Ah you didn’t say you olanned a stock and crew cascade.…
There had been a very serious incident at Huntingdon, where the line was closed for almost an hour. There were no trains of any type.
-
Our new ticketing site is now live! Using either this or the original site (both powered by TrainSplit) helps support the running of the forum with every ticket purchase! Find out more and ask any questions/give us feedback in this thread!
Regular Thameslink disruption: what could be done to improve performance and resilience?
- Thread starter Jan Mayen
- Start date
- Status
Ah you didn’t say you olanned a stock and crew cascade.…
There had been a very serious incident at Huntingdon, where the line was closed for almost an hour. There were no trains of any type.
Sponsor Post - registered members do not see these adverts; click here to register, or click here to log in
R
RailUK Forums
How many metros are as complicated as Thameslink though? Not just the number of branches and different stopping patterns, but in terms of linking up so much of the national rail network, importing delays from anywhere between Brighton and Scotland. How reliable is it realistically possible to get without lopping off some of the branches? Peterborough and Cambridge seem obvious targets with their interaction with LNER long distance services and Welwyn viaduct, but if capacity at Kings Cross no longer exists they are stuck.
Last edited:
bramling
Veteran Member
As I understand it, the line was opened around 1225. The gap was from 1240 to 1353 (at Huntingdon), and was partly caused by two successive Thameslink services being amended to run non-stop.
So there were trains, just not the type that passengers at places like Huntingdon or Sandy could use…
Last edited:
uglymonkey
Member
- Joined
- 10 Aug 2018
- Messages
- 708
Everything funnels through two tracks, St Pancras to Blackfriars, and south of the river is worse for disruption than North most days. These are my two main issues.
Last edited by a moderator:
The trains that come south from Peterboro’ had to get there first….
Reviving this thread so hopefully all the de-Thameslink chat can be contained here so that the other thread can return to running direct trains between Thurso and Catania via Thameslink, RER D and the Milan S Lines.
The main lines into St Pancras, King's Cross and London Bridge are all set up such that suburban services need to be funnelled into the Thameslink Core. Terminating services in the respective 'upper level' stations necessarily means cutting back existing upper level services. King's Cross only lost one platform but there have been a steady increase of LDHS services over the years making the station essentially full. Cutting back LDHS services makes neither financial nor political sense - it won't be allowed to happen. Resistance is futile.
That said, I do think the Thameslink network is too complicated, and the complication is in the pairing of north south services. I think the half-hourly basic structure is a problem - it means when a service (say Horsham - Peterborough) is a few minutes late, the choices are skipping stops to get the service back on path or stepping back by 30 minutes. Simplifying service pairings would provide more flexibility. If both sets of ECML outers (Cambridges and Peterboroughs) are paired with Brighton, then you can afford services to be 15 minutes late as when they get to south of the river you can just step everything back 15 minutes, i.e. the St Pancras - Brighton portion 'feels' on time. If you have a pool of drivers that all sign Brighton - Peterborough and Cambridge then that stepping back becomes possible. Likewise if all 4tph Bedfords are paired with Redhill then you can just step the whole thing back 15 minutes. With the current 30-minute pattern you can't step back - as if a Bedford - Brighton is 15-minutes late the driver doesn't sign Cambridge.
rmHawk765
Member
I am personally fully onboard with the idea of Thameslink, but we simply cannot achieve the level of timing precision and reliability the current network needs to function well whenever possible, at least not for now.
== Doublepost prevention - post automatically merged: ==
I fully agree with this, I feel the crossing-over when they go north of the river is completely unnecessary.
It clearly isn't completely unnecessary, otherwise the train planners would never have come up with it.
It means, for example that both routes south of the river retain a Thameslink service when there is disruption on one of the routes north of London. It means that Redhill has a service roughly every ten minutes to / from East Croydon, in conjunction with the Victoria service. It means that both routes have a fast link to Gatwick.
Bedford to St Pancras, and Stevenage to St Pancras both have to run on a quarter hourly basis. On the south side that doesn't work so well in the service mix.
Last edited:
Mgameing123
Member
Maybe you could build sidings at the stations outside of the core so that in case a service gets cancelled that train can take over the service? That could minimize the disruption to only those that travel from the core.
Or else simplify the service patterns in the core so that we no longer have those half hourly services to the most random destinations.
Or else simplify the service patterns in the core so that we no longer have those half hourly services to the most random destinations.
rmHawk765
Member
I'm pretty sure the Swiss literally expect some German services to always arrive late
Something I've heard about issues with Peterborough to Horsham not being able to recover is a lack of units being stabled at Horsham, so this could actually be a good idea.
Mgameing123
Member
But is definitely a good solution to the problem. That way making sure if a train enters the core too late it can be cancelled and turned into a standby train at some siding right before the core. That way disruption can be minimised.
Wouldn't that need stand-by drivers?
Mgameing123
Member
Yes it would but considering how bad reliability is on the line then it might actually be worth considering.
I was always surprised by the addition of routes in Southeastern territory (Sevenoaks and Rainham). What about killing them off and replacing them with some Southern territory services? I.e. Keeping things more closely aligned to the Brighton Main Line
whoosh
Established Member
- Joined
- 3 Sep 2008
- Messages
- 1,892
Standby drivers are too Treasury unfriendly!
What an idea!
But... the Treasury!
Mattplans
Member
Yes Rainham, no Sevenoaks/Orpington. Not sure Southeastern would ever provide 4tph via Catford?
But I agree, Rainham really has to go.
== Doublepost prevention - post automatically merged: ==
Considering we're discussing the implications of unnecessary branches, I'm not the biggest fan of this idea. I'd much prefer it to just go to Bellingham, Bromley South, Swanley or Orpington
Last edited:
The origin destination matrix data shows why. I posted this on the Sheffield-Brighton discussion recently.
One of the useful features of the origin destination matrix dataset is the regional tag, which makes it fairly easy to pick out by eye the biggest long distance Thameslink flows. I have identified the following South East-East of England flows bigger than 10k one way journeys:
38775 Gatwick-Cambridge
35426 Gatwick-St Albans
26884 Brighton-Cambridge
26191 Gatwick-Luton Airport Parkway
24694 Gatwick-Bedford
22222 Brighton-St Albans
17119 Brighton-Bedford
15249 Gatwick-Stevenage
13679 Brighton-Luton Airport Parkway
13886 Gatwick-Peterborough
12896 Brighton-Luton
12728 Gatwick-Luton
11375 Gatwick-Harpenden
There is a big asymmetry between the South East, where Gatwick and Brighton are the only really important stations outside London, and the East of England, where lots of stations on all routes have reasonable flows, especially with Gatwick.
These are historic, going back to the days when South Eastern ran into Blackfriars/Holborn Viaduct, but only via Elephant, not via London Bridge. Rainham is top of my list for pruning.
Land acquisition for new sidings near the core would be very expensive. The ECML already has Hornsey and the MML already has Cricklewood. Maybe it would help if the southern side of Thameslink could use Selhurst?
This depends on the location of the disruption. If the disruption is north of the river then Horsham ends up with surplus trains while Peterborough is short. If the disruption is south of the river then Horsham ends up short while Peterborough has a surplus.
Airport passengers (both Gatwick and Luton) need direct trains (or places to interchange outside the core) because interchange in the core by passengers with lots of luggage adds to dwell times and increases the headway between trains.
Piccadilly and Crossrail (and indeed Thameslink itself) have plenty of passengers joining and leaving with suitcases. Thameslink trains are designed for rapid access and egress. If airport direct trains are precious then we should be running Victoria and Piccadilly line trains into each other's tunnel. Everyone just needs to grow up.
London Trains
Established Member
- Joined
- 9 Oct 2017
- Messages
- 1,227
Beckenham Junction isn't a long, unnecessary branch like Rainham is though - only the short section between Herne Hill and Beckenham Junction isn't already served by Thameslink services, and as this service approaches the core from the Elephant and Castle direction, it would be relatively easy to spin them at Blackfriars in the event of disruption.
== Doublepost prevention - post automatically merged: ==
Route knowledge can be maintained by Southeastern running a once a day or even once a week service into the Blackfriars bays, I wouldn't let issues like this stand in the way of making wider improvements to Thameslink and Southeastern.
Last edited:
izvor
Member
There are already a handful of ECS movements in the day that do this anyway so it’s not an issue
Mattplans
Member
I don't really understand why Southeastern couldn't continue to run their BFR > BKJ service. Perhaps TL should just stick to one route? What's wrong with them just sending additional trains to Bellingham, Bromley South or Swanley? If there are additional trains to be run, it makes more sense for them to go on Thameslink's actual route instead of another spur....
Beckenham Junction is also part of some Thameslink driver's existing route knowledge being a diversionary route for the Catford Loop services. They couldn't run as efficiently as the Southeastern service and will need to do the Herne Hill to Beckenham Junction stretch at specific times to avoid the fast services which may not work with free slots in the core. It's also only three driver half shifts at most currently no where near a regular service.
There are two key sources of unreliability, lack of drivers and infrastructure unreliability on heavily used sections. The infrastructure unreliability will impact the service groups that use it no matter what services are or aren't Thameslink. It's possible some service groupings may allow a more effective service if the core has reduced capacity, though worth noting the last core failure the Orpington to Luton services were cancelled rather than turned at Blackfriars so additional services via Catford may not change the picture.
For lack of drivers, it's been said that productive days per driver has decreased across the industry. Many operators are running fewer services to address that and reduce rest day work requirements. Thameslink has suffered more as the end of route depot model has meant the decreases can't easily be shared across depots and were in a poor position to start with. Running additional services means the operators combined have to cover more services with the exact same train crew, which will only make things worse. Redistributing will only help if services can be covered more efficiently that way.
Taking the Rainham route to somewhere on the Catford loop as an example. On the Southeastern side the only depot who can do Rainham to London via Abbey Wood without a driver change is Gillingham. The Southeastern drivers there don't sign 700s or even 707s. The Thameslink drivers based there only sign 700s so can't drive Southeastern trains or the route knowledge to enter the same driver pool as the Southeastern ones. At the north end the Luton drivers who sign Rainham don't sign the Catford loop (and neither do the Gillingham Thameslink drivers). Similarly the Catford loop services are reduced compared to pre-covid suggesting Orpington/Ashford Thameslink drivers aren't in plentiful supply either. In the hypothetical scenario there were Southeastern trains on the Catford loop to replace, the driver person days of training required will be many years worth requiring additional trained headcount to cover training for others. That additional trained headcount could instead be put immediately to use on the Thameslink route and be based at depots for Gillingham work to eliminate train crew related cancellations.
In the current scenario there aren't any Southeastern trains on the Catford loop to replace. The Beckenham Junction unit or driver requirement is a scale less than the Rainham Thameslink route so doesn't really work. Southeastern would have to short form trains or use 700s which would need to be rotated for maintenance somehow adding a potential source of cancellations. With using 700s probably end up with the Thameslink Gillingham drivers without Thameslink work running those leaving the unreliability very much in place. The new Catford loop services won't have any drivers so would probably end up even more unreliable than the unreliable "Southeastern" Rainham service using Thameslink stock and former Thameslink drivers. I'm struggling to see how this would help unless there was a massive increase in trained drivers available and then again, well before those are in place the Rainham route could be made very reliable with a smaller increase in drivers.
Given the number of times driver knowledge gets brought up, it feels like simplifying the network into a smaller number of metro-type operations would bring significant advantages. Break down the 24tph Thameslink Operations into the following 4tph operations with each one having a driver pool who signs everything within that operation:
1. 4tph Peterborough/Cambridge - Brighton via Quarry line all day
2. 4tph Bedford - Gatwick/Horsham via Redhill all day
3. 4tph MML slow - Sutton all day
4. 4tph MML slow - Catford Loop (Orpington and Sevenoaks) (ideally) all day
5. 4tph Cambridge / Welwyn ECML slow - Cats / Tats (ideally) all day (this doesn't necessarily need to wait for CARS)
6. 4tph Bedford - East Grinstead / Three Bridges peak only
Then the Southeastern lines should also be simplified to that, say the Dartford Lines and the Herne Hill services (whether Victoria all-day or Blackfriars peaks) are all resourced by respective driver pools.
1. 4tph Peterborough/Cambridge - Brighton via Quarry line all day
2. 4tph Bedford - Gatwick/Horsham via Redhill all day
3. 4tph MML slow - Sutton all day
4. 4tph MML slow - Catford Loop (Orpington and Sevenoaks) (ideally) all day
5. 4tph Cambridge / Welwyn ECML slow - Cats / Tats (ideally) all day (this doesn't necessarily need to wait for CARS)
6. 4tph Bedford - East Grinstead / Three Bridges peak only
Then the Southeastern lines should also be simplified to that, say the Dartford Lines and the Herne Hill services (whether Victoria all-day or Blackfriars peaks) are all resourced by respective driver pools.
The fewer heavily used sections that Thameslink uses, especially on routes where it is the minority user, the better that is for Thameslink operation. Rainham is a classic example: if Thameslink withdrew from Rainham any infrastructure failures between London Bridge and Rainham could no longer impact on Thameslink.
The end of route depot model directly contributes to unreliability, as Horsham-Peterborough and Brighton-Cambridge demonstrate frequently. It is a fundamental flaw in the Thameslink operation. Unfortunately I can't see a fix for it that is either quick or cheap.
This won't happen because it removes through trains between Brighton and St Albans/Luton/Bedford, a direct service that is well used and has been established for almost 40 years.
I would keep Sevenoaks linked to Welwyn but run all day at 2tph. The long term objective would be to add another 2tph linking Orpington with a new station at Crews Hill New Town. In the peak these would replace the existing Orpington trains that mostly run to Kentish Town/West Hampstead.
No, no, and thrice no! The Cambridge stoppers must never go through the core. Be grateful that this discussion is not dominated by the Cambridge-Maidstone disaster that fortunately never was. Apart from the 4tph fast service Thameslink's northern ECML boundary has to be south of the 2 track section at Welwyn.
I can't remember the details of why the Cats and Tats got left out of Thameslink in 2018 (and we got Rainham instead), but, if the planners didn't think it would work before the experience of the last 7 years, they sure aren't going to think its workable now, and still without CARS.
== Doublepost prevention - post automatically merged: ==
I hope that 24tph has been quietly buried in the grave next to Cambridge-Maidstone. 20tph should be the limit to give firebreaks in the peak timetable. I do agree that having driver pools that cover more services would help, though it will be costly. In FCC days Cambridge drivers always worked to Peterborough, and vice versa, and the loss of that flexibility has been damaging.
Last edited:
If any suggestion for improving Thameslink's performance that doesn't negatively impact on other TOCs' services then the only option left on the table is continuing with the current mediocrity.
The original Bedford - Brighton pairing was purely out of operational convenience. There's very little revenue driver that says Brighton is a better pairing for Bedford than another destination. The only north-south origin-destination pair that any bearing on Thameslink service pairing is Cambridge - Brighton. Pairing with Redhill routes still preserves direct Gatwick connectivity for Bedford. For those where directly trains matter above all else they can't demand journey time as well - that's just greedy.
The current 30-minute patterns create 10/20 service intervals everywhere, that's not conducive to good performance. Especially on the Catford Loop, a 15-minute pattern would be much better for fitting in Victoria - Maidstones and freights. It doesn't take a mathematical genius to work out, that to have consistent 15-minute patterns north and south of the river every 4tph service north of the river needs to be paired with the same 4tph service south of the river.
== Doublepost prevention - post automatically merged: ==
Having driver pools that cover a substantial service group wouldn't be costly if that service group is high-frequency and low geographic coverage. Cambridge and Peterborough plus Brighton and Horsham would be expensive; Cambridge and Peterborough plus just Brighton would be very reasonable.
- Status