The contribution of the primary lining to the long term support of the tunnel has traditionally been discounted in the design of shotcrete lined tunnels. However, the development of alkali-free accelerators has led to the challenging of the view that shotcrete will suffer significant deterioration with time. It was recognised that the interaction between the primary and secondary linings raised several interesting design issues, and detailed studies were commissioned from specialists within RLE member firms, Arup and Halcrow. These studies confirmed that the primary lining would contribute to the structural support of the tunnel throughout the 120 year design life, but concluded that the design would be extremely complex due to the different stress and strain histories of the two linings. Based on these results it was decided to only consider some of the contribution of the primary lining by using a design philosophy which became known as “gray rock”. The primary lining would be modelled as part of the ground around the tunnel with relatively conservative parameters (c’=0, f=30º and E=10GPa).

Waterproofing

In the tender design, crack control to minimise water ingress was provided by reinforcement throughout the length of the tunnel. Since approximately two thirds of the tunnel length also required reinforcement to provide structural capacity, continuing with crack control reinforcement for the final third of the tunnel was more efficient than providing a waterproof membrane.

The redesign of the tunnel meant that structural reinforcement was no longer required in the vault and a waterproof membrane became the most economic option. The membrane was designed to be tied into a waterstop cast into the top of the concrete base slab. A conventional system was adopted incorporating a 500g/m² fleece and a 2mm PVC membrane by backgrouting, if the need should arise. The waterproofing work was subcontracted to EIGCC of France.

The Secondary Lining

The secondary lining design was again carried out using FLAC. The whole construction sequence was modelled, incorporating the new range of in-situ stress ratios and minor adjustments to other geotechnical parameters (the observed RQD values during the construction of the top heading allowed better corrections for the mass factor to be determined). The gray rock approach for the primary lining was only used when the secondary lining alone had insufficient capacity, generally in the low cover areas. The flat invert was subject to a rigourous design, including the use of beam-spring models. The effects of normal train loading and derailment impacts on the invert slab were considered. The final design had a lightly reinforced 600mm thick concrete invert and an unreinforced vault, 350mm thick.

The on-site RLE design team worked closely with the Eurolink engineers responsible for planning construction of the reinforced concrete invert. The contractor was keen to ensure that casting of the invert was completed as efficiently as possible. The concept of a continuously poured slab was developed, with reinforcement prefabricated into side cages and mesh sheets. The slab was placed from the centre of the tunnel, working outbye. An overhead boom pump allowed the steel fixers to place the reinforcement between the pump and the section where the concrete gang was working.

To enable quick and accurate placement of the invert reinforcement, it was important to lay the blinding accurately. To achieve this, a surface miner, more commonly used on road construction, was used to remove the final 300mm of chalk during invert excavation to within +/-10mm of the required level. The lean mix concrete blinding was then laid using a paving machine to similar tolerances. Both machines worked under the guidance of a laser system to provide level control. With this system, the final invert excavation and blinding of the entire 3.2km of tunnel took under 20 days. Although CTRL specification, following previous UK tunnelling practice, called for a maximum vault shutter length of 10m and a minimum concrete strike strength of 10N/mm², Eurolink had presented the case for adopting the continental practices of using a 12m long shutter and a reduced strike strength. The contractor carried out design checks to demonstrate that a striking strength of 7N/mm² still provided an adequate factor of safety. Eurolink also provided evidence of successful case histories. The team adopted a strategy of gradual change to ensure the proposed changes did not lead to reductions in quality or safety. RLE accepted the Eurolink proposal to order two 12m long vault shutters, with the proviso that only 10m of the shutters would initially be assembled and used. The shutter length was increased to 12m after the satisfactory completion of the first 100m of lining. The striking strength was gradually reduced to 7N/mm², which reduced the striking time sufficiently to allow a 24 hour cycle time to be achieved. These changes allowed the placing of 120m of vault lining per week, as opposed to the programmed rate of 80m.

Impact of Value Engineering and Design Development

The implications of the design improvements can be summarised as follows:

  • Volume of chalk excavated reduced by 21,000m³;

  • 5,900t of reinforcement saved by removing the reinforcement from the vault, and reducing the quantity of reinforcement percentage in the invert;

  • 35,000m³ of concrete saved by the reduction in the secondary lining thickness;

  • 350% increase in the rate for construction of the reinforced concrete base slab;

  • 50% increase in the rate for constructing the secondary lining vault.

Influencing Factors

The success of the development of the design is due to a number of factors, including:

  • Integrated on-site design team: The RLE design team was incorporated into the management team on site. This allowed good communications, and facilitated the processing of design changes though the RLE system. Potential problems and design queries were dealt with earlier and quicker than would have been possible with a design team located elsewhere. The design team were fully aware of the ongoing construction activities and were able to ensure that the design philosophy was being incorporated into the

    construction process as well as understanding the impact of design delays or changes.

  • Partnering: A partnering culture was supported by RLE from the outset and was adopted by all members of the project team. In general, partnering worked well, with the site being a friendly and relaxed place to work and with good cooperation between parties. All parties shared the same computer and paper filing systems and offices were mixed. There was a good level of trust and the adoption of partnering encouraged the development and sharing of ideas.

  • Contract: The contract adopted for the CTRL was the New Engineering Contract, Option C, which is based on an agreed target price for the works. A pain/gain mechanism was included so that the contractor stood to share in the saving if the actual cost was less than the target price, which encouraged all parties to make design improvement initiatives a success.

Safety

Contract 410, together with C350, the Medway Bridge, which was also awarded to Eurolink, had a good safety record relative to the rest of the construction industry. Safety was well managed on site, with a number of safety initiatives used during the construction period. These included safety walkabouts, audits by Miller Civil Engineering and the RLE Target Zero campaign. The number of reportable accidents per 100,000 manhours is 0.36 to date, compared to an overall average on the CTRL of 0.54 and an industry average of 1.32. However, a small number of serious accidents did occur, demonstrating the need for an ongoing culture of safety improvements in the industry.

Self-Certification

All the CTRL contracts rely on “self-certification” to manage the quality of construction. Self-certification places the responsibility for quality management with the contractor; a concept taken from the petrochemicals industry. The system worked relatively well, especially once the teething problems had been resolved. The term “self-certification” is perhaps misleading, as the role of quality is implied to be minimal. Although the project management team had less staff than a conventional resident engineers team, RLE personnel were heavily involved in quality, auditing and observing the management system in use. This allowed a more focused approach than inspecting every element of the contractor’s work.

Conclusion

This paper demonstrates the benefits of an integrated, on-site design team for the North Downs Tunnel. Significant time and programme savings have resulted from the application of Value Engineering principles and design refinement during construction. The use of the NEC form of contract and the adoption of partnering lead to an enjoyable and creative working environment. The project’s safety record has been considerably better than the industry average, although there is always room for improvement. Overall, the tunnel construction has been a considerable success, contributing towards Section 1 of the CTRL being under budget and within programme.

Related Files
The revised tunnel design