ITA - AITES WORLD

TUNNEL CONGRESS
21 - 26 April 2018
Dubai International Convention
& Exhibition Centre, UAE

ORAL PAPER
PROCEEDINGS
 Crossrail Farringdon station – lessons
learned, added value and innovatio

Petr Salak1, Adrian St. John2 and Angelos Gakis3 1km of SCL tunnels were built at depths of up to 35m below ground level with
1 MSc, CEng, MICE, CMgr, MCMI, Eur Ing, Design Manager at Dr. Sauer & excavation cross-sections varying from 25m2 to 110m2.
Partners, London, United Kingdom, psalak@dr-sauer.com
2 BEng (Hons), CEng, FICE, Eur Ing, Chief Engineer at BFK Joint Venture,
London, United Kingdom, Adrian.StJohn@kier.co.uk
3 Dr. Dipl.-Ing, MSc, DIC, CEng, MICE, Chief Geotechnical Engineer at Dr.
Sauer & Partners, London, United Kingdom, agakis@dr-sauer.com
ABSTRACT

Farringdon station, on the Crossrail project, was one of the most geotechnically
challenging tunnelling projects ever to be constructed in London.

Almost 1km of sprayed concrete lined tunnels were excavated, most of which in the
unpredictable Lambeth Group, with live interfaces to the box excavations and the
two TBM drives that were terminated at Farringdon while the main station tunnelling
works were underway. Key to the successful outcome was the optimisation of the Figure 1. Plan view of the main structures of the completed Farringdon station.
depressurisation and in-tunnel investigation design, the design and construction
of five additional, temporary tunnels, the optimisation of the excavation sequence The tunnelling works were executed mostly in the Lambeth Group formations
and support, the implementation of ground treatment (dewatering, chemical (approximately 85%) and to a lesser extent in London Clay and Thanet Sand units.
grouting) at the encounters with non-cohesive units and the overall geotechnical The encountered Lambeth Group units were the Upper Mottled Beds (UMB),
risk management. overlying the Laminated Beds (LTB) and the Lower Shelly Beds (LSB), followed by
the Mid Lambeth Group Hiatus (MLGH) and the Lower Mottled Beds (LMB). At the
The successful completion of the tunnelling works provides an opportunity to bottom, a layer of gravels separated the LMB and the Upnor Formation (UF).
assess the main design and construction aspects and draw conclusions that can
be useful for similar projects in the future. Lambeth Group clay units proved to be ideal for open face excavation tunnelling
method, providing sufficient stand-up time in the opened faces and only minor,
Key Words: Risk & probability analysis, Tunnels & tunnelling, Groundwater. localised instabilities. Interbedded sand lenses were however encountered in both
UMB and LMB. The sand lenses of the LMB were dry to moist due to the under-
1. INTRODUCTION drained pore pressure profile of the area, which is attributed to the historical water
abstraction from the lower aquifer. The sand lenses in the UMB on the other hand
Crossrail is delivering the Elizabeth line, a new railway for London and South were water-bearing in most of the encounters. The tunnels in Farringdon station
East. With 42km of tunnels and 10 new stations, all step-free, it is Europe’s largest were completed successfully despite encountering in two cases extensive, water-
infrastructure project. When the railway is fully operational in 2019, the Elizabeth bearing sand lenses. In both cases, the sand lenses were approximately 3 to 4m
line will increase London’s rail capacity by 10 per cent, reduce journey times and thick with a pore-water pressure of 0.7 to 1bar and exhibiting hydraulic continuity.
provide direct travel opportunities across the capital.
The top of the subsurface of central London is dominated by utilities and shallow Five faults were encountered during the tunnelling works as shown in Figure 2.
or deep foundations as anticipated in a city with 8.5 million inhabitants. Going one Soil properties were expected to be deteriorated in the faulted zones, however
level deeper, a dense network of existing tunnels, the “Tube”, provides the main no effects were noticed. To elaborate, the exposed tunnel faces were as stable
means for transporting the population of London daily through the city. As Crossrail inside the zones of influence of faulting as they were outside, and the measured
is the latest in the series of metro lines, the alignment had to go even deeper convergence of the SCL lining was similar.
to avoid these existing assets. Farringdon station tunnels broke new ground; the
tunnels were excavated mostly beneath the well-known London Clay, into the A description of the main units as encountered in the exposed tunnel faces is
Lambeth Group, a formation with unknown behaviour in exposed tunnel faces. presented in Table 1.

2. SITE SPECIFIC INFORMATION

The station included two ticket halls, two escalator inclines with lower concourse
tunnels, two platform tunnels, numerous cross passages and ventilation tunnels
as well as four stub tunnels (see Figure 1). The primary linings were constructed
using steel fibre reinforced sprayed concrete and the secondary linings cast in-situ
concrete in combination with a sheet waterproofing membrane. Overall, around

1 2
 Figure 2: Plan view of Farringdon Station showing the UMB Sand Lenses, the
faults, the boreholes as well as the Reference Design proposed deep wells.
 

Table 1: Description and SCL considerations of the geological units as

encountered in Farrington (from Gakis et al 2015)

3 4
 3. LESSONS LEARNT AND BEST PRACTICE the challenges alluded to above and the efforts made by all parties to derive benefit
from the OCI process, it is the authors’ view that far greater benefits could have
The following sections provide a commentary on selected topics from the been leveraged by engaging the Contractor earlier in the SCL design process.
construction of Farringdon station, with respect to lessons learnt and best practice. Many of the key design decisions had already been made by the time that BFK were
appointed, and any benefits arising from the Contractor’s alternatives could not
3.1. Division of Responsibility (DOR) and NEC3 Option C have been achieved without considerable abortive work and program implications.

As a result of extensive market engagement by Crossrail during 2009, the main 3.3. TBMs First
tunneling and civil engineering contracts were let as ‘construct only’. Whilst this
approach was probably the optimum arrangement in general terms, it created One of the great successes of the combined C300/C410 and C435 contracts
a complex Division of Responsibilities for the sprayed concrete lining design, (all delivered by BFK) was the effective and efficient alignment of SCL and TBM
particularly since the Client’s designer (Mott MacDonald) aimed to create programs. BFK’s ‘TBM first’ initiative was a bid-winning solution which enabled the
efficiencies in the SCL design by utilizing the primary lining to share load in the Client to combine contracts C300 and C410, deliver an £80m cost saving, 9-month
permanent case. This combination of factors posed a particular challenge to program saving, reduce ground movement by 30% and remove 50,000 lorries from
the teams; how could the design responsibility be clearly and effectively shared Central London’s roads. The ‘TBM first’ initiative was also deployed at Farringdon
between the Contractor and Client, where the former was responsible for temporary with great success, particularly in the light of the challenging and varied ground
works, sequencing, face sub-division, face stability and ‘temporary [stability] conditions. The TBM drives were extended to the eastern ticket halls before being
measures’, and the latter responsible for the permanent works which included the ‘turned and buried’ permanently. This TBM pilot tunnel provided the SCL team
sprayed concrete primary lining in both its permanent and temporary conditions. with a safe and robust environment from which probe drilling, depressurization,
In particular, at what point would the temporary shotcrete lining and associated chemical grouting and ground treatment could be carried out prior to the open face
trigger levels pass over from one designer to the other? Furthermore, for a variety SCL excavation for the platform tunnel enlargement. The benefit of such a tunnel
of valid complex reasons, a different approach to this conundrum was taken by cannot be overstated both in terms of safety and productivity terms.
each of the Crossrail SCL contracts (C300/C410, C435, C510, C360 and C305). At
Farringdon, the Contractor’s temporary works design (Dr Sauer & Partners - DSP) 3.4. Temporary SCL Structures
was considered to pass to the Client’s permanent works design (Mott MacDonald)
when the invert had been closed and the shotcrete achieved 28-day strength. There were a number of SCL structures which were required on a purely temporary
With the benefit of hindsight it could be argued that none of the contracts quite basis, and these act as worthy exemplars to demonstrate the benefits of Contractor-
achieved an optimum outcome, but the project teams worked hard to make each designed SCL works with a single SCL designer and Cat III checker.
one a success in its own right. It is the authors’ opinion that Best Practice for future The benefits of the temporary wraparound (PL2RC) and temporary connection
SCL projects – even those let as construct-only – means that there should only (CP1-CH1) are described in Gakis, Salak, St.John (2015a). The benefits of the
be one SCL designer. Given that SCL design is inherently method-related, will TBM reception chambers for the incoming C305 TBMs are described later in
most likely mean that the Contractor will take responsibility for the design if any of section 4.1. The designs delivered by the BFK/DSP team were healthier, and
that lining is to be relied upon in the permanent case. This will avoid unnecessary safer to construct, with a reduced carbon footprint, and produced significant cost
duplication, complexity and complication of roles and design responsibilities. and program benefits. The TBM chambers were the largest ever SCL tunnels to
By contrast, the use of the NEC3 Option C conditions of contract and a Geotechnical be built in London without a pilot tunnel and were safely constructed with no bar
Baseline Report proved to be of great benefit to the project. The NEC3 contract reinforcement and no additional thickenings around the junctions.
type actively promoted co-operation, collaboration and communication, and meant
that the project team had shared objectives, shared risks and a shared stake in 3.5. Depressurization Design
successful outcomes. The GBR provided an effective, objective measure by which
the Contractor and Project Manager could allocate risk and determine entitlement BFK/DSP team delivered a risk-based optimization of the reference investigation/
to a Compensation Event. This provided the Client with a competitive initial Target depressurization design, which led to a reduction of the total in-tunnel probing
Price and the mechanism to protect the Contractor if adverse ground conditions quantities by 70% and descoping of the 52 prescribed deep dewatering wells which
were encountered. were accompanied by associated access/logistics restraints (see Figure 2).

3.2. Early and Optimized Contractor’s Involvement (ECI vs. OCI) This optimized design was supported by additional information provided by an
extension of the instrumentation and monitoring boreholes to a final depth down
Upon appointment of BFK as the main contractor for Crossrail Farringdon Station to at least 4m below tunnel invert level, and the use of additional information from
(C435), the project team embarked upon a 12-week period of ‘Optimized Contractor boreholes & shaft construction. Additionally, the 3D geological model provided a
Involvement’ (OCI). In SCL terms, this period focused on achieving alignment of detailed conceptual model and a tool which drove the reduction of the geotechnical
assumptions, methods and sequence between the Contractor (and DSP, their SCL risks.
designer) and the Client (and Mott MacDonald, their SCL designer). At this stage,
the SCL design had achieved a maturity aligned to RIBA Stage D. Notwithstanding
5 6
 The implementation of the BFK/DSP design was based on a set of simple
performance criteria for groundwater flow and pore pressure control (Gakis, Salak,
St.John 2015b). These criteria were easily integrated in the site supervision work-
stream enabling rapid, objective decision making. This process was successfully
followed during the execution of the works in the multiple instances when water
bearing sand units were encountered, enabling decision making aided by simple
and reliable in-tunnel measurements.

3.6. SCL Tunneling in Sand

In two cases during the excavation of the Platform Tunnels, extensive, water-
bearing sand lenses were encountered. In platform tunnel, west, the sand lens
was encountered in a moderately unfavorable position, that being in the bottom of
the top heading and in the bench excavations. In platform tunnel, east however, the
position of the sand lens was significantly more unfavorable as it extended in the
crown of the planned enlargement and above.

In both cases, the sand lenses were approximately 3 to 4m thick with a pore-
water pressure of 0.7 to 1bar and exhibiting hydraulic continuity. The proposed
solutions included a combination of ground treatment methods with the available
SCL toolbox items.

In platform tunnel, west, it was decided to depressurize the sand lens through the
installation of numerous in-tunnel well-points through a constant pressure vacuum
dewatering process. The excavation of the enlargement in the affected area
commenced once a steady pumped flow rate was reached, after approximately
one month of continuous pumping. The face was split into four pockets and no
significant instabilities occurred as the tunneling works advanced with an average
rate of 1.8m/day.
Table 2: Details of in-tunnel depressurization and chemical grouting
In platform tunnel, east, it was decided to mitigate the significantly higher risk by installations in PTE.
a combination of in-tunnel depressurization and sodium silicate grout injections.
The former reduced the initial water pressure and the latter enhanced the sand 3.7. Permeation Grouting
by reducing its permeability and increasing its undrained shear strength. In total
approximately 100m3 of grout were injected, and the excavation works within Prior to the start of excavation works on escalator tunnel (ES1), permeation grouting
the affected area commenced once additional ground investigation yielded was required to mitigate the risk of water ingress and instability on the non-cohesive
satisfactory results. The top headings were split into 3 or 4 pockets, and spiles were upper units. Furthermore, a buried oil tank that was encountered alongside the
installed in every second top heading to further enhance crown/roof stability. The eastern pile wall of shaft SH-W3 also required treatment by means of permeation
implementation of all the measures proved to be successful, and no instabilities grouting to avoid the migration of contaminated material. The encountered oil tank
occurred as the tunnel was excavated with an average rate of 1m/day. obstruction was backfilled with weak concrete prior to ES1 construction. Other
structures such as foundation structure of the Thameslink Railway, Thameslink
Railway Tracks and redundant Fleet sewer, were in the proximity of the permeation
grouting block.

The Made Ground was not an in-situ material hence was highly variable, comprising
units from clayey sand and gravel to gravelly/sandy clay mixed with bricks, concrete
or similar. The River Terrace Deposits mainly consisted of clayey/sandy gravels
in a loose matrix or silty fine to coarse sand. Both formations were considered
permeable hence could be locally water bearing. Based on the information from
boreholes the assumed top level of London Clay was approximately at 102 mATD.
The permeation grouting had to be carried out from within the shaft SH-W3, by
installing Tube-A-Manchette (TAMs) PVC Pipes Ø 50mm through the unreinforced
7 8
 piles of the eastern pile wall. Two layers of TAMs were foreseen to achieve These displacements were back-calculated using 2D and 3D FEA to derive the
construction of the designed grouting zone. The success of the permeation effective grouting pressures acting above the tunnel crown (Gakis, Grau, Schwind
grouting was verified by three validation holes, prior to excavation works. In case of (2015)) and these validation outputs serve as a useful reference for future tunneling
unsatisfactory results, the permeation grouting was to be repeated until the criteria projects.
were met.
The most important findings were the following:
The target unconfined compressive strength of the grouted ground was 300kPa •T
 he 2D FEA calculated a crown displacement of 13mm prior to the rear
and target permeability was less than 5 x 10-7m/s. The target volume criteria for injections and 15.6mm after. This additional crown displacement of
grout injections were 8m3 or 8000l of injected grout. With assumptions of 30m3 approximately 2.5mm tied well with the observed monitoring results.
volume of treated block, 35% porosity of MG, 35% porosity of RTD and 75% degree •A
 grouting pressure of 50 kPa, applied in 4 consecutive stages (a total
of filling the voids (= volume to grout). of 200 kPa) was the assessed best estimate for the effective grouting
pressure in the 2D FEA.
Assessment of injection progress and monitoring results were presented, discussed •T
 he 3D FE back analysis produced similar results. The detailed simulation
and adjusted within the SRG / RESS. Movements of Crossrail and third-party assets of the stepped excavation and support of the platform tunnels as well as
resulting from the grouting operations were maintained within the specified limits. of the rear injection episodes, resulted in an additional 2.5mm of crown
displacement for an effective grouting pressure of 50 kPa (in 4 rear
3.8. Compensation Grouting injection episodes hence a total of 200 kPa).

Compensation grouting was employed to control and mitigate the surface 3.9. Proximity of adjacent SCL faces
settlements induced by tunnel excavations. Five shafts, supported by pre-cast
concrete rings, provided the vertical access for the installation of the tube-a- The project specifications for SCL tunnels stated that “Where two tunnels are
manchette pipes (TAMs). The boreholes that housed the TAMs were drilled using simultaneously under construction, there shall be at least 25m between the ring
predominately dry auger drilling. The drilling was instrumented, and a survey of closure point of each tunnel”. This specification presented challenges for the
the as-built positions was performed in all cases to ensure compliance with the contractor with regards to concurrent construction of the Eastbound Platform tunnel
prescribed tolerances (1:100 vertically and ±500mm at the distal end of the tube and the East Concourse tunnel, and construction program. After examining the
horizontally). interaction between the two tunnels in the 3D FE analysis models, it was concluded
that the minimum required distance between the closed rings of concurrent
The location of the grout shafts and the total coverage of the TAMs are shown in tunneling excavations could be reduced to 15m. This distance was regarded as
Figure 3. adequate for the two tunnels to advance simultaneously without impact on their
primary SCL linings. The construction of the tunnels was completed without any
exceedance of the deformation trigger levels.

3.10. Tunnel Construction Advance Rates

Tunnel advance rate can be quantified either as length per time or number of
advances per time. When the excavation sequence is divided in top headings/
bench/inverts, as in most cases at Farringdon station, it is easier to average the
advance rate in meters per day. As such, one meter per day advance rate implies
advance of an equivalent 1m long ring in one day.

The advance rate of a tunnel depends on numerous factors including but

not limited to:
Figure 3: Compensation grouting coverage at Farringdon station. The location of • Size/geometry/thickness of the tunnel
the grouting shafts as well as the extent of the TAMs are shown.
• Ground conditions/water pressures
Pre-heave of the ground was achieved through front injections ahead of the
advancing SCL tunnel face. This was followed by rear grout injections, which • Available tunneling plant
were performed above the completed tunnel sections with 5m minimum vertical
clearance. The rear injections induced an additional, average 2.5mm downward • Excavation and support sequence
displacement of the platform tunnel crown.
• Concrete quality/strength

9 10
 • Experience of contractor/ tunneling crews 4.1. Large tunnel without pilot tunnel

• Pre-support/mitigation measures There is limited case history in the UK demonstrating the successful construction of
large open face SCL tunnels without the use of a temporary pilot tunnel, temporary
• Unforeseen down time sidewall or temporary invert. Current practice is limited to a maximum of circa 8m
An analysis of the production rates at Farringdon station illustrated the benefit of diameter. The project team sought to challenge the conventional approach and
the ‘TBM first’ approach: deliver a more efficient solution.

• CH2, which was constructed with an SCL pilot was constructed at a long TBM reception tunnel STE1 was 9010mm high by 9410mm wide, with 400mm thick
average of 0.6 m/day. This comprised 1m/day for the SCL pilot and 1.2m/ steel fiber reinforced primary sprayed concrete lining. The primary lining thickness
day for the enlargement. was driven primarily by forces induced during TBM dismantling. DSP and BFK’s
preferred excavation sequence comprised 2 no. 1m long top headings, 1 no. 2m
• By contrast, PTW was constructed with a TBM pilot with a long average long bench and 1 no. 2m long invert, providing a minimum stagger of 1.5m between
of 1.0m/day. This comprised 12m/day for the TBM pilot and 1.1m/day for the faces. This solution provided the most effective compromise between face
the enlargement. The figures for PTE were 18m/day (TBM), 1.1m/day stability and rapid ring closure, thus minimizing in-tunnel convergence and surface
(SCL) and 1.2m/day (long average) for comparison purposes (see Figure settlement.
4). This illustrates the benefit of the ‘TBM first’ approach, and aligns with
the findings of St.John, Potts, Perkins, Sismondi (2015) for the other BFK The construction of STE1 without a pilot tunnel or temporary invert / sidewall
tunnels. delivered a number of efficiencies; it saved 8 days on the critical path compared
with a tunnel with a temporary pilot tunnel, and delivered savings of £300,000 in
direct costs alone. It also saved approximately 120m3 of shotcrete, equating to
a saving of 28tonnes of CO2. It also provided health and safety benefits to the
workforce and to nearby stakeholders; no temporary shotcrete to demolish meant
a substantial reduction in dust, noise and vibration. It also reduced the number of
lorry movements to and from the site, through a reduction in shotcrete volume and
reduction in wastage. These direct cost and savings were inevitably restricted due
to the short length of the STE1 tunnel. However, this new precedent, particularly
in such challenging ground conditions, will provide far greater benefits to the UK
tunneling industry and future projects.

4.2. Openings in previously constructed tunnels

A design philosophy which considered the contribution of the steel fibers in the
flexural strength of sprayed concrete linings was applied in five cases where
openings on existing SCL tunnels were required. Exploiting the strengths of non-
Figure 4: Average advance rates in platform tunnel east (PTE) and west (PTW) in linear 3D FE analysis and the plasticity characteristics of steel fiber reinforced
Farrington. concrete linings led to optimized designs beyond the common practice. No
additional introduction of shell thickenings or rebar reinforcement were required.
Quite apart from these productivity gains, the safety benefits of driving a TBM The avoidance of these measures delivered notable health and safety benefits
pilot tunnel (instead of an open face SCL tunnel) through challenging, faulted and as it eliminated the need for manual handling, work at height and steel fixing. The
varying ground conditions with the risk of encountering unpredictable saturated quantified savings for a single opening are summarized in Table 3.
sand lenses cannot be overstated.

4. INNOVATION

In several cases Farringdon station design and construction stepped beyond the
common practice applying unique, bespoke solutions that can provide valuable
reference for future projects in similar environments.

11 12
 Table 3: Savings in programme duration, material usage and environmental The advent of fiber optic strain technology (using Fiber Bragg Gratings, FBG) offers
impact when avoiding the installation of additional reinforcement and thickening a further method of monitoring tunnel behavior.
in one opening (after Gakis, Salak (2016))
As fiber optic strain technology is relatively new, the authors decided to install all
three monitoring methods mentioned above in parallel arrays within tunnel RTE2 to
allow direct comparison of all methods. The RTE2 was chosen due to the proximity
of the eastern ticket hall box, a Grade II Listed structure overhead, a buried TBM,
multiple SCL tunnels interacting with each other and active compensation grouting.
The escalator tunnel (ES2) was so close to RTE2 that the top heading of RTE2
and the invert of ES2 were in contact with each other. Therefore, a very detailed
understanding of tunnel behavior was needed.

FBG sensing cables were installed by Monitor Optic Systems at the extrados of the
shotcrete lining to measure radial strain within the lining itself. Strain and temperature
sensing cables are very small in diameter (2mm and 5.5mm respectively) but they
were proven to be robust enough to withstand shotcrete spraying pressures of up
to 7 bar. FBG monitoring system was remote automated system which allowed
data acquisition and review of data in real time.

The correlation between the direct stress measurements using concrete stress cells
and the strain measurements using fiber optics was investigated and the results
are presented and discussed in Salak, St. John, Gakis (2015). Fiber optics proved
In a hypothetical mined station with 20 openings, these savings would be translated to be a strong alternative to pressure cells, and the authors would not hesitate to
to 120 days in program duration, 90 tons of steel, 3000m3 of concrete, 3000m3 of specify and use them in the future projects for critical areas.
excavated material and 1000 tons of CO2 emissions. Figure 5 shows one of the
completed openings in Farringdon. 4.4. Geotechnical Risk Management

This solution maximized the use of robotic and remote construction methods and Due to the particularly high geotechnical risks related to the presence of the water-
showed excellent correlations with the predicted tunnel deformations. bearing units in the Lambeth Group and the geological faults, the geotechnical risk
management was a key aspect of the successful delivery of this complex station.
The BFK/DSP team focused on three elements:

• hazard identification
• reducing the risks to as low as reasonably practicable (ALARP)
• Updating the construction design and management (CDM) risk register
provided by the client at Contract Award to reflect the actual design
measures proposed by DSP, including implementation of depressurization
and toolbox items.

In order to illustrate and communicate to the team the spatial distribution of the
geotechnical risks, a hazard mapping was performed in several design and
construction phases as shown indicatively in Figure 6.

Figure 5: Completed opening and on-going excavation of “child” tunnel.

4.3. Fibre Optics Monitoring

Monitoring of a tunnel’s structural behavior during construction normally relies on

in-tunnel displacement monitoring, with designers occasionally prescribing the use
of pressure cells in critical areas. However, hydraulic pressure cells appear to be Figure 6: Mapping of Geotechnical related Hazards (detailed in Gakis, Salak, St.
fraught with difficulties which may render much of the data of questionable value. John 2015b)
13 14
 An optimized geotechnical risk management framework was developed, which 6. ACKNOWLEDGEMENTS
assessed all of the available data in order to minimize the residual risks at the
tunnel face. In addition to the investigation already carried out from the surface, a The authors would like to acknowledge the significant contribution from BFK joint
pattern of in-tunnel probing was prescribed throughout the SCL works. In addition venture, Crossrail and Dr. Sauer & Partners for providing the necessary resources,
to the data acquired through probing, face mapping data was integrated into an guidance and support in the preparation of this paper.
existing 3D Geological model that was originally developed in 2009 by the British
Geological Survey (BGS). This 3D model provided a useful day-to-day visual REFERENCES
geological database, and allowed the site team to make accurate predictions for
forthcoming SCL works (see Gakis et al 2015). The cycle of risk reduction presented Gakis, A., Salak, P., St.John, A. (2015a). Temporary Sprayed Concrete Lining
in Figure 7, started with the prediction based on the 3D geological model. After the Tunnels in Farringdon Crossrail Station. Proceedings of the World Tunnel Congress
excavation of each SCL tunnel, the data from the face mapping was integrated in 2015: Promoting Tunnelling in SEE Region. Dubrovnik, Croatia.
the 3D model. Thus, the risk was reduced as the level of knowledge and confidence
was constantly increasing. At the same time, in-tunnel probing provided additional Gakis, A., Salak, P., St.John, A. (2015b). Innovative Geotechnical Risk Management
data, validating the prediction from the 3D geological model. for SCL tunnels. Proceedings of the ICE Geotechnical Engineering, vol. 168. pp
385-395. Thomas Telford, 2015.

Gakis, A., Grau, B., Schwind, T. (2015). Tunnel Deformations caused by

Compensation Grouting at Crossrail Farringdon Station. Proceedings of the World
Tunnel Congress 2016: San Franscisco, USA.

Gakis, A., Salak, P. (2016) Efficient Design of Openings in SCL Tunnels. Tunnels
and Tunnelling International, October 2016, pp. 41-46.

Salak, P., Gakis, A., St.John, A. (2015). Taking the Strain, Coping with Pressure or
Shooting Targets? An Objective Comparison between Monitoring Solutions within
Sprayed Concrete Lined Tunnels at Crossrail Farringdon Station, London. Crossrail
Technical Papers Competition 2015.

Gakis, A., Cabrero, P., Entwisle, D., Kessler, H. (2015). 3D Geological Model of the
Completed Farringdon Underground Railway Station. Crossrail Technical Papers
Competition 2015 Winner.

Figure 7: Cycle of Risk Reduction

5. SUMMARY AND CONCLUSION

The solution for sprayed concrete lined (SCL) tunnels at the Farringdon station
site incorporated a unique combination of innovative design methods and an
integrated design and construction team to deliver major cost, time and logistical
benefits to the project. The geotechnical conditions were the principal challenge,
governing both the design and construction process. The tunnels were excavated
mainly within the Lambeth Group, a formation comprising stiff to very stiff over-
consolidated clays with interbedded, water bearing and pressurised sand lenses
up to 3m thick. The design was carried out by combining sophisticated non-linear
3D FE models, with the 3D geological model, integrating the most recently acquired
geological data. This paper presented the main design & construction steps that
led to the successful completion of the sprayed concrete tunnel works. Farringdon
station set new precedents for the UK tunnelling industry which can be profitably
applied to other projects to generate similar efficiencies and benefits in terms of
cost, programme, risk and health and safety.

15 16
 THANK YOU
FOR VISITING

ITA
AITES