Regular Thameslink disruptions over a week

[Purple Train]

The problem is the Elizabeth line core isn't nearly a quarter the reason they have disruption most of the time and the same often is for the Thameslink core. The issue is often the line at one end of the core when it comes to the Lizzy, Thameslink however has lines that aren't as near reliable as the GEML for example. And it feeds more than 2 lines either side of the core and needs pretty much all of them to be working relatively without much delays else it starts causing serious issues.

Absolutely, and I was talking with reference to the issue at City Thameslink rather than generally. Of course the Elizabeth Line has the advantage of having only three sections and some degree of operational separation, which I understand is far more difficult, if not impossible, to obtain on Thameslink. It isn't often that an issue with the GEML completely scuppers trains through Slough, for instance, but, taking Monday as an example, the failed RHTT at Stevenage affected services across the network, which were themselves grappling with the other five incidents!

It was 100% of the reason this morning!

Two days running!

 

[Railcar]

No, it’s the track circuit failure near City TL.

Are there track circuit failures at Drayton Park, South Acton or North Pole, where AC/DC changeovers occur?

 

[Mattplans]

Another day. This is insane.

 

[andos]

Looks like all the east grinstead trains will be cancelled again! My usual train this morning was cancelled due to driver shortage so I’m not having much luck.

 

[Bald Rick]

Are there track circuit failures at Drayton Park, South Acton or North Pole, where AC/DC changeovers occur?

Sometimes, but then they dont have 5MW trains on each line 16-20 times an hour.

 

[andy33gmail]

Another day. This is insane.

That explains why the Great Northern from Ely to Littleport was rammed last night, despite being late on a weekday! Can’t imagine what it must have been like nearer the origin

 

[Russel]

Sometimes, but then they dont have 5MW trains on each line 16-20 times an hour.

The core infrastructure should have been designed and built to cope with 16-20 5Mw trains an hour though.

Even if it was, something is clearly very wrong.

 

[jon0844]

The core infrastructure should have been designed and built to cope with 16-20 5Mw trains an hour though.

Even if it was, something is clearly very wrong.

I am sure it was designed just fine and there is something else causing the issue, which needs to be found (and quick) given the repeated instances.

This must be the worst possible type of incident that seemingly gets fixed and then fails again. The chaos this is causing is off the charts as there's no normal recovery plan in place where certain 9xxx services get cut to thin out the service. This is where there is almost nothing able to run, and far too many diagrams require trains to pass through to change crew so you can't just terminate them all short.

 

[RJ]

The core is shut at the moment so perhaps the issue is being rectified.

 

[uglymonkey]

Fingers crossed. Let's see what today brings!

 

[Bald Rick]

The core infrastructure should have been designed and built to cope with 16-20 5Mw trains an hour though.

Even if it was, something is clearly very wrong.

Well, yes, on both counts.

 

[AM9]

Well, yes, on both counts.

Isn't the overlap of ac and DC including both Farringdon and City Thameslink a further complication compared with the LO changeovers mentioned - also not helped by a standing start up the 1:27 on DC to Blackfriars which maximises the return currents on every train? Were it not for the DC-only Smithfield sidings, wouldn't a changeover at Blackfriars be a less problematic solution for the through route? Platforms 3 & 4 could still be DC only.

 

[Magdalia]

I come late to this discussion, partly because I find its title grating.

Disruptions on Thameslink are not regular, they may be frequent but they are not regular.

On the other hand disruptions are not random or independent.

The disruptions in the core in recent days are clearly a recurrent problem, and will go on until a solution is found.

I am reminded of a little radio programme by the mathematician Hannah Fry called Uncharted. One episode was about a sequence of mystery faults on the Singapore Metro, it is here for those interested:

BBC Radio 4 - Uncharted with Hannah Fry, 4. This Train Has Been Delayed

It’s 2016 and Singapore’s metro system is misbehaving wildly.

www.bbc.co.uk

4. This Train Has Been Delayed​

Uncharted with Hannah Fry
It’s 2016 and Singapore’s metro system is misbehaving wildly. The rail engineers and company officials are flummoxed. Can data save the day?

In the current circumstances, someone needs to be looking at the data, and, in particular, which units are in the core when the fault is triggered.

 

[jon0844]

As we get more and more advanced signalling systems, I wonder if these sorts of issues will become more common and be something we just have to get used to. Clearly from the above story, such issues occur all around the world - and there was the issue on the Northern line not so long ago.

Another issue is faulty PIS (Passenger Information) codes that can result in trains having to be cancelled, as if the train doesn't pick up the stopping pattern correctly it can lead to a train having to manually release doors at all stations - and that can lead to operational incidents. I believe guidance now is that you can't run a train like that for more than a short time. So a (however many) digit code can lead to a train not running at all!

Or, at best, you end up with trains going around with no working PIS so passengers don't know where the train stops, or where it is, or says 'Not In Service' as in the issue with the 379s right now.

 

[OscarH]

The 9J's (PBO-HOR) have similar issues, because the crew depot is in Horsham, trains have to be skipped East Croydon to Gatwick, leaving Redhill route with little or no service. Thameslink needs a better crewing locations but unlikely to happen

The crew depot might be in Horsham, but that doesn't stop them cancelling short at Three Bridges half the time and never making it there, I see that far more often than skipping on the Redhill bit to be honest

 

[Andy Orchard]

Thameslink "The Station Stop Skipping Company" . Thamestink do this trick regularly with Leagrave, Harlington and Flitwick . Tell you the train is cancelled but you stand at the station and watch it run through non stop.

But for some reason they never skip Luton Airport Parkway despite it having alternative competitor (EMR) services and also the "stop service" starting from Luton .

They very rarely used to skip St Albans City either, no idea why.........

 

[Magdalia]

ut for some reason they never skip Luton Airport Parkway

They very rarely used to skip St Albans City either, no idea why.........

The very large numbers of passengers using those stations is the reason: St Albans has 6.7 million entries and exits in 2024/25, Luton Airport Parkway has 4.7 million.

The crew depot might be in Horsham, but that doesn't stop them cancelling short at Three Bridges half the time and never making it there

There is also a crew depot at Peterborough, roughly half the time the driver will be from Peterborough not Horsham.

 

[Russel]

Disruptions on Thameslink are not regular, they may be frequent but they are not regular.

Frequent, but not regular?

 

[perstreperous]

Well, I have data as I usually do Sutton-Herne Hill [Thameslink] but, when that is up the creek, do Sutton-Balham [Southern] then London Underground (so its app records all the zone 2-3 tickets).

From years ago to about 4 weeks from now Thameslink was fine - I did the second journey 1 or 2 times a month at most. Then it started to slide (second journey more often than not) and, now, it's the second journey every day

 

[rwuk]

Frequent, but not regular?

Frequent - disruption happened 7 times this week
Regular - disruption happened between 13:00 and 13:30 on each occasion (i.e. irregular would be that there's no pattern to when the disruption occurs)

 

[Russel]

Frequent - disruption happened 7 times this week
Regular - disruption happened between 13:00 and 13:30 on each occasion (i.e. irregular would be that there's no pattern to when the disruption occurs)

This is the issue though, this week it's the core, next week it'll be East Croydon, the week after it'll be the core again, you get the picture, at what point do we admit that things aren't working?

 

[rwuk]

This is the issue though, this week it's the core, next week it'll be East Croydon, the week after it'll be the core again, you get the picture, at what point do we admit that things aren't working?

I was replying to your post which appeared to say "[How can the disruption be] Frequent, but not regular?".

I gave no opinion on when we should admit that things are or are not working. I have no opinion to give on that. The description of the disruption you write (abstracted to: this week in X, next week in Y, the week after in Z) is the very definition of not regular. Frequent, perhaps. But not regular.

A bit of a tangent, but there we go.

 

[Bald Rick]

Isn't the overlap of ac and DC including both Farringdon and City Thameslink a further complication compared with the LO changeovers mentioned - also not helped by a standing start up the 1:27 on DC to Blackfriars which maximises the return currents on every train? Were it not for the DC-only Smithfield sidings, wouldn't a changeover at Blackfriars be a less problematic solution for the through route? Platforms 3 & 4 could still be DC only.

Blackfriars would be more problematic. The issue is the negative return from DC trying to get to the earthed return of AC. Blackfriars is closer to more DC.

I am reminded of a little radio programme by the mathematician Hannah Fry called Uncharted. One episode was about a sequence of mystery faults on the Singapore Metro, it is here for those interested:

BBC Radio 4 - Uncharted with Hannah Fry, 4. This Train Has Been Delayed

It’s 2016 and Singapore’s metro system is misbehaving wildly.

www.bbc.co.uk

In the current circumstances, someone needs to be looking at the data, and, in particular, which units are in the core when the fault is triggered.

This has been done. You‘re not the only person who has suggested that podcast!

They very rarely used to skip St Albans City either, no idea why.........

Oh they do. Including this week (and in the last 20 minutes as it happens).

The very large numbers of passengers using those stations is the reason: St Albans has 6.7 million entries and exits in 2024/25, Luton Airport Parkway has 4.7 million.


There is also a crew depot at Peterborough, roughly half the time the driver will be from Peterborough not Horsham.

There’s also a crew depot at St Albans.

 

[bramling]

Oh they do. Including this week (and in the last 20 minutes as it happens).

I have seen the odd TL service run fast from St Pancras to Bedford. No real worry though, St Albans only has at least 9 other trains per hour to choose from…

 

[Minstral25]

The crew depot might be in Horsham, but that doesn't stop them cancelling short at Three Bridges half the time and never making it there, I see that far more often than skipping on the Redhill bit to be honest

You obviously haven't been dropped off at East Croydon with a hundred other people much. Looking at the stop shorts on RecentTrainTimes it's happened about 7-8 times in last 4 weeks, whereas Redhill has been skipped 50+ times (probably more but I got bored going back assessing individual journeys, which is difficult as they often run down the slow non-stop).

I'd assume if a PBO driver is driving they can stop for their break at Three Bridges, but a Horsham driver would need to use a Southern service back to depot. I'd assume all drivers are entitled to a break on finishing a PBO to HOR journey.

In any case it would make sense for the driver depot to be further into London at both ends.

 

[JonathanH]

In any case it would make sense for the driver depot to be further into London at both ends.

While I agree in part, it wouldn't make for easy start and end of service if the depots were at Hitchin and Redhill, rather than Peterborough and Horsham, both locations where additional sidings were built for Thameslink services. Neither Hitchin nor Redhill have stabling space for Thameslink trains.

Understanding the structure of the actual driver diagrams would be very interesting in knowing how the Peterborough to Horsham service actually works.

 

[riceuten]

TSGN have insufficient - and have always had insufficient driving staff to run the service - they rely on rest day working and presumably are happy to take a hit on performance bonuses if this if offset by driver training costs - if I had a pound for every TSGN manager who has told me that ‘we have xx drivers in training, when they graduate, then we can offer a better service’. This has never happened. And probably never will.

 

[OscarH]

You obviously haven't been dropped off at East Croydon with a hundred other people much. Looking at the stop shorts on RecentTrainTimes it's happened about 7-8 times in last 4 weeks, whereas Redhill has been skipped 50+ times (probably more but I got bored going back assessing individual journeys, which is difficult as they often run down the slow non-stop).

I was intrigued by this because I don't think I've ever been on one that goes fast through Redhill, but I have had my train from Ifield cancelled all the bloody time, to the point you start to wake up expecting it.

Some very crude analysis suggests both are actually roughly equally likely - so it's interesting how perceptions differ! I wonder if there's some reasoning to why, like certain times of date have one outcome or the other more likely, and it's just luck on our travel times.

My very crude analysis was to look at the Darwin data for every Thameslink service with headcode 9J since the 1st November, and then excluded ones which weren't Peterborough to Horsham (to exclude engineering work/early and late services which have odd patterns)

I got 132 total cancellations, and the following cancellations per stop (apologies for the TIPLOCs):

Tiploc: ALEXNDP: Cancelled:    5, Occurrences:   11, Percentage: 45.45%
Tiploc:  SALFDS: Cancelled:   73, Occurrences:  486, Percentage: 15.02%
Tiploc: EARLSWD: Cancelled:   71, Occurrences:  486, Percentage: 14.61%
Tiploc:    LHVN: Cancelled:  285, Occurrences: 2534, Percentage: 11.25%
Tiploc:  IFIELD: Cancelled:  285, Occurrences: 2534, Percentage: 11.25%
Tiploc: CRAWLEY: Cancelled:  285, Occurrences: 2534, Percentage: 11.25%
Tiploc: HNTNGDN: Cancelled:  284, Occurrences: 2534, Percentage: 11.21%
Tiploc: MERSTHM: Cancelled:  249, Occurrences: 2325, Percentage: 10.71%
Tiploc: COLSDNS: Cancelled:  254, Occurrences: 2384, Percentage: 10.65%
Tiploc:  PURLEY: Cancelled:   24, Occurrences:  233, Percentage: 10.30%
Tiploc:  HORLEY: Cancelled:  254, Occurrences: 2534, Percentage: 10.02%
Tiploc: REDHILL: Cancelled:  252, Occurrences: 2534, Percentage:  9.94%
Tiploc:    SNDY: Cancelled:  245, Occurrences: 2534, Percentage:  9.67%
Tiploc: BIGLSWD: Cancelled:  245, Occurrences: 2534, Percentage:  9.67%
Tiploc:  ARLSEY: Cancelled:  245, Occurrences: 2534, Percentage:  9.67%
Tiploc: STNEOTS: Cancelled:  241, Occurrences: 2534, Percentage:  9.51%
Tiploc: HORSHAM: Cancelled:  233, Occurrences: 2534, Percentage:  9.19%
Tiploc: HITCHIN: Cancelled:  228, Occurrences: 2534, Percentage:  9.00%
Tiploc: FAYGATE: Cancelled:   56, Occurrences:  631, Percentage:  8.87%
Tiploc: SCROYDN: Cancelled:    7, Occurrences:   90, Percentage:  7.78%
Tiploc: STEVNGE: Cancelled:  195, Occurrences: 2534, Percentage:  7.70%
Tiploc: ECROYDN: Cancelled:  189, Occurrences: 2534, Percentage:  7.46%
Tiploc:    PBRO: Cancelled:  185, Occurrences: 2534, Percentage:  7.30%
Tiploc: LNDNBDE: Cancelled:  183, Occurrences: 2522, Percentage:  7.26%
Tiploc: FRNDNLT: Cancelled:  179, Occurrences: 2534, Percentage:  7.06%
Tiploc: STPXBOX: Cancelled:  178, Occurrences: 2534, Percentage:  7.02%
Tiploc:    BLFR: Cancelled:  177, Occurrences: 2534, Percentage:  6.99%
Tiploc:    FNPK: Cancelled:  174, Occurrences: 2534, Percentage:  6.87%
Tiploc: WLWYNGC: Cancelled:    3, Occurrences:   45, Percentage:  6.67%
Tiploc: CTMSLNK: Cancelled:  148, Occurrences: 2294, Percentage:  6.45%
Tiploc:    GTWK: Cancelled:  161, Occurrences: 2534, Percentage:  6.35%
Tiploc:  THBDGS: Cancelled:  160, Occurrences: 2534, Percentage:  6.31%
Tiploc:  NORWDJ: Cancelled:   14, Occurrences:  236, Percentage:  5.93%
Tiploc:  WLWYNN: Cancelled:    0, Occurrences:    5, Percentage:  0.00%
Tiploc: POTRSBR: Cancelled:    0, Occurrences:    1, Percentage:  0.00%
Tiploc:    LTCE: Cancelled:    0, Occurrences:    2, Percentage:  0.00%
Tiploc:    KNGX: Cancelled:    0, Occurrences:    9, Percentage:  0.00%
Tiploc: KNEBWTH: Cancelled:    0, Occurrences:    5, Percentage:  0.00%
Tiploc: HATFILD: Cancelled:    0, Occurrences:    1, Percentage:  0.00%

I'm afraid it includes junk from different service patterns at the fringes of the day, or special-stop-orders on one day. Crawley and Ifield beat Redhill, but not by a relevant amount, and Horsham is very slightly below Redhill (seems like a few of what I assumed were stopping short are actually running fast between Horsham and Three Bridges).

 

[Minstral25]

While I agree in part, it wouldn't make for easy start and end of service if the depots were at Hitchin and Redhill, rather than Peterborough and Horsham, both locations where additional sidings were built for Thameslink services. Neither Hitchin nor Redhill have stabling space for Thameslink trains.

Understanding the structure of the actual driver diagrams would be very interesting in knowing how the Peterborough to Horsham service actually works.

As I said in an earlier post it hasn't been designed or built to allow it to happen, so it will not.

However, speculating if it had been built, if from Redhill then trains could start in both directions, some to Peterborough and some to Horsham with the Horsham trains eventually running through Redhill once service is running. It's just simple timetabling.

I was intrigued by this because I don't think I've ever been on one that goes fast through Redhill, but I have had my train from Ifield cancelled all the bloody time, to the point you start to wake up expecting it.

Some very crude analysis suggests both are actually roughly equally likely - so it's interesting how perceptions differ! I wonder if there's some reasoning to why, like certain times of date have one outcome or the other more likely, and it's just luck on our travel times.

My very crude analysis was to look at the Darwin data for every Thameslink service with headcode 9J since the 1st November, and then excluded ones which weren't Peterborough to Horsham (to exclude engineering work/early and late services which have odd patterns)

I got 132 total cancellations, and the following cancellations per stop (apologies for the TIPLOCs):

Tiploc: ALEXNDP: Cancelled:    5, Occurrences:   11, Percentage: 45.45%
Tiploc:  SALFDS: Cancelled:   73, Occurrences:  486, Percentage: 15.02%
Tiploc: EARLSWD: Cancelled:   71, Occurrences:  486, Percentage: 14.61%
Tiploc:    LHVN: Cancelled:  285, Occurrences: 2534, Percentage: 11.25%
Tiploc:  IFIELD: Cancelled:  285, Occurrences: 2534, Percentage: 11.25%
Tiploc: CRAWLEY: Cancelled:  285, Occurrences: 2534, Percentage: 11.25%
Tiploc: HNTNGDN: Cancelled:  284, Occurrences: 2534, Percentage: 11.21%
Tiploc: MERSTHM: Cancelled:  249, Occurrences: 2325, Percentage: 10.71%
Tiploc: COLSDNS: Cancelled:  254, Occurrences: 2384, Percentage: 10.65%
Tiploc:  PURLEY: Cancelled:   24, Occurrences:  233, Percentage: 10.30%
Tiploc:  HORLEY: Cancelled:  254, Occurrences: 2534, Percentage: 10.02%
Tiploc: REDHILL: Cancelled:  252, Occurrences: 2534, Percentage:  9.94%
Tiploc:    SNDY: Cancelled:  245, Occurrences: 2534, Percentage:  9.67%
Tiploc: BIGLSWD: Cancelled:  245, Occurrences: 2534, Percentage:  9.67%
Tiploc:  ARLSEY: Cancelled:  245, Occurrences: 2534, Percentage:  9.67%
Tiploc: STNEOTS: Cancelled:  241, Occurrences: 2534, Percentage:  9.51%
Tiploc: HORSHAM: Cancelled:  233, Occurrences: 2534, Percentage:  9.19%
Tiploc: HITCHIN: Cancelled:  228, Occurrences: 2534, Percentage:  9.00%
Tiploc: FAYGATE: Cancelled:   56, Occurrences:  631, Percentage:  8.87%
Tiploc: SCROYDN: Cancelled:    7, Occurrences:   90, Percentage:  7.78%
Tiploc: STEVNGE: Cancelled:  195, Occurrences: 2534, Percentage:  7.70%
Tiploc: ECROYDN: Cancelled:  189, Occurrences: 2534, Percentage:  7.46%
Tiploc:    PBRO: Cancelled:  185, Occurrences: 2534, Percentage:  7.30%
Tiploc: LNDNBDE: Cancelled:  183, Occurrences: 2522, Percentage:  7.26%
Tiploc: FRNDNLT: Cancelled:  179, Occurrences: 2534, Percentage:  7.06%
Tiploc: STPXBOX: Cancelled:  178, Occurrences: 2534, Percentage:  7.02%
Tiploc:    BLFR: Cancelled:  177, Occurrences: 2534, Percentage:  6.99%
Tiploc:    FNPK: Cancelled:  174, Occurrences: 2534, Percentage:  6.87%
Tiploc: WLWYNGC: Cancelled:    3, Occurrences:   45, Percentage:  6.67%
Tiploc: CTMSLNK: Cancelled:  148, Occurrences: 2294, Percentage:  6.45%
Tiploc:    GTWK: Cancelled:  161, Occurrences: 2534, Percentage:  6.35%
Tiploc:  THBDGS: Cancelled:  160, Occurrences: 2534, Percentage:  6.31%
Tiploc:  NORWDJ: Cancelled:   14, Occurrences:  236, Percentage:  5.93%
Tiploc:  WLWYNN: Cancelled:    0, Occurrences:    5, Percentage:  0.00%
Tiploc: POTRSBR: Cancelled:    0, Occurrences:    1, Percentage:  0.00%
Tiploc:    LTCE: Cancelled:    0, Occurrences:    2, Percentage:  0.00%
Tiploc:    KNGX: Cancelled:    0, Occurrences:    9, Percentage:  0.00%
Tiploc: KNEBWTH: Cancelled:    0, Occurrences:    5, Percentage:  0.00%
Tiploc: HATFILD: Cancelled:    0, Occurrences:    1, Percentage:  0.00%

I'm afraid it includes junk from different service patterns at the fringes of the day, or special-stop-orders on one day. Crawley and Ifield beat Redhill, but not by a relevant amount, and Horsham is very slightly below Redhill (seems like a few of what I assumed were stopping short are actually running fast between Horsham and Three Bridges).

I'd trust your numbers much better than my eyeballs looking through RecentTimeTrains. I wasn't specifically looking at Redhill but the route through Redhill in its entirety. Earlswood and Salfords are in fact much worse as not included the Bedford to Three Bridges service which are cancelled before 9J's call - I guess the low number of calls indicates early morning and late night trains. Purley and South Croydon rarely have 9J's call too.

From the statistics 73 trains are cancelled from Three Bridges to Horsham (unless some run fast through), whereas 253 skipped Redhill. I do note for Redhill that RTT often count a train as calling even if it goes through at speed, as happened with my train on Monday. I am not sure if that affects your statistics source (Darwin) as I do not have access to it.

Whichever way you look at it, Thameslink's performance to both Horsham and Redhill is terrible.

 

[OscarH]

I do note for Redhill that RTT often count a train as calling even if it goes through at speed, as happened with my train on Monday. I am not sure if that affects your statistics source (Darwin) as I do not have access to it.

It shouldn't - Darwin is the system that powers things like the station displays and other passenger facing info, it's quite common for cancellations to be put in this system and not in the systems that feed into RTT.

Whichever way you look at it, Thameslink's performance to both Horsham and Redhill is terrible

That is definitely the takeaway from it! It really is abysmal