Mott MacDonald was appointed in 2009 as the framework designer, for all the sprayed concrete lining (SCL) works for Crossrail, for all the tunnels, caverns and shafts across the project.

The SCL works within the scope specifically included:

a) All of the station tunnels (platform tunnels, passenger and non-passenger tunnels, lower concourses, escalator inclines, passages, openings and the linings for certain shafts) for Bond Street, Tottenham Court Road, Farringdon, Liverpool Street and Whitechapel Stations;

b) Cross passages between the running tunnels

c) A number of the connections to adjoining structures, SCL works around bored tunnels and support to LU links at Paddington, pedestrian link to LU at Bond Street and SCL works around bored tunnel soft eyes

d) SCL works associated with a number of shafts for construction between the stations; Fisher Street, Stepney Green, Mile End, Eleanor Street and Limmo Peninsula.

e) Crossover Caverns; the biggest ones to be constructed in London, in Fisher Street, Stepney Green and Whitechapel.

f) Lesser structures constructed in SCL, e.g., drainage sumps

g) Any temporary access works required during construction

Mott MacDonald developed the employer’s RIBA C+ design progressively through a Gate process to RIBA F1+ F2 to achieve detailed design for construction by providing all the drawings, commentary, construction schedules, constructability reviews, capital costs estimating, interfacing with other contracts in the project and third parties, support in the procurement of the works, Cat 3 checking and assurance required for any approvals under the Crossrail Act and any planning consents.

The design took into account both temporary and permanent conditions for the SCL systems.

In 2011, a further supplementary agreement was signed for Mott MacDonald to continue the scope into Stage G-H (Grip 5-6); Tender Assessment for contractor’s submissions, RIBA Stage J-K (Grip 6) Construction and RIBA Stage L (Grip 6) Construction Completion, which is currently underway.

Design Responsibility

Even though the original scope for Contract C121 was to take into account both temporary and permanent conditions for the SCL systems, this developed differently on the various construction contracts, but with Mott MacDonald retaining responsibility for the permanent SCL works for all the construction contracts awarded for the project.

Due to the fact that the primary lining formed part of the permanent ground support but the contractor had to rely on its performance to safely construct the works, and therefore had to assure themselves, some contractors decided to become responsible for the temporary SCL. The responsibility was then passed to Mott MacDonald either when the ring of the primary lining had been closed, concrete had achieved its 28-day strength or when the secondary lining was complete, depending on the contract. In these instances, Mott MacDonald carried out full independent check of the contractor’s temporary works to ensure they complied with the requirements for the permanent conditions.

For other contracts, Mott MacDonald retained the responsibility for both temporary and permanent SCL works and the contractor only carried out due diligence review of Mott MacDonald design without taking over any design responsibility.

All the permanent sprayed concrete linings were Cat 3 checked by Aecom.

Lining Configuration

The SCL linings comprise a primary lining, a waterproofing membrane and a secondary lining (see Figure 1 and Figure 2). Both primary and secondary lining are considered to be part of the permanent structure and therefore share the load in the long term.

The primary lining, which ranges in thickness between 200 and 350mm, includes a 75mm initial layer, which for design purposes is conservatively assumed to degrade in the long term and therefore not accounted for in the long term calculations. The primary lining incorporates steel fibre reinforcement to provide some tensile capacity and to enhance the ductility of the shotcrete.

At the junctions, the primary lining is further strengthened by a thickening layer, where any reinforcement required for the distribution of the long term stresses is installed. The presence of the thickening layer allows safe installation of the reinforcement under mature shotcrete.

Once the primary lining has been completed for a section of tunnel, a regulating layer is then applied to the primary lining to provide a smooth surface prior to the application of the waterproofing. This layer prevents steel fibres which are present in the primary lining from damaging the waterproofing.

Sheet waterproofing was specified in those areas of the project where water was anticipated and this included Farringdon station, Mile End and Eleanor Street; all excavated in the Lambeth Group. The sheet membrane was placed over a protective geotextile and compartmentalised at every construction joint and provided with waterbars and re-injectable tubes.

Sprayed waterproofing was specified in all the other SCL structures, where excavation was anticipated to be fully in the London Clay. The waterproofing was sprayed in two layers, to guarantee complete coverage, with a total thickness of 3mm but increased to 6mm in the junction areas and also at the connections with the shafts to ensure enough gap bridging capability if any differential movement between structures was to take place (Figure 3).

Due to either water inflow or dampness experienced in some of the structures where the excavation was close to the London Clay/Lambeth Group interface, the sprayed waterproofing suffered issues related to curing and blistering. Therefore in many of the deeper structures; Whitechapel, Liverpool Street and Stepney Green, the sprayed waterproofing was substituted in the inverts by a sheet membrane. A connection between the sprayed membrane and the sheet membrane was developed to avoid any potential water paths. Once the waterproofing is in place an in-situ invert is cast to a flat surface. This provides a suitable safe platform from which to complete the rest of the tunnel lining.

The cast in situ inverts for the tunnels were generally only steel fibre reinforced with the exception of the big crossover caverns and some of the egg shaped tunnels which due to their size or unusual shape generate bigger bending moments, requiring the inclusion of bar steel reinforcement.

The remaining secondary linings were either cast or sprayed. Where sheet membrane had been specified (Farringdon, Mile End, Eleanor Street), the secondary lining arch was built in situ with the aid of shutters using generally steel fibre reinforced concrete which also included 2kg/m3 of polypropylene fibres for fire protection purposes. Steel bar reinforcement was also required in the larger caverns, junctions and most unconventional tunnel shapes to accommodate long term stresses.

In areas where the sprayed waterproofing had been provided, a sprayed secondary lining was specified. This sprayed secondary lining was generally also steel fibre reinforced only, sprayed in two layers (S1 + S2) with a combined thickness ranging between 250 and 300mm. The secondary lining was completed by the application of a 50mm thick sprayed concrete layer containing polypropylene fibres for fire resistance purposes (fire proofing layer. FP).

Due to difficulties controlling the accurate thickness, profiling and bonding of this layer on some contracts, the polypropylene layer was increased in thickness, reducing in this way the total number of layers to just S1 and S2 and substituting the steel fibre reinforcement in the S2 layer by conventional reinforcement.

The setting out of the tunnels was carried out considering the described lining configuration to the theoretical design lines and design thickness.

Construction, excavation and workmanship tolerances and deformation allowances were considered by the Contractors in the final construction of the tunnels.

Design Methodology

The lining system proposed is a double shell with both linings considered part of the permanent load bearing structure throughout the design life of the tunnel.

The primary lining was designed to resist the external forces due to all short term ground loads applied taking into account the specific proposed construction sequence for the tunnel and any adjacent tunnel which is built simultaneously. The design also considered the effects of other transient loads such as compensation grouting and any surcharge loads applied at surface level during the construction works. It was also designed to resist a certain percentage of the long term ground loading apart from groundwater loads which the secondary lining was designed to resist.

The primary lining was designed with the aid of two-dimensional (2-D) and three-dimensional (3-D) numerical modelling techniques using the finite difference software package FLAC or FLAC3D, which is able to represent the strain-dependent stiffness of the ground and provides a reliable representation of the ground-structure interaction.

Two-dimensional numerical models of selected critical sections were used to confirm the stability of the excavation and to determine the tunnel lining internal forces. Examples of the 2D and 3D models carried out are illustrated in Figure 4 and Figure 5.

In general three stages for design are considered as follows; short term, which represents the construction stage under undrained conditions, intermediate stage which represents the constructed primary lining after two years of construction and prior to the installation of the secondary lining and the long term stage, representing the 120 years design life of the structure. The intermediate stage considered either undrained conditions in tunnels excavated within the London Clay B/A3 strata, partially drained in tunnels excavated within the London Clay A2 strata and fully drained in tunnels excavated in the Lambeth group.

The long term stage considers full drained conditions in all types of grounds, long term groundwater levels and allows for the degradation of the 75mm initial layer. Three-dimensional analyses were carried out for the largest crossover tunnel cross-sections, for instance, where the presence of the headwall had a positive influence on the final output of the lining forces. The models included the effects of all the adjoining structures and made allowance for tolerances at each construction stage.

A simplification inherent in 2D modelling is the relaxation of the ground ahead of an excavation face which is largely a 3D effect. The values used in the 2D modelling were accordingly calibrated against data from tunnels previously constructed in London Clay and 3D models.

Time dependent development of sprayed concrete strength and stiffness were also included in the models, using the strength gain curves from the Crossrail Sprayed Concrete Lining Specification and a relationship between the stiffness value and strength. Eurocode 2 permits a reduction of the sprayed concrete lining stiffness to take account of the creep and relaxation that occurs during early age loading before reverting to full stiffness prior to secondary lining installation. Plastic hinges were also allowed to develop, with a limit on the maximum rotation and the number permitted per cross section.

The effect of the compensation grouting was applied through a spreadsheet that allowed the application of factors to the calculated lining forces.

This spreadsheet was developed using a series of 2D (See Figure 7) and 3D models in which the effects on the tunnel lining forces was investigated for different lining thicknesses, proximity to the tunnel, tunnel shape and grouting pressure, and considering the reversal in stress path direction that takes place around a tunnel lining when compensation grouting is carried out.

An example of the effect can be observed in Figure 8. Figure 6 illustrates the exclusion zone that was specified for the construction of the SCL tunnels.

The design of the opening in the primary linings was also standardised following studies for different aspect ratios between parent and child tunnel and different tunnel shapes to obtain design charts which relate the concentration factor (see Figure 9) of the forces and moments in the area around the opening to the ratio of the child to parent tunnel.

The secondary lining was designed to resist the internal forces induced by its own self weight (no compression in the secondary lining), shrinkage effects, its share of the ground loads in the long term, the long term groundwater load and also temperature effects, mechanical and electrical loads, fixing loads and degradation of 75mm of the secondary lining due to the effects of a fire in a tunnel (accidental load case only).

Two dimensional and 3D computer structural models (see example in Figure 10) based on the centroid of the tunnel lining cross-section, allowing for required tolerances, and the application of all possible load combinations, were used to determine the secondary linings forces.

For the biggest cavern cross-sections and unconventional tunnel shapes, plastic hinges were also allowed to develop such that crack widths in the concrete lining due to rotation did not exceed 0.3mm in the serviceability limit state.

Testing of all the possible sprayed waterproofing membranes has not been able to demonstrate that they provide sufficient bond with sprayed concrete in the long term under saturated conditions and sustained loading, and therefore the secondary lining design did not allow for full composite action between the primary and secondary linings. It was assumed that the membrane allows full slip to take place between the linings.

The secondary linings were also designed to provide sufficient residual capacity to resist ground and hydrostatic loads after a tunnel fire represented by the RABT-ZTV (Eureka) time-temperature fire curve.

Fire testing of various secondary lining test panel configurations reflecting the fire proofing thickness and mix designs adopted on each contract were undertaken to demonstrate the linings’ ability to resist the RABT-ZTV fire curve and therefore obtain fire assurance.

Typical Performance Requirements

The sprayed concrete specification required the use of steel fibres of the deformed type Class I or II and a minimum steel fibre content of 30kg/m3 although a higher dosage may be required to meet the requirements for flexural strength as indicated in Table 1.

Lessons

  • The division of design responsibility in the temporary and permanent condition led to complicated interfaces between different design consultants and lengthy processes in order to gain approvals and compliance. The establishment of clear division of design responsibilities and a reduction in the number of parties involved on future projects would lead to both more optimal and economical SCL design solutions.
  • Further research and development is required into the properties of sprayed waterproofing products to ensure the bond strength between the membrane and the concrete can be relied upon in the long term to allow full composite action of the linings to be considered in future SCL designs.
  • There was a learning curve in the use of sprayed waterproofing membranes. The quality, particularly the smoothness of the regulating layer on which the membrane was sprayed and the final thickness achieved were checked independently, which proved essential to the final quality of the waterproofing.
  • Beam tests to measure residual tensile strength caused some difficulties in terms of clear understanding of the results. A guidance report was produced to aid proper interpretation of the results.
  • Appropriate hydrogeological and geotechnical parameters from comprehensive, well-managed ground investigations reduced geotechnical risk.
  • Maintaining the profile, tolerance and adherence of thin layers of shotcrete has proved challenging and therefore the tolerances for each sprayed layer should be reviewed and the number of layers minimised. Traditional use of profile bars to control the accuracy of final profiles has proved successful in conjunction with appropriate selection of working ‘targets’ to control spraying. There has been some success with the use of ‘semi-automated’ application of layers with equipment such as the Atlas Copco ‘Logica’ system, but this requires careful setup and control.
  • Fire testing of SCL materials incorporating typically 2kg/m3 of polypropylene (PP) fibres has indicated compliance with the specifications when tested under the Eureka fire profile with and without the inclusion of steel fibre reinforcement and profile bars. This has led to the opportunity for the incorporation of PP fibres within the secondary lining and elimination of the final 50mm surface fire proofing layer, which required appropriate cleaning and preparation of the lining surface to avoid debonding, accurate spraying of a thin layer and tight control of the final profile.
  • SCL is a very versatile construction method that allows tunnels of any shape, size and length to be constructed. However, the designer should consider when space proofing the structure that the most non-circular tunnels are likely to require heavy reinforcement which could pose health and safety risks to install and lead to quality issues due to problems with encapsulation of the rebar and shadowing