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

[Bald Rick]

Been off line for a bit.

It’s never just one train though, is it.

This example, which you specifically highlighted as incomptent, was ‘just one train” though. Had it been left to run late it would have caused delay to dozens of other trains, carrying thousands of passengers, but read on…

Firstly, it left Peterborough 18 late

Indeed. With the following service on time.

So intermediate passengers were delayed a further 12 minutes by this train being sent fast to Stevenage, than had it run booked calls.

Firstly, it left Peterborough 18 late, so given there’s

(1) 7 minutes worth of allowance within the journey

(2) Numerous places on the journey where there are long dwell times booked at stations ( 2.5 mins at Huntingdon, 1.5 mins at Stevenage, 1.5 mins at Finsbury Park, 1.5 mins at London Bridge, 2 mins at Coulsdon South, 3 mins at Redhill, just to name the obvious ones), there should have been opportunity for the train to have reached Horsham and be able to be more-or-less right time for its next journey.

But there’s no allowances between Peterboro and Hitchin. Assuming the Huntingdon call was done in 1 minute, and the others bang on (and I have never seen a Class 700 do a 30 second dwell time…) it would have been 16.5 mins late approaching Hitchin.

Which puts it behind the following Thameslink from Cambridge 9S33, and that puts it back to being (at least) 18 minutes late.

With 9J33 now following 9S33, we have two regulation decisions to make at Woolmer Green, with 1A63 and 1T31 approaching on the UF. In all likelihood 1A63 would go first, but 1T31 would wait (and be delayed). But waiting for 1A63 delays 9J33 further, however the allowances and a little catch up at Finsbury Park bring it back to 18L leaving there, still behind 9S33.

Unfortunately, That means it is going arrive at St P in a clash with 9P37 to Rainham, with the 9P marginally ahead, so that would go first. Another regulating decision to make. Then there is a 9R31 to Three Bridges right behind the 9P, another regulating decision to make. Hopefully the 9J goes first as that 9R is all shacks from East Croydon…

So 9J33 goes through the core, emerging from the shadow of 9P37 at Corbetts Lane Junction by now 20 mins late. It pulls back a few minutes through allowances to East Croydon, but has now caught up 1P33, which it follows to Purley, meaning it is back to 18L. It then follows 1R36 to Redhill, where it arrives, at best, 15 late (and in the path of 9R31, delaying that). That’s all assuming the Q pathed freight that is around st that time of day isn’t running today (it wasn’t). It then proceeds on to Horsham where it arrives 14-15 mins late.

It can’t be terminated at East Croydon, Redhill, Gatwick or Three Bridges as there are no spare platforms for it to hang around for in order for it to make a RT start back north on diagram, except at Three Bridges, but 9R31 hard on its tail lights needs that platform at Three Bridges as booked.

Its arrived at Horsham 15 late, and with a large slice of good will from the driver and slick working departs north as 9J48only 7 late. With some further time caught up (unfortunately delaying a couple of services), it arrives Blackfriars 4 minutes down, which is not great for all of the several hundred passengers on board (which there were, as I saw that very train tonight when it departed St Pancras on time to the second).

Unfortunately, in our alternative universe 9J48 is 4 minutes late at Blackfriars, which delays 9O50 to St Albans by 4 minutes, which delays 9T40 to Bedford by 4 minutes, which delays 1F60 to Sheffield and 1D61 to Nottingham each by 5 minutes. 1F60 then delays 1L50 to Stansted at Wigston, 1K75 to Crewe at Sheet Stores, 1V15 to New Street / Cardiff at Derby and so on.

I’d have probably done nothing and let it run.

And that would have been the wrong decision, as described above. Not only wrong, but it would have triggered many more decisions required on the service for the next 4-5 hours, any of which might have been got wrong (incompetent in your language) and all adding to the workload of control unnecessarily.

Second resort, turning it short at somewhere like Three Bridges or Crawley if the crewing arrangements allowed (inconveniencing less busy stations than Hitchin to Peterborough, and in most cases with more alternatives), or as a last resort run non-stop Peterborough to Hitchin but not omit Hitchin. All three of these options would have inconvenienced fewer people than what was done.

Incorrect. Running the train late would have caused substantially more delay to thousands of people, inconveniencing many, many more people than the few (as helpfully highlighted in the passenger data by Robski above) boarding between Huntingdon and Arlesey, who were delayed by an additional 12 minutes (totalling 30), and a total delay of 15 minutes for those boarding at Hitchin.

It wasn’t 2L at Hitchin, it passed pretty much on time, I know because I was on the platform and saw it come past.

And so we come to the real reason. Your individual journey was delayed by 15 minutes. As were those of some other passengers at Hitchin. But this was clearly the lesser of two evils compared to running it late.

 

[GordonT]

Given that Hitchin and Stevenage get a more frequent service than Arlesey, Biggleswade, Sandy, St Neots and Huntingdon, if skip stopping is needed to get back on track then would it be better to omit Hitchin and/or Stevenage?

The ability of time to be recovered or otherwise can depend significantly on the extent to which other trains are occupying the sections immediately in front of the late runner and potentially the time gained by skipping stop x can be lost again by being held outside the next stop y while the train in front serves it and then (metaphorically) running up its backside having skipped stop 'y'. (Edit - the previous post to this one crossed with mine and focusses on this in much better detail).

 

[Bald Rick]

And as far as I'm concerned "totality" means 100%, ( not 80%, nor 65%, nor 51%), or to put it another way, equal. So using the term "majority" doesn't cut it in my book. (British Rail had lots of failures but at least all fare paying passengers were treated the same. ie equal. Something that the modern railway is lacking IMHO).

So what would you have done to ensure absolute equality in this example?

 

[Steve Harris]

Yes and the thread said "public services need to best serve the majority of users".

And I used a publicly available definition which says public = totality (not majority, like you think it is) as a counter argument/point of view on what public means in the context of public service. If you've ever taken part in a debate, you will know that the opposing side can use a counter argument/view (which is what I did by quoting Wikipedia). If you don't agree with me that's fine, we can agree to disagree you know

 

[Horizon22]

And I used a publicly available definition which says public = totality (not majority, like you think it is) as a counter argument/point of view on what public means in the context of public service. If you've ever taken part in a debate, you will know that the opposing side can use a counter argument/view (which is what I did by quoting Wikipedia). If you don't agree with me that's fine, we can agree to disagree you know

I disagree with the premise. Public is totality under your definition sure and therefore the "majority of the public" is the "majority of the totality". Majorities are theoretically always based on >50% of the overall 100% cohort!

As an aside the totality of the population does not use rail transport - significantly less actually - so it doesn't quite work in this context.

 

[Hadders]

The ability of time to be recovered or otherwise can depend significantly on the extent to which other trains are occupying the sections immediately in front of the late runner and potentially the time gained by skipping stop x can be lost again by being held outside the next stop y while the train in front serves it and then (metaphorically) running up its backside having skipped stop 'y'. (Edit - the previous post to this one crossed with mine and focusses on this in much better detail).

Thanks to you and @Bald Rick for explaining. This is why I'm not a timetabling expert!

 

[bramling]

To me this looks like an excellent decision.

9J33 was an isolated incidence of a late start, with the following 9J35 running on time.

By running non-stop 9J33 covered almost 49 miles Peterborough-Stevenage in about 35 minutes for a start to stop speed in excess of 80 mph.

9J33 was on time through the core and all the way to Horsham.

A small number of people at intermediate stations lost a further 12 minutes having to wait for the on time 9J35.

Maybe you’re looking at this as glass half full whereas I’m seeing it as glass half empty. But I see it that people ended up being 30 minutes late reaching London when they could, and should, have only been 15 minutes late. A 30-minute delay is a lot worse than a 15-minute one.

There’s also the point that it was essentially an error to include Hitchin, when the train passed through virtually right time.

It is absolutely right that reliability and resilience in the GN has collapsed since May 2018, compared with what happened under WAGN and FCC. Much of this was predicted before 2018, especially on the Hitchin-Huntingdon section, but ignored.

And still seems to be ignored. The general flakiness of the service on the Cambridge side is an issue as well. It was said on here at the time that the combination of Hitchin-Cambridge and the single sections north of Littleport would prove to be a massive issue, and here we are.

 

[bramling]

But there’s no allowances between Peterboro and Hitchin. Assuming the Huntingdon call was done in 1 minute, and the others bang on (and I have never seen a Class 700 do a 30 second dwell time…) it would have been 16.5 mins late approaching Hitchin.
Which puts it behind the following Thameslink from Cambridge 9S33

As it happens, 9S33 was 4 minutes late at Hitchin, so that doesn’t necessarily follow (perhaps 9S33 should have been run non-stop from Cambridge to Hitchin as a result of that 4 minutes?).

With 9J33 now following 9S33, we have two regulation decisions to make at Woolmer Green, with 1A63 and 1T31 approaching on the UF. In all likelihood 1A63 would go first, but 1T31 would wait (and be delayed). But waiting for 1A63 delays 9J33 further, however the allowances and a little catch up at Finsbury Park bring it back to 18L leaving there, still behind 9S33.

Unfortunately, That means it is going arrive at St P in a clash with 9P37 to Rainham, with the 9P marginally ahead, so that would go first. Another regulating decision to make. Then there is a 9R31 to Three Bridges right behind the 9P, another regulating decision to make. Hopefully the 9J goes first as that 9R is all shacks from East Croydon…

So 9J33 goes through the core, emerging from the shadow of 9P37 at Corbetts Lane Junction by now 20 mins late. It pulls back a few minutes through allowances to East Croydon, but has now caught up 1P33, which it follows to Purley, meaning it is back to 18L. It then follows 1R36 to Redhill, where it arrives, at best, 15 late (and in the path of 9R31, delaying that). That’s all assuming the Q pathed freight that is around st that time of day isn’t running today (it wasn’t). It then proceeds on to Horsham where it arrives 14-15 mins late.

It can’t be terminated at East Croydon, Redhill, Gatwick or Three Bridges as there are no spare platforms for it to hang around for in order for it to make a RT start back north on diagram, except at Three Bridges, but 9R31 hard on its tail lights needs that platform at Three Bridges as booked.

Its arrived at Horsham 15 late, and with a large slice of good will from the driver and slick working

Which is another thing we don’t really seem to see on GTR. The rest of the above seems to read as a very good addendum to the post a little while ago where you implied that Thameslink can never run reliably.

departs north as 9J48only 7 late. With some further time caught up (unfortunately delaying a couple of services), it arrives Blackfriars 4 minutes down, which is not great for all of the several hundred passengers on board (which there were, as I saw that very train tonight when it departed St Pancras on time to the second).

Unfortunately, in our alternative universe 9J48 is 4 minutes late at Blackfriars, which delays 9O50 to St Albans by 4 minutes, which delays 9T40 to Bedford by 4 minutes, which delays 1F60 to Sheffield and 1D61 to Nottingham each by 5 minutes. 1F60 then delays 1L50 to Stansted at Wigston, 1K75 to Crewe at Sheet Stores,

So not content with transferring a delay from the GN side to the South Central, Thameslink has managed to transfer it to a third route as well.

And so we come to the real reason. Your individual journey was delayed by 15 minutes. As were those of some other passengers at Hitchin.

No it wasn’t, as I was waiting for 2R35. However plenty of people were on the platform at Hitchin waiting for the Horsham train, and were rather mystified and unimpresssd to see it sail through empty at the right time.

I can, just about, see a logic to running fast Peterborough to Hitchin, though I don’t agree with it, however skipping Hitchin would be viewed as an error of judgement by anyone who doesn’t view people attempting to use a train service with contempt.

 

[Horizon22]

As it happens, 9S33 was 4 minutes late at Hitchin, so that doesn’t necessarily follow (perhaps 9S33 should have been run non-stop from Cambridge to Hitchin as a result of that 4 minutes?).

5 minutes is normally the threshold a controller on a metro route would be roughly looking to “be aware” of an issue and 10 minutes for “intervention likely”.

So yes it does follow. I feel like you are deliberately trying to find fault in very detailed explanations & reasoning because you were unfortunately personally impacted.

Which is another thing we don’t really seem to see on GTR. The rest of the above seems to read as a very good addendum to the post a little while ago where you implied that Thameslink can never run reliably.

I don’t read it like that personally. Just that Thameslink runs through some of the most congested parts of the railway network. And back to the “it was better into Kings Cross”, some of that section north of Finsbury Park is included in that.

Being susceptible to potentially more sources of service disruption does not mean Thameslink “never runs reliably” as you so put it, which is provably untrue.

So not content with transferring a delay from the GN side to the South Central, Thameslink has managed to transfer it to a third route as well.

No it didn’t. That was the hypothetical in your “do nothing” scenario that you were suggesting.

The fact that the service alteration was made meant that none of these secondary impacts happened. And yes imparting disruption onto neighbouring routes is best avoided, so this sounds like you’re arguing against your own point!

No it wasn’t, as I was waiting for 2R35. However plenty of people were on the platform at Hitchin waiting for the Horsham train, and were rather mystified and unimpresssd to see it sail through empty at the right time.

I can, just about, see a logic to running fast Peterborough to Hitchin, though I don’t agree with it, however skipping Hitchin would be viewed as an error of judgement by anyone who doesn’t view people attempting to use a train service with contempt.

Okay so in reality this is arguing about the Hitchin stop, as opposed to the principle of the decision which you previously decried as “incompetent”?

Although now people are being treated with “contempt”? As I said earlier there are not always good options and you peak the ‘least bad’

 

[Magdalia]

I see it that people ended up being 30 minutes late reaching London when they could, and should, have only been 15 minutes late. A 30-minute delay is a lot worse than a 15-minute one.

Hitchin passengers were only 15 minutes late, by using 9S33.

As it happens, 9S33 was 4 minutes late at Hitchin, so that doesn’t necessarily follow (perhaps 9S33 should have been run non-stop from Cambridge to Hitchin as a result of that 4 minutes?).

Skip stopping on 9S33 was unnecessary: it was on time at London Bridge.

however skipping Hitchin would be viewed as an error of judgement by anyone who doesn’t view people attempting to use a train service with contempt.

The decision on which stops to skip had to be made before 9J33 departed from Peterborough, not as it was approaching Hitchin. The signallers needed the option of running 9J33 on the fast line, in order to maximise the opportunity to regain time. That's done by skipping all the stops to Stevenage, the only intermediate stations with a platform on the fast line are St Neots and Biggleswade. I note that none of your comments have specified whether 9J33 was on the fast line or the slow line when it passed through Hitchin.

 

[sharpener]

Been off line for a bit.



This example, which you specifically highlighted as incomptent, was ‘just one train” though. Had it been left to run late it would have caused delay to dozens of other trains, carrying thousands of passengers, but read on…




Indeed. With the following service on time.

So intermediate passengers were delayed a further 12 minutes by this train being sent fast to Stevenage, than had it run booked calls.




But there’s no allowances between Peterboro and Hitchin. Assuming the Huntingdon call was done in 1 minute, and the others bang on (and I have never seen a Class 700 do a 30 second dwell time…) it would have been 16.5 mins late approaching Hitchin.

Which puts it behind the following Thameslink from Cambridge 9S33, and that puts it back to being (at least) 18 minutes late.

With 9J33 now following 9S33, we have two regulation decisions to make at Woolmer Green, with 1A63 and 1T31 approaching on the UF. In all likelihood 1A63 would go first, but 1T31 would wait (and be delayed). But waiting for 1A63 delays 9J33 further, however the allowances and a little catch up at Finsbury Park bring it back to 18L leaving there, still behind 9S33.

Unfortunately, That means it is going arrive at St P in a clash with 9P37 to Rainham, with the 9P marginally ahead, so that would go first. Another regulating decision to make. Then there is a 9R31 to Three Bridges right behind the 9P, another regulating decision to make. Hopefully the 9J goes first as that 9R is all shacks from East Croydon…

So 9J33 goes through the core, emerging from the shadow of 9P37 at Corbetts Lane Junction by now 20 mins late. It pulls back a few minutes through allowances to East Croydon, but has now caught up 1P33, which it follows to Purley, meaning it is back to 18L. It then follows 1R36 to Redhill, where it arrives, at best, 15 late (and in the path of 9R31, delaying that). That’s all assuming the Q pathed freight that is around st that time of day isn’t running today (it wasn’t). It then proceeds on to Horsham where it arrives 14-15 mins late.

It can’t be terminated at East Croydon, Redhill, Gatwick or Three Bridges as there are no spare platforms for it to hang around for in order for it to make a RT start back north on diagram, except at Three Bridges, but 9R31 hard on its tail lights needs that platform at Three Bridges as booked.

Its arrived at Horsham 15 late, and with a large slice of good will from the driver and slick working departs north as 9J48only 7 late. With some further time caught up (unfortunately delaying a couple of services), it arrives Blackfriars 4 minutes down, which is not great for all of the several hundred passengers on board (which there were, as I saw that very train tonight when it departed St Pancras on time to the second).

Unfortunately, in our alternative universe 9J48 is 4 minutes late at Blackfriars, which delays 9O50 to St Albans by 4 minutes, which delays 9T40 to Bedford by 4 minutes, which delays 1F60 to Sheffield and 1D61 to Nottingham each by 5 minutes. 1F60 then delays 1L50 to Stansted at Wigston, 1K75 to Crewe at Sheet Stores, 1V15 to New Street / Cardiff at Derby and so on.




And that would have been the wrong decision, as described above. Not only wrong, but it would have triggered many more decisions required on the service for the next 4-5 hours, any of which might have been got wrong (incompetent in your language) and all adding to the workload of control unnecessarily.




Incorrect. Running the train late would have caused substantially more delay to thousands of people, inconveniencing many, many more people than the few (as helpfully highlighted in the passenger data by Robski above) boarding between Huntingdon and Arlesey, who were delayed by an additional 12 minutes (totalling 30), and a total delay of 15 minutes for those boarding at Hitchin.



And so we come to the real reason. Your individual journey was delayed by 15 minutes. As were those of some other passengers at Hitchin. But this was clearly the lesser of two evils compared to running it late.

Thank you for this masterly analysis, must have taken some time to chase down the many ramifications and shows the importance of recovering the timetable if at all possible esp while Woolmer Green exists as a bottleneck.

Hitchin passengers were only 15 minutes late, by using 9S33.


Skip stopping on 9S33 was unnecessary: it was on time at London Bridge.


The decision on which stops to skip had to be made before 9J33 departed from Peterborough, not as it was approaching Hitchin. The signallers needed the option of running 9J33 on the fast line, in order to maximise the opportunity to regain time. That's done by skipping all the stops to Stevenage, the only intermediate stations with a platform on the fast line are St Neots and Biggleswade. I note that none of your comments have specified whether 9J33 was on the fast line or the slow line when it passed through Hitchin.

This too.

Would just add that predictions based on recovering lost time by speeding up passenger handling at station stops are pie in the sky. Have commented in the past on apparent lack of urgency by platform staff in these situations. When I visited Korea with a 20+ trade mission many years ago we were warned repeatedly that the train would stop for only 60 secs and our experience was exactly that. So it can be done, just not in the UK.

 

[Richardr]

(British Rail had lots of failures but at least all fare paying passengers were treated the same. ie equal. Something that the modern railway is lacking IMHO).

What exactly does that mean?

 

[Nicholas Lewis]

Been off line for a bit.



This example, which you specifically highlighted as incomptent, was ‘just one train” though. Had it been left to run late it would have caused delay to dozens of other trains, carrying thousands of passengers, but read on…




Indeed. With the following service on time.

So intermediate passengers were delayed a further 12 minutes by this train being sent fast to Stevenage, than had it run booked calls.




But there’s no allowances between Peterboro and Hitchin. Assuming the Huntingdon call was done in 1 minute, and the others bang on (and I have never seen a Class 700 do a 30 second dwell time…) it would have been 16.5 mins late approaching Hitchin.

Which puts it behind the following Thameslink from Cambridge 9S33, and that puts it back to being (at least) 18 minutes late.

With 9J33 now following 9S33, we have two regulation decisions to make at Woolmer Green, with 1A63 and 1T31 approaching on the UF. In all likelihood 1A63 would go first, but 1T31 would wait (and be delayed). But waiting for 1A63 delays 9J33 further, however the allowances and a little catch up at Finsbury Park bring it back to 18L leaving there, still behind 9S33.

Unfortunately, That means it is going arrive at St P in a clash with 9P37 to Rainham, with the 9P marginally ahead, so that would go first. Another regulating decision to make. Then there is a 9R31 to Three Bridges right behind the 9P, another regulating decision to make. Hopefully the 9J goes first as that 9R is all shacks from East Croydon…

Firstly thanks for undertaking the forensic deep dive as it does demonstrate the amount of interactions these long distance TLk trains have and perhaps thats a downside of the through journey opportunities that have been created as there is limited opportunity to tolerate much late running. On a point of order 9R's off peak are Sth Croydon, Purley, Redhill in 21mins and 9J's are Coulsdon Sth, Merstham & Redhill in 20mins. What the clever signallers at 3Bs regularly do though is have a late running 9Jxx down the Fast line rather than tottering on the Slows as booked (they run Fast line in the Up direction from Stoats so work that one out!!). As an aside this should make it easy for either service group to back cover the Redhill line stations (incl Earlswood/Salfords). Sometimes they do just that but other times they don't and that is the inconsistency we suffer on the Redhill lines.

 

[Bald Rick]

As it happens, 9S33 was 4 minutes late at Hitchin, so that doesn’t necessarily follow (perhaps 9S33 should have been run non-stop from Cambridge to Hitchin as a result of that 4 minutes?).

Controllers have many skills, but one thing they don’t have is 20/20 clairvoyancy.

What is drilled in to controllers is to use the information that is available to them at the time they need to make a decision. The decision to run 9J33 fast was taken at least 15-20 minutes before 9S33 was due to depart Cambridge. There was no certainty that 9S33 would be late (and if so how late it would be), and therefore the controller would have quite correctly assumed it would be on time.

 

[JonathanH]

What the clever signallers at 3Bs regularly do though is have a late running 9Jxx down the Fast line rather than tottering on the Slows as booked (they run Fast line in the Up direction from Stoats so work that one out!!).

In 2018 they did run down the fast line off-peak, but other changes since then have freed up the path on the slow line for them, meaning all trains at East Croydon for Redhill leave from the platform 5 / 6 island.

They run fast line in up direction from Stoats Nest Junction as there isn't a gap for them on the slow line.

 

[bramling]

Controllers have many skills, but one thing they don’t have is 20/20 clairvoyancy.

What is drilled in to controllers is to use the information that is available to them at the time they need to make a decision. The decision to run 9J33 fast was taken at least 15-20 minutes before 9S33 was due to depart Cambridge. There was no certainty that 9S33 would be late (and if so how late it would be), and therefore the controller would have quite correctly assumed it would be on time.

I wasn’t implying that the control should have foreseen that 9S33 would be late, as clearly that’s never going to be the case.

However the whole chain of potential reactionary issues was on the premise that 9S33 would have run ahead of 9J33, which wasn’t a given.

 

[Nicholas Lewis]

In 2018 they did run down the fast line off-peak, but other changes since then have freed up the path on the slow line for them, meaning all trains at East Croydon for Redhill leave from the platform 5 / 6 island.

They run fast line in up direction from Stoats Nest Junction as there isn't a gap for them on the slow line.

OK thanks for that insight but its a dawdle through the 20mph at Sth Croydon although we are seeing more frequent use of the Slow to Slow Reversible at Sth Croydon which is 40mph. This ought to be the booked route as that give quicker jcn clearance at Sth Croydon as well as reduced point to point times.

 

[Bald Rick]

However the whole chain of potential reactionary issues was on the premise that 9S33 would have run ahead of 9J33, which wasn’t a given.

It wasn’t a given, but would have been what the controller would have to assume would happen, given the information available to them at the time the decison had to be made.

Besides, even if 9J33 did end up in front of 9S33 (delaying latter further into its journey), there would have been another chain of reactionary issues.

 

[Horizon22]

I wasn’t implying that the control should have foreseen that 9S33 would be late, as clearly that’s never going to be the case.

However the whole chain of potential reactionary issues was on the premise that 9S33 would have run ahead of 9J33, which wasn’t a given.

No but in making the alteration the TOC controller could also request the relevant regulation with Network Rail to ensure things went in the order requested which would make sense, given one train was running non-stop.

 

[bramling]

No but in making the alteration the TOC controller could also request the relevant regulation with Network Rail to ensure things went in the order requested which would make sense, given one train was running non-stop.

There’s a lot of coulds and mights about all this. What we do know is that all passengers between Peterborough and Stevenage *did* experience effectively a cancellation, and for Hitchin passengers at least this was completely unnecessary.

Given that an 18-minute delay isn’t massive in the grand scheme of things, especially with so much allowance during the journey, if the network can’t handle this without a significant intervention that renders the train useless to pretty much everyone between Peterborough and London (*) then it’s probably fair to say the setup isn’t fit for purpose.

* Given that Peterborough and Stevenage passengers tend to use LNER, and as evidenced by the fact the train passed through Hitchin with only a small handful of people on it. So this whole journey was only really useful for Peterborough-Stevenage journeys).

 

[Magdalia]

So this whole journey was only really useful for Peterborough-Stevenage journeys).

9J33 was very useful for all of those passengers that had an on time journey between Stevenage and Horsham.

 

[Horizon22]

Given that an 18-minute delay isn’t massive in the grand scheme of things, especially with so much allowance during the journey, if the network can’t handle this without a significant intervention that renders the train useless to pretty much everyone between Peterborough and London (*) then it’s probably fair to say the setup isn’t fit for purpose.

You keep saying this but on what basis are you making this argument? I would say for any regional / metro operator 18 minutes is a considerable delay worthy of an alteration and the extensive reply by @Bald Rick is proof that something would need to be done. For a intercity train, yes 18 minutes might be absorbable.

 

[Nicholas Lewis]

I see tonight 9J67 2054 PBO-HRH was 15L start from PBO ran fast to SVG arrive 3E. Then 9J69 also suffered a 14L start but held to its normal calling pattern now RT on departure from Redhill. It may have caused impact to other services perhaps but does show the level of recovery thats in the TL schedules to deal with late running. By the way im not suggesting any of it should be removed its very sensible and helped me on more than one occasion.

 

[mono1989]

I see tonight 9J67 2054 PBO-HRH was 15L start from PBO ran fast to SVG arrive 3E. Then 9J69 also suffered a 14L start but held to its normal calling pattern now RT on departure from Redhill. It may have caused impact to other services perhaps but does show the level of recovery thats in the TL schedules to deal with late running. By the way im not suggesting any of it should be removed its very sensible and helped me on more than one occasion.

As a TSM for TL this is normal for us when dealing with crewing issues.

To prevent multiple full cancelations we "step back" arrivals at both Peterborough and Horsham to allow trains to run albeit with a some 15+ late starts.

We tend not to run successive trains fast missing the same stops as this would create long gaps on certain parts of the network however as detailed above, we do have to try and get trains as righ time to the core as possible as we interact with multiple TOCS/ROC's.

I've seen LNER trains as far as Edinburgh have there late arrival attributed to late running TL trains owing to regulation issues out of Kings Cross.

We do understand that running a train fast does cause annoyance and it is normally a last resort but sometimes it is unavoidable.

 

[Robski]

Yesterday's performance by Thameslink was quite amazing (in a bad way).

9J74 21:55 Horsham to Peterborough started from Three Bridges due to a fatality. Ran on time as far as Finsbury Park where things started getting weird as the train was diverted into P5 for no apparent reason. At Potters Bar it gets put on the down slow line, and eventually is parked up at Welwyn GC platform 1.

We’re then informed that the onward driver was left behind in London(!) and that we are awaiting their arrival on a following train (2C72 23:27 ex Kings Cross) - at least, that's what I think was said as the PA system was defective as usual.

9J74 is delayed by about 30 minutes as a result of this and just to rub it in even further the service controller decides to pull all the stops between Stevenage and Peterborough! We couldn’t hear the driver inform us of this wise decision by the grand coalition of the utterly deranged because of the PA problem which was reported but as usual nothing was done about it.

Train leaves Stevenage empty after some passengers take it upon themselves (myself included) to inform other passengers about the amended calling pattern, left to stand in the cold for the next delayed Peterborough train 45 minutes behind (on top of the 30 minute delay!) - had they stayed on, they would have been stranded in Peterborough. By my estimates there were around 100 to 150 passengers waiting for the next train, plus quite a few who decided to use a taxi to get to their destination.

I have a lot of words to describe the actions taken by Thameslink that night. None of them are nice.

 

[Steve Harris]

Yesterday's performance by Thameslink was quite amazing (in a bad way).

9J74 21:55 Horsham to Peterborough started from Three Bridges due to a fatality. Ran on time as far as Finsbury Park where things started getting weird as the train was diverted into P5 for no apparent reason. At Potters Bar it gets put on the down slow line, and eventually is parked up at Welwyn GC platform 1.

We’re then informed that the onward driver was left behind in London(!) and that we are awaiting their arrival on a following train (2C72 23:27 ex Kings Cross) - at least, that's what I think was said as the PA system was defective as usual.

9J74 is delayed by about 30 minutes as a result of this and just to rub it in even further the service controller decides to pull all the stops between Stevenage and Peterborough! We couldn’t hear the driver inform us of this wise decision by the grand coalition of the utterly deranged because of the PA problem which was reported but as usual nothing was done about it.

Train leaves Stevenage empty after some passengers take it upon themselves (myself included) to inform other passengers about the amended calling pattern, left to stand in the cold for the next delayed Peterborough train 45 minutes behind (on top of the 30 minute delay!) - had they stayed on, they would have been stranded in Peterborough. By my estimates there were around 100 to 150 passengers waiting for the next train, plus quite a few who decided to use a taxi to get to their destination.

I have a lot of words to describe the actions taken by Thameslink that night. None of them are nice.

That sounds like a comedy sketch straight from the muppets !!

I could understand the skip stopping if the next service was right behind 9J74, but it wasn't. And turfing customers out at that time of night IMO shouldn't be the done thing. No doubt the excuse was that the driver was going out of hours.

 

[Horizon22]

Yesterday's performance by Thameslink was quite amazing (in a bad way).

9J74 21:55 Horsham to Peterborough started from Three Bridges due to a fatality. Ran on time as far as Finsbury Park where things started getting weird as the train was diverted into P5 for no apparent reason. At Potters Bar it gets put on the down slow line, and eventually is parked up at Welwyn GC platform 1.

We’re then informed that the onward driver was left behind in London(!) and that we are awaiting their arrival on a following train (2C72 23:27 ex Kings Cross) - at least, that's what I think was said as the PA system was defective as usual.

9J74 is delayed by about 30 minutes as a result of this and just to rub it in even further the service controller decides to pull all the stops between Stevenage and Peterborough! We couldn’t hear the driver inform us of this wise decision by the grand coalition of the utterly deranged because of the PA problem which was reported but as usual nothing was done about it.

Train leaves Stevenage empty after some passengers take it upon themselves (myself included) to inform other passengers about the amended calling pattern, left to stand in the cold for the next delayed Peterborough train 45 minutes behind (on top of the 30 minute delay!) - had they stayed on, they would have been stranded in Peterborough. By my estimates there were around 100 to 150 passengers waiting for the next train, plus quite a few who decided to use a taxi to get to their destination.

I have a lot of words to describe the actions taken by Thameslink that night. None of them are nice.

There's a few things to unpick here.

Firstly, being placed into P1 at Welwyn is an attempt by the signaller to get the train "out of the way", presumably knowing there was going to need to be an unbooked crew relief at Welwyn. Probably following the disruption, the driver was maybe not on their booked diagram, and was due to finish.

Next "left behind in London" is probably somewhat paraphrased - it is likely the driver was displaced following the fatality so wasn't where they needed to be and had to "pass" up to Welwyn.

The only reason I can see for pulling the stops is a driver being out of hours because 9J74 is meant to terminate and go into the sidings which would normally just be absorbed. However 5J74 appears to be cancelled on RTT, so it is hard to tell what actually happened to the stock or whether the sidings were full for some reason (related to the incident, perhaps).

 

[Horizon22]

I could understand the skip stopping if the next service was right behind 9J74, but it wasn't. And turfing customers out at that time of night IMO shouldn't be the done thing. No doubt the excuse was that the driver was going out of hours.

Not really an "excuse"; the driver might effectively refuse to drive the service and then nobody would be served.

But I agree it should not be the "done thing", and most controllers would avoid doing so.

 

[Bikeman78]

Not really an "excuse"; the driver would effectively refuse to drive the service and then nobody would be served.

Is getting home 30 minutes late instead of 15 minutes late really that different? Drivers on that route must be used to being a lot later than that! I've been on late running evening trains and never known the crew to insist that stops get pulled.

 

[Horizon22]

Is getting home 30 minutes late instead of 15 minutes late really that different? Drivers on that route must be used to being a lot later than that! I've been on late running evening trains and never known the crew to insist that stops get pulled.

Without being there, it's very hard to know what the communications were between driver / resource controller / service controller / signaller.

Especially as the driver was back "in London" so those discussions probably happened - "well I'll drive it but..."

Each circumstance is different. Ideally they were on their booked train, but with a fatality, crew and fleet are totally displaced and trying to track everything, prioritisating and managing resources gets very hard. It is also difficult late at night where you have very limited (if not zero) spare resource.