moondog

Last time I went to NTE (from China), I connected in CDG in both directions...on the outbound to an airplane, and on the inbound using a train with an AF flight number. IIRC, the plane was faster, but not much in the context of a 15 hour trip, and the CDG rail connection was quite painless.

roberino

Agree on AMS, FRA and ZRH. Those are examples of how rail and air infrastructure should be integrated. It seems like the two bodies actually talk to one another and make effort to sync up.

roberino

Once you factor in waiting for your luggage after the final leg versus taking your own luggage off a train I'd imagine that the difference becomes less still. Last week I went from Sheffield to Rennes. The AF routing was faster from MAN-RNS (5:30hr), but factor in time to get to MAN (1:15hr), checking in 2 hours ahead (because, COVID), time for the plane to get from the runway to the terminal (never seems to get factored in), time to retrieve luggage, then time to get from an airport to somewhere civilised and suddenly the journey by train (Sheffield-St Pancras-Gare du Nord-Gare du Montparnasse-Rennes) makes a lot more sense.

NickB

It seems to me that it is somewhat more challenging in the French context than in a number of others. Many of the examples you give are of smaller countries with a higher population density and intercity services that can sustain a high frequency regardless of the airport.

This is more problematic in France, with a much smaller population density but also a high degree of centralisation, which makes intercity services other than to/from Paris rather thin and affects the economics of high frequency services. It does not help that CDG is to the North of Paris, with Lille as the only major domestic destination in that direction. It seems to me that to have high frequency services, you would need to integrate the CDG station on lines to/from central Paris and make CDG the final point beyond Paris on TGV originating South of Paris (so a Bordeaux-Montparnasse-CDG or Lyon-Gare de Lyon-CDG).. Needless to say, the travaux pharaoniques and astronomical costs that this would imply means that it is a complete non-starter.

tbaiyun

The poor Grand-Est was again left behind. (just kidding, no offense). Major domestic destinations can also include Reims, Metz, Nancy, Strasbourg, with possible international extension to Luxembourg, Mannheim, Stuttgart, Basel. In fact, CDG is a valid transfer point for itinerary Nancy-Lille, if someone prefer not taking that fancy stairways between Gare de l'Est and Gare du Nord.

I thought the main reason why they couldn't put more trains from CDG is that they only built 2 platforms (4 rails) under T2?

Goldorak

For sure, there will be a capacity problem for the TGV station in CDG. We are not there yet, but it can happen fast.

NickB

I probably was not clear enough as I am not sure whether you understood what I was saying. What I meant is that, if you look at the structure of the network, if you look at the TGV line that goes through CDG, northwards there is only one major domestic destination: Lille. All the other destinations (including the Grand-Est destinations) are all southwards of CDG. So, CDG does not work well as a point on inter-regional networks as the only destination at one end is Lille and there is only so much traffic that needs to go to Lille from elsewhere in France. In an ideal world, you'd prefer an inter-regional hub to be in a location that provides for a range at destinations in both directions to allow for greater frequencies on the common segments (so, from that perspective, somewhere like Massy-TGV would work better because in one direction you have destinations in Western and South-Western France and in the other destinations in Northern, Eastern and Southeastern France).

You are only going to have so much traffic that wants to travel Strasbourg-(CDG)-Lille. If there was another significant destination, then you could alternate Strasbourg-(CDG)-Lille with Strasbourg-(CDG)-XXX to allow for greater frequencies to CDG without having to send gazillions of TGVs to Lille where the traffic does not warrant it.

I took the example of Bordeaux-CDG or Lyon-CDG but the problem is exactly the same with Strasbourg-CDG. The number of passengers on these services are not going to be huge. For higher frequencies to be viable, you must have the other points on the route to attract sufficient passengers and you are not going to have high enough numbers of passengers on a route like Strasbourg-CDG-Lille to warrant a high frequency. The only way to justify high frequencies on these kind of routes would be to have a central Paris stop. A Strasbourg-CDG-Paris Nord (or Paris-Est, it does not matter) would have a much greater likelihood to be viable at high frequencies .than Strasbourg-CDG-Lille. The problem is that, for this to work, the segment between CDG and Paris-Nord or Paris-Est would have to be fast enough not to make the Strasbourg-Paris Nord or Paris Est via CDG much longer. With the current rail infrastructure and the constraints in the Parisian area, this is plainly impossible. So, I don't think that we are likely to see a high frequency service on these routes (even if the problem of having enough capacity at CDG itself could be solved)..

San Gottardo

Except that CDGthat "detour" doesn't matter too much because depending on how you plan the network because it would be really small. From the TGV Est the bifurcation to CDG is at Claye-Souilly, which is something like 4-5 minutes driving time for the TGV to CDG. And it's still better than having to change stations in Paris, and would give some good connection to CDG.

A possibiliy would be to have
Line 1: hourly Bruxelles-Lille-CDG, that then alternate between Bordeaux and Lyon (and onwards to Marseille/Montpellier). Or you split the train at CDG if you want to serve Lyon and Bordeaux hourly.
Line 2: every two hours Strasbourg-CDG, that then alternatate between Nantes and Rennes. Or you split the train at CDG if you want to serve Nantes and Rennes hourly. This would be a "detour", but as explained above it would be short and definititely much shorter and more comfortable (just walk across the platform) than changing stations in central Paris.

It would improve cross-country TGV connections and provide better-than-today connections to CDG.

Of course that it is all a bit visionary from where we are today where SNCF isn't even able to spread CDG connections evenly throughout the day, has large gaps in the schedule, let alone sync them to AF departure/arrival waves (just in case someone thinks that would be preferrable - I don't).

NickB

As I said, the limited number of beyond destinations is just another factor that does not help. Yes, you can have trains back-tracking but it adds complexity to the operations and also creates a potential pinch point in the network in the few kilometres south of CDG. The 4-5 mins means adding something in the region of 20-25 mins to total travel time. Not the end of the world but not negligible either, especially if you add the additional time that the TGV Paris bypass line also involves.

And that is the real issue: do you really think that there is demand for a Strasbourg-CDG-Rennes every 2 hours and ditto for Strasbourg-CDG-Nantes? That would mean somewhere in the region of 3K to 4K passengers daily around 25K passengers weekly, give or take a few thousands, on each of those lines. Do you think that this demand exists or would most of these services be Marie-C�leste services with trois pel�s et un tondu?

San Gottardo

• Zero time lost for the trains going from the LGV Nord (Bruxelles, Lille) via CDG to the LGV Sud-Est (Lyon) or LGV Atlantique (Bordeaux) line
• 20 minutes lost for the trains going from the LGV Est (Strasbourg) via CDG to the LGV Atlantique (Nantes, Rennes) line

Those 20 minutes are compared to if the train went Strasbourg-Rennes without detouring via CDG. But more time gained and lots of comfort gained compared to arriving at Gare de l'Est and having to squeeze into the crowded and smelly Metro in Paris with your luggage instead of just walking across the platform.

EDIT: and, by the way, already today some (all?) of those cross-country TGVs have some ridiculously long stops on the Paris bypass line. 16 minutes in Marne-la-Vallee, or 14 minutes at CDG.... so seriously, that 20 minute "loss" you are mentioning may in fact not really be a loss of time compared to today where similar times are lost by just standing around at some station for... well not much really.

What do you mean by that?

You argue like the SNCF - and my point exactly is to try to think differently. Not "how many people would go every two hours from Strasbourg to CDG and then to either Rennes or Nantes?". But: "CDG is a connection point with synced schedules, so we connect people from Bruxelles, Lille, Strasbourg, Reims, Lyon and beyond, Tours, Bordeaux, Le Mans, Rennes, Angers and Nantes to each other with either a nonstop train or an easy platform change to a connecting train in less than 10 minutes". I believe there would be enough traffic to justify Lyon/beyond to the CDG hub once an hour, Bruxelles/Lille once an hour, and all other destinations every two hours.

And one other data point: Checking for a random date, there are today seven TGV departures on the LGV Est from Strasbourg that do not go to Paris, but to Nantes, Rennes, Bruxelles. Except that they are timed randomly: two within ten minutes of each other/four within three hours of each other, then nothing for five hours, then again two within 15minutes of each other, then again nothing for three hours.... why not simply make them depart at 7am, 9am, 11am, 1pm, 3pm, 5pm, 7pm to the hub, where they then meet trains from the other spokes and provide connections throughout the day? So I am actually not increasing capacity, I am just planning it differently. In a hub and spoke way.

But, helas, the SNCF doesn't think in terms of hubs and spokes like the NL, SNCB, DB, SBB and several others do. I understand the challenge that you cannot run a full and dense schedule between Aubusson and Busseau-sur-Creuse. France is too sparsely populated to irrigate it with train lines like Switzerland. But on the trunk routes, I think it should work. Especially as fast train connections also create new demand.

NickB

What I meant by that is that the Paris Bypass line is very slow (approx 45 mins between Massy-TGV and CDG) whereas the radial lines into Central Paris terminal stations are much faster, which means that, coupled with the fact that those inter-regional TGVs are on thinner routes and make more stops, the time saving overall can be negligible or even negative (some connections with change of stations in central Paris can be quicker than direct trains via the bypass line). Agreed that this is not the only consideration at stake and the ease of not having to change stations in Paris is a significant bonus in terms of comfort; also, in the longer term, the GV Southern Bypass should ameliorate that (whenever this is eventually built) but , for now, the time factor is not a major plus for those inter-regional TGVs.


OK. If I understand you well (and apologies if I don't), it seems to me that you have massively moved the goalpost here. In your earlier post, you were talking of a Strasbourg-CDG-Nantes every two hours, alternating with a Strasbourg-CDG-Rennes every two hours. Now, you are talking of a grand total of 7 TGVs on the Grand Est line to all destinations but using a connecting logic rather than direct trains. If we just keep Rennes and Nantes as direct destinations from Strasbourg (with the connections to Lille, Bordeaux, etc... no longer being direct but through connections at CDG or Massy), we are now talking of 4 trains a day to Nantes at four hours interval and 3 trains a day to Rennes at 4 hours interval. That is one heck of a massive difference with your earlier proposal of a train every two hour to each of these destination.

But let us also look at those so called random timings of TGV departures from Strasbourg. Approx half of these are in the 6-8am time slot. Now, this is not random. This is because it is where peak demand is. What you are proposing is to take away trains at time of peak demand to move them to times of low demand. Prima facie, that does not strike me as the smartest of moves.

Another issue with your system, which is theoretically very attractive on paper, is that it is much more crucially dependent on reliable connections. It is premised on the notion that, when I arrive in CDG or Massy-TGV in my Strasbourg-Nantes train, I will have a timetable-coordinated train to Lille or Bordeaux, as the case may be, within the next few minutes (let us say within 15 mins). Now, what happens if my train is 20 mins late? In the current point-to-point oriented system, well I will arrive at my final destination roughly 20 mins late. If the connection-based system (which incidentally is more of a nodal interchanges logic than hub and spoke), I miss my connection and I'll have to take the next one. Now, if you are in a system that can sustain hourly frequencies, that is annoying but not the end of the world: the next train is never more than an hour away so, all in all, you will have lost one hour or less. But let us not forget that we are now talking of the next train to Nantes being four hours later.

So you need a transport system which is capable of eliminating delays or you have to have a system where trains wait on each other but that becomes rather nightmarish in terms of complexity.

Incidentally, the SNCF does operate trains based on a logic of equal intervals between trains but it does so on the TER network rather than the TGV network. And, if you look at the TER network, you will notice that the distances (overall as well as between stops) on long distance TER trains operating under such systems tend to be similar to the kind of distances you have on intercity trains in NL, BE or CH. Perhaps this is more than just a coincidence...

Now, does that mean that everything is perfect and that there is no room for improvement? Of course not. Regional inter-connections have never been a priority at the SNCF and this reflects the inheritance of a structure based on regions with a head station in Paris. But that, in itself, is also a reflection of how the country as a whole is structured. There is definitely room for improvement in terms of regional interconnections. Whether that improvement means adopting systems that work well in smaller, more decentralised states with a denser population is, however, far from obvious. I would have thought that there is room for a regular, reasonably dense service on the Lyon-CDG-Lille-Bruxelles axis. For the others, I don't think it is that obvious.

maalloc

From what I've gathered, there isn't much that can be done here.
The Paris-Lyon segment is almost saturated and any trafic increase towards Lille, Marseille or Perpignan is kind of limited by the central bottleneck.

stimpy

There are normally regular TGV trains from Marseille to Brussels that bypass Paris. Yes they occupy both the busy Marseille-Lyon line and part of the way the busier Paris-Lyon line, but they do, or did run several times per day. I used to take that train often just on the Marseille-Lyon segment and sometimes the Lyon or Le Creusot to Brussels segments. These trains are often sold out in the summer time and holiday times with tourists, but at other times easy to book.

But I still fly regularly LYS-CDG-ex EU if I choose AF. With baggage and especially with family would never consider taking the train to Paris. As stated earlier for those of us who live in the countryside it just doesn't make sense.