How would a national regular interval timetable be planned?

[ChiefPlanner]

I expected to be shot down to be honest !

Horses for courses , know your markets and your routes and resources is what I say.


Anyway - in another life when what became the 2008 timetable on West Coast , was preceded by the 2004 timetable which took several years to nail down. Some very long and challenging meetings with all sorts of issues. Got there in the end though. So where did we start .....and opponents of HS2 listen carefully.


THE COVENTRY CORRIDOR - (New Street to Rugby Trent Valley actually) , once some hard decisions were done there , services were "extended" towards that there Euston - north to Glasgow and the big Northern Cities with the pattern through Rugby hardwired and treated like a bowling alley.

The WCML slow lines welded in - with the starting proviso of a half decent Northampton service (extended on of course) , with 2 class 4 freights and a class 6 as needed) , as the actual base.

So a pure Takh would have been very challenging , despite the access rights issued at the start of the WCML upgrade.........

 

[Bletchleyite]

Though off-peak the WCML is pretty close to a Takt. It might not have the connectional aspect in the same way, but it does have the clockface bit.

 

[30907]

First, apologies for the error on Bern-Zuerich upthread: caused by not looking at the bottom of the page in the ERT! I actually thought it was half hourly too.

Back on topic - the Swiss system works because the country is polycentric and the top rank trains are are key part; Germany is if anything even more polycentric and AFAIK the proposed Deutschlandtakt 2030 will work the same way.
The UK is heavily London-centric and the top rank trains would mostly (not entirely) not be key links between the Taktknoten (key connection centres).
I think this complicates the exercise - a case of tail wagging dog?

 

[Bald Rick]

As has often been said by train planners in this country, the Swiss system works as all their major cities are multiples of 30 minutes apart

 

[MontyP]

Pretty much every main line in the country now runs to a standard hourly pattern - all the intercities, most inter-regional (e.g. TPE, Liverpool-Norwich) and most suburban. The thing that is missing is the connectivity at interchanges such as Doncaster .

 

[Bletchleyite]

As has often been said by train planners in this country, the Swiss system works as all their major cities are multiples of 30 minutes apart

That didn't happen by accident. It happened because they don't do improvements to infrastructure purely to gain speed - they do it to get closer to that ideal. In some cases they deliberately slow trains down by adding stops to achieve it.

 

[Meerkat]

I am still not seeing what the great advantage is that overwhelms the disadvantages.
Am I correct in thinking all trains have to come in to an interchange, then wait til the last one in has sat for the minimum connection time for that station, then they all leave again? That’s a lot of platforms and a lot of dwell time....

 

[Bletchleyite]

I am still not seeing what the great advantage is that overwhelms the disadvantages.
Am I correct in thinking all trains have to come in to an interchange, then wait til the last one in has sat for the minimum connection time for that station, then they all leave again? That’s a lot of platforms and a lot of dwell time....

All trains where it would be useful, yes. There's little point in a train connecting with itself going back the other way. So at some points you might have two times of the hour when it happens, one one way and one the other.

The great advantage is a smooth connectional journey from anywhere to anywhere without waiting around.

The advantage of having it mid-journey is resilience and punctualilty, as you don't just put layover at the end as the UK has a habit of doing.

 

[Meerkat]

Aren’t you just making connections shorter by slowing down trains - is there actually any time gained?
What proportion of passengers actually use connections, compared to those going direct who will suffer demand cutting significant slowdowns?

 

[Bletchleyite]

Aren’t you just making connections shorter by slowing down trains - is there actually any time gained?

Not necessarily. You also speed up trains by removing stops. The point is your timetable is designed around connectivity rather than raw speed, because that connectivity gives you speed on a connectional journey and punctuality and reliability on a direct one.

What proportion of passengers actually use connections, compared to those going direct who will suffer demand cutting significant slowdowns?

In countries where Taktfahrplaene are established and successful, connections are the done thing, just as nobody thinks twice about making connections on the London Underground, because they are easy, consistent and reliable. In the UK people prefer direct services because connections are often awkward, badly-timed and unreliable. This causes people to do some journeys by other modes or to drive to the mainline railhead to avoid the change - both things I'm sure you would agree would be best if they stopped.

 

[Meerkat]

Not necessarily. You also speed up trains by removing stops. The point is your timetable is designed around connectivity rather than raw speed, because that connectivity gives you speed on a connectional journey and punctuality and reliability on a direct one.



In countries where Taktfahrplaene are established and successful, connections are the done thing, just as nobody thinks twice about making connections on the London Underground, because they are easy, consistent and reliable. In the UK people prefer direct services because connections are often awkward, badly-timed and unreliable. This causes people to do some journeys by other modes or to drive to the mainline railhead to avoid the change - both things I'm sure you would agree would be best if they stopped.

How do you remove stops?
Everything is slower - the connection time is shortened by slowing things down so they connect.
I think people like direct trains because they don’t have to get out their seat and faff about on a cold station. The details of the connection are less important.
On the important lines services are frequent enough for connection times to be less important, and then the branches should be made to connect - I don’t see that you need an inflexible overall plan to make that happen??
Ps wouldn’t having all services connecting at the same time be absolute chaos and grossly overcrowded?

 

[Bletchleyite]

How do you remove stops?

By deleting them from the timetable.

Everything is slower - the connection time is shortened by slowing things down so they connect.

As I said it is not that simple.

I think people like direct trains because they don’t have to get out their seat and faff about on a cold station. The details of the connection are less important.
On the important lines services are frequent enough for connection times to be less important, and then the branches should be made to connect - I don’t see that you need an inflexible overall plan to make that happen??
Ps wouldn’t having all services connecting at the same time be absolute chaos and grossly overcrowded?

I'd really suggest going to Switzerland and seeing how it works so as to make a more informed view, even if that view is still against. It's a beautiful (if a bit expensive) country well worth a visit.

 

[Meerkat]

By deleting them from the timetable.



As I said it is not that simple.



I'd really suggest going to Switzerland and seeing how it works so as to make a more informed view, even if that view is still against. It's a beautiful (if a bit expensive) country well worth a visit.

closing stations??
It seems to involve a lot of padding, just using that padding in stations. Padding hasn’t been popular in the past.
I have been to Switzerland, though driving (anyone who likes driving MUST have a go at the alpine passes, just awesome!). It is beautiful in the mountains, but everyone seems to ignore that the valleys are REALLY ugly (motorway, electric railway, canalised rivers, bland blocks of flats and industrial estates - lots of concrete) and in summer the ski infrastructure is pretty minging (huge car parks and steel/concrete structures). I would drive there a lot but for France - it’s like the Midlands...a huge dull expanse you have to cross to get to the interesting places!

I don’t think we will agree on he Takt. I can see why you like it, but I just see it as not enough gain to justify significant disruption, and a lot of spending that could achieve more targeted at capacity.

 

[Bletchleyite]

It seems to involve a lot of padding, just using that padding in stations. Padding hasn’t been popular in the past.

Padding is necessary for punctuality. Even without a Takt we need more, not less of it.

Closing stations? Not necessarily. Changing stopping patterns.

 

[MarkyT]

Once you get to much higher frequencies on most routes the planned synchronisation seen at a handful of major stations in Switzerland under 'takt' becomes largely irrelevant anyway. I don't disagree with it at all though. SBB and partner companies were able to do it quite easily with the facilities they had available there and although it's quite resource intensive, it produces a very reliable easy to use service that is very popular; the Swiss use trains and other public transport more than in any other country in Europe I believe. Long waits can still occur at some more minor hubs though for certain journeys so no way is it a universal panacea. Local 'takts' in extremities of the network, based around traditional times of longer distance expresses would be a good idea in UK. They wouldn't be around top of the hour though, just as they aren't actually at minor junctions in Switzerland.

 

[dm1]

The point of the Takt isn't that everything is at the top of the hour everywhere, but rather that it's at the same time every hour - if two nodes are 15 minutes apart on the same line there's no reason why the takt can't be phase shifted by 15 minutes between them as well. It's the regularity that's the important part, rather than at what time past the hour it happens.

Even two hourly where that is sufficient is a perfectly legitimate takt interval. In Switzerland many of the interbational EC trains are scheduled as such.

 

[Bletchleyite]

It also doesn't require all stations to be 30 minutes apart - that's oversimplistic. Nor do all trains have to connect - just the ones where it's useful. Connecting a train with its counterpart going back the other way is fairly pointless.

If we use the south WCML as an example (and I know this won't be feasible pre-HS2), what you might want to do is set up, for example, that an all-stops Euston-Northampton meets an IC to Manchester at MKC, so that passengers from local stations south of MKC can travel onwards to Manchester, and so that people having travelled fast from Euston can change for Wolverton and Northampton.

Or, in the days when there was a separate Scotland-Brum service, you might want that, at Crewe, to meet the Trent Valley stopper. It wouldn't be important to make that meet in both directions, because all the main WCML origins have direct services to Birmingham and Euston anyway. You might also want it to meet a fast Euston service calling at Rugby and/or MKC because those weren't at the time served by a direct service from Scotland.

 

[BrianW]

Hello again ...

A good reminder, thank you, that it's hard to connect everything with everything, whether direct or by changing; also balancing speed with frequency. Choices need to be made. It's also of course a bit of a nuisance to have to recast timetables whenever one service is speeded up, eg with electrification of GWML or MML and esp if hoping to weave in with XC.

I seem to recall an attempt to 'badge' Birmingham NS as 'the heart of the country' with West Coast electrification- all finely crafted connections thrown out by delay of 'connecting' services. Also hard to connect into and out of services. Are there stats regarding take up of particular connections? Difficult of course also to know unserved 'demand'. As others have observed if connections are not good, it'll be the car, or nothing.

I stopped taking train from Stroud to go north- ok in the morning but coming back the hoped for 'connection' had left Cheltenham to connect on time at Swindon for the much greater number of London-bound passengers.

Frankly I'm amazed timetables work at all. London tube great, yes, but frequent, regular and reliable- only really needing to 'connect' at eg Earls Court, Aldgate-ish. It was no surprise to me the Thamesink problems Farringdon-Blackfriars a year or two back. Dare I mention Ordsall Chord here- is it really only one tph Victoria-Piccadilly? That's a lot of hassle for the gain.

 

[Senex]

A lot of work in preparation for a national Taktfahrplan is being done in Germany, and the infrastructure improvements needed are very much a focus of railway planning in the current Bundesverkehrswegeplan. It's really interesting to see how complicated it is trying to do this sort of job on a large system in a large country, as opposed to the much more compact Swiss system where these ideas have been most fully developed. For drafts of how Taktfahrpläne might look for the various German states, see https://www.bmvi.de/SharedDocs/DE/Artikel/G/BVWP/bundesverkehrswegeplan-2030-deutschlandtakt.html.
A more recent iteration of the long-distance version than that in the list given is the one attached.

 

[yorkie]

Unlocked to enable @squizzler to be able to post

 

[squizzler]

Thanks for re-opening the thread. I had been wondering if there was much to be said for re-examining the benefits of a national Taktfahrplan in light of the recent discussions over reducing train frequencies on some routes after covid. I asked for this thread to be reopened - rather than start a new one - as it mainly seemed to take place before the implications of covid on the train services had emerged (not that it has done so yet!). It is the anniversary of the thread, and yet so much has changed in the meantime!

• The government is suggesting a reduction of train frequencies after Covid
• The replacement of franchises by concessions, and the end of ORCATS revenue distribution, seems likely to reduce the duplication of services over the same section of line by rival TOCs and deliberate favouring of short frequent trains, again meaning fewer departures from a given station.
• The rush hour is much diminished in importance, so the timetable does not need to be designed around making sure units are in the correct place for an hour or two twice a day.
• The reduced traffic means that the services which remain can be diagrammed to interconnect better.
• The slogan 'build back better' is in the public consciousness and given as a reason for electrification and better signalling. A true regular interval timetable will also mean a lot of infrastructure work.

All this makes me think that a regular interval railway might be what we need to aim at now, even if it was not the optimal solution a year ago. I also wonder if a regular interval timetable with easy interchange at key stations could be an easier sell to the general public than electrification when non-technical people can be persuaded that hydrogen and batteries can do the same thing. Of course it should not be one or the other - electrification and Taktfahrplan go together well!

 

[Bald Rick]

Don’t forget that a substantial part of the railway already operates on a regular interval basis. That some people don’t like some of the connections is unfortunate, and often because they don’t see the bigger picture. (You can’t connect everything with everything).

Nevertheless rewriting the whole Network timetable is a task that would take 5 years or so to implement, in stages.

 

[squizzler]

It would certainly be a long term commitment, but on the other hand it might take that long to get back the traffic we expected this year. Arguably that is what might make it a good time to start planning.

It will also take that sort of time to put in the necessary infrastructure interventions. I will understand the true Taktfahrplan principles are more than just a train every hour. It also means cross platform changes with multiple trains arriving and departing at the same time, meaning more platforms might be needed at some stations. Even line speed improvements are done by the Swiss to allow trains to connect.

 

[ChiefPlanner]

Don’t forget that a substantial part of the railway already operates on a regular interval basis. That some people don’t like some of the connections is unfortunate, and often because they don’t see the bigger picture. (You can’t connect everything with everything).

Nevertheless rewriting the whole Network timetable is a task that would take 5 years or so to implement, in stages.

Some of us know how complicated and even soul destroying timetable planning can be. There is almost a "takhtfarplan" cult in certain groups. A pure "takht" is almost like the geographical concept of pure location of perfect distribution of cities and settlements on an idealised Great Plain chessboard *. Life ain't like that and you have to cater for actual journey flows and real people decisions on where and when they want to travel. A classic example given to me some years ago , was the North and West (Newport to Crewe) , where a simplistic view was "fast" trains for the interurban with a few slow trains connecting into an out of places like Hereford and Shrewsbury. In real life you have important school and commuting flows which called for stops in the "fast" trains at certain times of day , and where the slow trains were wrongly placed for them , with operating and other constraints / considerations to manage.

A standard pattern timetable - say Worcester to Leamington via Snow Hill - has worked very well for more years than I care to consider - but it does not slavishly "takht" with a whole range of other , non related services - but best efforts made to fit in with and connect where realistic needs are identified. The Welsh effort through the Arriva bit a longtime ago had to be tinkered with on a much greater geographical area , with lots of constraints and issues - but done for "compelling" reasons.

Even the London Underground - an impressive operation - has all sorts of subtle changes to "break" a perfect timetable - greater dwell times in normal passenger journey times and a ramp down from high peak to off peak and late evening.

All very well pointing (as has been done many times -trust me) for Switzerland and so on , Holland and bits of Germany - but to quote a well respected railwayman "ain't neccessarily so" - life is much more complicated.

By the way - timetables cost a lot of resources and money.

(* for geographers - Christaller and Loesch)

 

[Bald Rick]

Some of us know how complicated and even soul destroying timetable planning can be. There is almost a "takhtfarplan" cult in certain groups. A pure "takht" is almost like the geographical concept of pure location of perfect distribution of cities and settlements on an idealised Great Plain chessboard *. Life ain't like that and you have to cater for actual journey flows and real people decisions on where and when they want to travel. A classic example given to me some years ago , was the North and West (Newport to Crewe) , where a simplistic view was "fast" trains for the interurban with a few slow trains connecting into an out of places like Hereford and Shrewsbury. In real life you have important school and commuting flows which called for stops in the "fast" trains at certain times of day , and where the slow trains were wrongly placed for them , with operating and other constraints / considerations to manage.

A standard pattern timetable - say Worcester to Leamington via Snow Hill - has worked very well for more years than I care to consider - but it does not slavishly "takht" with a whole range of other , non related services - but best efforts made to fit in with and connect where realistic needs are identified. The Welsh effort through the Arriva bit a longtime ago had to be tinkered with on a much greater geographical area , with lots of constraints and issues - but done for "compelling" reasons.

Even the London Underground - an impressive operation - has all sorts of subtle changes to "break" a perfect timetable - greater dwell times in normal passenger journey times and a ramp down from high peak to off peak and late evening.

All very well pointing (as has been done many times -trust me) for Switzerland and so on , Holland and bits of Germany - but to quote a well respected railwayman "ain't neccessarily so" - life is much more complicated.

By the way - timetables cost a lot of resources and money.

(* for geographers - Christaller and Loesch)

Wise words as ever, Guv.

It also helps when the Swiss have their main cities 60 minutes apart

 

[Bletchleyite]

Wise words as ever, Guv.

It also helps when the Swiss have their main cities 60 minutes apart

Of course, what they do is design the infrastructure to reliably support the extended Takt (which doesn't just include speeding stuff up, it can also include artificially slowing it down too, e.g. inserting small local stops into ICs) rather than building something seemingly random and cramming as much into it as we can.

It's a holistic approach.

 

[ChiefPlanner]

It would certainly be a long term commitment, but on the other hand it might take that long to get back the traffic we expected this year. Arguably that is what might make it a good time to start planning.

It will also take that sort of time to put in the necessary infrastructure interventions. I will understand the true Taktfahrplan principles are more than just a train every hour. It also means cross platform changes with multiple trains arriving and departing at the same time, meaning more platforms might be needed at some stations. Even line speed improvements are done by the Swiss to allow trains to connect.

Best start doing some financial business cases for that then ? - trust me , there will not be any money around for a long , long time.

 

[geoffk]

It would certainly be a long term commitment, but on the other hand it might take that long to get back the traffic we expected this year. Arguably that is what might make it a good time to start planning.

It will also take that sort of time to put in the necessary infrastructure interventions. I will understand the true Taktfahrplan principles are more than just a train every hour. It also means cross platform changes with multiple trains arriving and departing at the same time, meaning more platforms might be needed at some stations. Even line speed improvements are done by the Swiss to allow trains to connect.

Cross platform interchange would be important but achieving it at many major stations would mean track layout alterations and resignalling. I've seen it work well in Gemany and the Netherlands but without it you still have passengers dashing over bridges and through subways.

 

[JamesRowden]

Perhaps, as one of the ways to fill the holes in connectivity, we should look at running express buses to connect with long distance trains with through ticketing. A big advantage being that you don't need to build a new railway line to run a connecting route to the existing long distance route. At the moment the bus operators aren't encouraged to run express buses to connect with trains as generally the bus services are not advertised on railway journey planners or maps, and the bus operators don't get a proportion of the long distance rail fare which may not have been bought if the express bus service didn't exist.

 

[Bletchleyite]

Perhaps, as one of the ways to fill the holes in connectivity, we should look at running express buses to connect with long distance trains with through ticketing.

That's certainly something places with a bitty rail network such as Wales should be doing. It's inexplicable that they actively decided (it's minuted somewhere) not to build Trawscymru into this.