Over most of the 20th century, the rail industry was very national in character. Development projects were usually led by (nationalised) railways, who often involved all major makers in one way or another (say, combining the best elements of rival producers' prototypes). Technology, especially high-speed, was more – say – Made in Germany than made by Siemens, Henschel, AEG or Talbot. Symbolic for this was the practice that train type families were usually named for their domestic version (e.g. "TGV").

To be precise, the rail industry was national in developed countries: developing and third-world countries were export opportunities. And most of these areas were like marked territory for certain exporters: South America for the USA, (former) African and Asian colonies for France, the Balkans and the Near East for Germany; later respective client states for the USA and East Bloc countries.

Thus significant export competition in the rail industry is relatively new (and still of limited importance). Even the outsourcing of manufacture is of limited extent compared to other industries.

Advantage TGV: the nineties

The emergence of high-speed rail was first to ignite more open competition. And in that field, Made in France reigned supreme.

TGV derivatives were sold to international operators (Eurostar, Thalys), Spain (AVE S-100 and S-101), AMTRAK in the USA (Acela Express), South Korea (KTX-1), and were intended for the Texas Triangle (a project later killed by then governor, now outgoing US President George W. Bush).

In the same time, the export successes of rivals Made in Germany (ICE), Japan (Shinkansen) and Italy (ETR 500) numbered exactly... zero. Well, apart from Made in Italy, with Pendolino tilting trains sold to seven countries; but nowhere did these run at the 250 km/h they do in Italy. On the other hand, on the (next emerging) European locomotive export market, it was an inverted picture with Made in Germany reigning supreme.

At the same time, the European rail industry went through a brutal structural change.

This was the shift from, in the example on top, Made in Germany to made by Siemens. The transformation eliminated a lot of traditional companies, either by merger or bankruptcy. The industry in formerly 'communist' states was particularly badly hit, as their domestic markets de-facto disappeared for a few years from 1990. But this was the opportunity for the remaining West European giants to outsource.

The new free(er)-market environment was (and is) plagued by incoherences: complaining customers who didn't get what they wanted, producers strained by tight delivery schedules and requirements changing even during production, and disputes and miscommunication between subcontractors, subsidiaries and company centres.

Cut-throat Competition: The Zeroes

Towards the end of the nineties, the most coveted price for high-speed rail exporters was the Taiwan High Speed Rail (THSR). For this, the main contractor for the French TGVs, Alstom, wanted to play it safe: it allied with Siemens. This was a big thing: in almost every field of railway technology, France and Germany originated two different and usually incompatible philosophies, and the technology firewall was guarded anxiously. The project to produce a joint European high-speed train (to realise economies of scale) foundered on this – twice.

However, for export, there was readiness to create hybrids. Already in Spain the previous decade, the TGV-derived AVE trains ran on track and with signalling made in Germany. For THSR, Alstom and Siemens, after becoming preferred bidders, created the hybrid Eurotrain: the more powerful tractor heads of two ICE 2 sets were attached to the lighter middle cars of a TGV Duplex set.

The Eurotrain at München-Laim. In a demonstration run on 4 May 1998, it easily achieved 316 km/h. Photo from André Werske's HSR site

At the very end of 1999, the preferred bidder Europeans were in for a shock: THSR went to a Japanese consortium. They filed (but lost) a lawsuit against the awarding, then filed (and won on the final instance) another for compensation, which heralded the coming of bidding duels until the last legal instance as regular competition practice in the industry. In the next decade, Alstom's financial voes worsened (though the worst problems were in non-rail sectors) until the state had to save it from bankruptcy in 2004.

Meanwhile, junior bidder Siemens Transportation itself just climbed out of heavy losses. Then it got into heavy trouble with two products: the Combino trams (fissures in the aluminium body), and the S-103 high-speed trains for Spain. The latter were derived from the German ICE 3. However, in the spirit of the new cutthroat competition, the other contractors of that Made-in-Germany joint production refused to supply parts or sell licenses, so Siemens first had to re-develop the missing elements in-house...

Travelling this bumpy road, companies became providers of complete trains. Lately also in Japan. Symbolic for the change was also the appearance of brand names for train type families (for example, Siemens christened its ICE 3 derivatives "Velaro", Bombardier introduced "TRAXX" for a popular electric loco previously denoted with the class number at the German Railways).

A review of the changing fortunes on the high-speed ( >= 250 km/h) export market in this decade shows Alstom ahead overall only because it acquired Pendolino maker Fiat Ferroviaria, but Siemens's Velaro is the most-sold family, and lots of others gained a share:

• Siemens: Velaro to Spain (S-103, 26 sets), China (CRH3, 60 sets), Russia ("Sapsan", DC/AC multi-system EVS1 4 sets, DC-only EVS2 4 sets)

The newly delivered first EVS1 Sapsan is towed from St. Petersburg to Moscow across the station of Tosno, Leningrad Oblast, 22 November 2008. The Velaro RUS, like its Chinese sisters, is wide-body. Unlike in other Velaros, the front-end passengers don't get to peek above the driver's head. Photo by Konstantin Menshikov from SpbRailFanClub (where you can also see pictures of the two older Russian domestic-production 200 km/h trains)

• Alstom: TGV Duplex to Morocco (frame agreement for 18 sets) and Argentina (as part of an an entire project, 8 sets); AGV to Italian private operator NTV (see here and here, 25 sets)

• Alstom (ex Fiat Ferriviaria): non-tilting Pendolino derivatives to Spain (S-104, 20 sets, S-114, 13 sets), China (CRH5, 60 sets)

• Bombardier: Regina to China (follow-on order on the slower CRH1, 20 sets), ZEFIRO also to China (the world's first high-speed sleeper trains, 20 sets)

• Hitachi+Kawasaki: Series 700 Shinkansen derivative to Taiwan (700T, 30 sets)

• Kawasaki: Series E2 Shinkansen derivative to China (CRH2, 60 sets)

• AnsaldoBreda: V250 to the Netherlands and Belgium (19 sets)

Prototype of the V250 for NS Hispeed on exhibition in Berlin at the InnoTrans 2008. The train, though tested at Velím in the Czech Republic last August, is long delayed due to numerous problems. Photo by Coen Ormel from Flickr

• CAF: RENFE S-120-derived trains to Turkey (HT65000 "Hızlı Tren", 12 sets)

• Hyundai Rotem: KTX-II derivatives to Turkey (no firm order but ensured by a factory built)

Now, the last two should really bug Alstom.

Technology Transfer: the (not just) Asian Challenge

Builders in China, Japan and South Korea tended in the past either to develop technology for purely domestic use or to work with technology imported from one of the three big international manufacturers – the German-based trainmaking operation of Canada's Bombardier, France's Alstom or Germany's Siemens.

Now trainmakers in Asia's three biggest economies benefit from rapidly increasing domestic rail investment, using it as the springboard for an export drive that leaves many in the European heart of the industry uncomfortable.

It was stupid of the FT to lump Japan with the other two (Made in Japan already got as far as Spain in the past), but massive domestic investment is a significant point. The article then goes on to claim that the Asian competition is appearing just when Europeans do the same, which I can't agree with (it doesn't appear on the same level at least), but more on this later. What bugs Alstom's boss is, first, new competition using exported technology:

Mr Mellier says that, after designs are bought from Hitachi or Kawasaki of Japan – both of whom have supplied technology used in Japanese Shinkansen trains to China – or Bombardier, the technology is "made Chinese".

Alstom's boss cunningly infers rivals should be concerned, but exposes how his company is affected, too:

Korea's market is closed. Hyundai Rotem, the country's biggest trainmaker, has begun to produce its own high speed train after a technology transfer deal by which it used Alstom's TGV design for Korea's 300kph KTX express.

The fruit of years of testing with the HSR-350x prototype (which I showed here): roll-out for the first Rotem-developed and -built KTX-II on 25 November 2008. Photo from Korail News

It's not only the high-speed train, but Rotem managed to establish a foothold right at the door of Europe, in Turkey, where it built a factory and netted a number of major orders recently.

What's more, it's not just the Asians. For its exports to Spain, Alstom teamed up with CAF of Spain as junior partner. But, CAF developed its own platform, first by tilting the balance and having Alstom as junior partner (variable-gauge S-120, S-121), then selling Turkey an entirely in-house production.

But, as Mellier makes explicit, Alstom has the idea on how to deal with the problem of nurturing rivals:

companies ... should not sell their latest technology.

(My bold) Indeed that's Alstom practice:

Koreans have a 300kph domestically produced, single-deck train, but Alstom markets its double deck, 320kph Duplex and 360kph single deck AGV.

But, that may still not be enough. Rotem already announced its next prototype, and there is also the example of the other Spanish maker, Talgo.

Talgo built its Talgo 350 high-speed train (AVE S-102 and S-112 at Spanish Railways RENFE) in cooperation with Bombardier: they produced the low-floor middle cars (a high-speed first) on their own, while Bombardier helped out with the power cars (based mainly on those of the German ICE 2). But last September, Talgo shocked rivals by announcing the project Talgo AVRIL, an in-house development that would beat the present products of all rivals on speed (more below), weight and comfort.

Concept image showing the Talgo AVRIL's duckbill-nosed power car, which now has a passenger compartment and equipment moved underfloor, and a car-joining Jacobs bogie. There will be more non-low-floor cars with motorised bogies in the middle of the train, a strange configuration they named "salad". Image via Alta Velocidad

Now back to Mr. Mellier, who is most concerned about China, with its strong demands for technology transfer:

"They will use them, adapt them, aggregate them to [form] a Chinese technology based on foreign technology being leased by them," he says.

Indeed if you look at Alstom's main success in China, electric locos from its PRIMA platform (180 HXD2 double and 500 single locos), the Chinese co-producer's product page won't even name Alstom. Also, when China invited bids for high-speed trains, Alstom was so cautious as to not even offer the TGV, only the Pendolino (and got to supply the CRH5).

And indeed rival Siemens would have a few stories to tell. In 2004, it was temporarily expulsed from the high-speed train tender, until it agreed to fuller technology transfer. Its place in a follow-on tender was then made hostage to the Tibet/Olympics attendance dispute with the German government. After supplying the Transrapid for the Shanghai Maglev, it discovered that a Chinese maker made an attempted copy; and a dispute over technology ownership also contributed to the death of an extension project last December.

Protected home markets?

Now Mr. Mellier isn't solely concerned about technology.

"I sincerely believe that all the many tenders for rolling stock and signalling will be for Chinese companies and the access for non-Chinese companies will be kept to a bare minimum."

Well, Siemens thinks otherwise: it is in talks about 100 more CRH3 trains. [UPDATE: on 16 March, the order was placed with the Chinese partner, Siemens supplies the electronics & bogies for c. 18% of the total order value.] Still, as a trend, on one hand, this may be a fair description of Chinese policy. For example, a follow-on order for the CRH2 was not for Kawasaki but only its domestic partner. On the other hand, I think it is deviously misleading for those unfamiliar with the specialities of the industry. FT may think that it is special that

None of these Asian countries [China, Japan, South Korea] has in modern times allowed the import of a wholly foreign-built, foreign-designed train.

...whereas this is quite normal in the developed world, where almost all imported trains have significant (most often 50%+) domestic added value. The demand on foreign suppliers to team up with local manufacturers may be explicit or implicit, but it is pretty much the norm.

Politicians are often motivated by securing or creating jobs. However, there are some practical benefits, too. Trains aren't replaced every three years, they are a long-term investment – and for maintenance, locally available expertise and spare part production capacity is a big plus. (Being scrapped after the first major breakdown is the usual fate of more advanced train exports to the Third World.)

Then again, look at FT's examples:

Japan's Hitachi is building its first European order, a series of 225kph trains for high-speed domestic services between London and Kent in the UK, Europe's most open rail market.

One of Southeastern's class 395 "Javelin" trains runs at 225 km/h from Ashford to London-St. Pancras on a 13 December 2008 presentation run. The class 395 is based on Hitachi's A-train platform. Photo by Jamie Wiseman from Daily Mail

Two China State Railways subsidiaries won an order to supply three 200kph trains to Grand Central, a British company, in 2007, while CSR Nanjing Puzhen was on a shortlist to supply 200 diesel train carriages to the UK announced on December 22.

Hyundai Rotem was on the same British shortlist and has supplied trains to the Athens metro.

(My bold) The FT's implication being, Mr. Mellier may be protesting too much on the still protectionist Continent, while the UK's shiny free markets show the true competitiveness of his products vs. the Asian rivals...

Well, while I'd agree that Alstom's boss is protesting too much, I think there are some other factors here. One is that no major UK-based train-making company remained (all factories are foreign-owned), another is that train operators considering cheap bids from makers with few references are playing a risky game.

The Dutch and Belgian state railways made a really bad bet when going for the cheapest offer with the V250 (see above) – but not as bad as Danish state railways DSB. They also went for cheap when choosing the same maker, AnsaldoBreda, to supply the 200 km/h IC4 trains. The first of those entered service on 1 December 2008 – almost five years late!

Thus, in countries with significant domestic rail technology, it's not just simple protectionism (and, heh, corruption) that is at work: familiarity with the technology and work style, and trust in existing supply partners and product lines has a strong role, too.

As an example to counter the FT's implication, there is the world's largest trainmaker, Bombardier (which is Canada-based, but its train technology is centred in what used to be Germany-based ADtranz). The company supplied major parts for TGVs, ICEs and Talgos as junior partner, but its own project for a high-speed train, the ZEFIRO, never got off the ground here: no European railway took the risk of trusting in something existing on drawing boards only.

In fact, it seems that railways in these countries use international tenders merely to make their longtime suppliers more accommodating. As a nice pair of recent examples, the head of France's SNCF suggested that Siemens Velaros do have a chance in 2007, and the German DB indicated last year that it might buy Alstom AGVs – but in the end, SNCF ordered 80 improved TGV Duplexes (apparently named RGV 2N) in June 2007, while DB ordered 15 Velaro Ds in November 2008.

On the other hand, the European makers have problems with their domestic markets, too. Back to a point I flagged on top: while China and some others invest massively in railways, the Europeans minus Spain... not so much, especially in the high-tech (high-speed) sector. The size of the above mentioned DB order is meagre. And both SNCF's and DB's ordered trains aren't cutting-edge.

Alstom's ambitions in particular were long hampered by SNCF's unwillingness to order the AGV: they preferred the lower price and reliability of the TGV (and the higher capacity of the double-deck Duplex) over (relatively minor) improvements in performance.

Raising top speed

Still, the race is on to offer ever faster trains.

China itself announced its intention to develop 380 km/h trains, I devoted a full diary to my rather sceptical analysis. After listing the challenges, I even noted that I don't know if even the CRH3 (Siemens Velaro CN) truly do the announced world's fastest service at 350 km/h; and insinuated that Chinese authorities may ignore some safety precautions.

Now I know more: according to German Wikipedians, the trains are certified for only 300 km/h, but authorities just allow them to go "overspeed", with 340 km/h regularly achieved.

So, notwithstanding unending European problems with ERTMS/ETCS Level 2, 350 km/h is on the horizon. Intent on joining the club is Rotem of South Korea (with the government-sponsored Hemu-400x project, I reported), and – despite unmet expectations with the recent FASTECH 360 prototypes (see here) – Kawasaki of Japan (efSET).

Concept drawing of the efSET. Note how the sidewind-resistant duckbill nose was evolved into something more streamlined again. Image from Kawasaki Heavy Industries

Alstom's AGV is aiming for even higher: 360 km/h. Alstom announced that during trials on 13–14 December last year, the AGV reached 364.4 km/h, and finished tests necessary for the only 300 km/h version to be delivered to NTV. However, I am not sure what to make of this: for type approval, you need tests at +10% (i.e. 396 km/h), so I don't know if they plan to do those at a later date, or would do them only after the reception of firm orders for financial reasons, or encountered some problems.

Finally, there is the above mentioned announced threat to everyone from Spain, the Talgo AVRIL. To cut Madrid–Barcelona times to 1h45m(!), the targeted service top speed is 380 km/h. With Talgo's expertise, this should be considered with more seriousness than China's ambition – but not without doubts. Talgo named its current top product Talgo 350, but it is certified for only 330 km/h (with insufficient power and ride quality among the likely reasons).

:: :: :: :: ::

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