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

[saismee]

Doesn't excuse how it took GTR nearly an hour to realise their driver won't turn up on time though!

Or at all... did they even schedule anyone?

 

[Failed Unit]

Or at all... did they even schedule anyone?

I must admit it felt very 2018.

At departure time, waiting a few minutes isn’t rare we just assumed were getting held for a Peterborough train.

Then the departure time was updated to 2243 (we assumed control had a plan, a driver was coming from somewhere either by tube or train). This kept getting pushed back by about 2 minutes until the 2250 arrival from Cambridge appeared, which we then hoped that after the driver had changed ends they would take out train. (Which was now “delayed”)

Just like 2018, I have no idea if GTR had a driver, knew who the proposed driver was or even where they where and how they would get to us. It is of course also possible that a driver change was planned on route and it is that driver that was “missing” unfortunately this still happens to frequently with GTR. It was a bad night for GTR with unrelated issues hitting Thameslink services as well. But it is in fact a bad week for GTR.

 

[MikeWM]

It's been over 4 weeks since I last posted the 4-week rolling figures - are we finally seeing improved punctuality of evening services?


(chart shows Kings Cross to Ely weekday evening services, with very poor within-5-minutes punctuality (best 58%, worst 25%) for all services during the evening peak).

So, on the whole, no, unless you're travelling after 2000. The 1739 has improved a bit, the 1809 has got worse, the 1912 is a disaster area.

 

[Magdalen Road]

This is the 1512 Kings Cross to Kings Lynn for rolling 12 weeks and 4 weeks (second image). Black squares are all cancelled services. Very poor.
Edit as requested to provide alt text (20:23 29/06/25)
Image is of a calendar represented by blocks of seven days arranged in stacks of weeks, for a 12 week period. The colours of the blocks range from green to red to black. There are three shades of green, one amber, one orange and two shades of red indication of degrees of lateness. Black is cancelled. The chart is for the last 12 weeks performance of the 1512 service from Kings Cross to Kings Lynn. The performance is 50% of the service ran. In other words bloody awful.
The chart can be reproduced by using the website on-time trains.co.uk

 

[saismee]

This is the 1512 Kings Cross to Kings Lynn for rolling 12 weeks and 4 weeks (second image). Black squares are all cancelled services. Very poor.

What website is this?
Also slight tangent, is there still a sign that says "Magdalen Road" at Watlington?

 

[Magdalen Road]

What website is this?
Also slight tangent, is there still a sign that says "Magdalen Road" at Watlington?

Ontimetrains.co.uk
Yes there still is, I think on the signal box. The remnants of station and platform were fairly recently removed / renovated.

 

[MikeWM]

This is the 1512 Kings Cross to Kings Lynn for rolling 12 weeks and 4 weeks (second image). Black squares are all cancelled services. Very poor.

Which should form the 1710 KLN to KGX, which is why *that* service is currently cancelled more than 50% of the time (which, as I've pointed out many times recently, leaves almost an hour-long gap in service from Cambridge North to Cambridge, right in the middle of the evening peak...)

 

[jdp30]

I suspect a many would be quite happy for the service to run to/from King’s Cross on a permanent basis. It’s noteworthy that we were told on here that diverting to/from KX would be a simple matter during disruption. So either that was incorrect, arrangements have changed, or perhaps it was a lie…

100% - I'm in one of the FB rail user groups for a station between Hitchin and Peterborough and there's constant calls to just go back to KGX services.
I'm a huge fan of the ECML TL program but they don't help themselves - On Saturday a points failure at East Croydon led to no southbound services for several hours on the PBO branch.

Realtime Trains | Departures from Sandy around 2300 on 14/06/2025

Train information at Sandy around 2300 on 14/06/2025. From Realtime Trains, an independent source of train running info for Great Britain.

www.realtimetrains.co.uk

Looks like there was nothing southbound between 18:25 and 22:55?

 

[Taunton]

There are plenty of instances of terminating northbound trains at Letchworth and southbound trains at Three Bridges.

But reversals at Letchworth or Three Bridges can't work if there are traincrew changes at Cambridge or Brighton.

One of the fundamental issues with services being joined together like Thameslink is that previously, with separate Kings Cross to Cambridge and London Bridge to Brighton, you could get two, and possibly even three, round trips in a shift, efficiently. Now, with the long run taking 2.5 hours each way, you can't get more than one round trip per shift. Sure you can fill in with an odd more local run, but that introduces any disruption elsewhere. Or you can crew change en route, which rapidly turns the service to maggots even faster if there is any disruption.

 

[GN Boy]

100% - I'm in one of the FB rail user groups for a station between Hitchin and Peterborough and there's constant calls to just go back to KGX services.
I'm a huge fan of the ECML TL program but they don't help themselves - On Saturday a points failure at East Croydon led to no southbound services for several hours on the PBO branch.

https://www.realtimetrains.co.uk/se...025-06-14/2300?stp=WVS&show=call&order=actual

Looks like there was nothing southbound between 18:25 and 22:55?

A conversation that semi-frequency crops up is why not reinstate the stoppers from KGX to PBO, perhaps replacing the current Letchworth terminators. I’m sure there are plenty of logistical reasons why this is unfeasible but it would aid passengers and staff alike.

 

[bramling]

A conversation that semi-frequency crops up is why not reinstate the stoppers from KGX to PBO, perhaps replacing the current Letchworth terminators. I’m sure there are plenty of logistical reasons why this is unfeasible but it would aid passengers and staff alike.

I’d just replace one of the hourly Horsham to Peterborough services with a King’s Cross service, which would address most of the difficulties. We will be told there isn’t platform space at King’s Cross, but that’s a choice not a certainty.

== Doublepost prevention - post automatically merged: 19 Jun 2025 ==

I suspect the large number of passengers travelling ECML and Horsham to core stations would think differenty.

I’m not sure they would, especially once they’ve suffered a few occasions (as most will) having turned up at somewhere like Sandy and found there’s no trains for 3 hours, or perhaps found no train to get them home - especially now the evening King’s Cross fast services no longer run.

We’re now seven years in and still these services can’t seem to be made to work reliably.

 

[GordonT]

Do the services from Kings Cross whose first stop is Cambridge (and Cambridge non-stop to Kings Cross) normally achieve an entirely non stop journey in practice?

 

[jdp30]

I suspect the large number of passengers travelling ECML and Horsham to core stations would think differenty.

I’m not sure they would, especially once they’ve suffered a few occasions (as most will) having turned up at somewhere like Sandy and found there’s no trains for 3 hours, or perhaps found no train to get them home - especially now the evening King’s Cross fast services no longer run.

We’re now seven years in and still these services can’t seem to be made to work reliably.

Vast majority would prefer to go back I'd wager - PBO branch just gets screwed over. A reliable train to Stevenage or Finsbury park gives most of the benefit of the TL program anyway in exchange for a 5-15 minute same platform connection to the Core. How many 10 minute connections do you need to reach the same average journey as regular hour long delays?

Population in some of these towns is/has already been increasing rapidly, we should be talking about an extra 1tph to King's Cross, not 4.5 hours with no service because of an issue ~50 miles away...

 

[Horizon22]

One of the fundamental issues with services being joined together like Thameslink is that previously, with separate Kings Cross to Cambridge and London Bridge to Brighton, you could get two, and possibly even three, round trips in a shift, efficiently. Now, with the long run taking 2.5 hours each way, you can't get more than one round trip per shift. Sure you can fill in with an odd more local run, but that introduces any disruption elsewhere. Or you can crew change en route, which rapidly turns the service to maggots even faster if there is any disruption.

The only way around this in disruption is "stepping up" to later departures which limits your time at the (new) short-of-destination terminal (in this case Kings Cross). Of course this leads to crew displacement.

== Doublepost prevention - post automatically merged: 19 Jun 2025 ==

Population in some of these towns is/has already been increasing rapidly, we should be talking about an extra 1tph to King's Cross, not 4.5 hours with no service because of an issue ~50 miles away...

An unavoidable issue where you have services running though a "core". You impact one end of a branch service and the other end gets shafted. The service allows less platform usage at London terminals in normal operation (benefitting long-distance operators too), but in disruption the impacts are widespread. The same issue occurs with the Elizabeth line at say Canary Wharf or Abbey Wood which has a very limited service because a set of points have gone wrong at Slough. Of course Thameslink has a lot more branches, so the impacts can be worse.

 

[WeGoAgain]

Do the services from Kings Cross whose first stop is Cambridge (and Cambridge non-stop to Kings Cross) normally achieve an entirely non stop journey in practice?

Yep, but they don't tend to run at peak times, so a lot less rail 'traffic' about.

 

[Failed Unit]

It's been over 4 weeks since I last posted the 4-week rolling figures - are we finally seeing improved punctuality of evening services?

View attachment 182200

So, on the whole, no, unless you're travelling after 2000. The 1739 has improved a bit, the 1809 has got worse, the 1912 is a disaster area.

Depends how you look at it. some of the “2C” services 25% is double what they manage. Let’s see if December improves things.

 

[MikeWM]

Depends how you look at it. some of the “2C” services 25% is double what they manage. Let’s see if December improves things.

I fear it is going to be even worse than now in the down direction if they stick to the pattern of the proposed timetables - which have the fast services almost catching up to the semi-fasts between Hitchin and Cambridge, and the semi-fasts almost catching up to the stoppers - to the degree that the semi-fast is due to arrive at Cambridge just 4 minutes after the stopper, and then the fast just 4 minutes after the semi-fast. It looks almost designed to cause continual delays to the fasts.

 

[jdp30]

An unavoidable issue where you have services running though a "core". You impact one end of a branch service and the other end gets shafted. The service allows less platform usage at London terminals in normal operation (benefitting long-distance operators too), but in disruption the impacts are widespread. The same issue occurs with the Elizabeth line at say Canary Wharf or Abbey Wood which has a very limited service because a set of points have gone wrong at Slough. Of course Thameslink has a lot more branches, so the impacts can be worse.

I understand this, and I'd agree with you for the first hour or so at most in this case, but it wasn't exactly a total lost of power in the core or something disastrous to providing services north of the river. Hard to describe 4 hours with no southbound trains between Peterborough and Hitchin caused by a points failure at ECR as "unavoidable".

As mention in the thread we were told KGX diversions would happen smoothly as required, but they don't seem to.

It's a relatively routine type of failure miles south of the river, why was there seemingly no attempt to set up even a shuttle to Stevenage?

 

[bramling]

I fear it is going to be even worse than now in the down direction if they stick to the pattern of the proposed timetables - which have the fast services almost catching up to the semi-fasts between Hitchin and Cambridge, and the semi-fasts almost catching up to the stoppers - to the degree that the semi-fast is due to arrive at Cambridge just 4 minutes after the stopper, and then the fast just 4 minutes after the semi-fast. It looks almost designed to cause continual delays to the fasts.

The extension of the fast services to Cambridge is another farce in the making. It means local passengers have to change at Letchworth, whilst taking away trains which give people from stations further in the chance of a seat. And it will be problematic from a performance point of view.

== Doublepost prevention - post automatically merged: 19 Jun 2025 ==

I understand this, and I'd agree with you for the first hour or so at most in this case, but it wasn't exactly a total lost of power in the core or something disastrous to providing services north of the river. Hard to describe 4 hours with no southbound trains between Peterborough and Hitchin caused by a points failure at ECR as "unavoidable".

As mention in the thread we were told KGX diversions would happen smoothly as required, but they don't seem to.

It's a relatively routine type of failure miles south of the river, why was there seemingly no attempt to set up even a shuttle to Stevenage?

This is the thing, we are 7 years in, and this stuff is still happening. Four hour gaps to destinations when something goes wrong, and often failing to recover the the rest of the day.

 

[Horizon22]

I understand this, and I'd agree with you for the first hour or so at most in this case, but it wasn't exactly a total lost of power in the core or something disastrous to providing services north of the river. Hard to describe 4 hours with no southbound trains between Peterborough and Hitchin caused by a points failure at ECR as "unavoidable".

As mention in the thread we were told KGX diversions would happen smoothly as required, but they don't seem to.

It's a relatively routine type of failure miles south of the river, why was there seemingly no attempt to set up even a shuttle to Stevenage?

Capacity at Kings Cross probably has something to do with it - you divert a SB train in, and it needs to form a NB departure. If it was the booked train, it would be sitting there 2-3 hours. And by then the next Southbound train is in etc etc.

So you would have to run the train back NB with some sort of amended schedule. This would rely on a whole host of factors including whether the driver is able to do it (if they are based at a depot south of the river, they may be out of hours) and that time of day, spare driver availability etc. If a bunch of drivers are stuck south of the river, this might not always be the easiest to facilitate. Driver displacement becomes a real issue. Network Rail will understandably be concerned keep trains in the platform for an indefinite amount of time give the number of long-distance operations.

You can potentially run the trains, but building a plan on the fly whilst dealing with the rest of it is never easy. There are contigency plans for this purpose, but they rarely take into account crew and fleet issues because it is exceptionally difficult to build that in. Shuttles are often part of contingency plans but you do still run into the resourcing issues and they can take time to set up. Thameslink also operates across several NR routes, which no doubt adds a little bit of difficulty to the matter too. If it's a "routine failure", before going full on into a shuttle plan, you'd want an estimate for a line closure. If it's just an hour or so, you wouldn't bother because it would take just as long (if not longer) to get a reliable shuttle going.

You've only got 1 or 2 controllers trying to deal with all of this, so it's not always as simple as it seems. Working on building some "off the shelf plan" for Kings Cross diversions is probably helpful, but would need considerable work. You also have the issue on the Southern side where there aren't loads of suitable turning points - London Bridge doesn't have much capacity, you might have one bay at Blackfriars but probably not and otherwise you're up to Kentish Town.

That Thameslink has so many extensive branches, many of which are linked to one and other ultimately makes the recovery harder. Adding a ECML connection was always a risk.

 

[choochoochoo]

This weekend will be interesting. Hot weather. Shortage of drivers. Hopefully there's no big events on.

Although Sunday the Cambridge-Brighton's are only going as far as kings cross.

 

[Horizon22]

This is the thing, we are 7 years in, and this stuff is still happening. Four hour gaps to destinations when something goes wrong, and often failing to recover the the rest of the day.

The suggestion could be made that if it still going on after 7 years, it is almost impossible to recover with the constraints given.

 

[Failed Unit]

Capacity at Kings Cross probably has something to do with it - you divert a SB train in, and it needs to form a NB departure. If it was the booked train, it would be sitting there 2-3 hours. And by then the next Southbound train is in etc etc.

So you would have to run the train back NB with some sort of amended schedule. This would rely on a whole host of factors including whether the driver is able to do it (if they are based at a depot south of the river, they may be out of hours) and that time of day, spare driver availability etc. If a bunch of drivers are stuck south of the river, this might not always be the easiest to facilitate. Driver displacement becomes a real issue. Network Rail will understandably be concerned keep trains in the platform for an indefinite amount of time give the number of long-distance operations.

You can potentially run the trains, but building a plan on the fly whilst dealing with the rest of it is never easy. There are contigency plans for this purpose, but they rarely take into account crew and fleet issues because it is exceptionally difficult to build that in. Shuttles are often part of contingency plans but you do still run into the resourcing issues and they can take time to set up. Thameslink also operates across several NR routes, which no doubt adds a little bit of difficulty to the matter too. If it's a "routine failure", before going full on into a shuttle plan, you'd want an estimate for a line closure. If it's just an hour or so, you wouldn't bother because it would take just as long (if not longer) to get a reliable shuttle going.

You've only got 1 or 2 controllers trying to deal with all of this, so it's not always as simple as it seems. Working on building some "off the shelf plan" for Kings Cross diversions is probably helpful, but would need considerable work. You also have the issue on the Southern side where there aren't loads of suitable turning points - London Bridge doesn't have much capacity, you might have one bay at Blackfriars but probably not and otherwise you're up to Kentish Town.

That Thameslink has so many extensive branches, many of which are linked to one and other ultimately makes the recovery harder. Adding a ECML connection was always a risk.

That is interesting. They can do it instantly when the Moorgate branch dies (although no issues with which platforms they can use) can at 717 fit in front of a 5 car 8xx?

I guess it is the just continues its diagram exactly,

 

[bramling]

Capacity at Kings Cross probably has something to do with it - you divert a SB train in, and it needs to form a NB departure. If it was the booked train, it would be sitting there 2-3 hours. And by then the next Southbound train is in etc etc.

So you would have to run the train back NB with some sort of amended schedule. This would rely on a whole host of factors including whether the driver is able to do it (if they are based at a depot south of the river, they may be out of hours) and that time of day, spare driver availability etc. If a bunch of drivers are stuck south of the river, this might not always be the easiest to facilitate. Driver displacement becomes a real issue. Network Rail will understandably be concerned keep trains in the platform for an indefinite amount of time give the number of long-distance operations.

You can potentially run the trains, but building a plan on the fly whilst dealing with the rest of it is never easy. There are contigency plans for this purpose, but they rarely take into account crew and fleet issues because it is exceptionally difficult to build that in. Shuttles are often part of contingency plans but you do still run into the resourcing issues and they can take time to set up. Thameslink also operates across several NR routes, which no doubt adds a little bit of difficulty to the matter too. If it's a "routine failure", before going full on into a shuttle plan, you'd want an estimate for a line closure. If it's just an hour or so, you wouldn't bother because it would take just as long (if not longer) to get a reliable shuttle going.

You've only got 1 or 2 controllers trying to deal with all of this, so it's not always as simple as it seems. Working on building some "off the shelf plan" for Kings Cross diversions is probably helpful, but would need considerable work. You also have the issue on the Southern side where there aren't loads of suitable turning points - London Bridge doesn't have much capacity, you might have one bay at Blackfriars but probably not and otherwise you're up to Kentish Town.

That Thameslink has so many extensive branches, many of which are linked to one and other ultimately makes the recovery harder. Adding a ECML connection was always a risk.

So that summarises as “Thameslink cannot operate reliably”, which was the concern of many before May 18. GTR’s poor control seems to exacerbate things further.

 

[Horizon22]

That is interesting. They can do it instantly when the Moorgate branch dies (although no issues with which platforms they can use) can at 717 fit in front of a 5 car 8xx?

I guess it is the just continues its diagram exactly,

Yes I imagine the slight turnaround time difference presents no real issues at Kings Cross vice Moorgate. Or it's just held outside for a few minutes.

 

[bramling]

That is interesting. They can do it instantly when the Moorgate branch dies (although no issues with which platforms they can use) can at 717 fit in front of a 5 car 8xx?

I guess it is the just continues its diagram exactly,

Yes that’s exactly it, a Moorgate train generally doesn’t have a crew relief south of Finsbury Park, and the running time to King’s Cross isn’t wildly different.

What continues to really annoy me is how we were told on here pre-2018 how easy it would be to divert trains to King’s Cross, yet it clearly isn’t.

 

[Horizon22]

So that summarises as “Thameslink cannot operate reliably”, which was the concern of many before May 18. GTR’s poor control seems to exacerbate things further.

It operates reliably most of the time, so I would disagree. The issue is disruptive events can be more disruptive, or more accurately, more widespread than they were before.

== Doublepost prevention - post automatically merged: 19 Jun 2025 ==

What continues to really annoy me is how we were told on here pre-2018 how easy it would be to divert trains to King’s Cross, yet it clearly isn’t.

Which authority said this? GTR, or forum users here? Because practically it is easy to divert a train or two to Kings Cross yes, but formulating a robust service plan with ongoing diversions for several hours is a totally different matter.

 

[Failed Unit]

Yes that’s exactly it, a Moorgate train generally doesn’t have a crew relief south of Finsbury Park, and the running time to King’s Cross isn’t wildly different.

What continues to really annoy me is how we were told on here pre-2018 how easy it would be to divert trains to King’s Cross, yet it clearly isn’t.

Oh yes I remember that former forum member that could never be wrong.

Some pairs seem to work. The peak only Sevenoaks - WGC is often dumped at Blackfriars. However I have known an incident at Kentish Town destroy the services as the Bedford trains block the platforms so the ones heading to Finsbury Park just end up sandwiched between them.

However even if Kings Cross isn’t used. Terminating at Finsbury back and sending Hornsey works. I have seen them dump trains at WGC before now (no idea what the did with the train next, probably empty to Hornsey) However at least they could use other trains to at least get to the capital. The MML side has being rubbish this week as well, it doesn’t Recover either speaking to people that live on it.

 

[Horizon22]

However even if Kings Cross isn’t used. Terminating at Finsbury back and sending Hornsey works. I have seen them dump trains at WGC before now (no idea what the did with the train next, probably empty to Hornsey) However at least they could use other trains to at least get to the capital. The MML side has being rubbish this week as well, it doesn’t Recover either speaking to people that live on it.

Yes terminating and back to depot at least clears the line. However that evidently will lead to large gaps in the opposite direction later on.

 

[bramling]

It operates reliably most of the time, so I would disagree. The issue is disruptive events can be more disruptive, or more accurately, more widespread than they were before.

Yes that's the problem - there's more chance of a disruption, and significant difficulty in recovering from these disruptions.

Which authority said this? GTR, or forum users here? Because practically it is easy to divert a train or two to Kings Cross yes, but formulating a robust service plan with ongoing diversions for several hours is a totally different matter.

Someone who used to post here, and I believe was, or is, a Fleet Control Manager for GTR.