Could the Heart of Wales timetable be improved?

[Furryanimal]

Last season , with a Saturday train arriving into Llandovery at 1310, I travelled to watch my rugby team, Newport, play at Church Bank.
Okay,I had to wait until 1805 to leave but Llandovery have a nice clubhouse.
If I want to go by train next week( 4th January) I have to leave at the crack of Dawn,be in Swansea by nine and arrive in Llandovery more than four hours before kickoff.
And coming back could mean having to leave the game early!
Don’t like this new timetable at all and I might add that last season the train i caught to Llandovery was quite busy with rugby fans( including quite a few who got on at Llanelli for a lower division game further up the line )and tourists from America who spotted I was wearing a Green Bay Packers cap.
Not happy..so I either go to Carmarthen and get a bus ( long trip timewise as well)
or don’t go.
And I could go from Swansea to Llandovery for free having a bus pass.
Annoyingly ,if I could drive it would only take me 80 minutes to get from Cwmbran to Llandovery!

 

[louis97]

According to Baxenden Bank's helpful table above, the planned dwell is from 1517 to 1533, which is fairly typical at a crossing point. The southbound service is planned to arrive at 1528, which means that the crew from that only have nominally 5 minutes to get off that, cross over the bridge, take over and set up the northbound unit and depart. That already seems quite tight to me.

That is the passenger timetable, the working timetable has the required 8 minutes as per the Planning Rules. The southbound service is booked to arrive at 1525 in the working timetable.

 

[craigybagel]

Do any HoWL trains end up with standing passengers? Or is usage so low that never happens?

Back when I used to work the line (until about 5 years ago) full and standing was the norm on Saturdays in December with Christmas shopping traffic. About 90% of the passengers then were travelling for free on Welsh bus passes though.....

Also under the old 4 a day timetable the first northbound would arrive at Shrewsbury f&S on a 150. The changes required to fit in the 5th service and the extra short workings messed up the timings for the school/college traffic that filled that service (both to Craven Arms for Hereford and to Shrewsbury) put an end to most of that traffic travelling by rail

Obviously they are removing the late night SHR-LLO and return due to low usage. Does anyone know if it did get used in some capacity or was it literally empty each time it ran? I’m assuming the return back to Shrewsbury would likely be empty

Anecdotally from talking to the crews that worked it - it might have a handful on to Craven Arms, but it was pretty much empty both ways along the branch.

According to Baxenden Bank's helpful table above, the planned dwell is from 1517 to 1533, which is fairly typical at a crossing point. The southbound service is planned to arrive at 1528, which means that the crew from that only have nominally 5 minutes to get off that, cross over the bridge, take over and set up the northbound unit and depart. That already seems quite tight to me.

Don't forget the rather slow process to exchange tokens as well. NSTR signalling is cheap, but it's not the quickest or most user friendly.

 

[Jan Mayen]

What would be the most frequent service possible (assuming all five passing loops are in use)?

 

[Lewisham2221]

What would be the most frequent service possible (assuming all five passing loops are in use)?

With a quick calculation, I don't think hourly would be quite possible. So probably 1tp2h, or 3tp4h if you wanted to maximise frequency but use an irregular timetable.

 

[Dai Corner]

It seems to be a very expensive way to shift occasional heavy loads of passengers, some of whom aren't even paying and could travel for free by bus and others who would find a private hire coach or car sharing cheaper, quicker and more convenient.

Was the line built mainly to give the LNWR access to Swansea without having to share revenue with the GWR? That purpose disappeared in 1948.

Would anyone like to propose a timetable which would increase passenger-miles and revenue? Perhaps more short workings and fewer through trains?

 

[Western 52]

The southern section from Llandeilo and Ammanford to Llanelli and Swansea could probably support a better service if it was reliable and priced sensibly. Maybe hourly or so, but I guess TFW would have a different view! Pathing could be an issue I suppose with the reversal at Llanelli.

 

[ChiefPlanner]

It seems to be a very expensive way to shift occasional heavy loads of passengers, some of whom aren't even paying and could travel for free by bus and others who would find a private hire coach or car sharing cheaper, quicker and more convenient.

Was the line built mainly to give the LNWR access to Swansea without having to share revenue with the GWR? That purpose disappeared in 1948.

Would anyone like to propose a timetable which would increase passenger-miles and revenue? Perhaps more short workings and fewer through trains?

It would be a challenge to improve the cost base - the HoW operating costs versus revenue was something like 5 to 1 ratio (and that after very considerable changes to signalling etc over the years) - hard to see how you could significantly improve that - I fully appreciate the value of the line and doing more (running more service over the more populated sections appears to make sense) - but the overall economics are not encouraging. It is the conundrum of many regional services I am afraid.

Going back in history - the line was shared LMS / GWR down the Vale of Towy as a joint line - but passenger traffic was a less important factor compared to through freight the length of the line , and especially for the significant coal flows of best Welsh anthracite from the Ammanford area - to not only Swansea Docks , but nationally to the coal depots in the South East .....all gone.

 

[Baxenden Bank]

What would be the most frequent service possible (assuming all five passing loops are in use)?

The longest gap between loops is Knighton to Llandrindod at 35 / 38 minutes or so. Assuming trains run end to end, that would be your minimum frequency - every 1h 10m ish, depending on exact working times / rules of the route rather than published times. An even two-hourly timetable is nearly possible, you would need to shave 'a bit' off the Craven Arms to Llandrindod times which are currently 60 minutes 'departure to arrival'. Whether that can then be fitted into the Marches line is another issue.

Obviously if money was available you would move loops to optimal locations for your desired frequency, including dynamic loops (short sections of double track). That is what was done on the Cambrian line to allow the hourly frequency.

 

[Harpo]

At danger of drifting into speculation/fantasy, the route to regular frequencies would probably be through not having everything calling everywhere.

A fast/semi fast in one direction (loop-to-loop) passing a stopper in the other, then reversing the direction of the calling pattern with the next pair, might enable regular service patterns. Faster services to/from the busier locations might even bring growth?

Central Wales Metro?

Ain’t gonna happen though!

 

[Dai Corner]

Going back in history - the line was shared LMS / GWR down the Vale of Towy as a joint line - but passenger traffic was a less important factor compared to through freight the length of the line , and especially for the significant coal flows of best Welsh anthracite from the Ammanford area - to not only Swansea Docks , but nationally to the coal depots in the South East .....all gone

Presumably the passenger service was improved as line capacity was released from coal but the demand just wasn't there and it exists more for political reasons than anything else.

The Welsh Government may have to make a hard decision, particularly if rail infrastructure is devolved in the future.

 

[anthony263]

An hourly Swansea to Ammanford service with 1 train everyb2 hours continuing north to Shrewsbury and possibly Crewe would be ideal

 

[ChiefPlanner]

exists more for political reasons than anything else.

I think it was the late George Thomas MP , later SoS Wales who pointed out that the line ran through 5 marginal constituencies that helped save it - there were closure attempts in 1967 and again in 1972 I believe.

 

[Wilts Wanderer]

Whilst the dwell times at the passing loops will significantly add to the end to end duration, given the sheer number of stations does the contemporary door open and close procedure for the guard (check local door is at platform, unlock local door, open local sliding door, check all doors are on platform, unlock doors - start hustle alarm, watch remote doors close, start hustle alarm on local door, watch it close, ding, ding*) add significant time compared to the old heritage DMUs, where people would be getting off before the thing even stopped?
(Unless of course passengers didn’t close the doors behind them)

*I’m no guard and have probably forgotten something.

Traditionally the HoWL has been planned with ‘dot stops’ at most/all of the request stops, I.e. an allowance for deceleration and acceleration but nothing for the dwell itself, on the assumption that relatively few actual stops would be made on any given service. In practice the sectional timings also contained slack and the usual problem was that trains would run progressively earlier, requiring careful working by the crew to moderate their speed to avoid getting ahead of advertised times.

What appears to have happened with the new timetable is that every request stop has been allocated 30 secs dwell time - which given the sheer number of stops, adds quite a bit of overall allowance to the timetable. I would be curious as to just how slowly a train now has to run to maintain correct times at the stops once a certain number have been passed. The assumption now seems to be that every stop will be made, which of course is completely unnecessary, given the light usage of the line.

 

[craigybagel]

The longest gap between loops is Knighton to Llandrindod at 35 / 38 minutes or so. Assuming trains run end to end, that would be your minimum frequency - every 1h 10m ish, depending on exact working times / rules of the route rather than published times. An even two-hourly timetable is nearly possible, you would need to shave 'a bit' off the Craven Arms to Llandrindod times which are currently 60 minutes 'departure to arrival'. Whether that can then be fitted into the Marches line is another issue.

Obviously if money was available you would move loops to optimal locations for your desired frequency, including dynamic loops (short sections of double track). That is what was done on the Cambrian line to allow the hourly frequency.

It's frustrating how close it is to being able to run a 2 hourly service using the existing infrastructure, but it's hard to see how those 5 minutes can be shaved off Craven Arms to Llandrindrod. It's already 60MPH east of Knighton, the highest speed on the line, and even if it was upped to 75MPH a 153 isn't going to spend a lot of time at that speed with the gaps between stations and the need to stop at Bucknell level crossing. West of Knighton the long climb up to Llangynllo removes the point for any kind of speed improvement.

You could also get 2 hourly with just 1 extra loop, at Garth. This would allow trains to cross there, and at the existing loops at Knighton and Llandeilo. Hard to see anyone putting up the money for all of that though....

 

[Harpo]

Even further back, the GBTT request stop times were calculated to be the equivalent of the passing times for a non-stop train.

Zero dwell stops were timed in the WTT and the appropriate ‘p’, ‘q’, ‘r’ (etc.) codes added to put the earlier times in GBTT.

This way, trains simply got to the loops early and waited time.

 

[Wilts Wanderer]

Even further back, the GBTT request stop times were calculated to be the equivalent of the passing times for a non-stop train.

Zero dwell stops were timed in the WTT and the appropriate ‘p’, ‘q’, ‘r’ (etc.) codes added to put the earlier times in GBTT.

This way, trains simply got to the loops early and waited time.

That should occur anyway, it just becomes more problematic with 30 second dwells as after a string of stops the public/working differential has built up more quickly. Looking at RTT, all TFW have done is apply a minus-1 to each departure which is inadequate.

 

[Envoy]

How many people actually travel from the Heart of Wales to Llanelli & Swansea? Do those getting off/on at Swansea do so in order to travel to/from the east? Would people at the southern end (Llandovery southward) of the line prefer to have direct trains to/from Cardiff?

I think a survey should be done of residents at the southern end to see if they would be likely to use a direct service to Cardiff or do they favour things being left as they are with Llanelli and Swansea having the direct trains to HoW? If trains went direct from Cardiff to Pontarddulais it would save a heck of a lot of time and be more likely to attract tourist traffic doing a day tour of southern Wales & the Marches. I think that TfW should experiment this coming summer with a train leaving Cardiff at around 9.10am for Pontarddulais & the HoW. The unit could then return back down the HoW from Shrewsbury and the tourist traffic returning on any train back down the Marches.

If this were Switzerland the line would be electrified and the coaches would have large glass windows. Vegetation would have been cleared to open up the views. So, we are stuck with 153’s.

 

[70014IronDuke]

... What appears to have happened with the new timetable is that every request stop has been allocated 30 secs dwell time - which given the sheer number of stops, adds quite a bit of overall allowance to the timetable. I would be curious as to just how slowly a train now has to run to maintain correct times at the stops once a certain number have been passed. The assumption now seems to be that every stop will be made, which of course is completely unnecessary, given the light usage of the line.

Maybe this is why I have seen far more early arrivals of minus 2-3 minutes at the passing loops along the line on RTT since the new timetable has been implemented.

How many people actually travel from the Heart of Wales to Llanelli & Swansea? Do those getting off/on at Swansea do so in order to travel to/from the east? Would people at the southern end (Llandovery southward) of the line prefer to have direct trains to/from Cardiff?

Llandeilo latest annual pax flow

Swansea - 1105
Llanelli - 368
Cardiff - 212

Source: https://railalefan.co.uk/labs/flowstats/LLL/

Of course, with a direct service to Cardiff, the ratios might be reversed.

What is surprsing (at least to me) are the intra-HoWL flows.
Shrewsbury is the second largest flow for Llandeilo - at 699
Llandovery the third at 377

Of course, all these numbers are relative and (sadly) pitiful, but the southern end of the line surely has the most potential for genuine growth. However the long single line section from Llandeilo to Morlais Jcn was all too often causing crippling delays in the five-trains-a-day timetable once a train lost its path. Quite how much a new loop would cost at Pantyffynnon I don't know. I think there has to be space there, but in terms of justifying funding in the current or any climate, ... hmmmm.

 

[Dai Corner]

For comparison, First Cymru buses operate half hourly from Swansea to Pontardullais, hourly to Ammanford and at least every two hours to Llandeilo, though journey times are longer.

 

[ChiefPlanner]

Maybe this is why I have seen far more early arrivals of minus 2-3 minutes at the passing loops along the line on RTT since the new timetable has been implemented.

Llandeilo latest annual pax flow

Swansea - 1105
Llanelli - 368
Cardiff - 212

Source: https://railalefan.co.uk/labs/flowstats/LLL/

Of course, with a direct service to Cardiff, the ratios might be reversed.

What is surprsing (at least to me) are the intra-HoWL flows.
Shrewsbury is the second largest flow for Llandeilo - at 699
Llandovery the third at 377

Of course, all these numbers are relative and (sadly) pitiful, but the southern end of the line surely has the most potential for genuine growth. However the long single line section from Llandeilo to Morlais Jcn was all too often causing crippling delays in the five-trains-a-day timetable once a train lost its path. Quite how much a new loop would cost at Pantyffynnon I don't know. I think there has to be space there, but in terms of justifying funding in the current or any climate, ... hmmmm.

Re Pantyffynon extra loop - or reinstatement - in theory the 1892 box may have the old interlocking in place , or certainly the capability to reinstate it.....

 

[Envoy]

Maybe this is why I have seen far more early arrivals of minus 2-3 minutes at the passing loops along the line on RTT since the new timetable has been implemented.

Llandeilo latest annual pax flow

Swansea - 1105
Llanelli - 368
Cardiff - 212

Source: https://railalefan.co.uk/labs/flowstats/LLL/

Of course, with a direct service to Cardiff, the ratios might be reversed.

What is surprsing (at least to me) are the intra-HoWL flows.
Shrewsbury is the second largest flow for Llandeilo - at 699
Llandovery the third at 377

Of course, all these numbers are relative and (sadly) pitiful, but the southern end of the line surely has the most potential for genuine growth. However the long single line section from Llandeilo to Morlais Jcn was all too often causing crippling delays in the five-trains-a-day timetable once a train lost its path. Quite how much a new loop would cost at Pantyffynnon I don't know. I think there has to be space there, but in terms of justifying funding in the current or any climate, ... hmmmm.

I really question how they can arrive at these flows being as so many people buy split tickets? Then you have the fact that over 60’s can travel free in winter as far north as Shrewsbury rather than pay (a lot) to go via Cardiff. Even for those paying, it is considerably cheaper to go from say Llanelli to SHR via the HoW than via Cardiff and it takes only about 30 minutes longer as although the HoW is slow, it is a shorter route in miles. So, not all those people are likely to be going to SHR but merely using it as a staging point for onward travel on a new ticket.

Likewise, for eastward travel, it is not just the numbers going to Cardiff but everyone who wants to travel between HoW & Port Talbot and all places east thereof. So, to the 212 listed for Cardiff we must add the 225 for PAD for starters. I contend that the number of people desiring to travel between places east of Port Talbot and the HoW to be far greater than those travelling to Llanelli and places west thereof plus Swansea. Buses could easily cater for these people - including Llandeilo to Carmarthen - in the event that HoW trains ran directly between Cardiff and Pontarddulais and the H0W. Using this shorter direct route would boost tourist traffic as well as be helpful to the locals at the southern end of the line. People at the Shrewsbury end would also be more inclined to use the H0W route as a day trip down to Cardiff where they would stay for 2 to 3 hours before returning via the Marches. Using the slow Llanelli route just does not stack up.

 

[Topological]

The idea of a Cardiff service using the Swansea District Line does seem to have merit. Perhaps even a morning train that reaches Cardiff in the peak then returns up the HoW before coming back to Cardiff at a suitable end of day time. There would also be a peak working out of Cardiff going part way up the HoW.

Does the path for the Welsh Show work?

 

[70014IronDuke]

The idea of a Cardiff service using the Swansea District Line does seem to have merit. Perhaps even a morning train that reaches Cardiff in the peak then returns up the HoW before coming back to Cardiff at a suitable end of day time. There would also be a peak working out of Cardiff going part way up the HoW.

Does the path for the Welsh Show work?

I'd say the answer to that is maybe and not really.

With the new TT, it is unlikely that the exact Welsh Show path(s) used previously would work, but I'm sure some sort of paths will be found, assuming TfW decided they want the specials running next summer, though these may need some longish waits at the various crossing points.

However, the real problem is that the times for these paths are not what would be required for a basic Llandovery or Llandeilo to Cardiff peak-hour commuter service, which is the most logical place to start. In fact, with the current track layout and TT, I can't see how it could be done, as the current 2V22 06.44 Llandovery - Swansea effectively takes the path required for any Cardiff train.

The only alternatives as as see it would be

a) Divert 2V22 to Cardiff. That is not going to go down well with the current clientel or
b) Run a Carmarthen - Swansea service via the District Line and stop it at Llangenach to pick up passengers from 2V22 at around 07.40. This, of course, would be nothing like as attractive for Cardiff passengers, and would introduce a significant risk of missing the connection because of late running by 2V22. (And I'd say this is not an insignificant risk.)

And in the reverse direction, a train leaving Cardiff c 17.30 would likely find its way forward off the SDL blocked by the current 17.48 ex-Swansea having to clear the long section to Llandeilo. It could be worked around, but the track layout at the southern end of th eline does not make it easy.

 

[Topological]

I'd say the answer to that is maybe and not really.

With the new TT, it is unlikely that the exact Welsh Show path(s) used previously would work, but I'm sure some sort of paths will be found, assuming TfW decided they want the specials running next summer, though these may need some longish waits at the various crossing points.

However, the real problem is that the times for these paths are not what would be required for a basic Llandovery or Llandeilo to Cardiff peak-hour commuter service, which is the most logical place to start. In fact, with the current track layout and TT, I can't see how it could be done, as the current 2V22 06.44 Llandovery - Swansea effectively takes the path required for any Cardiff train.

The only alternatives as as see it would be

a) Divert 2V22 to Cardiff. That is not going to go down well with the current clientel or
b) Run a Carmarthen - Swansea service via the District Line and stop it at Llangenach to pick up passengers from 2V22 at around 07.40. This, of course, would be nothing like as attractive for Cardiff passengers, and would introduce a significant risk of missing the connection because of late running by 2V22. (And I'd say this is not an insignificant risk.)

And in the reverse direction, a train leaving Cardiff c 17.30 would likely find its way forward off the SDL blocked by the current 17.48 ex-Swansea having to clear the long section to Llandeilo. It could be worked around, but the track layout at the southern end of th eline does not make it easy.

Could 2V22 split? I am not familiar with the track layout coming off the bottom of the HoW, but that might help with the morning.

Likewise, a join in the evening could be performed, but that does induce risk.

IF the only location for split/join is Llanelli then that would make a big dent in the benefit of the Cardiff service.

 

[Anonymous10]

The southern section from Llandeilo and Ammanford to Llanelli and Swansea could probably support a better service if it was reliable and priced sensibly. Maybe hourly or so, but I guess TFW would have a different view! Pathing could be an issue I suppose with the reversal at Llanelli.

Unless you ran it to Carmarthen with a suitable connection at Llanelli for Swansea or just didn't call at Llanelli?

 

[Meerkat]

Am I totally misremembering in thinking that one of the line's reasons for survival is/was the importance of having a diversionary route for traffic to/from the steelworks?
It does seem almost the definition of a basket case route - infrequent tiny trains with two crew over a very long piece of infrastructure. Got to wonder whether the money would be better spent on the bus service and convert the line to a bike ride through the hills (though with the same issue of not being able to see anything due to the trees), and particularly how expensive an infrastructure failure would need to be to kill the line off.

 

[fishwomp]

It does seem almost the definition of a basket case route - infrequent tiny trains with two crew over a very long piece of infrastructure. Got to wonder whether the money would be better spent on the bus service and convert the line to a bike ride through the hills (though with the same issue of not being able to see anything due to the trees), and particularly how expensive an infrastructure failure would need to be to kill the line off.

Closing it would be costly. The Welsh Govt would need to find a way to blame central Govt, and vice-versa, and just ignore those rural commun constituencies.

Would be better to consider what interventions could make the line more useful to reduce the subsidy.

More passengers:
- low catchment area population
- high tourism potential

Both need:
- Good connections - currently hard to do day returns / round robin from anywhere far from Swansea/the Marches line..
- Faster journeys (platform door theatre, timing of request stops, passing loops convert to ETCS)
- Marketing, promotional tickets, reasonable fares - abolish peak hour tickets for anything via that route?

Less cost:
- single man operation (!), call it a tram train experiment.. but you're still not going to make a dent in the cost.
- reduce rolling stock cost - can't see how that can get lower than a 153.

 

[Envoy]

I note that rail tours seem to find slots for specials using the District Line & the HoW. Here is one on 29 March 2025:>

The Heart Of Wales Explorer - Confirmed Timings Added - Pathfinder Railtours

Take a journey through the Heart of Wales on board our special train: Joining Stations: Eastleigh, Southampton Central, Romsey, Salisbury, Warminster, Westbury, Trowbridge, Bath Spa, Filton Abbey Wood, *Severn Tunnel Junction.

www.pathfindertours.co.uk

With everyone on board we glide along the South Wales Mainline, through Newport, Cardiff and industrial Port Talbot, as our excursion heads for the Swansea District Line. From here we bear right and join the 90 mile Heart of Wales line, meandering diagonally across the rural heartland of Wales. The line follows the Loughor River to Pantyffynnon then into the Towy Valley between the market towns of Llandeilo and Llandovery. A climb up the Afon Bran valley leads to Cynghordy viaduct and on to Sugar Loaf summit, before descent to Llanwrtyd Wells, famous for its annual Bog snorkelling World Championships!

Let’s hope that Network Rail get some people out now in winter chopping down the trees and bushes that block the views. I guess that by 1 March this is not allowed due to the bird nesting season.

 

[Harpo]

I note that rail tours seem to find slots for specials using the District Line & the HoW. Here is one on 29 March 2025….

Subject to any traction constraints such as loco speed restrictions (e.g. 25mph late 80s, lifting to 40mph), it has always been possible to path specials on the Central Wales line, planning a way through unoccupied sections between loops to create a complete journey, but possibly with some longer waits at loops.

I think it’s also possible still, to lock a train away at Llandrindod leaving both loops clear?