Ideas to solve Thameslink reliability by reducing its network

[Bald Rick]

Are there four platforms at Sutton, with two of those for the Wimbledon direction?

Strictly speaking, there is only one platform to Wimbledon, Platform 2. And only 1 platform from Wimbledon, Platform 1. Neither of them permits a train from Wimbledon to reverse and go back where it came from.

On top of this base off-peak frequency peak extras could be added as needed, originating from Blackfriars northbound or Kings Cross southbound.

That would mean terminating trains in the core, which is an absolute no-no.

I will most likely post this again in a new thread in a few days with my changes to Southern, Great Northern, SWR, EMR and Southeastern also explained, but here is my idea for Thameslink.

Changes I have made:

The biggest change I have made is to convert the service north of the core to a 6tph base timetable instead of 4tph. This has many benefits, including that Brighton services can be equally spaced every 15 mins from the core southwards and still run to both Bedford and Cambridge, and skip-stopping services to Luton (currently the Rainham services) that provide additional services to some stations but skipping others will not be needed anymore, simplifying service patterns.

Sutton via Wimbledon services run into the Blackfriars bays, as these services are unnecessary anyway since the Northern Line to E&C is much faster than Thameslink from the Morden and Tooting areas.

My plan requires a new 12 coach Platform 3 at Reigate to be built.

New services run to West Croydon via Tulse Hill and Crystal Palace, providing new connectivity along the corridor of the 68/468/X68 bus routes which are very popular.

Services run all stations to both Welwyn and Hertford, with both lines having an equally spaced 4tph that alternates between Thameslink and Moorgate. Some platform extensions will be needed on these routes.


Service pattern (stations in core not shown as it would be repetitive):

Bedford to Brighton 2tph (12 coach)
(Flitwick, Harlington, Leagrave, Luton, Luton Airport Parkway, Harpenden, St Albans, Brent Cross, West Hampstead...London Bridge, East Croydon, Gatwick Airport, Haywards Heath, Wivelsfield, Burgess Hill, Hassocks, Preston Park)

Cambridge to Brighton 2tph (12 coach)
(Royston, Baldock, Letchworth, Hitchin, Stevenage, Finsbury Park...London Bridge, East Croydon, Horley, Gatwick Airport, Three Bridges, Balcombe, Haywards Heath, Burgess Hill)

Bedford to Horsham 2tph (12 coach)
(Flitwick, Harlington, Leagrave, Luton, Luton Airport Parkway, Harpenden, St Albans, Brent Cross, West Hampstead...London Bridge, Norwood Jct, East Croydon, Purley, Coulsdon South, Merstham, Redhill, Earlswood, Salfords, Horley, Gatwick Airport, Three Bridges, Crawley, Ifield, Littlehaven)

Bedford to Reigate 2tph (12 coach)
(Flitwick, Harlington, Leagrave, Luton, Luton Airport Parkway, Harpenden, St Albans, Brent Cross, West Hampstead...London Bridge, Norwood Jct, East Croydon, Purley, Coulsdon South, Merstham, Redhill)

St Albans to Sutton 2tph (8 coach)
(Radlett, Elstree & Borehamwood, Mill Hill Broadway, Hendon, Brent Cross, Cricklewood, West Hampstead, Kentish Town...E&C, Loughborough Jct, Herne Hill, Tulse Hill, Streatham, Mitcham Eastfields, Mitcham Jct, Hackbridge, Carshalton)

St Albans to West Croydon 4tph (8 coach)
(Radlett, Elstree & Borehamwood, Mill Hill Broadway, Hendon, Brent Cross, Cricklewood, West Hampstead, Kentish Town...E&C, Loughborough Jct, Herne Hill, Tulse Hill, West Norwood, Gipsy Hill, Crystal Palace, Norwood Jct)

Welwyn to Sevenoaks 2tph (8 coach)
(Hatfield, Welham Green, Brookmans Park, Potters Bar, Hadley Wood, New Barnet, Oakleigh Park, New Southgate, Alexandra Palace, Hornsey, Harringay, Finsbury Park...E&C, Denmark Hill, Peckham Rye, Nunhead, Crofton Park, Catford, Bellingham, Beckenham Hill, Ravensbourne, Shortlands, Bromley South, Bickley, St Mary Cray, Swanley, Eynsford, Shoreham, Otford, Bat & Ball)

Hertford North to Orpington 2tph (8 coach)
(Bayford, Cuffley, Crews Hill, Gordon Hill, Enfield Chase, Grange Park, Winchmore Hill, Palmers Green, Bowes Park, Alexandra Palace, Hornsey, Harringay, Finsbury Park...E&C, Denmark Hill, Peckham Rye, Nunhead, Crofton Park, Catford, Bellingham, Beckenham Hill, Ravensbourne, Shortlands, Bromley South, Bickley, Petts Wood)


Order of services and intervals between each service in the core (not the actual times):

00 Bedford to Brighton
05 Hertford to Orpington
07.5 St Albans to Sutton via Mitcham
10 Bedford to Horsham
15 Cambridge to Brighton
17.5 St Albans to West Croydon via Crystal Palace
20 Bedford to Reigate
25 Welwyn to Sevenoaks
27.5 St Albans to West Croydon via Crystal Palace
30 Bedford to Brighton
35 Hertford to Orpington
37.5 St Albans to Sutton via Mitcham
40 Bedford to Horsham
45 Cambridge to Brighton
47.5 St Albans to West Croydon via Crystal Palace
50 Bedford to Reigate
55 Welwyn to Sevenoaks
57.5 St Albans to West Croydon via Crystal Palace

This is another reason why my idea which changes the service north of the core to a 6tph base timetable (6tph to Bedford (fast), St Albans (slow) and the ECML) is a good idea.

It might seem a good idea. But it won’t work on the MML. Unless you want to extend all EMR journey times by at least 6 minutes, probably more. Also the St Albans turn back is 4tph max.


There’s a reason why the timetable is the compromise it is - and that’s because it’s a very, very, difficult exercise that took many people the best part of a decade to develop.

As an aside, why has it always been (even before the ECML TL) that all Bedford services call at all stops to Luton?

They don’t, in the peak, some run fast-ish.
Off peak providing Flitwick and Leagrave with 4tph is justified by the custom. Less so at Harlington but it does make for a standard repeating timetable which helps with Bedford platforming.

 

[Magdalia]

This is another reason why my idea which changes the service north of the core to a 6tph base timetable (6tph to Bedford (fast), St Albans (slow)

There's not enough reversing capacity at Bedford or St Albans for 6ph. At Bedford there's not enough platform capacity. At St Albans 6tph in a single reversing siding is not practical: checking RTT each reversing move is booked 13 minutes. But 4tph at each of Bedford, Luton and St Albans is feasible.

 

[Bald Rick]

At Bedford there's not enough platform capacity.

There is enough for 6tph at Bedford - it happens every morning and evening. Indeed it used to be 8tph!

Off peak is admittedly more difficult because of the freight paths.

 

[Magdalia]

There is enough for 6tph at Bedford - it happens every morning and evening. Indeed it used to be 8tph!

Off peak is admittedly more difficult because of the freight paths.

With two dedicated platforms and no other trains in the way it is possible to turn round 6tph, though has 8tph been done with 12 car class 700s, which require longer turnrounds than 8 car class 319s? Given that the dwell in the St Albans reversing sidings is 9 minutes for an 8 car class 700, then the dwell for a 12 car will be 1 or 2 minutes longer, and very tight to have a total processing time of less than 15 minutes, which would be essential for 8tph in 2 platforms.

Moorgate does 8tph in 2 platforms but the trains are only 6 cars. Moorgate has done 12tph with class 313s in the past, but I expect they had shorter setup times than class 717s.

 

[Bald Rick]

With two dedicated platforms and no other trains in the way it is possible to turn round 6tph, though has 8tph been done with 12 car class 700s, which require longer turnrounds than 8 car class 319s?

yes, done regularly with the 700s.

Given that the dwell in the St Albans reversing sidings is 9 minutes for an 8 car class 700, then the dwell for a 12 car will be 1 or 2 minutes longer,

Others may correct me, but the 9 minutes at St Albans is a legacy from the 319s, when drivers had to do a shuffle outside via the cab steps. It’s also about fitting the turnaround in with available paths.

A 12 car 700 can easily turn round in 6-7 minutes with the same driver. I’ve seen it done in 4. Bedford is typically planned at 14, but several examples at 10.

 

[Magdalia]

A 12 car 700 can easily turn round in 6-7 minutes with the same driver.

I’ve seen it done in 4.

I don't believe 4 minutes. When I was a young and fit commuter it still took me 10 seconds per car to walk the length of a train, so the walking time alone for a 12 car train is at least 2 minutes.

And turnround time is only part of the equation: there's reoccupation time too. That depends on the positioning of the signals and whether there are conflicting moves. Being able to reverse between the running lines, as at St Albans and Luton, is a big advantage.

 

[nw1]

That would mean terminating trains in the core, which is an absolute no-no.

I was thinking they would then go ECS somewhere (or come out of ECS) which would mean they wouldn't have to reverse.

It might seem a good idea. But it won’t work on the MML. Unless you want to extend all EMR journey times by at least 6 minutes, probably more. Also the St Albans turn back is 4tph max.


There’s a reason why the timetable is the compromise it is - and that’s because it’s a very, very, difficult exercise that took many people the best part of a decade to develop.

Based on the recent discussions here, I do wonder why they didn't do something like this pattern though:

4tph Bedford-Brighton
4tph Luton-Wimbledon loop
4tph St Albans (or Kentish Town?)-Catford route
4tph Cambridge or Peterborough-Three Bridges or Horsham via Redhill

as that would be more consistent and simpler, and thus, presumably, easier to timetable. As someone else said, an even 15-min service down to Brighton would make sense to me and keeping the old 15-min Bedford-Brighton route would be the most logical way of doing this.

They don’t, in the peak, some run fast-ish.

Sorry, meant the standard off peak pattern.

Off peak providing Flitwick and Leagrave with 4tph is justified by the custom. Less so at Harlington but it does make for a standard repeating timetable which helps with Bedford platforming.

I don't know Flitwick but it looks like a large village/small town so surprised it would warrant a metro-style frequency. Leagrave is a suburb of Luton so I can see 4tph might be justified. However I take your point about the ease of timetabling, and Bedford has the Corbys now anyway so they have a faster option.

 

[JonathanH]

As someone else said, an even 15-min service down to Brighton would make sense to me and keeping the old 15-min Bedford-Brighton route would be the most logical way of doing this.

On the contrary, having off pattern departures from Brighton makes fitting in the rest of the services more straightforward. Departures from Brighton were never 15 minutes apart when all the trains went to Bedford, although they were 15 minutes apart between Bedford and Gatwick Airport off-peak once Redhill was dropped.

As others have noted, Cambridge to Brighton is a successful combination from a passenger perspective, possibly more lucrative than running 4tph Bedford to Brighton.

Of course, the primary issue that we are looking to fix is the inherent reliability issue of running anything through Thameslink from the GN. Changing the destinations at the southern end doesn't really help. Indeed, it can be argued that having one GN and one MML destination on both the Redhill and Brighton routes should help with resilience.

 

[Magdalia]

I don't know Flitwick but it looks like a large village/small town so surprised it would warrant a metro-style frequency.

Flitwick is main station for Ampthill, and is also an alternative to Arlesey for Shefford.

 

[Richardr]

Flitwick always seems the most used of the stations between Luton and Bedford. Leagrave is important as it is the nearest station, a short walk from the Luton and Dunstable hospital.

 

[Minstral25]

There’s a reason why the timetable is the compromise it is - and that’s because it’s a very, very, difficult exercise that took many people the best part of a decade to develop.

Exactly - along Redhill route we had hours of conversations with the team and got some good compromises but ultimately some bad. Almost 6 months of rubbish timetable to get the Drivers and training sorted after it was supposed to be implemented didn't look good but shows how difficult it is.

Of course, the primary issue that we are looking to fix is the inherent reliability issue of running anything through Thameslink from the GN. Changing the destinations at the southern end doesn't really help. Indeed, it can be argued that having one GN and one MML destination on both the Redhill and Brighton routes should help with resilience.

MML through to Thameslink isn't much better but agree changing it all round would lose the little bit of resilience there.

Any changes are more likely to be cosmetic like 1 or 2 non-destination station stops being swapped than major route changes.

 

[CarrotPie]

My thoughts for the Great Northern network:

• Moorgate lines use 717s to run local services
• The rest of Great Northern uses 379s and 387s to run semi-fast/express services
• Thameslink uses 700s through the Canal Tunnels to run local/semi-fast services

All lines on the map represent 2tph

 

[Basil Jet]

My thoughts for the Great Northern network:

• Moorgate lines use 717s to run local services
• The rest of Great Northern uses 379s and 387s to run semi-fast/express services
• Thameslink uses 700s through the Canal Tunnels to run local/semi-fast services

All lines on the map represent 2tph

Where did this map come from? It shows half of the Moorgate Welwyns skipping Hatfield.

 

[SE%Traveller]

I wasn't looking at the current timetable, which is awful given what could be achieved. I was looking at optimal times for running both 4tph fast and 4tph stopping services along the line.

With optimal paths, from Denmark Hill to Bromley South, fast services take 12 minutes and stopping services take 24 minutes.

If there were 4tph Thameslink stopping services evenly spread every 15 mins, then a fast service leaving Denmark Hill directly before a stopping service will catch up to the next stopping service.

It's a nice idea but others with greater knowledge than I advise that 6tph is the max that can be regularly achieved on the Catford Loop outside peaks. Main consideration is freight (which isn't timetabled to use in in the peaks) but then there is also the Victoria Dartford trains too threading in for the stretch Nunhead to Peckham Rye.

Otherwise I think they could just schedule on 15 min intervals with the Fasts and Slows timetabled to depart Bromley South to London at the same time as they do with the SE Blackfriars and Thameslink Luton trains. They do the same via Herne Hill too, Slows leaving at the same time as the Fasts from Maidstone East and Chatham with the latter overtaking the former at Shortlands.

That said i suppose the Freight itself could be the cause of the 10/20 difference...

 

[CarrotPie]

Where did this map come from? It shows half of the Moorgate Welwyns skipping Hatfield.

I made the map myself; it's based on this, which I made from the May 2021 GN timetable.

Plus the GN semi-fast Peterboroughs also call at Hatfield, so it gets more service than now anyway.

 

[CarrotPie]

Having GN run semi-fasts to Cambridge and Peterborough still gives lots of places a London service if/when TL services don't run

(also GN semis can make extra stops to replace TL and/or TL can divert to KGX)

 

[Hadders]

My thoughts for the Great Northern network:

• Moorgate lines use 717s to run local services
• The rest of Great Northern uses 379s and 387s to run semi-fast/express services
• Thameslink uses 700s through the Canal Tunnels to run local/semi-fast services

All lines on the map represent 2tph

It looks like you've obliterated the fast service from Stevenage to London.

 

[CarrotPie]

It looks like you've obliterated the fast service from Stevenage to London.

Admittedly it does now make stops at Knebworth and Finsbury Park, but LNER and Lumo run non-stop to KGX anyway, so two extra stops shouldn't be a problem

 

[Trainbike46]

My thoughts for the Great Northern network:

• Moorgate lines use 717s to run local services
• The rest of Great Northern uses 379s and 387s to run semi-fast/express services
• Thameslink uses 700s through the Canal Tunnels to run local/semi-fast services

All lines on the map represent 2tph

It looks like you've obliterated the fast service from Stevenage to London.

and the usefulness of Cambridge-Brighton is lower if you link it to the slow cambridge service instead of the semi-fast service

 

[Basil Jet]

I made the map myself; it's based on this, which I made from the May 2021 GN timetable.

Plus the GN semi-fast Peterboroughs also call at Hatfield, so it gets more service than now anyway.

I'm looking at the May 2021 GN timetable, and I can't see any Moorgate Welwyns skipping Hatfield. There are trains skipping Brookmans Park and Welham Green, so you have misplaced a notch from Hatfield to Brookmans Park.

 

[Hadders]

Admittedly it does now make stops at Knebworth and Finsbury Park, but LNER and Lumo run non-stop to KGX anyway, so two extra stops shouldn't be a problem

LNER doesn't operate between Stevenage and Kings Cross in the morning peak, and would have insufficuient capacity if too many passengers switched from Thameslink. Lumo are pick-up and set down to/from the north only at Stevenage.

Also, the fast service between Stevenage and London has already been slowed down in recent years and adding an additional call at Knebworth would only make it slower.

 

[CarrotPie]

Changes to the service:

• TL Cambridge-Brighton resumes its current stopping pattern, but with Ashwell and Morden replaced with Welwyn GC and Meldreth
• TL Peterborough-Horsham resumes its current stopping pattern
• GN Peterborough Express' calls at Potters Bar, Hatfield and Welwyn GC have been moved to the Cambridge service, making this non-stop Stevenage-KGX
• GN Cambridge semi-fast now runs all stations from Hatfield, with an additional stop at Potters Bar.

Changes to the map:

• Made text bigger
• Removed duplicate names at Welwyn North and Knebworth making them appear darker
• Swapped back Horsham and Brighton as they were assigned incorrectly

Thoughts on the Welwyn GC-Sevenoaks TL stopping at Harringay and Hornsey?

== Doublepost prevention - post automatically merged: 27 Sep 2022 ==

I'm looking at the May 2021 GN timetable, and I can't see any Moorgate Welwyns skipping Hatfield. There are trains skipping Brookmans Park and Welham Green, so you have misplaced a notch from Hatfield to Brookmans Park.

Thanks for that. I've also removed the stops at Harringay and Hornsey from that semi-fast Welwyn GC, as per real life.
I'm not going to attach a third version of the map until I've added the MML, and possibly the services south of the Thames.

 

[JonathanH]

Thoughts on the Welwyn GC-Sevenoaks TL stopping at Harringay and Hornsey?

No. Putting more stops blocks up capacity on the route for no good reason.

TL Cambridge-Brighton resumes its current stopping pattern, but with Ashwell and Morden replaced with Welwyn GC and Meldreth

Putting calls at Welwyn Garden City on services otherwise running fast is asking for trouble. Essentially your suggestion is that a train runs immediately behind the train up to Welwyn Garden City then overtakes while the stop is taken. That eats up fast line capacity and has to work right time every time. It is perhaps unfortunate that Welwyn Garden City can't have fast services to London / Thameslink but that is how it is.

 

[CarrotPie]

No. Putting more stops blocks up capacity on the route for no good reason.


Putting calls at Welwyn Garden City on services otherwise running fast is asking for trouble. Essentially your suggestion is that a train runs immediately behind the train up to Welwyn Garden City then overtakes while the stop is taken. That eats up fast line capacity and has to work right time every time. It is perhaps unfortunate that Welwyn Garden City can't have fast services to London / Thameslink but that is how it is.

Would it perhaps work if said service also called at Hatfield, giving others even more time to pass it, then rejoined the fasts at Marshmoor?

 

[CarrotPie]

Or I assume a simple Welwyn - (Finsbury) - KGX could theoretically run in the peaks (though I won't add it)

 

[JonathanH]

Would it perhaps work if said service also called at Hatfield, giving others even more time to pass it, then rejoined the fasts at Marshmoor?

Whichever way it is done it essentially requires two paths on the fast lines instead of one. It is why, on intensively used routes, fast to slow to fast switches are rarely timetabled until something else has branched off.

 

[dodecahedron67]

Surely the easiest way to ease up some (8-car) trains for a few switchings would be to swap the now new southeastern Cannon St - Gravesend and Rainham TL over so it becomes Cannon St - Rainham and TL - Gravesend?

 

[Jason48]

I thought that Thameslink were giving up the Rainham route eventually. Only keeping the pathways from London Bridge to Luton. Thus returning pathways back to Southeastern from North Kent East Jn to Gillingham

 

[cle]

I thought that Thameslink were giving up the Rainham route eventually. Only keeping the pathways from London Bridge to Luton. Thus returning pathways back to Southeastern from North Kent East Jn to Gillingham

They should both run from Cannon St, to be honest.

The Luton-London Bridge path is 12 car so a bit of a waste on an 8 - it could bounce a Victoria service if Windmill is full (or, in an ideal world, 1tph East Grinstead, 1tph Uckfield!) - or take on the new Maidstone fast, with something else from Charing Cross. Or the Epsom fasts out of LB.

 

[CarrotPie]

I think Thameslink should ditch Rainham and replace it with Maidstone East (SE already runs MDE-BFR). This frees up a few 8-car units for use always.

Unrelated, I feel like extending the Gatwick Airport service could be extended to Horsham, giving more capacity, with the Bedford-Horsham service that would create being extended via Arundel to Littlehampton in the peaks.