I remember standing 800m down at the bottom of the Sedrun shaft when it was first finished,” says Heinz Ehrbar, AlpTransit chief engineer.” I was thinking about how we meet up with the next sections. In this small deep space how do we even know which direction to go in?”

The answer was found of course and with astonishing accuracy. The final breakthrough from Faido to Sedrun on 15 October was just 80mm off horizontally and 10mm vertically. All the previous breakthroughs in the different sections were also well within an allowable maximum error of 250mm horizontally and 125mm vertically. “They did a fantastic job,” says Ehrbar.

Achieving that tested the limits of modern surveying instruments and the teams using them, both those working daily with the contractors and on the TBMs and the AlpTransit Gotthard’s own small oversight team of four. The client worked with consultants from the VI-GBT, “Konsortium Vermessung Gotthard- Basistunnel” made up from Grunenfelder und Partner, BSF Swissphoto, Studio Meier and Gisi e Bernasconi.

The Sedrun shaft was one of the major challenges for the survey team led by Adrian Ryf, the head of geomatics at AlpTransit Gotthard. It certainly needed something more than a glance at a compass, which would not work there anyway.

“We had three things to do at Sedrun. One was to transfer coordinates down the long shaft accurately. The second was to determine and maintain direction and the other is to measure the height,” says Ryf.

Contradictorily the team first of all turned to a traditional method, the plumbline. But the modern means of doing this, an “optical plumb,” proved difficult at first because there was fog in the shaft caused by the humidity. The tiny red LED light, which is the target below, could not be seen.

So the traditional physical line was used. But it was a plumb on a grand scale, dropping the full depth of the shaft on three steel wires. Steel plates were used at the bottom, weighing around 370kg, to form a pendulum bob.

“It had to be left overnight to settle down” says Ryf. The base remained swinging very slowly but averaging the end points gave a centre point, theoretically transferring a coordinate from the top which was measured relative to fixed survey points at the end of the short 1km tunnel above.

But it did not work precisely. The problem is the Alps themselves, which are such a large mass of rock that they exert a gravitational effect, or to state it another way, they make the earth’s shape irregular.

The plumb does not hang straight towards the centre of the earth but anything up to 30mm off line, a big error.

Fortunately says Ryf, there is long experience of this issue in Switzerland and a big database of the geophysical characteristics of the mountains. The Swiss Federal Office of Topography even sells a specialist software to allow the correct position to be interpolated. The accuracy achieved on Gotthard has helped confirm the accuracy of that database, too, he says.

Once the physical plumb was complete, it was found that the ventilation could be reversed to achieve clear air in the shaft, allowing the optical method to be used too. “Results were very similar,” says Ryf.

But the optical method also needs the corrections calculated as, just as with any other surveying instrument, work begins by calibrating it against a vertical plumb.

In the second task at the shaft, obtaining the depth, things were a little easier says Ryf, since modern optical distance measurement is highly accurate and it was necessary only to measure vertically.

But then came the question of direction for the tunnels. To find that from plumbline transfers would need a longer distance between two points than the 8m or so at the bottom of the shaft, or even the 35m that became usable once a second shaft had been finished.

Instead the team turned to the latest gyroscopic direction finding instruments. These were first used for checks in the Channel Tunnel but have improved since.

“Inside a vacuum a small gyroscope spins at a very high speed, around 20,000rpm. Its rotation axis will align with that of the earth, which gives you a very accurate direction.” Though again, he points out, gravitational corrections must be run.

Such instruments are not much bigger than a total station, he says, though they are expensive at around EUR 120,000 (USD 167,000). Instruments were used that are owned by the ETH Zurich Technical University, where Ryf was a former lecturer, and one at Munich University.

The gyroscopes were also used to correct readings along the tunnel made with total stations, primarily, since this is Switzerland, Leica instruments. An error noted on the Channel Tunnel was that of small refraction effects near the tunnel walls, caused in that case by the colder air layer there and noticeable only on very long distances.

Here the rock was hot but just in case measurements were regularly taken in the centre of the tunnel “when it was clear of activity such as during holidays,” says Ryf.

An even newer technology, the Inertial Measurement Unit, was also tried out. “The IMU is a very accurate version of the orientation detector in the latest iPhone. It detects and measures accelerations and rotations. This was done for the first time to this level of accuracy in civilian application, though the military have used it.” It requires some complex calculation, however, he says.

Outside the tunnel a permanent network of reference points set up in 1995, and used for the tunnel survey, was rechecked in 2005 by the client team, using simultaneous GPS measures at 28 separate points. Leica supplied the GPS systems though Trimble is also used.

The exercise was needed not only to apply the latest technology and the better accuracy it could achieve but to see if tectonic movements of the Alps had changed any points. The Alps move around 1mm annually and relative movements between its parts can also have an impact.

Results showed the reference points to be sound and unchanged.

Another task for the survey team outside has been to monitor three significant reservoirs lying close above the route of the tunnel. Settlement effects or rock water loss could endanger the structures, even at the great depths of the tunnel.

“We installed an automated system on the dams and in their valleys with GPS and total stations” says Ryf. Each night the total stations have taken measurements from prisms at the side of the valley to check for any convergence or divergence. A telephone link sends it back to the consultant BSF Swissphoto Regensdorf near Zurich, where the data is processed automatically and sent to AlpTransit, which operates a special website for the monitoring.

“We have collected an enormous amount of data since 2001,” he says. Researchers are finding this useful though so far no dangerous movement has been found.

One other task carried out by the surveyors has been to establish a series of accurate survey reference points in the tunnels once they are lined. Set every 50m on both sides of the tunnel and accurate to 1mm, they provide a reference for the trackwork, which is just beginning.

Accuracies of just millimeters were achieved in breakthroughs Adrian Ryf was the survey head for AlpTransit Gotthard These concrete arch dams above the tunnel line were constantly monitered – Photo Swissphoto Readings in the tunnels had to be corrected for gravity effects – Photo Swissphoto