What could an “ideal” Bristol service recast look like?

[WTT Gremlin]

Would there be enough capacity for all of this between Bristol and Bradford-on-Avon, bearing in mind you would need to fit in Weymouth, Oxford and London trains too?

Also can the single track Melksham line handle 2tph to Bristol in addition to the existing Swindon - Westbury service (which should ideally be 1tph).

If flighted then probably, though would likely cause issues elsewhere/require more interventions.

As I previously mentioned without a recast of the Great Western Mainline, the South Wales Mainline, and the Cut-Off, then the amount of fundamental WTT changes at Bristol will likely be limited.

Sorry I wasn't clear by "double Melksham" I meant the entire line.

 

[brad465]

Would there be enough capacity for all of this between Bristol and Bradford-on-Avon, bearing in mind you would need to fit in Weymouth, Oxford and London trains too?

Also can the single track Melksham line handle 2tph to Bristol in addition to the existing Swindon - Westbury service (which should ideally be 1tph).

Not only does the branch need double track, but at one or both ends a spur needs adding to prevent a reversal being required. There definitely used to be one at each end though so this isn't something that requires extreme work. The Trowbridge end though probably makes more sense as those trains can also serve Bradford-on-Avon, while a steep slow curve at the Chippenham end could be too obstructive on the mainline.

At a minimum the branch needs double track at each end, so trains don't have to wait on the mainline, while Melksham could also do with doubling, especially if it were to host terminating trains doing a direct service to Bristol, providing accessibility stuff can be delivered cheaply (i.e. with unicorn dust).

I would also consider having a station at Bathampton that is positioned to allow express services to pass: there's already a loop on the up side, while perhaps on the down side the junction could be extended to allow a train calling there to be routed at the same time a down express needs to pass through.

 

[PTR 444]

Not only does the branch need double track, but at one or both ends a spur needs adding to prevent a reversal being required. There definitely used to be one at each end though so this isn't something that requires extreme work. The Trowbridge end though probably makes more sense as those trains can also serve Bradford-on-Avon, while a steep slow curve at the Chippenham end could be too obstructive on the mainline.

At a minimum the branch needs double track at each end, so trains don't have to wait on the mainline, while Melksham could also do with doubling, especially if it were to host terminating trains doing a direct service to Bristol, providing accessibility stuff can be delivered cheaply (i.e. with unicorn dust).

I would also consider having a station at Bathampton that is positioned to allow express services to pass: there's already a loop on the up side, while perhaps on the down side the junction could be extended to allow a train calling there to be routed at the same time a down express needs to pass through.

Another idea I have, which is perhaps more unconventional and may be completely impractical, is to reinstate both chords but keep the line single track and have Melksham served by a one-way loop service to/from Bristol/Bath. Would go Bath to Melksham via Bradford-on-Avon, then return to Bath via Corsham.

Fitting the Swindon - Westbury service alongside this might be a challenge though.

 

[The exile]

Would there be enough capacity for all of this between Bristol and Bradford-on-Avon, bearing in mind you would need to fit in Weymouth, Oxford and London trains too?

Also can the single track Melksham line handle 2tph to Bristol in addition to the existing Swindon - Westbury service (which should ideally be 1tph).

If the fast Bristol - London via Badminton service were reinstated (it did -just - get off the ground in 2019/20) then paths could potentially be gained between Bristol and Bathampton by slowing the via Baths. OK, the fastest Bath /Chippenham - Bristol journey times would be lengthened, but actual travelling time for most would be unaffected due to the increase in frequency.

== Doublepost prevention - post automatically merged: 22 Mar 2026 ==

Also can the single track Melksham line handle 2tph to Bristol in addition to the existing Swindon - Westbury service (which should ideally be 1tph).

Melksham line would need doubling (but it probably does anyway!)

 

[PTR 444]

If the fast Bristol - London via Badminton service were reinstated (it did -just - get off the ground in 2019/20) then paths could potentially be gained between Bristol and Bathampton by slowing the via Baths. OK, the fastest Bath /Chippenham - Bristol journey times would be lengthened, but actual travelling time for most would be unaffected due to the increase in frequency.

== Doublepost prevention - post automatically merged: 22 Mar 2026 ==

If Bristol - Bath - Oxford called at all stations, you could potentially speed up the “Via Bath” London trains by making Didcot and Swindon alternate 1tph calls. The same could then be done for the South Wales - London trains.

 

[NCT]

The existing network - work-in-progress towards full MetroWest, isn't bad. Taking one corridor at a time:

Filton

Serves two long distance directions (well, three, but practically only two) and local branches. You'd want to use the recently delivered four-tracking to full potential.

Local services - 2tph to Avonmouth (of which 1tph to Seven Beach), and 2tph to Henbury, calling at all stations on their respective corridors

Cardiff direction - 2tph, from Bristol first stop Filton Abbey Wood so that station retains northbound connectivity

Cheltenham direction - 2tph Cross Country (hopefully uncontroversial) and 2tph Gloucester all stations (ideally both tph rather than just the one would continue to Worcester if an inexpensive node solution can be found at Worcester)

Swindon direction - there's no capacity at the Paddington end for the desired Bristol superfasts, so alas it's a 0tph here.

Bristol suburban stations of Ashley Down and Filton Abbey Wood get 4-6 tph, hopefully at relatively even intervals.

Between Temple Meads and Stapleton Road there is 10tph over 4 tracks - ought to be a doddle.

Bedminster

Portishead really ought to be 2tph unless the extent of single track presents a hard constraint.

There ought to be a 2tph stopping service towards Weston-Super-Mare.
On top of that a 2tph semi-fast service to Taunton via Weston (this serve gives Highbridge & Burnham, and Bridgwater, 2tph, and gives Nailsea and Backwell a 2tph top-up)

I'd then run 2tph Cross Country to Exeter. Exeter would be a Swiss-style takt node with 2tph from Bristol and 2tph from Paddington. Of these 4tph, 2tph would go to Plymouth (and beyond), ideally 1tph each from Bristol and Paddington, if takt can be made to work. This means there's a 2tph opportunity to travel between Bristol and Plymouth.

A total of 8tph - should be doable.

Bath

Towards Swindon - 2tph London (uncontroversial) and 1tph Oxford (2tph would be nice but let's walk before we run)

Towards Bradford and Westbury -
- 2tph to Salisbury (I'm not specifying where those services go beyond Salisbury. I want Salisbury to be a takt node so that if 1tph terminates at Salisbury as today it would connect into the Romsey Rocket to provide a combined 2tph opportunity to travel to Southampton)
- 2tph (ideally) to Castle Cary (of which 1tph to Weymouth).

It's probably the Castle Cary service group that would pick up the local stops between Temple Meads and Westbury.

7tph, should be doable.

I'm sceptical of the value of joining these up, especially where they require a reversal, but Exeter can probably be joined up with Cardiff or Cheltenham.

Agree only to an extent. There is probably no revenue reason for linking anything up, or for any particular pairing. However, Temple Meads' layout is a bit tricky. The temptation would be to link the as many of my 10tph Stapleton Road with 8tph Bedminster, and let everything from Bath direction terminate. My worry is I'm not sure platforms 12, 13 and 15 are sufficient to turnaround my 7tph Bath direction services, though you can shunt to the sidings to the west.

Some of them need to be longer trains. They also need to be not the "inter city" trains, where I'd just give Birmingham-Bristol-Exeter-Plymouth 1tph, though again with a long train.

I think only the London and Plymouth - Birmingham trains need to be intercity, the rest should all be regional. The 80x fleet needs to be concentrated on London intercity routes to minimise/eliminate 5-car workings. For the other services, given current typical service lengths, there's a great deal of train lengthening opportunity before we start to talk about frequency increase beyond the 2tph building blocks.

 

[The exile]

If Bristol - Bath - Oxford called at all stations, you could potentially speed up the “Via Bath” London trains by making Didcot and Swindon alternate 1tph calls. The same could then be done for the South Wales - London trains.

There would be no need to do so to counteract my proposed slowing as that would really only affect London journeys from Bristol (which would have gained fast services via Badminton). The main “losers” would be Bath /Chippenham - but even there, the faster journey times to Bristol would be longer but compensated for by shorter waits because of increased frequency.

 

[lachlan]

What Melksham needs is a direct service to Bath Spa and Bristol. I'd go for a 2tph Melksham-Bath Spa-Bristol if it could be fitted in. If single track is involved it needs to be self-contained. Depending on timings these could also be the trains to serve Keynsham.

Bath to Swindon via Melksham is a longer route than via Corsham - and any trains diverted that way are trains that can't call at a reopened Corsham but still take up space at Bath and Swindon.

What Melksham needs is a more frequent service for connecting to an improved Chippenham service

This would be a very bad idea. it would introduce significant performance risk on East West Rail for very little benefit. For most journeys from west of Oxford to Cambridge and east thereof, via London is still going to be the best route, especially now that the Elizabeth Line is open.

It's 45 minutes from Chippenham to Oxford on the direct Saturday train, and if the stated EWR journey time of 1hr35 is accurate even with a change it will likely be quicker via Oxford. Similar story to Bletchley and Milton Keynes and avoids the faff of changing in Bath.

 

[The exile]

Bath to Swindon via Melksham is a longer route than via Corsham - and any trains diverted that way are trains that can't call at a reopened Corsham but still take up space at Bath and Swindon.

So -Oxfords call at Corsham and Melksham gets a Bristol - Chippenham (or Swindon if space permits) stopper in addition to the Trans Wilts.

 

[brad465]

[Bath]

Towards Swindon - 2tph London (uncontroversial) and 1tph Oxford (2tph would be nice but let's walk before we run)

Towards Bradford and Westbury -
2tph to Salisbury (I'm not specifying where those services go beyond Salisbury. I want Salisbury to be a takt node so that if 1tph terminates at Salisbury as today it would connect into the Romsey Rocket to provide a combined 2tph opportunity to travel to Southampton)
2tph (ideally) to Castle Cary (of which 1tph to Weymouth).

It's probably the Castle Cary service group that would pick up the local stops between Temple Meads and Westbury.

7tph, should be doable.

The 2tph to Castle Cary ought to be all to Yeovil Pen Mill, then 1tph to Weymouth and maybe the other goes to Yeovil Jct. This will require double track somewhere along the Yeovil-Castle Cary section though; perhaps a dynamic loop centred on Sparkford (with a reopened parkway station at the A303 intersection if demand warranted it).

 

[The exile]

2tph (ideally) to Castle Cary (of which 1tph to Weymouth).

It's probably the Castle Cary service group that would pick up the local stops between Temple Meads and Westbury.

I know they’ve only just acquired the second tph, but it won’t be long before a half hourly service for Oldfield Park, Keynsham and eventually Saltford will be insufficient.

 

[Magdalia]

Bath to Swindon via Melksham is a longer route than via Corsham - and any trains diverted that way are trains that can't call at a reopened Corsham but still take up space at Bath and Swindon.

I go back to the numbers, five of the top 10 flows are on the Swindon-Chippenham-Bath-Bristol axis.

606 Bath Spa-Bristol
270 Bristol-Cardiff
254 Bath Spa-Chippenham
178 Bristol-Cheltenham
164 Bath Spa-Trowbridge
157 Bath Spa-Swindon
149 Bristol-Taunton
141 Bristol-Chippenham
141 Bristol-Swindon
139 Bristol-Clifton Down

Those passengers are not going to want or use a slower service via Melksham. Having read the comments I think a reopened station at Corsham is a far better option, especially as the station could be a sort of Melksham Parkway.

It's 45 minutes from Chippenham to Oxford on the direct Saturday train, and if the stated EWR journey time of 1hr35 is accurate even with a change it will likely be quicker via Oxford. Similar story to Bletchley and Milton Keynes and avoids the faff of changing in Bath.

But the numbers of people making these through journeys will be tiny. The important thing is to have the Bristol-Bath-Swindon-Oxford service. The few people going through to/from EWR destinations change at Oxford.

Agree only to an extent. There is probably no revenue reason for linking anything up, or for any particular pairing. However, Temple Meads' layout is a bit tricky.

One of the reasons that I didn't want to be prescriptive here is that it is very dependent on what works for the existing infrastructure.

I think only the London and Plymouth - Birmingham trains need to be intercity, the rest should all be regional. The 80x fleet needs to be concentrated on London intercity routes to minimise/eliminate 5-car workings. For the other services, given current typical service lengths, there's a great deal of train lengthening opportunity before we start to talk about frequency increase beyond the 2tph building blocks.

There is much to agree with here! Longer trains is the way to increase capacity. The Swindon-Chippenham-Bath-Bristol axis can shift so many people because it already has the long trains.

 

[NCT]

I know they’ve only just acquired the second tph, but it won’t be long before a half hourly service for Oldfield Park, Keynsham and eventually Saltford will be insufficient.

I think train lengths can do much of the legwork before we get into frequency increase territory. Being a two-track railway I think timetabling will get tricky once you get beyond 2tph for these stations. Without the ability to run Bristol superfasts, the case for pathing out Bristol - London via Chippenham (there would be a genuine risk of this happening if the Keynshams have more than 2tph) would be very challenging to make.

 

[PTR 444]

If Bristol - Oxford doesn’t warrant 2tph, could the other 1tph run semi-fast to Reading instead?

That way, you have a proper semi-fast which calls at all stations between Bath and Didcot Parkway, including those which are planned for reopening. The main driver is that it would give Grove/Wantage a direct link with Didcot for connections to London, something which is absent from the Oxford service.

 

[NCT]

But the numbers of people making these through journeys will be tiny. The important thing is to have the Bristol-Bath-Swindon-Oxford service. The few people going through to/from EWR destinations change at Oxford.

This I don't agree with. I see Swindon and Bristol as an obvious extension of the Cambridge - Oxford knowledge arc, and I don't see Oxford being an obvious break point in travel patterns. The way Oxford station is set up favours trains running through rather than a load of services terminating from both ends.

I don't get this mentality that Cambridge needn't bother with the great unwashed west of Oxford or Peterborough.

 

[The exile]

I think train lengths can do much of the legwork before we get into frequency increase territory. Being a two-track railway I think timetabling will get tricky once you get beyond 2tph for these stations. Without the ability to run Bristol superfasts, the case for pathing out Bristol - London via Chippenham (there would be a genuine risk of this happening if the Keynshams have more than 2tph) would be very challenging to make.

Train lengths absorb crush loads at peak times but do nothing to make the train the “natural choice” for travel - for short distance journeys (as most from those stations are), only turn-up-and-go cuts the mustard. 8tph between Temple Meads and Bathampton (stopper leaving asap after fast) should surely be do-able.

 

[NCT]

Train lengths absorb crush loads at peak times but do nothing to make the train the “natural choice” for travel - for short distance journeys (as most from those stations are), only turn-up-and-go cuts the mustard. 8tph between Temple Meads and Bathampton (stopper leaving asap after fast) should surely be do-able.

8tph stopper, or 8tph total? Latter on a 4+4 pattern might just be doable under ideal conditions, former no chance. Temple Meads platform availability, Bristol East Junction, Bathampton Junction are all timetable constraints, not to mention services coming from a wider range of locations all with their own constraints. The chance of being able to timetable perfect headways at either end is very slim.

 

[The exile]

8tph stopper, or 8tph total? Latter on a 4+4 pattern might just be doable under ideal conditions, former no chance. Temple Meads platform availability, Bristol East Junction, Bathampton Junction are all timetable constraints, not to mention services coming from a wider range of locations all with their own constraints. The chance of being able to timetable perfect headways at either end is very slim.

8tph total (each way).
Fasts would be:
2tph Paddington
1tph Oxford
1tph Portsmouth

Stoppers:
2tph as now (roughly)
1tph Chippenham direction via Melksham
1tph Chippenham direction via Corsham (subject to station reopening) -as far round the clock face from the Oxford as possible to keep things balanced.
However, I am all too aware of the restraints caused by platform congestion at Temple Meads.

 

[lachlan]

I go back to the numbers, five of the top 10 flows are on the Swindon-Chippenham-Bath-Bristol axis.



Those passengers are not going to want or use a slower service via Melksham. Having read the comments I think a reopened station at Corsham is a far better option, especially as the station could be a sort of Melksham Parkway.

Indeed - no new curves needed. Even increasing the Melksham trains to one per hour would be a big improvement.

 

[The exile]

8tph stopper, or 8tph total? Latter on a 4+4 pattern might just be doable under ideal conditions, former no chance. Temple Meads platform availability, Bristol East Junction, Bathampton Junction are all timetable constraints, not to mention services coming from a wider range of locations all with their own constraints. The chance of being able to timetable perfect headways at either end is very slim.

Those sorts of constraints were of course the rationale behind DB’s S-Bahn concept until the money began to run out - entirely separate tracks for local urban networks. Sadly that is an unattainable dream here!

 

[JohnRegular]

Indeed - no new curves needed. Even increasing the Melksham trains to one per hour would be a big improvement.

If the existing service could go hourly, have good connections at Chippenham/Trowbridge, and most importantly run reliably, the train starts to look like a reasonable option between Bristol/Bath and Melksham.

However, a handful of cancellations or missed connections will be enough to put off many potential commuters. A direct service would be a huge upgrade.

 

[cle]

In addition, once East West Rail is open, the Oxford trains could be extended to Milton Keynes Central, Cambridge, or even Norwich. I know there are some aspirations for EWR services to run beyond Oxford to Cambridge, albeit by doing so it increases the prospect of delays.

There is a good case (once Wellington and Cullompton stations are open) to have this start back at Exeter, rather than having multiple trains that require changes both at Bristol and Oxford.

East West Rail is open now as far as Milton Keynes. Unfortunately, as discussed elsewhere, Chiltern are unable to provide a passenger service at present. Whether Chiltern actually gets to operate the service remains to be seen.

*Any* service between Oxford and Milton Keynes would be most welcome and an improvement on non-existent 2-car 196s. Whilst a 5-car 800 or 802 would be too big for the bay platform at MK, with a bit of imagination an extension to Northampton might be possible whether it is 1tp2h or 1tph.

The last post is the most pertinent. We don't even have spades in the ground for Cambridge, or even Bedford station. I agree that a tph extended to Northampton - and hourlies Oxford-Bristol - are far more likely than EWR to Norwich, or even Cambridge at this point - or in 15 years.

The other way - perhaps 1tph will go onwards (it was historically Reading) - maybe Didcot or Swindon would be helpful and less risk than combining with the Bristol-Oxford. But Oxford-MKC will basically be a shuttle, I think it will be pretty reliable. And MKC is more the goal than Bedford, for now.

 

[Magdalia]

This I don't agree with. I see Swindon and Bristol as an obvious extension of the Cambridge - Oxford knowledge arc, and I don't see Oxford being an obvious break point in travel patterns. The way Oxford station is set up favours trains running through rather than a load of services terminating from both ends.

I don't get this mentality that Cambridge needn't bother with the great unwashed west of Oxford or Peterborough.

We don't have an arc, we only have a corridor, and it goes between Oxford and Cambridge.

The priorities for delivering economic growth in Cambridge are local connectivity for commuting and connectivity with London and Oxford. The purposes of East West Rail are to support economic growth in Cambridge by delivering local connectivity for commuting, especially from Cambourne and Tempsford, and connectivity along the corridor, especially with Oxford.

Having through trains to Swindon and Bristol is not going to be a significant driver of economic growth in Cambridge, and not having through trains to Swindon and Bristol is not going to be a significant impediment to economic growth in Cambridge.

On the other hand, for delivering growth in Oxford, trains to/from Swindon, Bath and Bristol are important for local connectivity.

We don't even have spades in the ground for Cambridge, or even Bedford station.

Without the objective of delivering economic growth in Cambridge, the eastern leg of East West Rail would have died a quiet death a while ago.

 

[cle]

Yes, but a Bristol-MKC is far more realistic than Bletchley-Cambridge right now...

 

[freetoview33]

deleted

 

[Manutd1999]

I've thought a bit more about this and come up with the following ideas. The main aims are a move towards everything running at half-hourly intervals, a more clearly defined intercity/regional/local split and increased frequency for suburban stations.

Intercity:

• 2ph XC and 2ph London trains remain unchanged.
• Voyager fleet permitting, extend at least 1p2h of the Manchester XCs to Exeter/Paignton

Regional:

• Run Bristol-Gloucester as a standalone half-hourly service, with no continuation to/from Westbury
• Add an hourly Bristol-Oxford and hourly Bristol-Frome (extending 1p2h to Weymouth), both running fast to Bath
• Replace the Cardiff-Exeter with a half-hourly semi-fast Taunton-Cardiff. This gives Bridgewater/Highbridge an improved service, plus increases frequency from Bristol to Cardiff to 3ph.
• Cardiff-Portsmouth becomes a regional express, skipping Filton/Patchway/Severn Tunnel Jn.

Then introduce an expanded MetroWest local service, with the following routes served 2ph, calling all stations enroute:

• Weston-Super-Mare - Severn Beach/Avonmouth
• Portishead - Yate
• Westbury - Bath - (reversal at Temple Meads) - Henbury

Combined, this gives a 'turn up and go' frequency along the core route from Parson St/Bedminster through to Filton. It also enhances cross-Bristol connectivity from Bath/Keynsham to Filton and north Bristol

 

[The exile]

I've thought a bit more about this and come up with the following ideas. The main aims are a move towards everything running at half-hourly intervals, a more clearly defined intercity/regional/local split and increased frequency for suburban stations.

Intercity:

• 2ph XC and 2ph London trains remain unchanged.
• Voyager fleet permitting, extend at least 1p2h of the Manchester XCs to Exeter/Paignton

Regional:

• Run Bristol-Gloucester as a standalone hourly service, with no continuation to/from Westbury
• Add an hourly Bristol-Oxford and hourly Bristol-Frome (extending 1p2h to Weymouth), both running fast to Bath
• Replace the Cardiff-Exeter with a half-hourly semi-fast Taunton-Cardiff. This gives Bridgewater/Highbridge an improved service, plus increases frequency from Bristol to Cardiff to 3ph.
• Cardiff-Portsmouth becomes a regional express, skipping Filton/Patchway/Severn Tunnel Jn.

Then introduce an expanded MetroWest local service, with the following routes served 2ph, calling all stations enroute:

• Weston-Super-Mare - Severn Beach/Avonmouth
• Portishead - Yate
• Westbury - Bath (reversal at Temple Meads) - Henbury

The combined service pattern gives a 'turn up and go' frequency along the core route from Parson St/Bedminster through to Filton. It also enhances cross-Bristol connectivity from Bath/Keynsham to Filton

I assume that you’ve just not mentioned the second train to Gloucester that currently runs, rather than cutting it.

 

[Manutd1999]

I assume that you’ve just not mentioned the second train to Gloucester that currently runs, rather than cutting it.

Forgot about that... But yes, it could be a half-hourly Bristol-Gloucester-(Worcester)

 

[brad465]

I've thought a bit more about this and come up with the following ideas. The main aims are a move towards everything running at half-hourly intervals, a more clearly defined intercity/regional/local split and increased frequency for suburban stations.

Intercity:

• 2ph XC and 2ph London trains remain unchanged.
• Voyager fleet permitting, extend at least 1p2h of the Manchester XCs to Exeter/Paignton

Regional:

• Replace the Cardiff-Exeter with a half-hourly semi-fast Taunton-Cardiff. This gives Bridgewater/Highbridge an improved service, plus increases frequency from Bristol to Cardiff to 3ph.

Something still needs to serve Wellington and Cullompton when they reopen, so terminating everything at Taunton won't be an option, unless you want to annoy London passengers by making those services call there instead. You could maybe make it 2tph to Taunton, where 1tph is extended to Exeter (Bristol, WSM, Highbridge, Bridgwater, Taunton, Wellington, Cullompton*, Exeter), with the other 1tph all stations between Bristol and Taunton.

*I've left Tiverton out as calling at all three increases pathing challenges with the faster services, while I can't see many passengers from the two new stations wanting Tiverton Parkway, given the latter station's main purpose is a railhead. Also, if more XC services are extended down to Exeter, these will cater for the Bristol-bound Tiverton passengers.

 

[Magdalia]

• Voyager fleet permitting, extend at least 1p2h of the Manchester XCs to Exeter/Paignton

Regional:

• Replace the Cardiff-Exeter with a half-hourly semi-fast Taunton-Cardiff.

This doesn't work because of the future need to serve the new stations at Wellington and Collumpton. In the near future Tiverton Parkway won't be the only station between Taunton and Exeter.