Most of the bids put in for the 0resund immersed tube tunnel section of the 0resund Fixed Link were based on conventional immersed tube technology. Large steel or concrete elements would be constructed section by section in a huge open-air casting basin excavated on the shore of a convenient estuary or sheltered bay. When all elements were complete and made ready, the casting basin would be flooded and the elements towed out and sunk into position.
This was essentially the method chosen on projects like the Medway Tunnel in southern England. Adapting this technology for 0resund was one option - but there were many obvious problems.
'The main one was the sheer size of the project,' says Laing project director Mike Hooper. 'We needed a casting basin that could produce more than 3km of elements efficiently. A small basin that was used several times was one alternative - but that would have meant big gaps in production, with associated labour costs.'
Adapting an existing dry-dock somewhere in Europe was another option. After all, as Laing tender planner Andy Willard points out, once an element is afloat 'it's a sea-going vessel' and it can be towed remarkably long distances at quite low cost. 'A dozen different scenarios were on the table at one time or another, some of them quite innovative,' he adds.
'And even as we came up to tender submission in 1995 we still had two serious alternatives up our sleeves.'
One was a traditional casting basin, to be used three times in all. The second was originally triggered by the realisation that the projected contract period was remarkably tight, given the likely extremes of the Scandinavian winter.
'I've seen the sea frozen for 200m out from the shore,' Willard reports. 'Producing high quality concrete out of doors was not an attractive prospect.'
Dumez-GTM, with its long experience of producing concrete bridge segments, and of incremental launching of large bridge decks, believed its expertise could be adapted to the 0resund project.
But it was only six weeks before tender submission that 0TC committed to the GTM solution'.
Three key decisions had been taken, all based on existing technology and experience but never combined together on a project of this nature before. The tunnel would be formed by 20 concrete elements, each 175m long and made up of eight precast concrete segments. These segments would be produced at ground level under cover in climate-controlled conditions in a high-technology state-of-the-art factory. And each 2,700m3 segment would be cast in a single pour, without the usual cooling pipes designed to minimise differential thermal expansion as the cement hydrated.
Secondly, the segments would be slid out of the casting shed and post- tensioned together to form the element, all still above sea level. Then the whole 57,000t mass of concrete would be slid in one piece into the ground level section of a two-stage float-out basin surrounded by high bunds. (see box)
The basin would be pumped full of water until the element, now fitted with watertight bulkheads and buoyancy tanks, began to float. It would then be moved to the deeper section of the basin, the water would be lowered back to sea level, and the element could then be towed out to sea.
'The client could see lots of benefits in these proposals,' Hooper says.
Not least of these was the potential quality improvements and lower risks associated with concrete production in factory conditions. '0TC did consider using a moving weatherproof canopy which travelled along the casting yard,' says Marshall. 'But the greater quality and better working conditions of a proper factory tipped the balance.'
Given the extremely high quality specification for the tunnel units, the attraction to the contractor was at least as high as for the client. Normally the capital cost of such a facility would rule it out - but such was the sheer scale of the 0resund project that this time the numbers added up and both client and contractor could share in the benefits.
Some of the proposals were untested on the 0resund scale. These included the sliding of complete elements, and there were some anxious moments when the first segment was poured in November 1996. But by the time the final closure joint was made in February 1999 the only real problem had come when a watertight bulkhead failed on element number 13.
Such was the flexibility of the system that no serious delay occurred. Finishing works are now well advanced, and final handover is scheduled for June next year.