Caledonian Sleeper to mainland Europe?

[popeter45]

what i could see way more as an option (assume you had spare mark 5 carriages and added KVB too the class 92) would be a corridor train that rather than going all the way too Europe, ended up at the international side of St Pancras for direct connections too Eurostar/Virgin/etc
reduces scope only too adding juxtaposed control too somewhere in Scotland

 

[Zomboid]

lack of KVB also makes the French side tricky as well as nowhere in TVM land that can be used for a loco swap (i believe idea was for some AWS at Calais-Ville to enable that to be used)
if being realistic (as much as you can with such a idea), locomotive + carriages would not be the best option for such a service, either a multiple unit (like the class 79 in Norway) or pseudo Multiple unit (like the CR200J in china) would make more sense for tunnel/HS1 operations

It's pretty easy to find recent videos of 92s running on HS1, so whatever the signalling system is, there are 92s with it.

 

[Clansman]

If it's the highland towns keeping it in business, then how come the lowland sleepers are still run? And how come it makes a loss anyway? Fares are relatively high and whenever I've looked, it usually seems to be well booked.

As others have said it loses money because of railway economics. It serves little number of passengers across a large distance of rail, with too many expenses to maintain it by virtue of how it operates geographically and operationally in resources it needs.

It isn’t keeping the towns in business nor are the towns keeping the sleeper in business.

The Scottish Government has identified that tourists spend more money in the local communities served by the sleeper by virtue of travelling by it - to a point that merits the sleeper’s subsidy since the service serves local areas economically more than it serves the service itself.

Think of the sleeper in terms of not if it’s sustainable in isolation, but whether it contributes to the economy. In this case, the Scottish Government believes that in the case of the latter, it does. Ie tourists who are likely to travel by sleeper spend more in their destination than they do on the sleeper - the difference is the multiplier between money spent on the sleeper versus money spent in the destination against money spent if the sleeper wasn’t provided at all.

Long story short, it’s not a subsidy, it’s an investment in the economy.

The fares are high to extract maximum revenue. The more demand there is, the higher the price becomes. If the price was too high, the sleeper wouldn’t reach its sell out point, and the fact you see high prices with very few berths left is testament to the fact that this approach works.

 

[gingerheid]

And how come it makes a loss anyway? Fares are relatively high and whenever I've looked, it usually seems to be well booked.

because it has all the costs of running a train for just a dozen or so passengers per coach?

And it has almost all the overheads (in some respects perhaps even more) of a full scale long distance operator, but only one train a day on each route.

 

[zwk500]

It's pretty easy to find recent videos of 92s running on HS1, so whatever the signalling system is, there are 92s with it.

To run on HS1 itself and on the channel tunn needs only TVM, the French high-speed cab-signalling system.
To run onto the French conventional network would require the KVB train protection system fitted as well, which is also used ar St Pancras and Ashford but 92s apparently dont have it.

HS1 freight would presumably change from AWS/TPWS directly to TVM at Dollands Moor or Ripple Lane.

Although given freight does run through from France, where do freight trains change to a 92 on the French side?

 

[Falcon1200]

And how come it makes a loss anyway?

Because it is incredibly expensive to provide! The stock is specialist dedicated vehicles which make one journey per day, cannot be used for anything else, carries fewer passenger than day vehicles, and multiple locomotives and crews are needed to operate a small number of trains.

a corridor train that rather than going all the way too Europe, ended up at the international side of St Pancras for direct connections too Eurostar/Virgin/etc

That was tried with the direct HST between Edinburgh and Waterloo, when Eurostar used that terminal, and hardly anyone used it! Something else the proposers of the feasibility study should consider.

 

[Magdalia]

Being an old cynic, I started off thinking of this as basically a political project, through trains between Scotland and Paris or Brussels for a future independent Scotland belonging to the European Union.

The Scottish Government has identified that tourists spend more money in the local communities served by the sleeper by virtue of travelling by it - to a point that merits the sleeper’s subsidy since the service serves local areas economically more than it serves the service itself.

Think of the sleeper in terms of not if it’s sustainable in isolation, but whether it contributes to the economy. In this case, the Scottish Government believes that in the case of the latter, it does. Ie tourists who are likely to travel by sleeper spend more in their destination than they do on the sleeper - the difference is the multiplier between money spent on the sleeper versus money spent in the destination against money spent if the sleeper wasn’t provided at all.

Long story short, it’s not a subsidy, it’s an investment in the economy.

But this contribution, combined with a radio item on "coolcations" that I heard yesterday, got me thinking differently. Is there really a market for getting the French and the Belgians to come to Scotland by train to spend money in the local Scottish economy? If Europeans want a mountain holiday, aren't the Alps a much more attractive option than the Highlands of Scotland?

And the elephant in the room is borders, not the Borders Railway, but passports and security. If Eurostar can't make stopping at Ebbsfleet or Ashford viable because of the cost of border security, I can't see how passports and security are going to work for cross border sleeper trains.

 

[vuzzeho]

Surely this wouldn't run to the Highlands, though? Simply because of timing and demand, I imagine a Glasgow/Edinburgh service to Paris would probably be the most feasible. It could also serve as a WCML to Europe service, which might increase demand.

 

[lowe1409]

Surely this wouldn't run to the Highlands, though? Simply because of timing and demand, I imagine a Glasgow/Edinburgh service to Paris would probably be the most feasible. It could also serve as a WCML to Europe service, which might increase demand.

I imagine they would probably need to to make it feasible. Preston you have what I believe are the former postal platforms to the west which are isolated for border controls or Crewe could either isolate p12, or possibly have the space to add in a p13 for the International sleepers. Not sure if there's anywhere suitable nearer to Birmingham too. Biggest issue I see is your have to build a separate international terminal at every station it stops at for 1 train a night.

 

[marshlinker]

To run on HS1 itself and on the channel tunn needs only TVM, the French high-speed cab-signalling system.
To run onto the French conventional network would require the KVB train protection system fitted as well, which is also used ar St Pancras and Ashford but 92s apparently dont have it.

HS1 freight would presumably change from AWS/TPWS directly to TVM at Dollands Moor or Ripple Lane.

Although given freight does run through from France, where do freight trains change to a 92 on the French side?

There's AWS/TPWS and TVM interface at Barking, Ebbsfleet, Ashford International and Dollands Moor/ Sand gate. Also Fawkenham Junction (nowOOU).
Ashford International had KVB overlaid on the AWS TPWS system to allow class 374 to call.
I think at Calais the freight trains are signalled in to the yard by TVM to a complete stop for a loco change. Whether this would be acceptable for passenger trains in France I'm not sure.
At St Pancras one of the freight companies has a derogation to run trains without protection under special arrangements.

 

[gingerheid]

That was tried with the direct HST between Edinburgh and Waterloo, when Eurostar used that terminal, and hardly anyone used it! Something else the proposers of the feasibility study should consider.

In fairness the existence of the services was kept quite secret, and the length of time they took to crawl round to Waterloo was something else. However the idea sets you down a line of thinking that helps you realise what a doomed idea this is.

One might say that a moderately reasonable way of delivering the idea of an overnight service from Scotland to the continent would be to divert the existing sleeper to either King's Cross or St Pancras to assist (as much as reasonably possible) with an easy connection onto an early morning Eurostar service. Lets ignore a lot of practicalities and put the argument at it's absolute highest; lets assume that it was possible to either fit such sleeper trains in one of those stations at the relevant time, or that it was possible to organise carriages taken from existing trains into a separate shorter train to one of those places, or even that the existence of such a service would increase passengers numbers so much that there could be an increase in the number of services each night and that some of them could be shorter and fit in there. Even then... it's still an obvious no!

I think my sleeper train wish list would be not an even more expensive to operate and technically complicated operation to Europe, but a 7 day a week service and a parallel budget option on the existing routes. Given the number of coaches crossing the border each night, and given that they are not always that cheap, I feel this is where any opportunity there might be lies.

 

[marshlinker]

what i could see way more as an option (assume you had spare mark 5 carriages and added KVB too the class 92) would be a corridor train that rather than going all the way too Europe, ended up at the international side of St Pancras for direct connections too Eurostar/Virgin/etc
reduces scope only too adding juxtaposed control too somewhere in Scotland

You need to modify the Mk5s with some sort of retractable steps or modify a platform at St Pancras. The gap and height difference between UIC gauge and British is too large for passenger use. When the 395s used Stratfords international platforms for the Olympics temporary wooden platforms were placed on top over the sides.

 

[Bald Rick]

reposting from the other thread.

Well, the air + rail market for Glasgow / Edinburgh to Paris / Brussels / Amsterdam is roughly 15-20% of the air + rail market of London to Glasgow / Edinburgh.

A decent proportion of the Scotland - Amsterdam air market will be transfer passengers at Schiphol. Paris will have quite a few too. (Albeit so does Heathrow).

It’s reasonable to assume that if the same proportion of the city to city market willing to travel by sleeper between London / the South East and the Central Belt were willing to travel between the Central Belt and Paris / Brussels / Amsterdam by sleeper, there would be enough for about 2 sleeper coaches per night each way total.

== Doublepost prevention - post automatically merged: 8 Aug 2026 ==

There's AWS/TPWS and TVM interface at Barking, Ebbsfleet, Ashford International and Dollands Moor/ Sand gate. Also Fawkenham Junction (nowOOU).
Ashford International had KVB overlaid on the AWS TPWS system to allow class 374 to call.
I think at Calais the freight trains are signalled in to the yard by TVM to a complete stop for a loco change. Whether this would be acceptable for passenger trains in France I'm not sure.
At St Pancras one of the freight companies has a derogation to run trains without protection under special arrangements.

There‘s also different OLE settings for the tunnel.

 

[gingerheid]

Well, the air + rail market for Glasgow / Edinburgh to Paris / Brussels / Amsterdam is roughly 15-20% of the air + rail market of London to Glasgow / Edinburgh.

Adding in the coach market will make the size difference even bigger.

It’s reasonable to assume that if the same proportion of the city to city market willing to travel by sleeper between London / the South East and the Central Belt were willing to travel between the Central Belt and Paris / Brussels / Amsterdam by sleeper, there would be enough for about 2 sleeper coaches per night each way total.

But then it gets worse. If one was to be willing to take support for the service idea to the ends of the earth and try and put ridiculous ideas into practice then they might try and work out how to separate those two coaches from the existing services and get them through the channel. But even that wouldn't be suitable, as the demand would be too seasonal.

All the above said though... if we we having this discussion before Stratford International had been built, and at the time the Mk5 stock was being specified, then at the point political will to achieve mad ideas might actually have been enough

 

[HighlandStorm]

That was tried with the direct HST between Edinburgh and Waterloo, when Eurostar used that terminal, and hardly anyone used it! Something else the proposers of the feasibility study should consider.

I don’t think that’s relevant to direct trains or sleepers. There have been occasions I would have made use of direct Eurostar services from Edinburgh or Glasgow, but if I’m taking a train to a London terminus, I’d want to get off and break the journey at least a bit, not just shuffle over the platform to another train with no option to leave the station!

 

[brad465]

Time for this one again: "Consultancy, if you're not part of the solution, there's a lot of money to be made in prolonging the problem."

 

[Falcon1200]

I don’t think that’s relevant to direct trains or sleepers.

The service was made as convenient as possible with one same-station change, yet still hardly anyone used it, surely an indication of likely demand?

 

[Krokodil]

The service was made as convenient as possible with one same-station change, yet still hardly anyone used it, surely an indication of likely demand?

30 years ago, when Eurostar was far slower and you'd still need to change trains anyway.

 

[Bald Rick]

The service was made as convenient as possible with one same-station change, yet still hardly anyone used it, surely an indication of likely demand?

30 years ago, when Eurostar was far slower and you'd still need to change trains anyway.

30 years ago Low Cost aviation in europe was in its infancy.

 

[Krokodil]

30 years ago Low Cost aviation in europe was in its infancy.

Eurostar was also in its infancy and had a tiny market share. Now it is something like 89% of London-Paris traffic. Low cost airlines can't compete against HSR on routes under 4 hours.

 

[Bald Rick]

Eurostar was also in its infancy and had a tiny market share. Now it is something like 89% of London-Paris traffic. Low cost airlines can't compete against HSR on routes under 4 hours.

It can and does compete, but usually comes out as the minority provider.

In any event, I think we can agree that a Sleeper from Scotland to Europe will be rather more than 4 hours. Almost three times more.

 

[Krokodil]

It can and does compete, but usually comes out as the minority provider.

In the context of a sleeper, it would be a very long train if it was anything more!

In any event, I think we can agree that a Sleeper from Scotland to Europe will be rather more than 4 hours. Almost three times more.

You wouldn't want a four hour sleeper train, of course. Needs to be at least eight hours to be worthwhile. The Lowlander takes seven hours to get to Watford Junction from Glasgow. Three hours more to Paris? Assuming that it were to depart Glasgow around 22:30 (it would supplement the existing train, not replace it) then that means Watford Junction at 05:30 and Paris at 08:30 or so.

In the light of sleepers regaining popularity in Europe, there's certainly a market, but it would rely on EU membership and Schengen to be truly worthwhile. A Brussels portion would be popular for MEPs making their weekly commute.

The thing that sank Nightstar was that SNCB didn't have a locomotive powerful enough to both haul and provide ETS to the full set without double-heading. Less of an issue these days when more powerful locomotives exist on the spot hire market.

 

[Bald Rick]

Three hours more to Paris? Assuming that it were to depart Glasgow around 22:30 (it would supplement the existing train, not replace it) then that means Watford Junction at 05:30 and Paris at 08:30 or so.

From Watford you’re looking at easily 4 more hours to Paris, and only then if you can find a slot on HS1. I doubt SNCF would have it on LGV Nord with loco axle weights (something akin to a Class 92 or Vectron would not be permitted) snd certainly not south of Lille with the LGV morning peak in full flow.

The thing that sank Nightstar was that SNCB didn't have a locomotive powerful enough to both haul and provide ETS to the full set without double-heading.

The thing that sank nightstar was that it was completely uneconomical, partly due to the advent of low cost airlines. Thee is nothing to suggest that a Scotland - Europe sleeepr would be any different.

 

[Sir Felix Pole]

Eight members of SNCF's BB22000 'Broken Nose' class (1500V DC / 25kV AC ) were capable of running on LGVs at 200km/h, so loco hauled operation is not a complete no-no. They operated an express parcels service on the LGV Sud Est for a time.

I'm not sure what their current status is - they were mooted for the Portbou / Latour de Carol sleeper when it is diverted via Bordeaux due the ongoing engineering work on the 'POLT' line.

 

[Krokodil]

The thing that sank nightstar was that it was completely uneconomical, partly due to the advent of low cost airlines.

The final straw (and I've had this from someone then at Eurostar who was involved with testing both Nightstar and the NoL sets before the plug was pulled - they were weeks away from opening bookings) was the SNCB loco issues. The difference between single and double heading was enough to push it over the edge - the retrenchment of sleeper services in Europe came later - airline deregulation had barely been written into law when Nightstar was abandoned.

 

[marshlinker]

You wouldn't want a four hour sleeper train, of course. Needs to be at least eight hours to be worthwhile. The Lowlander takes seven hours to get to Watford Junction from Glasgow. Three hours more to Paris? Assuming that it were to depart Glasgow around 22:30 (it would supplement the existing train, not replace it) then that means Watford Junction at 05:30 and Paris at 08:30 or so.

Watford Junction to Paris North via the North London Line & HS1 is 324 miles.
By the time your sleeper has negotiated the crossovers to the WCML slow lines and then all the 30 mph restrictions through north london add in pathing allowances I reckon you'd struggle to get a journey time under 5 hours. So an arrival of 11.30 Paris time.
A class 374 averages 137 mph on a non stop St Pancras to Paris run.

 

[HSTEd]

I think you'd have more luck with an entirely bespoke 300km/h 400m sleeper EMU. Obviously it would cost a fortune to build such a thing, but all the infrastructure set up required to even attempt this concept would probably make that look positively reasonable!

Try and wrangle the first slot on LGV Nord in the morning.

 

[mike57]

As someone based south of the border, this is how I see it.

The current sleeper is a loss maker supported by Holyrood. I can see good reasons for continuing to support it, but surely the subsidy required for an international service would be far higher, and that would be on top of the existing subsidy to the domestic sleeper.

The London > major Scottish destinations market is huge in comparison to the Paris, Brussles (which are the two most likely southern end points) market. I dont know what proportion of journeys the sleeper contributes, but even if you just count rail journeys London > Scotland it must be small, so a similar percentage of Continent > Scotland journeys would result in a lot of fresh air being carried unless fares were budget airline low, which in turn would mean an even bigger subsidy.

Overnight sleeper to London followed by Eurostar is probably almost as quick if you dont want to fly.

Border formalities for 1 or 2 trains per day would need to be set up in potentially multiple Scottish locations for maybe 1 return service each day, and if you used the current model and did both sets of formatilies at the start of the journey you would need a French contingent at the Scottish end, which quite rightly the French would charge for.

This is before you get into the technical side, rolling stock, signalling, traction.

Why are they even 'throwing money' at it?

 

[popeter45]

I think you'd have more luck with an entirely bespoke 300km/h 400m sleeper EMU. Obviously it would cost a fortune to build such a thing, but all the infrastructure set up required to even attempt this concept would probably make that look positively reasonable!

Try and wrangle the first slot on LGV Nord in the morning.

dont think you even need 300km/h, in theory a 250km/h Stadler SMILE made too UK specs could work for way less
that or a tack on order too the HS2 fleet

 

[Tio Terry]

In the past, Border Force have ruled out any additional inland `port because the British government decided not to join Schengen.

Any new proposal would need to sort out how Border Controls would be handled before anything else.