Is the whole GB timetable interconnected?

[A S Leib]

How much effort is put into keeping LNER Edinburgh departures at King's Cross and (for most of the day) Bristol departures at Paddington on the hour and half hour?

 

[louis97]

Nope, as this doesn't give the optimum allocation of capacity through the pinch points on the network, you would start with those and work outwards

This is spot on, but it goes against what you said previously in a way. You can design a timetable around New Street, but then find out your Birmingham-Stansted trains present at the single track Syston South to East (other locations are available) at the same time and your timetable at New Street now no longer works.

New Street, the Thameslink Core and Castlefield Corridor are not really where you'd start if the whole timetable is to be re-written. You'd want to be looking at the network geographically and identifying as you've said the pinch points: Syston South to East, Hazel Grove High Level to Station, Ely North Junction, Worcester Foregate Street, etc. Once you have designed a timetable for those areas you can then look at the interface between them and toggle your timetable around the hour to see what fits best. It is unlikely you'll get a perfect fit and you'll need to amend each area slightly, but you'll be making the best use of the pinch points in each area.

 

[Djgr]

While there's a small amount of interaction at South Parkway-Hunts X and Chester, Merseyrail is as close to standalone as you're going to get.

Is there any at Chester given that Merseyrail uses a bespoke platform?

 

[IrishDave]

Check how little the Cross Country Voyager timings have changed over the last 20 years, I think you'll find they generally stay pretty fixed and timetable changes elsewhere are designed to work around them.

Absolutely, some of the CrossCountry Voyager paths are pretty similar to where they started in the Operation Princess recast in September 2002, and most have barely changed at all since the December 2008 recast (when the WCML was recast, and they disentangled Virgin West Coast from what was now Arriva CrossCountry).

What's interesting is I don't think those Voyager paths are the oldest unchanged paths (or at least broadly unchanged) on the network. The EMR Liverpool-Norwich paths haven't fundamentally changed since about 1992 - they might have been tweaked by minutes here and there but they're still fundamentally the same - because they cross WCML, MML, ECML and GEML and remain fixed when any one of those gets recast, but never do they get recast themselves.

== Doublepost prevention - post automatically merged: 2 Oct 2024 ==

This is spot on, but it goes against what you said previously in a way. You can design a timetable around New Street, but then find out your Birmingham-Stansted trains present at the single track Syston South to East (other locations are available) at the same time and your timetable at New Street now no longer works.

New Street, the Thameslink Core and Castlefield Corridor are not really where you'd start if the whole timetable is to be re-written. You'd want to be looking at the network geographically and identifying as you've said the pinch points: Syston South to East, Hazel Grove High Level to Station, Ely North Junction, Worcester Foregate Street, etc. Once you have designed a timetable for those areas you can then look at the interface between them and toggle your timetable around the hour to see what fits best. It is unlikely you'll get a perfect fit and you'll need to amend each area slightly, but you'll be making the best use of the pinch points in each area.

None of those "pinch points" are worked to maximum capacity in the same way, though. For the Thameslink core, for example, there are (in theory) 24 paths available each hour. The only way to achieve all 24 is to ensure parallel moves at Blackfriars Junction - without that, you're going to have a conflict somewhere and that'll knock out one or more paths. Therefore you have to start planning at least most of the south-east of England from Blackfriars Junction.

The single track bit at Syston is indeed a pinch-point, but it probably doesn't limit the number of trains you can actually run through Leicester - that will probably be a combination of the conflicting moves at Syston South Junction and at Wigston North Junction. If you design the timetable around New Street and then find the Birmingham-Stansted trains present at Syston at the same time, then you just add some pathing time to make it work.

Ultimately that's what pathing time is for - it's needed where a train must pass location A no later than a certain time, but must then pass a subsequent location B no earlier than another certain time. You're never going to be able to make a national timetable without pathing time, not unless literally every possible conflicting move is grade-separated!

 

[A S Leib]

Is there any at Chester given that Merseyrail uses a bespoke platform?

And does TfW Rail have its own platform at Bidston?

 

[Dr Day]

Cardiff Valleys? Or is the interface at Cardiff West and Leckwith Loop and potential for freight too significant?

 

[louis97]

None of those "pinch points" are worked to maximum capacity in the same way, though. For the Thameslink core, for example, there are (in theory) 24 paths available each hour. The only way to achieve all 24 is to ensure parallel moves at Blackfriars Junction - without that, you're going to have a conflict somewhere and that'll knock out one or more paths. Therefore you have to start planning at least most of the south-east of England from Blackfriars Junction.

I see where you are coming from about Blackfriars Junction, a great example of where you can design a really local timetable, even just Elephant & Castle/London Bridge to St Pancras, and then toggle other wider area timetables to match the paths and therefore making best use of the capacity available through the junction. This could then be hooked onto the timetable in the South east and dictate that timetable somewhat.

There are however junctions which are far more restrictive as a pinch point, that have a much wider impact due to the nature of the services passing through. Consider the service groups passing through Blackfriars Junction, there are many and that gives flexibility in what service group can go into each of the 24 paths available, the only limitation is whether they need a path to/from London Bridge or Elephant and Castle.

At Ely North Junction on the other hand you are far more restricted in what you can do parallel moves wise due to the layout of the junction and double red signalling approaching it means adding pathing time increases capacity used between the station and North Junction more than some other locations, the single lead junctions to both Kings Lynn and Norwich mean there is no parallel moves to/from these branches. Therefore designing a timetable around this will allow the best use of the capacity through the junction and reduce wasted capacity through pathing that could be avoided.

The single track bit at Syston is indeed a pinch-point, but it probably doesn't limit the number of trains you can actually run through Leicester - that will probably be a combination of the conflicting moves at Syston South Junction and at Wigston North Junction. If you design the timetable around New Street and then find the Birmingham-Stansted trains present at Syston at the same time, then you just add some pathing time to make it work.

You can add pathing time, but that eats up capacity, and that could be avoided if the trains just didn't pass near the pinch point.

Ultimately that's what pathing time is for - it's needed where a train must pass location A no later than a certain time, but must then pass a subsequent location B no earlier than another certain time. You're never going to be able to make a national timetable without pathing time, not unless literally every possible conflicting move is grade-separated!

Of course you can't produce a timetable without it, however every bit of pathing time in the timetable is capacity lost. If pathing time can be added as dwell time at a station with surplus capacity this is far better than added approaching a junction down the line.

 

[pokemonsuper9]

And does TfW Rail have its own platform at Bidston?

No, they share with Merseyrail there.

 

[Class15]

Designing a new timetable is a lot, lot easier than what some have suggested. On the North and West London lines, for example, the only real issue is the SN trains to Watford. The freight trains all have places where the can be put out of the way. NLL and WCML at Primrose Hill aren’t interconnected as the freight can simply wait for a path to become available on the next part of its route.

 

[Peregrine 4903]

Designing a new timetable is a lot, lot easier than what some have suggested. On the North and West London lines, for example, the only real issue is the SN trains to Watford. The freight trains all have places where the can be put out of the way. NLL and WCML at Primrose Hill aren’t interconnected as the freight can simply wait for a path to become available on the next part of its route.

How many timetables have you written and designed?

 

[IrishDave]

I see where you are coming from about Blackfriars Junction, a great example of where you can design a really local timetable, even just Elephant & Castle/London Bridge to St Pancras, and then toggle other wider area timetables to match the paths and therefore making best use of the capacity available through the junction. This could then be hooked onto the timetable in the South east and dictate that timetable somewhat.

There are however junctions which are far more restrictive as a pinch point, that have a much wider impact due to the nature of the services passing through. Consider the service groups passing through Blackfriars Junction, there are many and that gives flexibility in what service group can go into each of the 24 paths available, the only limitation is whether they need a path to/from London Bridge or Elephant and Castle.

At Ely North Junction on the other hand you are far more restricted in what you can do parallel moves wise due to the layout of the junction and double red signalling approaching it means adding pathing time increases capacity used between the station and North Junction more than some other locations, the single lead junctions to both Kings Lynn and Norwich mean there is no parallel moves to/from these branches. Therefore designing a timetable around this will allow the best use of the capacity through the junction and reduce wasted capacity through pathing that could be avoided.

I take your point on the lack of parallel moves at Ely North Junction, but I don't think that's where you would start a timetable, because even though it's restrictive, you're not actually hitting the theoretical maximum capacity through the junction. Designing the timetable around Ely North Junction suggests there's only one possible solution, which there isn't.

Ely North Junction has a junction margin of 3 minutes, so a maximum capacity of 20tph, or 10tph in each direction if you assume no parallel moves - but only 7.5tph (including freight) are planned over the junction (the half is because the Ipswich-Peterborough service is two-hourly). So there will necessarily be more than one way of timetabling Ely North Junction, and one of those solutions can be used to match up with the available capacity elsewhere - e.g. platform capacity at Ely, platform capacity at Cambridge, capacity at Shepreth Branch Jn, over the two-track section of the ECML through Welwyn, and down the WAML into Liverpool Street.

By contrast, the maximum capacity of Blackfriars Junction is 48tph, or 24tph in each direction assuming everything makes parallel moves - so if you want to hit 24tph in each direction, there's very little flexibility in how it's timetabled.

By the way, that's not to say Ely North Junction doesn't need to be upgraded - just because there's more capacity at Ely North Junction in isolation, doesn't mean that the whole system has the capacity to run additional trains over Ely North Junction. Turning the single-lead junctions into double-lead junctions will unlock capacity elsewhere that can't currently be utilised because it doesn't match up with the available capacity at Ely North Junction.

The point is, there are parts of the network that only have one timetable solution for that section - Blackfriars Junction is one of those key constraints, in a way that Ely North Junction just isn't - and those are where a national timetable has to start from.

You can add pathing time, but that eats up capacity, and that could be avoided if the trains just didn't pass near the pinch point.


Of course you can't produce a timetable without it, however every bit of pathing time in the timetable is capacity lost. If pathing time can be added as dwell time at a station with surplus capacity this is far better than added approaching a junction down the line.

Actually, I would disagree that pathing time is lost capacity - in fact, I'd argue that pathing time makes it possible to use capacity that would otherwise be unusable. For example, pathing time can be very effective at increasing capacity when it's used to equalise the speed of trains using a line: by (effectively) slowing down faster trains to travel at the same speed as slower trains, you enable more trains to use a section of line in the same amount of time.

This is perhaps most starkly shown on the North London Line, where freight trains have pathing time added to (effectively) slow them down to travel at the same speed as the stopping passenger trains run by London Overground. If the freight trains needed a clear run, you wouldn't be able to fit as many trains over the line, and your capacity would actually be less; whereas by adding pathing time, you increase the overall number of trains that can use the NLL.

And while I take your point about dwell time being better than pathing time where possible (which is indeed the approach taken for most CrossCountry paths), it isn't always - for NLL freight, you just want the freight trains to keep trundling along, rather than stopping and starting!

== Doublepost prevention - post automatically merged: 3 Oct 2024 ==

The top three would be paths through the Thameslink core, the Castlefield Corridor and Birmingham New Street.

I think it's one of the reasons the East Coast recast has come unstuck, as the new LNER timetable doesn't fit very well with TPE, Thameslink or XC respectively which are fixed by those constraints.

To come all the way back to this, I agree with those three as the top constraints, and if I had to add a fourth, it would be Leeds to Micklefield - one of those unassuming bits of two-track railway that really only has one timetable solution as soon as TPE want to run 4tph fast that are as close to evenly-spaced as possible. Another constraint that won't help the ECML recast...

 

[Class15]

How many timetables have you written and designed?

The point I’m trying to make is that there are some lines that can be “isolated” by using freight loops. So one part of the network can be redone but not redoing others. In my example the NLL could be rewritten but the WCML not changed because you could simply hold the freight in the Primrose Hill line. Granted this is not the case everywhere, but in some places, it’s not quite so bad. Agree though that creating a timetable is really tough.

 

[The Planner]

I take your point on the lack of parallel moves at Ely North Junction, but I don't think that's where you would start a timetable, because even though it's restrictive, you're not actually hitting the theoretical maximum capacity through the junction. Designing the timetable around Ely North Junction suggests there's only one possible solution, which there isn't.

Ely North Junction has a junction margin of 3 minutes, so a maximum capacity of 20tph, or 10tph in each direction if you assume no parallel moves - but only 7.5tph (including freight) are planned over the junction (the half is because the Ipswich-Peterborough service is two-hourly). So there will necessarily be more than one way of timetabling Ely North Junction, and one of those solutions can be used to match up with the available capacity elsewhere - e.g. platform capacity at Ely, platform capacity at Cambridge, capacity at Shepreth Branch Jn, over the two-track section of the ECML through Welwyn, and down the WAML into Liverpool Street.

By contrast, the maximum capacity of Blackfriars Junction is 48tph, or 24tph in each direction assuming everything makes parallel moves - so if you want to hit 24tph in each direction, there's very little flexibility in how it's timetabled.

By the way, that's not to say Ely North Junction doesn't need to be upgraded - just because there's more capacity at Ely North Junction in isolation, doesn't mean that the whole system has the capacity to run additional trains over Ely North Junction. Turning the single-lead junctions into double-lead junctions will unlock capacity elsewhere that can't currently be utilised because it doesn't match up with the available capacity at Ely North Junction.

The point is, there are parts of the network that only have one timetable solution for that section - Blackfriars Junction is one of those key constraints, in a way that Ely North Junction just isn't - and those are where a national timetable has to start from.


Actually, I would disagree that pathing time is lost capacity - in fact, I'd argue that pathing time makes it possible to use capacity that would otherwise be unusable. For example, pathing time can be very effective at increasing capacity when it's used to equalise the speed of trains using a line: by (effectively) slowing down faster trains to travel at the same speed as slower trains, you enable more trains to use a section of line in the same amount of time.

This is perhaps most starkly shown on the North London Line, where freight trains have pathing time added to (effectively) slow them down to travel at the same speed as the stopping passenger trains run by London Overground. If the freight trains needed a clear run, you wouldn't be able to fit as many trains over the line, and your capacity would actually be less; whereas by adding pathing time, you increase the overall number of trains that can use the NLL.

And while I take your point about dwell time being better than pathing time where possible (which is indeed the approach taken for most CrossCountry paths), it isn't always - for NLL freight, you just want the freight trains to keep trundling along, rather than stopping and starting!

== Doublepost prevention - post automatically merged: 3 Oct 2024 ==


To come all the way back to this, I agree with those three as the top constraints, and if I had to add a fourth, it would be Leeds to Micklefield - one of those unassuming bits of two-track railway that really only has one timetable solution as soon as TPE want to run 4tph fast that are as close to evenly-spaced as possible. Another constraint that won't help the ECML recast...

Pathing time keeps trains apart for headway or a junction margin, thats all. There is a strong argument depending on the line that a freight should be infront of a stopper. You never utilise 100% capacity either, so a 3 minute headway, junction margin or reoccupation would never realise 20tph.

 

[IrishDave]

Pathing time keeps trains apart for headway or a junction margin, thats all. There is a strong argument depending on the line that a freight should be infront of a stopper. You never utilise 100% capacity either, so a 3 minute headway, junction margin or reoccupation would never realise 20tph.

But was the Thameslink Core not intended to have 24tph on a 2½-minute planning headway? Admittedly it never actually got there, but that would have been 100% utilisation once it did.

 

[Harpo]

West Ealing to Greenford - 'Connected' or not?

 

[Class15]

West Ealing to Greenford - 'Connected' or not?

There’s a tiny amount of freight using the line, I’m afraid.

 

[The exile]

There is a strong argument depending on the line that a freight should be infront of a stopper.

If the stopper is running at 15 minute intervals or more frequently then “before a stopper” very quickly becomes “after a stopper” even when there’s no disruption.

 

[Mcr Warrior]

West Ealing to Greenford - 'Connected' or not?

First and last trains of the day on the line start from/finish at London Paddington, don't they?

 

[Harpo]

First and last trains of the day on the line start from/finish at London Paddington, don't they?

Aha! I think we need the rules for the radio game 'Mornington Crescent'!!

 

[Kilopylae]

But was the Thameslink Core not intended to have 24tph on a 2½-minute planning headway? Admittedly it never actually got there, but that would have been 100% utilisation once it did.

It's not difficult to argue that that Thameslink timetable was never going to work and therefore serves as a supporting example for why one can't plan to utilise 100% capacity.

 

[Esker-pades]

West Ealing to Greenford - 'Connected' or not?

Connected. Freight.

 

[zwk500]

Is there any at Chester given that Merseyrail uses a bespoke platform?

There's a tiny amount of empty/departmental moves that interact with Merseyrail. I'm not sure if there are any trains that use 7 as a through train between the Merseyrail services.

 

[Wilts Wanderer]

On occasion the Windsor-Slough branch is formed using a unit outstabled in the bay platform because the Main Lines are under possession. In that instance the branch is well and truly segregated.

 

[Sheridan]

There's a tiny amount of empty/departmental moves that interact with Merseyrail. I'm not sure if there are any trains that use 7 as a through train between the Merseyrail services.

There are definitely some TfW booked in 7, I’m sure I’ve been on a Holyhead-Manchester Airport early on a Sunday that went through 7 as booked. I’m guessing when Merseyrail is on a 15-minute frequency it’s not possible (or not worth the risk at least).