(Trivia) Examples of splits and joins

[Route115?]

Does any current SWR service arriving at Salisbury split into three parts that then continue to three different destinations?

I am not aware of any. In the new timetable the 1L60 19:20 Waterloo to Exeter leaves WAT with three units, one is datached at Salisbury and a second at Yeovil Jn, but it only goes to a single destination.

 

[Cricketer8for9]

Eastbourne where the Hastings/Ore trains leave 4 or perhaps sometimes 8 behind on the Eastbourne platform when reversing there.

 

[Hophead]

Southern's Arun Valley services to Portsmouth and Bognor split at Barnham on Sundays, but at Horsham the rest of the week.

 

[yorkie]

Thanks all; this is very helpful.

Should reducing or lengthening the train for capacity reasons at a mid point really be considered the same as splitting/joining in the sense of the original question? It’s not quite the same in terms of journey planning or timetabling.

In theory, it should be an irrelevance.

But in reality, it depends on how it's done; when not done correctly, it can cause issues (such as the Dalmuir example; admittedly this was shortening the train at the end of its journey, which makes it even more strange; see below again for the link, as well as a screenshot of an example bizarre journey produced by the error), so can be useful to know about.

Realtime Trains | 2121 Cumbernauld to Dalmuir | 27/12/2024

Real-time train running information for 2121 departure from Cumbernauld to Dalmuir on 27/12/2024. From Realtime Trains, an independent source of train running info for Great Britain.

www.realtimetrains.co.uk

Image from Trainline and shows a journey from Dalmarnock to Rutherglen, departing at 2216, with a journey time of 1 hour 19 minutes with "stops unavailable".

One of my favourite quirks of the timetable is the last train of the day from Gunnislake which joins a service from Penzance at Plymouth and runs through to Exeter. I think this is the only train on the Tamar Valley line that doesn't start or finish at Plymouth.

Realtime Trains | 2222 Gunnislake to Plymouth | 18/12/2024

Real-time train running information for 2222 departure from Gunnislake to Plymouth on 18/12/2024. From Realtime Trains, an independent source of train running info for Great Britain.

www.realtimetrains.co.uk

Oddly, RTT titles it as "2222 Gunnislake to Plymouth" but then shows a join and lists the stations through to Exeter; is this some sort of data issue? Anyone know what can cause this?

 

[pokemonsuper9]

Oddly, RTT titles it as "2222 Gunnislake to Plymouth" but then shows a join and lists the stations through to Exeter; is this some sort of data issue? Anyone know what can cause this?

If you look in detailed mode that is the 22:22 Gunnislake to Plymouth, if you go the other portion, it names itself the 2100 Penzance to Exeter St Davids, even when simple mode shows both services in one.

Realtime Trains | 2100 Penzance to Exeter St Davids | 18/12/2024

Real-time train running information for 2100 departure from Penzance to Exeter St Davids on 18/12/2024. From Realtime Trains, an independent source of train running info for Great Britain.

www.realtimetrains.co.uk

 

[barringtoncem]

I Don't know if it still happens but didn't the Kings X to Kings Lynn split at Cambridge as you had be in the correct portion of the Train if you needed Ely

 

[Magdalia]

I Don't know if it still happens but didn't the Kings X to Kings Lynn split at Cambridge as you had be in the correct portion of the Train if you needed Ely

Yes, but these did not have two portions with both continuing in traffic after the split. It still happens with 12 car trains.

For a long while the 1814 Kings Cross-Kings Lynn divided at Royston with only the front portion going to Kings Lynn and the rear portion calling all stations to Cambridge. But that does not happen now.

 

[Flange Squeal]

Some SWR services split at Woking for Alton/Basingstoke. Other splits on SWR have already been mentioned.

Just to expand on this for those unfamiliar with the service pattern, the Alton/Basingstoke splits on Sundays are every hour until mid-afternoon, from when the frequency on these routes increases to twice per hour. Towards London Waterloo, the portion from Basingstoke arrives into Woking platform 1 and the portion from Alton arrives a few minutes behind to attach. From London Waterloo, the formation arrives into Woking platform 5 where the Alton portion is split off the rear, with the front portion running to Basingstoke and the rear portion heading to Alton a few minutes behind. On Saturdays, engineering works that reduce capacity closer to London or see them diverted can often see these services attach/detach at Woking on that day as well.

 

[Tw99]

A unit gets removed from the 12 car 1752 Paddington to Didcot at Reading, but it just goes to the depot.

 

[duffield]

So, am I right to take it that we really have no current examples at all of trains that split, follow different routes and then later re-join and continue to the final destination? And that there have not been any for quite a few years now? Obviously, this did happen in the past (I used to work in timetabling in the 1980's and there were definitely some examples then on Network South East as was, and they were a pain in the ****. The timetable software would have a fit if a split/join train ran over midnight!).

So when did this last occur? Anything since 2000?

 

[DunsBus]

So, am I right to take it that we really have no current examples at all of trains that split, follow different routes and then later re-join and continue to the final destination? And that there have not been any for quite a few years now? Obviously, this did happen in the past (I used to work in timetabling in the 1980's and there were definitely some examples then on Network South East as was, and they were a pain in the ****. The timetable software would have a fit if a split/join train ran over midnight!).

So when did this last occur? Anything since 2000?

The Georgemas Junction split (from Inverness) and join (to Inverness) on the Far North Line continued well into the noughties. It ended when the rolling stock used changed from four-car class 156s to single-car class 158s.

 

[Harpo]

Odd TfW services split and combine Cambrian and North Wales portions at Shrewsbury.

Far more common though is the trademark TfW ‘set swap’ perfected on the valley lines at Cardiff and now a way of life at so many other places, presumably because traction is maintained in the middle of it’s routes. In fact, some swaps are temporarily timetabled post-Talerddig.

 

[Bill57p9]

Virgin Trains 1K25 from Holyhead currently joins 1R25 from Wrexham General at Chester (Mondays to Saturdays) before proceeding to London Euston.

On Sundays, 1J32 from London Euston to Wrexham General splits at Chester with 1D95 running on to Holyhead

 

[yorkie]

So, am I right to take it that we really have no current examples at all of trains that split, follow different routes and then later re-join and continue to the final destination? And that there have not been any for quite a few years now? Obviously, this did happen in the past (I used to work in timetabling in the 1980's and there were definitely some examples then on Network South East as was, and they were a pain in the ****. The timetable software would have a fit if a split/join train ran over midnight!).

So when did this last occur? Anything since 2000?

Probably not, though if it does happen I would want to know about it!

it may be worth you creating a historical thread if you are interested in the history of such services, as this thread is for current services only (I am asking for the purpose of testing a journey planner).

Another thing I would be interested in would be a service that joins with another, and then later splits. Do any current examples exist? I suspect some journey planners might not actually be able to cope with that combination.

Even if it's not a regularly timetabled thing, if anyone has any examples that are a one-off, these are useful to know; the links to the specific trains on Realtimetrains are very useful.

Thanks again to all who have contributed.

 

[arb]

I Don't know if it still happens but didn't the Kings X to Kings Lynn split at Cambridge as you had be in the correct portion of the Train if you needed Ely

Yes, but these did not have two portions with both continuing in traffic after the split. It still happens with 12 car trains.

For a year or two before Covid hit there was a northbound train in the evening peak that would arrive at Cambridge at split, with both halves remaining in service. The front part went all the way to King's Lynn (including calling at Ely) and the rear part just went to Ely. I don't recall for sure which train it was, but I'm thinking maybe the 17:35 from Cambridge (so 16:40-ish from King's Cross?), which was always one of the busiest heading north from Cambridge.

I suspect the reasoning had something to do with platform lengths: Ely definitely needed the capacity, but Waterbeach could still only take 4-car trains at the time.

 

[Ashley Hill]

There’s an Exmouth - Paignton 4 car 150 that splits at Exeter SD. One half carries on to Paignton, the other goes to Okehampton.
Wessex had two services that split and rejoined. One was the later version of the Penzance - Manchester which split/joined at Newport ,the 2nd half formed the Cardiff portion. Later the split/join took place at Bristol TM.
The other was a Penzance-Portsmouth service which joined and later split from a Cardiff Portsmouth service at Westbury.

 

[Tom]

Oddly, RTT titles it as "2222 Gunnislake to Plymouth" but then shows a join and lists the stations through to Exeter; is this some sort of data issue? Anyone know what can cause this?

Before the train gets processed out of the scheduling system into the real-time system, RTT will process the associations at a calling point level but won't adjust the origin/destination points of the train. That's due to change at some point in the future but we're working on stuff that is more commercially interesting at the moment.

Once it does the processing, it then would become the service from Penzance & Gunnislake to Exeter St Davids. So for the first running of that schedule tomorrow, that'll start appearing from about 1530 this afternoon.

As an aside, in terms of splits on SWR then plenty on the Weymouth services at Bournemouth (rarely associated, just drops coaches and is reflected in the activity fields), advertised at Southampton. Lots of splits at Salisbury to multiple directions but also drop coaches there and at Yeovil (and on the odd engineering work day, Westbury oddly is also listed sometimes). Sunday services to Basingstoke/Alton at Woking. There used to be something at Guildford on the weekend from what I recall too.

 

[Ashley Hill]

If you look in detailed mode that is the 22:22 Gunnislake to Plymouth, if you go the other portion, it names itself the 2100 Penzance to Exeter St Davids, even when simple mode shows both services in one.

The rear part from Gunnislake is locked out of use prior to departure from Plymouth to Exeter. The train from PZ is diagrammed a 3 car 158 and at 5 coaches is too long for Ivybridge.

 

[Deepgreen]

History now (fairly recent, though), but one of my regular commuter runs used to be a train from Victoria that split at Redhill into three portions - Tonbridge, Gatwick and Reigate. I was unaware of another day working in the UK that did so.

== Doublepost prevention - post automatically merged: 15 Dec 2024 ==

Anything that appears in the data as a split/join of a passenger service is useful to know about!

For example, this one has caused problems with journey planners:

Realtime Trains | 2121 Cumbernauld to Dalmuir | 27/12/2024

Real-time train running information for 2121 departure from Cumbernauld to Dalmuir on 27/12/2024. From Realtime Trains, an independent source of train running info for Great Britain.

www.realtimetrains.co.uk

View attachment 170984
(link and image show a train erroneously "splitting" at Dalmuir; it should be a termination of a passenger service; the stock then devices, with one portion being ECS and the other forming an entirely new service, but it is wrong in the data, which has caused all sorts of problems for journey planners)

It shocks me that errors like this aren't, or can't be, corrected immediately.

 

[Merle Haggard]

History now (fairly recent, though), but one of my regular commuter runs used to be a train from Victoria that split at Redhill into three portions - Tonbridge, Gatwick and Reigate. I was unaware of another day working in the UK that did so.

== Doublepost prevention - post automatically merged: 15 Dec 2024 ==


It shocks me that errors like this aren't, or can't be, corrected immediately.

Again historic - pre Covid; but down 3 x 4car 350s in the evening peak split at Northampton into rear 4 cars back to Euston, middle 4 cars to King's Heath and front 4 cars to New St.
Despite the internal display making clear which set was going where accompanied by set-specific automated announcements, passengers I tried to help avoid going back to Euston thought I was mistaken to the point of contradiction. I did try...

 

[Sonyman66]

Avanti 805007 arrives at Chester with the 12.48 from Holyhead and prepares to couple to 805008 which arrived with the 12.03 from Euston, the formation will return as the 14.32 from Chester to Euston..December 2nd 2024

 

[30907]

Anything that appears in the data as a split/join of a passenger service is useful to know about!

For example, this one has caused problems with journey planners:

Realtime Trains | 2121 Cumbernauld to Dalmuir | 27/12/2024

Real-time train running information for 2121 departure from Cumbernauld to Dalmuir on 27/12/2024. From Realtime Trains, an independent source of train running info for Great Britain.

www.realtimetrains.co.uk

View attachment 170984
(link and image show a train erroneously "splitting" at Dalmuir; it should be a termination of a passenger service; the stock then devices, with one portion being ECS and the other forming an entirely new service, but it is wrong in the data, which has caused all sorts of problems for journey planners)

Out of interest, what problem does it cause, given that the outward and return serve the same stations.

Sorry, ignore that, found your later post.

 

[sharpener]

I Don't know if it still happens but didn't the Kings X to Kings Lynn split at Cambridge as you had be in the correct portion of the Train if you needed Ely

Yes, but these did not have two portions with both continuing in traffic after the split. It still happens with 12 car trains.

AFAIR the split pattern at CBG was often that the front portion went forward to Ely/Lynn and the rear portion immediately formed a new service back to KGX, does that count? With or without a fresh incoming portion attached to the rear.

I don't think it happens like that now but I may be wrong. In the southbound direction there was always much panic on the platform when the doors were shut temporarily to enable the incoming set to couple up, as it is a busy station with a lot of occasional travellers and tourists who do not know the ropes.

It seems that neat operations like this are a thing of the past, now we are told we need to have yet two further platforms and to send trains to Cherry Hinton to reverse. Time was when P1/the crossover/P4 were utilised more intensively - though I could never understand why trains were not routinely signalled through to the far end if it was unoccupied so they did not have to clear the crossover after re-starting.

The ugly barriers now on P1/4 are also a total PITA as well, as they increase the distance you have to walk without reducing the congestion. Progress means deterioration it would seem.

 

[Magdalia]

AFAIR the split pattern at CBG was often that the front portion went forward to Ely/Lynn and the rear portion immediately formed a new service back to KGX, does that count? With or without a fresh incoming portion attached to the rear.

I don't think it happens like that now but I may be wrong. In the southbound direction there was always much panic on the platform when the doors were shut temporarily to enable the incoming set to couple up, as it is a busy station with a lot of occasional travellers and tourists who do not know the ropes.

This way of operating ceased in 2018 when Thameslink started to run through the Canal Tunnel. The timetable change meant that the time gap between the down departure and up arrival at Cambridge disappeared.

It seems that neat operations like this are a thing of the past, now we are told we need to have yet two further platforms and to send trains to Cherry Hinton to reverse.

They are a thing of the past because Cambridge has a lot more trains. The Cherry Hinton turnback is for East West Rail trains and will only happen if East West Rail happens. The current East West Rail proposal, now in consultation, only has one new platform, made possible by having the Cherry Hinton turnback instead.

Time was when P1/the crossover/P4 were utilised more intensively - though I could never understand why trains were not routinely signalled through to the far end if it was unoccupied so they did not have to clear the crossover after re-starting.

The platform 7/8 island reduced the need for this. I've been told that stopping down trains in platform 4 blocks platform 1 too because modern SPAD protection rules about minimum overlap distances mean that another train can't run into platform 1 while platform 4 is occupied.

The ugly barriers now on P1/4 are also a total PITA as well, as they increase the distance you have to walk without reducing the congestion. Progress means deterioration it would seem.

I don't understand this assertion. Trains still stop in the same places and there is no change in the distance that needs to be walked.

I have been using Cambridge station for more than half a century and can't recall ever seeing a train stop over the scissors crossover.

The barrier is there because more people using the station, over 10 million journeys in 2023/24, and that has increased the risk of crowd congestion around the main entrance/exit, which is difficult to adapt for the increased numbers of people passing through.

 

[sharpener]

I don't understand this assertion. Trains still stop in the same places and there is no change in the distance that needs to be walked.

I have been using Cambridge station for more than half a century and can't recall ever seeing a train stop over the scissors crossover.

Thanks for the additional detail @Magdalia. It may just be an impression but it seems to me the trains stop further from the crossover now, maybe connected with the overlap issue. Don't understand why this should be a problem, if points set to cross then a down train signalled for P1 could overrun safely onto the through road if unavoidable.

I was thinking more about the situation where a down is put in P1 so then they had to put the incoming up in P4 and then both trains have to cross before they can leave (seen it quite a few times) whereas if the first to arrive is put at the far end then neither hinders the other and they are free to leave any time.

Yes I have see a train sat on the crossover, it was a long steam-hauled special, came into P4 via the through road and the last carriage comprehensively fouled the crossover. As there were a lot of pax and grockles they couldn't shift it for ages so the station was at a complete standstill with my up train sat outside waiting.

There have been many changes at CBG over the years, the (unchecked) exit from P3 was very useful at busy times for example. Long gone are the days when I would send luggage home PLA to New Malden, I used to be quite good at the paperwork so other boys would ask me to do theirs too, a great wodge of blank PLA forms would just appear at random in the common room at the end of term.

 

[Magdalia]

It may just be an impression but it seems to me the trains stop further from the crossover now, maybe connected with the overlap issue.

Stopping points are very precise these days! Down trains running into platform 1 run right up to the signal which has not moved. The stopping point for down trains in platform 4 possibly moved a bit when it was extended for 230-240m trains.

I was thinking more about the situation where a down is put in P1 so then they had to put the incoming up in P4 and then both trains have to cross before they can leave (seen it quite a few times) whereas if the first to arrive is put at the far end then neither hinders the other and they are free to leave any time.

That rarely happens these days because most up through trains use platforms 7 and 8, with most down through trains using platforms 1 and 4.

Yes I have see a train sat on the crossover, it was a long steam-hauled special, came into P4 via the through road and the last carriage comprehensively fouled the crossover. As there were a lot of pax and grockles they couldn't shift it for ages so the station was at a complete standstill with my up train sat outside waiting.

I doubt that would be allowed now.

 

[arb]

I've been told that stopping down trains in platform 4 blocks platform 1 too because modern SPAD protection rules about minimum overlap distances mean that another train can't run into platform 1 while platform 4 is occupied.

Last Friday evening (13th), the 18:30 departure from Cambridge (to Norwich) arrived into platform 4, and was followed by the late-running 18:19 to Ely stopping in platform 1, before the Norwich train had left. (And off-topic, the departure screens were insistent that the 18:19 was going to depart first, but it didn't. Grumble).

 

[Class15]

I did not know Realtime trains does that (I always use detailed mode), it looks fun on the highland sleeper splitting into 3 (the Lowland isn't anywhere near as interesting).

The Lowland is interesting as the two portions go in opposite directions from Carstairs, which I’ve always thought is quite cool.

 

[NaZzAtAzEr]

History now (fairly recent, though), but one of my regular commuter runs used to be a train from Victoria that split at Redhill into three portions - Tonbridge, Gatwick and Reigate. I was unaware of another day working in the UK that did so.

== Doublepost prevention - post automatically merged: 15 Dec 2024 ==


It shocks me that errors like this aren't, or can't be, corrected immediately.

Reigate and gatwick airport still split and join at the peaks at Redhill

 

[Big Jumby 74]

I am not aware of any. In the new timetable the 1L60 19:20 Waterloo to Exeter leaves WAT with three units, one is datached at Salisbury and a second at Yeovil Jn, but it only goes to a single destination.

I can not comment on the current situation, only from a historical angle, which has been touched on up thread. Three way splits were not common by any means (in the base plan), two that were (SX) both involved the rear/third unit being taken out of service at Salisbury, so it wasn't a case of three passenger services being formed from one train. Those concerned, the 0920 from Waterloo, for Exeter & Bristol, the third unit went to depot at Salisbury for a short time (see below), the unit on this latter diagram was that which was crucial to the maintenance exam process each night.
The second three way split being the 1220 from Waterloo, again for Exeter, Bristol and Salisbury. The third unit out of service again, and shunted to platform 6 for the rear of the 1447 to Waterloo, the front portion (1447), co-incidentally being the aforementioned unit that was on the exam diagram (see above).

Three way passenger splits were not something that SW entertained as a rule (base plan wise) during my time, and simultaneous splitting and joining involving the same (passenger) train were an absolute no-no! I'm sure it did happen on odd occasions of disruption perhaps, as Control would do what needed to be done to resume normal service, but was never planned that way to my knowledge.