Yu Shyi-kun, Taiwan’s new prime minister, caused a minor stir last month when he made an official visit to the much delayed Pinglin tunnel project. He wanted to underline the government’s view of this key east-west road connection as a “highest priority” and insisted it should hit the latest completion date of 2005.
Whether that can be done remains an open question. As one of the world’s largest road tunnels – a dual two-lane motorway connection 12.9km long – Pinglin has already been delayed by difficulties with rock, water and even the tea planters in the remote mountains. Only 12% of the work had been completed when the original 1999 operation date passed, but progress since then has been much better and 35% of the two main 10.8m i.d. tunnels has now been completed. An exceptionally fraught pilot tunnel is also half completed. However, unexpectedly difficult geology, fractured rock and massive inflows of water continue, although the very worst ground is almost passed, and many more rock and water problems are expected.
Recent political pressure on the project leadership has been mingled with support for the engineers and contractors. Some Government ministers say they understand the tough conditions, as engineers continue to struggle with the multiple fault lines, crushed rocks and underground reservoirs. “Of course you can say ‘work harder'” adds one senior Taiwanese engineer on the client side. “But they are already working very hard.”
Long hours and a more systematic approach to the ground have been applied in the last three years, particularly as a result of a complete review of the project in early 1999, when a team of international tunnelling experts were invited in to examine the scheme. Their conclusions and advice, both technical and organisational, have been largely followed and have already meant significantly better progress. Despite the loss in 1999 of one of two big Wirth TBMs, after the worst of seven collapses, tunnel drives have been progressing consistently from the original eastern portals. New headings have allowed tunnelling to increase substantially.
Pinglin was never going to be an easy project. It is, to start with, hugely ambitious, aiming at a highway tunnel which was to be the third largest in the world when work began. The twin two-lane bores, each 10.8m i.d. with 60m between centres, will complete a 31km route though high mountains across the north, linking Taiwan’s capital Taipei with the western I-lan county. Currently the trip involves a tortuous journey through mountain villages and high tea plantations lasting two and a half hours; the complete route will make this half an hour, and give the island a critical fast link to its isolated west side. Several shorter tunnels, viaduct links and interchanges for the Beiyi Freeway have already been opened. But Pinglin remains the key to the whole connection. It includes, apart from the main tunnels, a smaller pilot tunnel of 4.8m diameter for drainage, ground treatment and an emergency escape route when the tunnel is operating, and also numerous cross connections.
“There are pedestrian links every 300m and vehicular passages every 1400m” says Wei-Chaung Lee, deputy project manager with RSEA Engineering Corporation, the contractor on the project since it began in 1991. Three sets of twin ventilation shafts complete the main excavation works, the deepest dropping to just over 450m from the surface and the others about 250m. These were not originally planned as access for construction but have become so, in order to speed up the work.
Geology
The whole 385km long island of Taiwan is the result of a complex three-way tectonic collision between at least two giant crustal plates; the Eurasian and the Philippine Sea Plate. The Eurasian plate plunges down eastwards underneath the Philippine plate, below Taiwan. However just off the north-east coast near the Pinglin project the Philippine plate plunges northward below the Eurasian plate. The resulting, mainly north-south ranges, of young mountains, up to 4000m high, are underlain by a series of sub-parallel eastward dipping thrust faults, running the length of the island, which lead to frequent earthquakes. The project area at the north of the island features some of the oldest rock, thin strips of Late Palaeozoic and Mesozoic metamorphosed rocks are exposed. The majority of the island is Cenozoic sedimentary rocks, both Tertiary and Quaternary: rock is often shattered, heterogeneous and difficult to predict. High cover to the route, rising to over 750m, means that investigation is difficult. “I would say the geology is several times more difficult than the Alpine and Apennine areas in the European Alps and Italy” declared Italian Professor, Sebastiano Pelliza, during the expert consultants meeting in January 1999.
Locally on the tunnel route a wide variety of sandstones and argilites predominate, with some hard and brittle types but also fairly competent at the western end. Most difficult is the Szeleng formation, which occupies most of the first 4km inwards from the eastern portals. Szeleng is a massive quartzite described as hard, brittle and abrasive by engineers on the project. “In the first part it is slightly metamorphosed and also frequently highly fractured and blocky, with clay gouge interleaving” says Lee. It is crossed by major fault zones. The tunnel route sits over a complex line of up-thrusting from the plate activity, almost exactly across one of the active plate boundaries. Further faults zones are expected.
Early history
The choice of using TBM, might also be considered ambitious. The contractor although experienced, had not carried out a TBM tunnel before. Taiwan’s government agency TANEEB (the Taiwan Area National Engineering Bureau) had settled on RSEA for the work by negotiated contract in 1991, in the then usual command economy way.
Two contracts had been agreed on a lump sum basis, one at US$66.4M for the smaller pilot tunnel and a larger US$537.1M one for eastern end works on the highway, which predominantly comprised the two main tunnel bores. TBMs seemed a good idea to the original design group because of their potential speed, allowing an estimated 15 years of conventional drill and blast work to be cut in half. TBMs also helped with a growing construction labour shortage in Taiwan, which like many Asian “Tiger” economies was expanding, but also limited foreign workers by quota. Major international consultants advising on the project, including the USA’s De Leuw Cather, and Parsons Brinckerhoff, Austria’s Geoconsult, local firms CECI, and Sinotech and Switzerland’s Electrowatt, were aware of the rock conditions from earlier ground investigation results, but it was thought that with ground treatment and drainage through the pilot tunnel it would be possible for the machines to work satisfactorily.
Sinotech, Electrowatt and Parsons then created a basic design for the TBM operation, which the former worked up into a detailed design. Work would start from the eastern end where access from the small coastal plain is easiest and head west. This also meant they could proceed upwards on the 2.5% tunnel gradient, an important consideration given large quantities of water expected. The first kilometer would be excavated by drill and blast in order to deal with very sheared quartzite, with machine launch after that. To cope with lack of TBM experience, one of the main tunnels (the eastbound bore) was to be carried out by an experienced subcontractor, France’s Spie Batignolles. RSEA would also get technical advice for the westbound bore.
The pilot tunnel was used to create a gallery 5m below and centrally between the main tunnels from which probing, grouting and drainage drilling could be carried out. However, because of slow drill and blast the 4.8m Robbins telescopic-shield TBM started at just over 500m in, during December 1992. It was due to finish by 1995, emerging at the western portal. But by early 1999, had still only excavated about 1600m, suffering ten stoppages, predominantly caused by the blocky nature of the rock jamming the cutterhead. Most of these incidents needed bypass work to clear the machine; these operations themselves were difficult and often had to re-start themselves because of bad rock. “Because of these difficulties the pilot tunnel has never been much more than a few hundred metres ahead of the main bores” says Lee. What work could be done for drainage and investigation was also difficult, because of the need to work outwards and upwards 50m laterally.
The main drives were also using telescopic TBMs, this time from Germany’s Wirth, delivered in spring 1995 after drill and blast had reached the launch chambers at 900m. Without much ground treatment there were going to be difficulties with these drives too, claimed Spie Batignolles. A dispute ensued, with the machines left parked outside the tunnels. Lee is not prepared to comment much on the details, saying only that RSEA terminated the contract in the autumn. Arbitration and legal wrangling has persisted since, which he says has been primarily in RSEA’s favour. Difficulties there were, with both drives, once they started in mid 1996. Inflows were frequent, even probing ahead proved difficult – the machines allow only for a shallow upwards angle and the blocky abrasive rock caused problems with heavy wear on bits and long times needed for the 30m holes.
Most major difficulties were on the westbound drive, which suffered seven major stoppages, the last of which, in December 1997, was a rock and water inflow over a 100m length, that buried the head of the machine and crushed part of the segmental lining behind it. Some 740lt/sec of water was recorded at 18 atmospheres of pressure. “That day I was on site” says Lee. “I remember in the morning we had more and more coming in and a crack developing in the crown of the tunnel. We tried to put steel rib supports inside the segments but by noon we knew we had a disaster coming. At 2pm we pulled back the men.” Clearing the machine took more than a year and it was heavily damaged. It had driven just 450m.
The other machine, 650m along, was at almost the same point in the rock and was halted. In early 1999 with three machines stopped – though the pilot tunnel continued by drill and blast – a team of foreign experts was invited to visit and comment on the operation. Apart from Professor Pelizza, these included Gunther Riedmuller from Graz university in Austria, Heinz Hagedon from the Gotthard base tunnel project in Switzerland, tunnel consultant Heinz Hofmann from Germany, and a representative from Japan’s Kajima.
A week long discussion ran to 40 pages of recommendations. Key to this was the flexibility of the operation be increased by various redesigns. The westbound TBM was almost certainly beyond use and should be abandoned with “slower but more flexible” drill and blast to continue. East bound the TBM would continue, but with drill and blast assistance to get out of the most difficult ground. Three main options were proposed; a top heading to form a canopy, a top heading and bench, or a centre pilot tunnel, although this would create problems with anchoring support. To add further flexibility, the cross connection tunnels should be fully used for excavation and to facilitate ground treatment. The passages also allowing plant and materials movement for lining work. Each main tunnel has an initial unbolted precast concrete segmental lining and an inner in-situ concrete lining with watertight membrane. Drainage channels are linked to the pilot tunnel at 1.5m intervals.
The group also agreed with new headings from the western end for all three tunnels; these would be by drill and blast because their downward slope would cause drainage problems. Further headings should be started from part way along, using shafts. Two of these are now being excavated from above rather than waiting for the originally planned raise bores; improved hoists speed excavation. It was suggested that ground conditions were slowly improving as progress westwards was made, but “major geotechnical uncertainties” remained, with ravelling and squeezing rock still likely to be encountered. With care the pilot tunnel TBM was advised to proceed. But the importance of increasing the speed of the pilot tunnel work, suggested continuing drill and blast “with increasing intensity” for the time being while TBM methods were studied. The tunnel has a “decisive role” for investigation and preventive rock mass improvement they said. Systematic core drilling ahead of the pilot tunnel was also suggested, using long hole core drilling up to 100m with a 20m to 40m overlap on alternating sides of the tunnel, allowing investigation but then as drainage to improve grouting operations.
The team also commented on the organisation of construction operations, calling for stricter management control systems and a new construction schedule on the basis of the new work. They also wanted to see improved training for the subcontractors and a more precise definition of the co-operation and responsibilities between the contractor, designers, client and subcontractors. That might also mean looking at the initial contracts. Possibly the lump sum conditions could be examined to find a more adequate basis for the work. It is possible that a normal contractor would have cut its losses and pulled off the project some time ago. Currently, there has been an increase in the price of US$77.8M on the main contract and US$31.7M on the pilot tunnel.
Current Progress
“Many of the notions in the experts’ report are now being implemented” says TANEEB representative Jawchang Laiw, vice director of the planning division. More money is available – he estimates US$110M at present – and says a number of measures to make the contract more flexible have been implemented, with better risk sharing, and training.
The westbound machine was dismantled in 1999 and drill and blast has pushed the bore forwards to 3164m at the end of March. The eastbound TBM has advanced to 2779m behind a top heading canopy of some 37m2 cross section, which has reached the 3317m point. On the western drives the main bores have reached 3096m on the westbound and 3211m on the eastbound, with the pilot tunnel pushing to 3737m. A subcontract for Kumagai is pushing the westbound tunnel outwards from the number two shaft. “It is expensive to work from here” says Laiw “when you have to haul all the spoil upwards through a 270m deep shaft. But it is a way to get ahead.”
The truck movements are not easy along the narrow mountain roads and the contractor has to be continuously aware of sensitivities to noise and dirt. There have been ongoing protests about the shafts anyway, because tea planters feel their crops will be ruined by exhaust fumes from the tunnel. Currently the deepest shaft is suspended while the issue is worked out. Shaft operations have allowed just under an extra km to be excavated on the westbound bore working outwards in both directions. Shaft based work on the pilot also began here in March reaching out towards the advancing TBM. The pilot TBM is working more steadily with a routine forward probe drilling of 10m to 30m carried out, which while slowing progress has made it more steady. Long hole drilling has been done five times and high pressure grouting 22 times. More skilled workers have been taken on and improvement made in the plant and equipment by the contractor. But there are more difficulties to come says Lee. One of the other recommendations of the experts’ group was further ground investigation, this indicates major difficulties in the ground just beyond the points currently reached. “We foresee another sixteen shear and fracture zones in the section ahead” says Lee “based on bore data from Sinotech”. One of the worst areas of all, a 300m long shattered rock zone around the Shihtsao Fault is “the most difficult fault zone of the whole project”. The experts estimated a completion date between four and six years depending on the measures implemented and the ground encountered. That means another three years and currently the completion date is set at 2005. But just what comes next is still to be determined. T&T
Related Files
Figure 3: Cross sections
Figure 2: Longitudinal section of the complex geology along the tunnel alignment
Figure 1: Alignment of the Pinglin Tunnel, situated in the north east of Taiwan
