The AU$880M (US$692M), 12.5km long Chatswood to Epping underground rail link currently under construction in Sydney, Australia, is one of the largest publicly funded schemes in New South Wales’ history. In underground construction terms, not only is the sheer scale of the project impressive, with its 25km of large diameter TBM driven running tunnel, but also the speed at which the tunnels are being bored. Currently, the first of two 7.2m diameter Robbins TBMs is carving through the alignment’s Hawkesbury Sandstone at rates of up to 200m per week!
Also notable are the geotechnical challenges facing the teams designing and constructing the vast caverns needed for the line’s three new underground stations.
When complete by mid 2008, the scheme will connect the city’s second business centre in Chatswood and the growing suburb of Epping via twin single-track tunnels.
The project and the people
Project construction is being carried out by the Thiess/Hochtief JV (THJV), which was awarded the US$667M design build contract in July 2002. Design is being provided by GHD and Parsons Brinckerhoff (PB), who were also responsible for the tender design.
Preparation work began in November 2002 with all major activity based around the project’s central M2 worksite. Being almost at the alignment’s mid point made it an ideal location to base the tunnelling operations. The site houses the 20m x 25m x 30m deep TBM launch/turning shaft, the concrete batching plant, and spoil operations.
US Tariffs are shifting - will you react or anticipate?
Don’t let policy changes catch you off guard. Stay proactive with real-time data and expert analysis.
By GlobalDataHere the 12.5km of twin tube tunnel is split roughly in two, starting with the 6.7km long drives west to Epping followed by the 5.8km long easterly drives to Chatswood. The schedule anticipates the Epping drives will be complete in July 2004, when the two TBMs will be partially disassembled and brought back through the tunnels while the cutterheads are transported above ground back to the M2 worksite. Upon “pull back” to the M2 worksite, the TBMs will be turned through 180° and launched towards Chatswood in September 2004. Tunnel boring is on a 24 hour a day basis with the TBMs predicted to reach Chatswood in mid 2005.
Rock and machines
The 12.8m centreline spaced tunnels run at depths between 15m and 60m through the Sydney Basin, consisting of the Mid Triassic Ashfield Shale, the Mittagong Formation and Hawkesbury Sandstone. Tunnel excavation is almost exclusively through the well documented ‘tunnel friendly’ Hawkesbury Sandstone, a medium to coarse grained quartz rich rock with strengths of around 35MPa.
The JV has chosen to bore the tunnels using two 7.2m diameter Robbins hard rock TBMs, procured under an optional buy back agreement. Some 10% of each 1200t machine was originally used on the Lesotho project in Africa, specifically the cutterhead support, the mainbeam, the gripper carrier and the gripper cylinder, the remaining 90% is all new.
The machines were taken from storage in France and transported to Newcastle, New South Wales, for refurbishment at the Forjacs shipyard in late 2002.
Each TBM was completely re-built under Robbins supervision to give 7,770kN of thrust and 1,475kNm of torque, while the 2,300kW cutterheads were armed with 35 x 17″ wedge lock design backloading cutters to cut the abrasive sandstone.
Roof drills, mesh installers and ring beam erectors were also provided to aid installation of the sparse initial tunnel support, designed to cope with the route’s six expected support classes of sandstone. This varies from spot bolting in the tunnel’s predominantly good ground (PRL-1) to a tight array of chemically grouted rockbolts and steel sets in the very occasional poor conditions (PRL-6) expected through the alignment’s six known fault zones. The JV manufactures any shotcrete required for roof support at the M2 worksite.
The tunnel invert is cast 6m behind the TBM and is an integral part of the machine’s operations. The casting crew do their work under the main beam of the TBM. The concrete is manufactured on site, delivered to the TBM in Jacon underground agitator trucks, discharged into a skip on the TBM and carried forward to the work station using a monorail crane.
Early progress
The first Robbins TBM, named ‘Babs Guulang’ by local schoolchildren (Aboriginal dialect for ‘Big Wombat’), was delivered to the M2 worksite in August 2003, assembled in 40 days and put to face on its journey to Epping in the second week of September 2003. TBM 2, or ‘Running Rosie’ followed six weeks later.
It was hoped that both TBMs, considering the good ground, would quickly reach excavation rates of up to 200m per week. For TBM 1, this was soon to prove difficult. Within 50m of starting the first tunnel drive THJV experienced significant difficulties with the cutter boxes becoming clogged, prohibiting the flow of muck through the cutterhead and onto the conveyor, limiting advance rates to 50m per week.
THJV project director, Steve Burns explained, “on starting the first machine there was a fair amount of water coming in from the face. This made the clay present in the rock matrix form a very sticky material that eventually plugged the cutterhead. This took a while to clean out and cost us a shift each time the cutterhead had to be cleaned. When the water inflow eventually reduced, things improved substantially, unfortunately the head plugged several times more whenever water was present.”
Another unwanted by-product was that the paste like material was causing higher than expected cutter wear, requiring frequent tool changes. The decision was taken between THJV and Robbins to modify the cutterhead by cutting out the 120mm + spaced grill bars in front of the muck buckets and increasing the size of the cutterhead openings. The cutter bars were removed at the first available opportunity and it was agreed that the muck buckets would be modified upon breakthrough into Macquarie Park Station.
During this time TBM 2 was being assembled in the launch chamber. The experience with the first machine’s muck handling was not lost on THJV, which immediately cut out TBM 2’s grill bars and partially opened the bucket, by removing the plate from the bucket’s leading edge, to leave the most open muck flow path possible. The machine went to face on the second drive in mid October and the results were immediate. In good rock with minimum water seepage TBM 2 bored instantaneous advance rates of up to 6m per hour, impressive for a 7.2m diameter machine.
TBM 1 achieved the first major tunnelling milestone on November 21st when it broke through into the Macquarie Park Station cavern and in doing so began another technically innovative part of the design – boring through the cavern bench with 1.7m of the cutterhead exposed.
Station bores
The breakthrough into the eastern side of Macquarie Park Station was within 32mm of the design alignment, thanks in part to the German PPS laser guidance system mounted in the TBM. The station cavern’s 200m length and 20m width had already been excavated by roadheader to a depth 5.5m above the cavern floor leaving a 5.4m high bench to be removed by the TBMs, with each drive separated by a central 5.6m wide pillar (see stations box). The central pillar is removed by excavator and rock hammer after both TBM’s have passed through the cavern.
This unusual method of removing the final 5.4m depth of the cavern was selected by the JV due to the tight nature of the construction schedule. It has been used at Macquarie Park and Macquarie University Stations, but will not be necessary at Epping and Delhi Road where there is sufficient time in the programme for the caverns to be fully excavated.
Following the cutterhead modifications, TBM 1 bored at significantly higher rates while maintaining survey accuracy. The machine exited the cavern a week later, and began the drive to Macquarie University Station.
More critical would be the drive through the station by the second machine, due to the minimal rock provided by the central pillar for the TBM grippers to thrust off. To allay fears that excessive gripper forces could cause the pillar to fail during the drive, the JV reduced the gripper thrust by half to 900t and the torque by 60% to 1500kNm. A bracing frame was also positioned within the already excavated TBM 1 down-track tunnel to resist the gripper forces.
The eastern section of the central pillar displayed good rock quality and posed little problem to TBM 2 as it passed through. However, it was known that a significant 10m wide fault zone in the western section would require additional bracing. The fault zone presented a major weakness when the gripper thrust was applied, and even with the bracing frame in place, block displacements due to the broken nature of the ground were a real issue. The decision was taken to provide additional resistance to the gripper thrust, and confine the fault by installing 2 x 360 UB 56.7 grade 350 beams, anchored through the pillar with eight DSI Arnall 63mm diameter GEWI anchors, 10.5m long with a 5m anchor zone and a 5.5m free length. The JV decided to install one beam and anchors on the TBM 2 side of the central pillar, and the bracing frame in the TBM 1 tunnel.
Initially, when the gripper was applied the bracing frame was not fixed to its optimum position and the gripper thrust moved the pillar towards the TBM 1 excavated tunnel, which was held only by the single beam, resulting in displacement that dynamically loaded the anchors. To remedy this, the bracing frame was realigned and TBM 2 passed tentatively through the fault zone with the anchors reducing the load transfer to the bracing frame by 50%.
The method proved successful and TBM 2 excavated the remainder of the station without incident and with only a 4mm maximum shear displacement measured within the fault zone.
On schedule
Both TBMs are now well on their way to the breakthrough at Macquarie University Station and when T&TI recently contacted the site the JV was still enjoying good production rates with a best week so far in excess of 200m. Project director, Steve Burns said that, “the TBMs are performing well and the whole TBM team is making best use of the Robbins machines, which appear to be well suited to the ground conditions. As you would expect TBM 1 is always the pioneer and any lessons we learn from TBM 1 are quickly implemented on TBM 2.”
The JV is busily planning for the final lining to start in September, following pullback of the TBMs from Epping to the M2 worksite. The forms for the 200mm thick concrete lining are currently being manufactured by Oestu in Hungary and are due to start arriving on site in May. The JV’s strategy of having total control of the concrete batching and delivery operation should stand them in good stead for the tunnel lining operation.
Related Files
Figure 4
Figure 3
Figure 1
Figure 2
