Delaying one train further to allow one to depart of time or delaying both?

[infobleep]

For various delay reasons I ended up waiting for the 18.02 to Haywards Heath from East Croydon. It eventually goes to Littlehampton.

Whilst waiting for it they started announcing it 2 minutes before it was due to leave. However I knew it was late. Soon they put up it was due 18.05. Then suddenly the board switched to showing the 18.06 to Bognor Regis and this due in after it at 18.06! The Bognor train arrived in on time and departed on time.

I can understand why they did this because that train is non stop to Three Bridges and the Littlehampton train stops at Horley and Gatwick Airport and then Three Bridges.

This however caused the Littlehampton train to depart 7 minutes late. It had only been 3 minutes late leaving Clapham Junction. So it added 3-4 minutes delay. It invariably lost more time on route.

It got me wondering. What is better, to delay both trains or hold the Littlehampton train for longer, so only one train is delayed.

If the Bognor train leaves late it could cause issues around Barnham. However if the Littlehampton train is delayed by 7-9 minutes by the time it reaches Hove, it may cause or get further delayed on the stretch from Hove. However it it may already be out of slot, were it to be 3-5 minutes late around Hove.

 

[lincolnshire]

Been on a train into Kings Cross where we ended up been 15 mins late, the other trains passed us and arrived on time, sometimes its done because if like our train ended up 15 mins late and the rest on time then you only have to pay delay money for one train instead of the knock on effect of having to pay out more money for more trains with only slight delays.

Its all down to money payout for delays.

 

[gimmea50anyday]

Happens to 1E63 07:08 NCL-LIV regularly. If 1E01 the up scotsman is late at Newcastle 1E63 gets looped at Tyne Yard or Ferryhill to let it pass, but this then causes a 10 min or more delay to 1F63. This then delays 1P19 at Northallerton and 1A60 at Thirsk who both have to await 1F63's passing. On arrival at York the punters transferring off 1F63 onto either 1A60 or 1P19 risk missing their connections due to the lack of connection time which 1E01 has ate up in its delay. This is significant as 1F63/1A60 is the earliest off-peak arrival into London from Newcastle (1E01 is full fare) as is 1P19 the earliest off-peak airport connection.

And the gain to 1E01 for all this delay and disruption? With no connections en route or station stops just one single solitary minute into the cross....

 

[Welshman]

It also happens occasionally at Chester if the 1710 or 1810 VT Euston-Holyheads are late. The 1932 or 2034 ATW stopping services are allowed down the coast in their proper slots, with the Eustons behind.

As the ATWs stop at more stations on the coast, and there is no chance for the faster VTs to overtake, they trundle behind, running from signal to signal to Holywell, where they usually come to a halt due to the long section from there to Rhyl.

So what was a 10-minute late arrival at Chester can turn into a 30-minute late arrival at Holyhead

 

[Antman]

Been on a train into Kings Cross where we ended up been 15 mins late, the other trains passed us and arrived on time, sometimes its done because if like our train ended up 15 mins late and the rest on time then you only have to pay delay money for one train instead of the knock on effect of having to pay out more money for more trains with only slight delays.

Its all down to money payout for delays.

I think maybe you are being a little bit cynical?

I remember a signalman telling me that generally it was better to further delay a late running train rather than delay a train that is on time although there are sometimes other factors to be considered.

 

[gimmea50anyday]

Being a stopper service, i would have thought there would have been people on that VT who wanted that connection, but then where do you draw that line between holding the connection for the late runner, and letting it go thereby adding to the delay?

After operation princess departures from Brum were xx.00 euston, xx.03 bournemouth and xx.06 cov stopper. if coming from the east coast route if the train was just a few minutes late the stopper would be let out first, or off the west coast again just a few minurptes delay at WVH the stopper would go first, the XC was never allowed to pass the stopper either at new street or international. This then caused the XC to then be +30 late at coventry and was repeated often. In this case I cannot see any legit reason why the XC wasnt given a passing opportunity when they were available. The delay then snowballd through Oxford, on the GWML being rescinded to the slows, then again T Basingstoke instead of being behind the weymouth it would be stuck behind the bourmenouth stopper. Often a turnround at Brokenhurst or Southampton would then be required just to get the set back on diagram right time.

Or again at xx.33 reading shuttle and xx.36 cov stopper, again with this one at Wolves if the XC was more than a few minutes late the stopper would leave first leaving the reading trailing behind, thereby causing the delay into new street and therefore missing its slot on the chiltern route at Tysley. Again this snowballs at Oxford resulting in an unscheduled Didcot terminator stop to get back on schedule. This repeatedly happened amid much frustration to staff and punters alike.

 

[westv]

This used to happen quite often when I used to get the 17:34 from Liverpool St to Wickford. The previous 17:32 to Chelmsford and beyond went on time and then we'd wait... and wait....and wait.... and wait until the late running 17:30 to Norwich left.
Now happens on the service Kings X to Hull. On several occasions we have arrived at Doncaster on time (or sometimes slightly early) but then we have to wait...and wait.. and wait for the Newcastle (?) non stopper to pass so ending up leaving Doncaster late.

 

[Sunset route]

For various delay reasons I ended up waiting for the 18.02 to Haywards Heath from East Croydon. It eventually goes to Littlehampton.

Whilst waiting for it they started announcing it 2 minutes before it was due to leave. However I knew it was late. Soon they put up it was due 18.05. Then suddenly the board switched to showing the 18.06 to Bognor Regis and this due in after it at 18.06! The Bognor train arrived in on time and departed on time.

I can understand why they did this because that train is non stop to Three Bridges and the Littlehampton train stops at Horley and Gatwick Airport and then Three Bridges.

This however caused the Littlehampton train to depart 7 minutes late. It had only been 3 minutes late leaving Clapham Junction. So it added 3-4 minutes delay. It invariably lost more time on route.

It got me wondering. What is better, to delay both trains or hold the Littlehampton train for longer, so only one train is delayed.

If the Bognor train leaves late it could cause issues around Barnham. However if the Littlehampton train is delayed by 7-9 minutes by the time it reaches Hove, it may cause or get further delayed on the stretch from Hove. However it it may already be out of slot, were it to be 3-5 minutes late around Hove.

It's all down to individual judgment calls, taking into account PPM and stopping patterns if known unless control advise anything different. There is no tech available to assist the signaller controlling Windmill Bridge/East/South Croydon and with the other approach junctions (Norward/Norward Folk) controlled by one signaller and (Selhurst) controlled by another. All three signallers need to know how each other works when passing trains through Selhurst triangle for a smooth passage.

 

[Deepgreen]

It's all down to individual judgment calls, taking into account PPM and stopping patterns if known unless control advise anything different. There is no tech available to assist the signaller controlling Windmill Bridge/East/South Croydon and with the other approach junctions (Norward/Norward Folk) controlled by one signaller and (Selhurst) controlled by another. All three signallers need to know how each other works when passing trains through Selhurst triangle for a smooth passage.

Strangely enough, I was pondering this very topic and location again this morning, as my train (the 0850 from East Croydon to Victoria, 0752 ex-Horsham), running on the up Vic. fast line, is usually (as it was this morning) held to let the parallel ex-East Grinstead working on the up Vic. slow line cross to the fast ahead of us at Selhurst. We both run fast to Clapham Junction and both are usually a couple of minutes late at this point. However, my train, being on the fast line already, would be able to clear the junction more quickly if allowed a clear road, whereas the East Grinstead one has to slow for the crossover anyway before re-accelerating and would seem logically to best be the second train here when both are running parallel.

 

[Sunset route]

Strangely enough, I was pondering this very topic and location again this morning, as my train (the 0850 from East Croydon to Victoria, 0752 ex-Horsham), running on the up Vic. fast line, is usually (as it was this morning) held to let the parallel ex-East Grinstead working on the up Vic. slow line cross to the fast ahead of us at Selhurst. We both run fast to Clapham Junction and both are usually a couple of minutes late at this point. However, my train, being on the fast line already, would be able to clear the junction more quickly if allowed a clear road, whereas the East Grinstead one has to slow for the crossover anyway before re-accelerating and would seem logically to best be the second train here when both a running parallel.

Without being there this morning I couldn't say, but if the up East Grinstead was formed of anything more than an 8car then it would block Cottage Junction if there was a following London Bridge service just to keep your train moving. All of the slow line junctions block each other these days with the longer trains we run.

 

[455driver]

What was the return working for the delayed train?
Would it still be able to leave on time if it was delayed further on its current trip?

What was the return working for the on time train(s) that would be delayed if the one already delayed was given a run?
Would they still be able to leave on time if they were delayed on their current trips?

Same questions apply to the train crews about their return workings, (legally required) PNBs etc!

It isnt a simple question about one or two trains!

 

[Bodiddly]

More often than not (in my experience) the 18.11 Glasgow QS-Inverness is late getting in to meet the 21.33 to Aberdeen. The Aberdeen is very rarely kept waiting and on one occasion I have passed it at Welsh's Bridge just outside the station. It is very frustrating! To give them their due though, Scotrail always lay a taxi on to stations on the Aberdeen line.
A quick check on Recent Train Times shows in the last 100 weekdays, 60% of trains on this service were over 10 minutes late! (there is a 7 minute connection between services)

 

[Deepgreen]

Without being there this morning I couldn't say, but if the up East Grinstead was formed of anything more than an 8car then it would block Cottage Junction if there was a following London Bridge service just to keep your train moving. All of the slow line junctions block each other these days with the longer trains we run.

My train was a five car, and the East Grinstead one an eight car. These are the booked formations for these two trains. I am not raising a significant issue here - it was just that the previous mention of this location jogged my mind back to this almost daily occurence.

 

[bramling]

For various delay reasons I ended up waiting for the 18.02 to Haywards Heath from East Croydon. It eventually goes to Littlehampton.

Whilst waiting for it they started announcing it 2 minutes before it was due to leave. However I knew it was late. Soon they put up it was due 18.05. Then suddenly the board switched to showing the 18.06 to Bognor Regis and this due in after it at 18.06! The Bognor train arrived in on time and departed on time.

I can understand why they did this because that train is non stop to Three Bridges and the Littlehampton train stops at Horley and Gatwick Airport and then Three Bridges.

This however caused the Littlehampton train to depart 7 minutes late. It had only been 3 minutes late leaving Clapham Junction. So it added 3-4 minutes delay. It invariably lost more time on route.

It got me wondering. What is better, to delay both trains or hold the Littlehampton train for longer, so only one train is delayed.

If the Bognor train leaves late it could cause issues around Barnham. However if the Littlehampton train is delayed by 7-9 minutes by the time it reaches Hove, it may cause or get further delayed on the stretch from Hove. However it it may already be out of slot, were it to be 3-5 minutes late around Hove.

At the end of the day it's a judgement call for the signaller, based on a potentially very long list of possible factors.

Some examples of the sorts of things to be considered;

- Will one or other of the trains have an adverse effect on others?
- Length of turnround times at end of journey before next working. If one has a tight turnround then don't delay that one.
- What do the drivers do next and when? If one of the drivers has a tight PNB then not a good idea to delay that one.
- What do the trains do next? If one of them goes down a depot then that may be the one to delay, if the driver has sufficient time.
- Platform availability at the terminus?
-What are the chances of either service making up the time?
-Can the time be made up another way?
- Do other of the trains need to join in a particular order to another?

And many more things.

The worst thing to do is just to consider PPM, as what happens is if one trains runs 5 minutes late it tends to go to the bottom of the pile. I've been on trains which have turned up just a few minutes late, then at every conceivable conflict and junction point other stuff has been given priority, turning a 5-min delay into 20 late or more.

 

[LNW-GW Joint]

NR's Train Management System running in the ROCs is supposed to solve all this.
Roger Ford has a piece on it in the March Modern Railways.
NR is planning a trial, integrated with ETCS and C-DAS, between Norwich-Lowestoft/Great Yarmouth.
(C-DAS is the Connected Driver Advisory System, the driver's end of all this).

Until this nirvana, I thought NR was supposed to manage the network to "minimise overall delay".

 

[exile]

Apparently at one time in the USA a late train was to be given priority - until it was 24 hours late when it was to be regarded as the following day's train.....

 

[455driver]

At the end of the day it's a judgement call for the signaller

TOC control can also request a certain order of priority.

 

[infobleep]

At the end of the day it's a judgement call for the signaller, based on a potentially very long list of possible factors.

Some examples of the sorts of things to be considered;

- Will one or other of the trains have an adverse effect on others?
- Length of turnround times at end of journey before next working. If one has a tight turnround then don't delay that one.
- What do the drivers do next and when? If one of the drivers has a tight PNB then not a good idea to delay that one.
- What do the trains do next? If one of them goes down a depot then that may be the one to delay, if the driver has sufficient time.
- Platform availability at the terminus?
-What are the chances of either service making up the time?
-Can the time be made up another way?
- Do other of the trains need to join in a particular order to another?

And many more things.

The worst thing to do is just to consider PPM, as what happens is if one trains runs 5 minutes late it tends to go to the bottom of the pile. I've been on trains which have turned up just a few minutes late, then at every conceivable conflict and junction point other stuff has been given priority, turning a 5-min delay into 20 late or more.

Thanks for your reply. How does the signalman remember all of this or is this information available on a screen for him to look at?

I find all of this fascinating.
--- old post above --- --- new post below ---

What was the return working for the delayed train?
Would it still be able to leave on time if it was delayed further on its current trip?

What was the return working for the on time train(s) that would be delayed if the one already delayed was given a run?
Would they still be able to leave on time if they were delayed on their current trips?

Same questions apply to the train crews about their return workings, (legally required) PNBs etc!

It isnt a simple question about one or two trains!

I'm sure it's not. It fascinates me because it's so complicated. I couldn't be a signaller because I don't have the right mind for it but I think I'd find the job interesting. I could be wrong through as I've never been inside a working signal box or centre.

 

[Sunset route]

At some point it becomes 1st come 1st serve when you have to combine the two London routes in both directions, as holding for correct PPM regulation don't work when your junction blocking.
--- old post above --- --- new post below ---

Thanks for your reply. How does the signalman remember all of this or is this information available on a screen for him to look at?

Your be surprised how little information we have to work with, paper simplifies with hand written admendments and your own memory for what works and doesn't work and this is at locations where your pushing 1200 and 1400 train movements per 24hr period. Length of trains, train, drivers and guards diagrams are not usually known apart from terminal working where you have stock or platform workings. But waiting in the wings is TMS isolated so the information should radically improve in the not to distant future.

 

[infobleep]

At some point it becomes 1st come 1st serve when you have to combine the two London routes in both directions, as holding for correct PPM regulation don't work when your junction blocking.
--- old post above --- --- new post below ---


Your be surprised how little information we have to work with, paper simplifies with hand written admendments and your own memory for what works and doesn't work and this is at locations where your pushing 1200 and 1400 train movements per 24hr period. Length of trains, train, drivers and guards diagrams are not usually known apart from terminal working where you have stock or platform workings. But waiting in the wings is TMS isolated so the information should radically improve in the not to distant future.

Fascinating. Just think 50 years ago it was probably even more primitive with regards information to go on.

 

[WessexEclectic]

But waiting in the wings is TMS isolated so the information should radically improve in the not to distant future.

Has this system been trialed yet or is this a roll out? Also am I correct in saying that this system had been developed elsewhere?

TIA

 

[Sunset route]

Has this system been trialed yet or is this a roll out? Also am I correct in saying that this system had been developed elsewhere?

TIA

This is the Hitachi system as part of the Thameslink core contract, what it looks like with our panels, God only knows?

 

[gimmea50anyday]

In my experience, over reliance on ARS and not actually watching whats going and regulating accordingly is the problem.

In the big green NX panels of old it was easy to see what was going on. Moving that info to a 14" VDU puts a lot of info into a small space which then also makes that info smaller and harder to see....

 

[carriageline]

I think maybe you are being a little bit cynical?



I remember a signalman telling me that generally it was better to further delay a late running train rather than delay a train that is on time although there are sometimes other factors to be considered.

Depends. If the already late train is less than 5 late, thus making PPM then generally it's better to keep it on time, or try make it improve time, all being equal. Of course stopping pattern can change that.

Same also can be said if you know a train that is over 5 late can be made less than 5 late upon terminating, give it preference.

But the OPs post sounds right to me. Running the stopper Infront of the fast could of made the fast more than 5 late too, all that stopping and starting!
--- old post above --- --- new post below ---

In my experience, over reliance on ARS and not actually watching whats going and regulating accordingly is the problem.

In the big green NX panels of old it was easy to see what was going on. Moving that info to a 14" VDU puts a lot of info into a small space which then also makes that info smaller and harder to see....

Couldn't be further from the truth, and quite frankly I don't think you know what your talking about
--- old post above --- --- new post below ---

Thanks for your reply. How does the signalman remember all of this or is this information available on a screen for him to look at?

Quite frankly, like my learned colleague said, it's years of experience and knowledge passed down. More often than not, it's finding out what works well, and at other times being caught out and promising your self you won't do that again! We have a screen that shows how late a train is/where's it going/where it's come from, and that's about it. Although we can interrogate it to give us more detail if need be.

And then you have to factor in we often only have 20-30 seconds to make these decisions, with (pretty much) only a rough idea of stopping pattern, and how late It is to go by. Sometimes we do make mistakes, unfortunately.

 

[infobleep]

Depends. If the already late train is less than 5 late, thus making PPM then generally it's better to keep it on time, or try make it improve time, all being equal. Of course stopping pattern can change that.

Same also can be said if you know a train that is over 5 late can be made less than 5 late upon terminating, give it preference.

But the OPs post sounds right to me. Running the stopper Infront of the fast could of made the fast more than 5 late too, all that stopping and starting!
--- old post above --- --- new post below ---



Couldn't be further from the truth, and quite frankly I don't think you know what your talking about[emoji38]--- old post above --- --- new post below ---



Quite frankly, like my learned colleague said, it's years of experience and knowledge passed down. More often than not, it's finding out what works well, and at other times being caught out and promising your self you won't do that again! We have a screen that shows how late a train is/where's it going/where it's come from, and that's about it. Although we can interrogate it to give us more detail if need be.

And then you have to factor in we often only have 20-30 seconds to make these decisions, with (pretty much) only a rough idea of stopping pattern, and how late It is to go by. Sometimes we do make mistakes, unfortunately.

When you make a mistake, is it just accepted your human or because your a signalmen dealing with thousands passengers, do they expect you to be nearly always on the ball so to speak. It does seem quiet pressured but perhaps it actually isn't like that at all.

 

[carriageline]

When you make a mistake, is it just accepted your human or because your a signalmen dealing with thousands passengers, do they expect you to be nearly always on the ball so to speak. It does seem quiet pressured but perhaps it actually isn't like that at all.

Depends on the nature of the delays really, how often you make them etc. Your not really hounded for the odd couple of minutes here, but if it becomes a regular thing you will get pulled up on it

 

[D1009]

I'm sure it's not. It fascinates me because it's so complicated. I couldn't be a signaller because I don't have the right mind for it but I think I'd find the job interesting. I could be wrong through as I've never been inside a working signal box or centre.

If you're that interested, Simsig will teach you a lot about it.

 

[bramling]

When you make a mistake, is it just accepted your human or because your a signalmen dealing with thousands passengers, do they expect you to be nearly always on the ball so to speak. It does seem quiet pressured but perhaps it actually isn't like that at all.

At the end of the day, the most important thing is everyone gets home safely as a result of your shift. That is the no.1 priority, it's always worth emphasising that signallers make safety-critical decisions as well as regulating, and the former is always the priority. Authorising trains past failed signals, protection etc always takes precedence.

As a signaller you are never going to get it right all the time, even the right decision one moment can suddenly become the wrong decision when something unexpected happens. Obviously someone repeatedly flagging up would receive appropriate support. We often criticise signallers from the point of view of being on the ground, however before doing so it's always worth remembering you don't know what else could be going on in the box.

It certainly can be pressurised, but it's one of those things where experience comes in. What is massively daunting on day one will become second-nature over time.

 

[carriageline]

At the end of the day, the most important thing is everyone gets home safely as a result of your shift. That is the no.1 priority, it's always worth emphasising that signallers make safety-critical decisions as well as regulating, and the former is always the priority. Authorising trains past failed signals, protection etc always takes precedence.



As a signaller you are never going to get it right all the time, even the right decision one moment can suddenly become the wrong decision when something unexpected happens. Obviously someone repeatedly flagging up would receive appropriate support. We often criticise signallers from the point of view of being on the ground, however before doing so it's always worth remembering you don't know what else could be going on in the box.

Your last point is bang on the money, thanks for realising that! I don't think people realise how often we can have the phone glued to our ear, normally filling out some sort of paperwork that's going to stop trains going into a group of our lads, and at the same time trying to run trains.

It's why I sometimes get frustrated with drivers sending SG when they see a double yellow they know they don't need, and I'm busy dealing with a fault or failure and I've forgotten about them.

 

[Boysteve]

I do not normally complain about signalling decisions as a passenger. Usually I blame the lack of information. The only time I have criticised a signaller was many years ago at Crewe. It was probably 2001 or 2002 and I was on a marginally later Sunday evening VT HST from Euston to Manchester. The journey had been slow using slow lines though Buckinghamshire and then via Northampton. At one point I thought we were going to be very late but our trusty HST stormed though the Trent Valley and although Crewe was not a booked station stop we approached Basford Hall with over 20 minutes to go before scheduled arrival in Stockport. Also the Crewe to Manchester stopper was due to depart in about three minutes time. Imagine my surprise when we stopped at the junction to let the stopper depart ahead of us and we then crawled through to Stockport. Whilst stationary no Southbound train came through the Up Fast. Arrival into Stockport and Piccadilly was over 20 minutes late. I maintain this was a huge balls up.

Quite often I get frustrated with the Manchester to Piccadilly trains but I blame the lack of information. Try catching a 10 minute late XX:06 ex-Cleethorpes to the airport. Does this go before the XX:16 stopper? Does the stopper go first? If the stopper goes first is it held at Slade Lane to the TPEx take the Airport branch first? In my experience all three chances are exactly 33.33333% likely. You ask the platform staff and some say they do not know which will get to the airport first, the others lie. Simple guidance for airline passengers would certainly help but I will not hold my breath with such a complex congested section of track involved!