Speakers Roy Brunsden of the Highways Agency, John Ropkins of John Ropkins Ltd, and Doug Allenby of Edmund Nuttall Ltd, were introduced by BTS meeting chairman, Roy Slocombe. Roy Brunsden opened the presentation. Prior to the works about to be described, Junction 15A consisted of a two lane bi-directional underpass beneath the M1 with roundabouts on each side accessing the slip roads to the motorway and the Swan Valley Development (Figure 1). As part of the original planning consent for the development, Swan Valley was committed to increase the capacity of the junction by providing a second underpass. This would enable a dual carriageway to run between the north and south roundabouts on the A43, thus reducing delays on the road between Northampton and Towcester.
In order to minimise disruption to the motorway’s daily 112,000 vehicles, the new underbridge was tunnelled under the M1 using a ‘box jacking’ technique; it was the first contract to be let by the Highways Agency using this method of construction.
A number of factors favoured the use of a jacking technique at this location as follows:
Contract procurement placed greater emphasis on quality rather than price, the split being 70:30. Partnering was also encouraged between the five organisations (client, Highways Agency (HA); contractor, Edmund Nuttall, URS Corporation, Halcrow Group and John Ropkins) involved in the scheme.
To cater for calculated minor settlement the motorway surface was resurfaced overnight in the zone to be affected by the works in advance of jacking. The whole jacking process took five weeks and the motorway remained open throughout with the surface being monitored as necessary. The project was completed and opened to traffic in spring this year.
The system
John Ropkins considered and illustrated the jacked box tunnelling process (Figure 2). An anti-drag system is fitted top and bottom to the box structure which is preceded by a short shield which, depending on size, may be compartmentalised. A berm is placed at the exiting point to prevent distortion of the embankment in this region. The box is supported on a jacking base and shoved forward by rams in either 150mm or 300mm increments. The anti-drag system is composed of steel ropes and is fed through slots in the shield.
Installations using the technique have been constructed at Didcot, Oxfordshire, West Thurrock, Essex, Silver Street and Lewisham, London and Dorney, Berkshire in the UK and at Boston, US. The emphasis throughout has been to control and minimise ground disturbance arising from the tunnelling process. Maximum cumulative settlement can normally be limited to between 30mm and 80mm depending on the nature of the ground and the size of the box.
A further aim has been to simplify the tunnelling operations as far as possible thereby optimising control of the process and minimising duration. On most projects the period during which surface movements occur is limited to the duration of the tunnelling operation, which is typically two to five weeks, depending on the size of the box.
The experience accumulated on the above projects provided the basis for the design and construction of the M1 Junction 15A Improvement Scheme. The key requirement from HA, was that the underbridge be installed without disrupting traffic on the motorway. To this end HA included in the contract charges that would be imposed on the contractor should closure of one or more traffic lanes or the hard shoulder of the motorway become necessary. Also imposed were limits on settlement and the distortion of the motorway surface.
The design developed to satisfy the HA requirements consisted of a single piece box of a length chosen to avoid the need for construction work to be carried out on the M1. To form this underbridge the box had to be installed at depths of as little as 1.6m in order to achieve an acceptable vertical alignment for the new A43 carriageway. A cellular shield was required to support the pulverised fly ash (PFA) and clay embankment fills and the underlying natural clay in the tunnel face. The estimated maximum jacking load of 4,500t was comfortably within the capacity of four standard jacking rigs. A sheet pile jacking pit was required to retain the working area and to provide sufficient reaction to the jacking forces. The top anti-drag system was restrained by tendons anchored to the rear of the jacking base and by the jacking pit side walls. A carefully controlled method of entering the tunnelling shield through the sheet pile headwall was devised using hydraulic props. Finally a system for remote monitoring of the motorway surface was developed.
In practice
The construction processes were then discussed by Doug Allenby. The original motorway embankment consisted of clay fill overlying natural boulder clay. After its opening in 1959 excessive settlement of the clay fill occurred over a 20 year period, which necessitated major reconstruction in the early 1980s. At that time a depth of some 3.5m of clay under each carriageway was removed and replaced with carefully laid and compacted PFA.
During the precontract site investigation nine cable tool boreholes were sunk together with dynamic probes and window sample holes to determine carriageway construction and the configuration of the PFA. Based on these, together with visual observations and learned advice, it was concluded that the PFA might range from being weakly cemented clay-spadeable with a stable face, through to well cemented, which could require machine excavation.
The clay fill material was typically a firm to stiff sandy clay with some fine gravel, and the underlying boulder clay was firm to stiff with occasional pockets of very gravelly clay or fine sand. The watertable was about 1.5m below the top of the jacking base.
The box was installed from a jacking pit on the east side of the M1 (Figure 3). The minimum cover was 1.6m at the edge of the hard shoulders, and the maximum cover was 2.2m under the central reservation; in addition the existing motorway drains had to be sealed and diverted. Great care was exercised in excavating, preparing and blinding the formation level for the jacking base and it was unfortunate that a boulder bed, typically 0.75m thick, occurred around this level: it is shear along this interface, which transfers the jacking forces from the box into the ground.
The level and plane of the top underbridge surface of the jacking base determines the alignment of the box during installation; a high order of dimensional accuracy is required. Reinforcement density was typically 127kg/m³ in the C40 concrete used in the jacking base.
The underside of the box had a steel plate soffit shutter profiled to suit the bottom anti-drag rope layout. The 45m long box for the underbridge was cast in three 15m long bays to minimise shrinkage: each bay comprised a floor pour and a side wall/roof pour. Reinforcement densities were typically 220kg/m³ and a C50 concrete pump mix with 35% of the weight of cement replaced by PFA was used. High quality proprietary EFCO plywood-faced shuttering panels were used and again allowable construction tolerances were tight. The dimensional accuracy of the box is checked immediately prior to its installation to establish trimming bead clearance and to cast guides down each side to direct the underbridge on its correct alignment during the initial stages.
Given the uncertainties in the ground investigations it was decided that as tunnelling was so close to the surface of the M1, hand excavation and close boarding must be available should difficulties arise. To meet these requirements an open face cellular shield with three working levels (each with seven compartments) was provided. The top level compartments were designed for hand excavation and the middle and lower levels for machine excavation as well as hand excavation. The top and middle level decks were positioned at the PFA/clay and clay/boulder clay interfaces respectively. Compartments were dimensional to facilitate timbering particularly in emergencies. The shield was constructed in reinforced concrete with steel plated cutting edges attached to the perimeter, working decks and vertical dividing walls. Trials were carried out prior to casting to determine the most suitable excavating equipment and Mecalac 360° excavators were chosen.
At the outset of installation the underbridge with shield was advanced to the jacking pit headwall. The shield’s cutting edges were progressively entered through the steel sheet piling by propping the piles with the face rams, then cutting them into discrete panels to be removed when the shield’s cutting edges were fully buried and the face supported.
Full face excavation then commenced by carefully excavating 300mm off the face and jacking the underbridge forward while maintaining a minimum cutting edge penetration of 450mm. Seven miners excavated the top level compartments and two Mecalac excavators, types 10MSX and 12MSX respectively in the middle and lower compartments. A dedicated Mecalac excavator type 14MXT equipped with a 30kW Webster Transverse Cutting Unit was used to excavate through the boulder bed and clear obstructions ahead of the invert cutting edge.
The settlement monitoring instrumentation was mounted on columns beside the motorway. The maximum settlement of the road surface recorded was 28mm. Four jacking rigs were installed giving a total thrust of 7,200t. Jacking loads were carefully calculated and the jacking operative was provided with a graph detailing the jacking thrust for each metre of advance.
Doug Allenby said that the natural reaction of any self-respecting box jack tunnel engineer when confronted with 1.6m cover under the M1 Motorway is to ram the shield hard into the face to prevent collapse; that explained why the actual jacking thrust exceeded the predictions during the early stages of installation. By carefully observing the settlement plots it was possible to trim closer to the cutting edges so reducing the jacking thrust, while simultaneously maintaining control of the face. The reason for the higher thrust right at the end comes from “classical tunnelling”; the breakthrough was made with the face only half excavated!
Construction of the piling works, jacking base, jacked underbridge and shield together with fitting out, took 35 weeks, working day-shift hours. Installation of the underbridge took one week through the sheet piled headwall and a further three weeks to its final position working 24hrs/day seven days per week. A further 12 weeks were spent completing the wing walls and finishing within the underbridge. The work of upgrading the adjacent roads and the two enlarged roundabouts proceeded concurrently and entailed considerable detailed traffic management and extensive use of nightime possessions. The underbridge was opened to traffic on 29 March 2003.
Related Files
Figure 1
Figure 2
Figure 3
