Experiences of service recovery (or lack of) on GTR Great Northern

[bramling]

The Fen Line certainly appears to be suffering.

Welwyn Garden City south seems to be better, but can’t recover even small delays, they are not snowballing as much. Morning only

My unscientific view.
X23 and x53 Moorgate - They are on platform 4 and generally depart on time now they no longer have a conflict with the arriving London service,
x00 and x30 Sevenoaks - They are generally on platform 3, so no longer have to wait for the Cambridge service
x08 and x38 Kings Cross - Tend to be late, probably Welwyn North Congestion
x10 and x40 Moorgate - Tend to be late because of the Kings Cross service, sometimes gets let out first and delays the Kings Cross service even more,
x10 and x40 Cambridge 0 Does seem to vary but not conflicts at Welwyn Garden City,

I have notice not many people use the King Cross service now, probably down to it only being 7 mins behind the Sevenoaks, more likely to be a 700 (so you don’t get the less cramped seating arrangement) and the time difference is not that great. Must be frustrating for the poor cafe owned at the station.

On the latter point, I’d imagine many Welwyn GC passengers will prefer the Sevenoaks service because it starts at Welwyn, so they know they will have not just a seat, but be able to have pick of the seats. Never underestimate how valuable this is! I bet there’s even a few who turn up very early to get their favourite seat in declassified first.

 

[Failed Unit]

On the latter point, I’d imagine many Welwyn GC passengers will prefer the Sevenoaks service because it starts at Welwyn, so they know they will have not just a seat, but be able to have pick of the seats. Never underestimate how valuable this is!

I have used the 0800 a few times, it is normally standing before New Southgate. Even stations such as Hatfield and Potters Bar are not going to wait any more for the following Kings Cross service.

Heading North I still prefer Kings Cross for the same reason. (Although at the time I travel, it isn’t a 700 which is doubles the attractiveness.

Shame the 1827 can’t run because of congestion at Welwyn North, although I can’t see why they can’t just run it to Welwyn Garden City.

In the old timetable, although the Sevenoaks services started at WGC - the station had no advantage to other stations on the line as apart from the 0732, they service had to wait for the Cambridge train to clear the signal section.

 

[philjo]

I wonder if one longer-term solution might have to be a bay platform somewhere on the Cambridge section. We seem to have reached a point where the Shepreth BJ to Cambridge section, and Cambridge station itself, is now a significant additional bottleneck. Letchworth is the only place where there is space, although it doesn’t help the built lift towers in the space.

I attended a joint meeting of the rail users groups for the Cambridge branch stations in 2018/2019 when it was mentioned of the possibility of converting the down side siding at Royston into a bay platform to allow up trains to start from there just after an up fast from Cambridge had been through. Part of the Car park would need converting to a platform but it seemed more feasible than other locations suggested such as a 3rd platform at Meldreth which would require the purchase of land.

It is a pity that the lifts at Letchworth were built in their current position and blocking the potential to use additional platforms as the station built in 1912 was designed to have 4 platforms. Terminating trains could then use the other platform and possibly turn back from there. I think there were structural issues with the original lift shafts preventing their reuse or allowing new lifts to be added alongside it.

 

[Magdalia]

It does feel like the original May 18 timetable seemed to work better

It is too early to tell, there has only been 5 days of weekday service with the revised track and signalling. I'm confident that the new layout at Hills Road will make a difference for down trains approaching Cambridge, because up departures from platform 8 no longer conflict with down arrivals into platform 7 or into platform 4 via the through road.

I wonder if one longer-term solution might have to be a bay platform somewhere on the Cambridge section. We seem to have reached a point where the Shepreth BJ to Cambridge section, and Cambridge station itself, is now a significant additional bottleneck. Letchworth is the only place where there is space, although it doesn’t help the built lift towers in the space.

I don't see any of this as being feasible or desirable. The southern approach to Cambridge has been a bottleneck since 2018, trains queueing to get into Cambridge in the evening peak isn't new. The original Cambridge South proposal included 4 tracks all the way between Shepreth Branch Junction and Cambridge, and hopefully we will still get that eventually as part of East West Rail. But since 5 January we have the new signalling and that should help to manage any queue of trains.

That is not to say it will get better of course, but obviously GTR performance teams will be noticing this discrepancy between the booked time to complete a split and the actual time taken - which as I said is sometimes 100% worse - and want to know what exactly is going on. It could be a range of factors from incorrect CIS/PIS causing passenger confusion, technical issues completing the split, lack of driver familarity etc.

What has changed in 2025 is the time taken to split trains at Cambridge. This isn't a timetable thing it is a class 379 thing.

What has also changed is the time taken to clear terminating down trains out of the platform at Letchworth. From the few times I have watched this recently, it needs more staff, one person for an 8 car train takes too long. This may indirectly be a train thing too. In the days of class 365s and class 317s the driver used to do the front half of the train; that doesn't seem to happen now, from what I have seen the driver stays in the cab and the platform staff have to do all of the train. so it takes twice as long.

And of course platform overtime is a significant cause of queues building outside both Cambridge and Letchworth. If the platforms are still occupied with overtime for splitting or clearing out then the next train can't get in.

 

[Magdalia]

It is a pity that the lifts at Letchworth were built in their current position and blocking the potential to use additional platforms as the station built in 1912 was designed to have 4 platforms.

Since Letchworth platforms were lengthened for 12 car trains additional platforms are no longer feasible.

 

[MikeWM]

9, 6, 6 for splitting tonight. I'm still not at all sure why splitting 379s is any different from splitting 387s, which they seemed to manage to do most of the time in 5 minutes, but hopefuilly they'll work it out eventually.

Today was probably the most 'average' day ever - no issues on the line, no recovery needed from earlier incidents, no extremes in temperature, no strong winds or driving snow or heavy rain or hot sun. Yet once again none of the main five evening services (1T48, 1T50, 1T52, 1T54, 1T56) got to Ely on time, indeed only one (1T56) got here within 5 minutes of its scheduled arrival.

 

[jon0844]

It could be a range of factors from incorrect CIS/PIS causing passenger confusion, technical issues completing the split, lack of driver familarity etc.

I have heard that Cambridge staff have acknowledged there is a problem with their CIS, relating to explaining the splits and joins. They are recommending more announcements on the platform and by train crew to aid the process. I have not heard how their system announces such things compared to systems using the Infotec displays?

 

[jon0844]

What has also changed is the time taken to clear terminating down trains out of the platform at Letchworth. From the few times I have watched this recently, it needs more staff, one person for an 8 car train takes too long. This may indirectly be a train thing too. In the days of class 365s and class 317s the driver used to do the front half of the train; that doesn't seem to happen now, from what I have seen the driver stays in the cab and the platform staff have to do all of the train. so it takes twice as long.

And of course platform overtime is a significant cause of queues building outside both Cambridge and Letchworth. If the platforms are still occupied with overtime for splitting or clearing out then the next train can't get in.

Shutting down a 700 is dead easy. Quick to walk though, few hiding places, and you're on at one end and off the other.

With a 387 or 379, the need to exit the train causes delay and also introduces risk (if someone were to try and hide and walk back through the carriage to what has been locked off) which means you can't rush.

If you encounter someone who causes you hassle and refuses to comply to get off, it becomes even worse. Obviously the same people would create an issue with a 700 too, but at least staff can use the guard panel to notify the driver to ask for help notify the signaller etc.

Two people really should be terminating anything but 700s, and if there are known issues then have travel safe officers on hand to help get people off too.

 

[Failed Unit]

Today with have the classic example of the inability to recover delays.

0753 Welwyn Garden City - Moorgate arrives late.
0800 Welwyn Garden city - Sevenoaks waiting.

At 0800 the 0753 still isn’t ready to go. What to they do? Hold the 0800 which stops less frequently to follow it. This decision has now made at least 4 trains late. Not by much but the 0800 will miss its slot through the core now. Both following 0808 King’s Cross and 0810 Moorgate will also be stuck in the queue.

I know it is easy for us in the armchair to say “let the train with least stops go first”. I have no idea why it so rarely happens

 

[jammmmmsssssss]

I have heard that Cambridge staff have acknowledged there is a problem with their CIS, relating to explaining the splits and joins. They are recommending more announcements on the platform and by train crew to aid the process. I have not heard how their system announces such things compared to systems using the Infotec displays?

But splits were happening in a timely manner with the 387s for years and the delays only started when 379s started splitting

 

[jon0844]

But splits were happening in a timely manner with the 387s for years and the delays only started when 379s started splitting

Do the 379s have an issue with auto announcements on board? Or should that be, are they set up properly?

Perhaps they don't work quite the same as the 387s - or the method of activating the messages is different and drivers haven't been fully trained - and that's why Cambridge staff want better announcements made.

 

[MikeWM]

Do the 379s have an issue with auto announcements on board? Or should that be, are they set up properly?

Generally, the announcements are certainly different on the 379s - I'm fairly sure that there are more of them and that they announce the next stop while much further away.

I've not yet been on a splitting 379 so I'm not sure what is announced.

 

[jon0844]

Given how long the 379s were going around saying 'not in service' and other issues relating to the displays (especially on the outside) I so wonder if this is a factor.

377s and 387s seem pretty good at announcing splits and telling people what coach they are in*. I haven't been in a 379 that has split or joined yet.

* People do still get confused and not listen of course, but the more people that do understand and do what they need to do the better and quicker the overall process will be.

 

[jammmmmsssssss]

I've not yet been on a splitting 379 so I'm not sure what is announced.

I have. I was right at the front and didn't hear any announcements (except for the stops being listed about five times)

 

[MikeWM]

I have. I was right at the front and didn't hear any announcements (except for the stops being listed about five times)

Tonight's (1T48) has no automated announcements and no scrolling information *at all*, at least not in the front coach between Cambridge North and Ely, which was quite pleasant. The driver did make manual announcements.

Today's splitting times : 7, 7, 8.

Though today may be considered the first 'not totally atrocious day' since the timetable change. 1T54 got to Ely *on time* (!!!) and 1T52 was only 1 minute late. Really that shouldn't be an achievement, but given the last month it really is.

I still got home 10 late though as I went for 1T48, which (as above) still took too long to split at CBG, but the main issue was getting caught behind a late Thameslink service which picked up a delay somewhere around Purley, two hours earlier.

 

[Bikeman78]

I have heard that Cambridge staff have acknowledged there is a problem with their CIS, relating to explaining the splits and joins. They are recommending more announcements on the platform and by train crew to aid the process. I have not heard how their system announces such things compared to systems using the Infotec displays?

It looks like two of the three evening trains being discussed use platform 1 at Cambridge. So hardly anyone is going to try to board the unit that detaches because it is furthest from the entrance. It really shouldn't take very long to clear out the stragglers that boarded the wrong unit at Kings' Cross. I've seen a 10 car double decker split at Rotterdam way faster. Rear six cars terminated with staff boarding at the rear and middle of the unit to usher people off. The split took place after maybe two minutes, with the last few people emerging from the front door of the rear set a few seconds later.

 

[MikeWM]

Today's splitting times : 10. 11, 6.

Yes, it is probably getting a bit tedious my posting these every day so I'll probably stop for a bit. Though it is very tedious indeed getting to Cambridge North every day to see the 1822 listed as on time, then going to 'delayed', then showing up 5-6 minutes late. Every single night. It is really getting rather irritating, especially on these cold January evenings.

 

[Bikeman78]

Today's splitting times : 10. 11, 6.

Yes, it is probably getting a bit tedious my posting these every day so I'll probably stop for a bit. Though it is very tedious indeed getting to Cambridge North every day to see the 1822 listed as on time, then going to 'delayed', then showing up 5-6 minutes late. Every single night. It is really getting rather irritating, especially on these cold January evenings.

Not at all. Always good to have the figures written down. If people aren't interested, they don't have to read them. Has anyone watched the splitting process to see why it takes so long?

 

[Intercity_225]

How long do the joinings at Cambridge in the morning peak take?

 

[rmHawk765]

Has anyone watched the splitting process to see why it takes so long?

Judging by what has been said here, I'd imagine it might have something to do with the PIS seemingly making zero announcements regarding the splitting. That said the 387s have a more advanced TMS (the 379s being a downgrade in this department might be unusual to crews) which may play a part in it.

 

[MikeWM]

How long do the joinings at Cambridge in the morning peak take?

Looking at this morning on RTT, 5, 3 (!!) and 6.5 minutes.
Yesterday 6, 5, 5.

Which is - interesting. You'd expect splitting to be rather easier than joining, but it seems the opposite is currently true.

 

[RailwayRookie]

Looking at this morning on RTT, 5, 3 (!!) and 6.5 minutes.
Yesterday 6, 5, 5.

Can you give an example of how you extrapolate the times from RTT, not nitpicking (maybe later) just curious.

 

[MikeWM]

Can you give an example of how you extrapolate the times from RTT, not nitpicking (maybe later) just curious.

Just looking at the arrival and departure times at Cambridge for 1T07, 1T09 and 1T11 (morning) and 1T48, 1T50 and 1T52 (evening).

Yes there could be other factors at play, not just the time required for the actual split. But for 1T48 in particular, as I tend to be staring at it on Traksy as I'm standing in the cold at Cambridge North, it isn't that there are other services in the way so the delay isn't likely to be due to eg. being held at a signal while something else passes.

 

[RailwayRookie]

Just looking at the arrival and departure times at Cambridge for 1T07, 1T09 and 1T11 (morning) and 1T48, 1T50 and 1T52 (evening).

Yes there could be other factors at play, not just the time required for the actual split. But for 1T48 in particular, as I tend to be staring at it on Traksy as I'm standing in the cold at Cambridge North, it isn't that there are other services in the way so the delay isn't likely to be due to eg. being held at a signal while something else passes.

Thank you for the response. As you say other factors will also be at play but for an Apples to apples comparison thays the best method you will get short of being there.

Also of note is that the signal aspects shown are vague at best. I dont if its only since the resignalling but theres large discrepancies between whats on the map and on the group currently so thats not a great method either

 

[Intercity_225]

Looking at this morning on RTT, 5, 3 (!!) and 6.5 minutes.
Yesterday 6, 5, 5.

Which is - interesting. You'd expect splitting to be rather easier than joining, but it seems the opposite is currently true.

How intriguing! Are the splitting times as bad when the morning peak trains split at Kings Cross?

 

[saismee]

Also of note is that the signal aspects shown are vague at best. I dont if its only since the resignalling but theres large discrepancies between whats on the map and on the group currently so thats not a great method either

Depends on which map you use. OTT is entirely missing Cambridge South and is displaying Ely with all routes selected simultaneously, while also having gibberish signal states. Traksy doesn't show many aspects and I don't know if the ones displayed are correct. Signalmaps.co.uk seems to be the most accurate with signal states and layout, though it seems to be fussy on slow/inconsistent networks.

 

[bramling]

Realtime Trains | 2C65 2025 Cambridge to London Kings Cross | 15/01/2026

Real-time train running information for 2C65 2025 departure from Cambridge to London Kings Cross on 15/01/2026. From Realtime Trains, an independent source of train running info for Great Britain.

www.realtimetrains.co.uk

Another more medial one tonight.

2 late leaving Cambridge, lost a further 4 mins by Stevenage.

Now to be fair tonight is a horrible night so some of this will no doubt be due to defensive driving, don’t have an issue with this. However the train didn’t have Cambridge South to call at, had it done so then that it would have been a further couple of mins behind.

One wonders if the run times between Cambridge and Hitchin for the stopping services are sufficient. Clearly right-time presentation at Woolmer Green is very important. It is something of a shame we don’t seem to be able to achieve 1990s 317 times though.

 

[Magdalen Road]

Data from one annual season ticket's worth of commuting (Jan 25 - Jan 26).
111 commutes (outbound and return) Ely to King's Cross
56 outbound commutes delayed with a total cumulative delay of 533 minutes
37 return commutes delay with a total cumulative delay of 519 minutes. One of the returns was via sardine LNER to Peterborough and then sharing an Uber because it was unlikely multiple trains' worth of passengers would fit on the next EMR or XC.
A grand total of 17.5 hours of commuting delays.
So the return commutes are delayed less frequently but for longer when they are delayed.
Pretty poor, in my opinion.

 

[jammmmmsssssss]

Depends on which map you use. OTT is entirely missing Cambridge South and is displaying Ely with all routes selected simultaneously, while also having gibberish signal states. Traksy doesn't show many aspects and I don't know if the ones displayed are correct. Signalmaps.co.uk seems to be the most accurate with signal states and layout, though it seems to be fussy on slow/inconsistent networks.

Traksy and OTT's signal aspects in the Cambridge area are completely wrong at the moment. Signalmaps has been the only one that I've used for a while as it's updated way faster than the others after layouts change and it shows all of the signal aspects and routes set. It seems to require decent internet to get it open, but once it's open, it doesn't require that good of a connection

 

[bramling]

I see GN has gone sour again this evening, I believe a passenger incident of some kind at Welwyn North (not a fatality). Seems elements of the service are starting to disappear already, 45 mins in.

Looking at what’s going on at Welwyn GC at present, one change which would actually be quite useful nowadays would be crossovers to allow down trains to run direct into Welwyn platforms 1 and 2, giving some extra flexibility during disruption.