Professional Documents
Culture Documents
Crossrail at Canary Wharf
Crossrail at Canary Wharf
Proceedings of the Institution of Civil Engineers Civil Engineering Special Issue 165 May 2012 Issue CE5 Pages 5057 http://dx.doi.org/10.1680/cien.11.00039 Paper 1100039 Received 01/08/2011 Accepted 23/12/2011 Keywords: piles & piling/railway systems/transport planning ICE Publishing: All rights reserved
proceedings
Hoe Yeow BSc, PhD, CEng, MICE Associate director, Arup, UK Duncan Nicholson MSc, CEng, MICE Director, Arup, UK
3 4
Cliff Bryant FCIOB, MAPM Executive director, Canary Wharf Contractors Ltd, UK Martin Westbury BEng, CEng, MICE Operations director, Expanded Ltd, UK
Innovative thinking and willingness to break with the norm were two of the main ingredients to successful delivery of the first Crossrail station at Canary Wharf in London. For example, minimising the environmental impact of construction activities in an urban environment led to the use of an innovative silent piling technique, which in turn led to significant cost savings and cut a year off the programme. In addition, a flexible approach to design development allowed continuous improvements in layout to be made to generate further cost and programme savings, with additional savings arising during construction through adoption of the observational method.
Existing station New station Surface line Tunnel Tunnel portal Acton Main Line Ealing Broadway West Ealing Hanwell Southall Hayes and Harlington West Drayton Shenfield Manor Park Ilford Seven Kings Forest Gate Brentwood Maryland Harold Wood Stratford Gidea Park Whitechapel Romford Farringdon Bond Street Paddington Chadwell Heath Goodmayes
Taplow
Tottenham Court Road Liverpool Street Canary Wharf Custom House Abbey Wood Woolwich
DLR
Blackwall Basin
Heron Quays
River Thames West India South dock 0 m 500
50
Civil Engineering Special Issue Volume 165 Issue CE5 May 2012
Achieving more for less at Canary Wharf Crossrail station, London Yeow, Nicholson, Bryant and Westbury
1. Introduction
Canary Wharf Crossrail station is the first station on the Crossrail cross-London railway project to be constructed and, at the time of construction, the only one let as a design-and-build contract with developer Canary Wharf Group. It recently won the 2011 Ground Engineering award for technical excellence due to innovation in its design and construction techniques. Three years from inception, it is ahead of the development programme. Early involvement of the developer in the project allowed significant optimisation of the station box design from the original Crossrail reference design scheme. Innovative thinking, local knowledge and willingness to break with the norm enabled the station to be delivered at 58% of the cost of the reference design. This paper describes the success of the Canary Wharf Crossrail station project, outlining the early design work undertaken by the developer and its project management group Canary Wharf Contractors Ltd (the client), together with designer Arup and architect Tony Meadow Associates. Further design refinement was also undertaken once main contractor Expanded Ltd was appointed and executive architect Adamson joined the team.
which are up to 40storeys high. The east, north and west retaining walls were formed from 310 Giken tubular piles of 1214mm diameter, installed using two Giken silent pilers. Reinforced concrete (RC) piles were formed by boring through the bottom of the tubular piles and the whole pile then reinforced and concreted. The southern station retaining wall was formed by a line of 1180mm diameter hardfirm secant piles at 875mm centres from the dock bed. This saved a year on the construction programme from Crossrails original proposal to fill the dock, install retaining walls then excavate the material now contaminated with the dock silt. Along the station centreline are 105 tension piles of 1200mm diameter and 30 plunge-column piles of 2100mm diameter to facilitate the top-down construction. 2.1 Ground and groundwater conditions The ground conditions were assessed from existing exploratory boreholes undertaken for the Canary Wharf development and
Billingsgate market Canary Wharf tower
Figure 2. Aerial view of Canary Wharf Crossrail station (CWCS) site looking east shows the high urban density of the Canary Wharf development
Figure 3. Plan and longitudinal section of the 265m long station box, which includes retail areas and a roof park above the station concourse and platforms
51
Civil Engineering Special Issue Volume 165 Issue CE5 May 2012
Achieving more for less at Canary Wharf Crossrail station, London Yeow, Nicholson, Bryant and Westbury
specifically for the Crossrail station project (Arup, 2008a). These revealed that the site typically comprises a downward geological sequence of dock sediment, made ground, River Terrace Deposits, Lambeth Group clays, Upnor Formation, Thanet Sands and Chalk. To the east of the station box, both London Clay and the Harwich Formation are revealed in the boreholes above the Lambeth Group. Figure 4 shows the geological cross-section along the station box. The hydrogeology of Canary Wharf includes two distinct aquifers called the upper aquifer and lower aquifer. The upper aquifer lies within the River Terrace Deposits and Harwich Formation. The lower aquifer comprises the Lambeth Group sands (Reading and Upnor Formations), the Thanet Sand Formation and the underlying Chalk Group. They are separated by an aquiclude that is predominately formed of low-permeability clays of the Lambeth Group. 2.2 Client and stakeholder focus With the client and stakeholders (all trades, consultants and manufacturers) having a common objective and strategy, the focus for project delivery centred on risk management, transfer of each element to the party best geared to manage risk and delivery on time, within cost and to the required quality. Providing this under a fixedprice design-and-build contract gives cost and time certainty to the ultimate client Crossrail, to which the completed asset transfers. This held true when the strategic spending review was instigated due to a change of UK government. Two key events drove the whole schedule the arrival of the tunnel-boring machine in mid-2012 and Crossrails operational date of December 2017. In Crossrails reference scheme, the proposed approach to construct the station box resulted in late delivery and cost over-runs. The client initially reviewed the design brief and quickly identified that the proposed ventilation system greatly increased the box size. Reducing the box size brought large cost savings, and further value engineering of the construction methodology eventually led to the development proposal now under construction. The qualities of materials were enhanced to match Canary Wharf standards and all finishes were designed to meet the Building Research Establishment (BRE) grade A environmental green standard.
Extent of station box
100 Terrace gravels Dock silt/made ground
90
85
80
Upnor formation
(a)
75
70
Thanet sand
(b)
65 Chalk
60
Figure 4. Geological cross-section along the station box determined from existing borehole records
Figure 5. Comparison of the Crossrail reference scheme with 360m long station box (a) and developer Canary Wharf Groups scheme with a 259 m long station box incorporating retail development (b), which effectively saved 360million
52
Civil Engineering Special Issue Volume 165 Issue CE5 May 2012
Achieving more for less at Canary Wharf Crossrail station, London Yeow, Nicholson, Bryant and Westbury
building occupiers was a high priority. Construction of the cofferdam using a silent piler system was the prerequisite of such a waterretention system. At an early stage of the design development, to manage the design and construction risks, the developer commissioned a trial of the Giken piling system (Giken, 2008) to explore its use for the project (Arup, 2008b). The system utilises reaction force derived from fully installed piles to press in subsequent piles hydraulically. The piling equipment is mounted on top of the reaction piles and self-moves to the next position by gripping the pile being pressed-in. Where needed, the crush system was used to auger through the bands of rock in the Lambeth Beds. This utilises the unique simultaneous press-in and augering process of the Giken system to penetrate hard ground, which avoids the use of extensive temporary works to construct the cofferdam in the dock. A test frame was designed and constructed on site to install three 218m long 914mm diameter Giken tubes to confirm constructability in similar ground conditions, rate of production and water-tightness for bored piling within the tubular pile (Figure 6). The tubes were pushed through approximately 3m of dock silt, 5m of sand and gravel (Terrace Gravel and Harwich Formation) and 35m of stiff clay (Lambeth Beds). The trial provided valuable installation data to extrapolate and confirm the push-in force needed to install the 1214mm diameter tubes required for the cofferdam. The rate of installation of up to three tubes per day was also confirmed during the trial, with each tube taking an average of 3h to install. After removal of the spoils in the piles, the bores were left open for 7days and they showed little sign of water ingress. This confirmed the water-tightness of the tubular pile toed into the Lambeth Clay for subsequent installation of the RC piles using a dry-bore construction method. The trial information was then provided to the tendering contractors. The trial cost 250000 but the savings in time and construction cost were significant. Compared with the tender design scheme, it allowed construction without importing tonnes of fill for the doubleskinned sheet pile cofferdams, removed the North quay cut-off wall and reduced the area of dock taken to construct the station box. With this trial, the client was able to bring the contingency in line with industry norms and commit to programme certainty, which was key to the tunnelling machine arrival date of mid-2012. In addition to significantly reducing construction risks it reduced the overall programme time. Extensive construction experience in the area also led to the decision by the design team to dewater the lower aquifer to maximise piling in dry bore and to enable excavation beneath the interlocking pile cofferdam using a contiguous wall. A total of 14 wells (3 existing and 11 new ones) were installed into the underlying Chalk to reduce the lower aquifer water level by up to 20m with an average pumping rate of approximately 130l/s. Finally, a profiled base slab was also introduced to enable arching action to develop and provide a better structural form for the station box.
Figure 6. One of three 218m long, 914mm diameter Giken trial piles being pushed silently into the ground the trial cost 250000 but resulted in significant programme savings
4. Tender-stage development
Two enabling work schemes were produced for tender. The first scheme used double-skinned sheet pile cofferdams and linked with the existing cofferdam system along the south-east side of the proposed station box and a new cut-off wall along the North Quay (Figure 7). The majority of the remaining south side cut-off was provided by the existing basement structures at Canada Square, south
53
Civil Engineering Special Issue Volume 165 Issue CE5 May 2012
Achieving more for less at Canary Wharf Crossrail station, London Yeow, Nicholson, Bryant and Westbury
As part of the tender process, the contractors and designer were provided the opportunity to undertake design development to improve the designs to suit the contractors construction method
Temporary propping
Adams Place
Bank of America
HSBC
KPMG
10
20
30
40
50 m
Figure 8. Second tender design option with a propped Giken pile cofferdam
Station box
Adams Place
Bank of America
HSBC
KPMG
10
20
30
40
50 m
Figure 9. The alternative design proposed by the contractor used a mixture of Giken piles with tied-back anchor piles and a secant-pile wall to the south
54
Civil Engineering Special Issue Volume 165 Issue CE5 May 2012
Achieving more for less at Canary Wharf Crossrail station, London Yeow, Nicholson, Bryant and Westbury
of the proposed station box. This enabled the station box construction work to be carried out from the base of the drained dock in relatively open conditions. The second scheme used cofferdams made from interlocking tubular Giken piles jacked down to the Lambeth Beds clay and then extended using RC piles to form a contiguous wall, with temporary propping across the station box (Figure 8). In this case the tension piles would be installed from a temporary deck at dock water level. As part of the tender process, the contractors and designer were provided the opportunity to undertake design development to improve the designs to suit the contractors construction method. This allowed the contractors an insight of the design intent and gave the designer an understanding of the preferred construction method.
the stakeholders, client and suppliers that has stood the test of time and occasional challenges as the design and methodology developed. This allowed open dialogue and risk sharing between purchaser and supplier. Key-process pictorial planning allowed the strategy to be quickly understood and enhanced by all. Further design development was undertaken by the contractor to determine the clutch design of the Giken piles and the anchor tie to support the cofferdam wall. For the clutch, the contractor opted for the PP clutch instead of the PT clutch. Despite having a larger total cross-sectional area, the PP clutch offered three smaller grouting compartments instead of one large compartment as in the PT clutch (Figure 11). The two outer compartments were designed to be grouted up using grout socks while the middle compartment was grouted up without a grout sock. At this stage, full-scale testing
Anchor piles
Tie
Figure 10. Three-dimensional view of the main components of the retaining structures
PT interlock
PP interlock
Figure 11. PT and PP clutch arrangements offered by the Giken system the contractor opted for the latter due to a larger number of grouting compartments
55
Civil Engineering Special Issue Volume 165 Issue CE5 May 2012
Achieving more for less at Canary Wharf Crossrail station, London Yeow, Nicholson, Bryant and Westbury
of the tie was also undertaken by the contractor to confirm the load capacity of the tie. The station box and its surrounding existing structures were analysed using two-dimensional finite-element software to capture the complex soilstructure interaction behaviour. This enabled the station box to be designed to allow for the influence of the construction of the station on these structures. Due to the dock water behind the cofferdam wall on the north side, the whole structure was noted to sway to the south during construction. A 20mm sway movement was predicted but only 10mm was measured during the work.
6. Construction phase
Construction of the station was undertaken using a top-down construction sequence. The major stages were n dewatering of the lower aquifer n probing for unexploded ordnance for piles installed from barges n installation of the cofferdam steel tubular piles and RC piles on the east, north and west sides of the site n installation of the anchor piles on the east, north and part of the west sides of the cofferdam n construction of the berm of part of the west side cofferdam n prestressing the anchor tie n draining the dock n clearing dock silt and installing piling platform n probing for unexploded ordnance for piles installed from drained dock bed n construction of the secant-piled wall on the south side, the plunge-column piles and the tension piles n undertaking top-down construction by excavating and installing level 3, 4 and 6 slabs and, at the appropriate time, constructing the level 2, 1 and 0 slabs n flooding the site after completion of the station box structure.
About 100000m3 of dock water was pumped out during the dock draining work. During excavation, a total of 135000m3 of soil was excavated from the station box and about 35% of this material was reused to create ship-impact berms and a flood-storage reservoir south of the station box. During construction of the station box, a comprehensive array of monitoring instruments was used to capture construction-induced displacement of the walls and the anchor piles, movement of the capping beam and tie forces in the anchor ties. Sixteen sets of monitoring arrays comprising inclinometers, load cells and survey prisms were installed in selected wall piles, anchor piles and their ties, with an additional six sets of inclinometers and prisms installed in the bermed cantilever wall and the south secant wall. Additional prisms were also installed at selected locations on the capping beam to monitor its movement. The layout of the monitoring system is shown in Figure 12. The instruments were carefully read prior to construction activities to establish baselines. The baselines for individual instruments were read in less than a week due to constraints of site activities although a longer duration was preferred. Subsequent measurements were taken at a set monitoring frequency and at critical stages of the construction work, and these measurements were compared with the corresponding trigger values to confirm that the performance of the structure was within design values. At the same time, the excavation was carefully recorded to provide a clear sequence record of the works. This allowed the observational method (Nicholson et al., 1999) to be used, with monitoring data being reviewed and ground stiffness parameters back-analysed after excavation to the 4 slab to modify the soil parameters and construction sequence of the headwalls. This led to the removal of the lower-level berm and inclined props during excavation to the final base-slab level. It is estimated that the associated cost saving was approximately 500000 in addition to substantial programme savings.
Station box
Figure 12. The comprehensive monitoring array allowed the observational method to be used, which in turn avoided the need to install a lowerlevel berm and inclined props during excavation to final base-slab level
56
Civil Engineering Special Issue Volume 165 Issue CE5 May 2012
Achieving more for less at Canary Wharf Crossrail station, London Yeow, Nicholson, Bryant and Westbury
6.1 Approach to risk Risk to the client is seen as a manageable commodity, as demonstrated on this project; each element of the project was broken down into separate work streams to enable the team to understand the deliverables, methodology and scope to minimise risk. Where a methodology had not been previously employed at Canary Wharf, desk studies ensued, followed by site visits to other works where the technology or methodology had been employed; in extreme cases, full-scale mock-ups were implemented. Once the technology or method had been proven and the team was satisfied, the project could be undertaken to the required quality, time and cost requirements. Specialist trades, manufacturers and suppliers were engaged through tender to validate the clients way of working. Their knowledge was then applied to refine the works further. Continuing the process through the tender period, post-award and throughout the detailed design further reduced the cost and increased the contingency.
7. Conclusions
The involvement of the developer early in the design development stage of Canary Wharf Crossrail station has significantly reduced the cost of delivering the first station box for the Crossrail project from about 860million to 500million, with a 150million contribution from Canary Wharf Group. This was achieved by a team with extensive local experience in the design and construction techniques of underground structures in Canary Wharf. Early design development relied heavily on the existing cofferdam on the south side of the station site to provide the necessary water cutoff and this has proven to work well. This early works explored many design scenarios with different levels of confidence in the make-up of the south-side water retention. The project also allowed an innovative method of construction
well suited to an urban environment where environmental impacts in the form of construction noise and vibration need to be kept to a minimum for successful implementation. The developer invested significantly in a silent piling trial and adoption of this new construction technique minimised construction risk and reduced the programme by a whole year. The key to achieving the ambitious construction programme was to minimise heavy marine operations by draining the dock for the installation of the south retaining wall and the internal tension and plunge-column piles. The contractor also utilised part of the existing cofferdam as its access onto the dock after it was drained. This eliminated the need for imported fill to create an access ramp. The station has many constraints and the layout and construction sequence evolved over time. This flexible approach to design development involved continuously testing the design to see if additional layout improvements could be made, together with cost and programme savings. The interactive involvement of the client, designer and contractor was important throughout this process. During construction, establishing baseline readings for the monitoring instruments was another important activity that took significant efforts to achieve. Without confidence in these baseline readings, the programme and cost savings achieved from the change in the construction sequence for the headwalls would have been difficult to realise. The success of the project has been acknowledged by the UK government, Crossrail, the client and stakeholders. It has set standards adopted by Crossrail and has been acknowledged by peers through awards for engineering excellence, safety and the environment.
Acknowledgement
The authors would like to thank all who worked on the project for their efforts in the delivery of this first station for the Crossrail project.
The success of the project has been acknowledged by the UK government, Crossrail, the client and stakeholders. It has set standards adopted by Crossrail and has been acknowledged by peers through awards for engineering excellence, safety and the environment.
References
Arup (2008a) Geotechnical Interpretative Report. Arup, London, UK. Arup (2008b) Giken Piling Trial Report. Arup, London, UK. Giken (2008) Construction Revolution Silent Piling Technology. Giken Seisakusho Co. Ltd, Kochi, Japan. Nicholson D, Tse CM and Penny C (1999) The Observational Method in Ground Engineering Principles and Applications. CIRIA, London, UK, Report 185.
57