Train Driving Techniques and Instructions

[dk1]

Moderator note - split from:

Strikes - Who's driving and staffing?

Recently during the strikes I've had lots of time to stand and watch. I've noted that trains pulling into stations (Especially BHM) are a lot slower, and sometimes marshalled into stations with platform staff using hand signals towards where stopping points. At Wolverhampton trains are...

www.railforums.co.uk

I quite agree. But it is situation and location specific, and doesn’t need a 10mph ‘restriction‘ at every red signal.

During the RHTT season all red signals have an instruction to travel at no more than 10mph at the warning magnet as opposed to 20mph at other normal times of the year. This instruction has been in place for the best part of 20 years at my TOC.

 

[ComUtoR]

I’m sure we all understand the need to drive defensively but....

In general I'm in agreement. However, when dealing with this from a performance perspective, you still need to look at why the signal was red. Are you being held for a passing train which is running late ? Was the Signal held due to a fault ? Did the driver need to acknowledge a broadcast, was the Signaler asleep ? Etc etc.

Not forgetting the there are restrictive before the red too. This will also will performance. Those times when you get checked down but the signal is off before you get here etc will still destroy sectional timings.

The speed at the magnet and the defensive driving policy is a tiny tiny speck in performance management.

 

[Bald Rick]

SWR

During the RHTT season all red signals have an instruction to travel at no more than 10mph at the warning magnet as opposed to 20mph at other normal times of the year. This instruction has been in place for the best part of 20 years at my TOC.

Thanks both. I’m doing some work on why actual running t8mes vary and was not aware of this, so that’s helpful.

By increasing the time it takes to reach the appropriate stopping point you also start to increase the risk of the driver getting distracted and spadding at low speed.

You also increase the risk of the train in rear seeing restrictive aspects, and on busy railways you get a domino effect. I was delayed this morning by a train 8 miles in front of me that was slow away from a station, with 2 trains between us.

The speed at the magnet and the defensive driving policy is a tiny tiny speck in performance management.

I‘d argue it’s more than tiny, but agreed a small part.

 

[jamesst]

Thanks both. I’m doing some work on why actual running t8mes vary and was not aware of this, so that’s helpful.




You also increase the risk of the train in rear seeing restrictive aspects, and on busy railways you get a domino effect. I was delayed this morning by a train 8 miles in front of me that was slow away from a station, with 2 trains between us.





I‘d argue it’s more than tiny, but agreed a small part.

15mph at my toc 200yards back from the signal regardless of the time of the year. We're actually told not always to assume the aws magnet is the correct distance.

 

[Bald Rick]

15mph at my toc 200yards back from the signal regardless of the time of the year. We're actually told not always to assume the aws magnet is the correct distance.

And thus is the frustration - ie different policies at different companies (accepting there might be some rolling stock variation).

I once had a situation where three different TOCs had three quite different driving policies for approaching the same buffer stops with the same rolling stock.…

 

[43066]

once had a situation where three different TOCs had three quite different driving policies for approaching the same buffer stops with the same rolling stock.…

Out of interest where was that?

 

[dctraindriver]

While rule book states 20 at the magnet the PDP states 15. I’ll stick to the PDP. It’s annoying at times however I feel there’s enough padding to run to time.

 

[TurboMan]

15mph at my toc 200yards back from the signal regardless of the time of the year. We're actually told not always to assume the aws magnet is the correct distance.

Similar to GWR, which is 20/20 (20mph at 200m from the signal, stopping 20m from the signal) or 10/20 in low adhesion (10mph 200m from the signal, stopping 20m from the signal). So no reference to AWS magnets.

It's not so much that the magnet can be at an incorrect distance, it's more that there is an assumption that all magnets are 200yds/183m from the signal when the current standard specifies that the magnet is placed 'at a distance that equates to at least three seconds train running time, calculated at the maximum permissible speed applicable at the AWS track equipment' (see 3.1.7 of RIS-0775-CCS). That means that where linespeed is 30mph (for example), the AWS magnet can be placed 40m from the signal (assuming I've got my distance/speed/time triangle correct!).

 

[baz962]

I'm trying to get my head around this fascination. 20/15/10 at the magnet. The difference between those speeds are not going to impact the service. If you come up on the red quicker, you just sit there longer. And surely then , the difference between 5/10 mph won't make much of a difference. I replied to someone who thought a particular line was 110 mph on one thread that there was actually 120 mph for around 3 miles and someone else replied saying that would save seconds. So I'm not sure that going 10 mph instead of 20 mph for a couple hundred ish metres will impact any service. All that said I'm going to protect my license before I care about a few seconds anyway.

 

[Dieseldriver]

I'm trying to get my head around this fascination. 20/15/10 at the magnet. The difference between those speeds are not going to impact the service. If you come up on the red quicker, you just sit there longer. And surely then , the difference between 5/10 mph won't make much of a difference. I replied to someone who thought a particular line was 110 mph on one thread that there was actually 120 mph for around 3 miles and someone else replied saying that would save seconds. So I'm not sure that going 10 mph instead of 20 mph for a couple hundred ish metres will impact any service. All that said I'm going to protect my license before I care about a few seconds anyway.

Quite.
Personally, I’ve always thought 15 or 20 at the magnet in a passenger train is a sensible maximum speed to be travelling at whether it’s part of the driving policy or not.
I do wonder when Drivers moan about it what speed they want to approach a red signal at? 40 over the magnet in full service?
As for the lower speeds mandated in some TOCs policies for leaf fall, as a Driver you can utilise that to be your best friend if you conform to the policy and then slide through the red at slow speed.

 

[Haywain]

I'm trying to get my head around this fascination. 20/15/10 at the magnet. The difference between those speeds are not going to impact the service. If you come up on the red quicker, you just sit there longer. And surely then , the difference between 5/10 mph won't make much of a difference. I replied to someone who thought a particular line was 110 mph on one thread that there was actually 120 mph for around 3 miles and someone else replied saying that would save seconds. So I'm not sure that going 10 mph instead of 20 mph for a couple hundred ish metres will impact any service. All that said I'm going to protect my license before I care about a few seconds anyway.

It may only be a few seconds, but each few seconds adds up and can become minutes.

 

[baz962]

It may only be a few seconds, but each few seconds adds up and can become minutes.

No because it's rare that you will encounter loads of reds and you still have to stop anyway. You just sit there longer. The only time it would make any difference is if you encountered dozens of red signals and they all had poor line of sight from the magnet.

 

[12LDA28C]

It may only be a few seconds, but each few seconds adds up and can become minutes.

Irrelevant when approaching a signal at danger. Safety is the number one concern on the railway, not punctuality.

 

[Haywain]

No because it's rare that you will encounter loads of reds and you still have to stop anyway. You just sit there longer. The only time it would make any difference is if you encountered dozens of red signals and they all had poor line of sight from the magnet.

I'm not referring purely to signals, it's just one of many contributory factors to timekeeping.

Irrelevant when approaching a signal at danger. Safety is the number one concern on the railway, not punctuality.

I don't believe I suggested otherwise.

 

[Bald Rick]

I'm trying to get my head around this fascination. 20/15/10 at the magnet. The difference between those speeds are not going to impact the service. If you come up on the red quicker, you just sit there longer. And surely then , the difference between 5/10 mph won't make much of a difference. I replied to someone who thought a particular line was 110 mph on one thread that there was actually 120 mph for around 3 miles and someone else replied saying that would save seconds. So I'm not sure that going 10 mph instead of 20 mph for a couple hundred ish metres will impact any service. All that said I'm going to protect my license before I care about a few seconds anyway.

I’m afraid this is not correct. It makes a big difference. I’ve seen very detailed modelling on this, based on actual data, and the difference is stark.

in simple mechanics terms, the difference between 10 mph and 20mph over 300 metres is half a minute. Whilst it might seem to be only half a minute to you, such additional delay can quickly become a couple of hundred delay minutes across a complex network, particularly on approach to busy junctions and termini.

 

[the sniper]

in simple mechanics terms, the difference between 10 mph and 20mph over 300 metres is half a minute. Whilst it might seem to be only half a minute to you, such additional delay can quickly become a couple of hundred delay minutes across a complex network, particularly on approach to busy junctions and termini.

Magnets are theoretically normally only around 183m off the signals and if you're still doing 10mph at the signal you'll waste more than half a minute filling out an RT3185...

 

[Bald Rick]

Magnets are theoretically normally only around 183m off the signals and if you're still doing 10mph at the signal you'll waste more than half a minute filling out an RT3185...

there is that yes!

 

[Egg Centric]

Well, what is the proposal for changing this then?

We have two "sets" of organisations with completely different incentives.

• TOCs do not want their drivers to be risking SPADding
• Network Rail (it appears) sound like they're modelling, expecting, and wanting the drivers to be driving a bit more aggressively to prevent delay.

Not joined up, and don't see how it can be with the current structure...

I can't express any view on which is "better" (although my general view is railway is far too risk averse as can be seen from the tiny number of fatalities) but this needs to be synchronised one way or the other, no?

 

[Bald Rick]

Well, what is the proposal for changing this then?

We have two "sets" of organisations with completely different incentives.

• TOCs do not want their drivers to be risking SPADding
• Network Rail (it appears) sound like they're modelling, expecting, and wanting the drivers to be driving a bit more aggressively to prevent delay.

Not joined up, and don't see how it can be with the current structure...

I can't express any view on which is "better" (although my general view is railway is far too risk averse as can be seen from the tiny number of fatalities) but this needs to be synchronised one way or the other, no?

I think the point is that the timetable is constructed on a set of assumptions about driving technique, which then seem to be changed by different operators, without then playing that back in to the timetable. And then we all wonder why performance gets worse.

But the annoying thing is where different operators change their driving policies in different ways to deal with the same issue. This is ’preferential engineering’ of operational practice. The laws of physics do not change depending on which TOC is running a train!

 

[Magdalia]

in simple mechanics terms, the difference between 10 mph and 20mph over 300 metres is half a minute.

I think in quarter miles of 440 yards/400 metres both from railway mileposts and laps of an athletics track. These times for quarter miles are ingrained in my mind:

10 mph 90s
15 mph 60s
20 mph 45s
30 mph 30s
45 mph 20s
60 mph 15s
75 mph 12s
90 mph 10s
100 mph 9s

In my extensive experience of commuting the starts and stops are the biggest contributors to variations in end to end journey times, because that is when the train is moving slowly. At big terminal stations which platform the train departs from/arrives at can make a huge difference. On the old Kings Cross layout Ps 6-9 straight out onto the fast line were almost always much faster to passing Finsbury Park than the others. Ps 10 and 11 were a stagger on the slow line to Holloway and about a minute slower.

In recent years one of the biggest contributions to start time variation has been running brake tests. Bizarrely, the fastest up road starts from Cambridge are from platform 4, even though the distance is the longest: from platform 4 the running brake test can be done on the through line so the driver can go for full power at the 90mph board, from all other platforms the running brake test is only just starting at that point.

 

[ComUtoR]

I’m afraid this is not correct. It makes a big difference. I’ve seen very detailed modelling on this, based on actual data, and the difference is stark.

in simple mechanics terms, the difference between 10 mph and 20mph over 300 metres is half a minute. Whilst it might seem to be only half a minute to you, such additional delay can quickly become a couple of hundred delay minutes across a complex network, particularly on approach to busy junctions and termini.

I'm going to have to disagree again. As long as the Signal is Red the Driver cannot pass it (under normal circumstances) The delay is for as long as the signal is at Red. It doesn't matter what the speed is on the approach to it. The Driver isn't going past it. I could take a whole minute between the magnet and then stopping. If the signal doesn't clear then I'm still going to be delayed until it does.

Now that this has been bumped to its own thread, its important to highlight that one of the reasons why Drivers will slowly approach a Red is because in many circumstances it will step up (yes, I know). I'm not talking about "approach control" but where a Driver knows they are following another service in front so can reduce their speed and trickle towards a red and maintain a reasonable distance without stopping their train. This is seen more on restrictive than towards a Red.

London Bridge and stations through 'The Core' are different because Drivers do not feel comfortable with closing up so will hang back till the get 'Two Yellows' Then I would agree that the Professional Driving Policies and driving techniques can cause delays. (still very minimum)

I will however, *fight you to the death regarding attribution for "approach controlled signals"



*bring forth the Lirpa

 

[O L Leigh]

I’m afraid this is not correct. It makes a big difference. I’ve seen very detailed modelling on this, based on actual data, and the difference is stark.

in simple mechanics terms, the difference between 10 mph and 20mph over 300 metres is half a minute. Whilst it might seem to be only half a minute to you, such additional delay can quickly become a couple of hundred delay minutes across a complex network, particularly on approach to busy junctions and termini.

I find this quite interesting.

The problem surely is that the actual delay incurred is dependent on how quickly the signal clears up from red, and that’s not always something that a driver can predict. If you get up to the signal and have to stop, the length of the delay is out of the hands of the driver and is decided instead by the signaller. Also, if you have to stop, it takes longer to get going again than if the signal clears before stopping.

I understand that the data can show certain things, and that’s terribly interesting. But when it comes to actual driver behaviour, we cannot always consider these things.

 

[Horizon22]

I’m afraid this is not correct. It makes a big difference. I’ve seen very detailed modelling on this, based on actual data, and the difference is stark.

in simple mechanics terms, the difference between 10 mph and 20mph over 300 metres is half a minute. Whilst it might seem to be only half a minute to you, such additional delay can quickly become a couple of hundred delay minutes across a complex network, particularly on approach to busy junctions and termini.

But surely you'll be ending up at 0mph anyway - (in your scenario that is; @ComUtoR has explained well why a driver might not go down to 0mph on metro routes) - and be waiting? In fact a lower speed - approaching a crawl - might be preferable as you've never reached 0mph and can start moving quicker. The modelling must be based on certain assumptions about X secs until the signal clears to a single yellow?

These are incredibly fine margins, and whilst I agree there's a performance impact, I'd argue its probably to the same degree of split-second but incorrect signaller regulation or just sheer forgetfulness sometimes!

 

[Bald Rick]

I find this quite interesting.

The problem surely is that the actual delay incurred is dependent on how quickly the signal clears up from red, and that’s not always something that a driver can predict. If you get up to the signal and have to stop, the length of the delay is out of the hands of the driver and is decided instead by the signaller. Also, if you have to stop, it takes longer to get going again than if the signal clears before stopping.

I understand that the data can show certain things, and that’s terribly interesting. But when it comes to actual driver behaviour, we cannot always consider these things.

But surely you'll be ending up at 0mph anyway - (in your scenario that is; @ComUtoR has explained well why a driver might not go down to 0mph on metro routes) - and be waiting? In fact a lower speed - approaching a crawl - might be preferable as you've never reached 0mph.

These are incredibly fine margins, and whilst I agree there's a performance impact, I'd argue it’s less than poor signaller regulation or just sheer forgetfulness sometimes!

I haven’t explained this very well. It’s not so much the delay it causes for that actual train*, but the delay caused by taking longer to clear the section or junction behind so that subsequent routes are free and can be reset for following trains. Especially at complex layouts with lots of conflicting movements, or where there are very short headways. Hence why ATO in the Thameslink core makes such a difference.

I agree poor regulation is also an issue. Been plenty of that this week as the new timetable actually gets operated for the first time!

* in some cases it is, particularly where approach control is involved.

 

[Egg Centric]

It does seem quite paradoxical that professional driver policies are potentially going to take discretion away from drivers AND lead to their eventual replacement with computers that won't have to adhere to them. Rather unfair too.

 

[Horizon22]

I haven’t explained this very well. It’s not so much the delay it causes for that actual train*, but the delay caused by taking longer to clear the section or junction behind so that subsequent routes are free and can be reset for following trains. Especially at complex layouts with lots of conflicting movements, or where there are very short headways. Hence why ATO in the Thameslink core makes such a difference.

I agree poor regulation is also an issue. Been plenty of that this week as the new timetable actually gets operated for the first time!

* in some cases it is, particularly where approach control is involved.

But surely the point remains that if you're approaching a red, and eventually reach 0mph you'll clear that section in the same amount of time? I suppose it depends how long your section is or the headways and how much overlap there is in turns of the driver aiming for 20mph or 10mph whilst between the sections.

You'll get to the point where these sort of time margins will be beyond almost all drivers, and thus is only really resolved by an ATO operation or more slack timetabling / extra capacity.

 

[O L Leigh]

I haven’t explained this very well. It’s not so much the delay it causes for that actual train*, but the delay caused by taking longer to clear the section or junction behind so that subsequent routes are free and can be reset for following trains. Especially at complex layouts with lots of conflicting movements, or where there are very short headways. Hence why ATO in the Thameslink core makes such a difference.

Understood.

But the delay to the following train is going to be dictated more by how long the preceding train takes to clear the previous signal’s overlap than how long it takes to get up to the red. Also, the train will be quicker to clear a section if it doesn’t have to stop as opposed to starting away from a stand.

I don’t know much about the Thameslink core. How does the signalling compare to elsewhere on the network?

 

[Bald Rick]

But surely the point remains that if you're approaching a red, and eventually reach 0mph you'll clear that section in the same amount of time?

No, not at all, because…

it depends how long your section is or the headways and how much overlap there is in turns of the driver aiming for 20mph or 10mph whilst between the sections.

to use a real example: on Thameslink, a train approaching a station stop (which is a quasi red signal for this example) red signal at St P low level on ATO can clear the section behind up to a minute quicker than a train being driven manually and cautiously.

If there was a conflicting move behind that train, then that slower driving has put an extra minute into the second train.

But the delay to the following train is going to be dictated more by how long the preceding train takes to clear the previous signal’s overlap than how long it takes to get up to the red.

Agreed, but its not always an overlap; sometimes it’s a fouling point for a conflicting move.

 

[Egg Centric]

But surely the point remains that if you're approaching a red, and eventually reach 0mph you'll clear that section in the same amount of time? I suppose it depends how long your section is or the headways and how much overlap there is in turns of the driver aiming for 20mph or 10mph whilst between the sections.

You'll get to the point where these sort of time margins will be beyond almost all drivers, and thus is only really resolved by an ATO operation or more slack timetabling / extra capacity.

Understood.

But the delay to the following train is going to be dictated more by how long the preceding train takes to clear the previous signal’s overlap than how long it takes to get up to the red. Also, the train will be quicker to clear a section if it doesn’t have to stop as opposed to starting away from a stand.

I don’t know much about the Thameslink core. How does the signalling compare to elsewhere on the network?

Have you guys every played SimSig? It's intuitively extremely obvious why Bald Rick is right here if you give it a go - partly cause of clearing the previous train's overlaps but much more importantly about clearing junctions (not just the ones the stopping train has passed, but also the ones its *own* overlap is holding in a certain way until it is proved stationary)

(Well he said it with the right terms before I posted this but after I wrote it so nvm - just play SimSig!)

 

[Bald Rick]

I don’t know much about the Thameslink core. How does the signalling compare to elsewhere on the network?

That depends what you mean. It’s 4 aspect, short sections, largely 30mph, 75 seconds maximum technical headway. With the ATO on it becomes a different beast. It’s just about possible to get two trains in a platform - tail end of one departing (at linespeed) and front end of the second entering (at or near linespeed).

A late train in ATO with nothing in front of it will easily recover 3-4 minutes more between Spa Road and Kentish Town compared to being driven manually. Accelerating at full power into a station platform with a red at the end takes a little bit if getting used to!