Bank station capacity upgrade (BSCU) is a congestion relief and interchange improvement project for London Underground (LU) in the City of London. The project is currently entering the construction phase, working to relieve the issues highlighted in Figure 1.

The infrastructure improvement proposed as part of the Transport and Works Act Order (TWAO) submission includes: ¦ Improved below ground interchange capacity including a new southbound running tunnel and platform for the Northern line,

¦ New station entrance on Cannon Street,

¦ Step-Free access from platform to street for the Northern line and DLR,

¦ Moving walkway for improved journey times to the Central line,

¦ Additional lifts and escalators to ease congestion and improve accessibility,

¦ Improved fire and evacuation routes.

Bank Bloomberg Place (BBP) is the sister project to BSCU and is also currently in construction. As part of the planning permission for the new 3.2 Acre Bloomberg European headquarters (Figure 3), provision was made for a station box to the Waterloo and City line within the developer’s basement. The structure was constructed on behalf of Bloomberg by Robert McAlpine, to be leased by TfL.

The box fit-out and the connection to LU tunnels was then procured through a design and build contract with Hochtief UK. The new entrance provides step free access to the Waterloo and City line as well as reducing congestion and improving emergency evacuation times for the wider station.

One of the many challenges for both projects is the potential impact on a number of third party stakeholders throughout design and construction. This is particularly apt for an area in which approximately 80 per cent of buildings are within a conservation area, with historic national landmarks and international financial centres in close proximity. In defining the alignments for the new tunnels a key consideration was whom and what the projects would impact, and how LU would either avoid, mitigate or manage these interfaces. Building foundations in particular would require attention. Alignment

The need for additional underground space at Bank was clear, however delivering this space while minimising/avoiding impacts with respect to ground movement, underground obstructions and noise and vibration in this densely-populated area would require detailed planning of the alignment.

An alignment envelope was identified and an initial desktop study conducted to provide information on building locations and their foundations.

Ultimately the project agreed that trying to accommodate or avoid underground obstructions would be counter intuitive. Avoiding foundations would result in a longer tunnel and therefore additional excavation and ground movement. As a result the alignment that was straightest and smoothest (and hence held the greatest benefit with respect to line speeds) was chosen. The ground movement, noise and vibration and underground obstruction risks would be addressed by the tunnel design.

The Bank Bloomberg project did not identify any potential underground obstructions during the alignment definition and therefore could proceed with low residual risk of potential, unknown foundations – particularly considering its proximity to the existing LU tunnels (which had no record of underground obstructions being encountered).

FEASIBILITY

Desktop study

Once the alignment was defined, the potential impacts on each building along the route could be estimated. A desktop study was conducted to look into the foundation structures for each building above the alignment of the tunnels. Detailed research of public archive records, as well as discussion with building owners yielded greater detail in the foundations that may be intercepted across the route.

Where as-built pile depths could not be found, back calculation and engineering judgement was used to define conservative estimates on potential depths, types and the materials used. The foundations could then be modelled within the project workspace to better quantify the risk of interception and potential impact.

A single set of ‘pile general arrangement’ drawings was then created. This would collate all sources of research and information into a single location, and would enable the future construction team to quickly identify any potential interceptions.

Any assumptions made at this point would be recorded within the project assumptions register, for resolution or confirmation and consideration within other aspects of the project design.

Scope definition

A physical clash between the tunnels and foundations would instantly create an interface with the building. These could be dealt with at the tunnel face or potentially through works at surface level (such as underpinning).

However additional interfaces could still arise through noise and vibration impacts in the permanent state caused by train movements in the tunnel. The submitted TWAO had committed to minimise noise and vibration. To achieve this with a standard track-form required a 1.5m clear separation from any pile. Achieving 1.5m separation within the tunnelling works would be extremely challenging, particularly with the residual risk of unknown piled foundations along the route. As such, noise and vibration would be best addressed through the use of a high performance track form, reducing impact at source.

Any additional ground movement impacts through interception would also need to be considered in building damage assessments; however this could not be progressed until potential solutions were drafted. The interception design would therefore progress while aiming to minimise excavation volumes as far as reasonably practical and maintaining a dialogue with the engineers responsible for these assessments.

From the work done to date and review of the model, the pile interceptions could be divided into four scenarios:

¦ A redundant pile is encountered; previously separated from the building above (e.g., temporary works),

¦ A redundant pile is encountered; not separated from the building above (e.g., historic piles not broken out),

¦ Live pile encountered; where interception would reduce ultimate capacity but maintain design load through a reduced factor of safety,

¦ Live pile encountered; where interception would reduce ultimate capacity substantially, meaning load would need to be shared to neighbouring structures or through a newly created transfer structure.

The risk profile for this work could now be quantified based on three factors, including the impacts above.

Likelihood of interception was assessed against a 500mm and 150mm offset from the tunnel springline. This would account for potential plan tolerance or out of verticality. The confidence in the data would be assessed against its source, or whether it agreed with previous modelling data at tender stage.

The author presented this data to engineering management for Dragados and LU. For each intercepted building the author proposed an overall level of risk then suggested where additional surveys could be conducted, or more information could be requested from building owners to mitigate these risks. Where a designed solution looked the most likely scenario further information would be gathered on the capacities and design loads of live piles.

Concurrently, the planning and consents team were negotiating with building owners to facilitate the TWAO application in preparation for a public inquiry legal agreements were reviewed by all parties within the project to ensure commitments made did not preclude any initial solutions considered. LU’s view was to achieve separation wherever reasonably practical. This would not only simplify the design; but limit any potential ongoing relationship with third parties with respect to noise, vibration and future support, construction, demolition and maintenance – something third parties would likely be keen to avoid.

CONCEPT DESIGN

Load transfer interceptions

With factors of safety for building foundations usually between 2.5 and 3.5, and often found to be higher with the more detailed analysis and information on soil conditions that can be conducted and found today, the transfer option can often be discounted. This assumes however, that the building owner is willing to accept a limit on future, more aggressive re-utilisation of their piles. In cases where future construction is planned or desired, or where factors of safety would be reduced to unacceptable factors of safety, transfer options should be explored.

Initially, back analysis was carried out to validate the ability of the foundation system to redistribute load. If a pile is cut or undermined, the load within the pile may temporarily move towards surrounding piles; and the pile cap and slab must have sufficient capacity to allow for these changes. While the piles themselves may have a great deal of spare capacity, this redistribution could lead to settlement and potential cracking of the structural slab. This may in turn lead to intolerable serviceability concerns and hence a solution must be found without any temporary loss of support.

Similarly, if neighbouring piles do not have sufficient capacity to carry the load lost by the interception; the settlement effects may be too large for the building once again limiting any temporary loss of support.

A ‘cut then support’ option (Figure 7) would assume any load redistribution or settlement affects from the interception were acceptable; allowing the pile capacity to be temporarily reduced and eventually replaced by a transfer structure within the tunnel excavation, restoring capacity. A compressible fill material could be used to ensure any future load change in the pile (and settlement or heave) would not impact on the LU tunnels.

The initial tunnel excavation could proceed similarly to a pilot tunnel, through localised breakout of the foundation. An enlargement can then be created to mechanically fix to the pile and provide an end bearing section of equal capacity to the intercepted pile length while incorporating the ‘physical separation layer’ to allow for future movement. This would also allow the initial tunnelling works to proceed quickly and hence minimise ground movement.

For a ‘support then cut’ solution (Figure 8), large initial excavations would be required to enable installation of temporary pile supports. Unlike the ‘cut then support’ option the remaining works could not proceed until the temporary support structure is created. Additionally, the large excavated volume could potentially introduce buckling effects onto the pile, leading to cracking and hence serviceability issues.

Both options while structurally feasible would require detailed analysis with respect to ground movement, constructability, the building’s structure/ damage impact and presentation to building owners.

While these options were tabled for discussion and further work was completed, the author’s primary focus switched to non-transfer design options.

Pile separation

A single, simple solution to the remaining pile interception scenarios was clear. If remaining capacity was not required (redundant piles) or a reduced factor of safety could be tolerated, the end state of our solution would be to cut and separate the tunnel from the pile to ensure no load is transferred. While there may be no designed connection to the piled foundation, load may be transferred without a gap or movement joint between the two structures.

Demolition of the above building may result in heave at the pile’s toe, and further development of the building may result in settlement. A compressible fill material between the tunnel and pile structure could be used for this purpose, enabling physical separation (Figure 9).

The next steps would be to progress the detailed excavation and support sequence, review the constructability and determine the extent of potential movement (and hence separation requirement) for the final design of live separation piles

DETAILED DESIGN (PILE SEPARATION)

Pile Movement

To determine the physical separation and then the additional local breakout required about each pile the degree of future settlement and heave that may occur must be calculated. The capacity of a typical pile was calculated initially, both before and after interception.

This would validate any reduction in the factor of safety and confirm if they are acceptable.

The pile’s potential settlement and heave could then be calculated by assuming the pile is first fully unloaded (demolition) and subsequently loaded to its new (now reduced) ultimate capacity mimicking the worst case load, and hence movement, scenario.

Any potential heave could be estimated using empirical ground movement calculations; however it is highly unlikely that the building would heave due to relaxation to the same magnitude that it may settle from loading. Therefore any settlement that can be accommodated by the separation material could similarly accommodate heave. Building damage

Settlement of the building and its foundations during the excavation can initially be calculated empirically, with further detailed analysis conducted if deemed required (following LU guidance for staged assessment). Previous experience from the Jubilee Line Extension and Channel Tunnel Rail Link has found that piles largely settle with the ground and horizontal movements can conservatively be estimated to match Greenfield soil movements. Much of this movement would occur ahead of the tunnel face.

One important factor to consider with this movement was the potential for buckling, particularly if the piles were (when considered as a concrete, unreinforced column) classed as slender. In these instances a check on buckling capacity was completed to ensure any lateral movement of the pile would not lead to cracking and therefore serviceability issues. Any instances where buckling would lead to damage would have to mitigate ground movement further, potentially through reduction of the excavation section size.

Material requirements

Upon defining a worst case settlement figure and initial discussions with the construction team and LU; several requirements for the separation material were defined:

¦ Durability; the material should last for the design life of the new tunnels. Any degradation due to the surrounding clay could lead to a void behind the lining and a potential weakness to water ingress.

¦ Ability to spray against (SCL); the material would ideally need to robust enough to withstand the sprayed concrete lining being applied; to ensure as far as practically possible that the excavation sequence can continue as normal without the need for additional plant or material

¦ Compressibility/elasticity; to act elastically within the loads required while maintaining a minimal thickness to prevent the need for overexcavation

¦ Ease of installation; with many of the interceptions occurring at the tunnel crown, the weight and handling of the material would be important to ensure effective and safe installation Several options were found and an initial material was chosen based on research into joint fillers used in reinforced concrete construction. Based on the settlement requirements and the material properties a 150mm thick Flexcell compressible fill material was selected for confirmation by the contractor pre-construction.

Excavation and support (SCL)

Development

Any design must ensure safety and efficiency in its development. Using the 3D model and input from the construction team, some of the key CDM risks could be further identified and eliminated or mitigated. As mentioned above, working from height and the working area would be a key risk to the installation of any separation.

Furthermore, any installation would require supported ground above any working operatives. The detail itself was developed with these risks in mind. Finally, several areas were heavily congested with foundations (see Figure 12). Ensuring the construction team could work around and cut/breakout these piles to progress the excavation, while ensuring plant, material and personnel are able to safely move underneath would be an important consideration for the design.

Taking these requirements into account the author created a detail that would locally enlarge to allow for future separation, while supporting the ground in the immediate vicinity of the pile to ensure the safety of personnel underneath the crown. All breakout would be completed mechanically, eliminating any risk of operatives working under unsupported ground.

Step one creates the local enlargement as part of the primary lining. This envelops the pile with an initial lining sufficient for temporary ground support. This allows the pile to be broken back partially to continue the excavation safely. The supporting layer would fail in shear if any long term movement is experienced, maintaining physical separation. Step two involves breaking the back and fixing the separation layer to the pile directly or to the surrounding temporary support. The separation layer is slightly oversized to allow for any potential lateral movement of the pile. Step three then installs post drilled fixings to allow for continuity in the lining and additional support for the fresh sprayed concrete. The remaining primary lining is then sprayed to standard profile.

The excavation and support sequence offered flexibility in both breakout and safe stop. By allowing the excavation to proceed past the piles, the excavation can progress to an extent; however a clear limit is placed on the advance length beyond any temporarily supported ground/piles. This ensures full ground support is sprayed before the excavation proceeds mitigating risk of further ground movement resulting in potential failure of the temporary support.

The design could now be endorsed by the construction team. Further lessons learned would capture best practise and issues faced from the industry to ensure the design was as robust as possible. The work that the Bank team had done to date was also shared with other areas of LU, including Crossrail 2.

CONSTRUCTION

Through colleagues in the tunnelling team some feedback was obtained from Crossrail C300/C410 (BFK) on pile removal. Their approach largely agreed with design; however there was a slightly different approach to their application of final primary lining around the physical separation.

By introducing steel mesh around the physical separation membrane, the sprayed concrete would be better supported in its temporary state; remaining self supporting while the concrete gains strength. BSCU has not yet encountered piles but BBP did unexpectedly intercept two unknown piles. Using the knowledge obtained at BSCU, the author was able to assist in the design and construction of the interception. The tunnelling methodology here was squareworks but the general principles of the design were similar.

Ultimately both instances of construction experience have been a useful exercise in validating the design with first hand experience and they were successful in their implementation. It will be useful to see how future interceptions feedback on our design.

CONCLUSION

Tunnelling in London has been a regular occurrence for over a century. Previous tunnelling projects have perhaps avoided these obstructions; but this is becoming increasingly difficult. The current and future demand for capacity coupled with the more densely congested nature of the underground (as well as above ground) space means that obstructions and impacts such as these will occur more and more often.

Additionally, LU’s business plan hinges on moving customers quickly and efficiently through its stations and tunnels. With Crossrail 2 in the planning stages, and several station capacity upgrades entering feasibility or concept through the Future Stations Programme, the demand is evident in the near future. TfL’s responsibility to its customers is to deliver an efficient and effective design with a sound business case for these projects.

Accomplishing these goals while ensuring third parties and stakeholders are not adversely impacted by the work will require new solutions. It is the responsibility of projects such as this one to pioneer the best engineering-driven and cost-effective design and construction. Innovative approaches such as interception design overcome these modern day challenges and our industry must continue to promote these with the view to providing reference material and knowledge for future projects.

The pile interceptions work at BSCU and BBP is something in which the author is truly proud to be involved. It is not only the engineering achievements tabled here that will stand future tunnelling in good stead, but the approach taken to overcoming these challenges. Collaborative meetings of engineers and construction managers from consultants, contractors and the client, yielded solutions that were innovative and forward thinking. Building more than 1,200m of tunnel in the City of London is no easy task but the team at Bank have consistently risen to the challenge the construction phase will surely follow suit.