JUST 10 years from now, travellers will be taking high-speed trains under the Alps. The mountain range that stretches from Zurich to Milan and dominated the history of the continent for thousands of years will become nothing more than a long gap between stations.
This year, work will begin to drive a railway tunnel defiantly through the base of the mountains. Stretching 57 kilometres between the Swiss towns of Erstfeld and Bodio, the Gotthard Base Tunnel will be the longest and deepest in the world, taking passengers more than 2 kilometres down, to where heat from the Earth鈥檚 depths raises the temperature to nearly 50 掳C.
Going under the mountains will cut an hour and a half off the journey from Zurich to Milan, and dramatically reduce road freight traffic in Switzerland鈥檚 narrow valleys. And if the forecasts are right, the tunnel should also be surprisingly cheap, at just a quarter of the 拢11 billion budget for the 53-kilometre Channel Tunnel.
Advertisement
But taking the travelling public deeper than they鈥檝e ever been taken before is a tricky business. Nobody knows exactly what kind of rock lies hundreds of metres beneath the surface. Or how to ventilate such a long and deep tunnel without vast investment. And should anything go wrong, nobody knows the best way to get people out.
Since the Swiss people voted for the project in a referendum in 1987, this last question has moved into sharp focus. In the last four years there have been five serious fires in alpine tunnels. All were started by incidents involving heavy goods vehicles in congested road tunnels鈥攈azards that won鈥檛 apply to the new rail tunnel. But in each of those fires, poor evacuation procedures were to blame for making a bad situation worse.
Building the tunnel is the job of AlpTransit Gotthard (AG), a subsidiary of the government-owned Swiss railway system. It will comprise two tubes bored 45 metres apart, connected by cross tunnels every 325 metres. Unlike the Channel Tunnel, there will be no additional safety tunnel, and that has some commentators worried. 鈥淚 prefer a small evacuation tunnel and I always look when I take a tunnel to see if it has one,鈥� says Franz-Josef Ulm of MIT, who investigated the damage caused by a serious fire in the Channel Tunnel in 1996. In that incident, passengers escaped via the safety tunnel, and there were no serious injuries.
Christopher Kaue, AG鈥檚 safety officer, says that if there is an accident in one of the Gotthard tubes, people can easily escape via the other tube鈥攁t least, once the trains have stopped running. 鈥淚t鈥檚 possible to get people to a safe haven within minutes,鈥� he says. If a fire breaks out, Kaue says that, ideally, a train will continue to one of the two emergency stop stations along the tunnel, where exits and lift shafts will be ventilated and pressurised to keep them free of smoke.
But that is the only active ventilation planned for the tunnel. In another controversial decision, AG will rely on the 150-kilometre-per-hour winds created by the trains to keep air in the tunnel cool despite the hot rock it passes through. Alan Beard of Heriot-Watt University in Edinburgh, who studied the 1987 underground fire in London鈥檚 King鈥檚 Cross tube station, says he hopes that AG鈥檚 ventilation plans work. If they don鈥檛, simulations of how quickly fires would spread could also be unreliable. 鈥淚f a surface is already warmed up, it鈥檚 easier to get a fire going,鈥� he says.
In a fire, trains would have to stop to allow passengers to escape, stopping the ventilation. So if a burning train can鈥檛 reach the nearest emergency stop station, passengers will have to walk down a hot, unventilated main tunnel to the nearest cross tunnel, which could still be many kilometres from the nearest evacuation shaft.
From statistical analysis of previous disasters, AG predicts this kind of scenario would only happen 鈥渙nce every 3000 years鈥�. But not all experts think AG鈥檚 approach is valid, and Kaue admits the figure is based on very limited data. 鈥淭here haven鈥檛 been a large number of accidents in Europe,鈥� he says.
Geology presents AG鈥檚 engineers with another set of unknowns. But they have learned the dangers of skimping on research from the experience of workers on the Seikan tunnel under the Tsugaru Strait in northern Japan. That project took 17 years instead of 9 because too few pilot tunnels were dug and the main tunnel flooded because it proved to be too close to the sea above. Today the tunnel relies on pumps that run 24 hours a day. If they fail, the tunnel will flood within a week.
Floods may seem unlikely in the Gotthard Base Tunnel, which will run 500 metres above sea level. But workers got an unpleasant surprise last year when they were confronted with pressurised jets of sand and water from the water table, as the pilot tunnel they were boring through solid granite suddenly ran into a kind of sand called sugar-grain dolomite. 鈥淚t was pretty nasty,鈥� says Herbert Einstein of MIT, who works on logistics for the project. Luckily further pilot studies show the dolomite doesn鈥檛 take up much of the planned route. 鈥淭hey have better conditions than that, though not as good as they hoped,鈥� says Einstein.
As engineering projects become more ambitious, the potential consequences of any disaster worsen and the cost of safety precautions rockets. Active ventilation or digging a safety tunnel could cost so much that the project would become economically unviable. And experts don鈥檛 agree on the best way to make a tunnel safe or on what level of risk is acceptable.
With safety standards not even attempting to keep pace with future designs (see opposite), it鈥檚 increasingly being left to the companies that finance and build new tunnels to come up with their own answers to these questions.