HS2 Tunnelling: The Client’s Perspective

ALL IMAGES COURTESY OF HS2 LTD

PROJECT DESCRIPTION AND UPDATE

There are 44km of twin bore tunnel and 8km of cut-and-cover or Green Tunnel comprising some 19% of the 225km length of HS2 high-speed rail route, between Euston, in central London, and Birmingham. These are being constructed by four different Integrated Project Teams each comprising of HS2 Ltd, the Contractor and the Designer, as follows.

• SCS: Skanska/Costain/Strabag – the London Tunnels;
• Align: Bouygues, Sir Robert McAlpine and VolkerFitzpatrick – the Chiltern Tunnels;
• EKFB: Effage/Kier/Ferrovial/BAM –the Green Tunnels; and,
• BBV: Balfour Beatty/Vinci – Long Itchington Wood and Bromford Tunnels.

Typical bored tunnel cross sections vary between 7.55m and 9.1m internal diameter (ID), respectively; sprayed concrete lined (SCL) cross passages (CP) are around 4m ID and between 12m and 30m in length. Large diameter SCL tunnels, up to 11.25m in diameter across the axis, have been constructed; and, cutand- cover tunnels are generally double-cell precast concrete with each cell around 8.5m x 8.5m.

THE LONDON TUNNELS

Starting at Old Oak Common station (OOC) and going west, all the tunnelling assets were described in turn, starting with the SCL stub tunnels constructed to the east of the OOC box. These are circa 100m-long and will be the launch tunnels for the tunnel boring machine (TBM) drives to Euston station. The tunnels are complete and the TBMs have been assembled ready for a launch in 2026. Access for the stub tunnel construction is via the Atlas Road Logistics Tunnel (ARLT), constructed by a 6.2m diameter, 850m long earth pressure balance machine (EPBM) – a refurbished Crossrail machine – and completed in November 2023. This tunnel connects the SCS logistics hub with the very eastern end of the OOC station box and will be instrumental in the delivery of the Euston Tunnels.

There are further SCL tunnels, 370m-long, at the west end of the station box. These connect to a cross over box outside the station and are up to 13m in diameter to allow for diverging tracks. These tunnels were driven with a full-face pilot tunnel circa 5m-diameter then enlarged to full face. Typical lining thickness are 300mm-400mm for the primary lining and 275mm to 400mm for the secondary. At the time of the BTS presentation, the primary lining here was 75% complete and on hold awaiting a utility diversion. West of the crossover box are some further short SCL tunnels and a shaft making up the launch arrangements for the Northolt Tunnel East.

Northolt Tunnel East (NTE)

The NTE tunnels are 5.6km-long, 8.1m ID and were driven by an EPBM machine in semi open mode for a majority of the drive. They passed through one ventilation shaft at Westgate and terminated at a reception shaft at Greenpark Way.

There are 14 CPs up to 30m in length, comprising of a primary lining, circa 250mm-thick, a sheet waterproofing membrane and a cast insitu secondary lining of 250mm thickness. With the exception of the last CP no ground treatment was required.

Greenpark Way shaft is the reception shaft for both the NTE tunnels – driven west, and also the NTW tunnels driven east. The TBMs were received, in turn, into a reception can and then removed from the shaft. All have been successfully received.

Northolt Tunnel West (NTW)

The NTW tunnels were launched from a portal structure 250m-long at West Ruislip. A shallow launch methodology was used with additional temporary kentledge incorporated at the portal.

The tunnel is 8.8m in diameter and was driven by an EPBM in closed mode for 7.8km to Greenpark Way shaft, via intermediate shafts at South Ruislip and Manderville Road. There are 20 CPs of similar construction to the NTE ones. Due to the ground conditions a mixture of dewatering and ground freezing was employed.

Copthall Tunnel

The final section of tunnelling for SCS is the Copthall Tunnel, a cut-and-cover tunnel approximately 1km in length, west of the Northolt tunnel portal. This is a twin-track, bi-directional, cast insitu cut-and-cover tunnel with a maximum overburden of about 16m.

Originally this was to be a cutting, but the tunnel was introduced partly to assist with the re-use of the spoil from NTW. It is the only twin-track bidirectional tunnel on the project and consequently has several permanently open air-shafts for ventilation and aerodynamic considerations.

The Chiltern Tunnels

The Chiltern Tunnels are the longest rail tunnels in the UK at 16km in length, and 9.1m ID. These were driven by Variable Density slurry TBMs, at a depth of up to 90m, though chalk with the added pressure of crossing under the M25 within 300m of the launch.

The tunnel passes through five shafts up to 63m deep at Chalfont St Peter, Chalfont St Giles, Amersham, Little Missenden and Chesham Road.

A single design used for these shafts comprised a circular diaphragm wall shaft with connection adits; they have headhouses designed to look like agricultural buildings and successfully blend into the surrounding countryside. There are 37 SCL CPs in the Chiltern Tunnels. All of the CP zones were grouted prior to excavation with some requiring a significant grouting campaign due to the 2bar water pressure. The twin-tube tunnel is complete and has been handed over to HS2 with the shafts not far behind.

THE GREEN TUNNELS

EKFB are building three Green Tunnels at Wendover, Chipping Warden and Greatworth varying in length between 1.4km and 2.5km.

Predominantly for environmental reasons these three tunnels are all constructed using the same precast ‘Matiere Arch’ methodology, with a dedicated precast factory being established at the Stanton precast facility in Derbyshire for casting all 10,000 elements.

The structure consists of a 5-piece arch – 2x outer walls, a central wall, and 2x arch roofs with a cast-in-place invert tied to the walls with double hoop bars.

Maximum overburden is around 10m, and two of the tunnels carry permanent road diversions over the top. Following precast arch installation, a sheet waterproofing membrane is rolled out over the top and welded to a sheet installed beneath the blinding and then backfill commences in carefully controlled layers.

BBV TUNNELS

Long Itchington Wood Tunnel

Long Itchington Wood Tunnel is an 8.8m diameter 1.5km-long tunnel, bored using a VD slurry machine through mudstone with 3 SCL CPs. Upon completion of the first drive, the machine was transported back to the start and used for the second drive before being repurposed as the second TBM for the smaller diameter Bromford tunnel.

Bromford Tunnel

The Bromford Tunnel is a 7.55m-diameter tunnel bored using a VD slurry machine up to 35m below ground level on the approaches to the HS2 terminus station at Birmingham Curzon Street.

The tunnel is 5.8km-long and has a single intermediate shaft, consisting of a D-Wall circular shaft with sprayed concrete adit connections for ventilation, services and personnel access. One of the key interfaces for this tunnel is the Bromford Viaduct carrying the M6 Motorway above, which sits within the 1mm settlement contour for 700m of the drive. Following five years of negotiations with the asset owner, National Highways, both drives passed this asset with no impact.

INVERTS AND WALKWAYS

Within the tunnelling contractor scope is the first stage invert concrete and the walkways. Align and SCS, respectively, elected to pour their invert concrete concurrently with the tunnel construction by using a moveable bridge under which a section of invert (including the carrier drain and catch-pits) was cast.

This methodology has significant programme benefits, allowing TBM operations to carry on unhindered as the invert pours were cast below the bridge.

Align elected to use a slip-forming machine to cast the walkways, obtaining excellent productivity, whereas for logistical reasons SCS are using a moveable formwork bridge.

SCHEME DEVELOPMENT

The increase in quantum of tunnelling as the scheme developed was described. In 2012, before the Hybrid Bill was submitted, the amount of cut-and-cover and bored tunnels was 40% less than the current scheme. Tunnels that have subsequently been added include:

1. The entire Northolt Tunnels East – the original alignment was to be at surface level between OOC station and Greenpark Way shaft in Greenford, west London;
2. Bromford Tunnel – as the original alignment was to thread its way through to Curzon Street at grade, negotiating several network rail lines and the M6;
3. The Chiltern Tunnels have been extended by 2.7km; and,
4. New cut-and-cover tunnels at Greatworth and Copthorne.

AERODYNAMICS

A short description was given on some of the aerodynamic issues that affect tunnel design for a high-speed railways, including:

1. Control of aural comfort of passengers – this is the sensation you feel in your ears as a train enters a tunnel;
2. Pressure loading on the walls of the tunnel structures and equipment from the passage of trains; and,
3. Micro-pressure waves. These are sonic booms that, unless mitigated, can be heard at the exit portal as a train enters the tunnel (possibly many km away) unless mitigated. Mitigation of these is achieved with a porous portal at the tunnel entrances.

GEOLOGICAL SETTING

The geology of HS2’s Phase 1 alignment between Birmingham and London was described, showing it spanning between older Triassic Marl and Sandstone beds to younger Tertiary sediments – resulting in quite varied geological conditions for the three areas of HS2’s bored tunnels. Ground Investigation (GI) to examine this geology included 6085 boreholes, drilled to a total depth of 160km.

South

The NTW tunnels are bored through the London Clay Formation, Lambeth Group and Chalk Group, whilst NTE and Euston Tunnels are bored in London Clay. In the varied NTW geology, the Lambeth Group consists of only Reading Formation – the Upper and Lower Mottled Beds, which contain sand channels.

These sand channels are common in the mottled clays, having been deposited as sands, silts and clays during formation of rivers in the Tertiary Period. They were likely deposited in a flood plain environment in meandering and overlapping channels, which led to variable ground conditions.

Challenges

In the first 2km, the NTW TBMs encountered challenges in the Lambeth Group, in particular, the sand channels in the face and above the crown had the potential to unravel above the TBM. The Contractor managed the tunnelling in this ground using their standard operations by running the NTW TBM in full EPB mode, constantly supporting the face, and using tail-skin grouting to fill the anulus around the tunnel lining.

Data from the TBM and ground monitoring is reviewed at the Shift Review Group (SRG), which includes spoil-grout reconciliation. Reconciliation compares the weight of the excavated material for a ring, to the theoretical weight and the volume of grout used in the anulus. Where data suggests more excavation than expected, checks are made for gaps in the grouting or for unravelled sand zones with probing and secondary grouting utilised where necessary.

In these conditions, SCS used some additional tunnelling measures including:

• having conditioners suited to granular ground available;
• applying bentonite under pressure around the TBM shield to support the ground above the crown; and
• longer probing was used at a higher frequency to check for zones of unravelled sand.

The reconciliation process is complemented by the surface ground movement monitoring, which continues for at least four months and goes through a robust close-out process. Ground movement above these tunnels was minimal, even given the potential for settlement above zones of unravelling sand – the volume loss maintained to 0.4% or less.

The ground conditions were also challenging for some CPs, with chalk in the inverts and sand channels in the face and above the crown. These 4m-diameter, 11m-long SCL tunnels faced risks of water ingress from these beds and collapse of the sand. Various methods of ground stabilisation were considered, including dewatering, grouting and freezing. On the basis of safety and programme certainty, SCS elected to freeze the most onerous conditions, a decision endorsed by the HS2 Independent Tunnelling Expert Panel. Of the 21No. CPs in Lambeth Group, 12No. were frozen, 5No. dewatered and 4No. were untreated.

Central

The Chiltern Tunnels are bored through most of the different formations of the Upper Cretaceous Chalk Group, from the older firm, blocky Zig Zag Chalk to the younger moderately dense Newhaven Chalk.

The water table varies from almost 80m below ground level, to being at ground level at the location of the River Misbourne.

Challenges

Challenges tunnelling in chalk include:

• Variation in the geology;
• High permeability and groundwater pressure; and,
• Soluble rock.

In the case of the Chiltern Tunnels, there is quite a variation in the chalk materials through which the tunnels were bored, with varying density of chalk, numerous layers of marls, hardgrounds and flints, including tabular flints, which can have implication for wear on TBM cutterhead tools.

Discontinuities in the chalk give the rock a high bulk permeability and in the Chiltern Tunnels, where the tunnels are deep, the associated water pressures are high. Align reported that they were unable to undertake some of their planned cutterhead interventions in the chalk, due to high groundwater ingress, and all CPs required grouting.

A further challenge of tunnelling in chalk is the soluble nature of the rock, which can be subject to the process of dissolution. The resultant solution features are a risk to tunnelling, including in the Chilterns, providing an additional dimension of variable ground for these HS2 tunnels.

Align used a VD TBM to deal with the variable ground conditions. The presence of a screw and crusher was considered invaluable to deal with the hard flints encountered, which sometimes made up over 10% of the face.

North

The Bromford and Long Itchington Wood tunnels are built through the Sidmouth, Tarporley and Branscombe Formations of the Mercia Mudstone Group. The water table is generally at ground level.

In engineering terms, the mudstone is categorised as a weak rock and has variable material properties, which is a function of its state of weathering. Grading schemes, such as that found in CIRIA C570 ‘Engineering in Mercia Mudstone’, are used to distinguish engineering properties – Grade I & II are structured rock, whilst Grade III & IV are more weathered.

Discontinuities in the structured mudstone give the rock a high bulk permeability, which has caused some historic difficulties in tunnelling with high water ingress. Also, the low plasticity of the mudstone means there is a risk of rapid softening upon wetting, which is a particular risk in TBM tunnelling at the cutterhead and in the muck away system.

Challenges

The Mercia Mudstone grading varied both vertically and horizontally along the Bromford Tunnel. BBV reported challenges with the variability and plasticity of the mudstone. In particular, they cited that the ‘sticky’ material in the TBM cutterhead:

• caused blockages in the slurry network;
• required continual adjustment of the TBM Operating parameters; and,
• requiring regular stopping and cleaning.

TBM PERFORMANCE DATA

The first TBM launched in May 2021 and HS2 has been gathering data ever since. Total HS2 tunnelling was compared to other London tunnelling projects, and it was noted that HS2 has bored more tunnelling than the Jubilee Line Extension, Tideway, London Power Tunnels Phase 1 and 2, and Crossrail.

HS2 has now driven nearly 75km of TBM tunnel, plus about another 2km of SCL-lined tunnel which equates to about 80% of all the Phase 1 tunnelling.

The HS2 team developed a TBM tracker that enabled the tunnelling rates for all drives to be tabulated. There are many factors that influence TBM rates so it was noted that a direct comparison between each drive shouldn’t necessarily be made but it provides interesting data nonetheless.

The rates provided are rolling averages and it was noted that each drive was not far away from achieving the programmed progress rates they set out to achieve.

Similarly, TBM utilisation was presented with the same caveat that there are many different influences on when a TBM is available or not. A graph was presented that showed a range of TBM availability between just over 30% and just over 80%.

The HS2 tunnels team has also tracked volume loss. Generally speaking, the values are well below the contract value of 1%, with the highest values being for the Northolt Tunnel East drive (semi-open mode). The volume loss values for the Chiltern and Bromford tunnels, in Chalk and Mudstone, were generally too small to measure except for a small spike towards the north end of the Chilterns.

FUTURE TUNNELLING

Euston tunnels

The Euston tunnels will be launched in Q1 2026 from the stub tunnels at OOC. These tunnels are 7.55m-diameter and 7.3km-long with 18 CPs.

Both EPBM machines are now nearly fully assembled in the stub tunnels, which has been a challenge for the SCS team with access through the station box having to be coordinated with TBM and gantry delivery and availability. The roof is now completed over the top of the TBM assembly area so the only way in or out is now the ARLT.

The TBM drives do not end in a traditional reception chamber. Instead, they will finish inside an SCL chamber where the machines will be broken apart and disposed of through the ARLT.

It was explained that to achieve the level of train service required, there needs to be two Up Lines into Euston station for the single Down Line. This bifurcation needs to take place below ground due to alignment constraints, requiring the construction of the Euston Caverns.

Euston Caverns

The Down Line will finish just north of the Euston Cavern Shaft in an SCL reception chamber. This SCL tunnel will then be extended up to the open cut box outside Euston station. A large cavern will be driven from the shaft to facilitate the split of the Up Line into two SCL tunnels (Up Line East and Up Line West). There will be three caverns of reducing size (15m, 13m and 11m) and the TBM will be driven into the smallest one at the north end. Both TBMs will be dismantled in the SCL tunnels and returned through the ARLT.

To minimise ground movement, the outer lining of the largest cavern will be driven as series of interconnecting ‘stacked drifts’, SCL tunnels circa 2.5m ID driven from the crosscut adit. These are subsequently backfilled with concrete to form a large pipe-arch. When the arch is complete, the cavern is excavated in traditional heading bench and invert methodology.

This will be a critical piece of construction due to the shallow cover to the West Coast mainline above, and the delicate Network Rail retaining wall to the side. Construction of the cavern is due to commence around 2027.

Q&A

Following their well received presentation to the joint meeting of BTS and BGA, in October 2025, at the ICE in London, the speakers – Martyn Noak and Mark Lemmon – took questions from the packed audience.

Q: Noemi Barrington, Project Manager SCS: Can you share a few elements of innovation you have implemented across HS2 on the tunnelling?

A: HS2 has an innovation team that manages and shares innovation ideas across the programme. In tunnelling, we have done research into reduced carbon sprayed concrete lining (SCL), although that wasn’t progressed. We did, however, install low carbon concrete as temporary works in some of our sites. We have gone diesel-free at some construction sites, i.e., Canterbury Works in London Borough of Brent. EKFB have also been looking at innovation in the Green Tunnels, including efficiencies in installation.

Q: How do you manage uncertainty on a project like HS2?

A: While clients are sometimes seen as risk-averse, i.e., using new products or the approach to the way things are done, HS2 has a pool of Subject Matter Experts who can review new ideas that maybe depart from Standards or Specifications for consideration.

Q: Peter Townsend, Mott MacDonald: How have the HS2 GBRs performed on various packages against Contractor’s claims? Have you found any unexpected ground conditions not covered by the GBR?

A: The HS2 GBRs are part of the current live contract documents and were jointly drafted during the Integrated Project Team phase, by the Contractors and HS2. These GBRs are currently in use in the contracts to assess compensation events raised by the Contractors.

The GBRs have been used to assess some compensation events. There have been ground conditions encountered, raised by Contractors, that are not in the GBR but in these circumstances the contract has mechanisms to assess those claims by reviewing physical conditions information available to the Contractors at contract award, such as, for example, the Ground Investigation data and publicly available information.

Q: Phillipa Halton and James Lawrence, Geological Weathering: Why do you think the weathering grade in the Mercia Mudstone and Chalk was so highly varied?

A: The variation could be for a number of factors. In both the Mercia Mudstone and the Chalk Group the grading was taken from logging of borehole samples and there is a possibility the variation was due to spacing of boreholes, sample quality, and/or logging interpretation.

The main reason for variation in geology such as in the Mercia Mudstone and Chalk Group is the presence of jointing, where percolation of water leads to weathering and erosion of the rock near the joints. This is supported by the visibly deep weathering under the River Misbourne in the Chalk

Q: Charles Allen: Could the panel give an update on the status of the cost and programme of the HS2 Civil Engineering Works?

A: The HS2 Project is currently going through a reset on cost and programme, the conclusion will be announced to the DfT and Government at the end of 2025.

Q: Phil Quelch, Jacobs: If you were doing it again, as a Client, what would you do differently?

A: One aspect that could be approached differently would be the Ground Investigation. The way it was set out by HS2 was for an initial phase of GI be undertaken by HS2, which was spaced at least to the minimum requirements of Eurocode 7.

The plan was for the Contractors to come onboard and supplement that GI with their own GI, focusing on the areas of highest risk. We could have done more GI upfront, which might have helped with cost certainty, however, over the 200km+ alignment this would have been a significant cost, before the Project had Royal Assent.

Q: One of the Problems with London Clay reuse has been in embankments. How have you managed the known risk of moisture content issues and subsidence?

A: London Clay is not being used for any embankments on HS2. London Clay tunnelling spoil from Euston Tunnels and Northolt Tunnels will be placed sustainably but not in any engineering structures.

Q: Rosa Diaz, Mott MacDonald: Considering this project is the first time Variable Density (VD) TBM machines have been used in the UK, do you think they were the right choice? How do you think they performed?

A: They were used effectively in the Chiltern Tunnels, as shown in the geology presentation, particularly dealing with the high proportion of flint. Also, on the first 1km of the Chiltern Tunnels, under the M25, it was effective at minimising impact on third party assets. There is less evidence that the VD machine was utilised effectively at Bromford and Long Itchington Wood, but we are waiting for BBV to report on that via papers, etc.

Q: HS2 length in tunnels – What percentage of the journey to Birmingham is in tunnels?

A: 20%, so about 10-12minutes.

Q: What are your considerations for changes as a Client?

A: As a Client, HS2 is now tending to resist change as the knock-on effect to other aspects of the work, i.e., follow-on contracts, leads to significant costs. Proposed change is compared to Specification and Standards, reviewed by Subject Matter Experts. Acceptable change has to be safety critical or demonstrate good value.