The Madrid-Barcelona line (thick blue line; dashed for the section opened today). Enhanced cut-out from Trainspotting Bükkes's Rail map of the Iberian Peninsula

When the Spanish government considered building the 621 km line, it judged that a travel time at or under 2½ hours is necessary to compete with air traffic. This led to a very ambitious goal of running trains at 350 km/h. However, the story of the construction of this line is a compendium of all the faults one should avoid in high-speed rail construction, even if none of them were fatal for the project.

• First, there was the now notorious ETCS Level 2.

ERTMS/ETCS Level 2 is a new all-European train control and signalling system based on wireless technology. In the nineties, the Aznar government decided to equip the line with this system alone, even though at the time, it existed only on paper.

But ETCS Level 2 trials have been plagued with problems for a decade, normal wireless communication was just not reliable enough for permanent use in everyday rail traffic. The latest versions achieved stable performance on some lines (in Switzerland and Italy) only recently.

Several new European high-speed lines have been the victims of the ETCS Level 2 problem, but Madrid–Barcelona has been the first. After authorities finally saw that it won't work any time soon, they had to retrofit the not-yet-opened line with standard signalling. Mere 200 km/h conventional trains took up service on the first section, the 442 km until Lleida (and 39 km of connections to Zaragoza and Lleida) one year late, on 11 October 2003.

AVE/RENFE 252 028 (of Siemens's Eurosprinter family) with a gauge-changing Talgo VII set on the Barrancos de Fuentelices Bridge near Alhama de Aragón (near Calatayud, south-west of Zaragoza), 18 May 2005. Photo by F.JOLLY from RailPictures.Net

• Second, there was geology.

The Aznar government rushed planning and construction tenders, in the course of which certain problem zones were overlooked. Near Zaragoza, the line was built on a plain with dolomitic rock underneath, which is prone to implosions of cavities, and also smaller ground motions. When the problem surfaced in the form of track shifts, a speed restriction had to be instated, and geology had to be investigated in earnest.

The line descends from the mountains towards the Mediterranean on the hilly sides of a valley. This route was chosen despite geologists' warning that the valley side is unstable and moves. The Lilla, Can Magre and Puig Cabrer tunnels (altogether 3.8 km) suffered cracks and subsidence, and had to be stabilised with expensive methods.

An AVE S-103 from Lleida to Madrid on the bridge- and tunnel-rich section along the Jalón river (south-west of Zaragoza), Alhama de Aragón (near Calatayud), 18 May 2005. The duckbill-nosed Talgo 350 trains (rated for 330 km/h) began service earlier that year (on 26 February). Photo by F.JOLLY from RailPictures.Net

• Third, there was the dispute with Barcelona.

There was a seemingly never-ending dispute between the Spanish government, Catalonia, and the city of Barcelona about the route. One major point of contention was whether the line shall pass by the airport (it does in the end), another was about where the continuation of the line towards France should be tunnelled under Barcelona (there are fears that the Sagrada Familia could be affected).

The dispute, and the resulting very messy tendering process (re-started a number of times) caused most of the four year delay in reaching the city. In fact, when the above-mentioned tunnels were repaired and only the city access remained, they first built a station in the fields near Tarragona for the true-high-speed trains, and a facility for the S-120 gauge-changing trains (see previous high-speed rail diary). Thus a slower (max 200, later 250 km/h) direct service to Barcelona started in May 2006.

Station in the green: An S-102 as inaugural VIP train arrives in station Camp de Tarragona, 18 December 2006. From the next day, it was terminal station for non-gaguge-changing high-speed trains in regular service. Photo by Jaume Sellart/Efe from 20minutos.es

• Fourth, there was rushed construction.

Digging on the surface in a city is bound to meet upon unexpected challenges, and tight construction schedules as prescribed by the Aznar government can cause organisational chaos. And indeed, earlier last year, Zapatero's transport minister saw that the construction companies aren't on track to meet their deadlines.

But then Zapatero's minister outdid Aznar's: she decided to push the companies to work full-throttle around the hour. Of course, the result was irresponsible and shoddy work, which led to accidents and further delays. This culminated in the collapse of Barcelona's commuter traffic, after a tunnel implosion damaged a neighbouring commuter line (see kcurie's account: 1, 2, 3).

The final tally for the now opened line is €7,083 million. For the entire line until the border (including the in-construction Barcelona–Figueres–Perpignan/France line), the sum grew from the 1997 estimate of €6.6 to €10.8 billion (though that also involves currency effects). However, the specific cost of €13.5 million/km is still well below that of other new high-speed lines (those mentioned in the previous diary range from €15 to €61 million/km).

Map of the winding Lleida–Barcelona section. The blue sub-sections, including the problem tunnels on the north–south part, were finished earlier, but in interim service only until a point north-east of Tarragona. The red/yellow parts through the airport were built last. Taken from ADIF brochure [pdf!]

But now the nightmare is over. Also, in the meantime, line speed was increased step-by-step:

• 250 km/h from May 2006,
• 280 km/h from 16 October 2006,
• 300 km/h from 7 May 2007.

Despite still not going full-speed (nor any time soon), the fastest schedule with the S-103 trains is 2h38m (check full schedule in via_libre article, one of my sources for the diary, hat tip to Migeru & his contacts).

As a final critical note, I think the Spanish Railways repeats a bad policy the operators of almost every over-budget high-speed line failed with: hoping passengers will tolerate heftier ticket prices. Madrid–Barcelona costs up to €120.40 (though promo tickets on the web cost a third of that).

At any rate, RENFE expects 6.1 million passengers in the first year, growing to 7.8 million in 2011.

:: :: :: :: ::

AGV

Maker Alstom presented a first 7-car prototype of the fourth-generation TGV, re-named the Automotrice à Grande Vitesse (AGV, c. "high-speed self-propelled carriage") on 5 February (also see mention in the Salon).

This Reuters photo from rfi managed to NOT capture a certain politician who loves the limelight

The AGV is intended for regular service up to 360 km/h, and it is the most advanced high-speed train around.

• The train's electric motors are of a new type (the permanent-magnet synchronous motor) that is even lighter while tolerating even higher power than the current state-of-the-art (asynchronous motors). This has been tried only in test trains before: the testbeds for the AGV itself, including the world rail speed record holder, and the Fastech 360 prototype in Japan.

• This is the first high-speed train with driven Jacobs bogies. (A normal bogie is a steel frame with two (or more) axles that can rotate relative to the carbody. A Jacobs bogie however is between two neighbouring carbodies, the end of each resting on separate rotating joints.) This has been tried before only in lower-speed vehicles and again the AGV's testbeds.

I note that company PR still makes much of a supposed relative "stability" of Jacobs bogies, though the 'harmonica effect' in the accident they want to remind of was caused by hitting an obstacle, and the TGV also uses (another kind of) dampers against (another kind of) carbody swings. However, there is a real benefit in reduced weight per train length (12 bogies suffice for an 11-car, 189.9 m AGV, while the 8-car, 200.32 m ICE-3 has 16), wider carbody, more even track loading, and reduced wear.

With two other major innovations, the AGV caught up with rivals:

• Distributed power: instead of traction by two locomotives at the two ends, every second bogie along the train is powered. This allows higher accelerations and greater passenger capacity at the same train length. (The catching-up is with all Japanese Shinkansens, their Taiwanese and Chinese derivatives, the German ICE-3, and it's derivative for Spain, the S-103/Velaro-E.)

• To reduce noise emissions and the danger from side winds, the aerodynamics design moved away from aesthetics. For the nose, that meant that both the front and sides are concave, reminding of a duckbill. (In that, the AGV follows the newer Shinkansens and the S-102, S-130 Talgos from Spain.)

Photo from Alstom press release, where you find eight more photos

To underline Alstom's technological lead: the AGV could have been ready years ago, and the delay wasn't for technology: in the first half of this decade, French state railways SNCF had doubts about the extra benefits in top speed and efficiency being worth the extra price, and wasn't forthcoming with orders; this at the same time Alstom had financial troubles and wasn't able to build a demonstrator on its own. Another sign of being ahead is German state railways DB's inclusion of Alstom in its latest high-speed tender, with a purchase reportedly considered seriously.

:: :: :: :: ::

Check the Train Blogging index page for a (hopefully) complete list of ET diaries and stories related to railways and trains.