Timetables - Headway definition

[WesternGiraffe]

On a hypothetical stretch of railway, there is a regular clock face service. The longest section is 5 miles.

Trains of Type A, cover this longest section in 10 minutes, whereas trains of type B which occasionally run as required, take 12 minutes to cover the same section.

In a simplistic view, are the values used for headway in timetable planning deemed to be the values for travelling over this long section?

Am I correct in thinking that headway is the minimum amount of time after a train has passed that the next train can be timetabled to pass?

If it is presumed that the timetable shows that a train of type A runs, followed by a Train of Type B and then a train of Type A, a minimum frequency for trains of Type A would be every 22 minutes?

I realise that in a real world scenario there are a number of other factors that can affect headway, but I wanted to make sure that my understanding of this term was correct.

 

[steamybrian]

Speaking as a retired BR timetable planner then what you have said is correct however...

....................train A...........train B...... train A

station X .......1000..............1010......... 1022

station Y.........1010.............1022..........1032


....it leaves no margin for late running or delays. Timetables have built in recovery time.
It would be better if train B departed at say 1015 arriving Y at 1027 thus train A departed at 1030 thus giving service A a half hourly service departing station X at 00 and 30 mins past hour.
I admit tight timetabling is done in suburban areas but recovery time is built in.
Hope this is helpful

 

[edwin_m]

The OP's question implies a five mile signalling block section, which would most likely be treated as an Absolute Block headway, shown as AB in the Timetable Planning Rules. This requires the spacing of trains to be the travel time of the first section plus 2min for the communications between the signal boxes, though there are probably exceptions in certain places. Thus if the Type A left at 00 the Type B couldn't follow it until 12 and the next Type A wouldn't be until 26. This would allow for a viable 30min interval for Type A with a Type B running in between as necessary.

However, in areas with multi-aspect signalling, the headway depends on the signal spacing, number of aspects, line speed and various other factors. It is the minimum interval between consecutive scheduled times of trains passing a timing point, before the second train starts encountering adverse aspects due to the first one.

All the above are strictly known as planning headways, which are usually a bit more than the technical headway possible with the signalling.

 

[John R]

In determining the headway, surely you would need to add at the very least the time needed once the train has left the section before the one behind doesn’t get checked by restrictive signal aspects?

 

[WesternGiraffe]

Thanks for the confirmation. Some of the other considerations, not mentioned earlier I had buzzing around in my head were such things as length of train, 775, intermodal versus 2 car DMU, gradients, stopping patterns, intermediate blocksignals or MAS equivilant, weight of train, route knowledge and lots of others.

One question that does come to mind though is that if there's a level crossing within the section, splitting the long section into at least 2, is the multiple sections taken individually or the entire section as a whole? I'm sure it must be 2 as the signal protecting the crossing is a "stop" signal or is this another occasion when allowances come into play as generally services will only ever be checked and not stopped?

 

[zwk500]

In determining the headway, surely you would need to add at the very least the time needed once the train has left the section before the one behind doesn’t get checked by restrictive signal aspects?

The normal way to calculate the planning headway is to calculate the time before the following train will see the least restrictive aspect it would normally see.

So there is time to clear the overlap, and time for the following driver to see the signal.

Full methodology is publicly listed in the National Timetebale planning rules on NRs website.

== Doublepost prevention - post automatically merged: 21 Jun 2026 ==

Thanks for the confirmation. Some of the other considerations, not mentioned earlier I had buzzing around in my head were such things as length of train, 775, intermodal versus 2 car DMU, gradients, stopping patterns, intermediate blocksignals or MAS equivilant, weight of train,

All of that does need to be taken into account.

If there is a soignificant difference between traffic types you may need to specify separate headways if planned on a number of minutes basis.

For absolute block sections, the headways are defined as the limit of each section plus an allowance for signallers actions, so the headway does not need to allow for the runtime differences separately.

route knowledge and lots of others.

Route knowledge not really.

One question that does come to mind though is that if there's a level crossing within the section, splitting the long section into at least 2, is the multiple sections taken individually or the entire section as a whole? I'm sure it must be 2 as the signal protecting the crossing is a "stop" signal or is this another occasion when allowances come into play as generally services will only ever be checked and not stopped?

If there are multiple signals between the timing points, the intermediate signals are accounted for by writing the headway as a number of minutes to elapse between trains at each timing point.

 

[WesternGiraffe]

Re Timing Points, looking at the WTT there are different timing points over the same set of sections, which is, as far as I can tell directly related to the type of service e.g local vs express or passenger vs freight.

What governs which location is a timing point?

== Doublepost prevention - post automatically merged: 21 Jun 2026 ==

And by "route knowledge", I was thinking of occasions where a driver may be on a diversionary route requireing to be conducted, and an additional amount of time allowance for the conductor to instruct the driver.

Although having sat back and considered this further, I'm not sure if this is still permitted current practice?

 

[zwk500]

Re Timing Points, looking at the WTT there are different timing points over the same set of sections, which is, as far as I can tell directly related to the type of service e.g local vs express or passenger vs freight.

What governs which location is a timing point?

A mix of factors. If it's an absolute block post or location where a headway changes it will tend to be a mandatory timing point. Junctions with significant crossing movements (or far from another station) will also be mandatory.

Some locations are mandatory only for certain traffic or lines as you mention and this will be govered by the layout and traffic mix.

Signalling can also drive which tiplocs are required, as complex areas will require route codes at several junctions on approach to get correct routings, especially if ARS.

Sometimes locations are made mandatory just to break up a very long section.

There are also locations that were mandatory for historic reasons, but on changing the junctions for whatever reason the tiplocs were not adjusted and so its just lasted through.

And by "route knowledge", I was thinking of occasions where a driver may be on a diversionary route requireing to be conducted, and an additional amount of time allowance for the conductor to instruct the driver.

Any additional time needed for this would be planned as a longer runtime and would not directly impact the headway values. Although most likely any briefing would be done stationary.

Although having sat back and considered this further, I'm not sure if this is still permitted current practice?

Route conducting is absolutely still common practice.

 

[The Planner]

Add in the fact you can time absolute block as TCB and not need the +2 if its colour light and not someone pulling levers and block bells. As with previous answers, locations at the top of hills etc can be mandatory if the landscape can dictate how trains perform.

 

[800002]

Drawing it all together, then.
If there is no intermediate, mandatory, timing point between point A and point B, then 'headway' shall be classed as AB+2, absolute Block plus 2 minutes.
Train 2 cannot pass / depart Point A until 2 minutes after Train 1 has arrived / passed Point B.

If there are intermediate, mandatory, timing points between Point A and Point B then headway shall be based upon the time it takes for Train 2 to arrive at Point A, showing the least restrictive aspect (A green 'signal'), after Train 1 has past Point A.
The cycle continues from mandatory timing Point to mandatory timing Point. The headway value can increase or decrease based on this calculation.

To remember, there may be stopping points/ stations between two blocking points or intermediate mandatory timing points which affect how long the train is in the 'block section' for.

There can be huge block sections of double track which can take considerable time to clear given the Track and Train characteristics.
10+ minutes of 'headway' is not uncommon.

 

[zwk500]

Drawing it all together, then.
If there is no intermediate, mandatory, timing point between point A and point B, then 'headway' shall be classed as AB+2, absolute Block plus 2 minutes.
Train 2 cannot pass / depart Point A until 2 minutes after Train 1 has arrived / passed Point B.

Will normally be classed but yes.

If there are intermediate, mandatory, timing points between Point A and Point B then headway shall be based upon the time it takes for Train 2 to arrive at Point A, showing the least restrictive aspect (A green 'signal'), after Train 1 has past Point A.
The cycle continues from mandatory timing Point to mandatory timing Point. The headway value can increase or decrease based on this calculation.

Essentially yes. Although typically the technical headway is calculated for all signals in a section and then the timing points adjusted if necessary.

To remember, there may be stopping points/ stations between two blocking points or intermediate mandatory timing points which affect how long the train is in the 'block section' for.

Correct.

There can be huge block sections of double track which can take considerable time to clear given the Track and Train characteristics.
10+ minutes of 'headway' is not uncommon.

Headways of 10+ minutes are fairly uncommon because lines resignalled to multiple sections will Typically be busy enough to require a lower headway.
Block sections where some or all trains take more than 10 minutes to clear the section are reasonably easy to find.

 

[800002]

Will normally be classed but yes.

Apologies, yes.
There are of course exceptions to most rules on the railway (planning wise).

Essentially yes. Although typically the technical headway is calculated for all signals in a section and then the timing points adjusted if necessary.

I was thinking more broadly across consecutive blocks, yes.
Eg, bletchely to Bedford has multiple 4 minute headway sections, then a five minute section.

Correct

Phew! (Thank you).

Headways of 10+ minutes are fairly uncommon because lines resignalled to multiple sections will Typically be busy enough to require a lower headway.
Block sections where some or all trains take more than 10 minutes to clear the section are reasonably easy to find.

Fenny stratford - Woburn sands can be 9 minutes (it isnt In all circumstances)
Mantle wood - Aylesbury is 9 minutes.
Hamilton - Newton is 6 minutes
Busby - Busby Jn is 10 (following all stations stopper)
Hunterston Jn - Largs is 10 in one direction / 7 in the other

I admit I may have pulled 10 minutes out of thin air, they are very much the exception.

Talking of exceptions, there are such circumstances of AB+0, where the following train can pass point A simultaneously with train 1 is passing / departing Point B.

The AB sections take much more work as they are solely dependant on train As speed and ability to clear the block section.

I've seen a 20+ minute AB section (effectively) as it was:
Climbing gradient, from a stand.
Hauled by steam with a full rake of passenger stock.
(Possibilities are endless).
A diesel would likely be clearing it in 8 minutes.

== Doublepost prevention - post automatically merged: 24 Jun 2026 ==

As I probably should have done earlier.

Headway is defined in this Network Rail Rules document:

https://sacuksprodnrdigital0001.blob.core.windows.net/operational-rules/Operational%20Rules/TPR%20documents/TPR%202026/V4/TPR%202026%20V4%20NAT.pdf

Page 7, section 1.5
This is the National Train Planning Rules
Alas there is far too much information to officially share while using my mobile device.

 

[zwk500]

Apologies, yes.
There are of course exceptions to most rules on the railway (planning wise).

Hahah yes. I used to joke with colleagues that the railway's attitude planning rules was to make a rule and then immediately create an exception for it.

Fenny stratford - Woburn sands can be 9 minutes (it isnt In all circumstances)
Mantle wood - Aylesbury is 9 minutes.
Hamilton - Newton is 6 minutes
Busby - Busby Jn is 10 (following all stations stopper)
Hunterston Jn - Largs is 10 in one direction / 7 in the other

I admit I may have pulled 10 minutes out of thin air, they are very much the exception.

They do exist - IIRC Shrewsbury to Gobowen is 11 minutes because it's got an intermediate block signal. But they are not too common. Most numerical headways are 3-5 minutes.

Talking of exceptions, there are such circumstances of AB+0, where the following train can pass point A simultaneously with train 1 is passing / departing Point B.

Correct, as @The Planner mentioned. This can be because the signalling is actually Track Circuit Block rather than Absolute Block and therefore no signaller actions to offer and accept the trains are required, or because the timing point is not quite at the section limit and so the signaller's 2 minutes is included while it moves from e.g. the box to the station.
There's also situations where the signal box is at one end of the station, so trains in one direct are AB+2 based on arrival but in the other direction AB+2 based on departure time.

The AB sections take much more work as they are solely dependant on train As speed and ability to clear the block section.

Correct

I've seen a 20+ minute AB section (effectively) as it was:
Climbing gradient, from a stand.
Hauled by steam with a full rake of passenger stock.
(Possibilities are endless).
A diesel would likely be clearing it in 8 minutes.

On a technical point, it's not 100% correct to describe an AB headway as being 'x' minutes because of the fact that the time between trains at any given point is not fixed. Whereas if the headway is 4 minutes, trains at Point A and Point B only need to be 4 minutes apart, regardless of whether or not train 1 takes longer through the section than train 2 normally would (this is when you'd use 'pathing time' to allow train 2 to keep 4 minutes behind).

As I probably should have done earlier.

Headway is defined in this Network Rail Rules document:

https://sacuksprodnrdigital0001.blob.core.windows.net/operational-rules/Operational%20Rules/TPR%20documents/TPR%202026/V4/TPR%202026%20V4%20NAT.pdf

Page 7, section 1.5
This is the National Train Planning Rules
Alas there is far too much information to officially share while using my mobile device.

Yes, I referred to that document in post #6. It is, as you say, quite a lot to consider.

 

[800002]

Hahah yes. I used to joke with colleagues that the railway's attitude planning rules was to make a rule and then immediately create an exception for it.

That's the way we handled (the lack of) SRTs, Sectional Running Times, back in the day!

They do exist - IIRC Shrewsbury to Gobowen is 11 minutes because it's got an intermediate block signal. But they are not too common. Most numerical headways are 3-5 minutes.

Yes, wholeheartedly agree.
Apologies, I like to work in extremis.

On a technical point, it's not 100% correct to describe an AB headway as being 'x' minutes because of the fact that the time between trains at any given point is not fixed. Whereas if the headway is 4 minutes, trains at Point A and Point B only need to be 4 minutes apart, regardless of whether or not train 1 takes longer through the section than train 2 normally would (this is when you'd use 'pathing time' to allow train 2 to keep 4 minutes behind).

Yes. Quite accurate, on the technicality.
It's important to ascertain those points / locations to which it is 'mandatory' to time two services at.
Remembering you're unable to apply a stated headway to a train at a location which it is not physically timed at. (Usually visible on a graphical display where it appears a train is closer than it is due to scaling and perspective.

Yes, I referred to that document in post #6. It is, as you say, quite a lot to consider.

I thought i read that (short memory, sorry).
We can see it now as rhe diagrams are quite interesting.
(Who doesn't love a good NR diagram!)

 

[zwk500]

That's the way we handled (the lack of) SRTs, Sectional Running Times, back in the day!

Ah, the joy of planning on TRTs in TPS... 'Lock everything, and document it twice'

Yes, wholeheartedly agree.
Apologies, I like to work in extremis.

Quite alright!

Yes. Quite accurate, on the technicality.
It's important to ascertain those points / locations to which it is 'mandatory' to time two services at.
Remembering you're unable to apply a stated headway to a train at a location which it is not physically timed at. (Usually visible on a graphical display where it appears a train is closer than it is due to scaling and perspective.

As you say, although if memory serves there was a headway change at Penkridge for quite some time when that station was only a conditional TIPLOC for stopping trains. I think it's now been made a mandatory TIPLOC though. And there might be one or two AB boxes that haven't yet made it into BPlan (Onibury on the Marches if memory serves).

 

[The Planner]

Ah, the joy of planning on TRTs in TPS... 'Lock everything, and document it twice'

Quite alright!

As you say, although if memory serves there was a headway change at Penkridge for quite some time when that station was only a conditional TIPLOC for stopping trains. I think it's now been made a mandatory TIPLOC though. And there might be one or two AB boxes that haven't yet made it into BPlan (Onibury on the Marches if memory serves).

Penkridge was put in as its sort of a route boundary. Back in train plan days you never had TRT issues, you got people to input the SRT directly and as long as it was put in A Plan before you uploaded you were fine.

 

[800002]

Ah, the joy of planning on TRTs in TPS... 'Lock everything, and document it twice'

I know exactly what you're referring to! ; -)
It was really something one needed to be consistently aware of.

at Penkridge for quite some time when that station was only a conditional TIPLOC for stopping trains

It's now mandatory, yes. For everything. I think it's now a boundary point between routes.
But not 100% (TPRs required to verify)

^^as kindly indicated above. *edit

 

[Geep]

The OP mentioned the incidence of level crossings in one post. These come in a wide variety of types, some with stop signals, some without. Nearly all private crossings do not, in fact, have stop signals, and not all public road crossings with signals necessarily affect the signalling system, strange as it may seem. During times when the Absolute Block system was the primary signalling system, intermediate level crossings where public roads crossed could be protected by stop signals worked either by a gatekeeper, often living on site in a crossing keeper's cottage, or by what in the North-east of England was known as a gate box. In other cases, the only 'stop' signal was the red roundel on the gates, when the gates were closed against rail.

Usually, such intermediate locations were made aware of the working of trains by having indicators or copy block instruments showing if trains were in the section or not, and such intermediaries had the responsibility of not opening the road gates whilst a train was in section. It was a relatively cheap system which worked adequately well, but was not foolproof, and mistakes could be and were made.

The point to bear in mind is that would be no effects on the headway calculation unless there was a proper signalbox with a working block system and stop signals.