ORR P-Code data releases.

[YorkRailFan]

P-coded cancellations​

Periodic (4-weekly) statistics on resource availability shortage pre-cancellations (i.e. changes to train services caused by non-availability of staff or rolling stock) that are included in a revised timetable, and therefore may not be appearing in operators’ cancellations scores.

Operators have been providing ORR with the number of resource availability shortage “P*-coded” pre-cancellations since railway period 11 (8 January 2023 to 4 February 2023). The factsheet and data table present data for both on the day cancellations and resource availability shortage “P*-coded” pre-cancellations by cause for each operator.

For information on any revisions, please see our Revisions log.

Our statistical practice is regulated by the Office for Statistics Regulation (OSR).

If you have any questions or feedback on these statistics, please contact [email protected]

More passenger rail performance data on punctuality, reliability and causes of delay are available on our passenger rail performance page.

Latest factsheet​

Passenger rail performance: Cancellations data factsheet, rail period 8 (15 October 2023 to 11 November 2023)​

Date published: 01 Dec 2023
Date next published: 05 Jan 2024

Key results​

• In period 8, 12 operators recorded resource availability shortage “P*-coded” pre-cancellations. Of these, eight were due to the operator having resource availability shortage, three were due to Network Rail resource availability shortage, and one was due to both operator and Network Rail resource availability shortage (TfW Rail).

Historical factsheets​

• 2023-24​

  • Passenger rail performance: Cancellations data factsheet, rail period 7 (17 September 2023 to 14 October 2023)
  • Passenger rail performance: Cancellations data factsheet, rail period 6 (20 August 2023 to 16 September 2023)
  • Passenger rail performance: Cancellations data factsheet, rail period 5 (23 July 2023 to 19 August 2023)
  • Passenger rail performance: Cancellations data factsheet, rail period 4 (25 June 2023 to 22 July 2023)
  • Passenger rail performance: Cancellations data factsheet, rail period 3 (28 May 2023 to 24 June 2023)
  • Passenger rail performance: Cancellations data factsheet, rail period 2 (30 April 2023 to 27 May 2023)
  • Passenger rail performance: Cancellations data factsheet, rail period 1 (1 April 2023 to 29 April 2023)

• 2022-23​

  • Passenger rail performance: Cancellations data factsheet, rail period 13 (5 March 2023 to 31 March 2023)
  • Passenger rail performance: Cancellations data factsheet, rail period 12 (5 February 2023 to 4 March 2023)
  • Passenger rail performance: Cancellations data factsheet, rail period 11 (8 January 2023 to 4 February 2023)

Data tables​

• Periodic​

  • Table 3128 - Cancellations by operator and p-code

P-coded cancellations | ORR Data Portal

Most operators are P-coding trains due to a shortage of train crew, there are exceptions (most notably GC due to the reliability of the 180s) and some operators P-coding trains due "to a lack of Network Rail resources" with TFW being a prime example of this.

 

[Narrowgauge725]

So Northern and TPE are the worst not a lot of hope for the north of the country!

 

[YorkRailFan]

So Northern and TPE are the worst not a lot of hope for the north of the country!

There hasn't been hope for us in decades.

 

[Snow1964]

There is now new ORR data
Cancellations by Operator is table 4.3
Cross Country is worst with 7.5% cancelled
Compare that to 7 Operators less than 2% cancelled

https://dataportal.orr.gov.uk/media/1kdfjw4u/performance_stats_release_jul-sep-2023.pdf

 

[Djgr]

Quite frankly it's appalling.

 

[Andymo]

The p coded cancellations are the real cancellation rates and are appalling.

 

[YorkRailFan]

Periodic (4-weekly) statistics on resource availability shortage pre-cancellations (i.e. changes to train services caused by non-availability of staff or rolling stock) that are included in a revised timetable, and therefore may not be appearing in operators’ cancellations scores.

Operators have been providing ORR with the number of resource availability shortage “P*-coded” pre-cancellations since railway period 11 (8 January 2023 to 4 February 2023). The factsheet and data table present data for both on the day cancellations and resource availability shortage “P*-coded” pre-cancellations by cause for each operator.

For information on any revisions, please see our Revisions log.

Our statistical practice is regulated by the Office for Statistics Regulation (OSR).

If you have any questions or feedback on these statistics, please contact [email protected]

More passenger rail performance data on punctuality, reliability and causes of delay are available on our passenger rail performance page.

Latest factsheet
Passenger rail performance: Cancellations data factsheet, rail period 9 (12 November 2023 to 9 December 2023)
Passenger rail performance: Cancellations data factsheet, rail period 9 (12 November 2023 to 9 December 2023)
Date published: 05 Jan 2024
Date next published: 26 Jan 2024

Key results
In period 9, ten operators recorded resource availability shortage “P*-coded” pre-cancellations. Of these, all were due to the operator having resource availability shortage (none were due to Network Rail resource availability shortage).

P-coded cancellations | ORR Data Portal

The cooler weather will have helped GC as their 180s won't be overheating as much, Northern and TPE still very high, along with TFW. I have noticed that LNER have been cancelling quite a few trains due to "a fault on this train" as well.

 

[Snow1964]

The latest data has been released today
covering 12 Nov to 9 Dec

TPE adjusted (including P-Code) cancellations 13.5%
Northern 10.5%
TfW 8.0%
Grand Central 7.6%
West Midlands 7.2%
LNER 6.2%
East Midlands 5.3%
Scotrail 2.4%
Chiltern 2.1%
C2C 1.2%

https://dataportal.orr.gov.uk/media/eg2dtcwo/passenger-rail-performance-cancellations-data-period-9.pdf

 

[Nicholas Lewis]

The latest data has been released today
covering 12 Nov to 9 Dec

TPE adjusted (including P-Code) cancellations 13.5%
Northern 10.5%
TfW 8.0%
Grand Central 7.6%
West Midlands 7.2%
LNER 6.2%
East Midlands 5.3%
Scotrail 2.4%
Chiltern 2.1%
C2C 1.2%

https://dataportal.orr.gov.uk/media/eg2dtcwo/passenger-rail-performance-cancellations-data-period-9.pdf

Top 2 OLR and then number 3 run by the Welsh Government not a good advert for the public sector ownership

 

[ComUtoR]

Top 2 OLR and then number 3 run by the Welsh Government not a good advert for the public sector ownership

Is it ever ? A common theme across this forum is a call to punish a TOC and get it back under public ownership. Take the franchise off them etc etc. All evidence points to the Government not being able to run a proverbial in a proverbial.

Probably worth of its own thread but has the Gov every successfully run the railway ?

Most operators are P-coding trains due to a shortage of train crew, there are exceptions (most notably GC due to the reliability of the 180s)

Why does the reliability of the 180s change the availability of traincrew ?

 

[Waterlemon]

Top 2 OLR and then number 3 run by the Welsh Government not a good advert for the public sector ownership

Just shows who owns them doesnt really matter. Will only improve with investment.

 

[chris_in_salop]

Is it ever ? A common theme across this forum is a call to punish a TOC and get it back under public ownership. Take the franchise off them etc etc. All evidence points to the Government not being able to run a proverbial in a proverbial.

Is it really? A much more vocal theme on this forum is that we could reach railway nirvana if only we privatised everything. Witness the barely disguised glee every time something goes wrong at TfW.

But nationalisation doesn't have to equal micro-management plus austerity. Other models are available.

 

[YorkRailFan]

Why does the reliability of the 180s change the availability of traincrew ?

The report from the ORR says that the P-Coding is due to lack of staffing or availability of rolling stock.

i.e. changes to train services caused by non-availability of staff or rolling stock

P-coded cancellations | ORR Data Portal

 

[ComUtoR]

The report from the ORR says that the P-Coding is due to lack of staffing or availability of rolling stock.

Gonna tread a little carefully here. Please bear with me.

Your post implied that 'GC' aren't high on the shortage of crew statistics because they had reliable trains. This is what threw me. Have I inferred wrongly and you meant something different ? Surely each column should be taken independently ? Are you adding part/full P code cancellations ?

As a follow up. Is train reliability linked to these P codes ? Would a low MTIN unit have fewer P codes ?

I had a very brief look at one of those links and it seems that P codes only apply in 3 circumstances. "Shortage of traincrew", "shortage of rolling stock" and "shortage of Network Rail resources" Those are very broad and reading about P coding on this forum and trying to breakdown the dark arts of delay attribution. I'm worried that a narrative is being created.

Cheers in advance.

 

[Javelin_55]

I had a very brief look at one of those links and it seems that P codes only apply in 3 circumstances. "Shortage of traincrew", "shortage of rolling stock" and "shortage of Network Rail resources"

Trains can be P-coded for any reason, as long as the cancellation is agreed before 10pm the day before. These will be attributed under delay code PG.

If operators are able to provide a breakdown of which trains are P-coded specifically for different reasons, then it will rely on the operators keeping their own notes on why. And having worked with a number of control staff in a few TOCs, I'd be very surprised if that data is halfway accurate. Not a reflection on them, just the workload they have.

I was under the impression no-one gave a monkeys why trains were P-coded anyway, so I'm interested that this data exists in the first place.

 

[Snow1964]

I was under the impression no-one gave a monkeys why trains were P-coded anyway, so I'm interested that this data exists in the first place.

P-codes were originally intended for exceptional events, floods, snowdrifts, fire or bomb evacuations closing lines etc. and allowed items beyond railways normal control to be excluded from cancellation statistics.

So you were right, no one cared, but then some Operators saw it as a way to manipulate statistics so regulator now wants to know why as trying to clamp down on those who don't provide enough staff, or make enough trains serviceable.

Why are statistics now accurately required , it's because if a train is in timetable 12 weeks before, and tickets are sold, it disrupts the person with a ticket. Regardless of if cancelled month before, week before, day before or hour before, all of them disrupt their travel plans and any related consequences eg events or meetings being attended. Passengers can't ask a football match or concert to change the time.

If control staff do not understand the consequences (and think they are running trains to make railways life easy, rather than stopping passengers getting to paid for events on time), then there is a serious training gap.

 

[YorkRailFan]

Gonna tread a little carefully here. Please bear with me.

Your post implied that 'GC' aren't high on the shortage of crew statistics because they had reliable trains. This is what threw me. Have I inferred wrongly and you meant something different ? Surely each column should be taken independently ? Are you adding part/full P code cancellations ?

As a follow up. Is train reliability linked to these P codes ? Would a low MTIN unit have fewer P codes ?

I had a very brief look at one of those links and it seems that P codes only apply in 3 circumstances. "Shortage of traincrew", "shortage of rolling stock" and "shortage of Network Rail resources" Those are very broad and reading about P coding on this forum and trying to breakdown the dark arts of delay attribution. I'm worried that a narrative is being created.

Cheers in advance.

I was talking on the full P-Code list. GC used to be the highest in previous months. GC has often tweeted that trains are cancelled "due to a fault on this train", sadly GC deletes tweets after 24 hours so it's not possible to quote or share a link.

 

[CyrusWuff]

Ideally the figures should be weighted according to the number of services operated, otherwise the likes of Caledonian Sleeper, Grand Central, Hull Trains and Lumo look a lot worse than Northern given that cancelling two trains out of ten looks a lot worse than 100 out of 2300 percentage wise.

 

[Falcon1200]

I was under the impression no-one gave a monkeys why trains were P-coded anyway, so I'm interested that this data exists in the first place.

There is one overriding reason why train cancellations are P-coded; It improves the performance figures!

If control staff do not understand the consequences (and think they are running trains to make railways life easy, rather than stopping passengers getting to paid for events on time), then there is a serious training gap.

As a (former) Controller responsible for both agreeing P-coded timetable alterations (requested by a TOC) and initiating them myself (for Network Rail [NR] Incidents) Control staff are well aware of the implications, for passengers, of simply removing trains from the applicable timetable the night before. However due to the serious performance implications of not doing so the pressure was always on from opposite (TOCs) and above (NR) to ensure that this was done, regardless (for the latter in particular) of how busy Control might be dealing with the incident necessitating the alterations in the first place.

Personally I was never happy with the practice, IMHO P-coded changes should only apply from Day 3, eg agreed by 2200 on a Monday to apply from the Wednesday onwards, not Tuesday; So that the altered timetable could be properly advertised to passengers. But I was not in a position to change, or ignore, the process!

 

[Horizon22]

Ultimately it's likely impossible to know for definite a breakdown of the core reasons behind each operator's P-Coding, because "PG" is a delay code in itself, and hence it 'covers up' the 'real' reason for the cancellation. There might be some underlying data, but I imagine this is hit and miss.

Personally I was never happy with the practice, IMHO P-coded changes should only apply from Day 3, eg agreed by 2200 on a Monday to apply from the Wednesday onwards, not Tuesday; So that the altered timetable could be properly advertised to passengers. But I was not in a position to change, or ignore, the process!

It is useful to be used, especialy when you may otherwise end up with scores of "cancelled" and "delayed" during serious incidents for something known to impact multiple daily timetables. But it is being grossly overused by some operators.

 

[Bertie the bus]

P-codes were originally intended for exceptional events, floods, snowdrifts, fire or bomb evacuations closing lines etc. and allowed items beyond railways normal control to be excluded from cancellation statistics.

Fairly similar to Do Not Travel notifications and now they are issued because Christmas Eve and New Year's Eve fall on a Sunday.

The railway has a never ending capacity to take something that has a genuine and useful purpose and abuse it for its own gain.

 

[jayah]

Ultimately it's likely impossible to know for definite a breakdown of the core reasons behind each operator's P-Coding, because "PG" is a delay code in itself, and hence it 'covers up' the 'real' reason for the cancellation. There might be some underlying data, but I imagine this is hit and miss.



It is useful to be used, especialy when you may otherwise end up with scores of "cancelled" and "delayed" during serious incidents for something known to impact multiple daily timetables. But it is being grossly overused by some operators.

Whatever happened to the earlier practice of introducing a revised permanent timetable for a period of weeks / months until the shortage can be rectified?

 

[AndrewE]

Fairly similar to Do Not Travel notifications and now they are issued because Christmas Eve and New Year's Eve fall on a Sunday.

The railway has a never ending capacity to take something that has a genuine and useful purpose and abuse it for its own gain.

Not "the railway" but the smaller units looking at bottom line, regulator sanctions or just published statistics. I hate management accounting!
Just make the remit "to run the railway in the interests of passengers!"

 

[Horizon22]

Whatever happened to the earlier practice of introducing a revised permanent timetable for a period of weeks / months until the shortage can be rectified?

Well that's long term planning, as opposed to short-term cancellations. Also creating a whole bunch of new diagrams / rosters & evenly spaced timetable is no easy task.

I don't know but is someone like Northern or TPE cancelling the same train at the same time, every single day? Or is it more likely it depends on the shortage of crew on final rosters each day/week and once spares have been marked up? I'd have a hunch its the latter.

To be honest it doesn't matter; a cancelled train should be cancelled with whatever reason it is decided to be and shown on journey planners, not unceremoniously removed overnight (with the exception of such incidents P-coding was meant for!).

 

[800001]

P-codes were originally intended for exceptional events, floods, snowdrifts, fire or bomb evacuations closing lines etc. and allowed items beyond railways normal control to be excluded from cancellation statistics.

So you were right, no one cared, but then some Operators saw it as a way to manipulate statistics so regulator now wants to know why as trying to clamp down on those who don't provide enough staff, or make enough trains serviceable.

Why are statistics now accurately required , it's because if a train is in timetable 12 weeks before, and tickets are sold, it disrupts the person with a ticket. Regardless of if cancelled month before, week before, day before or hour before, all of them disrupt their travel plans and any related consequences eg events or meetings being attended. Passengers can't ask a football match or concert to change the time.

If control staff do not understand the consequences (and think they are running trains to make railways life easy, rather than stopping passengers getting to paid for events on time), then there is a serious training gap.

It is not control staff who decide to Pcode, the instruction comes from further up the chain.

Yes there will be some control staff who do not think of the customers, but 99% of control staff are there to do the job, which is getting people A-B on there booked train.

 

[infobleep]

It is not control staff who decide to Pcode, the instruction comes from further up the chain.

Yes there will be some control staff who do not think of the customers, but 99% of control staff are there to do the job, which is getting people A-B on there booked train.

So playing devil's advocate, are people higher up the chain thinking of customers?

 

[Falcon1200]

So playing devil's advocate, are people higher up the chain thinking of customers?

I will just say that if your work performance rating depends to a large extent on your department's performance figures, you will do everything (legally) possible to make those figures look as good as possible.....

 

[Snow1964]

I will just say that if your work performance rating depends to a large extent on your department's performance figures, you will do everything (legally) possible to make those figures look as good as possible.....

The infamous bad targets and unintended consequences.

Instead of setting targets that optimise the image, grow the business and customer income, someone even further up the chain sets a dumb target that can be manipulated to make someone look good instead, even if bad overall for business growth or profitability.

Business textbooks on bad business always mention this, too many poor quality middle managers competing with each other, rather than delivering business strategy

 

[jayah]

Well that's long term planning, as opposed to short-term cancellations. Also creating a whole bunch of new diagrams / rosters & evenly spaced timetable is no easy task.

I don't know but is someone like Northern or TPE cancelling the same train at the same time, every single day? Or is it more likely it depends on the shortage of crew on final rosters each day/week and once spares have been marked up? I'd have a hunch its the latter.

To be honest it doesn't matter; a cancelled train should be cancelled with whatever reason it is decided to be and shown on journey planners, not unceremoniously removed overnight (with the exception of such incidents P-coding was meant for!).

It isn't long term planning. In the past, revised timetables have been implemented in a short number of weeks and operated for much longer.

Instead we have the nonsense of mass P coding of services hours before they are due to run, while the long term timetable keeps getting increased.

This matters very much. Nobody knows what service will run until 10pm the day before. A reliable, yet reduced timetable, is infinitely better for all concerned.

 

[Horizon22]

It isn't long term planning. In the past, revised timetables have been implemented in a short number of weeks and operated for much longer.

Instead we have the nonsense of mass P coding of services hours before they are due to run, while the long term timetable keeps getting increased.

This matters very much. Nobody knows what service will run until 10pm the day before. A reliable, yet reduced timetable, is infinitely better for all concerned.

I'm sure a planner can tell me the exact distinction, but seeing as "MTP" doesn't exist, and its more than just a week it's neither here nor there. Also, there is still - as I said above - a lot to consider when trying to standardise an "emergency timetable".

That's because P-coding is being misused from its original purpose. Trains should just appear "cancelled" based on the LTP (long term plan) timetable.