GWR Bristol - Oxford service

[Snow1964]

The half-hourly 'superfast' services were timetabled to run via Bristol Parkway and Hullavington and would therefore have interacted with a different subset of crossings (many of the 13 must be foot crossings, or at least I can only recall road crossings at Steventon and just south of Appleford on this route).

I might be wrong, but fairly sure the route via Badminton (opened 1903) was in an era where all crossings of a main line were bridges.

Almost certainly any crossings are on older section, and Oxford line

 

[Falcon1200]

just south of Appleford on this route).

Rail traffic at Appleford surely reduced with the reduction of the Didcot/Oxford stoppers from half-hourly to hourly? The Oxford/Bristol service would not increase this above historical levels (and there is no longer an incessant flow of coal trains to and from Didcot Power Station either!)

Increased rail traffic over level crossings seems a strange way to object to improved rail services, presumably on the basis of inconvenience to road users? Or is there a belief that more accidents will occur, if so is there data from other routes where train frequency has increased to support this?

 

[Snow1964]

Or is there a belief that more accidents will occur, if so is there data from other routes where train frequency has increased to support this?

It's a basic mathematical theory (not comparison based) let's say as example there is an incident every 200,000 trains (I don't know the numbers, and clearly will vary by crossing anyway).

let's say 30 trains per day x 363 days = 10,890 trains per year
So risk says incident likely 18 years 4 months

Now if run extra trains, say 5 extra each way per day.
(30+5+5) x 363 = 14,520 per year
So risk now says incident likely every 13 years 9 months

You get idea, more trains per day is in theory more risk, so Network Rail suddenly decide upgrade is required to restore incident risk rate to 18y 4months.

But of course the theory has flaws, what if someone previously cut services, perhaps coal trains to former Didcot power station. What base year are they using. Lots of other variables, example did previous era train have high level headlight, was it noisier. Why is an IET on a Bristol-Oxford seen as different to a Paddington-Oxford IET on same crossing?

Of course £1.75m is lot more than cost of some new wicket gates and better or bigger signs, so what exactly is needed if service starts that is deemed not been needed upto today. Perhaps some improvements on paper improve safety (or maybe won't in practice if someone will still misuse crossing). But still a mystery where that amount came from (was it plucked out of thin air)

 

[fishwomp]

The road crossings are Causeway and Stocks Lane, just west of Steventon and Appleford.

Both of which are barrier controlled - at Appleford you have to press a button to notify the signaller that you want the gate to be lifted - it's a rarely used road/track to the appleford sidings (which the trucks for it don't use as they have an alternative) and a bridleway.

 

[Falcon1200]

so Network Rail suddenly decide upgrade is required to restore incident risk rate to 18y 4months.

Interesting detail, thanks, but is it not more that the ORR have deemed the LC upgrades necessary, rather than Network Rail? And if those ipgrades are necessary, someone has to pay for them!

 

[Snow1964]

Interesting detail, thanks, but is it not more that the ORR have deemed the LC upgrades necessary, rather than Network Rail? And if those ipgrades are necessary, someone has to pay for them!

I think the debate is more along lines of if ORR deems upgrade required, then Network Rail needs to do it as early as possible.

But if deemed doesn't require doing at current time, but suddenly will need doing if extra train a hour starts, why should extra train have to pay both track access fees and for network upgrade

Its a bit like the principle of if you regularly use a supermarket, then it continues as is, but if you are a new customer then supermarket wants you to contribute to cost of installing an extra till because it is now busier.

 

[Stompehh]

The need for level crossing upgrades was raised by Network Rail in their submission and then taken into consideration by the ORR. The ORR didn't directly come to that conclusion themselves.

 

[The exile]

It's a basic mathematical theory (not comparison based) let's say as example there is an incident every 200,000 trains (I don't know the numbers, and clearly will vary by crossing anyway).

let's say 30 trains per day x 363 days = 10,890 trains per year
So risk says incident likely 18 years 4 months

Now if run extra trains, say 5 extra each way per day.
(30+5+5) x 363 = 14,520 per year
So risk now says incident likely every 13 years 9 months

You get idea, more trains per day is in theory more risk, so Network Rail suddenly decide upgrade is required to restore incident risk rate to 18y 4months.

It’s also not just a matter of the stats - the greater the density of train service, the greater the likelihood that crossings stay shut after the passage of one train because the next one’s coming (and so on). Particularly at half-barrier or unbarriered crossings people then get tempted to take the law into their own hands.

 

[jazzy]

Apologies if I’ve missed it when scanning this thread but do we know what timings these new services would be? Am I right in thinking this represents an extra Bristol, Bath, Chippenham, Swindon train every two hours?

 

[Bald Rick]

It's a basic mathematical theory (not comparison based) let's say as example there is an incident every 200,000 trains (I don't know the numbers, and clearly will vary by crossing anyway).

let's say 30 trains per day x 363 days = 10,890 trains per year
So risk says incident likely 18 years 4 months

Now if run extra trains, say 5 extra each way per day.
(30+5+5) x 363 = 14,520 per year
So risk now says incident likely every 13 years 9 months

You get idea, more trains per day is in theory more risk, so Network Rail suddenly decide upgrade is required to restore incident risk rate to 18y 4months.

But of course the theory has flaws, what if someone previously cut services, perhaps coal trains to former Didcot power station. What base year are they using. Lots of other variables, example did previous era train have high level headlight, was it noisier. Why is an IET on a Bristol-Oxford seen as different to a Paddington-Oxford IET on same crossing?

Of course £1.75m is lot more than cost of some new wicket gates and better or bigger signs, so what exactly is needed if service starts that is deemed not been needed upto today. Perhaps some improvements on paper improve safety (or maybe won't in practice if someone will still misuse crossing). But still a mystery where that amount came from (was it plucked out of thin air)

Level Crossing risk is much more complex than that. I have posted on other threads the stripped down simple version.

 

[Roger B]

I've skimmed the ORR document (https://www.orr.gov.uk/sites/defaul...al-agreement-directions-letter-2026-05-05.pdf).

It states (para 2) that "The application proposed the introduction of services initially every two hours from the Subsidiary Change Date (SCD) in May 2026".

However, I'm thinking that that's unlikely, given the level crossing risk discussed above.

Does anyone know how long it will take to overcome this challenge, and when services are expected to resume (over and above the Saturdays-only service in the winter time-table)?

Many thanks

 

[cle]

Between this and the EWR shambles - not to mention Botley Road - all a bit of a sh-tshow for rail in Oxford recently. We really can't get out of our own way in this country.

 

[800301]

I've skimmed the ORR document (https://www.orr.gov.uk/sites/defaul...al-agreement-directions-letter-2026-05-05.pdf).

It states (para 2) that "The application proposed the introduction of services initially every two hours from the Subsidiary Change Date (SCD) in May 2026".

However, I'm thinking that that's unlikely, given the level crossing risk discussed above.

Does anyone know how long it will take to overcome this challenge, and when services are expected to resume (over and above the Saturdays-only service in the winter time-table)?

Many thanks

As planned from Monday we’ve heard now

 

[officewalla]

ORR signed off on the risk assessments according to one particular grapevine that reached my way. Upto NR to implement any changes required now if that is the case!

 

[Class707SE]

Services seem to be on Real Time Trains!

 

[cle]

Sorry which ones - the seasonal weekend ones or new ones?

 

[Class707SE]

Sorry which ones - the seasonal weekend ones or new ones?

New ones - https://www.realtimetrains.co.uk/service/gb-nr:G15058/2026-05-18

 

[Phillipimo]

Services seem to be on Real Time Trains!

I note that in the hours they run, Bath to Bristol will have roughly a train every 10 minutes which is very impressive. The weekday timetable also seems suitable for Bristol commuters which should help relieve other services.

 

[cle]

Ok so these are happening with immediate notice, and the level crossing investment / shakedown is a future-issue to solve in parallel?

 

[Benjwri]

Ok so these are happening with immediate notice, and the level crossing investment / shakedown is a future-issue to solve in parallel?

The level crossing issues have already been sorted, at least in the short term, as reported by RAIL.

Great Western Railway can operate Oxford-Bristol services from Monday (May 18) after level crossing concerns were addressed in time.

The Office of Rail and Road approved plans for the two-hourly service to run as an 18-month trial on May 5, but had stipulated they could only start once Network Rail had installed mitigation measures at level crossings along the route.

NR had said the GWR services, running six days a week, would lead to a “considerable” increase risk levels at 13 level crossings with Bathampton and Canalside 1 (near Royal Wootton Bassett) being the main concerns.

The former has an average of seven daily users but has been the scene of two recent near misses. In the last week vegetation has been cleared to improve visibility and clearer markings installed.

Canalside 1 crossing is used 53 times a day. Temporary safety steps have been installed to improve access while CCTV will also be installed and an awareness campaign launched.

Line speed will be reduced from 125mph to 100mph.

Both crossings will have Overlay Miniature Stop Lights installed though no timescale has been given.

 

[Class707SE]

Line speed will be reduced from 125mph to 100mph.

Will this not affect timings of other services? Or is it too small/short section to affect things?

 

[John R]

Will this not affect timings of other services? Or is it too small/short section to affect things?

Eastbound traffic will shortly be restricted to 80 for the diverge to the usual approach to Swindon, and traffic from the Badminton route is limited to (I think) 75, so I suspect the effect will be limited in that direction.

Similarly on the westbound track towards Badminton there is a 75 restriction, and I’m not sure for Chippenham bound trains whether they will have got much above 100 by that point anyway.

So not as bad as a 100 stretch in between two longer 125mph sections.

 

[brad465]

Level Crossing risk is much more complex than that. I have posted on other threads the stripped down simple version.

Can you post a link to at least one of those posts so I can read it please? (the forum search function isn't ideal for finding this)

 

[Class707SE]

Eastbound traffic will shortly be restricted to 80 for the diverge to the usual approach to Swindon, and traffic from the Badminton route is limited to (I think) 75, so I suspect the effect will be limited in that direction.

Similarly on the westbound track towards Badminton there is a 75 restriction, and I’m not sure for Chippenham bound trains whether they will have got much above 100 by that point anyway.

So not as bad as a 100 stretch in between two longer 125mph sections.

Ah understood. Thanks for explaining.

 

[Snow1964]

GWR has recently published its performance figures for 4 weeks in April

There were 237 short forms on IET services (about 9 per day)

If they couldn't find enough stock last month, where are the IETs going to come from for these extra trains (and there were short forms on Devon and Cornwall locals too, so don't say released by 175s)

https://www.gwr.com/-/media/gwr-sc-website/files/publications/performance-report/gwr-performance-report-period-2701.pdf

 

[The exile]

GWR has recently published its performance figures for 4 weeks in April

There were 237 short forms on IET services (about 9 per day)

If they couldn't find enough stock last month, where are the IETs going to come from for these extra trains (and there were short forms on Devon and Cornwall locals too, so don't say released by 175s)

https://www.gwr.com/-/media/gwr-sc-website/files/publications/performance-report/gwr-performance-report-period-2701.pdf

Well, the diagram that starts with the 06.55 Bath - Bristol and then goes to Gloucester at 07.40 (no idea what it does / did after that) was an IET up to last week and is now booked to come off depot coupled to another Turbo, so there’s one for a start.

 

[GilfachFargoed]

GWR has recently published its performance figures for 4 weeks in April

There were 237 short forms on IET services (about 9 per day)

If they couldn't find enough stock last month, where are the IETs going to come from for these extra trains (and there were short forms on Devon and Cornwall locals too, so don't say released by 175s)

https://www.gwr.com/-/media/gwr-sc-website/files/publications/performance-report/gwr-performance-report-period-2701.pdf

Why not?, released by 175s is the correct answer.

 

[Class707SE]

Only 2 IETs are needed for this service. As mentioned, the 175s will help.

 

[sir_gummerz]

GWR has recently published its performance figures for 4 weeks in April

There were 237 short forms on IET services (about 9 per day)

If they couldn't find enough stock last month, where are the IETs going to come from for these extra trains (and there were short forms on Devon and Cornwall locals too, so don't say released by 175s)

https://www.gwr.com/-/media/gwr-sc-website/files/publications/performance-report/gwr-performance-report-period-2701.pdf

Theres several bristol to Worcester services that previously had IETs, and now use a turbo instead. At least based on what I've seen over the last week

 

[WelshBluebird]

Theres several bristol to Worcester services that previously had IETs, and now use a turbo instead. At least based on what I've seen over the last week

And where is that turbo coming from given those also end up being in short supply?