The Cambrian point is mentioned somewhere in the explanatory documents - staying with 158s until 2025, so no timetable recast on which to consult.

The Ebbw Vale, Borderlands and Conwy Valley routes are I think omitted as well - probably because there's no tangible change there either; Pembroke Dock isn't properly included except as an origin/destination, with the journeys going "off-table" at Whitland. Ebbw Vale has already gone over to a new timetable that roughly matches the previous draft; this route is alluded to, as most of the Maestegs will run through - and it looks like there will be 3x 3-car & 1x 4-car diagram for EBB-MST. I think the 4 car diagram has shifted from Ebbw Vale to Maesteg peak coverage, but might have misremembered something there.

An interesting point in the blurb is additional workings to Tenby in the summer. That would be tight, as it takes something like 26 minutes to cover the single line WTL-TEN, so along with robustness margins and time there isn't any slack in an hourly service. Missing out the request stops in the additional hours would save a bit of time, as would 197 timing, but this would make a marginal difference. Going a bit off-pattern and not running every single hour could make it feasible, though. Running through to PMD wouldn't work hourly, but trains are much emptier after TEN so that wouldn't be a major issue.

I think I read that Mk4s not running to SWA^» (Swansea - next trains) was partly down to the fuel range of the 67s. However there's also the issue that breaking at SWA rather than CDF^» (Cardiff - next trains) would break the through journeys between CDF and MFH - bearing in mind that with MFH being 2-hourly, it would lose all its CDF through journeys in one direction (broadly speaking). Retaining the direct service to CDF is probably seen as a much bigger political necessity than hourly to MAN. However it would be good if all MAN services could connect well, which doesn't happen now; I _think_ this will improve in the new tt but not checked properly yet.

Some of those PMD-CDF journeys are running very close to Londons west of SWA as well as being very close behind departing SWA; I think the West Wales resignalling is due for completion soon (this spring/summer perhaps but I've lost track) but even so it looks very tight. The CDF-PMDs also have tight turnarounds at CDF - the down workings are in a sense fine if delayed as they have very long turnarounds at PMD, but the knock-on delay passing at TEN would mean the up train would potentially miss the London connection at SWA and then cause a knock-on delay turning around at CDF. In practice maybe any significantly delayed down workings would end up being held back at WTL to avoid the knock-on delay, at the expense of then being very late onto the branch (thus Delay Repay costs) but recovering during turnaround at PMD.

These hour gaps SWA-CDF aren't great politically, even if on paper the hourly IETs^▸ (Intercity Express Train) provide enough seats to match demand. On the other hand, the way trains are flighted in the down direction means there isn't a proper half-hourly down service anyway. In the longer run maybe an additional local service would be useful (providing half-hourly for Pyle, proper metro frequency SWA-GWN-LLE etc) but this would require more funding, not to mention being lower priority in terms of the badly-needed getting the basics right first.

Hopefully any issues with splitting/joining 197s will be reduced by then as there will be a lot of it! The IETs were awful in the early days, with the CMN-PAD^» (Paddington (London) - next trains) often leaving SWA late, but they mostly do it very quickly now...

Where the CDF^» (Cardiff - next trains)-CMN runs are removed, connections could be better - it looks like the CMN - FGH/MFH shuttles have long connection margins for the Londons. I wonder if the shuttles could be retimed, even if it means running off-pattern, except perhaps the that arrives at CMN just before 9am. These workings appear to lay over at CMN for a long time rather than doing anything else - so perhaps the reason is purely cost-cutting or just an idea that an IET^▸ (Intercity Express Train) is more than enough provision in those hours.

I note that the change to the HoW^‡(resolve) pattern also means the ad-hoc SWA^» (Swansea - next trains)-CMN workings between HoW diagrams have been lost, but they weren't vital for capacity provision.

While some of these changes are perhaps because of paths/capacity not being available yet or a sensible avoidance of more transitional issues, there has perhaps been a demand for reduction of costs compared to the earlier draft.


It's not the splitting that bothers me, but that it's the 2 car portion that goes through. The colour scheme makes clear that only the Manchesters split at SWA ("orange" shading for the 5 car section) - the portion that goes through to MFH/FGH returns to Cardiff as 2 car throughout, which isn't ideal. I think if the CDF terminators also split/attached at SWA, things would become more complex at SWA; doing it every 2 hours means there's a detachment one hour, then an attachment hte next hour. Diong it every hour would make things messier and delays would propagate easily.

For the SWA-CDF part, some journeys are quiet enough for 2 cars, but many could do with 3 - especially bearing in mind that the non-Manchesters include peak workings (first MFH-CDF and 17xx CDF-MFH). These peak workings both run behind IETs which helps, but can still be quite busy as 2 car, especially when the GWR^▸ (Great Western Railway) is disrupted, and the mid-morning workings can be crowded on weekends & school holidays... although ironically that's mitigated here by removing them completely, so passengers have to use the IET either side!

In order to keep things simple it looks like the 2 car would always be at the Manchester end (i.e. 2 cars continuing when splitting at SWA, and 3 cars when later evening workings split at CMN)... it wouldn't be practical to vary this, as the split/join plan would become much more complex, so it has to be always 2 or always 3 continuing past SWA.


I think that's possibly part of it - the Cardiff pull is a big selling point... but the presence of hte GWR through services covered that need and was part of hte justification. Another factor is that it reduces the number of diagrams needed for Swanline & Pembroke Dock combined. The original draft needed 3 for PMD (as now) and another 3 for hourly Swanline. This arrangement appears to need 4 for the 2-hourly through service, plus some juggling to provide the peak extras. I also note that 153s are to remain on CRE^» (Crewe - next trains)-SHR^» (Shrewsbury - next trains) stoppers - same shading as HoW - perhaps corresponding to one fewer HoW diagrams being needed?

Overall I wonder if this is TfW trying to under-promise so that it can over-deliver later. It could be argued that this is a valid response to recent over-promising, under-delivering, as anyone who has wandered through the underpass at Queen Street will note!

I can't see 1L11 getting to CDF^» (Cardiff - next trains) at 0835 behind that stopper! Especially if it's operated by 153s initially - those timings could only possibly work with a 197, and even then it's tight. Strange that the stoppers are often leaving SWA^» (Swansea - next trains) 2 mins behind the Londons (and the timings at Llansamlet suggesting this is being maintained throughout that stretch)... have the long 2-aspect sections between Landore and Neath that previously required 5 min headways in the rules been removed? (Not rhetorical - no idea.)

I'm not a fan of the 5 car Manchesters splitting to 2 cars (as per the previous draft) at SWA. Maybe with the recently increased IET^▸ (Intercity Express Train) presence beyond Swansea the capacity need is better managed, but it's not ideal, as those 2 car sets end up on the journeys that start/terminate at CDF. It's less of a problem on the down where they're generally following IETs, but on the up it's not great, or indeed on the down if IETs are disrupted. There is a general issue with near-hour gaps CDF-SWA (to spell it out, on the main line between Wales's two main cities) - on the down because of the way trains are flighted, and on the up where the newly-proposed gaps arise.

Good to see the increase on the Milfords, but it is at expense of improvements to the Fishguards. Perhaps this is recognising differences in demand, and it also means more Milfords to/from Manchester. The idea would be the previously proposed 2 hourly to each, plus a 2 hourly CMN-MFH shuttle, but that would require another unit & crew.

The evening down Swanline service much is better than in the earlier draft, with existing last train restored. Plus there's an up Swanline leaving SWA around 23:30, a need that is mentioned in local media from time to time; I think in the previous  draft that only stopped at main stations.

Interesting to see the Pembroke DOcks and Swanlines merging - perhaps on the basis of feedback from previous draft, noting that PMD was the only Pembrokeshire branch not to have a regular Cardiff service, although it could be argued that the PMDs connnected best with the Londons - and connecting into an IET is perhaps preferable to a through 2 car journey. (Can of worms re IET comfort opened...)

The Swanline and HoW^‡(resolve) improvements are very unfortunate. Swanline and south end of HoW are underutilised, and need a service more suited to the populations served in order to grow - huge lost potential. With recent cuts in bus services there's also the possibility of abstracting former bus demand.  As it stands the service levels keep numbers low which can then be used as a reason not to improve things. Certainly the late evening SWA^» (Swansea - next trains)-LLV hasn't looked full when I've seen it, but with few daytime trains to provide the outbound traffic, and a high cancellation rate, it's not exactly been allowed to grow demand. Presumably this comes back to funding, of course, but not good when we need to encourage public transport usage.

The benefits of this go well beyond a quicker journey. I've just checked RTT^» (Real Time Trains - website) and the 16:18 PAD^» (Paddington (London) - next trains)-SWA^» (Swansea - next trains) being 'superfast' gives it a much better path:

• It leaves CDF^» (Cardiff - next trains) before the stoppers that is current follows, giving it much better timings west of CDF, along with avoiding the performance risk associated with delays to those stoppers. (OTOH^▸ (On The Other Hand) if it is 10 mins late it could still be stuck behind them)
• It will go through to CMN, replacing the 1548 through journey. In fact the SWA-CMN part of the journey is in the same path, as the 1548 had a half-hour wait at SWA, so a much quicker through journey. The 1548 also runs close to a Milford Haven service, so it is in a way pointless except for those who really wanted the through train, as the quickest option for CMN is to transfer to the Milford.
• It's currently booked a 5 car, and I assume that won't be changing as the CMN leg doesn't show a split on RTT (assuming it's correct). It's usually fine west of CDF that side of 6pm but removing the RDG^‡(resolve) & SWI^» (Swindon - next trains) stops would probably help with loading on the busier part of the journey.

I'd think of 'essential' things like public transport and food/essentials/non-specialist shops as different from a leisure destination in terms of the need to accept a range of payments. It's a long time since I last paid for a journey with cash, but there are those who need to travel but would struggle with paying with a card. Children or those with 'basic' accounts spring to mind. With prepayment cards being more widespread, it may become easier. Many young/basic accounts now come with Electron/Solo type cards which must always authorised even for transactions below the floor limit (AIUI^▸ (as I understand it)) to prevent the account going below zero*, but even now I think there would be many situations where cash would be used out of necessity, e.g. parent giving child some cash for a journey or someone using last cash set aside before payday.

* ISTR^▸ (I seem to recall/remember) the railway historically didn't accept these, partly for consistency because it was impractical for on-train purchases. With modern technology perhaps things have moved on now.

The reports of issues near Avonmouth correlate reasonably well with the equinoctial spring tides...

https://ntslf.org/tides/hilo?port=Avonmouth

Perhaps this data needs to be fed into Journeycheck!

I think a lot of the pattern changes come down to the way train-trams are to be deployed. Perhaps joining the dots between the following:

• Cardiff Bay extension will be the first 'tram' destination.
• The train-tram depot is in Taffs Well. (Placed here for land availability or political reasons, and service pattern following from this, or located based on service pattern decision...?)
• RCT valleys the most likely (or at least vaguely possible) next 'tram' mode extensions? (Hirwain, Rhondda Fawr etc?)
• The eventual proposed plan (or at least one of them) would have 6tph to CDF^» (Cardiff - next trains) and 6tph to CBD from the RCT axis - so the CDB trains are all extras compared to existing pattern. (Again cause or effect? Already decided to run train-trams from CDB that way so extra services needed, or running to CDB because of desire for extra trains? Worth noting more trains needed in total because train-tram shorter than Sprinter.)

 

I wonder if the CYS-CDF^» (Cardiff - next trains) journeys include people who book an advance ticket to CDF, and then a day return to CDF - either because the advance fares are only available on the main line flows, or because CYS isn't on their radar, until they look into it nearer the time and realise there's a decent local service that drops them right where they want to go.

I think the difference between Llandaf and Ebbw Vale will be down to the direct trains; the difference is less stark if the 'big 4' Cardiff station figures are aggregated. Someone travelling from Llandaf to Cardiff Bay or somewhere near but not immediately adjacent to Queen Street station will go to the relevant station. Someone from Ebbw Vale might choose to avoid the change (or extra change for CDB) and walk from Central. If I were doing that journey on a season ticket I'd buy the CDQ/CYS ticket (same price) or maybe even CDB (slightly more typically, unless recently changed) to keep options open; maybe if I lived in Ebbw Vale and worked in Cardiff Bay I'd usually walk from Central, but maybe use the train to the bay if it's raining, or walk to Queen St for shopping after work before going home from there. But I assume most would just pick a 'usual' station.

The number of journeys to CDF^» (Cardiff - next trains) & CDQ is intriguing- the sort of suburban journey that usually fares badly in competition with bus and foot. It's different if making that journey to connect with the main line, where there's no extra cost (same fare as CDF from most places last time I checked) but clearly rail is competing well here if people are willing to pay for such a short journey. Perhaps evenings out where rail is seen as a viable, safe alternative to walking or taxi contribute to this. Perhaps there's a lesson for the many places where closely-spaced suburban stations do badly in the stats, although Cathays has relatively unique access to education & work destinations and a local population dominated by a demographic that uses public transport a lot.

Range of changes also interesting... local flows and established student flows like Trefforest/Pontypridd doing well. Bristol has held up quite well - students who live near enough to travel home regularly, inter-university work flows, day trips (I could imagine BRI^» (Bristol Temple Meads - next trains) being a popular destination for students esp the overseas ones who like to see sights) etc.  Llantwit Major and Llandaf have dropped badly - places where the flows will be the CYS-bound work commute, and fairly affluent so car ownership would be high?

A lot of issues over recent years to stifle demand as well - not just Covid, but major reliability issues during one of the leaf fall seasons (2018/9 ish), and a lot of closures for Metro upgrades (including today, in fact). No Rhondda origins in the top 20 but I wonder how the recent, long Treherbert closure would have affected things.

Several valleys (re)openings in that era (thinking about it, not sure OTTOMH if Cathays was a 're') as industry declined, meaning less capacity used by freight and more commuting needed as people had to look further for work. I'd guess they knew student & related employment numbers would be a significant part of the demand, but also AIUI^▸ (as I understand it) student numbers are substantially higher now.

The eventual Metro frequency increase might distribute demand a bit, but I'm a bit apprehensive - the train-trams are about the same length as a Pacer, so even with doubled frequency the total capacity won't be much of an increase on the pre-Covid peak, if we simplistically say train length (metres not cars) is a direct measure of capacity. I wonder if the sparks effect would offset the post-Covid commuting loss, such that the capacity is eaten quickly.

Often my commute destination and still the same - narrow platforms (extended for 6 car trains that never happened), narrow exits with ramps leading to cramped gate huts (one a bit bigger for booking office). Would be fine if it were a wayside halt, but as the peaks flow both in and out, rather than one way morning and the other in the evening, there are clashing flows when trains arrive. People arriving at platforms often stand right by the exit, and people waiting to get on are blocked off by people getting off as there isn't space for both. The up platform's gates used to be part-time effectively, depending on staffing, but now there's remote assistance from the other side. People whose tickets don't work therefore block the flows while working out what they need to do to escape.

The strip of space for the station is fairly narrow (bearing in mind it was a reopening) but there are dead spaces between the platform railings and boundary walls which might allow for a bit of widening.

Is this perhaps an algorithm fail? Some route guesstimate quirk that's incorrectly split the 'Birmingham Stations' data between actual destinations.

Busy 'off-peak' trains towards Barry in summer (with less mid-route changeover at Cardiff than usual) indicate that there are strong day tripper flows, not to mention the complaints on social media about overcrowding. The Barry Island one shows the value of direct services - Pontypridd, Merthyr etc vs places like Caerphilly that might be expected to produce similar demand. I wonder if the switch of routes will affect numbers, or if the established passengers will accept the change and new flows will develop.

Ebbw Vale is interesting - the Newport opened a while ago but had a very patchy service. When that beds in it there's probably a lot of untapped demand for Newport, given how nearby it is, but there's competition with frequent buses.

I don't see huge numbers of people with luggage on the VoG trains but there is probably some traffic to justify the Airport reference. I imagine a lot of long-distance passengers for the airport use the bus from Cardiff; it was originally suggested that there would be a fast VoG service but it never materialised, and would be difficult to path. Perhaps also some inbound commuting (airport staff & college) from origins that aren't immediately served by local buses?

Cathays pops up a lot... and it wouldn't surprise me if its commuter flows have held up better than some others post-Covid, given the student demand - both to Cathays, and commutes from Cathays to Trefforest. Although many of the students live in Cathays, anecdotally there do seem to be a lot who commute, along with university & associated industry employees.

A few years ago I overheard someone asking the person at the other end of the phone line to do something in their house while they were out. A potential intruder was given the opportunity to learn where they lived, when they would be away, how to find the spare key, and how to avoid a reaction from the dog...