An engineer could get nervous on metro station works in Singapore after last year's Nicoll Highway collapse. But on the new Harbour Front station for the Circle Line they are taking things steadily.
'Safety rather than speed is the first priority, ' says Poh Eng Cheun, senior project manager for client Singapore's Land Transport Authority (LTA).
The job is not one to take lightly.
It is not every day engineers dig away beneath a station and its two live railway lines, while keeping everything running and passengers relatively unaware of the work going on beneath them.
The project will connect the last section of Singapore's new Mass Rapid Transit Line into the fairly substantial metro network the island state has built up. Harbour Front, which opened last year, is the terminus for the North-East Line at the moment, and will be one of the changeover points on the system.
This requires the station to double in size, extending southwards by 140m, to create a new platform level with space for the new lines and passenger interconnections.
'It means essentially underpinning the station and then mining beneath its concourse - with the majority of the work underground, ' Poh says.
Work is further complicated by the fact that the station will sit beneath a large offi ce and retail development, Vivo City. Construction of the threestorey building has begun and much of the work will depend on the station being in position below.
'We have to create stub walls as part of the top slab, ' Poh says.
Beneath the mixed development the station is being built top down.
Undermining is a key part of the job. Fortunately this part of the island does not have marine clay layers found elsewhere and the ground is relatively firm, mainly mudstones and some moderately weathered siltstone. LTA senior inspector of works Edward Marsh adds that compressive strength is about 5Mpa.
Design is by German consultant Meinhardt. The method is to create a number of 9m high headings in the ground which are then used to build full height columns and separation walls. Once these are pressure grouted to support the concourse slab above, the intermediate 'pillars' of ground can be removed.
The initial drifts are taken step by step, using first a roughly 4.5m square top heading mined out with an excavator-mounted roadheader which can carve its way through the relatively soft rock quite quickly.
The excavation is done 1.2m at a time and then support goes in: steel posts and top sections to form a frame. Poh says: 'They have a jack at the top to preload the frame against the concourse slab above. We have tight limits on any movement that is allowed - it has to be kept to less than 15mm.' Alert levels are 10mm and 13mm.
Instrumentation has been set up all around the excavation and in the station above. 'We have building settlement markers and prisms, I think some 110, set up all around and these are monitored day and night, ' Poh says. 'And there are strain gauges on the frames.' He explains that the twice daily measurement routine needed to spot any movement quickly is all done by hand. 'We found that the tight spaces, corridors and small rooms at the back of the concourse meant an automated system would be very difficult to put in.' In a few critical areas where the ground was weaker and some initial movement was detected, four readings are done every 24 hours and the placing of steel frames doubled to every 0.6m.
Shotcrete and passive rock bolting is used for support between the steel sets, whose frames are linked at the base with a truss beam which is important for the second stage bench excavation.
For this another 4.5m depth is taken out, 1m at a time. Each advance removes ground from underneath the steel frames. The truss beam provides crucial support until a second part of the steel frame goes in below and the whole unit, now 9m high, is rejacked to ensure the loads are correct and capable of supporting the truss for the next frame along.
When the average 18m deep heading is complete, a fi rst part of the 2m thick base slab is cast 'or even 3m in some places, very heavily reinforced' and then temporary props can be installed, preloaded to 100% of the working load.
Then columns and dividing walls can be cast. These vary from a series of 2.5m cross-section columns, to complete walls 700mm thick. All of them fi nish 500mm short and are then pressure grouted to complete the connection, 'ensuring that all the voids are filled', Poh says.
This has the secondary effect of compensating for any flaws in the previous station's concourse slab, where some areas of badly compacted concrete have been discovered.
The drifts are at 11.4m centres and the 7m or so of remaining 'pillar' ground between them is excavated once the permanent support is completed.
More conventional top-down work is under way for the remaining part of the new station, which accounts for 70% of its volume and contributes most of the 29,000m 3 of spoil. The con actor is using a Hitachi sliding arm excavator for the muckout, a relatively new machine type which is becoming commonplace on Singapore projects, allowing better manoeuvrability than the long arm machines.
Finally, Poh says, there is bottomup work at either end of the station.
There is also a 50m section of train tunnel to make as advance work to allow the line tunnels to connect easily. This too is being mined. It is quite important as it is on the critical path and likely to take to the end of 2005 to complete, Poh says.
He admits work has been a little slow at times since it began in early 2003, but it has to fi nish early next year because the developer has an agreement with LTA for his property and wants to get on. 'I think we should do it, ' Poh says.
However, the trains will not run for some years yet because the station is part of Phases Four and Five of the overall Circle Line project, and construction work for this has only just begun (see box).