Technical innovation, stubbornness and persistence were required in equal measure to achieve the recent breakthrough on the Gotthard base tunnel, winner of the prize for projects over £1bn at the NCE Tunnelling Awards. Adrian Greeman reports.
Genius is said to be 10% inspiration and 90% perspiration.
Likewise construction of the 57km long world record Gotthard base tunnel, which made its final breakthrough on 26 October, was a mixture of invention and ingenuity combined with sheer effort and determination.
In the more than 10 years that the excavation of its twin bores has been underway, designers and contractors have had to find new ways to deal with extraordinary problems.
But the Swiss, German, Italian and Austrian firms that have done much of the work also had simply to struggle their way through difficult and dangerous rock conditions, large-scale supply structures and logistics and, for Switzerland at least, unprecedented timescales.
The SwFr9.8bn (£6.4bn) Gotthard base tunnel was clearly going to demand a new level of complexity from the beginning.
Unlike the Seikan rail tunnel in Japan and the Channel Tunnel, pinnacles of 20th century tunnelling, this rail tunnel was to be at depth, running far below the snow-tipped mountains of the Swiss Alps with sometimes as much as 2,400m of solid rock above.
The enormous pressures caused by such a large overburden brought a number of difficulties. First was the temperature.
At these depths the geothermal heat of the earth is slow to dissipate which meant tunnellers were faced with high temperatures of as much as 50oC underground.
It was a serious problem not only for men and machines in the excavation and construction phase, but also for operations over the century and a quarter design lifespan of the tunnel.
“The Alps consist predominantly of two great ‘massifs’ of hard igneous rocks. These, the Aar and the Gotthard, are the backbone of the mountains.”
Significant extra ventilation and cooling was required for both.
Providing sufficient cooling was a major issue on all the contracts with clusters of 150mm diameter pipes needed to carry cold water many kilometres to the working areas to feed air-conditioning units and carry heat out to groups of cooling towers and holding basins at the tunnel portals.
For cooling alone, around £7,000 a day was typically being spent on electricity for pumps on such a contract.
The heat compounded another problem: water. Always an issue in tunnels, the mountain heights meant water might be found in cracks and faults running high above the tunnel, giving a potential pressure head of almost 200 atmospheres.
This could flood the tunnel quickly, possibly with hot water. Escape routes in the deepest sections would also be very difficult, especially at the centre where access was via an 800m deep shaft.
The huge rock load above the tunnel created other problems.
The high stresses carried the risk of rockbursts, where rock would suddenly shatter, explosively dislodging rock blocks up to several tonnes in weight into the bore space.
It also has a squeezing effect on the rock in many areas, so that tunnel bores had to be overcut and then more heavily reinforced than usual along much of the route.
This meant more materials, bigger machines and the need to consider higher levels of risk.
In two particular areas, the risks and problems in tunnelling were potentially much higher because of the nature of the ground, which is heavily cracked, faulted and crumbled.
“The Alps consist predominantly of two great ‘massifs’ of hard igneous rocks,” says Heinz Ehrbar, chief engineer for AlpTransit Gotthard, the company set up by Swiss Federal Railways to manage and construct the project. “These, the Aar and the Gotthard, are the backbone of the mountains.”
But between the great blocks are layers of softer rock and sedimentary deposits.
These have been ground like corn over the millions of years between the slow tectonic movements of the massifs, leaving rock fractured and damaged.
Various layers like this underlie the picturesque high valleys of the Tavetch intermediate massif and associated zones of rock in the Urseren-Garvera.
To the south, between the southern Gotthard massif and a third lower lying rock bloc, the Leventina Pennine area, there is also damaged ground.
But it is combined with another problem caused by the chemistry of the rock.
A kind of dolomite, it has a texture like crystalline sugar and in the presence of water it becomes a flowing powder, almost impossible to get through.
It fills the so-called Piora syncline which extends as a 300m thick band to the tunnel depth.
The special challenges of both areas were long known about and were a central part of discussions from the time when a base tunnel was first proposed in the 1940s.
So once the unique Swiss democratic process of referendums had agreed that the project should go ahead in the early 1990s, investigation of both these area became a high priority.
“We had to get a complete view of the risks,” says Ehrbar, who was with one of the main consultancies at the time, Zurich firm Electrowatt, which worked with Amberg Engineering on the southern section.
“You could hear the movements of the ribs which give suddenly like rifle shots. That can worry miners and we wanted them to become accustomed to that because it is part of the system working.”
On the northern part Swiss firms Gähler & Partners, Gruner, Rothpletz Lienhard and CES carried out the investigation and design.
The Piora was the most critical, so much so that in 1993 a 5km long tunnel was bored from a nearby valley to investigate, ending 300m above the main tunnel alignment from where a gallery would descend for probe drillings.
But even as this exploratory bore was being completed the worst was confirmed - a forward probe suddenly released a stream of high-pressure water and “sugar-rock” sand into the tunnel.
“It was a complete beach,” said one engineer at the time as fine white powder even spilled onto the highway at the tunnel entrance.
No one was hurt, despite the high pressure inflow, and this stopped once the drillhole became congested.
When the tunnel end had been sealed with concrete, new probes carried out by Canadian firm Morisette and Italy’s Rodio brought better news.
Cores to the main tunnel depth showed the Piora to be dry further down, and therefore still solid and without the feared 170 atmospheres or more water pressures.
The conclusion of the geologists was that a gypsum “cap” had formed in the Piora band at a higher level which was impermeable to the great head of water above.
The rock to be tunnelled would be stable dolomite, marble and dolomite-anhydrite.
While this was underway the Tavetsch zone was being investigated using directional drilling techniques borrowed from the oil industry. “We carried out some very long probes here with cores up to 1750m long,” says Ehrbar.
These too initially brought bad news since the probes became stuck several times in the rock, demonstrating the extent of the squeezing problems that would be faced in the cracked zones.
The drilling was nearly abandoned “until one last probe got through” says Ehrbar.
The extremely fractured ground was tested at the ETH Technical university in Zurich where a new type of triaxial cell was devised to simulate the pressures that would be found.
This ground would be a problem. It needed a lot of support but ordinary steel ribs would be crushed.
In the Alptransit offices in Lucerne are samples of thick steel profiles bent right back on themselves by just such forces.
Engineers believed a solution could be found, borrowing from German coal mining experience where seams are exploited up to 1km deep.
Special “collapsible” arch support sections had been developed for these, based on submarine technology where hull ribs are made with sliding friction joints.
The sliding arch sections “give” enough to allow the ground to relieve some of its stress slowly before the friction resistance builds up.
Special bolted clamps enable the precise levels of friction to be set.
But the method had never been applied on large cross section tunnels so theoretical and practical trials were needed.
Ehrbar remembers a final full-scale test in the central section of the tunnel itself where some of the mine workers were involved.
“You could hear the movements of the ribs which give suddenly like rifle shots. That can worry miners and we wanted them to become accustomed to that because it is part of the system working.”
Between 2km and 3km of this ground was expected on the alignment, requiring slow face excavation and a huge amount of support, sometimes with three arches every metre.
These challenges meant very different timescales would be needed for different sections of the tunnelling.
“But of course the sections were intended to be of approximately the same time span,”
There were to be five main divisions accessed by intermediate shafts or side adits - an advantage that a mountain tunnel has over a sea tunnel.
The sections were of varying lengths according to difficulty.
In the north there was some possibility of squeeze and heat to cope with, but tunnel boring machines (TBMs) could be used for rapid progress though mainly hard rock for two sections.
In the centre, accessed only by an 800m deep shaft at the village of Sedrun, was the very difficult ground and here only a few kilometres would be tackled.
The south was thought to be the most straightforward and a section from the southern portal near the town of Bodio was the longest.
The second southern section from a side tunnel at Faido, where the Piora test had begun, was also straightforward but extra work was necessary.
At this point was to be one of two great multi-function stations (MFSs), the major safety provision for the tunnels.
Here, trains could stop in an emergency or fire and discharge the passengers into refuge and escape tunnels.
Cross passages allow them to escape to the opposite tunnel bore for evacuation.
The MFSs comprise a large cavern and assorted side tunnels, with crossover link tunnels for foot passengers.
A network of smoke vents and air ducts was also needed.
There were also to be major track crossover tunnels at this point, partly to allow trains to switch tunnels for safety reasons, but also for maintenance needs so that sections of tunnel can be isolated in one bore.
The multifunction stations were also the major inlet points for powered ventilation of the tunnel used during routine weekly maintenance, and for power lines and additional equipment.
All this meant substantial additional excavation at the Faido section, to be done by conventional means as well as the TBM work.
A second MFS was built at Sedrun.
“But of course the sections were intended to be of approximately the same time span,” says Charly Simmen, a project manager with AlpTransit.
In the event that did not work out. Construction work proved far more difficult than thought in the south, but more straightforward in the centre.
Meanwhile, political and environmental issues held up the start of the northernmost contract.
The town of Altstadt, supposed birthplace of William Tell, wanted an extension to the tunnel to take the railway further within the mountain along the side of its valley.
An agreement to make provision for this in the future, by building a junction just within the base tunnel for additional bores, took some years to reach.
Despite this delay, the northern sections went exceptionally well.
While the portal section was suspended, the second started under a contract let to Austrian-Swiss joint venture of Strabag and Murer.
A 1.8km long side adit had been built early on at the little town of Amsteg in the narrow valley which leads into the Gotthard high road pass.
Two Herrenknecht TBMs were assembled in a 70m long cavern at the end, then driven southwards for 11.5km towards Sedrun.
Despite some difficult ground zones, speeds of as much as 56m per day were achieved in mainly hard granite.
In much of the Gotthard the lining needed is around 300mm thick but for Sedrun between 600mm and 900mm thickness of concrete was used.
“The only real issue was a jamming of one of the machines in 2003,” says Adrian Wildbolz, the AlpTransit supervising engineer for these sections.
Both machines hit hydrothermally disaggregated ground, like loose road chippings.
For the lead machine this meant using extra support to achieve slow but steady going, but for the other, the ground unravelled and jammed the cutterhead.
A rescue operation was devised with two separate hand dug and heavily grouted tunnels excavated from the first bore across the 40m separation.
From these the ground was sealed and an “umbrella” created with forepoling to create a chamber in front of the machine to allow the head to be cleaned and freed up. Nearly six months was lost.
But subsequent speeds were so good that the machines still arrived half a year early and Strabag was able to bid for the delayed 7.7km long section from the portal at Erstfeld when it was finally let, with drives beginning in 2007.
Using the same TBMs it made fast progress, breaking through in mid-2009.
In Sedrun, meanwhile, the work also went well. Access was difficult, from two 800m deep shafts at the end of a 1.1km long tunnel.
The second shaft was added during the tender stage for both safety and logistical reasons, to speed difficult site access and deliveries and aid evacuation.
As anticipated, the ground on the 2km northern heading from the shafts was exceptionally difficult with substantial squeezing.
But the special sliding arch support system worked well and contractor group Transco-Sedrun, comprising Swiss contractor Implenia Bau leading Frutiger, Germany’s Bilfinger Berger and Italian Pizzarotti, made steady progress of around 1m a day.
The excavation here had to be much larger in diameter than elsewhere, up to 13m, to allow for the squeezing, and for the eventual inner concrete lining to the tunnels which must give the running tunnels long-term stability for the next 100 years.
In much of the Gotthard the lining needed is around 300mm thick but for Sedrun between 600mm and 900mm thickness of concrete was used. A design option for 1.2m was considered but not used.
Southwards the ground was not expected to be much better.
But it proved surprisingly sound after the first 600m. Hydraulic hammer excavation and then drill and blast made such good progress that an initial 4km to be done was extended twice to a final 6.5km, says Jakob Lehner, a supervising engineer for the contractor.
The extensions helped compensate for some major difficulties on the southern two contracts carried out by Tunnel AlpTransit-Ticino (TAT) a group led by Swiss contractor Implenia Bau (formerly Zschokke-Locher and Bati-Group) with the Austrian Alpine Bau, Italian firm Impregilo with Swiss subsidiary CSC Impresa Costruzioni, and Germany’s Hochtief.
Both contracts had been merged into one, a contractor’s alternative at the bid stage because logistics seemed to be the key issue.
Expected good ground meant the southernmost contract from Bodio would be the longest section at 16.5km, and even the Faido section which included the complex multi-function station would be 13.5km long.
The merging allowed two Herrenknecht machines to make both drives, with a reconditioning stop in the cavern area for the multifunction station.
It also justified the expense of an automated rail system.
“It is the great irony of the project that the most risky and complex sections we planned for should have been less difficult than thought, though not easy, and the easiest had so many problems and challenges,”
Nearly all material supply and spoil removal was done via the Bodio portal which meant supply trains would run up to 30km and take two and half hours for the journey in the final stages.
A centralised rail network with computer controlled points and CCTV monitoring was economically feasible, and necessary, to supply the TBMs. On other contracts, conveyors were used for spoil and manually controlled trains for supply.
To everyone’s dismay the rock in the south proved far worse than expected with convergence and rockbursts a particular problem.
This made it necessary to modify the TBMs to allow better rock support and additional mesh support for safety reasons.
The TBMs had been built at a smaller diameter of 8.8m “like formula One cars for speed” says the JV’s managing director, Olivier Böckli “but had to be beefed up”.
Similar problems slowed the drives northwards once the TBMs had stopped at the Faido cavern where the diameter was increased to 9.4m.
But the biggest problems were to be in multifunction station area itself.
From the moment the contractor began work with excavation from the 2.7km side adit at Faido the rock proved far worse than anticipated, demanding major support and sometimes complex partial face excavation.
As much as 70% of time was used installing supports says Böckli.
Worse still, part of the cavern suffered the same squeezing issues as Sedrun.
Convergence was over a metre in places and some parts had to be reprofiled as many as four times.
For one section, the special sliding arches were brought in, as used at Sedrun.
To ease the work the station area was redesigned with the train tunnel crossovers separated out from the emergency side tunnels and moved southwards where there was better ground.
The passenger exit routes were lengthened.
In all, the changes stretched a three-year construction period to five for the multifunction station.
“It is the great irony of the project that the most risky and complex sections we planned for should have been less difficult than thought, though not easy, and the easiest had so many problems and challenges,” says Heinz Ehrbar.
The only problem it did not have was in the once feared Piora zone, through which both TBMs passed incident free.
To help the schedules, the lengths of the drives for the Faido section were reduced by 2km, the addition taken up in the Sedrun contracts.
In the last week of October the TBM cutterhead broke through into the Sedrun section, completing excavation of the world’s longest tunnel.
The tunnel is expected to open in 2017.
The Gotthard base tunnel is one of two deep level tunnels included in a £19.2bn investment in Switzerland’s railway.
Swiss voters approved the project in the 1990s under the country’s comprehensive referendum system.
The project and funding from oil tax and heavy vehicle road tax, were both agreed.
The key purpose of the tunnels is to integrate Switzerland into the European high-speed rail network and to reduce truck-borne freight passing between northern Europe’s industrial centres and northern Italy.
By creating a deep level tunnel with a near flat vertical alignment rising to only 550m above sea level, the capacity of the near saturated rail network across the Gotthard pass can be hugely increased.
Gradients at only 6.7‰ mean the tunnel will allow speeds of up to 250kmh for high-speed passenger trains and 160kmh for capable modern freight trains which will also double their capacity to 4,000t without needing the two locomotives required on the 1,150m high pass.
Several other tunnels are part of Gotthard axis.
The Zimmerberg near Zürich was completed in 2000, and will be extended by another 11km.
A second is the Ceneri, south of Gotthard and now under construction. Other intermediate tunnels are posited. These will eventually put almost the whole line underground.
The Lötschberg axis in the west, an equally deep tunnel but shorter at 34.6km, has operated since 2000. It still awaits full completion. At present a third of its length is limited to a single bore.
Gotthard base tunnel will shave 50 minutes from the journey times between Zürich and the major Italian rail hub at Milan when it opens in 2017, with Ceneri two years later cutting out another 10 minutes.
Around 50 freight trains will use the base tunnel daily, in addition to the 150 on the existing route.
But as they are larger this virtually doubles the capacity in tonnage to 40M.t a year.