One traditional advantage a steel immersed tube tunnel has over the concrete alternative is that a simple pre-placed screeded layer of crushed rock or gravel is all that is normally needed.
Stiffer and more brittle concrete elements have normally been fitted with temporary legs which support them above the bed of the trench while sand is pumped in below through ducts to form the permanent bed.
These legs and ducts cause extra congestion in the slab reinforcing cage and put up the cost.
But a gravel bed seemed out of the question, unless it could be pre- placed with unprecedented accuracy. Only when Boskalis Westminster engineers proposed a technique they had developed for placing the gravel foundations to a quay wall in Singapore did the planning team reconsider.
Central to this new technique was a specially developed multi-purpose pontoon mounting a hydraulically 'heave compensated' gravel spreader unit. Using satellite navigation and sonar Boskalis Westminster claimed it could place a gravel bed to an overall accuracy of better than +/-25mm - without subsequent screeding.
This also promised an answer to the fedtemog problem, which had caused a lot of concern.
Fedtemog, a tangled mixture of seaweed and silt, can spread over the Baltic seabed at astonishing speed. The original nightmare scenario was to discover a substantial layer of soft fedtemog under the tunnel elements before sand fill began. A gravel bed, however, could be placed only a short time before an element was submerged, immediately after the bed of the trench had been vacuumed clear of all unwanted deposits.
To give the bed placing team the best possible chance of achieving the required accuracy it was decided to form the bed of a series of 500mm to 1,400mm deep bunds 1,650mm wide at the top, with 1,000mm valleys between. This had three main benefits, according to Symonds immersed tube design team leader John Busby.
'Obviously it eliminates the risk of adjacent passes of the laying operation overlapping each other and causing high spots. The valleys also act as an escape route for accumulated silt on top of the bunds, which would be squeezed out when the element landed.
'But the main advantage is that the bunds make the foundation 'softer', allowing local high spots to deform more readily as the element beds down.'
Crushed Swedish granite of a nominal 60mm size was the chosen material for the bed. In practice an accuracy of close to +/-15mm was achieved.