What is the best remedy for the Coryton line's performance?

[Daniel Boone]

Moderator note: Split from

South Wales 'Metro' updates

Thanks, and yes a 231 was dragged to Holyhead by a 37 on March 16th because of the Canton lathe problem. Crewe's was (and still is?) out of action as well whereas Holyhead's was new just in time for leaf fall last autumn. A 231 and 756 left Canton behind a 37 today, presumably heading for...

www.railforums.co.uk

Ty-Glas is back open and looks like it’s back to the usual routine on the Coryton line this morning, run fast and delays. Interesting to see how much one additional stop can make to the timetable, throw in some crowds of commuters and it’ll be back to Birchgrove terminus.

Out of interest, what do you think the best remedy is for the Coryton line's performance?

 

[VR494]

Well before Ty Glas shut it was almost always there that they turned round at, not Birchgrove.
Even when they had more than enough time to get to Birchgrove and back.

There are a few operational input factors effecting the service, these include:

• Wrong traction - door's too slow under current operation
• Timetable too tight, only 30 second dwell time.
• ARS issues at Heath Jct, if +3 late on a 2PXX the route is given to 2EXX and 2YXX services needing manual intervention from a signaller. There is a high workload on the panel, and can be missed while dealing with other trains.
• Ty Glas station having only 2 metre tolerance on the extention, 1 metre each end, meaning 756 trains only just fit. This needs a cautions and slow approach to the PSM.

 

[VR494]

I know it wasn't me you asked, but...

I just know what I've seen as a passenger and I probably don't understand how things work, but from my point of view it would help if it was possible to...

1) Not turn late runing trains round so early that they then sit for 15 minutes before heading back again
2) Let a Coryton train that's a few minutes late out onto the main line rather than having to wait until after the next train from Caerphilly has passed. There is quite a small window between the timetabled departure and the point at which the junction gets set for the main line, even though it's then quite a while before the train on the main line actually passes and the train from Coryton could have got out first without delaying it. It could then make up time because of the additional time in the timetable between Heath Low Level and Queen Street. Instead it's late off the branch, delaying the next one onto the branch, and then it repeats.

Reliability seems to have been a lot better since the 756s came in. It's not obvious to me why given that they don't seem to save any time on the Coryton branch due to the extended dwell times. Maybe it's because Ty Glas closed at the same time.

Before that there seemed to be a period when trains were allowed to run late on the Coryton Line rather than being cut short at the first sign of delay. Maybe it messed up the rest of the network too much but it didn't seem to cause the Coryton Line itself to collapse.

I don't suppose this would be acceptable after all that work extending the platform, but if having Ty Glas open does wreck performance is the answer to close it again?

== Doublepost prevention - post automatically merged: 4 Apr 2026 ==



Under current operation indeed. Which suggests to me that the problem is the operation, not the traction itself.

The company has stated it has intention to go driver open/close on these on CVL, this is still some time away.

== Doublepost prevention - post automatically merged: 4 Apr 2026 ==

I do wonder how much more it would have cost to have made it a bit longer. There seems to be plenty of space at the end.

It cannot be extended any further due to underground utilties, it just meets the RIS 3782 TOM regualtions.

 

[positron]

Out of interest, what do you think the best remedy is for the Coryton line's performance?

A passing loop on that line and the Penarth line so the turnaround times aren't so tight. Though Heath junction can probably be improved. Unfortunately I don't think there's room to extend the single lead junction onto the Coryton line.

One thing that could help dwell times a bit is if the flirts allowed the platform filler steps to come out when the doors are released rather than when the open button is pressed. I'm guessing the same 'issue' exists on 398s too.

 

[vikingsmb]

the junction between the rhymney line and the coryton line was moved in 1984, it was orginally double track for short length then single track (prior to 1984) and after 1984 single track

The line needs doubling between heath low level and say rhiwbina or even whitchurch

 

[31160]

A passing loop on that line and the Penarth line so the turnaround times aren't so tight. Though Heath junction can probably be improved. Unfortunately I don't think there's room to extend the single lead junction onto the Coryton line.

One thing that could help dwell times a bit is if the flirts allowed the platform filler steps to come out when the doors are released rather than when the open button is pressed. I'm guessing the same 'issue' exists on 398s too.

That is a very good idea with the step coming out, extend that to every class with the step thingy

 

[VR494]

That is a very good idea with the step coming out, extend that to every class with the step thingy

It can be done on the flirts but needs a software modification from Stadler. If the drivers release the doors, the steps already come straight out.

Regarding the Coryton line, tfw should have made this the initial tram-train route to Cardiff Bay. It is lines like this these 398 are designed for. They must have a good councillor to demand the use of inter-regional trains with toilets for stops within walking distance of each other.

 

[AdamWW]

It can be done on the flirts but needs a software modification from Stadler. If the drivers release the doors, the steps already come straight out.

Presumably in principle to save wasted ramp movement they could arrange it so that if a door button is pressed before the doors are released, the ramp extends but the door doesn't open until it can.

Regarding the Coryton line, tfw should have made this the initial tram-train route to Cardiff Bay. It is lines like this these 398 are designed for. They must have a good councillor to demand the use of inter-regional trains with toilets for stops within walking distance of each other.

Perhaps I don't understand how these things work but it seems a little far fetched to me that a local councillor would be able to successfully exert this level of influence over TfW's plans.
In any case it would be a pretty large slap in the face to commuters on the Coryton Line to not only fail to give them any ultimate increase in frequency unlike almost every other core valley route but also make everyone going to Central change trains too.

I don't see what stop spacing has to do with it anyway, as opposed to the potential total journey time (which I quite agree doesn't justify a 756).

 

[CapitalStar378]

Presumably in principle to save wasted ramp movement they could arrange it so that if a door button is pressed before the doors are released, the ramp extends but the door doesn't open until it can.



Perhaps I don't understand how these things work but it seems a little far fetched to me that a local councillor would be able to successfully exert this level of influence over TfW's plans.
In any case it would be a pretty large slap in the face to commuters on the Coryton Line to not only fail to give them any ultimate increase in frequency unlike almost every other core valley route but also make everyone going to Central change trains too.

I don't see what stop spacing has to do with it anyway, as opposed to the potential total journey time (which I quite agree doesn't justify a 756).

I've been home to Rhiwbina this weekend to see the folks and was surprised how busy the 756s were, obviously it's Easter but when a single 153 used to be sent up to Coryton, they were often barely 1/3 full. I think that having comfortable, spacious and quiet trains has increased ridership. 4tph would do even more to boost numbers, bring on the passing loop and M4 J32 Park & Ride too.

Got the train back from Queen St yesterday and there were delays due to an incident at Central. The Coryton train was changed to a Heath Low Level terminator, which meant it went one stop only. The next train was showing as cancelled on the NR app, asked the staff to check and they said it was going to be cancelled but they changed their minds. The Coryton service does seem to be the sacrificial lamb, this needs to be changed if TfW want to continue to attract more users.

 

[AdamWW]

I've been home to Rhiwbina this weekend to see the folks and was surprised how busy the 756s were, obviously it's Easter but when a single 153 used to be sent up to Coryton, they were often barely 1/3 full. I think that having comfortable, spacious and quiet trains has increased ridership. 4tph would do even more to boost numbers, bring on the passing loop and M4 J32 Park & Ride too.

It does seem to me that ridership has gone up significantly with the introduction of 756s, but I might be seeing what I want to.

I would expect 4 tph to make quite a difference - half an hour between trains isn't very attractive for short journeys.
Though having it (almost) stay half hourly until the end of service is a definite improvement.

Got the train back from Queen St yesterday and there were delays due to an incident at Central. The Coryton train was changed to a Heath Low Level terminator, which meant it went one stop only. The next train was showing as cancelled on the NR app, asked the staff to check and they said it was going to be cancelled but they changed their minds. The Coryton service does seem to be the sacrificial lamb, this needs to be changed if TfW want to continue to attract more users.

Yes indeed. It does seem that the Coryton Line will be sacrificed to avoid even a small risk of impacting the service on the rest of the network - even if that means turning several successive trains round short.
Once after two services were completely cancelled due to wider disruption, the following train was turned round at Ty Glas because it was 7 minutes late - to then sit there for 12 minutes before heading back.

 

[positron]

I suspect that the 756s running on Coryton line is only going to last as long as it takes for the passing loop to be added and then it'll be swapped to 398s, like you say that sort of line is perfect for the 398s (as envisaged obviously they're currently perfect for nowhere). I suspect it'll also help extension business cases for it to be able to be built to lighter rail standards.

Probably more discussion for the more speculative thread though: https://www.railforums.co.uk/threads/south-wales-metro-future.258538/


== Doublepost prevention - post automatically merged: 5 Apr 2026 ==

Regarding Cardiff west one thing I still don't understand is what the service pattern is that gets you 6tph to the bay, and 6tph to central but also 4tph on the city line. I can work out that you can get 3tph by just sending one of the treherberts the other way and looping both of them. But there's a missing 1tph somewhere.

 

[positron]

Any updates on any of the remaining infrastructure work needed for the original vision of Metro to be completed? I think there's Crwys Road station, an extended loop at Tir Phil, and a 2nd platform at Rhymney still go to on that line; and double tracking on and the new station at Butetown on the bay line. As far as I know everything else is done infrastructure wise?

Obviously there's the now committed projects: phase 1a of Crossrail and the added 'missing' track at Cardiff West, but they're planned for after the initial works finish as far as I understand.

 

[MikePJ]

Regarding Cardiff west one thing I still don't understand is what the service pattern is that gets you 6tph to the bay, and 6tph to central but also 4tph on the city line. I can work out that you can get 3tph by just sending one of the treherberts the other way and looping both of them. But there's a missing 1tph somewhere.

My understanding was that the plan was for the existing 2tph loop service on the City Line to be augmented by an 2tph extra Radyr-Central-Coryton service. That does require a passing loop on the Coryton line, but there was a mystery sum of money allocated to the “core valley lines” in the spending review announcement that included Cardiff West, so the two might be done at the same time.

 

[AdamWW]

My understanding was that the plan was for the existing 2tph loop service on the City Line to be augmented by an 2tph extra Radyr-Central-Coryton service. That does require a passing loop on the Coryton line, but there was a mystery sum of money allocated to the “core valley lines” in the spending review announcement that included Cardiff West, so the two might be done at the same time.

Initial electrification work put things in the way of providing a loop on the Coryton Line, but when it came to actually putting up catenary supports they seem to have been placed to give plenty of space for one. And the stuff that is in the way seems to be raised trunking that would have been used for 25 kV cabling through a permanently earthed section that never happened so presumably wouldn't cause any trouble if removed.

 

[positron]

My understanding was that the plan was for the existing 2tph loop service on the City Line to be augmented by an 2tph extra Radyr-Central-Coryton service. That does require a passing loop on the Coryton line, but there was a mystery sum of money allocated to the “core valley lines” in the spending review announcement that included Cardiff West, so the two might be done at the same time.

I agree and one of the MP in Cardiff even overtly says that 48m is partly for the loop. Issue with that 2tph from Coryton using the passing loop is that still leaves Treherberts reversing at Central which isn't ideal.

 

[AdamWW]

There are a few operational input factors effecting the service, these include:

• Wrong traction - door's too slow under current operation
• Timetable too tight, only 30 second dwell time.
• ARS issues at Heath Jct, if +3 late on a 2PXX the route is given to 2EXX and 2YXX services needing manual intervention from a signaller. There is a high workload on the panel, and can be missed while dealing with other trains.
• Ty Glas station having only 2 metre tolerance on the extention, 1 metre each end, meaning 756 trains only just fit. This needs a cautions and slow approach to the PSM.

One more thing - I don't think the 5 mph temporary speed restriction in Coryton station is helping much.

I don't know if it is temporary or if it's actually going to be permanent but they haven't put the proper signage up yet.
I hope the former.

== Doublepost prevention - post automatically merged: 7 Apr 2026 ==

There are a few operational input factors effecting the service, these include:

• Ty Glas station having only 2 metre tolerance on the extention, 1 metre each end, meaning 756 trains only just fit. This needs a cautions and slow approach to the PSM.

Is that "just fitting" as in a passenger door will be off the platform if more than a metre out, or is that to get the whole train fully alongside the platform?
If the latter, does it matter if the train overhangs a bit so long as the passenger doors don't?
There is of course a lot more train between the outermost doors and the ends than there is on a 153. (Getting a 2 car 153 to fit the old Ty Glas platform needed some precision).

 

[MikePJ]

TfW have finally published board meeting minutes from last year, up until December 2025 (after I nudged them with an FOI request) - here are a few highlights:

September 2025, TfW Rail board:

"Work continues with legal teams on drafting specification requirements for fleet procurement, including depot needs" [then followed by several redacted sections]

October 2025, TfW Rail board:

The Board discussed accelerating the Class 398 EIS [entry into service] critical path to March 2026 by streamlining
sign offs and isolating the Treherbert diagram to reduce training needs, with rest day working
potentially enabling readiness by mid-February. Software delays and testing risks were
acknowledged, particularly around multi-unit testing, brake tests, and pilot course assurance
scheduled for 18 November. Parallel deployment of Class 756s to Cardiff Bay was discussed as
a route to achieving performance targets, pending professional head sign offs and resource
checks. Toilet access compliance and timetable implications were raised, with shuttle services
and interspersed diagrams considered. [redacted section]
Strategic direction leaned toward accelerating rollout, removing Class 150s, and cascading Class 231s, requiring
ELT support and stakeholder engagement. Financial implications of recommended options
were noted, with a [redacted] acceleration cost advised.

October 2025, TfW board:

The Board discussed TfW’s rolling stock strategy with an emphasis on fleet resilience in the context of a
changing timetable to enable growth and reduce government subsidy. The Board noted that over the next three
years the timetable and associated fleet requirements were sufficient to deliver the advertised service, but
there is an element of risk when bringing in new fleet. Whilst resilience and contingency are required, a balance
needs to be achieved and any contingency will come at a cost which is unlikely to pay for itself.

There are then a number of redacted bullet points, but of the unredacted ones the most interesting are that the Board agreed that:

• 20 Class 153 Units are brought into operation (17 leased through Porterbrook and three TfW owned) as quickly as possible to provide risk mitigation between now and May 2027
• Proceed with all the reliability improvements necessary for the MKIV fleet including progressing the staffing/technical/night shift requirements for day-to-day resilience/reliability
• The Board noted the potential capacity issues in the early 2030s and particular issues around requirements for the latter phases of Crossrail. It was agreed to draft a letter to local authorities in the Cardiff Capital Region setting out the regional benefits and to discuss rolling stock strategy

December 2025, TfW board:

A recent meeting between TfW, CAF and CAF rolling stock owners which has led to five further units
being planned for acceptance before Christmas and a full repair of recent issues which will see six
additional units back in service by the end of January 2026.

 

[Envoy]

If they got on and extended the line the short distance from Coryton station to a new station to serve the under construction Velindre Hospital, then that would obviously mean that the timetable would have to be re-cast and trains run the full length of the branch. (Obviously the much more expensive extension over the Taff to connect with Radyr would be the big winner in terms of connectivity).

 

[anthony263]

If they got on and extended the line the short distance from Coryton station to a new station to serve the under construction Velindre Hospital, then that would obviously mean that the timetable would have to be re-cast and trains run the full length of the branch. (Obviously the much more expensive extension over the Taff to connect with Radyr would be the big winner in terms of connectivity).

Coryton to Radyr extension is near enough impossible ive seen a few videos done might look close on the maps however the amount of work including landscaping eyc means we looking at quite a bit of money

 

[Llanigraham]

It is interesting that people seem to think that a short extension beyond Coryton station to the new Velindre Hospital would increase passenger numbers to any great degree. Remember this is going to be a major cancer treatment centre, with patients who will be immunity supressed so they will not, and are not encouraged, to use public transport to get to and from their treatment. It puts them at too much risk. I take patients their regularly and even though their is a bus route right outside the current unit very few people use it for that reason, instead they use their own transport or one of the services provided by Welsh Ambulance.

 

[MikePJ]

It is interesting that people seem to think that a short extension beyond Coryton station to the new Velindre Hospital would increase passenger numbers to any great degree. Remember this is going to be a major cancer treatment centre, with patients who will be immunity supressed so they will not, and are not encouraged, to use public transport to get to and from their treatment. It puts them at too much risk. I take patients their regularly and even though their is a bus route right outside the current unit very few people use it for that reason, instead they use their own transport or one of the services provided by Welsh Ambulance.

It's not so much patients who generate hospital traffic as staff. Velindre employs around 4,000 people according to their annual report. If you live near a hospital you tend to be aware of a spike in traffic around shift changes too.

== Doublepost prevention - post automatically merged: 9 Apr 2026 ==

Coryton to Radyr extension is near enough impossible ive seen a few videos done might look close on the maps however the amount of work including landscaping eyc means we looking at quite a bit of money

Yeah, I still wonder if what we'll end up with is an extension from Velindre to a terminus in Forest Farm Country Park near the Radyr footbridge. Passengers could then walk across to change trains and it avoids the performance issues with a circular service. However, having walked that route I think you'd get a lot of opposition to running trams through a nature reserve.

 

[anthony263]

It's not so much patients who generate hospital traffic as staff. Velindre employs around 4,000 people according to their annual report. If you live near a hospital you tend to be aware of a spike in traffic around shift changes too.

== Doublepost prevention - post automatically merged: 9 Apr 2026 ==


Yeah, I still wonder if what we'll end up with is an extension from Velindre to a terminus in Forest Farm Country Park near the Radyr footbridge. Passengers could then walk across to change trains and it avoids the performance issues with a circular service. However, having walked that route I think you'd get a lot of opposition to running trams through a nature reserve.

Actually using tram-trains could be a better way to sell such an extension

 

[Envoy]

Any extension from Coryton / new Velindre station would not go through the Forest Farm Nature Reserve in order to reach Radyr. The line would continue in the old cutting done by the Victorians to pass under Longwood Drive (near Asda) and then west just on the south side of the M4 to reach a new bridge over the Taff and connect with the central valley lines and hence Radyr and make possible a complete circuit of Cardiff suburbs.

Go on Google Maps and drag the yellow man to the cutting to see that Google have walked through most of the cutting.

 

[MikePJ]

I was suggesting the use of Forest Farm as a terminus as an alternative to the high cost and complexity of a bridge over the Taff, which is the ideal solution but may not survive cost-benefit analysis.