BACKFILL PLACEMENT has emerged as the chief focus of investigations into a major tunnel collapse on the Chiltern Railway line near Gerrards Cross in Berkshire, UK, last month.
The tunnel, being constructed as part of a ú20.3M ($38M) 'air rights' project for supermarket giant Tesco, partially collapsed at 7.30pm on Thursday 30 June, blocking the railway line.
The alarm was sounded by the driver of a Chiltern line commuter service about to enter the tunnel.
Track owner Network Rail operations director Robin Gisby told NCEI that the backfill operation is its primary cause of concern.
'Something about the method of construction on this section is not right, ' said Gisby.
A 320m long section of Victorian railway cutting adjoining Gerrards Cross station was being 'tunnellised' to provide a town centre site for the Tesco store. The operation involved erecting a tunnel of precast concrete segments over the twin track London-Birmingham rail line and backfi ling to original ground level.
Work was being done under a design and build contract by Jackson Civil Engineering with consultant White Young Green and precast arch supplier the Reinforced Earth Company.
Jackson chief executive Richard Neall insisted that backfi ll operations were 'in line with what the design allowed.
'There was a strict loading regime in place, ' he said.
White Young Green chief executive John Purvis would only say that 'efforts are focused on the clean up operation' and would not comment on the contract at this stage.
The tunnel works as a three pin arch, composed of half span, 2m wide segments leaning against one another under self weight.
It was inherently flexible, allowing 200mm of vertical movement and was reliant on surrounding earth pressure provided by backfill for much of its long term stability.
The backfilling operation had reached formation level over approximately half of the site, allowing steelwork for the supermarket to be erected.
Fill operations were advancing towards the south portal at Marsham Lane bridge.
The 30m long collapsed section is around 60m north of Marsham Lane, with between a half and two thirds of the fill operation completed.
It is thought that an imbalance in the placement and compaction of fill either side of the tunnel combined with a surcharge of fi l over the tunnel's crown triggered collapse.
Geotechnical engineers expressed surprise at the apparent height differential between fill to the sides of the tunnel, and especially at the quantity of fill placed above the tunnel in the collapse zone.
'The collapse was probably due to too much load on the crown of the arch and not enough fi l on the sides, ' said the head of one specialist consulting firm.
'Units [appear to] have failed by creating a hinge in the concrete section. The hinge has rotated downwards, which is consistent with the crown of the arch moving downwards under excessive load and/or the sides of the arch moving outwards under too little lateral restraint.
'This problem would have been made worse if there was a signifi ant difference in the level of the fi on the two sides, ' he added.
A height difference would have resulted in uneven earth pressure on the walls of the tunnel, causing it to deform asymmetrically, or sway.
'At the section that is still standing the central hinge has bent down indicating an imbalance between the vertical and horizontal loads and a failure by outward spreading of the arch, ' the head of a specialist consultancy said.
He suggested that monsoonlike rainfall that hit Gerrards Cross the day before the collapse could have aggravated the situation.
'It is possible that rainfall increased the vertical load on the tunnel. At the same time, if the fill at the sides is not completely free draining, it could have led to a reduction in the passive pressure available from the fill on the sides.' A designer of similar precast concrete arch systems added that there was a relatively short transition zone between the area of the site that had been fully fi led and that where the collapse happened.
'Movement around the crown can be quite substantial - as much as 200mm up as you fill around the sides and then 200mm back down as you fill over the top. But that's not a problem if it's properly controlled.
'You could trigger failure if you go straight from full depth fil over one section straight to no fil; that is, a section that's fully fl exed next to one that's unflexed, ' he said.
To minimise the risk of differential defl ction over the length of the tunnel it is normal to raise fil levels evenly along the whole tunnel length, or to provide a gradual, long, ramped transition, he claimed.
Network Rail and Health & Safety Executive investigators expect it to take a month and a half to determine the cause of collapse. 'We're focusing on all aspects of design and construction, ' said a HSE spokesman.