France High-Speed Rail

After 30 Years, TGV Service Prospers Even as its Future is Questioned

» France’s high-speed rail network is more extensive than ever and attracts huge ridership — but the financial viability of new lines raises concerns.

Thursday, France celebrated the 30th anniversary of the opening of the high-speed link between Paris and Lyon by then-President François Mitterand, an occasion that redefined travel in Europe and encouraged countries around the world in invest in faster train service by offering train service at speeds above 150 mph for the first time. SNCF, the public national rail company, celebrated this evening at Paris’ Gare de Lyon, where services first originated.

The distinct orange and blue TGV trains that have rocketed through the French countryside at speeds of up to 320 km/h (200 mph) since 1981 have been extraordinarily successful in attracting travelers away from airlines and even the highways because of the quick journey times they offer between center-cities. And they’re supremely safe: More than 1.7 billion rides have been taken on TGVs, fatality-free. Because of France’s reliance on nuclear plants, the electric-powered service has provided the country a low-carbon travel alternative; when considering alternative routes people would have taken without high-speed service, new routes — including construction and operations — are carbon positive* in the long-term.

Because of a series of investments in new dedicated passenger lines, Paris is now three hours from the Mediterranean, two hours and twenty minutes from London and the German border, and one hour and twenty minutes from Brussels. 2,300 km of new lines already under construction or planned for opening within the next decade (see map at the end of the article) promise significant improvements that will bring Toulouse and the Spanish border within three hours of Paris.

The high cost of new rail lines, however, puts in question how much further expansion the French can afford.

Combined with relatively affordable fares, lower travel times have increased ridership on TGV trains in France to about 100 million passengers a year (141,000 a day), more than one and a half times the population of the country as a whole. Though SNCF’s one billion total annual riders is dwarfed by ridership on DB German rail (1.9 billion), France’s encouragement of the construction of new high-speed lines and SNCF policies that push all riders to fast trains, rather than segregating train speeds by the means of individual travelers, have allowed the company to control 50% of the European high-speed market, compared to 22% for DB, 11% for Spain’s Renfe (which has a longer high-speed rail network), and 10% for Italy’s FS.

83% of French people have ridden TGV trains, similar to the percentage of Americans who have flown and far higher than the percentage of Americans who have ridden Amtrak, high-speed or not (less than a third of them).

From its TGV services, SNCF has made significant profits, which have reached €900 million annually in recent years, much of which has been used to cross-subsidize losses on slower Intercités trains serving smaller cities and TER operations offering regional rail.

But the demands by President Nicolas Sarkozy — who recently saidThe TGV, it’s France” — on the national public rail infrastructure owner RFF to build more high-speed lines has changed the equation. Though new routes are usually partially funded by local, regional, and national governments, TGV offerings have been expected to pay back a portion of construction costs (and the much smaller cost of line maintenance) through fees on track usage. RFF covered 28% of the construction costs of the LGV Rhin-Rhône Branche Est, which will open for service later this year, for instance; those costs will be paid back through track fees eventually passed down to passengers on TGVs.

But debt accumulated to build the lines has reached €29 billion for RFF and €9 billion for SNCF; new lines, at a cost of €16-27 million per kilometer, will increase those sums substantially.

RFF has responded to this increase in debt by significantly increasing track fees, and it plans to do so by 40% between 2008 and 2012 — enough to wipe out SNCF’s margin of profitability on the TGV entirely (though the French government has said it would work to stabilize those charges after 2013). RFF will increase fees on the most popular TGV routes the most.

SNCF has responded by threatening to cancel routes with lower ridership (even though they are profitable if excluding the track fees devoted to construction) and it has said the loss of profitability will make it impossible for it to replace the original 1981 fleet of TGVs before 2020. Fares are increasing at 3.4% annually, twice the rate of inflation, and SNCF plans to charge users more on select routes even as it reduces customer service for others to a low-cost model over the next few years. Competition on international routes running through France, expected to begin later this year, will put another cog in the TGV’s future.

Other solutions, such as public-private partnerships for new routes, may reduce the burden on RFF, but they won’t help much for SNCF or riders because someone will have to pay for construction costs at some point. The new LGV Sud Europe Atlantique, to run from Tours to Bordeaux, will cover 55% of its construction costs through track fees, but the project’s PPP partner’s investment in the project will have to be paid back through higher track fees on trains through the corridor (the private investor will also get to keep all profits from the line). RFF’s 14% share of the costs of this project will have to be covered by riders on other lines paying track fees, since all track fees from this route will go to the investor, not RFF.

The irony of charging more track fees to pay for the construction of new lines is a lower degree of service on existing lines and less train travel, since SNCF must cover capital costs with operations profits and higher fares reduce demand.

Standing in the way is the inherent conflict between SNCF’s interests, which revolve around maximizing passenger revenue, and those of RFF, which are about minimizing debt from infrastructure construction and maintenance. These must be harmonized. SNCF’s President Guillaume Pépy has suggested that the current separation of operations from track ownership is a clunky, inefficient model that does not respond adequately to the needs of the railway. Other solutions, such as Germany’s or Spain’s, in which infrastructure is owned by a division of the national rail company, may be less amenable to future competition in services but may make for a less administratively complex system.

The fundamental question is whether France should continue to build new lines (and increase the debt of agencies like RFF), even if that means putting existing services out of the price range of some riders. The government has chosen to pursue that path, but it may not be a solution that is viable in the long term.

Nonetheless, the TGV remains a model for the rest of the world. SNCF has successfully demonstrated how to extend fast, safe, and environmentally friendly rail services to most of the country at prices that most of the population can afford.

Note: In the map above, which is just of high-speed lines, not all French rail lines, LGV refers to “ligne à grande vitesse,” or high-speed rail line.

* Carbon positive in this case means that the line will produce less carbon than would have status quo alternatives.

Image at top: Two TGV Duplex units linked at Marseille-St. Charles, from Flickr user Dmitri Sumin (cc)

57 replies on “After 30 Years, TGV Service Prospers Even as its Future is Questioned”

This sounds like what is going on is like a war game they are expanding a profitble system at a very fast pace and are gaining ridership but at the same time they are taking on too much debut building this vast system. This same process has always happended though out history when they first built the railroads and canals and some toll roads in that they are sucessful so they start building new branch lines and extensions as fast as possible so they take on debut to build it and end up over extending themselves and the whole system goes bankrupt on dubot even though it is profitble at the same time.

What they should do is take a breather for the time being and stop building a lot of these lines on the drawing board for the time being and wait a few years so that they don’t have to raise the track use fees to cover the costs of building the new track. Then after a few years when the debut has been lowered or brought under control they could start building again.

Have to agree Ocean Railroader, Beyond making a connection to Spain and Germany they seem to be extending themselves. Can’t really see the point of going everywhere for the sake of everywhere.

Would love to see HSR committment like that in US but you have to respect the fact we travel just as much by airplane as they do by HSR on a percentage basis. I’m realistic in my thoughts that we will not have the political will for HSR. However, What I do believe is absolute necessary in the states is a strong national network to provide car competitive travel by intercity rail, capacity issues and congested highways is begging for it and some states actually see the reason why 90-110 mph rail corridors need to happen.

Amtrak’s future looks brighter then France in that Amtrak has tons of untapped low hanging fruit such as intercity trains between very large cities that right now eather have very slow or unpreductible service or no train acess at all leaving a great deal of room to expand it’s ridership. France meanwhile though is a small county with a lot less major cities then the US so right now they are building the high speed rail lines to the hard to reach towns and cities.

Even now though it is such a major process to get a new Amtrak line up and running from a rail line that has no passanger rail or even to up grade a existing Amtrak line that Amtrak to get a high chance of getting the project done is going to go after the major rail lines first.

Moreover, let’s keep things in sight: France has 68 millions inhabitants in an area smaller than Texas, over which Paris exert a centrality unparalleled by any American city.

Forget Texas; the best geographies for HSR in the US are the Northeast and California. Both have their major cities on one line or on a simple system of one line with two branches; both have a dominant megacity; both have much larger second cities than Lyon will ever be (the Northeast has 3 separate second cities). In the developed world, the NEC should be the second busiest HSR corridor, behind Tokaido.

I once made analogies between various nations’ HSR systems and various parts of the US: existing Amtrak and proposed HSR.

France’s and Spain’s systems are like the states near Chicago, with Paris = Madrid = Chicago = hub with spoke lines. Even the size scales are similar. The Paris-Lyon, Madrid-Seville, and Chicago-St.Louis distances are rather close: about 300 mi / 500 km.

Italy’s and Japan’s systems are like the Atlantic and Pacific coasts: an approximately linear system with short branches along its length. The Aomori-Nagasaki and Boston-Atlanta distances are rather close: about 1100 mi / 1800 km, with the Turin-Naples distance being about half that.

Germany’s system is like the eastern half of the contiguous US: a 2D network and a rather patchwork sort of system.

France has political considerations to make as well.

Part of the great bargain of whole TGV project included a promise that every major region would be served in the future, and at the same time airports and highways were expected to pay more for themselves via fees and tolls.

Now, the French government can’t just stop projects at leave Southwestern France without any timetable for new services. This is especially important as other services were downgraded and received no major upgrades in lines, stations, signaling so that money could be funneled to TGV construction.

The British experience has shown that it is possible to extract a lot more money from rail passengers if the service is fast enough to compete with cars and airplanes. That might bring other considerations like pricing out the lower third income bracket of rail travel, but higher fares on fast-ish trains are doable.

Why were the early lines able to be built and operated profitably with lowish fares? Were the routes flatter and thus cheaper to build? Was line ridership much higher than is projected for the new lines? Were construction costs much lower and, if so, why did they increase so much faster than inflation?

There are a variety of issues to consider:

» All of the lines that were and continue to be built are operating profitably.

» Ridership on the first few lines was higher per kilometer than the newer lines, because they connected the biggest cities — Paris, Lyon, Bordeaux, Lille, and Marseille (which, with the exception of Paris, are by U.S. standards small, but they’re large enough to attract big ridership in France). Some of the newer lines go to smaller cities like Rennes, Le Havre, Limoges, etc., which have lower potential ridership. (Even so, others, like the lines to Toulouse and Strasbourg, should be equally successful as the initial projects.)

» Though the costs of the initial Paris-Lyon have been paid off over a long period, the other investments continue to weigh down on RFF/SNCF, which is why there is that €40 billion accumulated debt.

» Construction costs may have gone up, but the €15-30 million/km costs for the new lines aren’t so bad in my estimation.

» The question is how much the positive externalities resulting from the construction of the high-speed lines — in economic, spatial, safety, and environmental terms — justify subsidizing construction costs, and the degree to which passengers should be expected to pay for both operations and debt on initial capital.

15-30 million is a very wide range. Does that mean “roughly 15 million per km over flat ground that lends itself to rail construction” and “roughly 30 million per km on more challenging terrain” or “costs over any piece of land could be in that range because we’re never really sure with a massive infrastructure project that will take half a decade to build”?

Well, these are reasonable questions. Here’s a review of recent projects:

» LGV Rhin-Rhône: €16.4 m/km
» LGV Bretagne/PDL: €18.7 m/km
» LGV Est-Euro II: €18.9 m/km
» Contournement de Nîmes et Montpellier: €20 m/km
» LGV Perpignan-Figueres: €25 m/km

The explanation for the high cost of the Perpignan-Figueres line likely resulted from the huge tunnel that had to be built under the Pyrenees. The increase in costs from €16.4 m/km for Rhin-Rhône to €18.9 m/km for Est-Euro II to €20 m/km for the Contournement likely has much to do with the fact that there is inflation; the first was completed this year, the second two will be done in 2016. More minute differences must be compared project-by-project.

LGV Est-Euro II has a (I think) 6 km long tunnel, which adds, of course. It also has, if I remember correctly, extensive connections to the regular network, and some rebuilding/refurbishing of stations (such as Strasbourg).

The contournement de Nîmes et Montpellier also has quite a high density of civil engineering, compared to its short overall length.

Now, for the 302 km long LGV Sud Europe Atlantique, we get to €26.5 m/km, not counting in the €1 bn added by RFF for connections (38 km new lines) and upgrades of the current network.

I think here, the higher number is mainly due to inflation.

There are some questions that need to be asked here, and anywhere high speed rail is proposed:

Is the idea to shift trips to rail from air and highway, or add rail capacity on top on the shaky assumption of increasing travel demand?

Will short-haul air travel be made more expensive and less frequent at the same time as rail is made faster? Will highway speed limits be reduced as rail speeds increase?(In urban transportation, transit success is often largely due to reduced parking availability and/or increased cost. Why not do similar things with HSR?)

Is France subsidizing the competition for rail travel, for example by continuing to expand highways and airports with government funds or loan guarantees?

Are new lines really likely to be carbon positive given that the end of cheap oil, and the resulting economic fallout, will likely greatly constrain air and highway travel?

Are new lines really likely to be carbon positive given that the end of cheap oil, and the resulting economic fallout, will likely greatly constrain air and highway travel?

You’ve got two sides to that question ~ the operating carbon emissions and the capital works carbon emissions.

For infrastructure built to cope with prospective transport demand, the lower capital works carbon emissions of HSR versus highway roadwork for a given transport capacity would suggest that the capital works carbon emissions of the HSR would be lower, even if in retrospect it turns out that the projections were way off.

For operating emissions, even if the availability of HSR permits twice the transport to continue to take place in the middle of more severe oil price shocks than we have yet experience than would be possible with road and air alone, it would still be less aggregate emissions from HSR operations than from half as much oil-based road and air transport. Bear in mind that much of the emissions from the UK study are driven by emissions per kilowatt hour in the UK which are substantially higher than in France, and it still found substantially lower operating CO2 emissions on a per passenger basis.

It has been shown in several cases that with the introduciton of HSR, short distance air got considerably reduced (Paris-Lyon, Madrid-Barcelona, Paris-Marseille, Paris-London). And all that was not “helped” by government actions. Particularly for Paris-London, the price difference between Eurostar and airlines is not that big; in fact, in the lower price segment, Eurostar is more expensive, but there is, of course, the time factor, due to the rather long distances between the business centers and the airports.

In France, one can say that the trunk highway network is built. So, there will not be that much of a subsidy for new routes; expanding may be, but that can also be paid through the user fees (which are non-neglectable. Building up more air capacity can be expensive, and if airport capacity can be gained by the HSR replacement, it can be a cheaper to spend the funds on HSR.

It might logical to think that in a crowed place such as France there is not really going to be much suport for turning four lane and six lane highways into eight and ten lane highways do to the high cost of land. Also most likely between now and the next 30 years they most likely have the maxaum amout of cars and car owners that are phiscally able to exist around this time frame based off of the high costs of owning a car in Europe in general. So unlike in the US their great ages of highway widening and new highway routes most likely are over for the time being unless something happens. The US meanwhile still has pently of growing traffic and there are still a good number of unfisnihed highways and highway widening projects out there waiting to happen. Also land is cheaper to buy in the US for highways then it would be in France.

@ Ocean – Good point about the price of land. In some places it might make sense to close or downsize regional airports, sell the land, and use the money to build HSR or even just upgrade conventional mixed passenger and freight rail. I think you could find cases both in North America and Europe where this would make sense.

This is also a possibility for the land under some aging freeways and interchanges in urban North America that need to be demolished regardless of what replaces them. (Part of the land under the Turcot Interchange and connecting freeways in Montreal for example.)

My take is that things have changed significantly in the last 30 years, and that significant changes in thinking are needed.

That (closing airports or reducing highway speeds to force people into rail) would be abhorrent. Low-cost airlines made many regional airports explode in traffic.

Those flights do not compete with the same TGV routes, but offer options for travel to/from other European places where rail travel isn’t and will never be feasible on a realistic basis (Toulouse-Stockholm, Dublin-Marseille, Paris (Beuvais) – Athens etc.

Freight rail in Europe is generally negligible, priority for passenger trains being one of the reasons (proximity with ocean ports being the second main one).

Priority for passenger trains has *nothing* to do with the lack of freight service in Europe. You can tell this because Russia has priority for passenger trains and has a gobload of freight. (Arguably the lack of freight service has led to greater priority for passenger trains.)

Why doesn’t Europe-west-of-the-former-USSR have much railfreight? First is the prevalence of waterways, second is the national border crossing problems (border delays for freight were huge until *very* recently and transnational freight *still* must handle many signal systems), third is the archaic coupling system (making switching a pain and imposing short maximum train lengths), fourth is the tunnel and bridge clearances (preventing large container loads)…. add up all of this and you have plenty to explain the lack of rail freight.

The coupler situation in particular is ridiculous. The EU is trying to standardize signalling, and I hope after it does that it will standardize on a modern coupler.

Not in Europe, where the main lines are double-tracked. The issue there is maximum weight imposed by minimum speeds, both on grades and to avoid interfering with passenger traffic too much.

It is a whole bunch of reasons why you don’t see “super-heavy, xuper-long, super-slow” freight trains in Europe. There is freight, quite a lot of freight, actually, but there are those limitations.

For international freight, the border crossings are still a pretty sad issue; there is Schengen, but that’s for people; there is some common treaties for freight, but customs have still to be passed when crossing borders, and they are still not fully in the information age. The situation is improving, and at least for intermodal services, things are quite fast.

The next issue is different signalling systems. This is gradually overcome with multi-system locomotives on one hand, and the gradual introduction of ERTMS, even if it is on the lowest levels.

On the other hand, changing the locomotive at the border is not such a big deal, if it is done right, and doing a full brake check is also a check of the train, to spot problems. The issue is more that there are some operators having/causing issues, making rail simply unacceptable for shippers.

The couplers are only partly an issue, because there is also the limited axle load, preventing really heavy haul. There are, however, dedicated solutions where very heavy trains are run, and those block trains do have automatic couplers.

In fact, there were agreements and trials for an automatic coupler in the early seventies of the last century. Some railways did invest quite a bit, such as the Swiss Federal Railways, but some of the “big players” were too cheap and unwilling to work on their fleets. So, the whole thing lost momentum, and we still have the old style couplers…

The consequence is that for the normal network, we now have rather fast short and light trains. Which is not bad at all in a mixed operation, because a decent freight train can more or less keep speed with a local, therefore being able to use a same slot. In fact, I have encountered that some time ago in Germany, south of Offenburg on the 4-track, where an intermodal ran parallel to the local I was on, and it was able to catch up to almost the same car after a station stop… kind of fun…

So, (western) European rail freight is essentially light and fast, even for bulk.

Not in Europe, where the main lines are double-tracked. The issue there is maximum weight imposed by minimum speeds, both on grades and to avoid interfering with passenger traffic too much.

Dualling doesn’t eliminate the necessity of sidings. They remain crucial for operation and their length defines the maximum length of trains.

The main reasons for Europe’s low rail freight use are entirely geographical. Coastal shipping (not inland waterways) is one reason; additional reasons are shorter shipping distances and a different, truck-friendlier mixture of goods shipped from in the US (e.g. the US hauls much more coal).

Yonah linked to a study a few months ago subtracting such geographical issues from Europe’s freight modal split. This turns out to account for a majority of the difference: if the US had had the same above-mentioned factors as in Europe, its rail freight mode share would’ve been 15% rather than 37%. The difference between this 15% and Europe’s 7% share is attributable either to passenger dominance or the poor operating practices mentioned by Nathanael.

California HSR is viable largely because it stands to steal so much of the short-haul traffic in the state. Most of the volume in and out of Chicago is short-haul flights. If half the flights to O’Hare and Midway were replaced by HSR, you’d be able to close Midway. You’d also have something beyond HSR to Cleveland, or Detroit, or St Louis; you’d have regional trains stopping at small cities and towns, which would give them a degree of connection comparable to much larger cities. Connect those smaller places and they can reach each other and the larger cities too. That drives down the need for getting on the road, which reduces the need to accommodate so much car traffic. Same goes for trucks.

“Discount” carriers typically exist because they use a “cheaper” airport. Why is it cheaper? In many cases it was largely paid for by the traffic on the traditional airlines. No thought is given to essentially handing over this amortized value to airlines, when they did nothing to earn it and routinely impose costs on the public sector. One of the best examples of this is Porter Airlines in Toronto. Porter doesn’t use Pearson, which by Federal mandate has some of the highest landing fees in the world (which are used to subsidize operations at other Canadian airports). Porter uses Downtown (a.k.a. Island) Airport, the continued existence of which is an encroachment threat to the Toronto Islands, as well as a very real infringement on the use of waterfront land to the north, northwest and northeast of the airport. Porter probably wouldn’t exist if Downtown Airport were closed.

This whole explosion of “discount” carriers is kind of embarrassing, as well as being a distraction. Are they better-run businesses? In some cases, like Southwest, yes they are. But when your business model is predicated not only on an absence of labor cost but an absence of infrastructure cost (not paying landing fees at major airports, or choking the airspace with so much traffic that one delay ripples across a continent,) you’re extracting your efficiencies from the public purse. When Bostonians travel to Manchester or Providence to fly Southwest, somebody has to pay for the space on the freeway for them to do so. The choice to not have true HSR between Boston and Washington leaves a warped transportation market, where the private sector forces costs on the public purse in the form of more crowded airspace and more crowded roads, when all of that travel and more could be accommodated on trains, often at a lower cost. Yes, I would like to see some airports closed and some roads maybe narrowed, or at least not so full of traffic. If there’s going to be any kind of subsidy, let’s be clear and honest about it.

France isn’t really a crowded place. If it were a US state, its population density would rank 9th, behind Florida. In the UK, common wisdom is that France could build the LGVs cheaply precisely because it’s not very dense, whereas England has much less room.

A mid-sized French city typically mightn’t sprawl as much as a mid-sized North American one, which might make land acquisition easier and somewhat decrease opposition, due to there are fewer people effected (although I doubt French farmers are willing to sell their land, I guess the payments there would still likely to be less than the payments in suburbia).

I have a map in GMaps similar to the one above. If you’re not familiar with how the LGV lines link up with the existing network, you can see that by zooming in on one of the segments labeled “branch towards”. I don’t have a good map for the LGV Est II, so the route for that line is simplified.

The land acquisition is pretty cheap. The process is as follows: SNCF determines the route (might be RFF right now, I’m not sure). The farmers yell that it’ll put them out of business and demand outrageous changes. SNCF then approaches the farmers individually, and offers above-market compensation. The farmers sell eagerly. On the way, if the route slices through plots, SNCF will organize land swap deals.

Even in suburbia, there are usually enough ROWs that land acquisition is cheap. If you look at the proposed alignment of HS2 (available online if you follow links from Wikipedia), you won’t see any obvious reason why it costs so much.

I’d thought that Autoroutes (expressways) in France were tolled, is that not still the case (or only partially true)? Would that then make a difference in this case, i.e. tolls helping fund infrastructure improvements?

Nearly all autoroutes are tolled, and many are privatized.

However, airlines frequently get local subsidies by promising jobs and playing regions against each other. For example, Marseille pretended not to notice that Ryanair was violating French labor law at its base there, and complained once the national government noticed and told Ryanair to comply with wages and benefits requirements. (In general, Ryanair is as expert at extracting local concessions as Wal-Mart.)

Cross subsidies are well possible with vertical separation, but they have to be moved from railway operation to either:

1) infrastructure manager who could e.g. charge more for highly utilized lines at peak times and correspondingly lessen the fees in off-peak times and on less used lines


2) at department of transportation level when logical set of trains/lines is put to bid and bidders compete for least subsidy/highest fee within defined parameters of service

I like approach 2) most because it can provide best service with probably the highest level of transparency of public money. Note that 2) is well established in German Verkehrsverbünde where it also allows integrated schedules & fares regardless of operator, something that is usually said that only integrated railway udertaking can do.

At least in Germany your #2 only happens for regional/urban rail, which are subsidized. The long distance/high speed travel is supposed to be a completely free market and operationally profitable, and basically 100% is run by DB at this point.

The contracted urban/regional rail is run mostly by DB, with a 20% market share of private companies. So what happens is that DB gets rid of less profitable long distance lines (even if they are still profitable), cutting of smaller regions from long distance rail. Then the states have to come in and contract out and subsidize fast regional operations to replace that long distance line, which DB Regio is fairly likely to get.

So there are some people who say that the desire to create a free market for long distance rail has failed, with basically 100% market share of DB for hsr in Germany; with DB ‘picking the raisins’ and leaving the rest for states to subsidize — And ask that all rail operations be put to tender.

The existence of long distance but slow inter-regional service was an anomaly. Where signaling is modern and the networks dense, there is no reason, whatsoever, for the existence of train routs 400km long with 30 stops along the way.

Long distance frequent stopping trains do please a small share of passengers that take advantage of one-seat rides, but they are, in dense networks like that of Germany, an operational anomaly in which smaller stops get long-distance trains at the price of regional/local travel.

So DB is not cutting regions out of long-distance travel, with timed transfers, it is merely requiring a change of trains, often meaning a large majority of relations can be reached faster rather than slower.

This transformation will have losers, though: namely relations involving two small stations on a former single route now requiring 2 transfers.

Your assumption is rather wrong. These Interregio services were actually extremely popular for two main reasons. They reduced the number of interchanges which sped up journeys for quite a lot of people. And they were reasonably priced.
Most rail travellers in Germany are now worse off because of the decisions of the profit-driven Deutsche Bahn.

I don’t think it is correct to maintain these select cheaper Interregio services when they reorganized their network with much better, and faster, ICE services (and incidentally more expensive).

The number of passengers travelling faster on multiple connections (R + ICE + R) is far larger than those worse off in terms of travel time.

Well, the problem is that the long distance system is shrinking – today’s ICE routes are a subset of the former Interregio+ICE routes. While you can call that market-driven rationalization; I’d call it the desire of DB to cash in on subsidies by forcing states to replace former long distance routes with subsidized regional rail.

On a personal note, I rather like the regional express (RE) brand. It usually runs at 120..160Km/h, so much slower than ICE, but also much cheaper. On the weekend you can group flat rates and it makes cheap student travel very possible – albeit slow.

I don’t think, unless there is a market for it, that rail companies should provide money-losing long-distance but slow direct trains.

I can understand the rationale of having a subsidized Wiesbaden-Frankfurt rail service. However, there is no justification why people should be able to travel on subsidized trains from Bremen to Dresden, for instance.

Commuting might be something worth subsidizing, long-distance travel, not sure about.

But there is no real market, only a quasi-monopoly. And we’re not talking about obscure direct services, but rather about cutting off mid-sized cities from long distance travel altogether. That’s why within this current ‘market’ it may make sense to put long distance operations to bid, rather than continue trying this basically failed market.

ICE services are barely faster than Interregios, they are just more expensive. They run on the very same tracks.

You have a complete misconception of Germany when you assume that the ICE network with a few feeder services would be faster. The majority of Germans lives in mid-size and small towns. They have virtually been cut off from long-distance rail travel by Deutsche Bahn.

By checking the general schedule of DB, I highly doubt medium-size cities are deprived of meaningful rail service. That is not the case at all.

Cities like Fulda, Ingolstadt, Göttingen and Augsburg are served by frequent ICE stops.

Then, you have cities like Magdeburg in eternal complaint of having been left out high-speed lines that bypassed them. But they will often have frequent and reliable regional/IC connections to nearby rail hubs.

What DB needed to do is to push even more ICE Sprinters, long-distance non-stop trains that give real advantage on time travel over ICEs and make a matrix of travel times to have lower mean values.

connections to nearby rail hubs …

… where they waste valuable time by waiting for their connecting train.

Travelling via major hubs is slower in most cases and therefore a change for the worse.


As a french, I’m quite honoured to see an article about our railway politics in this website !

I just wanted to have you noticing that you forgot a project on your map : the “Ligne Nouvelle Montpellier-Perpignan”, which is the missing link between the French and Spanish highspeed networks !

This project is to be build far before the lines in Normandy or in Provence (Marseille-Nice) that you’ve drawn on your map !

Anyway, thank you for the article !

Yonah ??? Surely Lupus is correct, that it’s an oversight not to include the full Montpellier-Perpignan link to connect Spain with the TGV system?

I believe the Spanish are working to complete the connections from Barcelona to the French border as well.

It wouldn’t make any sense to spend heavily getting thru the Pyrenees without a link to the main TGV system coming fairly soon.

Yes, the Montpellier-Perpignan link is planned; you can see it in RFF’s list of new lines. This also includes the line from Paris to Orléans to Lyon and the Interconnexion Sud of Ile-de-France. The reason I didn’t include these projects is that a recommended alignment has not yet been selected. From my understanding, they will not be completed — and perhaps won’t even be in construction — before 2020.

Montpellier-Perpignan is not as critical, as it is not too long, and allows decent speeds already. But it would be good to have it available for high speed, for sure.

They say that Figueres-Girona will open next year, and that would then allow to connect through to Barcelona, shortening the trip time from Lyon to Barcelona to about 4 hours.

I had the impression that the increases of track fees were at least partly to keep DB Fernverkehr off French tracks, i.e. good old fashioned protectionism. I thought in an earlier article you described that RFF contracts the management of the tracks to SNCF anyway, possibly reducing some of the impact of increased track fees.

The closer integration of tracks and operations in Germany is usually heavily criticized there, because it can lead to monopolistic strategies. For example DB Netz (Network) can increase track fees, and give profits to DB Holding, which can be used to cross subsidy the operations that are now unprofitable because of the high track fees. This is a way to push smaller, private companies out of the market.

That is why I think railway holdings should be really broken up and train operation companies forbidden to enter the infrastructure market (and vice-versa).

To be clear, yes, a major percentage of track improvements are contracted back from RFF to SNCF, but the French system is a broken-up one: SNCF and RFF are independent entities. And they have a major conflict of interest.

In fact RFF was splitted from the SNCF. And by the time the two companies splitted, RFF kept the biggest part of the SNCF debt.
This was a “trick” to make the SNCF look more “competitive”.

I had a notion that a large part of the current spending on TGV extensions was a big part of the French stimulus program.

Some financial considerations may have been brushed aside in the haste to get the spending and employment efforts underway. Nonetheless, I have the feeling that the French stimulus program, including these new HSR lines as well as public transit projects in various cities across the country, will turn out to be a good thing.

In the short run, it will provide a big boost to employment in the on-going economic slump.

In the long run, these projects will leave the nation with valuable assets purchased when construction costs were relatively low (compared to the future costs) and the cost of financing (interest rate on government debt) was also very low.

Leave a Reply