The east-west alignment of the Umiray-Angat Transbasin main tunnel has a 0.14% gradient from the Macua portal. In such a shallow hard rock drive, the TBM vertical steering control makes it difficult to facilitate continuous drainage of water inflows back from the tunnel face without flooding.
A total of 16 boreholes were drilled for investigation of the long rock tunnel, mainly at areas of shallow cover. The deepest was only 350m, giving poor coverage for the tunnel considering the 1,200m maximum overburden. No in-situ stress measurements were carried out.
The geology of the tunnel, which is in an area of high seismic activity, consisted of a young (Late Eocene) volcanic sequence composed of 50-120MPa agglomerates, basalts and tuffs with inter-layered sedimentary rocks including limestone close to the Umiray portal area where overburden cover was low. Fig 2 shows the strata and the rates of advance monitored on a weekly basis.
The project suffered a series of delays. Table 1 summarises the monthly progress, listing the main causes as cave-ins at the TBM cutterhead, water inflows and squeezing ground.
The first cave-in or collapse at the TBM cutterhead occurred at chainage 4,330m at a fault zone that affected 44m of tunnel. This is assessed as being over 5 on the IMS ground classification system (McFeat-Smith, 1998). Here the cutterhead was totally buried under mylonitic debris from a collapse zone for 8-10m above and ahead of the face. The void walls were stabilised first by shotcreting then by injection of mortar and finally the debris treated with resin grouts while being excavated in successive 1.8-2.5m long sections. The final step involved consolidation grouting.
In this case the benefit of the double shielded facility was evident as the rock grippers could not be used to sustain any thrust and up to 40,000kN thrust was achieved by thrusting against the segments. The overtorque and high power facilities also proved invaluable for excavation of the collapse debris. A delay of 41 days was experienced in the fault zone.
The second cave-in came in late November 1999 when the TBM cutterhead entered a fault zone in a white limestone at an area of low cover. This time delay was 68 days, particularly frustrating for the JV, as it was forced to put back the planned breakthrough ceremony in December 1999.
On 11 October 1998, sudden and abnormal water inflows of about 600 litre/s were encountered at the TBM cutterhead at chainage 5508. Here andesite porphyry is inter-layered with sedimentary rocks in strata and there was reason to anticipate such high inflows. The water level rose quickly above the working platform at the TBM.
After a few days flows began to dissipate, allowing the miners to install two permanent pumps at the TBM working area, with one 400 and two 250mm diameter backup pipelines running the full length on both sides of the tunnel. This enabled the TBM to restart slowly. Later a sump, two auxiliary pumps and 11km of pipeline were added behind the back-up.
While the inflows at this point fell to about 50% of the initial volume, further high inflows were experienced on the drive resulting in peak inflows of 810 litre/s. However, as at chainage 5508m, the well-motivated TBM crew and skilled back-up team quickly installed water diversion facilities. Several electrical and mechanical breakdowns were caused by the high inflows which, with a range of delays due to flooding, management of the large water inflows and spoil disposal accrued to over 48 days.
Shortly after the first major inflow, the TBM crew met its first section of convergent ground at Chainage 6131m, on 11 December 1998. The cutterhead and shield were suddenly jammed by unexpected, abnormal short-term convergence in a very strong, competent basalt and metamorphosed volcanic sequence. A calculated convergence of 0.1m had been expected according to the known geological profile and the TBM overcutting facility was designed to accommodate up to 0.14m. However convergence was significantly above this, rendering the facility unusable. Attempts to free the TBM, including the use of maximum power were in vain and the cutterhead had to be dislocated from the main support, necessitating intervention for a special realignment.
Fortunately, the innovative design of the TBM meant the telescopic shield could be opened up, allowing access for hand mining and smooth, controlled blasting. As shown on Fig 4 this allowed a 1.6m high heading to be excavated and supported by shotcreting for a distance of 5m to free the TBM cutterhead. The total length of this first convergence zone was about 30m, which delayed progress by 31 days.
A second, slight convergence zone in hard rock was met on 16 March 1999 at chainage 8016m and was dealt with in the same way.
A third and final zone resulted in closure of the tunnel over a length of some 175m. A peak convergence of 0.37m was monitored, occurring in less than two hours. Sections of very loose soils were also found in this zone, which were stabilised by advance injection of resins. Hand mining again proved effective in minimising delays to the drive.
Comment
To obtain a perspective on the water inflows on this project it is worthy to note the writer’s experience gained from the driving of some 200km of rock tunnels, some of these “notably wet” tunnels. From this experience, a systematic method of predicting inflows has been drawn up as presented by McFeat-Smith (1998). More recently individual inflows through open joints and shear zones in both land and marine tunnels, yielding inflows of 75 litre/s with cumulative inflows of about 200 litre/s over sections of tunnel of several hundred metres in length. In the opinion of the writer inflow of the scale of these recent experiences (>200litre/s) are difficult to predict as well as to control on site and it is both fair and reasonable for contractors to expect additional reimbursement in contracts for management of such disruptive inflows.
Related Files
Fig 1: Sketch of hand mining work
Fig 4: Simple geological profile
Figs 2 and 3
