As part of its expansion programme, The Dorchester Collection acquired a property at 45 Park Lane, close to its prestigious hotel in London, with the plan of refurbishing the existing building to provide a new annex comprising a penthouse suite, 45 exclusive rooms and a restaurant and bar. The 45 Park Lane building required substantial foundation work, refurbishment and internal fitout to achieve the required standard of construction, and the contract was awarded to Kier Wallis in August 2008.

In the planning stages, The Dorchester Collection chose to use the facilities at The Dorchester to provide the building at 45 Park Lane with hot and cold water, power and communications. Following an ‘optioneering’ exercise open cut works were ruled out due to the substantial disruptions it would cause to the hotel’s operations and to the local road network. The decision was made to provide a service link between the two sites through a tunnel. A route passing under the hotel was considered but rejected because of the risk of damage to finishes. Instead a route along Deanery Street was adopted and discussions started with Thames Water, who maintain a sewer running above the planned line of the tunnel. Mott MacDonald provided a reference design for The Dorchester Service Link, prescribing 117m of 2.7m minimum i.d tunnel, and two 3.7m minimum i.d 12m deep shafts, with the contract being let as an ICE 2nd Edition Design and Construct form. The short length of the tunnel, the limited headroom at the shafts that would be excavated within basements, and the fact that the planned alignment was curved all lent to sprayed concrete lining (SCL) rather than segmental construction.

Part of the works would be completed within the basement of the 45 Park Lane building during the substantial main contract refurbishment works. While the main worksite would be located in the fullyoperational hotel basement. In an effort to mitigate these interface risks, and to use the existing relationship that had been formed through the main contract works, project managers Buro 4, the employer’s representative Mott MacDonald and The Dorchester Collection agreed to appoint Kier Wallis as the principal contractor.

In July 2009, following a competitive tender process, Joseph Gallagher, with Alan Auld Associates as its designer, was appointed as the specialist design and build sub-contractor to work “back-toback” with Kier Wallis to deliver the project. The submitted construction methodology included innovative solutions to deal with the difficult work site locations.

Both shafts were constrained by the existing basement arrangements and by the limited headroom present—minimum 2.2m in 45 Park Lane building and 2.0m in the Dorchester Hotel.

Design development and evolution
The site’s geological sequence comprised Made Ground overlaying Brickearth, with a layer of River Terrace Deposits leading into the London Clay—a typical geological sequence for the London Basin. Due to the presence of water bearing gravels directly beneath the basement floor it was decided that a form of ground treatment was required to reduce the permeability of the ground and increase the stability of the shaft excavation. A low viscosity, single shot Tam polyurethane grout was used. The grout reacted when it came into contact with water, to form a dense, solidified polyurethane foam.

The shaft internal diameters were dictated by the minimum space required for a safe method of construction. The depth of the shaft was derived from the requirement for one diameter of cover within the London Clay above the tunnel crown. This resulted in the shaft base level being 16m below ground level, which also provided a clearance of 6m to the invert of the Thames Water brick sewer in Deanery Street, which the tunnel alignment passed beneath.

The initial Mott MacDonald outline shaft design included precast concrete segments installed using underpinning methodology, with the lower section of the shaft formed using a primary and secondary SCL. During the detailed design stage Alan Auld Associates redesigned the shaft to be formed using steel liner plates. The SCL section was retained and designed to start 1m below the interface between the River Terrace Deposits and London Clay. The SCL was 150mm thick and reinforced with a single layer of A142 WWF. In the revised design the secondary SCL lining was replaced by a 200mm cast in-situ concrete reinforced with two layers of A252 WWF.

The primary lining in the 45 Park Lane Shaft was also revised to be formed entirely from steel segmental rings. No sprayed concrete work was prescribed in this shaft due to the difficulty of basement access during the extensive refurbishment works and to the severely limited space available for the necessary shotcrete equipment.

Following a detailed assessment of feasible tunnel construction methodologies, Joseph Gallagher opted for a primary SCL that allowed for mechanisation and settlement reduction. Mott MacDonald’s primary tunnel lining design was refined to an nreinforced, 150mm thick SCL (C30/37) designed to carry the full ground load, surcharge and groundwater load in the short term. Immediate application of an initial sealing layer of 50mm was prescribed to minimise deterioration of the ground due to its exposure and to provide safety to the tunnel operatives. To minimise settlement, the excavation sequence was designed as a 1m round top heading followed by a 1m invert advance staggered 1m behind. After 55m of tunnel excavation and following a detailed review of observed settlement it was jointly decided that the full face could be advanced in 1m rounds to speed up the closure of the primary lining ring. This also simplified tunnelling and increased the advance rate.

The secondary lining was constructed using a cast in-situ, steel reinforced concrete lining, which was cast outside hand-built segmental steel formwork. While it would be acceptable to use SCL from a structural performance point of view and while technically the rough finish would be acceptable for a services tunnel, it was considered that the steel formed wet cast lining would help provide the high quality product the client desired. The secondary lining was designed as a 200mm thick cast insitu concrete (C32/40) with two layers of A252 mesh and was designed against full groundwater load, surcharge, ground load in the long term and to meet fire and durability requirements.

Mott MacDonald provided an outline design prescribing a 120 year design life of the tunnel and both shafts. This was achieved by assigning an appropriately durable concrete mix design with a 40mm minimum cover to the reinforcing steel. All structural steelwork has a 40 year design life; so galvanised steel was used for the access ladders and landings.

Excavation and primary lining
The precast concrete segments originally proposed for the primary lining of the shafts were substituted by steel liner plates supplied by American Commercial. Each 406mm deep ring was comprised of 13 no. 8mm thick steel segments that proved to be versatile, small, light and easy to handle during shaft sinking. The headroom in the basement was not sufficient to use standard hydraulic jacks nor manoeuvre heavy kentledge to sink the shaft as a caisson; so conventional underpinning was chosen. Prior to commencement, the River Terrace Gravels underlying the basement slabs were consolidated by injection of polyurethane grouts via lances driven into the ground. Although it was later found that penetration of the polyurethane resin into the dense sandy gravels was limited, both shafts were carefully and successfully sunk in single ring increments using a degree of timber support where necessary to minimise overbreak.

The bottom half of The Dorchester shaft within the London Clay was lined with 150mm of fibreless shotcrete, reinforced with a single layer of A142 WWF. The SCL allowed the integration of the tunnel eye with the lining, avoiding the need to breakout through the steel rings, potentially restricting access further by the requirement for temporary propping in the shaft. Both shafts were sunk during a 12 hour dayshift at an average rate of 1m per day.

Excavation of the 117m long 3.1m i.d. tunnel from The Dorchester shaft began on 14 January 2010 and reached the shaft at 45 Park Lane on 5 March, with tunnelling production ranging between 2-5m advance per day for fives days per week. Initially an 8T Butor electric-hydraulic excavator was trialled but proved to be too large to manoeuvre while mucking in the small confines of the tunnel. Following removal of the Butor a conventional diesel-powered 1.5 T Taekuchi excavator was lowered into the shaft and almost immediately a marked increase followed in production.

As is common when tunnelling in such restricted work areas it was found that the maximum tunnelling advance rate was limited by the mucking and concrete delivery operations.

Excavated spoil was loaded at the tunnel face into a 1m3 capacity bottomdischarging skip on 18in (457mm) gauge rails which, when fully laden, were pushed manually down grade to the “Y” branch turnout at the pit’s bottom and hoisted to basement level by the 3.2T SWL runway beam hoist. Clearance below the hook to the landing area floor was tight at 1.4m, which meant that the skip had to be rolled out from under the beam before it could be lifted and emptied into the muckbay using a JCB 520 telehandler.

The capacity of the muck bay in the basement garage of The Dorchester was approximately 57T. During dayshift the spoil was carried up the spiral ramp by the telehandler to an 8yd3 (6.1m3) skip situated adjacent to the hotel kitchens at the rear of the building in Deanery Street, while tunnel excavation continued.

Readymix concrete for the SCL was delivered in loads of 4-5m3, which suited tunnelling production and the storage capacity of the KBM6 Muhlhauser remixer in the basement adjacent to the shaft. Each load was received at a small hoarded area on the pavement outside the hotel in Deanery Street and was discharged into the remixer via a small chute through a hole in one of the pavement lights adjacent to the rear Ballroom entrance.

Waterproofing
The client’s requirement was for leak free, un-drained tunnel and shafts that will use the benefits of excavation in the low permeability London Clay and minimise the influence on the water table. Over the last 10 years, throughout the tunnelling industry there has been a shift away from sheet systems to the use of sprayed membranes for waterproofing. There are two types of sprayed waterproof membrane, reactive resin based systems and non resin based membranes, the later of which was used in The Dorchester tunnel.

The non-resin based system is an ethylene vinyl acetate powder based product, where water is added during the spraying process. Before the application of the BASF Masterseal 345 sprayed waterproofing membrane a regulating layer was used to ensure the receiving surface was not too rough, and hence reduce the consumption of the membrane material.

On average a 4-5mm thick layer was applied by hand-spraying with an application rate of approximately 50m2 per hour. In both shafts the cast in-situ secondary lining concrete contained a hydrophobic, pore-blocking Pudlo admixture to provide watertight concrete. Joints in the shaft lining, in the waterproofing system itself and between the tunnel and shaft linings were identified as weak points and therefore extra attention was required to seal them against water ingress. Tam re-injectable grout tubes were used in all construction joints.

Testing and quality control
SCL testing consisted of the flow test, Hilti gun penetration test to check the early age strength development, cubes and cores to check the compressive strength. The sprayed waterproofing membrane was checked during the application with a simple depth gauge. After the application a full visual inspection took place and was recorded. Test patches were taken by peeling a 50mm x 50mm patch off of the membrane from the SCL surface, to check for the required minimum 3mm thickness.

Shaft and Tunnel Secondary Lining
Construction of the tunnel’s secondary lining commenced on 12 March 2010 and completed on 30 April. The 2.7m i.d cast insitu tunnel secondary lining was formed behind hand built segmental steel formwork fabricated by Specialist Formwork. Each ring was tapered to suit the two 30m radius bends in the tunnel and comprised 10 segments. The weight of each segment was kept to a minimum to allow safe manual erection and dismantling and then striking after each pour. The tunnel lining concrete pours were generally 20m3 in volume and production regularly achieved the planned 8m of tunnel lining per day.

Settlement
The design necessitated the use of sprayed concrete for the primary lining to minimise face loss, settlement and potential damage to adjacent buildings along Deanery Street adjacent to The Dorchester.

A series of monitoring points were installed on the road and pavement along Deanery Street prior to construction. These were measured by the contractor using precise levelling techniques at night when traffic was at a minimum.

Pre- and post- construction surveys of road and adjacent buildings have been undertaken by Plowman Craven and have continued on a monthly basis in order to confirm the cessation of any further tunnel induced settlement. During the project the volume loss was found to be a consistent 1 per cent and the change of excavation sequence was found to have no significant impact.


Schematic view of The Dorchester Service Link showing the Thames Water sewer above the tunnel and the London Ring Main below The working area at the top of The Dorchester shaft The Dorchester shaft under construction using steel liner plates Excavation and mucking during tunnel excavation: 1.5T excavator and 1T hand-pushed skip on rails Looking into the 45 Park Lane shaft following Masterseal application The finished tunnel lining