P Coded Cancellations

Hopefully this is the right forum to ask this question in....

Do any of the open data sites allow users to check if a train has been P-Coded as a cancellation, or, indeed, to report on all P-Coded cancellations for a particular operator/at a particular station.

I have looked on various open data websites and get different information from each - and I have no problem with that.

Some examples I have of the same train:
RTT - https://www.realtimetrains.co.uk/service/gb-nr:P02792/2025-02-22/detailed - shows as P coded in detailed view. Has a 'plan' type of VCN
Recent Train Times - https://www.recenttraintimes.co.uk/Home/Service?ServiceId=130952852 - shows as a planned cancellation
Charlwood House - https://charlwoodhouse.co.uk/rail/livetrain/P02792/22/02/25 - shows as cancelled by VSTP process (analagous to RTTs VCN)

Am I right that the VSTP Cancellation is a P code cancellation, and that if Charlwood House shows that as being cancelled after 10pm the night before (as in this case), the train is in the timetable of the day and is treated the same as if a train fails on the day as fdar as passenger rights go?

RealTimeTrains detailed search Advanced Options allows you to display only CAN coded trains for a location and time. Those are the planned cancellations (including VCN).

Thanks - I'm not sure that quite covers what I'm looking for - it may be I need to be more specific in P-coding. For example this train (https://www.realtimetrains.co.uk/service/gb-nr:P01840/2025-02-23/detailed) is shown as CAN, but has been replaced with this train (https://www.realtimetrains.co.uk/service/gb-nr:F41049/2025-02-23/detailed). In that case I'm not so interested. It may just be PG cancellations I'm looking for.

The timetable for passenger information purposes reloads at 0300 every night. If a TOC wants to PG/P-code a service(s), they have to send the request to Network Rail by 2200 and it be processed by then. So I am not 100% sure how this would show on every 3rd party system as the data may be processed differently.

I'm not sure whether you can explicitly see in RTT whether it is a cancellation attributed to PG without going into every service individually, but perhaps you can run code to scrape that data. Bare in mind that the RTT reasons don't always match the official attribution codes perfectly.

rob.rjt said:

Thanks - I'm not sure that quite covers what I'm looking for - it may be I need to be more specific in P-coding. For example this train (https://www.realtimetrains.co.uk/service/gb-nr:P01840/2025-02-23/detailed) is shown as CAN, but has been replaced with this train (https://www.realtimetrains.co.uk/service/gb-nr:F41049/2025-02-23/detailed). In that case I'm not so interested. It may just be PG cancellations I'm looking for.

Click to expand...

These are PD cancellations, which is a different TOPS/Trust code - https://wiki.openraildata.com/index.php/Delay_Attribution_Guide

Also it depends on how the TOC manages the cancellation - a service can be deleted from the CIS, or just shown as "cancelled" via a Darwin-connected system and either/or could be applied even if it is a P-coded cancellation. Network Rail processing the P-coding, does not always update all customer information systems, just the running information. So this might also might mean you miss some with the PG code if the individual operator has put it down to "shortage of train crew" (TG / TB for example).

So may not be as simple as you think to navigate!

rob.rjt said:

Am I right that the VSTP Cancellation is a P code cancellation, and that if Charlwood House shows that as being cancelled after 10pm the night before (as in this case), the train is in the timetable of the day and is treated the same as if a train fails on the day as fdar as passenger rights go?

Click to expand...

When a schedule is cancelled does not determine whether something is eligible for P-coding. The key consideration is when the alterations are agreed between Network Rail and operator, and communicated to the travelling public. At the extreme end it is perfectly possible for a schedule to remain uncancelled and unreported in the system for 8 days, and P-coded at the end of the 8th day quite legitimately, which almost never happens. Neither of the two criteria I described above can be determined in any way through open data websites.

The coding of cancellations (including the concept of "plan of the day") is purely a contractual exercise, with no consideration given to customer experience as it was never designed for that purpose. That is until one day some bright spark thought they would write it into customer facing contracts, expecting those who purchased tickets in advance with a given itinerary in good faith to recheck their journey status no earlier than the night before. Then again it is hardly unusual over the years for different versions of franchise agreements and national rail contracts to directly contradict delay attribution principles, so an event eligible for P-coding may well become ineligible a day/week/month later, and vice versa.

If you want a definitive answer in relation to compensation I doubt anyone would be able to give you one on this forum. All I can say is from a quick look at the line up for that day and how various things are coded, I have doubts over the accuracy of some of them, but without knowing Elizabeth line's specific contractual terms with DfT/TfL or what happened on the day I cannot give you any guarantee.

The 22:00 time isn’t strictly correct in terms of passenger information.

For those channels to be updated for the following day, the data needs to be in the timetable CIF, the cut off for which is far earlier in the day, by 11:00 in the morning. There is an emergency process to force an interim CIF to some downstream systems eg Darwin (Day A for Day B process) but this is very rarely used.

Whilst pre 2200 will work in terms of TRUST and performance regime; it won’t for journey planners and the like.

Frontera2 said:

The 22:00 time isn’t strictly correct in terms of passenger information.

For those channels to be updated for the following day, the data needs to be in the timetable CIF, the cut off for which is far earlier in the day, by 11:00 in the morning. There is an emergency process to force an interim CIF to some downstream systems eg Darwin (Day A for Day B process) but this is very rarely used.

Whilst pre 2200 will work in terms of TRUST and performance regime; it won’t for journey planners and the like.

Click to expand...

It wouldn’t be done in the CIF file, hence the TOC removing things from the passenger information when the timetable rebuilds.

Horizon22 said:

It wouldn’t be done in the CIF file, hence the TOC removing things from the passenger information when the timetable rebuilds.

Click to expand...

Agreed - but a good information team would cancel it in advance of the overnight TT build anyway

Frontera2 said:

Agreed - but a good information team would cancel it in advance of the overnight TT build anyway

Click to expand...

To be technical they would probably 'delete' rather than 'cancel', but yes indeed they would!

We’re very much in the cancel camp, if it’s P-Coded the day before, customers still deserve to know that the train(s) is cancelled and not just vanish. Plus our Delay Repay wouldn’t work automatically for deletions and nor would our alerts etc.