Revolutionary techniques are currently being implemented by contractor, Miller Civil Engineering, in a desperate bid to rectify last year’s much publicised tunnel collapse on Yorkshire Water‘s 10.6km-long, £200M ($280M) Humbercare flow transfer main project in Hull. Liquid nitrogen (LIN) is being used in a complex groundfreezing operation. (News Aug ’00)

The events of the 16th November 1999 has left the Lovat TBM Maureen, used to drive the 3.6m-diameter tunnel, trapped 150m west of the collapsed section with just 50m left to complete boring.

The exact reason for the catastrophic event is still to be ascertained.

All remedial work centres around Shaft T3 as the length of collapsed tunnel stretches 30m west and 65m east of this point. The term collapse in this case may be something of a misnomer. Over much of the near 100m length, the tunnel is still intact and has simply dropped into a void created by the underlying bed of fine windblown sand that poured into the tunnel under high groundwater pressure during failure of the lining. The effect of the displacement resulted in the tunnel dropping by up to 2.5m leaving a 60m diameter, 2m deep subsidence bowl around shaft T3.

Damage assessment

Immediately following the collapse, the team on site took dramatic steps to stabilise the collapsed area. Surrounding shafts T3 and T3A were backfilled to make an air- tight environment and 2 bar pressure of compressed air was applied from a bulkhead at Shaft T4. The remedial technique was soon replaced with the longer term more cost effective flooding of the affected area between Shaft T4 and the abandoned TBM. This gave the team time to investigate and implement an effective rescue strategy.

Holes were drilled through the tunnel crown to certify the length of damaged tunnel. At the ends where the intact tunnel was determined, thick grout plugs were introduced and the entire length back to Shaft T3 was grouted to fill the void between the top of the sand and the tunnel crown. A vertical freeze bulkhead was constructed through and around the tunnel at the point on the east side where the tunnel was discovered to be intact. This was achieved by drilling vertical holes from the surface down through the tunnel and freezing the ground with LIN. Water was then drained from the Shaft T4 and the tunnel entered from this point and cleaned out. A structural blockwork and concrete bulkhead was constructed and the freeze turned off.

Freezing process

The tunnel is being rebuilt using the sprayed concrete lining technique along the original alignment and when completed will have the same finished internal diameter as the original tunnel (3.6m). Excavation, to be carried out in five 20-m horizontal stages, has necessitated a sophisticated LIN groundfreezing method.

The initial stages comprise a vertical freeze wall through the tunnel drilled from surface about 20m east from shaft T3. A horizontal freeze is then applied around the tunnel periphery to allow excavation wholly within frozen ground. This is done by setting up a series of holes parallel to the tunnel alignment from the shaft out to the first vertical freeze wall (see diagram). The drilled holes are splayed out to increase the 5.6m diameter to 7.5m over the last few metres of the 20m advance thereby constructing a drilling chamber to allow horizontal drilling of the next 20m length.

The final 20m length on the east and west side will be ‘wrapped’ around the existing intact tunnel to allow connection of the sprayed concrete lining (SCL) and segmental tunnel. At the time of writing, exact details of the joint of the SCL tunnel to the intact tunnel were to be finalised.

The aforementioned splaying pattern requires extensive survey work to ensure 3-D positioning of the full hole length is accurate. This is achieved by Austrian specialist drilling contractor Insond using a down-hole survey tool.

The holes are then equipped with copper freeze pipes plumbed into the LIN system. Control of the LIN system, supplied by Linde, incorporates actuators to regulate flow of LIN to each hole to maintain a predetermined exhaust temperature. The safety systems include an automatic shut off of LIN supply if flow control detects a leak or if oxygen monitors detect low levels.

Tunnel reconstruction

Once the freeze had been completed around the tunnel periphery and at the vertical bulkhead, excavation began from the shaft heading west towards the TBM.

The excavation was originally being carried out by a mini- excavator with hand trimming. Miller has just introduced a Schaeff roadheader to speed up the process now that excavation space permits. Miners are dogged by stifling conditions as digging through the sands, gravel and original concrete segments. They also have to contend with a 4m-long trapped loco.

A 3-stage system has been adopted consisting of crown, bench and invert at 1m advances. The primary lining has to be applied against frozen ground and application problems were addressed during tests against frozen trial panels. The possible application problems were overcome by using an insulation layer against the ground and heating the mix of aggregates and water to a temperature of 40oC prior to batching with the latest accelerator technology. The heated mix is then applied.

The lining is reinforced with mesh, lattice arches and steel-fibre reinforcement in the mix.

Secondary lining will be applied once the 20m leg has been excavated and work is being undertaken on the next leg. It will comprise 450mm SCL onto the primary lining. Once in place and up to strength the freeze will be turned off.

Speculation without reason

The well documented event has drawn widespread speculation within the UK tunnelling fraternity. All manor of reasons for the calamity have been thrown into the pot ranging from difficult ground conditions, fluctuating groundwater systems, vibrations from construction to the grouting system used in the trapezoidal ring lining.

The fact is no solid reason has yet been formulated. Peter Chamley, project manager for the consulting engineers, Ove Arup told T&TI: “It’s simply too early to tell. We’ve drawn up an initial list of up to 10 possible reasons for the tunnel displacement. It could be just one of these, or a combination of several.”

“The bottom line is, what caused the leak that let the sand in? That’s what we’re focusing on.” He concluded, “We’re not going to know for certain until we’ve opened up the east side excavation which is where the collapse happened.”

Steve Tindall, project manager for client, Yorkshire Water, told T&TI of the team spirit that has helped minimise the potential fallout of such an incident, saying, “The teamwork approach, developed from Miller’s early involvement and encapsulated in our ‘Charter’ has enabled the team to focus on problem solving and delivering to time.” He concluded, “The tunnel incident was an extreme example of a problem and it was gratifying that the team together with new specialist sub-contractors and other participants felt able to continue with this approach to everyone’s benefit.”

Related Files
LIN groungfreezing scheme