TfW class 398 Stadler Citylink tram trains

[SWML9102]

I think hitting a bridge while raised would probably explain the damage. If you didn't know what a pantograph is supposed to look like I think it would be hard to figure it out from what was left.

Once the offending 398 was out of the way, the trains stuck behind it all came in with their pantographs down.

It seems a curious coincidence that this would happen on the first day - maybe some weird software bug raised the pantograph when it shouldn't have?

Certainly looks like that's struck something pretty solid! Either not dropped when it should have or raised earlier than it should. Not dropping could also be due to a trackside beacon fault as well as an on board fault?

The first trip I ever did on a 756 on the Taff Vale, the driver did an emergency stop on the approach to Pontypridd from the north. From what I could gather, he was seeing an indication that the panto had not dropped on approach to the non-wired section through Pontypridd station. It transpired that it was a false indication, panto had dropped and we were able to proceed. I presume the driver can also see if it does not raise when it should. Is it part of the driver's role to keep track of what the pantograph is doing, coupled with route knowledge of which sections are wired? And does he get any indication of a panto raising in an unwired section? All dependent of course on his desk indications correctly reflecting reality!

There seem to be a lot more potential failure modes in discontinuous electrification than a standard fully wired system.

 

[AdamWW]

These is a broken down tram train at Radyr this evening, as of 20:20 at least. On platform 2, lights off, no one home. I assume this is why our train home was delayed from Cardiff central, with a post departure announcement that we'd skip Cathays and Llandaff.

That will be the one with the modern art on the roof that in previous life was a pantograph.

 

[BillStampy]

398031 also now withdrawn from service early, leaving Cardiff Bay with 0 train service this evening

 

[Torfaen]

Certainly looks like that's struck something pretty solid! Either not dropped when it should have or raised earlier than it should. Not dropping could also be due to a trackside beacon fault as well as an on board fault?

If the pantograph didn't drop before the end of a section of OLE, wouldn't it go up to its limit of travel when there were no longer any wires above it, and surely that would that trigger it to drop automatically and give some indication in the cab?

On the other hand, failing to read a beacon that tells the train to drop the pantograph seems more likely than the train raising it somewhere there isn't a beacon, and I'd hope the onboard electronics meets functional safety standards (such as the ones adopted in the automotive industry) that would prevent this.

 

[fgw_43025]

398031 also now withdrawn from service early, leaving Cardiff Bay with 0 train service this evening

No doubt someone will say just teething problems and only to be expected when introducing a brand new fleet

 

[Torfaen]

Teething troubles? The early finish yesterday evening for 031 was apparently due to shortage of crew - despite the other two diagrams having been terminated hours previously. Pulling the Pontypridd to Cardiff Bay service seems almost a default response when something goes a bit awry on the CVL.

022 / 026 are out this morning, so whatever caused the issue with 017 yesterday thankfully doesn't seem to warrant grounding the fleet pending an investigation.

 

[Peter Sarf]

Every now and again parts of the 3rd rail sections of Thameslimk routes come into contact with a raised pantograph. Especially between City Thameslink and Blackfriars station.

I think no software to stop what I assume is driver error.

Do 398s have software etc to prevent such errors ?.
If not then maybe a newly qualified driver or just distraction. Discontinuous electrification must then multiply the chances of human error.

Anyway I gave up trying to follow proceedings yesterday on my creaking mobile so thank to various posters for info.

I plan to update the list when I get to my computer.

I had a sample of 398026 on Friday Cardiff Bay to Taffs Well and back :-

Kind of as I expected.

Compared to Stadler trams in Croydon they are nicer. More open feel. Maybe the (bland to some) colour scheme.

Seats more train like than tram. Not immediately uncomfortable. (Did not feel like the ironing boards in 700s or 377/6s and 377/7s).

The windows are nice - lovely views of the Taff and its valley.

Fairly good acceleration. Fooled by the lack of diesel thrash (150).

Quiet and ride OK - a little less refined than a FLIRT.

Conclusion - people are going to miss the FLIRTs (which were/are temporary anyway) for a while but I will take the planned frequency increase over that.

 

[positron]

Drivers shouldn't be raising them manually anywhere outside a depot to my knowledge. It's all controlled by trackside equipment.

My guess would be a beacon failure but then there is overrun protection where the pantograph should drop anyway so if it has hit a bridge it's a design failure somewhere.

 

[AdamWW]

Conclusion - people are going to miss the FLIRTs (which were/are temporary anyway) for a while but I will take the planned frequency increase over that.

You might be right. But I think it's easy to overestimate how discerning "normal" passengers are.

And they might also enjoy not waiting for guard controlled door operation or for slow gap fillers, being able to hear manual announcements when near the guard as there isn't the echo effect from the lag in PA announcements, and not having steps up and down to many of the seats.

== Doublepost prevention - post automatically merged: Sunday at 12:11 ==

My guess would be a beacon failure but then there is overrun protection where the pantograph should drop anyway so if it has hit a bridge it's a design failure somewhere.

The train came to a stop after Taffs Well. If it had failed to drop the pantograph approaching Taffs Well, it would have already passed under several bridges before reaching the station.

Perhaps those bridges are high enough. Or the problem was somehow not appreciated (is that possible?) until the train tried to put the pantograph up just before Radyr.

Anyway today trains are using their pantographs on that stretch again - including a 398 I saw.

 

[Tom125]

100% I absolutely love the drivers opening the door: I noticed on my first ride yesterday how quick the doors open. I can’t wait for the valleys to be turn up and go like a proper metro. I don’t really understand why the guard has to close the doors? Wouldn’t the guard be more efficient if they carried on checking/ selling tickets and helping passengers rather than checking the platform and closing the doors?

 

[MikePJ]

100% I absolutely love the drivers opening the door: I noticed on my first ride yesterday how quick the doors open. I can’t wait for the valleys to be turn up and go like a proper metro. I don’t really understand why the guard has to close the doors? Wouldn’t the guard be more efficient if they carried on checking/ selling tickets and helping passengers rather than checking the platform and closing the doors?

Yes, that would be hugely more efficient. However, the guards’ union (RMT) wants guards to maintain a “safety-critical role” and identifying that the doors may be safely closed is the one that best stands up to scrutiny. The reason the union wants guards to maintain a safety-critical role is because a) that way a train can’t operate without a guard, which keeps their members in jobs and gives the union power to call an effective strike b) traditionally, ticket inspectors and other customer-facing staff were considered more junior and paid less than guards, so they fear for their members’ pay if the role is reduced.

 

[Kernow Dave]

The pantograph has been mangled, don't know how

looks like a piece of a tree branch to me?

 

[Peter Sarf]

looks like a piece of a tree branch to me?

Good spot, I cannot be 100% sure as I am looking on a small screen but there is defitely part if a tree there or near.

So not a software problem - as we programmers liked to point out !.

 

[womump]

Here’s some Sunday evening fun. The attached photo was taken at 1635 today (18/10) from the Amsterdam flight back into Rhoose (CWL), looking south down into Taff’s Well, showing the TfW depot to the left of the image. The dedicated observer could probably deduce something from the number of (tram)train-like objects visible, cross referenced with those out on diagrams. Happy spotting.

 

[AdamWW]

Good spot, I cannot be 100% sure as I am looking on a small screen but there is defitely part if a tree there or near.

So not a software problem - as we programmers liked to point out !.

No. I can see there is something that looks like a tree branch in that photograph but it's actually a just a heavily mangled part of the pantograph.

== Doublepost prevention - post automatically merged: Sunday at 20:25 ==

Here’s some Sunday evening fun. The attached photo was taken at 1635 today (18/10) from the Amsterdam flight back into Rhoose (CWL), looking south down into Taff’s Well, showing the TfW depot to the left of the image. The dedicated observer could probably deduce something from the number of (tram)train-like objects visible, cross referenced with those out on diagrams. Happy spotting.

Nice photo!

That certainly is a lot of trams. Or trains.

 

[MikePJ]

398018 is on the Aberdare-Merthyr circuit today. https://www.realtimetrains.co.uk/service/gb-nr:G38222/2026-10-19/detailed

 

[s17663]

On the other hand, failing to read a beacon that tells the train to drop the pantograph seems more likely than the train raising it somewhere there isn't a beacon, and I'd hope the onboard electronics meets functional safety standards (such as the ones adopted in the automotive industry) that would prevent this.

In an interview with a TFW spokesman a few months ago, it was mentioned that the delays to the 398s introduction was partly due to issues with the trackside beacons.

 

[StripeyNick]

looks like a piece of a tree branch to me?

Definitely weren't any branches or such like.
I've been studying my photos of it and as AdamWW says, it's just the way the pantograph has mangled

 

[Torfaen]

In an interview with a TFW spokesman a few months ago, it was mentioned that the delays to the 398s introduction was partly due to issues with the trackside beacons.

The 756s also use the same trackside beacons - so haven't any issues with the beacons been sorted for some time now? Or is the on-train equipment that reads them different on the 398s?

 

[Peter Sarf]

Here’s some Sunday evening fun. The attached photo was taken at 1635 today (18/10) from the Amsterdam flight back into Rhoose (CWL), looking south down into Taff’s Well, showing the TfW depot to the left of the image. The dedicated observer could probably deduce something from the number of (tram)train-like objects visible, cross referenced with those out on diagrams. Happy spotting.

On Friday 16/10/2026 I could see 23 x 398s on Taffs Well. I identified 20 of them including 398008 coming on about 14:20 and 398013 leaving about14:00.

Inside the main shed could be a further six by my reckoning. That makes 29.

In service were three (ignoring 008 & 013. So up to 31 at/passing Taffs Well.

That is out of 33 (398032 at Asfordby, 398035 & 398036 yet to be delivered). So where are two more I wonder ?.

I have updated my list btw :-


== Doublepost prevention - post automatically merged: Monday at 09:55 ==

I feel a list coming on :-

398001
398002
398003
398004
398005
398006
398007
398008 Testing from 15/10/2026
398009
398010
398011 Testing from 15/08/2026
398012
398013 Testing from 19/08/2026
398014
398015
398016 Testing
398017 In service on 28/07/2026 6th, two in service 03/08/2026 (398017 and 398029 not coupled)
398018 In service on 10/08/2026 8th, Testing (multi ?) before
398019 In service on 14/10/2026 12th, Seen at Merthyr on 22/08/2026, presumably testing ?.
398020 In service on 15/10/2026 13th, Testing (multi)
398021 In Service on 18/08/2026 9th
398022 In service on 03/07/2026 4th (all day 23/07/2026)
398023 In service on 08/09/2026 11th
398024 In service on 25/06/2026 3rd
398025.In service on 02/09/2026 10th
398026 In service on 16/10/2026 14th, Testing from 18/08/2026
398027
398028 In service on 23/06/2026 1st
398029 In service on 31/07/2026 7th, also did first weekend service (Saturday 01/08/2026), two in service 03/08/2026 (398017 and 398029 not coupled). First use in service use of a 398 on NR metals 15/09/2026.
398030 In service on 14/07/2026 5th
398031 In service on 24/06/2026 2nd
398032.Old Dalby
398033
398034.Delivered on 14/05/2026 ?
398035.Not yet delivered
398036.Not yet delivered

First day with two in service 03/08/2026 017 & 029.
First day with three in service 19/08/2028 017, 029 & 031.
First rescue of a 398 (022) by two other 398s (possibly T&T as reversal required) 01-02/09/2026 - successful.
. . . 398031 leading to Taffs Well with 398028 on the rear.
First Sunday service using a 398 06/09/2026 398031.
First use in service use of a 398 on NR metals 15/09/2026 (398029)
Service use paused 19/09/2026 - 13/10/2026 presumed for a software update, testing resumed 09/10/2026.

That is all the past dates (before 01/08/2026) I can unearth.

Below is a list of 398s in the order they entered service (I have decided to make life more complicated !).
Order, Unit, date, (days elapsed)

01 398028 23/06/2026

02 398031 24/06/2026 (1)

03 398024 25/06/2026 (1)

04 398022 03/07/2026 (8)

05 398030 14/07/2026 (11)

. . . Milestone : First all day service on 23/07/2026 (398022)

06 398017 28/07/2026 (14)

07 398029 31/07/2026 (3), also did first weekend service (Saturday 01/08/2026), two in service 03/08/2026 (398017 and 398029 not coupled)

. . . Milestone : First day with two in service 03/08/2026 (398017 and 398029 not coupled)

08 398018 10/08/2026 (10)

09 398021 18/08/2026 (8)

. . . Milestone : First day with three in service 19/08/2028 (398017, 398029 & 398031).

10 398025 02/09/2026 (15)

. . . Milestone : First use on a Sunday 06/09/2026 (398031)

11 398023 08/09/2026 (6)

. . . Milestone : First use in service use of a 398 on NR metals 15/09/2026 (398029)

12 398019 14/10/2026 (36)

13 398020 15/10/2026 (1)

14 398026 16/10/2026 (1)

. . . Milestone : First use of 398s North of Pontypridd AND four in service.

 

[s17663]

The 756s also use the same trackside beacons - so haven't any issues with the beacons been sorted for some time now? Or is the on-train equipment that reads them different on the 398s?

I’m guessing it was a communication issue between the beacons and the 398s themselves. The issues had been solved as of June, but it could still be slightly temperamental given there have been issues with them in the past.

 

[800001]

Had it been confirmed anywhere that it was actually a fault with the 398?

Or could it actually of been an OHL fault/damage which then caused pantograph damage?

 

[AdamWW]

Or could it actually of been an OHL fault/damage which then caused pantograph damage?

I am certainly no expert but to me it looked like as if the pantograph hit something solid.

I can easily imagine how an interaction with a bridge could do that.

I can't visualise an overhead line failure having that effect unless somehow a gantry had moved itself.
And if that was the case I doubt that the trains stuck behind the failed one would have continued their way (albeit with pantographs down).

 

[s17663]

On a slightly different note, does anyone know if the 398s have run in double yet, or if they will in future? If so on which routes?

 

[John R]

On a slightly different note, does anyone know if the 398s have run in double yet, or if they will in future? If so on which routes?

They are planned to run double on all three routes, with alternate trains being double once the 4tph service to the valley heads is introduced.

 

[TheSWTS]

Does anyone know what happened yesterday? From the looks of it, 398029 failed, then 398020 which came out to replace it also failed?

 

[Torfaen]

Did 020 fail? Or did it simply come out of the depot too late for the service it was due to operate, and it was so near the end of service that there was no point in waiting at Radyr for it to pick up the remainder of that diagram?

== Doublepost prevention - post automatically merged: Yesterday at 09:14 ==

They are planned to run double on all three routes, with alternate trains being double once the 4tph service to the valley heads is introduced.

With the initial fleet size of 36, I can't see how they will have enough units for that. Currently I think the Merthyr/Aberdare circuit is 9 diagrams, and Treherbert to Cardiff 5 diagrams, so 14 in total for a half-hourly service. Double that is 28, and make half of them double sets, and you'd need 42. With modest accelerations they may save a couple of diagrams, but 36 is still way short.
Until the Cardiff West upgrade happens, I guess Merthyr/Aberdare won't be going to 4tph, so hopefully the priority will be to run 2ph with double sets?

 

[AdamWW]

Until the Cardiff West upgrade happens, I guess Merthyr/Aberdare won't be going to 4tph, so hopefully the priority will be to run 2ph with double sets?

Are you sure about that? I didn't think that was the case.

So far as I know the extra two trains per hour (on all three routes) are supposed to go to the Bay, so no additional capacity is required on the City Line.

It does presumably need the Bay line work to be completed though.

== Doublepost prevention - post automatically merged: Yesterday at 09:23 ==

Definitely weren't any branches or such like.
I've been studying my photos of it and as AdamWW says, it's just the way the pantograph has mangled

The bit that looks like a tree branch is in fact one of the two components (I don't know the technical term) that make contact with the overhead wire, as shown in the image. (This is a 756 but the pantographs are similar if not the same).

I believe they normally have replaceable carbon(?) contact strips which I think are just visible in this image. I couldn't see any sign of them on the mangled version so presumably they were another casualty of the incident - leaving the bare metal underneath which from the right angle in the right light does look a bit like a tree branch.

It is quite impressive how they got wrapped around the other parts of the pantograph. Whatever it hit, it must have been at some speed.

 

[positron]

Did 020 fail? Or did it simply come out of the depot too late for the service it was due to operate, and it was so near the end of service that there was no point in waiting at Radyr for it to pick up the remainder of that diagram?

== Doublepost prevention - post automatically merged: Yesterday at 09:14 ==


With the initial fleet size of 36, I can't see how they will have enough units for that. Currently I think the Merthyr/Aberdare circuit is 9 diagrams, and Treherbert to Cardiff 5 diagrams, so 14 in total for a half-hourly service. Double that is 28, and make half of them double sets, and you'd need 42. With modest accelerations they may save a couple of diagrams, but 36 is still way short.
Until the Cardiff West upgrade happens, I guess Merthyr/Aberdare won't be going to 4tph, so hopefully the priority will be to run 2ph with double sets?

That's not true Cardiff West is not the blocker for 4tph on Merthyr/Aberdare services.

It's the blocker for 4tph on the City line, how they plan to run those extra 2tph is still anyone's guess I can work out 3tph easily. I'm guessing that Llandaff might actually lose serivices to get the 4tph on city.

Anyway like you say I don't think every other service will be doubled up. I'm guessing it'll more be doubles in peak and then distributed through the day a bit with some hours just serviced by single units.

 

[AdamWW]

Yes, that would be hugely more efficient. However, the guards’ union (RMT) wants guards to maintain a “safety-critical role” and identifying that the doors may be safely closed is the one that best stands up to scrutiny. The reason the union wants guards to maintain a safety-critical role is because a) that way a train can’t operate without a guard, which keeps their members in jobs and gives the union power to call an effective strike b) traditionally, ticket inspectors and other customer-facing staff were considered more junior and paid less than guards, so they fear for their members’ pay if the role is reduced.

Actually I disagree.

If there is a guard on board, I think there is more benefit to having them on the platform than otherwise.
They are visible to passengers who might need to ask where the train is going or if they can board without a ticket because the machine is broken etc., and they are in a better position to wait for latecomers or yell at them to hurry up etc.

And in any case under driver controlled operation the guard doesn't check the platform after the doors are shut that's done by the driver using the cameras. The guard just gives the signal to close doors.