Delays and Headways

[Legolash2o]

I'm wanting to select a train that was delayed to see if it was caused by another train.

I will be attempting this by comparing the schedule and actual times and finding any overlaps.

The same could be done to predict a delay, so that if train A is delayed, then it would also impact train B, etc.

My question is, would you have to include headway and junctions margins for the calculation?

If train A arrived at 11:58 (3 mins late) and train B was due to be at that location at 12:00, it wouldn't overlap. However, if a train required a 3 min headway, it would by 1 minute. This excludes performance allowances, etc.

Hope that makes sense.

 

[DPQ]

Yes

It should be noted that

a) there is a tolerance in headways and junction margins of between 5 and 20%. So just because a planned headway is 3 minutes doesn't mean the actual technical headway is 3 minutes.

b) Not all delays are timetable related and won't be identified from a schedule. These could be many different factors.

 

[Legolash2o]

Thanks for the reply.

I'll add a tolerance and see how that works. Do you think it's safe to assume that a headway is 3 minutes by default and not specified within the Train Planing Rules?

My initial aim is to select a train that is late and then try and estimate what other trains would be impacted due to that train. That of course doesn't stop train B from being delayed by another issue (loco fault, crew changes, etc).

Thanks again.

 

[The Planner]

Yes

It should be noted that

a) there is a tolerance in headways and junction margins of between 5 and 20%. So just because a planned headway is 3 minutes doesn't mean the actual technical headway is 3 minutes.

b) Not all delays are timetable related and won't be identified from a schedule. These could be many different factors.

The tolerance is very rarely that low. Some headways are up to double their technical values. Junction margins will depend on how much they are rounded up by unless they are defaulting to the old 3 before 2 after rule.

 

[Legolash2o]

The tolerance is very rarely that low. Some headways are up to double their technical values. Junction margins will depend on how much they are rounded up by unless they are defaulting to the old 3 before 2 after rule.

Sorry, what do you mean by the 3 before 2 rule?

 

[The Planner]

Sorry, what do you mean by the 3 before 2 rule?

At a junction, if you are conflicting with another train, you either cross it 3 minutes before it or 2 minutes after it.

 

[Legolash2o]

Thanks!

 

[Wilts Wanderer]

At a junction, if you are conflicting with another train, you either cross it 3 minutes before it or 2 minutes after it.

Which has never made sense to me, as if you're crossing 2 mins after another train, from the other trains perspective its not achieving the 3-before principle!

 

[Legolash2o]

Which has never made sense to me, as if you're crossing 2 mins after another train, from the other trains perspective its not achieving the 3-before principle!

I'm glad I'm not the only one who thought that!

 

[Ianno87]

Which has never made sense to me, as if you're crossing 2 mins after another train, from the other trains perspective its not achieving the 3-before principle!

I'm glad I'm not the only one who thought that!

There is logic at most junctions in that the margin is measured relative to the train taking the main line / fastest route.

 

[Legolash2o]

Are there times where you need to apply a headway AND a junction margin?

 

[The Planner]

There is logic at most junctions in that the margin is measured relative to the train taking the main line / fastest route.

Exactly this.

== Doublepost prevention - post automatically merged: 26 Jun 2021 ==

Are there times where you need to apply a headway AND a junction margin?

If you are following something then diverging then yes. Though applying a headway is a given at all times.

 

[Ianno87]

Are there times where you need to apply a headway AND a junction margin?

Between two given trains you'd only apply one or the other.

If, for example you have two tracks operating the same direction (Fast Line and Slow Line) with crossovers between them, it would be:
-A headway if trains are diverging from or converging to the same line
-A junction margin if one train is crossing from Slow to Fast and the other from Fast to Slow Line

Slade Lane Junction would be a good example, I think.

 

[Legolash2o]

Exactly this.

== Doublepost prevention - post automatically merged: 26 Jun 2021 ==


If you are following something then diverging then yes. Though applying a headway is a given at all times.

Between two given trains you'd only apply one or the other.

If, for example you have two tracks operating the same direction (Fast Line and Slow Line) with crossovers between them, it would be:
-A headway if trains are diverging from or converging to the same line
-A junction margin if one train is crossing from Slow to Fast and the other from Fast to Slow Line

Slade Lane Junction would be a good example, I think.

Thabks. I think I understand.

If a train was on the fast line and a second train behind on the slow line wanted to move onto the fast line to change at the junction, it would have to use both a headway and a junction margin? As it's converging on the same line as the first train.

I may be over thinking this.

 

[Ianno87]

Thabks. I think I understand.

If a train was on the fast line and a second train behind on the slow line wanted to move onto the fast line to change at the junction, it would have to use both a headway and a junction margin? As it's converging on the same line as the first train.

I may be over thinking this.

No, in that case you would just apply a headway (although in some cases a diminished headway may be possible at the merge point if the second train is joining at a lower speed, subject to the standard headway then applying from the next timing point)

 

[Legolash2o]

Ah ok.

I was thinking of the Slade Lane Jn example where a freight train is heading south on the Up Slow but a passenger train on the Up Fast wanted to change onto the Up Slow (by passing over the Down Fast first before heading West at the junction.

 

[Ianno87]

Ah ok.

I was thinking of the Slade Lane Jn example where a freight train is heading south on the Up Slow but a passenger train on the Up Fast wanted to change onto the Up Slow (by passing over the Down Fast first before heading West at the junction.

I think you have the line designations mixed up - a train running Up Fast (from Piccadilly) to Up Slow (towards Stockport) does not need to cross or interact with the Down Fast - it would be on the easternmost track the whole way through the junction.

The lines north of the junction are arranged (West to East): DS/US/DF/UF.

South of the junction: Down Styal / Up Styal / DS / DF / UF / US.

 

[Legolash2o]

Ah you are correct!

I'll try again lol.

If we assume DS/US/DF/UF. Before the junction there's is a freight train on US and a passenger train is on UF. The freight train isn't using the junction and just carrying straight on South.

The passenger train needs to use the junction to head west. It needs to crossover either straight on the US behind the freight train (or temporarily on to the DF for a short section depending on the point layout). Then from the US, it would then take the junction onto the another route (UF).

I hope that explains it better.

 

[The Planner]

Its still just a junction margin if you are following the freight over the junction as the full headway wouldnt apply as you are diverging as Ian noted earlier.

 

[Legolash2o]

Awesome, thanks both!

 

[Spartacus]

Not only would you have to include headways and junction margins, but things like signal overlaps (something that for a long time barely got a mention in the train planning rules!), section lengths and train lengths. I’ve lost count of the number of long freights booked across Trent East Jn to stand at Sheet Stores, with another service booked over Trent East while it’s stood there. On paper there’s no clash, in reality the rear of the freight’s almost certain to be fouling the junction behind, and you certainly can’t pass Trent East to Sheet Stores before the previous train has passed Sheet Stores, no matter how many minutes apart they are at each.

 

[Ianno87]

Not only would you have to include headways and junction margins, but things like signal overlaps (something that for a long time barely got a mention in the train planning rules!), section lengths and train lengths. I’ve lost count of the number of long freights booked across Trent East Jn to stand at Sheet Stores, with another service booked over Trent East while it’s stood there. On paper there’s no clash, in reality the rear of the freight’s almost certain to be fouling the junction behind, and you certainly can’t pass Trent East to Sheet Stores before the previous train has passed Sheet Stores, no matter how many minutes apart they are at each.

Those examples really should be accounted for in the rules too.

 

[The Planner]

Not only would you have to include headways and junction margins, but things like signal overlaps (something that for a long time barely got a mention in the train planning rules!), section lengths and train lengths. I’ve lost count of the number of long freights booked across Trent East Jn to stand at Sheet Stores, with another service booked over Trent East while it’s stood there. On paper there’s no clash, in reality the rear of the freight’s almost certain to be fouling the junction behind, and you certainly can’t pass Trent East to Sheet Stores before the previous train has passed Sheet Stores, no matter how many minutes apart they are at each.

Overlaps are included in junction margins and headways. Your example is about train length which is a big elephant in the room as timing loads cannot account for it, you have to rely on the information from the FOC and planner knowledge. However more and more specific length rules are appearing in the TPR now. If any delay is getting Q coded then someone should be looking at it.

 

[IanXC]

Theres also the question of intermediate block sections, there are plenty which are not accounted for in the train planning rules, and so can't be planned against, but which are present and will clearly have an impact on how delays propogate.

 

[The Planner]

Theres also the question of intermediate block sections, there are plenty which are not accounted for in the train planning rules, and so can't be planned against, but which are present and will clearly have an impact on how delays propogate.

Interested to hear some examples.

 

[IanXC]

Interested to hear some examples.

A couple that spring to mind:
Yorkshire Coast, Beverley and Driffield both work Absolute Block, but have intermediate block sections dating from the closure of other signal boxes, looking at the timetable planning rules theres no account for this capacity, nor have I ever seen it used even when Charters etc run.
Tyne Valley, again some of the boxes have intermediate block sections which can't be planned against.

In both cases if there is late running these additional sections are used and trains appear to pass through the area quicker than the timetable planning rules allow.

 

[edwin_m]

A couple that spring to mind:
Yorkshire Coast, Beverley and Driffield both work Absolute Block, but have intermediate block sections dating from the closure of other signal boxes, looking at the timetable planning rules theres no account for this capacity, nor have I ever seen it used even when Charters etc run.
Tyne Valley, again some of the boxes have intermediate block sections which can't be planned against.

In both cases if there is late running these additional sections are used and trains appear to pass through the area quicker than the timetable planning rules allow.

So effectively there could be double the capacity of what is being timetabled. I wonder what happens if a box is closed and replaced by an IB section - does the route capacity suddenly drop? Indeed, do the TPRs actually say you can't use IBs, as they are effectively block section boundaries, although a planner wouldn't have timing information for them?

However, the default TPR headway for AB sections is transit time plus 2min, and since some of that 2min gets used by signallers ringing bells at each other there's not really much slack in AB sections that don't have IBs.

 

[Ianno87]

A couple that spring to mind:
Yorkshire Coast, Beverley and Driffield both work Absolute Block, but have intermediate block sections dating from the closure of other signal boxes, looking at the timetable planning rules theres no account for this capacity, nor have I ever seen it used even when Charters etc run.
Tyne Valley, again some of the boxes have intermediate block sections which can't be planned against.

In both cases if there is late running these additional sections are used and trains appear to pass through the area quicker than the timetable planning rules allow.

I suspect here that in these instances there's never been a particular need to make use of the extra sections (other than for a bit more timetable flexibility), so the rules don't reflect being able to use them, as they've probably never been challenged (as there's nothing to be gained if nobody actually wants to run extra trains).

So effectively there could be double the capacity of what is being timetabled. I wonder what happens if a box is closed and replaced by an IB section - does the route capacity suddenly drop? Indeed, do the TPRs actually say you can't use IBs, as they are effectively block section boundaries, although a planner wouldn't have timing information for them?

However, the default TPR headway for AB sections is transit time plus 2min, and since some of that 2min gets used by signallers ringing bells at each other there's not really much slack in AB sections that don't have IBs.

A knowledgeable planner could be able to estimate the passing time at IBs by reading off the timetable graph, and timing a following train accordingly, although it wouldn't technically be compliant with the stated rules (so would risk rejection or delay minutes being attributed). More likely, I suspect, to be done for trains on an STP basis if it's the only way of, say, getting an ad-hoc freight or empty stock move in.

If it was being done for the Long Term Plan, I suspect the "proper" route would be to formally propose a change to the rules to permit timing at the IB sections.

 

[Watershed]

So effectively there could be double the capacity of what is being timetabled.

Not necessarily double, but yes, increased capacity.

However, the default TPR headway for AB sections is transit time plus 2min, and since some of that 2min gets used by signallers ringing bells at each other there's not really much slack in AB sections that don't have IBs.

Transit + 2 mins is about the same margin, percentage wise, as the default 3 minute headway value on TCB lines.