.5m i.d. concrete lining. The integrity of an expanded lining relies on the hoop stresses that have developed. Once the key is removed, there is effectively a 4.9m diameter hole with no ground and no lining to support the tunnel. For this reason, the dismantling process has to be approached in a disciplined manner.

Ring dismantling was carried out using the dismantler shield. In order to maintain stability in the tunnel, the pilot tunnel was strapped for a minimum length of 30m ahead of the shield. The straps were cut only when the hydraulic jacks from the dismantling unit had been locked into the segments. One of the keys was then removed using a hydraulic jack, allowing the bottom three plates to be removed, beginning with the segment adjacent to the missing key. Plates were removed using a hoist located in the leading back-up sledges.

Once the invert places had been removed, the hydraulic jacks from the crown and down one side were withdrawn, allowing these segments to slide into the invert. Throughout this phase, operatives stood on a raised platform within the dismantler shield, protected from the falling plates. Holes in the dismantler shield enabled segments to be prised free with a nail bar should they become stuck. Once the ring had been dismantled, the leading back-up sledges were pulled farther up the pilot tunnel by two air winches attached to the tunnel lining. The dismantler shield was then pulled up to meet the wagons using two jacks. The dismantling time for an expanded ring was approximately 20 minutes, for bolted rings sometimes double that time.

Excavation

Excavation was performed using a hydraulic backhoe. However, due to difficulties in accessing the left-hand side of the face, hand trimming was at times required. A moveable platform to the left of the face provided a safe working platform for a miner to access the face to carry out this trimming. The same platform was also used for the erection of face support and the weekend box-up.

The mucking operation was affected by the requirement for the erection chamber to be constructed at the west end of the platform and for the muck to be removed through Jubilee Gardens. This meant that spoil had to be removed through the face, as the shield was driven back on itself, towards Jubilee Gardens. This was achieved using an arrangement of three belts. The first took muck from the invert towards the back of the machine, where it was deposited onto a second belt. This returned through the face and onto a third belt, in the leading back up sledges. From here, the muck was deposited in skips for removal to the Jubilee Gardens pit bottom. The second belt was retractable, allowing access to the right hand side of the face when retracted and allowing muck to travel into the pilot tunnel when extended.

To minimise settlement, excavation was carried out in a specific manner, including installation of temporary face support. Positive pressure was applied to the face using the five breasting plates mentioned earlier. Excavation for one ring took 45- 60 minutes.

As the shield was shoved forwards (using a combination of 24 possible rams), the secondary belt was retracted to prevent it from reaching the pilot tunnel. The forepoles and breasting plates, along with face support rams, were switched to low pressure for the shove. Maintaining a positive pressure on the face at all times was important in minimising settlement.

Shield cycle

The ring build could be carried out simultaneously with the dismantling and mucking operations, ground treatment activities permitting. Concurrent grouting took the form of grouting in blocks ahead of and behind the shield, thus compensating for the tunnelling works in that area. Excavation was allowed to proceed once the front block had been completed; shoving of the shield was only carried out when the back block was finished. The back block in turn could only start once the ring outside the skin had been grouted and the front block when the shield had been shoved. This caused an immediate breakdown in the operating cycle of the shield. Close communication between the shift engineer and the monitoring section kept this disruption to a minimum.

Conclusion

Construction of the Westminster Station tunnel, with its various contractual and engineering constraints, provided challenges from the design through to the construction phase. It combined the relatively standard type of equipment with the more unusual. The latter included the mucking arrangement, involving muck being removed through the face of the drive; the method adopted for erection of the shield; and the dismantler shield. The last mentioned meant that an expanded lining could be used for the pilot tunnel, enabling the running tunnel to continue at the accelerated rate made possible by the use of expanded rather than bolted PCC lining. The time saving was also reflected in the dismantling of the rings. However, it was the close teamwork of the gang on the shield, combined with the supporting sections, that ensured the various features of the shield were used to their full potential.

Related Files
Figure 2: Tunnel layout
Figure 4: Arrangement of belts
Figure 1: The Jubilee Line Extension and Contract 102
Figure 3: Section
Figure 5: Compensation grouting injection timing