Are Swiss trains genuinely reliable?

[dm1]

I spent a week in Switzerland in July and only encountered one delay, which was on a Railjet, albeit the cause of the delay was in Switzerland. It was major disruption too, but handled better than I’ve ever seen equivalent disruption handled in the UK.

The Railjet had departed Zürich HB as usual but only made it as far as Zürich Wiedikon before being stopped. Can’t remember exactly what it was now but it meant that the usual route to Sargans was blocked.
Within 10 minutes of stopping at Zürich Wiedikon the train had left again in the reverse direction to take a diversionary route to Buchs, running non-stop through the low level platforms at Zürich HB and then via Winterthur, Amriswil and St Margrethen. Arrival into Buchs was about an hour late, the diversionary route skipped Sargans so for those headed there like I was a trip on IR13 was also required, no difficulty whatsoever with using an advance (or whatever SBB’s equivalent is) for the Railjet on the IR13 to get to Sargans from Buchs.

It struck me how well prepared they were for this disruption and how quickly they were able to implement an unplanned diversion. The chances of the main Zürich to Sargans line becoming blocked whilst an Eastbound Railjet is between Zürich HB and Zürich Wiedikon (as opposed to before it departs Zürich HB) are very low and yet SBB were still fully prepared for that scenario when it happened. A UK equivalent would be a service leaving New Street but only making it as far as Smethwick before the line became blocked at Sandwell, then almost immediately reversing, going back through New Street and going via Aston and Tame Bridge Parkway to reach Wolverhampton instead, it likely wouldn’t happen.

So yes, Swiss railways are reliable. And on the rare occasions things go wrong they appear to be very good at handling it.

I this is a good example to mention a key tool in the reliability toolbox of the signalling and control staff - pre-planned contingency strategies. Since the whole railway network runs in a Takt, the timetable follows a standard hourly pattern, meaning it is feasible to pre-plan for almost any possible kind of disruption and develop strategies for each one, taking pathing constraints, diversion routes, and various other things into account. Obviously not every possible scenario will be covered, but mainlines (such as the line from Zürich to Sargans) being blocked and fatalities (yes, they happen in Switzerland too, with unfortunate frequency) are standard disruption scenarios, and the diversion via Winterthur and St. Margrethen is a typical measure that is taken in the situation described above. Those strategies may also include which lines to cancel and which lines should receive additional scheduled stops (e.g. most kinds of disruption to the line between Zurich and Bern will almost invariably lead to the IC1 and IC8 services making unscheduled stops in Olten to pick up disrupted connecting passengers).

As another example, the chronic unreliability of the inbound EC trains from Munich meant at some point SBB had a full-on strategy with different diversion routes and overtaking locations depending on exactly how many minutes late the EC train was arriving at St. Margrethen, in order to decide whether to make the IR13 scheduled to depart towards Zurich a few minutes after the EC should wait or not, which included pathing calculations about whether the late EC would subsequently get stuck behind an S-Bahn train in the Zurich area and by when exactly the IR13 would be able to catch up to its timetable should it depart late.

This means that instead of trying to make up these control strategies on the fly, they can take one of the existing strategies and apply it to the situation at hand, freeing their mental capacity to get the details right, because the big picture is already there.

Another key difference to the UK is the considerable flexibility the controllers have in dispatching spare or replacement trains to cover for delayed trains. There are agreements and contracts in place between almost all operators in Switzerland allowing last-minute allocations if a train happens to be in the right place at the right time during disruption. For example, there are numerous cases of a late ICE (nominally operated by SBB within Switzerland) from Basel being covered for by an SOB Traverso. If a driver and train are available, then they will be used to stabilise the service, and the contractual sorting out will happen afterwards if needed.

 

[Austriantrain]

British culture is pretty fluid and across our railway network there's quite a range of practices. It's a double edged sword but it does mean there exists good practice at home that the rest of the country can gravitate towards. For example, passenger behaviour and operating practices are generally pretty tight on London Underground and even on NSE areas of National Rail, and I would say these are towards the top end of the European range. It's easy to exaggerate the doom and gloom.

In terms of general funding levels and infrastructure km / vehicle km ratio the UK isn't going to approach Swiss levels any time soon, but Switzerland isn't the only model around. I think the Dutch model is a more suitable one for the UK to follow. Partly because British stations tend not have the space for Swiss number of platforms or junction grade separation in comparable locations. Where they'll have 8 platforms (platform faces) we'd have 4, where they'd have 2 islands we'd just have 2 platform faces. This isn't the end of the world and there are ways to work with what we've got. Where the Swiss would dogmatically have everything at half-hourly frequencies and each service group having dedicated platforms for overlapping dwells around nodes, the Dutch would quite happily run combined 15-minute frequencies so the same number of trains can be shoved through fewer platforms. This does means some branches don't connect quite as well but 'inner' stations get the frequency bonus. The British timetable structure isn't a million miles from the Dutch one, we just need to apply the model more consistently and more rigorously. That's achievable with incremental improvement.

The Dutch and the Swiss models are not in real world competition though. The dutch one evolved in a country that in large parts always was very densely settled and where building up a national „S-Bahn“-style network was rational and feasible. It is true that if most things run every 15 minutes, you don’t need to build major hub stations because connections are always short.

The Swiss Takt was designed in the late 1970s/early 1980s in what was still a very rural country. Having hourly train services all around was revolutionary; a 15-Minute-service would have been an absurdity and impossible to finance. If you have an hourly service, you have to time your connections, and the same is still true for half-hourly services if overall journeys tend to be short and often involve more than one connection. Although Swiss demographics have changed somewhat, it is still quite different from the Netherlands.

Also, Switzerland often having three tiers of services and not long, but somewhat longish nonstop runs mean that anything more than the half-hourly service would need a very large amount of four-tracking, which historically has never existed in Switzerland and mostly still does not. It is anybody‘s guess whether a large amount of four-tracking would be cheaper than extending stations over and over. I would be doubtful.

I agree though that the Dutch model would be suitable for large parts of the U.K., and there is large amounts of four-tracking too. It would still require a lot of infrastructure investment and the political will for it does not seem to be there. As it is, you have none of the two models and are also not on your way towards either of them, mainly because the UK has never quite figured out what it wants its railways to do and follow through with it even medium-term.

 

[Krokodil]

Most major cities in the UK couldn't manage a Takt. Partly because of the platform and junction capacity mentioned above (I imagine that you'd have to double the size of Birmingham New Street for example) but also because many cities don't have one central interchange, they have several. Manchester has Piccadilly and Victoria, Birmingham has services distributed between New Street, Moor Street/Snow Hill, and eventually Curzon Street. London meanwhile has a bunch of termini, some of which are quite poorly connected from the others. In those cities the Dutch model mentioned above is more appropriate. You're never going to make the 3tph between Euston and Birmingham or Manchester fit into a Takt timetable.

Out in the regions however, there is more scope. Chester is practically there already, with services converging around the xx:20 and xx:50 mark though the lack of through platforms and loss of most of the mid-platform crossovers mean that the extension of Crewe shuttles to Wrexham will be useless for passengers from the Holyhead-Manchester service. Shrewsbury can work too, and Bristol Temple Meads springs to mind as somewhere with potential.

 

[dutchflyer]

Just a short note from here in Nederland now: (tongue in cheek, I guess it would be named?!)
That Swiss Takt-modell-this was even quite clearly stated in the beginnings, was set up with that from here, NL in mind as example and as a kind of copy. I vaguely recall that in the spirit of ´state-railways always cooperate and help one another out!´ then some dutch timetable-makers were sent to Switzerland to help them out.
And-praise to the Brits across the North-Sea: even though the very basics/not yet national, start of the hourly ´Takt-model´ in central Nederland-pre war!-with timed ´all-around´ connections planned in it- was said to be (or at least described as such in later trade and enthusiast-publications) with that of the then Southern Region in mind and as example.
All around (like in Zürich) planned connections are now far less here in NL-Zwolle is the main one surviving-but cross-platform 2 mins connections (admittedly many of these only really do-able for those very much in the know) do exist in quite a few places. I often-coming from the airport-even use 2 of those within 12 mins to get home

 

[Bemined]

Even though it's a smaller station, Zutphen is a good example of a station that is designed around connections as well. The timetable utilizes the bidirectional functionality of the tracks to allow four trains to arrive at nearly the same time even though the approach to the station is only dual track on either side. All branch lines then offer a connection to the mainline intercity, while people getting off the intercity can change to any of the branch lines.

 

[Austriantrain]

Just a short note from here in Nederland now: (tongue in cheek, I guess it would be named?!)
That Swiss Takt-modell-this was even quite clearly stated in the beginnings, was set up with that from here, NL in mind as example and as a kind of copy. I vaguely recall that in the spirit of ´state-railways always cooperate and help one another out!´ then some dutch timetable-makers were sent to Switzerland to help them out.
And-praise to the Brits across the North-Sea: even though the very basics/not yet national, start of the hourly ´Takt-model´ in central Nederland-pre war!-with timed ´all-around´ connections planned in it- was said to be (or at least described as such in later trade and enthusiast-publications) with that of the then Southern Region in mind and as example.
All around (like in Zürich) planned connections are now far less here in NL-Zwolle is the main one surviving-but cross-platform 2 mins connections (admittedly many of these only really do-able for those very much in the know) do exist in quite a few places. I often-coming from the airport-even use 2 of those within 12 mins to get home

Thank you for this interesting historical perspective. I fully expect the Swiss to go towards a 15-Minute-frequency e.g. in the „golden triangle“ and I believe there are plans for it. For obvious reasons though, that will not be the case in much of the country.

I do believe though that on principle they will only do more than half-hourly where it is absolutely necessary, because the infrastructure costs are enormous and they consider the benefits not so large.

 

[NCT]

The Dutch and the Swiss models are not in real world competition though. The dutch one evolved in a country that in large parts always was very densely settled and where building up a national „S-Bahn“-style network was rational and feasible. It is true that if most things run every 15 minutes, you don’t need to build major hub stations because connections are always short.

The Swiss Takt was designed in the late 1970s/early 1980s in what was still a very rural country. Having hourly train services all around was revolutionary; a 15-Minute-service would have been an absurdity and impossible to finance. If you have an hourly service, you have to time your connections, and the same is still true for half-hourly services if overall journeys tend to be short and often involve more than one connection. Although Swiss demographics have changed somewhat, it is still quite different from the Netherlands.

Also, Switzerland often having three tiers of services and not long, but somewhat longish nonstop runs mean that anything more than the half-hourly service would need a very large amount of four-tracking, which historically has never existed in Switzerland and mostly still does not. It is anybody‘s guess whether a large amount of four-tracking would be cheaper than extending stations over and over. I would be doubtful.

I agree though that the Dutch model would be suitable for large parts of the U.K., and there is large amounts of four-tracking too. It would still require a lot of infrastructure investment and the political will for it does not seem to be there. As it is, you have none of the two models and are also not on your way towards either of them, mainly because the UK has never quite figured out what it wants its railways to do and follow through with it even medium-term.

I wasn't implying that the Swiss network should move to a 15-minute frequency. Was I was trying to say was this:

Where 2 sets of half-hourly services converge at a node:
- The Swiss application of 'takt' (or our understanding thereof) is both sets of services should arrive before 00/30 and depart after 00/30. Two directions x two sets of services mean four overlapping station dwells therefore four platform faces required. This is something the UK network generally cannot emulate with its existing infrastructure (most stations would need to double in size)
- In the Netherlands, at some nodes they are quite happy to run those two half-hourly services with a 15-minute (or so) offset. You only need 2 platform faces for the same amount of throughput per hour. This is something the vast majority of the UK network can emulate just by tidying up the timetable structure.

 

[Belperpete]

In the Netherlands, at some nodes they are quite happy to run those two half-hourly services with a 15-minute (or so) offset. You only need 2 platform faces for the same amount of throughput per hour. This is something the vast majority of the UK network can emulate just by tidying up the timetable structure.

The term "tidying up the timetable structure" sounds very begnine. In practice, unless you can magically strip time out of the schedules such as by making the trains go faster, what you are talking about in most cases is inserting extra time into the schedules. So that something that could run at say a 110 minute frequency runs at a 120 minute frequency. The extra ten minutes adds extra recovery time, which improves reliability. What is not to like?

Well, unless you can add those extra ten minutes into the terminal turn-round time (which may not be possible, e.g. if you don't have the platform capacity for trains to spend longer there), then you are going to have to add them into the journey time. Extending journey times does not sound so good. I understand this is effectively what happened on the Cambrian when it was converted to a regular(ish) two-hourly cycle, many journey times got longer.

I suspect that the Swiss takt model was likewise achieved at the expense of journey times. Swiss trains may be notable for their timekeeping and regular timetable, but less so for their speed. And unfortunately once you have locked into this takt model, there is little incentive to make minor upgrades to improve journey times, as in most cases all that would happen is that the train would spend longer waiting at its next node station.

 

[The exile]

The term "tidying up the timetable structure" sounds very begnine. In practice, unless you can magically strip time out of the schedules such as by making the trains go faster, what you are talking about in most cases is inserting extra time into the schedules. So that something that could run at say a 110 minute frequency runs at a 120 minute frequency. The extra ten minutes adds extra recovery time, which improves reliability. What is not to like?

Well, unless you can add those extra ten minutes into the terminal turn-round time (which may not be possible, e.g. if you don't have the platform capacity for trains to spend longer there), then you are going to have to add them into the journey time. Extending journey times does not sound so good. I understand this is effectively what happened on the Cambrian when it was converted to a regular(ish) two-hourly cycle, many journey times got longer.

I suspect that the Swiss takt model was likewise achieved at the expense of journey times. Swiss trains may be notable for their timekeeping and regular timetable, but less so for their speed. And unfortunately once you have locked into this takt model, there is little incentive to make minor upgrades to improve journey times, as in most cases all that would happen is that the train would spend longer waiting at its next node station.

In most cases “minor upgrades” to the journey times that could be made are negligible in comparison with the time already saved by having reliable connections - not just between trains but also between trains and buses. “Journey times” frequently seems to be measured by “time on train” rather than “time between closing my front door and arriving at my destination” - which is usually what really counts.

 

[NCT]

The term "tidying up the timetable structure" sounds very begnine. In practice, unless you can magically strip time out of the schedules such as by making the trains go faster, what you are talking about in most cases is inserting extra time into the schedules. So that something that could run at say a 110 minute frequency runs at a 120 minute frequency. The extra ten minutes adds extra recovery time, which improves reliability. What is not to like?

Well, unless you can add those extra ten minutes into the terminal turn-round time (which may not be possible, e.g. if you don't have the platform capacity for trains to spend longer there), then you are going to have to add them into the journey time. Extending journey times does not sound so good. I understand this is effectively what happened on the Cambrian when it was converted to a regular(ish) two-hourly cycle, many journey times got longer.

I suspect that the Swiss takt model was likewise achieved at the expense of journey times. Swiss trains may be notable for their timekeeping and regular timetable, but less so for their speed. And unfortunately once you have locked into this takt model, there is little incentive to make minor upgrades to improve journey times, as in most cases all that would happen is that the train would spend longer waiting at its next node station.

Some of the tidying up just involves shifting train times across the clockface so everything runs to a common symmetry. The Chester - Shrewsbury - Hereford already runs to a 'sort of' takt around xx20. Just shifting that to xx00 (or xx30) would make things easier in terms of, say everything around Manchester and Cardiff running to a common symmetry. Then there are low-hanging fruits like running London - Edinburgh and London - Bristol as symmetric timetables rather than 'half hourly at origin'. LNER's insistence on half-hourly at origin for the London - Edinburgh is one of the main reasons the Up direction doesn't work and doesn't work with GTR in the Dec 2025 timetable.

Yes, some extra time might be appropriate in some cases, but it needn't as bad as some make it out to be. Especially in a post HS2 world where the high-concentration high-value flows are migrated to dedicated point-to-point services, the remaining network will no longer have single dominant flows. The remaining network would be far more reliant on attracting a diverse range of journeys, and that means journeys involving interchanges. Getting interchanges to work and to work reliably becomes more effective at reducing generalised journey time across the board, and is in my view a more effective way of attracting passengers onto the railway, than chasing the small subset of no-change passengers by shortening isolated in-vehicle times.

It's not as though existing schedules are efficient. There are oodles of pathing time and excess dwell times all over the shop because we insist on shoehorning direct everywhere-to-everywhere trains into the network not designed for accommodating them. The same amount of slack (a 'slack budget' if you will) could be put to far better use for strategic / intended purposes.

 

[Austriantrain]

I suspect that the Swiss takt model was likewise achieved at the expense of journey times. Swiss trains may be notable for their timekeeping and regular timetable, but less so for their speed. And unfortunately once you have locked into this takt model, there is little incentive to make minor upgrades to improve journey times, as in most cases all that would happen is that the train would spend longer waiting at its next node station.

The Swiss were, when introducing the Takt, of the opinion, that somewhat slower journeys on direct trains were more than compensated by shorter journey times on connections. In a country where lots and lots of connections have to be made, I believe this is the right call.

That does not necessarily mean that it is the right call for every country. There is a difference if your Zurich - Geneva train needs 10 minutes more than the minimum (especially since the vast majority of passengers on such a train do not travel from Zurich to Geneva), or if London - Edinburgh (where a majority of people *will* be traveling end-to-end) takes half an hour longer.

Also, just as an aside: Swiss trains are not fast. Mainly, this is not because of the Takt but because of line speeds on their historic lines, which are low for topographical reasons. What the Swiss did, though, was to decide that with tiny exceptions, they do not need higher speeds (achieved by new HSL), because their country is so decentralised that these higher speeds would profit few people. BTW, that is not uncontroversial. There is a whole citizens' initiative complaining about the lack of fast connections to foreign rail networks (I will look it up later).

And as another aside: This is also why I believe that while the Swiss model is generally very useful for Austria, it needs to be tweaked. Even on domestic journeys, distances are much greater in Austria so that absolute journey times are more important here.

 

[NCT]

The Swiss were, when introducing the Takt, of the opinion, that somewhat slower journeys on direct trains were more than compensated by shorter journey times on connections. In a country where lots and lots of connections have to be made, I believe this is the right call.

That does not necessarily mean that it is the right call for every country. There is a difference if your Zurich - Geneva train needs 10 minutes more than the minimum (especially since the vast majority of passengers on such a train do not travel from Zurich to Geneva), or if London - Edinburgh (where a majority of people *will* be traveling end-to-end) takes half an hour longer.

I do believe that a connections-prioritised network design on the core north-south inter-city lines is only economically/politically tenable in a post HS2 world.

 

[Austriantrain]

I do believe that a connections-prioritised network design on the core north-south inter-city lines is only economically/politically tenable in a post HS2 world.

Same here. You certainly mean HS2 in full though, not the rump version?

 

[NCT]

Same here. You certainly mean HS2 in full though, not the rump version?

Absolutely. London / Birmingham - North West / Yorkshire / Scotland inter-city services all 'as fast as possible' (predominantly) segregated from the rest of the network.

The sections north of Preston and York would want to be integrated within a takt though, but that should be doable.

 

[The Pelican]

Personally I think the Takt system would be best pursued in a fairly polycentric UK reigon with lots of different grades of railway- Liverpool-Manchester-Leeds for example- and that there are suprisingly large areas of the country where it might not work particularly well- broadly the area North-West of London has a huge arc where nearly everything is a mainline, you cross the Great Western Mainline, Chiltern Mainline, the less-mainliney Chiltern Aylesbury line, then the West Coast, Midland and East Coast mainlines- in general in that area, connections won't tend to matter as much, because there's no other railways to connect to.
East West Rail will of course provide something to connect to, but that's only through one string of stations, and you'd need several East West rail style lines to create a grid of useful connections.

On the other hand, clockwise and anti-clockwise heading out from London from this area, the SWR Windsor Lines and the West Anglia Mainline are exactly the sort of branching networks where a Takt could do a lot of good.

 

[Austriantrain]

Absolutely. London / Birmingham - North West / Yorkshire / Scotland inter-city services all 'as fast as possible' (predominantly) segregated from the rest of the network.

The sections north of Preston and York would want to be integrated within a takt though, but that should be doable.

Makes perfect sense to me. BTW, that is why - as I wrote in the other thread where we got into each others hair - I always thought some elements of HS2 strange, for instance that there would be no chord from Leeds HS2 station towards York (at least as far as I know): While London - York as fast as possible is obvious, having separate Birmingham - Leeds and Birmingham - York trains never made a lot of sense to me: Birmingham - Scotland would have gone via Preston anyway, and a more frequent Birmingham - Leeds - York (-Newcastle) would seem much more useful than separate, less frequent trains (and Leeds HS2 could also have been used as an "overflow" for trains into Leeds from the North). While a time penalty on Birmingham - York, still much faster than today and very competitive.

But not only is this OT... HS2 seemingly will never extend that far anyway.

 

[NCT]

Makes perfect sense to me. BTW, that is why - as I wrote in the other thread where we got into each others hair - I always thought some elements of HS2 strange, for instance that there would be no chord from Leeds HS2 station towards York (at least as far as I know): While London - York as fast as possible is obvious, having separate Birmingham - Leeds and Birmingham - York trains never made a lot of sense to me: Birmingham - Scotland would have gone via Preston anyway, and a more frequent Birmingham - Leeds - York (-Newcastle) would seem much more useful than separate, less frequent trains (and Leeds HS2 could also have been used as an "overflow" for trains into Leeds from the North). While a time penalty on Birmingham - York, still much faster than today and very competitive.

But not only is this OT... HS2 seemingly will never extend that far anyway.

I still consider eventual full HS2 to be realistic within the normal British political pendulum, so it is part of a plausible medium to long-term solution mix for getting the GB network to be more reliable in the future (letting the WCML and ECML work less hard).

Birmingham - Leeds/York having separate trains makes perfect sense. The infrasructure is optimised for the largest London - Leeds / London - York 'as fast as possible' markets and a surface terminus station on the underdeveloped south side of Leeds city centre is far cheaper to build than an underground through station.

Birmingham - Leeds at 2tph and Birmingham - York - Newcastle at 1tph are still a good service level.

There's far more civilisation to the south of Yorkshire than to the north of it. Manchester will send 6tph into Leeds from the west, and 2 from Blackpool. These 6 going to York will be entirely sufficient for the 'Leeds to the further north' market. There is no additional 'overflow' to run. There will be a lot of trains running to York from the south, far more than the number of trains needed to service York - Newcastle. So again, there's no reliance on a more frequent service level from Birmingham.

 

[Austriantrain]

Birmingham - Leeds/York having separate trains makes perfect sense. The infrasructure is optimised for the largest London - Leeds / London - York 'as fast as possible' markets and a surface terminus station on the underdeveloped south side of Leeds city centre is far cheaper to build than an underground through station.

I was not speaking about an underground through station though, but of a chord that allows trains from Leeds HS2 to use HS2 to reach York. A bit of a difference on costs...

 

[The Pelican]

HS2's real problem was always that the two fast tracks would probably fill up too quickly, and then three mainlines would struggle to accommodate any further growth. A version with four tracks from Birmingham Interchange to Euston would have solved most of that, but apparently would have cost almost twice as much. Failing that, an A framed network with a high speed Nottingham to Kings Cross line would also have helped, but probably wouldn't have been built for an extremely long time. So in some ways I prefer the version as it was in 2021 or so, when it reached Manchester but not towards the East coast.

This relates to the general issue with the British railway system, which is that it's reasonably easy to keep the competing interests happy, as long as the generous provision that the Victorians left is still there to make e.g. regional, Intercity and freight trains able to be accommodated.

 

[The exile]

There's far more civilisation to the south of Yorkshire than to the north of it.

Far more population, maybe?!

 

[Tester]

Far more population, maybe?!

Ouch

 

[Krokodil]

I suspect that the Swiss takt model was likewise achieved at the expense of journey times. Swiss trains may be notable for their timekeeping and regular timetable, but less so for their speed. And unfortunately once you have locked into this takt model, there is little incentive to make minor upgrades to improve journey times, as in most cases all that would happen is that the train would spend longer waiting at its next node station.

Are there any interchanges in Switzerland where the trains converge at a non-standard time? Perhaps somewhere that is unevenly spaced from its neighbours on all sides so that the symmetry minute was xx:15/xx:45

 

[NCT]

Are there any interchanges in Switzerland where the trains converge at a non-standard time? Perhaps somewhere that is unevenly spaced from its neighbours on all sides so that the symmetry minute was xx:15/xx:45

There is a mixture of 00/30 nodes and 15/45 nodes in Switzerland. I've just been looking at Thun and Spiez - they don't appear to be 'nationally strategic' (if that's a phrase) nodes but there are a number of routes converging at both locations. They appear to be neither 00/30 nor 15/45 nodes, yet there seems to be something that works.

== Doublepost prevention - post automatically merged: 3 Nov 2025 ==

I was not speaking about an underground through station though, but of a chord that allows trains from Leeds HS2 to use HS2 to reach York. A bit of a difference on costs...

OK, that would be far cheaper but would make the network more complicated and in my view go against making the network more reliable. A 'Stourton link' was looked at as an add-on scheme and if memory serves me correctly had a pretty poor case.

Taking inspiration from Prorail's philosophy to rebuilding Utrecht Centraal - for a major urban hub you want each set of approach tracks to do one thing. Everything from HS2 terminates in the new HS2 shed; everything from the Transpennine route goes through to York and Hull.

There are enough trains from Manchester (and Blackpool) to run through Leeds to York and Hull; additional services to York from Birmingham would represent excessive train capacity and overload the network. You also introduce a many-to-many relationship across a key node which is very un-Dutch (and un-Japanese).

If there is a contrast between the Swiss and Dutch approaches then I think it is this (possibly slightly exaggerated for effect) -

- The Swiss build a lot of platforms and grade separation to allow trains to enable many-to-many relationship across key nodes and resilience is achieved through lots of flexibility and redundancy;
- The Dutch are making their network Japanese style simplified to restrict cross-node relationships to one-to-one, and reliability is achieved through simplicity.

Only the Dutch model would get past the UK Treasury.

 

[Richard Scott]

- The Swiss build a lot of platforms and grade separation to allow trains to enable many-to-many relationship across key nodes and resilience is achieved through lots of flexibility and redundancy;

Not everywhere, still plenty of single line sections, which can cause severe knock on chaos if there's even a slight delay. Experienced it on more than one occasion and chaos was fairly widespread not just on that section.

 

[NCT]

Not everywhere, still plenty of single line sections, which can cause severe knock on chaos if there's even a slight delay. Experienced it on more than one occasion and chaos was fairly widespread not just on that section.

So grass is greener and all that ...

 

[bahnause]

I suspect that the Swiss takt model was likewise achieved at the expense of journey times. Swiss trains may be notable for their timekeeping and regular timetable, but less so for their speed. And unfortunately once you have locked into this takt model, there is little incentive to make minor upgrades to improve journey times, as in most cases all that would happen is that the train would spend longer waiting at its next node station.

In order to implement the takt working in the nodes, numerous sections had to be extended or rebuilt. A key element of the concept was the new line between Rotrist and Mattstetten (1.7 billion Swiss francs). In the case of the line across the southern foot of the Jura, selective upgrades were carried out to enable the use of tilting trains and to achieve the necessary reduction in journey times. Other projects in the first stage included the Adlertunnel between Liestal and Muttenz (CHF 387 million), the Zimmberberg Bastunnel between Zurich and Thalwil (CHF 945 million), upgrades to Zurich HB, Basel SBB and Aarau (including a second city tunnel) and a whole lot more. This results in the following journey time reductions (based on the 2005/2008 timetable):

Olten–Bern from 40 minutes to 26 minutes
Zurich–Bern from 69 minutes to 56 minutes
Basel–Bern from 67 min to 55 min
Lucerne–Bern from 81 min (via Wolhusen–Langnau) to 60 min (via Sursee–Zofingen-NBS)
Olten–Biel/Bienne from 41 min to 32 min (via NBS and upgraded Inkwil-Derendingen line)
Zurich–Bern–Geneva from 2:56 hrs to 2:43 hrs
Zurich–Biel/Bienne–Geneva from 3:09 hrs to 2:42 hrs

The opening of the Lötschberg Base Tunnel in 2007 reduced the journey time between Visp and Bern to less than an hour, enabling the establishment of a full node in Visp; at the same time, the journey time from Zurich to Valais was reduced to less than two hours. In addition, a full node was established in Interlaken.

 

[Austriantrain]

There is a mixture of 00/30 nodes and 15/45 nodes in Switzerland. I've just been looking at Thun and Spiez - they don't appear to be 'nationally strategic' (if that's a phrase) nodes but there are a number of routes converging at both locations. They appear to be neither 00/30 nor 15/45 nodes, yet there seems to be something that works.

== Doublepost prevention - post automatically merged: 3 Nov 2025 ==



OK, that would be far cheaper but would make the network more complicated and in my view go against making the network more reliable. A 'Stourton link' was looked at as an add-on scheme and if memory serves me correctly had a pretty poor case.

Taking inspiration from Prorail's philosophy to rebuilding Utrecht Centraal - for a major urban hub you want each set of approach tracks to do one thing. Everything from HS2 terminates in the new HS2 shed; everything from the Transpennine route goes through to York and Hull.

There are enough trains from Manchester (and Blackpool) to run through Leeds to York and Hull; additional services to York from Birmingham would represent excessive train capacity and overload the network. You also introduce a many-to-many relationship across a key node which is very un-Dutch (and un-Japanese).

If there is a contrast between the Swiss and Dutch approaches then I think it is this (possibly slightly exaggerated for effect) -

- The Swiss build a lot of platforms and grade separation to allow trains to enable many-to-many relationship across key nodes and resilience is achieved through lots of flexibility and redundancy;
- The Dutch are making their network Japanese style simplified to restrict cross-node relationships to one-to-one, and reliability is achieved through simplicity.

Only the Dutch model would get past the UK Treasury.

The two models don’t achieve the same though. They almost might if everything runs every 15 minutes, but at less than that - the usual 30 Minute Standard - the Swiss way will achieve significantly shorter overall journey times especially if there is more than one connection involved.

I don’t dispute that it comes with a huge bill and that not everyone will want to pay it; for the U.K. it would probably unaffordable (not because you are so poor but because the country is so large, and as you rightly pointed out your stations don’t really lend themselves to Expansion) The Swiss do however, as they have stated several times at various referenda.

 

[Belperpete]

In order to implement the takt working in the nodes, numerous sections had to be extended or rebuilt. A key element of the concept was the new line between Rotrist and Mattstetten (1.7 billion Swiss francs). In the case of the line across the southern foot of the Jura, selective upgrades were carried out to enable the use of tilting trains and to achieve the necessary reduction in journey times. Other projects in the first stage included the Adlertunnel between Liestal and Muttenz (CHF 387 million), the Zimmberberg Bastunnel between Zurich and Thalwil (CHF 945 million), upgrades to Zurich HB, Basel SBB and Aarau (including a second city tunnel) and a whole lot more. This results in the following journey time reductions (based on the 2005/2008 timetable):

Olten–Bern from 40 minutes to 26 minutes
Zurich–Bern from 69 minutes to 56 minutes
Basel–Bern from 67 min to 55 min
Lucerne–Bern from 81 min (via Wolhusen–Langnau) to 60 min (via Sursee–Zofingen-NBS)
Olten–Biel/Bienne from 41 min to 32 min (via NBS and upgraded Inkwil-Derendingen line)
Zurich–Bern–Geneva from 2:56 hrs to 2:43 hrs
Zurich–Biel/Bienne–Geneva from 3:09 hrs to 2:42 hrs

The opening of the Lötschberg Base Tunnel in 2007 reduced the journey time between Visp and Bern to less than an hour, enabling the establishment of a full node in Visp; at the same time, the journey time from Zurich to Valais was reduced to less than two hours. In addition, a full node was established in Interlaken.

Exactly my point. These were not minor upgrades, they involved significant spends. Unfortunately in the UK we don't seem to like spending significant sums to get major improvements, we seem to prefer spending as little as possible. Hence all the arguments about tweaking the existing network rather than building a high speed line.

 

[bahnause]

There is a mixture of 00/30 nodes and 15/45 nodes in Switzerland. I've just been looking at Thun and Spiez - they don't appear to be 'nationally strategic' (if that's a phrase) nodes but there are a number of routes converging at both locations. They appear to be neither 00/30 nor 15/45 nodes, yet there seems to be something that works.

The easiest way to get an overview over the whole system and the nodes is the "Netzgrafik" (network graphic). It's almost a work of art.

 

[eldomtom2]

Unfortunately in the UK we don't seem to like spending significant sums to get major improvements, we seem to prefer spending as little as possible.

Probably because when we do spend significant sums to get major improvements, costs balloon and we end up paying far more than another country would.