Here’s one potential pan-European high-speed rail map, incorporating existing and likely future high-speed lines in France, Spain, Britain and Italy; the lines I’ve argued Germany should be building; and plausible and semi-plausible extensions into Eastern Europe.
Here’s a small version of the map:
For full-size 56 MB link, click here. Blue lines exist or are under construction, red ones are either under planning or proposed solely by me or by local activists.
The Polish network is fairly optimized, but the rest of Eastern Europe isn’t, relying on long-range international connections that may or may not flop due to a possible international trip penalty. I only took it up to a point, so yes, there’s that link via (North) Macedonia and Kosovo, but I drew the line at some point and did not add a line from Warsaw up the Baltics and under sea to Helsinki; the Baltic capitals just aren’t big enough, and the light at the tunnel, Helsinki, isn’t big enough either.
Note also that some cities gain through-tracks on this map that they don’t currently have, especially Paris. This is to be a four-track system connecting Gare du Nord to Gare de Lyon and Gare de l’Est to Gare Montparnasse; since there’s no chance of building the main station under Les Halles this side of the 1970s, the station would have to be at a somewhat skew location relative to city center, most probably around where Gare du Nord and Gare de l’Est are now. Additional cities with notable through-tracks: Milan, Rome, Munich, Florence; Madrid gets through-tracks but those are already under construction as part of the third Cercanías axis, at typically low Spanish costs, and Marseille gets through-tracks as is the plan for the mixed classical/LGV system for Provence.
The trip times are always net of station dwell times and short timed connections at major junctions, so they can be added across the map. In Germany I sat down and figured out frequencies, running consistent stopping patterns every half hour; this doesn’t work Europe-wide, as some places are too low-density and have to make do with hourly patterns, like Eastern Europe (and, if it keeps its baroque fare system, Spain).
Which Spanish branch line would connect to the Strait of Gibraltar crossing, if such a thing were ever constructed?
I… did not think about it. The proposal I saw was to go well to the west of the strait to avoid kilometer-deep water, and in that case the line has to hook up via Seville, not Málaga.
I haven’t looked at Great Circle distances. How many people are close enough to the Straits, to want to take a train from Spain to Morocco or vice versa. There will be airplanes.
Ferries currently exist on those routes plus Madrid has a strategic interest in linking Ceuta and Melilla (only one of which has an airport inside its territory) to the mainland…
Which doesn’t answer the question of how many people want to go between Spain and Morocco by train.
If you are worried about there being an airport within the territory that implies there is some sort of political problem between the Spaniards and the Moroccans. Ferries work better than trains when it comes to avoiding each other. Ask elderly Berliners how well having rail connections worked when there is a political problem.
WIkipedia says there are 85,144 in Cuerta and 86,384 in Melilla. How much money do you spend on 171,528 people. Who may or may not want to go to mainland Spain often. A question you haven’t answered.
As Spain is the former colonial power of both parts of Morocco and Western Sahara, there is obviously a potential for conflict between the two countries. Add to that that Spain continuously accuses Morocco of trying to “get back” (technically Ceuta and Melilla have been Spanish longer than Morocco existed) the two cities, their inhabitants not wanting to go to the Spanish mainland should raise all sorts of alarm bells among so-inclined politicians in Spain. Besides, both cities have one representative each in the Spanish parliament which only has 350 seats total which makes them over-represented by about a factor of 2…
Note that none of this changes finances or geology, but those are some very convincing political reasons to keep subsidizing ferries and the airport/heliport in either city and they might some day be good reasons for an otherwise rather pointless tunnel….
Ask older former West Berliners how well having rail connections work when there is a political disagreement. I’d much rather have ferries. … ask Danish survivors who made it to Sweden during the war.
East Germany kept the S-Bahn long after it ceased making them any money because it gave them an important foothold in West Berlin. France is investing in its former colony so they can build their HSR system…
For many years southern Spain (‘Andalus’) was part of the muslim empire – as was Marocco. There is still a cultural bond between these countries – even if it was surpressed during the fascist rule.
Since Marocco is the most progressive country in the arab cultural region there is undoubtly a big demand for a rail connection – both passenger and freight.
But I suppose this is beyond the comprehension of people that have never left their own tiny bubble…
TKO / Stockholm
How much is “big” ? Herbert argued that there are two enclaves on the northern shore of Africa. How many, give us number not something like “big”, of high speed rail seats 171,528 people fill a day.
The economies of Ceuta and Melilla are very dependent on laborers who commute daily from Morocco; both countries want to keep the flow of people and goods moving, and have no interest in closing off the enclaves (unless you’re a Sub-Saharan immigrant).
However, Morocco accounts for 9% of Spanish exports and 5% of imports by value. Imports are mostly agricultural products, so probably a larger % by tonnage. Therefore a rail connection accross the Strait could be very interesting from a freight point of view – once there is continuous standard gauge to the French border, the entire EU single market will be open to Moroccan fresh produce.
On the other hand, passenger traffic accross the strait is very seasonal: there is a huge peak in the summer, when hundreds of thousands of Arab-Europeans from France and beyond drive all the way to Tarifa/Algeciras, and cross by ferry to spend their vacations with their families. Perhaps night trains with carriages for automobiles could attract part of that market. Providing one-seat rides between Madrid/Andalusia and Tangier/Ceuta could also attract year-round weekend trips.
Moving produce by rail can be done, we do it in North America, transcontinental. Building a two track freight railroad across Spain to avoid a ferry sounds expensive. And then across France because all the all the comments I read about it say it sux. The freight has to move in days not weeks because the lettuce gets a little limp after ten days or so.
If we can’t do electric trains we will need to have high speed rail links between Europe and Africa, because high speed rail is the only vaguely viable alternative to long distance flights.
Even if I’m stuck at home for the next year tourism is still going to be important.
There will always be long distance flights. You can fly airplanes using bio fuel and it’s already been certified. Or you can make synthetic fuel, the industrial chemists have been fooling with that for almost century. Dump carbon and hydrogen, from almost any source in one end and almost anything you’d want comes out the other. Or pay for carbon capture and sequestration. There will always be long distance flights.
This is the first comment after comment number 69…
At any rate, Spain’s high speed network is already standard gauge as is the entire rail network of Morocco. Morocco is already building a domestic hsr network. Now for freight alone a train ferry might be an alternative, but I think a number of reasons why a tunnel could have a business case twenty, thirty years down the line have already been pointed out. Note also how long distance flying has plummeted into an abyss. Who’s to say it’ll return in the same way? And synthetic fuels are technologically feasible but horrendously expensive, even if you derive them from coal – ask east Germany, apartheid south Africa or others…
Biofuels are flawed as they need massive amounts of land. Aside from cross Atlantic flights which would be difficult to replace with high speed rail why isn’t that generally the better option?
10 percent of the gasoline I buy in the winter is ethanol made from maize. They might be using other things but most of it is made from maize. When most of the road fleet is electric there will be plenty of capacity.
Corn based ethanol is a make work project for the Caucus State Iowa. It makes neither economic nor ecological sense
The subject was whether or not there is enough land, not if makes sense. it is heavily subsidized but the subsidies aren’t so high we would stop flying if they disappeared. And the reason we don’t suck carbon dioxide out of the air and make hyrdogen from water and run it into a sythetic fuel plant is that it’s cheaper to grow things than it is to do that. And any of it is cheaper than figuring out how to run high speed trains to Denver.
As long as oil derived fertilizer stays cheap…
The Norwegians make fertilize using excess hydropower. The process to make synthetic ammonia to make synthetic fertilizer isn’t particularly picky about the source of the hydrogen it uses. But then you could just be injecting the hydrogen into a synthetic fuel process directly. … there is always going to be airplanes.
Fixed nitrogen is only one of the three essential plant nutrients and the easiest one to provide in large quantities…
I know what the numbers on the bag fertilizer mean, too. Fixing nitrogen with a variant of the Haber-Bosch process is what most people have in mind when they consider the fossil fuels in synthetic fertilizer. There are people around here who are happy to give me a few buckets of um um organic nitrogen for free. Makes the compost heap vent “steam” real nice. It’s good thing, I can tell it is using the enormous quantity of carbon I dumped in it along with the buckets of stuff I dumped in for it’s nitrogen. Don’t get that sorta kinda right it goes slimy and stinky.
I believe if my calculations are correct the Orient Express in this version takes around 12.5 hours. Shows how far trains have come since the initial almost 3 day time table in the 1880s.
My only serious question at this time is if it would not be worth it to build a more direct connection between Strasbourg and Stuttgart to speed up Paris-Munich? Or do the Mountains make a tunnel not worth it?
Considering that you have repeatedly criticized current official HS2 plan, why there isn’t through-running in Birmingham?
What happens to LGV Interconnexion Est when there is through-running in Paris?
Is Kraków-Lviv viable?
Nice pan-European network, Alon, better then the awful TEN-T maps the EU makes…
Some quick notes on Italy:
-Florence through running station and tunnels are under construction for 3/4 years, but construction was halted for several reasons and a project reassessment. The Belfiore HS u/g station is here: https://goo.gl/maps/Bm4WcqJYKWBkqYLo9, 1 km from the existing one and poorly located for connections with local s-bahn like system (but there is a tram stop). There is an advanced proposal to link the old and new stations with a people-mover (airport style). More serious people propose to build a stop for local trains next to the HS u/g Belfiore, but there are actual geometrical difficulties.
– A through running tunnel for Milan, with a u/g station in the big square in front of the existing Milano Centrale, has been on planning documents for a while now. Maybe one day it will be built, depending on traffic. Rome already has a through running station quit centrally located: Tiburtina. Italo used to stop just here, but people preferred Termini, so they switched. Maybe a tunnel with a u/g Station under Termini itself could be a nice investment in the far future to improve the whole network performance.
-The whole triangle in Sicily (so also the Catania-Palermo inland route through Enna that you’re not suggesting) is goindg to be built at 250 km/h standard. Some sections are already u/c. There is a 10y horizon and a 5bn budget for the whole triangle (both existing and upgraded lines at 250/200 km/h).
-The Florence-Genoa-Nice-Marseille line is going to be ridiculously expensive. The existing XIXth century line is already mostly in tunnels and partly 1 track east of Savona. A HSL will probably be built 5-10 km inland (the coast is either heavily developed or protected areas), so 90% tunnel, the rest viaducts. East of Genoa part of the line has already been moved inland and double-tracked.
-Salerno to Reggio Calabria will be another tremendously expensive line, mostly in tunnels (geography is extremely rugged on the Tirrenian coast of Calabria). Calabria is the poorest Italian region (GDP par capita is less then half Lombardy or ER), not really populated. There is potential for tourism and already now some HS train continue south along the existing double truck line. The only reason to built it would be if it is to continue on a bridge or tunnel to Sicily. The bridge was a major highlight of Berlusconi platform. Even some early works were performed. It has been put on hold, because the budget was 10bn and counting and was to be connected to crappy infrastructures on both sides. Maybe a train-only tunnel would come in a more reasonable budget.
-The whole Adriatic coast rail line is going to be progressively upgraded to 200kn/h operation, at least from Bologna to Ancona or Pescara. There are no major cities, but a continuous built stretch and it is already quite an important market for both work/leisure travel (the biggest today out of the main HS network).
How would the economics of the connection to Dublin work? It’s 100+ km, more than double the distance that the channel subtends, and it carries the same international penalties as the chunnel to a way smaller metro area than Paris.
Same language, though. London-Dublin air traffic is comparable to pre-Chunnel London-Paris air traffic, around 4-5 million per year.
Twice the distance twice the cost and while it’s an interesting exercise, use the money someplace else that serves more people at lower cost. All destinations do not have to have rail. That’s what airplanes are for.
All London airports were having capacity problems as of 2019. Replacing replaceable flights would alleviate that. And it would cut down emissions. Also: the relationship between length and cost is not linear. Modern TBMs should actually get cheaper the more they’re used…
London Dublin is the busiest city pair in what was the European union late 2019…
Interesting plan. Some things that struck me in the parts of Europe I’m familiar with…
Utrecht to Groningen: This one is my biggest issue. I don’t really see the point of a 45-minute journey, which I assume would be non-stop considering you need an average speed of well over 220kph and getting into and out of Zwolle is really slow. Neither Utrecht nor Groningen are particularly big; significant value from this route comes from the population centres and timed transfer opportunities at Amersfoort and Zwolle. Major value also comes from traffic from Rotterdam and the Hague, which don’t seem to get a connection under this plan, despite the Rotterdam-Utrecht journey being pretty slow (this would also allow connections from Belgium to the northern Netherlands).
Rotterdam/The Hague in general: Connections to this area seem quite poor, especially given it’s more populous than Amsterdam.
Amsterdam to Brussels: It already takes 75 minutes to get from Amsterdam to Antwerp and you’re not proposing any new lines. How did you arrive at this time?
Northern England: you should consider incorporating ‘Northern Powerhouse Rail’ / ‘High Speed 3’ / whatever they’re calling it thesedays. It’s planned to be quite well-integrated with this network and probably makes timed interaction/coordination that bit harder to plan in Northern England.
Rest of England: can’t help thinking this misses some important flows, particularly in the London – Birmingham – Bristol/Cardiff triangle. But that’s obviously not a huge issue.
Murcia–Valencia–Barcelona; Nantes: same issue as above
I think Utrecht-Groningen is 1-stop (at Zwolle), not nonstop?
Amsterdam-Antwerp has Thalys slowness (why would you ever run locomotives on a line with this many stops?) plus a 4-minute dwell at Rotterdam.
Rotterdam-Utrecht can work fine as a timed low-speed connection.
Why are you connecting Groningen to Utrecht and not to Amsterdam? I’m asking because there has been a lot of planning for a HSL-Noord, Zuiderzeelijn, Lelylijn, which would run Groningen – Drachten – Heerenveen – Emmeloord – Lelystad – Almere – Amsterdam Zuid. This would add a lot of new connectivity for internal intercity traffic and such a line would be better used than the one you drew, because it doesn’t parallel an existing line. The existing lines in the North have plenty of capacity, so I’d rather add a new connection.
Amsterdam Zuid is supposed to become the future (international) high speed rail station by the way.
Utrecht – Amsterdam Zuid on the existing 4-track railway takes 22 minutes with max speed of 140km/h, which includes a 2 minute stop at Bijlmer Arena station, which already has passing tracks. The speed can be increased to 200km/h, so then a 15 minute time should be feasible. So I don’t think a new line is necessary there.
I do like the line to Germany via Venlo. The line to Germany is always assumed to have to through Arnhem (like the current ICE), but combining Amsterdam – Utrecht – Ruhr and Rotterdam – Eindhoven – Ruhr in one line makes sense, also considering the services that should be run.
Doesn’t Utrecht-Amsterdam have capacity problems? That’s why I was assuming new tracks, but upgraded tracks work too if there’s enough room on them.
I was conflicted between Amsterdam-Groningen and Utrecht-Groningen. I picked Utrecht because its location is more central, so there are convenient connections onward to not just Amsterdam but also Rotterdam and Eindhoven.
Utrecht to Amsterdam Centraal has capacity problems for sure, because of the overstretched line from Bijlmer to Amsterdam Centraal. I don’t think Utrecht to Amsterdam Zuid is that bad, especially by Dutch standards; it has four tracks and only stops in Bijlmer, which is probably below capacity at the moment. The third and fourth tracks were quite a recent addition IIRC.
Firstly, I feel my previous comment understated how impressively comprehensive and detailed a plan this is.
Re Utrecht to Groningen: I just don’t see the case for a full-blown high-speed line on this route. It currently supports two* 120′ mid-length intercity trains each hour which maximise connections on a half-hourly takt with the one-hour mark at Zwolle (I hope this unwieldy sentence is comprehensible!). The route is quite slow and you’re right that this could and should be improved. But building an alignment for the speed required is going to require significant compromises which ignore the many connections and thus limit its (already low-ish) ridership.
Amersfoort is a good-sized city in and of itself (larger than Zwolle), and also useful for connections to het Gooi to its west (pop. 200k wealthy people who travel a lot). These are roughly timed to pulses at xx:30 and xx:00. Your plan instead requires either doubling back to Utrecht or a 3-seat ride via Zwolle.
I also can’t help feeling this misses the point of Zwolle as a hub for timed connections. The trains into Zwolle are timed to maximise connectivity around a pulse every xx:15 and xx:45. This makes sense because there isn’t really a primary flow through Zwolle – but your plan means that the (Amsterdam C?-)Utrecht-Groningen flow is prioritised over everything else, with what must be a short stop at Zwolle. This short stop means that trains to/from Amsterdam Zuid and Flevoland (Almere/Lelystad), to/from Twente, and to/from Leeuwarden will have to wait for more time before and after the high-speed train to allow for transfers, slowing down these journeys. This slows down the majority of the passenger flow through Zwolle.
As I mentioned, the current approaches into Zwolle are also pretty slow for geometry and points reasons. If you used these, it would add a few minutes onto the journey and require the rest of the line to be even faster to compensate. One solution would be to bore a straight tunnel under Zwolle, but this is both ludicrously expensive and makes the connections outlined above worse. Your solution seems to be to bypass Zwolle with what I assume has to be a station on the outskirts, but that’s really bad for the aforementioned connections (this might not actually be the case; sorry if I’ve assumed wrong).
* = as in, one through train and two half-trains which peel off at Zwolle
Frequency depends on speed, no? I’m possibly thinking too much about the 10-minute frequency between Amsterdam and Eindhoven, with every third train requiring a transfer; if trains run that fast to Zwolle and Groningen then plausibly the supported frequency will be very high too, every 15-20 minutes, and then the timed connections don’t necessarily need to be timed with international connections. (For the same reason, the maps I’m producing for the Northeastern US don’t have a knot in New York, only in cities where trains only come every half hour.)
So your solution for Zwolle would be turn-up-and-go frequency on the high-speed line and maintaining a timed pulse (whether at xx:15/xx:45 or some other time; you’re right that international synchronisation doesn’t matter here) for the other flows? It could work. Would it mean that all trains on the high-speed line have the pattern Utrecht – Zwolle – Groningen, or would some of them only do half this route and go to other destinations? And would all trains continue to Amsterdam, or could some go to Rotterdam and/or the Hague?
The line from Amsterdam to Eindhoven and Limburg is a nice comparison. However I would say it’s just an inherently busier route than Utrecht to Groningen; it has a greater density of potential passengers along its whole length, including a lot of shorter trips between the individual cities. This route is really optimised for high-intensity service, given it has four sizeable cities each 30-50km apart and a number of close, economically integrated mid-sized towns at either end. It’s 80km between Utrecht and Zwolle and 90km between Zwolle and Groningen, so this line doesn’t have the same density. There’s also not much past Groningen to run or connect onto, unlike past Eindhoven where trains can run on to the Limburgish towns and pick up more passengers there.
I just feel that perhaps adding 15 minutes to this journey – allowing a slightly less expensive alignment, reliable half-hourly pulsed connections at Zwolle, and possible stops at Amersfoort – might be a better use of the large amount of money and political capital it would take to build this line.
Why connect Gare du Nord to Gare de Lyon ? Lille and Lyon are already connected in 3 hours through interconnexion est with reasonable frequency (8 trains a day) and both Lille and Lyon have the potential for many connections/extensions to northern Europe and southern France/Europe Sure, if you want to do trips like Arras – Dijon today you probably need to go through Paris as frequencies are so bad, but there is a lot of room for improvement here.
Gare du Nord – Montparnasse would make more sense. For Lille Bordeaux, going via Paris is currently much faster (4:15 vs 4:45/5:45) and offers much better frequencies than the only 2 direct trains. Projected Interconnection sud wouldn’t improve significantly journey times since it would add a stop in Orly. With a tunnel Nord Montparnasse, Lille Bordeaux could be done in less than 3:30.
Anyway, it looks as unlikely as the tunnel to Ireland…
8 trains per day is not reasonable frequency.
It is for Coburg…
I meant reasonable relatively to other transversal routes like Lille Bordeaux or Strasbourg Rennes/Nantes with only 2 trains a day and many radial routes run by SNCF. Otherwise I agree, we should have 1 train per hour on Lille Lyon, same than what currently have Paris Lille and Paris Lyon, so no need to connect gare du nord and gare de Lyon.
Would love to know what the travel demand is like between countries in the Balkans. The distances are long enough that through service won’t get much use (8h30m Milan-Istanbul vs a 3h flight) so this would depend on a lot of people traveling between within the former Yugoslavia, Hungary, Greece, Romania, and Bulgaria. Right now, there are 6 buses a day each direction Belgrade Zagreb (5h30m), and 4 daily trains Ljubljana Zagreb (2h5m, but also IIRC not very full and heavily subsidized); don’t think an hourly train would ever run full but maybe full enough?
That train would be overnight.
There are still some trains carrying private vehicles from central Europe to Turkey that don’t even have revenue stops in the Balkan…
The Madrid high-speed interconnector has supposedly been under construction for 10 years, although I can’t find info as to whether it was ever completed. I find it hard to believe that high speed trains will use a Cercanias tunnel, as they have different gauges.
In fall 2019 there was heavy construction on the Cercanìas trunk with temporary shutdowns…
It was not completed; it’s the third axis, planned for completion in the new decade.
Josh, Herbert, Alon: I think you’re talking about three different projects! To clarify:
There was work in the Recoletos tunnel (built 1933-1967), which was never planned to have traffic as heavy as what it has today. Works included complete substitution of catenary (from classic to rigid), track (from ballast to slab) and signalling (ASFA to ASFA Digital). This is one of the current Cercanías trunks, used by C-1, C-2, C-7, C-8 and C-10 trains. It has Iberian gauge track and 3kVDC electrification.
You might be thinking of the future Cercanías east-west tunnel between Canillejas and Príncipe Pío, with intermodal interchanges at Avenida de América and Alonso Martínez. As of 2020, that tunnel only exists on paper.
This is the third north-south tunnel between Atocha and Chamartín, built for high-speed trains: standard track gauge and 25kVAC electrification. It’s basically finished, but the opening date has been pushed back over and over again with no given explanation. The last announced date is June 2020. It’s currently semi-operational, used only for transfers of rolling stock between the Northern and Eastern HSLs.
Has the new underground station for high speed trains at Atocha been put on hold?
Why would it?
It wil be built in a later phase. According to current plans, trains from Valencia/Alicante/Murcia will use the tunnel, without stopping at Atocha, and stop at Chamartín – either terminating there or running through to cities further north. But there will be little traffic until the new trains arrive 2-3 years from now to relieve Renfe’s chronic shortage of rolling stock.
Atocha is way closer t the center of Madrid than Chamartin which is pretty away from most things of interest either for business or leisure travel…
You’re mostly right, but: there’s also significant business density along Paseo de la Castellana (mostly between Nuevos Ministerios and Chamartín) – Nuevos Ministerios and Plaza de Castilla are the 3rd and 5th busiest Metro stations, after all. Plus there’s massive TOD planned between Chamartín station and Fuencarral.
Part of the motivation to send trains from the Eastern HSL to Chamartín is avoiding capacity problems at Atocha, especially with the jump in traffic that liberalization is expected to bring.
Why would liberalization increase traffic?
The main obstacles to the growth of HSR ridership are Renfe’s chronic shortage of personnel and rolling stock, and high fares pushing price-sensitive travelers towards flying, driving and buses. The entry of new operators into the Spanish market will definitely help with the former and (hopefully) also with the latter.
In the long term, rerouting the Madrid-Barcelona HSR to enter the city from the north could make a lot of sense. Right now, there is very asymmetric demand, with the overwhelming majority of traffic entering the city from the South. However, this connection would balance demand almost perfectly, with all trains from Catalunya running through to Andalusia or Lisbon, and all traffic from Valencia/Alicante running through to the Northwest. Moreover, this would only require 24 km of new track between Torrejon de Ardoz and Charmatin, almost all above ground, and allow for a station at Barajas Airport T4.
One of the original plans was for the Barcelona line to enter Madrid from the north, with an intermediate stop at the airport. It would have indeed made a lot of sense, but it was shot down in the environmental review stage. IIRC it had something to do with the impact of tunneling in the Jarama river area.
I’m skeptical that almost all could be done above ground in the future, for several reasons: most of theland is already developed now, you’d still need ~8 km of tunnel under the runways, and I’m not sure the existing ROW between Torrejón de Ardoz and Chamartín can be easily 6-tracked.
You’re right about needing to tunnel under the runways but that seems worth it to allow all trains to run through Madrid and make Barajas Airport a short trip from almost all of Spain.
If we can dream of through stations in Paris, does it still make sense to have track on either side of the Rhine between Strasbourg and Basel?
The map is not correct for the left side of the Rhine, no high speed railway is under construction or even planned, there is only a legacy railway that allows speeds up to 200 km/h.
A high speed line would make sense though, it would accelerate trains Lyon Strasbourg and leave more capacity for freight and regional trains. Yes, TGV from Lyon to Frankfurt or Stuttgart could take the legacy railway Mulhouse Freiburg for around to 50km to join the new line Basel Karlsruhe but that would mean skipping the largest city in the area, Strasbourg and not much faster.
By the way, there is a another small mistake for France. There is no high speed line between Dijon and LGV Sud-Est so it should be in red not blue. A project exists (Rhin Rhône Ouest) but no construction is planned so far.
The Dijon-Sud Est branch is blue because I’m proposing to run it on existing track at low speed. Blue is any line that exists or is under construction, even if it’s not high-speed, hence for example Switzerland. Hence also the lower average speed proposed on the Paris-Dijon city pair.
Kind of? It’s a pretty dense area, and Strasbourg-Basel is already pretty fast – it’s not greenfield HSR, and to a large extent neither is the line on the German side.
https://www.youtube.com/watch?v=QqZCk-t0r9E this German language YouTube video goes a bit into detail about the German line from Basel to Mannheim…
The German line has been overwhelmed by traffic decades ago and expanding takes ages. This is right in the heart of the “blue banana” and the shortest way from northern Italy to the north sea ports…
Rail Baltica and the Tallinn Helsinki tunnel have two things going for them: EU interest/funding and established cross border connections. Helsinki is the busiest destination out of the nit very busy Tallinn airport and there are loads of ferries. It’d also establish a land route bypassing Russia which has geostrategical advantages.
Why did autocorrect replace “not” with “nit”? What does that even mean in English?
Nit? It’s the egg of a louse. Removing nits from your hair requires very careful combing, hence nitpicking. But about 100% of modern usage of nitpicking is metaphorical.
Anyway, why don’t they (Rail Baltica and the Tallinn Helsinki tunnel) show up on your map?
Would the Oslo-Copenhagen market have anything “on the way” to make it viable? Also, why don’t Danish and Finnish domestic hsr plans show up on this map?
Just checked, of course there would… The busiest non-capital airport in the Nordic countries and a city important enough to get a spur out of Stockholm in your map…
By the way, Oslo-Copenhagen is busier than Oslo-Stockholm (though not by much, and the difference may be explicable by ground based competition or price). But Oslo’s three busiest routes are all domestic. Why are none of them on this map?
Stavanger, Trondheim, and Bergen are small cities. The reason Norway has so much domestic air traffic is that with the terrain and city sizes it wasn’t cost-effective to upgrade any of these city pairs to what Sweden has.
I think at two million pax p.a. It should at least be considered. After all, Norway has insane oil wealth now which means a problem in the future…
The terrain in Norway is seriously challenging, probably insurmountably so in the case of Oslo to Bergen.
Norway has quite a bit of experience building ridiculous road tunnels (I think both the deepest and the longest are or are planned to be in Norway) and they are virtually all refinanced through tolls. Officially the toll ends once the original cost has been “paid back” and there indeed some tunnels in Norway that have become toll free due to this. I’m not sure how honest the accounting is, but to me it seems Norway would be one of a very few countries with the required tunneling expertise and they would probably not be shy about asking Austria or Switzerland for help if need be. Norway is quite aggressive in its targets to make domestic transportation carbon free and unless there is an innovation in electric planes right around the corner that I haven’t heard of, this further shifts the business case in favor of HSR for Norway… Especially since Norway has the problem (to a lesser extent but still) that faces the Gulf Oil Monarchies: What when the oil runs out. It appears however, that political will is quite low as the stub line from Oslo to its airport seems to have been deemed a financial failure…
The quality and quantity of Norwegian bridging and tunnelling is indeed something to beyond. In the tinest northnmost places you can see that the oil money was spread around, and that durable things were constructed (well, sea level rise will wipe away lots of it.) And the costs are surprisingly low. Quality, Swiss-level concrete porn abounds!
And they’re still at it!
Not sure about Finland, but the Danish plans are not HSR. I’m depicting medium-speed trains in Switzerland because otherwise it would be really awkward to depict high speed lines out of Zurich and Basel but nothing connecting Zurich and Basel, whereas Aarhus is out of the way; I could add it to the map with the hour concept, but that would raise awkward questions like what about connecting it with Hamburg.
https://en.m.wikipedia.org/wiki/Copenhagen%E2%80%93Ringsted_Line there’s this…
What do you think of a HH tunnel?
https://en.m.wikipedia.org/wiki/High-speed_rail_in_Finland this article indicates there was some movement under the Rinne government. I don’t know what his successor did with those proposals…
This is… fascinating. Great work, Alon.
I’d like to do some minor nitpicks/additions regarding the Spanish network, which is the one I’m familiar with.
-The through-tracks in Madrid are unrelated to the third Cercanías axis; see my earlier response.
-The Valencia-Castellón section is not really HSR, just a classic line equipped with a third rail so that AVEs can run on it. There is however a new 64-km HSR section between Tarragona and Vandellós. In the same corridor, there’s an almost finished HSL from Alicante to Murcia, and under construction from Murcia to Almería.
-The under-construction line from León to the north actually ends at Pola de Lena, some 40 km south of Gijón.
-The Zaragoza-Pamplona line is correct on the map, but Pamplona-Irun is not: the current plans are that the line from Pamplona will continue either to Bilbao (joining the Vitoria-Irun line at Ezkio station) or westward to Vitoria. However, if the French take their LGV to the Spanish border, Spain will build a bypass of Donostia, cutting the Vitoria-Irun trip time by around 15′.
Again, these are just minor nitpicks. The high-res map is amazing and I could spend hours looking at all the details! Madrid-Paris in under 6 hours is so promising…
What do you think of a Pyrenees Base Tunnel near Canfranc?
I’m skeptical. There could be demand for freight – Airbus in Toulouse and Opel in Zaragoza, for instance. But the current line’s maximum grade (1.99%) would be freight-friendly if electrified, and I’m not sure there’s demand for long-distance freight: the Barcelona-Perpignan line is way below capacity anyway. As to cross-border passenger traffic, I don’t think it would be worth the expense either. Ridership would obviously be higher than its current value of zero, but the population of the area isn’t huge, and Toulouse isn’t getting a direct link to Paris anyway.
There are also fuzzy plans for a Barcelona-Andorra medium-speed line – perhaps that one could indeed be extended to Toulouse.
There is no current line across the border. Both sides have been inching towards reopening the connection, but thus far Canfranc remains a dead end…
Fair enough; what I meant is the current line + the work needed to reopen the international link, which is at least an order of magnitude cheaper than a new base tunnel.
It is, but nearly 2% grade is not nothing and on this line there is an odd “hole” on the straight line from Paris to Toulouse and onwards via Canfranc to Zaragoza and thence Madrid…
The shortest reasonable route from Paris to Madrid is via Tours, Bordeaux, San Sebastian, Pamplona, and (almost) Zaragoza. It should be possible in 5:00 hours. A tunnel from Canfranc to Pau might cut 20 minutes, but almost certainly wouldn’t warrant the cost. Toulouse isn’t really on the way from Paris to anything, although trips from Bordeaux and the Basque Country towards Montpelier and Marseilles benefit from a link.
Paris Toulouse by itself is a pretty busy market…
And a Pyrenees Base Tunnel has at the very least entered discussion in Spain
Right, but a Pyrenees Base Tunnel almost certainly wouldn’t tie in at Toulouse. It would be much more likely to cross between Pau and Canfranc.
If you’re building the Base Tunnel already it’d be dumb not to extend the line to Toulouse
Is HS2 worth it considering the anglosphere’s incompetence? That money could be used to build light rail for every city in the uk and nationalise and upgrade and electrify the railways to 140 mph.
It’s really hard to say, since the costs are opaque (100+ billion pounds is a high-end estimate including stuff nobody counts elsewhere, like rolling stock – 80 is more comparable). But then the alternatives are also really expensive, like the WCML CBTC debacle.
But that wouldn’t alleviate the current capacity crisis…
Couple of questions: first, how do you propose to do Ingolstadt Munich in roughly fifteen minutes? The Nuremberg Ingolstadt line is already at 300 km/h v(max) and takes roughly half an hour. The other thing: which time is the accurate one for Nuremberg Prague (props for including it in this map) 110 or 75 minutes? Also why this route and not Munich-Prague? And finally how do you intend to reduce Nuremberg Erfurt to 45 minutes and what about the Coburg trains?
Cool map! Why don’t you include the existing high speed line between Warsaw, Kraków, and Katowice though?
If you have a tunnel connecting England to Dublin, then wouldn’t a high-speed line from Dublin to Belfast be worth it? There’s 700k in the Belfast metro area, and under this plan it would not only have a connection to Dublin, but also London within HSR range. Upgrading the existing line to higher speeds could also be a possibility.
I’d think a line to Cork would also be a good idea…
And I don’t know how high travel demand is from Scotland to the Emerald Isle but a tunnel in the north would obviously be shorter and may be worth it even in the presence of a tunnel further south…
Because of the gauge difference, any continuation of the HSR line from Dublin would need new tracks or dual-gauge tracks on the urban approaches to Belfast or Cork. Ireland isn’t geographically enormous (the state of Indiana is bigger), and you could probably get a lot of speed gains just by electrifying existing lines and upgrading the rolling stock, with an occasional curve realignment here or there. In particular, the line up to Belfast apparently needs a lot of work and is currently slower than driving.
Also, the Irish government has other rail investments it should be focusing on – in particular, a crossrail-type tunnel from Dublin Connolly (which serves lines to the north, south, and northwest) to Dublin Heuston (which serves lines to the west and southwest).
There is a scheme, currently shelved, to build a tunnel for the city’s DART commuter rail system from Heuston across the city, joining up with the line to Belfast just north of Connolly station. This tunnel would include a station at the current Dublin Pearse station on the line south of Connolly but would not serve Connolly directly. Pearse is the second-busiest station in Ireland after Connolly and is well-situated for commuters, but it only has two platforms and no room for intercity service. If the plan were updated to connect the tunnel to Connolly as well as Pearse, you could extend the intercity trains that terminate at Heuston to Connolly. Then, passengers arriving on the HSR from Britain would have an easy transfer at Connolly to every line of the DART network and intercity trains to every major city on the island.
Wouldn’t Ireland benefit from a big “regauging day” like they did in the Southeastern US after the Civil War?
This would mean Ireland spending billions of euros on track and billions of euros on an entirely new fleet of rolling stock just to run mostly-empty through trains from a single tunnel to Great Britain. Ireland can’t buy continental rolling stock anyway because its loading gauge is too small. So I don’t really see any tangible advantages from this.
Why would moving tracks from 1600 mm to 1435 mm cost billions of €? And why would replacing a few bogies (heck in Spain they have variable gauge bogies that can change gauge while moving) cost billions? Anglosphere incompetence? Besides, Ireland should spend some money on increasing its loading gauge anyway and if a tunnel through the Irish Sea is ever built, it would be madness to build it to the narrowest possible insular loading gauge…
Fjod, can you explain what you mean when you say “Ireland can’t buy continental rolling stock anyway because its loading gauge is too small”? Irish gauge is broader than standard, so I don’t see how its loading gauge can be narrower than Europe’s. If anything, Ireland can’t buy European trains because their carbodies are too narrow.
Herbert, there is little sense in switching from broad gauge to standard in Ireland. It would reduce capacity.
Even in the maximum-build scenario, Ireland will only have a connection from Dublin to Wales and from Belfast to Scotland. The country is too small to warrant much of a domestic HSR network. Even now, it only takes 2 hrs and 50 mins to get from Dublin to Cork, and that could be reduced with line upgrades. In the unlikely event they do end up building HSR to Cork or Belfast, it is fine to keep that standard, Shinkansen-style, with dedicated tracks on the station approaches. I suppose they could also build the HSR to Irish gauge and use existing approach tracks, but that’s unlikely because it would raise costs and prevent through-running to Great Britain.
Maybe it wouldn’t cost billions to perform the basically useless task of switching the gauge of the entirety of Ireland’s ~3500km of track and 700 or so carriages (plus freight) in one day (I don’t even know how you’d get the workforce for this…) while leaving the rest of the rail infrastructure intact, but it would definitely cost a large number of millions. But you’ve kind of highlighted the unnecessariness of your plan by mentioning variable-gauge bogies; surely it would be a much easier solution to just install these on the (very few) freight trains that use both the British and Irish networks if such a tunnel were to be built. As for passenger rail, the lack of demand means it doesn’t make operational sense to run many (if any) trains from Great Britain to non-Dublin Ireland anyway (Belfast might be a different matter). So what actually would be the benefits of regauging the entirety of Ireland?
Mike – this is quite confusing but loading gauge is a different thing to track gauge (commonly abbreviated as just ‘gauge’). Loading gauge is about the height and width of the trains, not the distance between the rails (this is track gauge).
Britain and Ireland have much smaller loading gauge, both in height and width, than continental Europe. This means that structures in Ireland (bridges, platforms, viaducts, some earthworks) are narrower and shorter than in continental Europe. I don’t know how uniformly small Ireland’s loading gauge is, but the UK has some sections of track with loading gauges that are as high as continental loading gauges (but not as wide obviously, because platforms get in the way). But for the sake of cross-network compatibility, the UK and Ireland buy passenger trains that are shorter than their European counterparts. This probably makes their trains a little more expensive, because off-the-shelf designs have to be modified for British and Irish standards.
Why wouldn’t the be demand for a London-Cork service? Or for a Glasgow-Dublin service?
If the Southeastern U.S. could do it in the nineteenth century (and obviously back then it was deemed worth the expense) why couldn’t Ireland do it today?
Spain built its HSR to international standard gauge even though naysayers might say there’s no demand from Madrid to France and people can just switch trains in Barcelona… Or worse yet, Perpignan
Most passengers from London would go to Dublin, and most passengers from Cork also go to Dublin. So it makes some sense operationally to run specialised HSR trains (with higher crashworthiness, luggage storage, catering etc) to Dublin, then run specialised intercity trains to Cork. The minority of passengers who need to transfer – not just to Cork but to suburban Dublin, Galway, Limerick etc can do so. If you really feel the need to run a few through trains a day (if there’s even demand for that), they can just be variable gauge (or run on standard-gauge HSR if Ireland wanted to build it). Scotland-Belfast-Dublin likewise – although a tunnel between Scotland and NI would go through a 300m-deep underwater trench filled with millions of tonnes of munitions and nuclear waste, and thus is exceptionally risky. Ireland should probably concentrate on easier wins like electrification and short bypasses first.
There is an obvious answer to your second question which is that the southeastern US had millions of former slaves, poor and seasonal workers who could and would do this menial job. 21st century Ireland does not. The US south also had more than one interconnection with the rail network of the rest of the US, which meant that there was a major inconvenience to through-running – and no variable gauge technology to solve this problem. There were also political reasons for the US to do this: unifying a previous breakaway region. This leads onto another point, which is that is a politically ridiculous suggestion for Ireland to spend all that money to change its standard to align with the UK’s.
Irish gauge was set by Parliament in London. 1435 mm isn’t the “English” standard, it’s the world’s
C’est un peu un rêve mais un tunnel en dessous la mer adriatique entre l’Italie et l’Albanie n’aurait il pas une meilleur chance que le tunnel à Helsinki? C’est a peu près la même distance un dessous l’eau mais 1. C’est très près de la fin d’une ligne éxistante en Italie, 2. Ca traverse un nombres baissé de frontières comparé aux autres lignes (2 pays entre la Turquie et l’Italie contre 4 dans votre plan) 3. C’est plutôt au milieu du réseau qui veux dire il y a peut-être plus de passagers qui passent à travers (enfin, entre Istanbul et les plus grandes villes d’Italie il y aurait surement pleins de gens).
Mais, c’est déjà un tunnel hyper long et en plus il y a pas mal de collines dans les pays d’à côté. Que pensez vous?
*pas mal de collines dans les pays d’à côté
Despite Albania being an Italian puppet state / protectorate / colony during WW2 and despite Albania experiencing something of a tourist boom, I do not think there is a chance of a tunnel under the Adriatic see (to my knowledge pretty deep around there) this side of 2050. Are there even flights from Tirana to Italy’s secondary cities?
Bien sûr, mais je pense que la valeur augmente quand le tunnel n’est pas uniquement pour l’Albanie. Pense à une ligne de Istanbual qui utilise le tunnel, et peut-être une ligne qui passe plutôt vers Sofia-Bucarest.
Tirana-Istanbul is already at the upper bounds of the distances at which daytime HSR can compete with aviation… Maybe HSR sleeper trains could work, but outside China (which has its own circumstances) they’re not a reality yet…
According to flightconnections.com (one of my favorite sites) there are flights from Tirana to 15 Italian cities. (Not right now of course…)
According to Wikipedia, several of those flight connections were only planned to start in April… But yes, the two busiest destinations out of Tirana are Italian, only the third place is Turkey, so there are indeed links to both former colonial overlords…
Nice. As a local I would love to see Zagreb/Ljubljana to Munich, or Vienna, or Budapest. Or all of them. The rail passenger traffic is definitely not there at the moment, given the low speeds. But the bus/car traffic is there. All the ex-Yugoslavian countries are very connected to west Europe, with huge numbers of immigrants in DE, AT, and further north. I wish that we grant contracts to build high speed networks to Chinese companies. Those companies are eager to get a foothold in Europe and are competing with low prices. Who knows maybe even below cost? I say let them, we end up with cheap railways. Croatia is upgrading its rail network at the moment, but it is not very ambitious, with 160km/h being the top speed everywhere. Serbia is having the line to Budapest built to 200 km/h. Slovenes don’t seem to care much about speed and are focused on increasing capacity to Koper harbour and not much else.
I’m not convinced it’s worth it. Zagreb or Budapest to Istanbul is 5 hours with fast HSR, and the biggest cities in between are Belgrade (1.7m), Bucharest (1.8 m), and Sofia (1.7 m). This seems very, very marginal. These lines would be dependent on traffic from these (small, poor) intermediate cities to Istanbul, Vienna, and Northern Italy. It would attract very little end-to-end traffic, and I think the investment is best put elsewhere. I just don’t see a scenario where it makes much sense to attach Western European HSR to other networks (Turkey, Russia, Morocco, etc.)
Eventually someone will find out a way to make high speed freight work.
Granted, I’d see its role more in north-south movement of fruits that taste a lot better when harvested ripe (currently some of them are carried via air freight) but I wouldn’t totally discount usage cases for “Madrid Istanbul in under 80 hours”…
There isn’t enough high value freight that has to move fast to do it.
The usage case of a bridge isn’t determined by the number of people swimming through the river
The problem with most high speed rail for freight is that the lines aren’t long enough yet, the question is what is air freighted from Spain or Germany or Italy to London? Because that’s the only line that’s long enough to be worthwhile not just using a lorry.
I wouldn’t bet that HSR freight has the same exact market as air freight. When air freight was introduced, it created its own market. The only longer running experiment with HSR freight I know of, TGV La Poste, carried mail….
And they don’t carry high value, time critical mail by high speed rail anymore.
The one generation of TGV La Poste were retired at the end of a service life of a bit over 30 years. Not a long life for a train, but not terrible. All ICE generations had their first refurbishment earlier than that…
Because the platoons of accountants on both sides of the transactions who have access to decades of detailed information used these things called computer to analyze it and decided that it doesn’t make sense. That doesn’t stop crayonista from scrawling lines. Who are using their superpowers to see how much it could be possible.
Over here the platoons of accountants mostly think deficits outside pandemics, depression, and wars are immoral. And it’s not a depression if it’s only affecting Greece and other countries that Germans look down on.
Numbers are number and they don’t have a political bias. What numbers to look at does and the accountants don’t decide which numbers are going to be looked at. Or which way to look at them. And there are politically motivated accountants who understand what kind of advice you want to result but that isn’t bookkeeping, that is political calculus.
Numbers never exist in a vacuum. And if they do, they don’t mean anything…
Accountants make whole careers out of comparing one number to another. …. that’s the way double entry bookkeeping works. The debits have to equal the credits. The individual numbers entering into the books don’t have a bias. And there are agreed upon standards for what goes where. It’s not the accountants who decide to call it something else and book it in the wrong place. And then cherry pick numbers to make Dear Leader look good.
Who counts the value of a ton of CO2 and how?
It would attract end to end traffic, just that traffic would be overnight. The fact that the destinations on route are weaker aids the sleeper train case. So there would be a train an hour 7am to 9pm, and then 8 trains an hour in a 60-90 minute period overnight. That’d let you do the nightly maintenance in two tranches.
We do not know yet whether high speed sleeper trains would work. We only have a few of them in China, but Chinese aviation is put under constraints (airspace closure, no flying in inclement weather and so on) that European aviation is not under… And Chinese trains have different price points. I’m not saying they can’t work, but I am saying, the bet is a bit high to stake on something that might work but has not been tried…
More importantly, Chinese high-speed sleepers connect enormous cities. Beijing and Guangzhou are each similar in size to London and Paris combined. Meanwhile, between them there are enough huge cities, like Wuhan, that there’s justification for building high-speed rail the entire way. Over here it’s different – I think there’s justification for building at least the western parts of the network I put up, but London-Rome or whatever probably isn’t big enough for sleepers like Beijing-Guangzhou.
A big problem with the business case for sleeper trains in general is: What do they do during the day?
And while historically sleepers had single cars coupled and uncoupled going hither and thither and yon (thus making a business case viable for long and thin routes) but there are two major problems with this: First, coupling and uncoupling and marshalling take a lot of labor and thus cost a lot of money. We can probably not automate this for some time yet. The other big problem is speed. The fastest locomotive, the ÖBB Taurus / Siemens EuroSprinter goes 230 km/h in revenue service and while it has beat the 300 km/h in a trial run, that was without pulling any load.
Are those insurmountable problems? Probably not. But those are still problems. And I think trying to start HSR sleepers on existing routes (e.g. London-Spain) should be done before building HSLs whose only business case would be sleepers or aviation costing orders of magnitude more than it does today…
And hanging out in a hotel bar is more fun than being cooped up in a claustrophobic sleeping compartment. . Or even using the conventional toilet in your spacious hotel bathroom compared to the one on a train. People who can afford trips that long are going to fly
Do overnight trains have to use sleeper cars and just use regular cars?
There are decent amount of red-eye flights and overnight coach buses.
Seats on trains are generally more comfortable than seats on airplanes or buses.
Sure it won’t be comfortable as hotel bed but as long as prices are affordable there could be some demand for overnight HSR travel.
they aren’t running the buses and planes because there is a market for travel in the dead of night. They are moving equipment to meet the next day’s demand and if they are doing that anyway they might as well get a bit of revenue out of it.
I would be perfectly happy with boarding a HSR train at 11pm, sleeping 8 hours in a small compartment, and getting off at 8am the next morning well rested in some distant destination. That would be much preferable to spending half a day of usable daytime on flying the same distance. It would also save a night’s hotel cost.
Easy to comtemplate arithmetic to think about is that the median airline has that plane in the air earning revenue 12 hours a day and when you use a seat for 3 hours you are paying 6 hours of rent on it. You can’t do much of anything with a sleeper during the day when people are awake and to use it for a ten hour overnight trip you have to pay a full day’s rent on it. It will cost too much even with some sort of low or no carbon fuel for airplanes. Dead dinosaur juice is gonna be reallly reallly cheap forever because demand was weaking even before the current crisis revealed that free market zealots who don’t want to spend no munney value money more than they do live people. And more than they do dead ones too.
A sleeper train could become a compartment seated train in the daytime. Sure it’s not as efficient as a normal modern train but it’s better than nothing.
And taking a short flight, even with today’s flying experience and spending some time in a normal sized hotel room will be cheaper and more pleasant that a night on a sleeper train. The fantasy is amusing but normal people aren’t going to want to do it. and it will cost too much.
Austrian Railways are quite happy with the usage their sleeper trains are getting. Notably the most expensive categories sell out first…
And people take cruises to nowhere. I’m not under the illusion that there will ever be regular Transatlantic service with multiple ports and multiple lines again. It’s not going to be a mass market. It’s fun a fantasy, there will be airplanes that normal people use.
Flying effectively takes most of the day. A sleeper train does not.
Let’s say I’m flying to Edinburgh from London and live an hours drive from Heathrow, leave home at 8am, arrive at airport parking at 9am, get to terminal at 9:30am, flight departs at 11:30am, lands at 12:30pm, get tram into Edinburgh at 1:15pm, arrive in Edinburgh city centre at 2pm.
Sleeper experience – leave home at 9:30pm, get to Euston for 11:30pm (so allowing double time for public transport), then arrive at Edinburgh city centre at 7:30am, have a shower and am ready for the day at 8am.
Why would you fly to Edinburgh when the high speed train get you there faster than flying? You haven’t thought about this too well have you?
Some flights between London and Edinburgh are cheaper than the train…
@Herbert Even Chinese sleeper trains cannot operate everyday, they can only operates over the weekend, a bit similar to what London tried with their tube network, presumably for more maintenance work mid-week
And even for Guangzhou-Beijing which is a long journey and see a lot of night time trains, the total number of overnight trains being offered even including trains from beyond are only in the number of single digit. Compares to a total of three digit number of trains on the line in the daytime.
As for usage of them in the daytime, in China some sleeper trains are being offered as bed-seat to passengers on some routes on some days but it’s obviously a downgrade in comfort level that people tend to avoid unless other trains are full or they don’t know better.
@Alon Levy I don’t think most sleeper trains in China stop or serve a lot of passengers along middle cities in the middle of night.
You can make them work if you make sure the tracks aren’t open near neighborhood to avoid excess noise, and if you move maintenance window of each track section into time that are not used by high speed trains in the night time I guess. But the problem is, historically even when air travel was less popular and night train was the norm, there was still much fewer users of night train than daytime transportation options, that I am not sure about the ability for a route constructed mainly with usage by night train in mind to success
I agree with you that in the future the EU should be looking at Eastern HSR lines – especially as they could guide an integrated standard network.
However, I’m not so optimistic about Chinese companies. The reality is, construction costs seem to be local to geography, not companies. Alon I think did a post about this recently, comparing construction costs in Vietnam – where Japan and China both charge many times what they pay domestically to build metros.
It’s not “many times,” it’s maybe twice as high. Very few places are off the global median (which is $250 million/km) by a factor of more than 2, it’s just that New York is one of these places and I write about it a lot.
If you get a current daytime train from London to Edinburgh that normally takes about the same amount of time as flying. In the future a daytime London to Edinburgh high speed daytime train probably would be more sensible as it’d only take a couple of hours, but still. If you take those sorts of timings for London to Edinburgh and apply them to London to Istanbul, then you add another 3 hours to the flight time meaning the journey takes all daytime . On that basis a high speed sleeper taking 15.5 hours leaving London at midnight and arriving at 3pm (5pm local) would give you more useful time in Istanbul than the flight that would land at 5pm (7pm local).
Now maybe that’s too marginal to be of benefit, but if you did Paris to Istanbul where the flight would take nearly as long but the train would be a couple of hours quicker then it’s a different story.
Plus mostly the flights aren’t as convenient as my example, if the flight departs at 9am (or earlier) or lands after 10pm then I’m going to sleep worse in my hotel than on a sleeper train as I arrive super late or have to get up super early.
Problem is: a train has more seats to fill than an A320. And even if you have an “all luxury sleeper” option, that’s a minimum of hundreds of high priced tickets every time the train runs. Business people by and large want to be able to go every day… Yes, I know there are business people on once weekly Ryanair flights, but they don’t party premium fares…
True, but if the train left London at 8:30pm UTC, then Paris at 11:30pm UTC+1 it would arrive into Istanbul at ~3pm UTC+3, that means people who wanted to go to Athens could also change at Nis at presumably ~11am UTC+2. That means that train could take people from Paris, Brussels, London, Birmingham and Manchester who wanted to go to both Istanbul and Athens as well as potentially places in the former Yugoslavia (though you might want to avoid stopping between Paris and Belgrade or Novi Sad for flexibility).
There’s two flights a day with BA from London to Istanbul, so there is demand I’m sure.
I”m going to a hazard guess that the demand for sleeper service to Scotland is roughly the same as what is currently offered. And Istanbul is not in Scotland and if you wanted to talk about Turkey why did you bring up Edinburgh?
Because the London Edinburgh Sleeper runs now. The London Istanbul sleeper is an example that could run with a Europe wide high speed rail network.
The main flaw with the current U.K. sleeper trains is the cost (because they only have one or two people in a room) as it is a lot more expensive than flying, the daytime train or driving. If they had 4 person sleeper cabins I’m sure it’s market share would rise.
And sure a 15 hour sleeper isn’t dramatically quicker than flying, but an 8-10 hour one is. And a 15 hour one is better from a climate perspective.
I think a market can exist, but the experiment costs in the range of billions to run… Maybe it’s also a case of “running one makes no sense, running a dozen does”…
If the point is to get an undisturbed night of sleep, doing it 8 hours doesn’t do that. and you still have to pay 24 hours rent.
Sleeper trains run now. that is what the demand is. People take sea cruises. People take land cruises across the U.S., Canada, Australia and Russia. That doesn’t mean it’s a viable transportation option. Not when we can do something, whatever that is, to make net-carbon-neutral jet fuel.
Austria’s Sleeper Trains have a growing ridership which includes business travelers as well as leisure and even a bit of VFR…
As I said, on the Austrian sleepers (which have lines entirely or mostly outside Austria so it’s questionable they’re a “political” enterprise) the most expensive categories sell out first. For some reason few people wish to sit in a seat through the night even for 29€ Sparschiene prices…
Damn! I arranged for a big (1m odd) print of this for my wall. Unfortunately (or fortunately?), we’re now in shutdown in New Zealand, but I look forward to it arriving later.
Ooh, wow! Send a pic when it arrives?
Overnight HSR aside, how much short-haul flying would this plan, if implemented, actually reduce? There aren’t a lot of city-pairs that look like they would be converted from a high air/rail split to the opposite: London-Dublin, Stockholm-Oslo/Copenhagen, maybe Naples to the Sicilian cities. Then obviously there are a lot of places in Eastern Europe, but I don’t know how much travel volume they have in the first place (given poorer economies and a greater degree of national fragmentation). And there are some borderline cases that would become overwhelmingly rail-dominant (e.g. London-Edinburgh, Berlin-Frankfurt, etc.). But what leaps out at me is that there is relatively little low-hanging fruit to pick off, especially in comparison to North America.
Most of the busiest air routes in Europe from ten or twenty years ago already have HSR. To get much above 2 mio. Pax p.a. You need to combine several airports (such as Dublin to all London airports or Frankfurt to both CGN and DUS)… The crux is moving VFR and leisure away from air travel, but both those markets seem more price than time sensitive. And leisure is to some degree free in the choice of destination in a way that business or VFR isn’t…
– Given the UK to Ireland tunnel distance, would Stockholm to Helsinki be a more worthwhile highspeed route? It should be able to connect Helsinki to all other Nordic countries within reasonable high speed train travel time
– Is Baltic rail not viable in high speed train format?
* I mean, the blog have also mentioned Baltic capitals not being large enough, but even with Saint Petersburg and Helsinki throw into the mix, and connection to other Nordic countries, is it still not sufficient?
Returned to this article via link (partly because I’m planning a trip to Eastern Europe post-corona). Also have last week’s air piracy on mind.
If the regime falls in Belarus and Russia, isn’t Warsaw-Minsk-Moscow a pretty viable route? Or is the distance too far?
It might be viable, but population density in Russia is really low, the cities are just too far apart.