At present, Ascott, Finstock and Combe each get one peak-time train in each direction per day. (Shipton gets a couple more in the evening.) That's the "Oxfordshire Halts" service which forms part of the peak variants. I wouldn't see any reason to change that and I don't think the NCLTF are proposing changing it either - they're tiny little halts with no parking and, in the case of Finstock and Combe, some distance from the settlements they serve.

Regular readers will remember that the North Cotswold Line Taskforce - a body led by county councils, with Network Rail and GWR involvement - has proposed a new service pattern for the Cotswold Line, backed up by infrastructure enhancements.

Recap

The core pattern involves 2tph from Worcester, with 1tph calling at Vale of Evesham stations; all at Moreton; the other 1tph at Kingham/Charlbury; and all at Hanborough. Additional 'Oxford Metro' services bring Hanborough up to 4tph. To support this, redoubling is proposed between Wolvercot Junction (near Oxford) and Hanborough, and between Evesham and Pershore.

You can see the plans here: http://researchbriefings.files.parliament.uk/documents/CDP-2020-0004/NORTH-COTSWOLD-LINE-TASK-FORCE-STRATEGIC-BUSINESS-CASE-DEC-2019_ISSUE_110120-(002).pdf

The key pages are p5, a diagram showing the service pattern; p46, the options considered (option 5 is the one selected); and p50, which has a sample weekday timetable.

Although Worcester benefits with two faster services to Oxford/London per hour, most stations on the line won't benefit - they remain at 1tph, perhaps with minimal journey time improvements. Alternating between "Vale of Evesham stoppers" and "Oxfordshire stoppers" breaks connectivity between the two areas. Broadly, the only stations to benefit significantly are Worcester, Moreton and Hanborough; other stations, including the two busiest at Charlbury and Evesham, see no improvement.

Context

Worcester has a long-standing aspiration for faster trains to London. The merits (or otherwise) of this have been extensively debated on this board over the years and I don't propose to rehash them.

In Oxfordshire, housing is being developed at Salt Cross (nearest station, Hanborough: 2200 houses), West Eynsham (Hanborough, 550), East Chipping Norton (Charlbury/Kingham, 1200), and North Witney (Charlbury/Hanborough, 1400). Improvements to the A40 will provide better bus services for the Witney/Eynsham corridor, but longer-distance traffic is expected to go by train.

Further west, new homes are planned at Long Marston (Honeybourne: up to 3500), in northern Pershore (700), and on the southern edge of Worcester (Worcestershire Parkway: 2000+).

There's also the issue of the Honeybourne-Long Marston-Stratford line, which Warwickshire County Council is keen to reopen.

Scoping out an alternative

I believe strongly that the proposals are not good enough as they stand. They offer nothing to most of the line, and indeed reduce through travel options. Housing developments in the Vale of Evesham and West Oxfordshire will generate significant demand that, outside Hanborough, these plans don't address.

They also do nothing to address the biggest gap in the Cotswold Line service at present - the limited journey opportunities from the Vale of Evesham. There is a significant travel market from Evesham and Pershore to Birmingham, as well as to other towns within Wychavon district and elsewhere. The NLCTF proposals have this as a "Future extension option" conditional on Honeybourne-Stratford reopening.

So I've been looking at whether something better is possible. The constraints I've assumed are:

• Through services from Paddington, departing/arriving Oxford at the same time, served by IETs
• Infrastructure enhancements identical, or similar to, the NLCTF proposals
• 2tph base service including faster services for Worcester
• Oriented around principal travel patterns (Vale of Evesham<->Worcester; eastern stations<->Oxford and London) but retaining other journey opportunities where possible
• Potential direct services from the Vale of Evesham to the West Midlands

The alternative service pattern

Happily - it is indeed possible! In brief:

• 1tph principal stations Oxford-Worcester, as at present
• 1tph principal stations Oxford-Moreton, then Evesham, Worcestershire Parkway, Worcester
• 1tph Vale of Evesham stopper, all stations Honeybourne-Worcester
• 2tph 'Oxford Metro' Oxford-Hanborough

This gives the following timetable:



Single-line utilisation at the eastern end (Charlbury-Hanborough): up xx13-20 [H], down xx23-xx30 [C], up xx43-50 [H], down xx53-00 [C]

Single-line utilisation at the western end (Norton Jn-Pershore): up xx07-xx14 [P], down xx15-23 [N], up xx29-xx35 [P], down xx36-44 [N], up xx50-xx57 [P], down xx58-06 [N].

(Blue = single-line sections. Purple = Oxford Metro services. [H]/[C] etc. = where services cross. All section timings are from current WTT as per RealTimeTrains. All times departure unless otherwise shown.)

Going further

I've concentrated on the regular plan between Worcester and Oxford as this is where the biggest infrastructure constraints (and proposed spending) are.

The Vale of Evesham stopper is expressly intended to continue to Droitwich and Birmingham. Considering how best to slot this into the existing Worcester-Birmingham pattern would be an exercise in itself.

There are many options for what the other trains do west of Shrub Hill. The NLCTF plan envisages increasing Malvern to hourly, which would be easy to add onto the 2tph Worcester-Oxford service. They are also proposing extending half of the services (the Oxfordshire stoppers) to Kidderminster, which seems a dubious proposition to me, but nonetheless Kidderminster extensions could be achieved with this plan too. Needless to say, the more trains that can continue to Foregate Street as the city centre station, the better.

In the event of the Honeybourne-Stratford line reopening, the Vale of Evesham stoppers could be extended to Stratford with no further infrastructure work required on the Cotswold Line.

This proposal vs NCLTF proposal

The service level at each station:

• Oxford: no difference
• Hanborough: 4tph east, 2tph west - no difference
• Charlbury: 2tph - extra 1tph, retains all existing through services
• Kingham: 2tph - extra 1tph, retains all existing through services
• Moreton: 2tph - no difference
• Honeybourne: 1tph east, 2tph west - extra 1tph west, retains all existing through services
• Evesham: 3tph - extra 2tph, retains all existing through services
• Pershore: 2tph - extra 1tph, retains all existing through services
• Worcestershire Parkway: 3tph - extra 1tph
• Worcester Shrub Hill: no difference

Journey times Shrub Hill-Oxford are 1hr15 for the principal stations, 1hr06 for the fast. This compares to the NCLTF proposal of 1hr08 and 1hr04. In other words, Worcester still gets an hourly fast service (2 minutes slower than the NCLTF proposal), which hits the magic 2hr00 from Paddington (2hr01 up). I suspect there's the opportunity to shave another couple of minutes off here and there.

The enhanced service at Evesham and Worcestershire Parkway has the potential to be a gamechanger. With a 20-minute interval at Parkway, connecting onto CrossCountry services becomes much more attractive: Worcester gets a fast regular connection onto the Cardiff service, for example. Meanwhile, Evesham effectively gets a turn-up-and-go service to Worcester at half the travel time of the X50 bus. Evesham, the biggest settlement between Oxford and Worcester, has a lot of untapped potential.

Notes

This is a standard hourly pattern. The Cotswold Line does of course have a different pattern in the peaks and I'd anticipate that would continue. Conversely, the service could be thinned down in the evenings.

I looked into omitting Evesham from the "fasts", but the constraints of the single line mean there's no time advantage in doing so.

The Vale of Evesham stoppers continue to Honeybourne principally because this is a practical place to turn the train. In order to pick up passengers here you'd probably need to build out an extra platform face on the Long Marston side. Alternatively, you could run ECS here, or not go to Honeybourne at all but reverse into the Evesham siding instead.

In the interests of honesty, what downsides does this proposal have? Two that I can see. First, the single line between Pershore and Norton Junction is intensively worked, with trains regularly crossing at Pershore. There are two longer recovery periods per hour, but ideally some modelling would look at the likely impact of delays here, and consider whether moving the single line section would help. Second, Worcester departures are regular to Evesham (every 20 minutes) but have 40/20 intervals to Oxford. This is inevitable given that the two services need to arrive at Oxford on the half-hour, but one has fewer stops. I don't see it as a major issue - it happens all round the network, day in, day out.


So. Shoot me down. What have I missed?

5 is Ely.

Which I really should have recognised as I walked past that spot on my way to work every day for a year!

Foxfield?

I swear I should know R but...

The OS has always been rather ornery on canal names. It called the Llangollen and the Montgomery the "Shropshire Union" well into the 90s, long after the post-war names had been widely adopted. Even today it uses the "River Dun Navigation" for a stretch of waterway that everyone calls the Sheffield & South Yorkshire Navigation.

9 is Hebden Bridge on the Rochdale Canal - one of my favourites.


Heh - there really is no right answer on what to call the constituent bits of the GU system.

The Grand Union Canal Company was effectively formed by the Regent's Canal buying the Grand Junction and the 'Warwick' canals. But despite that, people have still continued to call the Paddington-Limehouse bit the Regent's Canal. The "Grand Union" usually describes the Brentford-Birmingham main line and its various branches, though a few people still call it the Grand Junction.

Canal and river names can be confusing, but spare a thought for the publishers of the Nicholsons waterways guides, who recently(ish) issued their Severn, Avon & Birmingham book with a beautiful picture of Pulteney Bridge over the River Avon. It looked great. Shame it was the wrong Avon.

Yes - the Devon Coast-to-Coast (NCN 27) between Meldon and Lydford, and then for a short way on the run-in to Tavistock.

Yes - exactly there.

(I cycled past it a couple of months ago while riding the new King Alfred's Way.)

Stakeholder email this morning:

I like this. Could name a few of them but for my "one each at first" it does, of course, have to be 6 - Charlbury.

It's 8.7m long, and travelled up and down a 500m track in a desolate wasteland. This has cost their investors $400m.

Should we tell them you can get the same experience in the UK for one pound 40?

Or John Prescott's wonderful "The Green Belt is a Labour achievement, and we intend to build on it."

Easier to manage conflict in Utrecht, though, where there's high-quality cycle infrastructure everywhere. The issue with the BBRP is that it's pretty much the only good-quality, safe cycle route for miles around, so everyone from high-speed Stravaites to toddlers on balance bikes gets funneled onto the one path.

Meanwhile, here in Oxfordshire, the county council's "cycling champion" (Suzanne Bartington, the excellent councillor for Witney North and a rail advocate too) has just resigned, citing a lack of support from her Cabinet.

There are some good pics of the work to be done at https://www.cotswoldcanals.net/photo_index.php?cid=1b

(You can also browse through the site and explore the remaining derelict sections of the canals. There'll be a couple of tricky railway crossings heading east from Stroud, but that's currently allocated to "Phase 3"...)