The Cliffsend Underpass in Kent, UK, is the world’s longest jacked structure at 126m. A deck with a single clear span of 20m is mounted on a precision slide track built in cramped tunnels, the piling of which was a difficult project on its own.

Project overview and timeline
The underpass project is part of the East Kent Access Phase Two highway project in Kent. “This forms the final link from the Thanet Way (A299) highway on the north Kent coastline from the end of the M2,” says Mark Pritchard, design manager for the VolkerFitzpatrick Hochtief contracting JV (see Figure 1,). “Which then turns into the A2 going down to the Dover area. This is the last ‘wrap around bit’ on the eastern coast. So it’s a strategic link for Kent,” adds Pritchard.

Kent County Council (KCC) awarded the contract for works to the JV with a scheduled start date of October 2009. Although the project value is listed as GBP 65M (USD 103.1M), Pritchard says that it was part of a GBP 87M contract from KCC that covered other aspects such as land acquisition and an interface between KCC and Network Rail.

The underpass accounts for a third of the value of the GBP 65M project and is the focus of these works.

The contract end date is September 2012, though it is expected that March will mark the end of work on site. Also involved is construction of 9km of new dual carriageway and a road bridge over rail.

Underpass situation
The village of Cliffsend hugs Pegwell Bay some 2km southwest of Ramsgate in Kent, UK, and is bisected by the Ramsgate to Ashford main railway line. A level crossing divides Foads Lane to the south of the line with Foads Hill to the north. The Cliffsend underpass was driven directly under this point where road meets rail. “Because of the geometry and topography of the site,” says Pritchard, “we had to drive the underpass through at a very acute angle to the crossing. Normally when going under existing rail, you would go square to the track to minimise distance. Here you can’t, so the zone stretches about 100m-long, though 25m wide.

“To make matters worse, although we generally had very competent chalk to work with, there was one band of Thanet Sands to deal with. And of course it passed directly below the worst point—the level crossing. […] ‘Brilliant’ was the initial thought, how on Earth could we prevent the railway from settling?”

Jacked deck solution
Jacobs was the internal designer for KCC and they worked with BAM Nuttal. An outline design was put to Network Rail, though the underpass tender was put out as a contractor design item. Jacobs’ solution was the indicative, but not the detailed final design. “We had jacked box experience prior to this job,” says Pritchard. “So we got on the tender list and engaged both Atkins and Jacked Structures for their particular expertise. We got them to come up with a solution to satisfy KCC, and more importantly Network Rail.

“The design we inherited was an old solution. Two 12x8m boxes jacked in side by side about 6m below the railway. They would probably be staggered so they can be done simultaneously. There were big concerns hanging over this indicative solution. Normally jacked boxes aren’t as long as this (final length 126m) structure and you would go square, and also not 6m below the tracks. We had to reposition a pumped foul sewer at that level by lifting it up to a higher point—fortunately the pumping chamber here could be made shallower by 1.5m.

“But the main concern was digging the two 12x8m holes with the fill on top and maintaining face and track stability. That was the fundamental problem of the entire job: how to install this massive structure without impacting rail, especially as rail services would not be halted or even slowed for us. Jacked Structures and Atkins came up with an alternative for us that incorporated Jacked Structures’ jacked deck.”

The deck
The jacked deck is a concept developed by James Thomson, CEO of Jacked Structures. Thomson tells T&TI that experience and burned fingers led to the belief that improvements could be made to standard box jacking. Thomson explains that the original idea they presented was to create a slide track inside jacked boxes at the side, and then jack in a single-span deck, with a cutting shield at the front, to form the roof of the structure.

The track would be set in the top jacked box at each side of the deck running the entire length of the underpass.

Thomson adds, “Friction is the absolute concern with jacking. Jacking a deck rather than two boxes means that all of the friction we must deal with comes from the ground above, and a bit to the sides, and quite an amount from the base as well—as when you are jacking a complete box—and also, there is a chance of two boxes interfacing and causing friction. That is very dangerous. The jacking load needed to get the deck in is also of course much lower than a full box.”

Pritchard adds that the 3 to 3.5m face of the deck is much less to manage than a 12m face. He continues, “It is much wider than two boxes, but it is divided into manageable segments by the shield at the front. And the vision was for it to sit in the boxes, which are still big, the size of a room, but manageable. They are concreted together and a slide track installed on top. So you have the walls, the roof and you excavate out the rest.

“We didn’t go with the box sides in the end, we went with 3m tunnels as it is a more known technology. Before we got the contract, we spoke with Network Rail to explain the benefits of the deck. We summed this up as: reduced face height, foundations in place so it isn’t going up or down and the already huge resistance, which would result in a phenomenal force on the bottom of a box, making the required jacking forces astonishing. The design, which included workshops with Atkins, Jacked Structures and Mammoet de Boer, took a total of 18 months from October 2009 through to the final installation.”

Jacking
Jacking 2,500t deck sections is not the smallest of challenges that can emerge from a project like Cliffsend. The specialist jacking subcontractor, Mammoet, developed a skidding system to move the sections into position in front of the thrust blocks, which are equipped with ground anchors by Keller. Ordinarily you would push against the ground behind the unit; this was not possible as a large area has been cleared to get the units in. “The skidding system lifts the cured unit up 50mm off the ground and moves it with a kind of cam system that puts pins in as it goes,” says Pritchard. “Relatively speaking it is not a huge weight, considering the 10,000t fill weight. The legs of the unit slide through a gap in each thrust block.

“The jacks on each block can provide a 7,000t jacking capacity and there are intermediate jacking stations between units one and two, two and three and three and four. These have another 7,000t capability giving a total 35,000t of jacking force. Interjacks were between all units except the last two. You push from the back for as far as you can, then you deploy the interjacks and jack one block off the next.

“The interjacks had a stroke capacity of 120mm, so we moved forward in increments of 100mm. A bigger stroke of 300 or 400mm is possible from the back. The steel edge of the shield cuts into the face so there is never a void at the front. At points the chalk was too hard so we had to dig up the front with the agreement of Atkins and Network Rail but the perimeter, the sides and the top did the final cut. We never cut above or to the sides so as to minimise overbreak.

“All of these solutions have been used before, but I don’t think they have ever all been used together.”

Tunnels and slide deck
The tunnels housing the 128m-long slide tracks were not specially designed in any way. The drives were executed by Murphy with EPBMs. Abutment walls were piled down from the tunnels. Four 50mm diameter piles of segmental rebar were sunk 14m down through each segment. Some 530 piles were completed in three months with a Keller-operated Clem 702 piling rig (and segmental augers) that, Pritchard says, had virtually 0mm clearance in the 3m diameter tunnels. “If I were doing this again,” Pritchard says, “I would bore the tunnels with an extra 0.5m diameter, just to save the pain of working in such a constricted environment, it is definitely worth the initial investment for when you get to the piling stage.

“The tunnels would be fully grouted as any permanent tunnel, and following the drilled holes and piles through the bottom, and the installation of the track, the top bit is cut off with peckers and drillers as the jacked deck advances.

“Threaded bolts support the slide track to a 0.5mm tolerance. The 60mm conventional bridge bearings we used were put to the limit. If you didn’t have the 0.5mm tolerance on the alignment of casting beam…well, it would be almost impossible to replace one. The bearings were tested to double load. Some showed signs of distress by the end of the jack, probably due to the guidance system pushing against the chalk at the side. In the open air the deck was moving all over the place; it moved 50mm to the side at one point. The bearings showing the most distress had a bit of sideways movement, but all were intact at the end. Some look like they are almost going to delaminate.”

Steel ropes and sheets
Thomson’s view that time spent on the concept pays dividends later is echoed by Pritchard on the issue that both tunnellers gave as the primary concern: friction. Pritchard says, “The anti-drag system looked exactly like the concept; it was precise. Basically with jacking a box or other structure, there is a risk of micro lateral movement of the soil at the interface between the deck and the overlying soil due to frictional forces leading to settlement and heave. To overcome this issue, you need to employ an anti-drag system.

“The most common form is a steel rope system. You anchor the ropes at one end, in our case the western wall head, and 19mm-diameter ropes lie all across the deck in two-diameter centres, so half the concrete is covered. Lubrication is easily pumped in and so friction is reduced. The problem we encountered was the sheer lengths of rope needed. Usually boxes are 30-50m long and we are more than twice that. There was no room to store the required rope on the drum, you just couldn’t do it.”

Thomson came up with the drag sheet solution at Cementation Trafalgar. The sheets predate the ropes and served the Cliffsend situation well as they were thinner.

A grade 550 steel was used. The design called for a 4mm thickness but this can’t be rolled very well, so two lots of 2mm steel were used. The force was seen as the limiting factor as the sheets could break. Corus-made sheets were bolted to the soffit with a standard grease pumped in. There was a backup system of bentonite delivered through a pipe network between the sheets but it was not needed.

Rail concerns
Network Rail’s position on the project was that there would be no interruption to regular services at all. This meant there could be no possessions. A 60mph (97kph) speed limit was imposed, but Network Rail was uncomfortable with this and a tighter 40mph (64kph) was imposed, though over a smaller section of the route so train times were not noticeably affected. The JV said its position was that although it was contracted by KCC, this was a Network Rail job and managing that side was the difficult concern of the job.

The road was shut to cars during works and was due to reopen a few days after T&TI visited the site.

On the concept
Thomson concludes, “A recent T&T poll showed that nearly one in three tunnellers believe that the use of NATM in soft ground is fundamentally unsafe. I share this view. NATM and SCL have been widely promoted for use in soft ground even in high-risk situations despite some notable failures and problems.” Thomson continues, “The main things that stand out with the jacked deck method, as far as I am concerned, are that it is broken down into a number of units so there is always a structural lining where we are working. Nothing will ever collapse and fall on the workers or cause a major settlement.

“You could theoretically have a slip at the face but you can control that. At worst some amount of material may be lost, but not huge amounts and it certainly wouldn’t be a disaster. This is in summary a very safe method of doing a difficult job: putting a large structure under an existing facility.

“If you are going under a railway, or maybe creating underground structures under somewhere like Oxford Street or Regent Street, safety must be the primary consideration. You can’t afford to make any mistakes or you will have an even greater disaster on your hands.

“The issue in part is that people just don’t yet realise the potential of what we can do with this technology. The jacked deck concept opens up the potential to install without surface disruption even longer and wider spans than those achieved at Cliffsend.”


The jacked deck units were moved by a ‘skid system’ that was developed by Mammoet de Boer Cliffsend in the Ramsgate area of east Kent The jacked structure with shield. It is the world’s longest at 126m The foot of the jacked deck rested on a slide track. As the structure advances, the crown and side of the tunnel is broken out to reveal the track Visible section of tunnel, deck foot and slide track as T&TI visits