The A3 Hindhead project located in Surrey, UK is a 6.7km dual carriageway trunk road that includes a 1.83km tunnel being delivered under a Highways Agency Early Contractor Involvement (ECI) contract.

During the preliminary design phases both TBM and Sprayed Concrete Lining (SCL) methods were considered for construction, and a comparison between the designs is included in the article. The differences in construction methodology including planning, and program are also discussed.

Project description

The A3 Hindhead project is one of the schemes in the UK Government’s Targeted Programme of Trunk Road Improvements. The project will complete the dual carriageway link between London and Portsmouth and remove a major source of congestion, particularly around the A3/A287 traffic signal controlled crossroads (figure 1).

The project will deliver quicker, more reliable journeys on a safer road, and remove much of the present peak time “rat-running” traffic from unsuitable country roads around Hindhead. The centre of Hindhead will be freed from the daily gridlock that blights the area, with the result that the project will bring benefits to road users, local residents, and the highly prized local environment. Construction of the project commenced in January 2007, and the tunnelling is planned to commence in January 2008. The tunnel is scheduled to open in 2011, with the scheme completed in March 2012.

Planning phase

The A3 Hindhead project has been planned since 1983 when it was included in the Government’s Trunk Road program. The challenge for this project was balancing environmental impact with user economic benefits. Striking the right balance between these competing issues has taken over 20 years to resolve. Almost all the scheme lies within an ‘Area of Outstanding Natural Beauty’ (AONB) with Site of Special Scientific Interest (SSSI) that are part of the Wealden Heaths SPA – an EU Designation that prohibits development except for “no choice” and national economic or safety considerations. The route passes through National Trust owned land which is classified as “in-alienable”, that is, if the National Trust so chooses, it can only be compulsorily purchased with the approval of the UK Parliament.

In the 19 years since the project was first mooted, and the commencement of the current phase of project development in October 2002 there has been a sea change in planning and approval thinking whereby economic benefits originally had primary importance and now environmental damage mitigation is uppermost. The key to a successful planning process in this case was to provide a tunnel under the environmentally sensitive areas, and to close the old A3 thereby reuniting sections of the SSSI/SPA. Other environmental benefits included the removal of intrusion and severance within the AONB, and the removal of air, noise and light pollution.


The geology of the Hindhead area comprises a sequence of fine grained sedimentary deposits laid down during the Lower Cretaceous period in near shore transgressive marine conditions on the margins of the subsiding Weald Basin. The tunnel is within the Hythe Beds – a 90m thick sequence within the Lower Greensand Series formation.

The tunnel at the south passes through Upper Hythe A and B which are similar units with an increasing number of sandstone bands with depth, described as ‘medium dense thinly bedded and thinly laminated, clean to silty and clayey fine and medium sand with subordinate weak to strong sandstone, cherty sandstone and chert’

The majority of the tunnel passes through the more competent Upper Hythe C and D, and Lower Hythe A units, described as ‘Weak, locally very weak to moderately strong, slightly clayey fine to medium sandstone with occasional thin beds of clayey/silty fine sand’. The sandstone within Upper Hythe C/D and Lower Hythe A has typical UCS values of between 2 and 5MPa and is heavily fractured with six joint sets including the sub-horizontal bedding with mean fracture centres varying between 190 and 815mm. The tunnel is above the historically observed water table, with the maximum predicted water table exceeding the invert level in only one location by a depth of less than 1m (figure 2).


The horizontal alignment for the tunnel was determined based on road design considerations and environmental constraints resulting in a reverse curve through the tunnel with a minimum radius of 1050m.

The vertical alignment was determined based on geological constraints with the desire to minimise the length of tunnel through the sand at the southern end, to keep the tunnel above the water table. The tunnel passes beneath the Devil‘s Punch Bowl which is a re-entrant, primarily spring-sapped valley system with erosion feeding backwards from the Hythe Bed/Atherfield Clay interface at the valley base. The crossing of the punchbowl provides a cover constraint to the tunnel, and the cover changes rapidly from around a minimum of 16m to a maximum of 58m within a horizontal distance of 130m.

Cross section

The Hindhead tunnel layout comprises twin 2-lane bores with cross passages at 100m nominal centres. Each bore has two 3.65m lanes, with full batter curbs and 1.2m wide verges on each side of the tunnel. The verge width is sufficient to allow for sight-lines due to the horizontal curvature of the tunnel, to accommodate electrical services and also to provide wheelchair access to the cross passages and emergency points at 100m nominal centres along the tunnel (figure 3).

The vertical traffic gauge provided is 5.03m with an additional clearance of 250mm to the Equipment Gauge to allow for flapping tarpaulins and other transitory gauge infringements. These requirements result in a horseshoe shaped tunnel structure with a 10.6m i.d and an excavated diameter of 11.6m, with a face area of 96m2.

Tunnel excavation and support

The presence of the sand layers, in one location up to 2m thick, led to the selection of the Sprayed Concrete Liner (SCL) method whereby shotcrete is sprayed at the face following each excavation advance. Standard hard rock tunnel support techniques such as pattern bolting were not considered suitable due to the sand layers and the very low bond stress negatively impacting the effectiveness of rock reinforcement.

Four basic support types have been designed for the standard tunnel cross sections with minor variations required at cross passage junctions and Emergency Point niches. There is one main support type for the sandstone section, with three support types covering the section through sand and the transition from sand to sandstone.

Excavation and support types are specified based on tunnel chainage and have been designed to cover all expected ground conditions. It is not proposed that support types be selected based on geological inspection. The Hythe beds have 6 joint sets and an average joint spacing of less than 200mm. The UCS values are typically 2-5MPa. This material is expected to act as a continuum, and given the heavily fractured nature of the material, and presence of sand layers, meaningful variations in rock quality are expected to be difficult to detect. A suite of ‘additional measures’ discussed below have been designed to account for any local stability issues. Geological inspection and mapping of the open excavation, and monitoring results will be used to determine the advance length that may vary between 1m and 2m, with 1m advances specified at critical locations such as beneath surface structures and roads.

Support Type 1

At the northern end of the tunnel (Chainage 3120 to 4650), excavation is in rock (UHC/D, LHA) and Support Type 1 is specified throughout. The tunnel is generally excavated with a full-face heading followed at a distance by the bench excavation. Due to the generally stable nature of the ground and tunnel location above the water table, a closed invert is not required and the horse shoe shaped primary lining is supported on elephants feet (figure 4).

In addition to the sequences and support requirements for each of the support types, contingency ‘additional support measures’ have been specified. The requirement for the contingency measures will be triggered by geological inspection and mapping, and monitoring results, and will include:

Spiling – Self drilling GRP tubular spiles will be installed when ground conditions result in excessive overbreak, or instability in the crown. Spiling is detailed as mandatory for approximately 30% of the Support Type 1 excavation in areas of known potential crown overbreak such as where the tunnel crosses Fullers Earth bands, and where a 2m thick layer of sand and shattered rock intersects the crown.

Additional face support measures – Geological inspections will inform the selection of additional face support measures such as sealing layers, or face dowels.

Probe Drilling – Probe holes drilled ahead of the face have been specified for the entire length of Support Type 1 to relieve any potential hydrostatic pressure from perch water ahead of the face. If water is detected in the probe holes then additional holes will be drilled to drain any water

Grout Stabilisation – Microfine cement or chemical grouts will be used to stabilise running sand bands, or other local areas of instability caused by sandy layers. It is proposed to conduct site trials prior to construction to determine the optimum method and materials for stabilisation of sandy materials

Invert strut or rib bolts – Convergence monitoring will be undertaken during construction with a three stage trigger limit system. Unplanned convergence resulting from worse than expected ground conditions will be addressed by installation of an invert strut at bench level.

Support Types 2 – 4

At the South end of the tunnel (Chainage 2880 to 3120 (m001)), excavation is in sand (UHA/B) and support types 2, 3 and 4 are specified. The excavation will be carried out on dayshift only due to constraints on working hours and will be made stable with the use of a steel pipe umbrella and face dowels. The pipe canopy comprises 12m long, 114mm diameter tubes at 400mm centres with an overlap of 4-5m. The advance length is a maximum of 1m for these support types.

Support Type 2 has sandy material (UHA/B) in the heading only, with the heading elephants feet supported on the sandstone material (UHC/D). This means self-drilling GRP face dowels are required in the heading only, and the heading can advance ahead of the bench. The face dowels are 12m long with a 4m overlap and are installed with the same drill jumbo used to install the pipe canopy.

Support Type 3 has a full face of sandy material with the elephants feet of the bench supported on the sandstone material. As the heading elephants feet are not supported on sound material, the heading must be advanced with the bench, with a 2m separation provided to maintain face stability. Face dowels are required for both the heading and the bench (figure 5).

Support Type 4 has a full face of sandy material that extends below the tunnel, and therefore a closed invert is required. The heading must be advanced with the bench, with the invert closed a maximum of 6m behind the face.

Design and methodology

The excavation sequences outlined above are designed to control strains in the ground so that as much as possible of the ground load bearing capacity is used and the strains are maintained at levels that minimise yielding.

A principal innovation with the support measures is the design of primary lining as permanent. This is possible due to a number of advances in tunnelling technology in recent years. Firstly, non-alkaline accelerators are now available with no loss in shotcrete strength with time. A recent innovation is the use of 3-D scanning survey equipment that provides excellent shape control for both excavation and spraying, and allows shotcrete lined tunnels to be constructed without lattice girders. Historically the inclusion of lattice girders meant the primary lining had to be considered temporary due to the corrosion potential of the steel lattice girder within the primary lining. Spiling is envisioned in several locations due to adverse soil layers. This will be carried out with self-drilling Glass Reinforced Plastic (GRP) dowels, again with no adverse durability issues. The sprayed concrete will be reinforced with steel fibres as is required for safe installation, however the design does not rely on the flexural capacity of the steel fibres, and the lining is designed as plain concrete. This is possible due to the curved shape of the section with all moments resisted by axial forces within the lining.

Chemical hazard alert notices

Towards the end of the preliminary design period, the UK Health and Safety Executive (HSE) issued three Chemical Hazard Alert Notices (CHAN’s) relating to a reduction in the allowable exposure levels of Respirable Crystalline Silica (RCS) and NOx during construction (Table 1).

All project development up to that stage had been based on excavation of the tunnel with diesel equipment. It was not possible to achieve these proposed limits with the diesel based construction methodology. The lower CHAN limits produced conflicting demands, as the lower NOx limits require increased ventilation, but increased ventilation produces more dust causing problems with respirable silica.

This resulted in the consideration of an alternative method of construction, and a preliminary design prepared for a tunnel constructed using an EPBM.

One of the benefits of utilising a TBM was that it was no longer necessary for the vertical alignment to follow the optimal material for ground support. This meant that the tunnel could fall from south to north, allowing the deletion of the low-point sump. Another advantage of TBM construction was that a single-pass lining could be utilised. A fully gasketed segmental lining fire hardened with polypropylene fibres was proposed.

The main disadvantages with the TBM options were the openings required at cross passage junctions and for Emergency Points. The junctions are significantly more expensive for a TBM tunnel, and where the benefit cost ratio (BCR) for the Emergency Points was greater than 1 for the SCL tunnel, the cost of the opening for a TBM tunnel resulted in BCR < 1, however once a safety measure is proposed to the Emergency Services it is very difficult to withdraw it at a later date.


There was a small benefit to the program for the TBM option however this was minimal due to the fact that the SCL tunnel is proposed to be excavated concurrently from 4 faces, whereas the TBM had to excavate the first bore, and then be turned to excavate the second bore. The tunnel length of 1.8km was insufficient to give the TBM option a significant program advantage. A tunnel of 2.5km length would have been required to give a program advantage for the TBM of around 12 months.

An interim guidance note “Occupational Exposure to Nitrogen Monoxide in a Tunnel Environment” based on the ALARP principals applied to NO exposure limits prepared by the British Tunnelling Society (2006), was used to develop a modified SCL construction methodology such that NO limits of 3-5ppm and the revised CHAN limits for NO2 and RCS could be achieved. Additional measures include increased ventilation, more electric plant including an additional face conveyor and diesel plant conforming to the Stage III emissions standards in the Non-road Diesel Engines Directive. The modified SCL solution was still more economic than the TBM alternative.


The development of the planning and design of the A3 Hindhead project has been outlined in this paper. The conclusions drawn are:

  • Projects through environmentally sensitive areas need to provide special mitigation measures in order to achieve planning consent
  • The design of the tunnel support includes a permanent primary lining possible due to; availability of non-alkaline accelerators; use of 3-D laser scanning survey technology for control of robotic spraying equipment to achieve tight shape tolerances avoiding the need to use lattice girders; and the use of self-drilling GRP spiles in areas of poor ground
  • Alternative designs have been prepared for an SCL tunnel and a TBM tunnel with the following conclusions:
  • Below a length of 2.5km the TBM option had insufficient program and cost advantages
  • The TBM option had a lower assessed commercial risk due to the insensitivity to adverse ground conditions; however this was insufficient to overcome the increase in cost
  • A modified SCL methodology that using ALARP principles for airborne pollutants in accordance with the BTS best practice document was developed


This article was based on a paper prepared with joint authors Tony Rock of Mott MacDonald and Paul Hoyland of Balfour Beatty. Thanks also to Paul Arnold of Highways Agency for his constructive comments.

Project location Figure 1 – Project location Geological long section and vertical alignment Figure 2 – Geological long section and vertical alignment Typical tunnel cross section Figure 3 – Typical tunnel cross section Support Type 1 in Sandstone section Figure 4 – Support Type 1 in Sandstone section Support Type 3 typical for the sand section Figure 5 – Support Type 3 typical for the sand section Table 1: Details of Chan’s