Cotswold Line redoubling: 2008 - 2011

[whole diary]

[Btline]

Btline, more single-line sections, whatever the length and wherever they are on the line, equal more potential for delay. It doesn't matter where the timetable says trains will pass - it's what happens when they hit problems that counts and the more bits of single line, the worse the problems get. And if Industry Insider had to factor another section into his calculations, then he certainly wouldn't have any hair left!

What do you mean? Redoubling Pershore to Honeybourne would make no difference to the timetable planning, as trains will not be timetabled to pass on that stretch. It does matter where trains pass, because they have to pass on double line, thus the timetable has to work with this. Trains will pass at Evesham. Under the current plans, Evesham will be at the end of a passing loop. My idea would be to put Evesham in the middle of the passing loop (thus making the loop "dynamic" in both directions).

What it would mean is less delays, i.e. Northbound services which will have to wait at Evesham for delayed southbound services (or have 15 mins padding, which is no better).

About trains breaking down - this can happen anywhere, and there would still be disruption (esp at peak times) if a train broke down on the double line. Yes - the route would not be blocked, but I am sure the lack of bi- directional signalling would make passing the broken train a long and slow process.

How often do trains break down here compared to trains get stuck at Evesham? Relatively few I expect.

And I am certain that NR^» (Network Rail - home page) only chose the current option to minimise costs (new platforms etc), which is not their fault.

[Not from Brighton]

I use the Cotswold line for business and leisure (approx twice a month) from Worcester, I'm not a commuter.  For me my priorities for the line are:
1. - Punctuality
2. - Frequency
3. - Journey Time

I doubt there are many daily commuters from Worcester to London (prepared to be shot down!), so I'd expect that most of my fellow travellers might have similar priorities.  If I have to be at a meeting at a certain time then that's when I NEED to be there.  Punctuality has to come first.
The current journey times beat car travel (for me) and although I find all the stops soul destroying, it doesn't stop punctuality coming first.

So I support NR^» (Network Rail - home page)'s efforts to improve punctuality without necessarily improving journey times.

[Chris from Nailsea]

Welcome to the Coffee Shop forum, Not from Brighton - and thanks for your input to this topic!

In principle, I'd agree - if I need to get to a meeting somewhere, I want to be sure I'll get there on time: whether it takes two hours or three hours is a pain, but not quite so critical.

 

[willc]

What do you mean?

What I meant was - and I thought it was pretty clear - was that it doesn't matter where trains are timetabled to pass under normal circumstances. In theory, the current timetable is fine when all goes well, even if many journey times leave much to be desired.

But when things go wrong on a busy line with a series of single-line sections, then problems pile up. Have another of these - big or small - then when there is a problem it will exacerbate resulting delays.

The long double track stretch will help to minimise knock-on delays, as trains on the other line can just keep moving. If a single-line section is blocked/occupied then they can't - and the delays pile up, particularly between Moreton and Evesham.

There will still be crossovers at Moreton and new ones at Honeybourne (hopefully Evesham too), so with single-line pilot working, you would still be able to keep things moving if a train did break down and block one line.

And such situations are very much the exception - the fundamental problem and the one that Network Rail set out to solve was everyday punctuality and reliability and they concluded that 32 miles of double track in the middle of the line was the way to achieve this. Fewer single-line sections equals less waiting around.

As for 'dynamic' loops, Ascott to Moreton was one long before anyone coined the phrase, and they aren't really very dynamic. They are just long loops. If the single line at either end is occupied, then you still have to wait to use it. Most trains are currently timetabled to meet at Moreton and near Shipton anyway, which might not appear the most effective use of the double section. I would hope that in future train meets can be arranged a bit further east than Evesham, to give some extra time margin, as passing at the station would just be replicating the current Shipton/Ascott scenario, where a train only a minute or two late starts to delay one going the other way.
 

[Andy W]

What do you mean?

What I meant was - and I thought it was pretty clear - was that it doesn't matter where trains are timetabled to pass under normal circumstances. In theory, the current timetable is fine when all goes well, even if many journey times leave much to be desired.

But when things go wrong on a busy line with a series of single-line sections, then problems pile up. Have another of these - big or small - then when there is a problem it will exacerbate resulting delays.

The long double track stretch will help to minimise knock-on delays, as trains on the other line can just keep moving. If a single-line section is blocked/occupied then they can't - and the delays pile up, particularly between Moreton and Evesham.

There will still be crossovers at Moreton and new ones at Honeybourne (hopefully Evesham too), so with single-line pilot working, you would still be able to keep things moving if a train did break down and block one line.

And such situations are very much the exception - the fundamental problem and the one that Network Rail set out to solve was everyday punctuality and reliability and they concluded that 32 miles of double track in the middle of the line was the way to achieve this. Fewer single-line sections equals less waiting around.

As for 'dynamic' loops, Ascott to Moreton was one long before anyone coined the phrase, and they aren't really very dynamic. They are just long loops. If the single line at either end is occupied, then you still have to wait to use it. Most trains are currently timetabled to meet at Moreton and near Shipton anyway, which might not appear the most effective use of the double section. I would hope that in future train meets can be arranged a bit further east than Evesham, to give some extra time margin, as passing at the station would just be replicating the current Shipton/Ascott scenario, where a train only a minute or two late starts to delay one going the other way.
 

Will, the maximum impact ' delay' that can be caused by a single line (excluding a train failing on that section) is directly proportional to the length of the line plus the number of stops on the single section.

At present the maximum delay on the Evesham section is when a train is waiting at Moreton for a late running train to leave Evesham. In that scenario not only is it a long section (13.5 miles) but also the train stops at Honeybourne - add 5 mins (?) including slowing & accelerating.

There is a not dissimilar problem Evesham - Norton Junction - (10.6 miles) with Pershore in between.

If they doubled Pershore - Honeybourne - (10.5 miles) they would have the following advantages:-

The maximum delay would be Honeyboune - Moreton - (8.5 miles) so less that the delays today with no station stops - rather that the current plan which is Evesham - Norton (10.5 miles PLUS the stop at Pershore).

I hope that is clear.

Now that is a total of 10.5 miles, 3.5 miles less than the 13.5 miles Moreton - Evesham that is being done - so the improvements that would be needed at Pershore could be offset by doubling less track!!!

However I would go further - rather than double Ascott - Charlbury - I'd double Charlbury - Hanborough.

If this was done you'd not have Finstock, Coombe &  impotrantly Hanborough on single line. At present they are adding 3.7 miles of double line from Ascott - Charlbury. The distance from Charlbury - Hanborough is 6.2 miles so I'm suggesting an extra 2.5 miles i.e. I've already saved 3.5 miles of doubling Evesham - Honeybourne so use that budget.

So what now is your maximum delay on the eastern end - you've 2 single sections 3.7 miles Charlbury - Ascott (with one station Ascott)
and 3.7 miles Hanborough - Wolvercote - far better that what they are doing - 9.9 miles with 3 stations, 1 of which is frequently used.

Now there would be additional cost for the stations but the advantage is huge.

As you say it's about punctuality and that ensures that delays are kept to a minimum.

You see William, like many things in life, it's not so much about length but how you use it!!

[IndustryInsider]

As promised here is an example of how the Cotswold Line timetable could be improved following the redoubling of track just east of Evesham station through to Moreton-In-Marsh, and also between Ascott and Charlbury.  I MUST STRESS THAT THIS IS IN NO WAY AN OFFICIAL DOCUMENT, JUST MY ATTEMPT TO SHOW WHAT COULD BE DONE.

It is based on the following assumptions and aspirations:
1) It assumes the removal of token working along the line.
2) It makes near maximum use of track capacity, but is fully workable within the single line constraints that would still exist.
3) It is also fully workable in terms of stock diagrams.
4) It assumes the use of HST^▸ (High Speed Train)^s on most services, backed up by some Turbo^s where appropriate, and that Adelante^s are no longer available for use.
5) It attempts to address the following issues, compromising where appropriate:
     a)     An increase in off-peak frequency to 1 train per hour (tph) throughout the majority of the day.
        b)   A more ^clockface^ pattern of trains, so that customers will find the service easier to interpret and plan for.
        c)     An increase in peak services, especially at the western end of the line, to offer a more attractive commuting service between Honeybourne, Evesham and Pershore to Worcester.
        d)   A decrease in journey times where possible:
                1)   Most off-peak trains from London-Worcester are between 5 and 15 minutes quicker than they are now. Most off-peak trains from Worcester-London are between 5 and 10 minutes quicker than they are now.
                2)   Peak trains are also quicker. The two principal Hereford-London trains would be 15 and 7 minutes quicker than at present. The two principal London-Hereford trains would be 15 and 18 minutes quicker than at present.
                3)   Headline timings of 2hrs 3 minutes from London to Worcester (Shrub Hill) and 2hrs 56 from London to Hereford. Still slower than timings from the previous decade, but with additional stops still included.
        e)   Services between London and Hereford are increased from 5 per day to 7 per day and are much better spread to give a roughly 3-hourly direct service.
        f)   Services between London and Great Malvern are increased from 10 to 16 per day.
        g)   Services between London and Worcester are increased from 16 to 20 per day, with all except two running through to Foregate Street.
        h)   Additional trains in the peak hours running between Moreton-In-Marsh and Didcot Parkway which has been discussed as an aspiration for the new service.
        i)   No stations get a reduction in trains. I think ALL stations have an overall increase in the number of train stopping, except for Combe and Finstock which retain the current service on 1 train a day in each direction.

There are limitations in so much as the London Midland service between Worcester and Hereford would need to be altered (although no reduction in frequency would be needed), and also the services down from London to Oxford would usually run at xx:12 rather than xx:22.

Your comments on the timetable are of course welcome!

Click on the links below for details:

Down direction: http://img131.imagevenue.com/img.php?image=95999_Cotswold_Line_-_Down_122_764lo.jpg
Up direction: http://img194.imagevenue.com/img.php?image=96008_Cotswold_Line_-_Up_122_43lo.jpg

[willc]

Andy,

It's not about length, it's about simplicity. How on earth adding another single-line section is supposed to help matters is quite beyond me. This isn't the West Highland Line, with three or four passenger trains a day and the odd freight or two in the very large gaps between passenger trains - it's a busy line now, which is likely to get busier.

Just to repeat for the umpteenth time - Network Rail computer-modelled every conceivable combination they could, multiple loops and all, including what happens if things go wrong - and ended up binning the rest. Why? Because the chosen scheme is simple and operationally straightforward and most likely to produce the reliabilty and punctuality they want.

Single-line; loop; single-line; loop; single-line, loop; single-line - of random lengths (the combination you seem to favour if I'm reading it correctly) isn't a recipe for reliability or fast recovery from problems - lots of switches between double and single track equals lots of potential for conflicts the minute trains aren't in their booked path.

Would you want to be the person in the signal box trying to sort out the morning or evening peak service when something goes wrong, with the sort of track layout you suggest? I wish you luck.

On a more positive note, I think Industry Insider deserves congratulations for his efforts, which make a lot of sense and show what the NR^» (Network Rail - home page) plan could deliver.

I'll forgive him for not trying to weave in Moreton-Didcot shuttles off-peak too, since trying to run them on a regular hourly pattern looks a bit of a challenge. And hopefully Turbo (3) can become new 2+2-seated DMU^▸ (Diesel Multiple Unit) (4).

I'd make the following observations:

Out of London - there really needs to be a through service from London somewhere in the gap between the 15.51 and the 17.22 - preferably a 16.51 - that gap is just too long and puts too much pressure loading-wise on the other two trains. And not entirely sure if it would be worth that 19.25 from Didcot, though if it made a slick connection from a Bristol or Cheltenham train at Didcot it could be handy for anyone who missed the 18.22. In this direction perhaps also a morning westbound train from Shipton (stop by the 07.12?) and a mid-afternoon call (by the 14.10? a long-standing tradition that got axed recently) to give people the chance to get to Worcester and the West Midlands that way and back from Oxford or London earlier. Definitely on the right lines in terms of morning services into Worcester, though that early train towards Worcester should call at Honeybourne since the current timing constraint would be removed. But I would say that the 19.22 should remain the last Hereford train - 23.30 is too late to reach Hereford - an hour earlier feels better and you would be in time for the last buses from the city centre. The 07.12 should stop at Hanborough to allow out-of-Oxford commutes - or get a Moreton shuttle moving out of Oxford as the Cathedrals Express is running in - to give a Hanborough arrival before 9am.

Out of Hereford/Worcester. 04.52? Are there really that many insomniacs in Hereford? But the roughly three-hour journey times are good psychologically/marketing-wise. And again at Shipton I'd try out a call by the 09.25 ex-Malvern to give a direct off-peak journey into London from the Wychwoods, even if there was to be an Oxford/Didcot shuttle calling regularly (Shipton and Milton-under-Wychwood have a rather bigger population than Honeybourne). I would also observe you probably would want to do a bit better for late-afternoon/early-evening Hanborough to Oxford services, even with the looming pressure of the trains going the other way - as was noted by others earlier, people do travel out of Oxford to work at Hanborough and need to get home. And why not run that 17.45 Malvern-Moreton in passenger service to Oxford, even if it is held a while to allow the 18.22 to reach Charlbury? You can probably take a couple of minutes off a non-stop Oxford-Charlbury HST^▸ (High Speed Train) too, as they can manage 12 or 13 minutes without too much trouble given a clear run through Wolvercot junction - padding has slipped in here recently.

[IndustryInsider]

Thanks for the feedback, Will. Your suggested tweaks are by and large very sensible - and I'm sure others could think of more. I'm pleased that the general base of the timetable fits quite well with you though.

As for the 04:52ex Hereford, it's hardly going to be full and standing of course, but it was included as the stock stables at Hereford overnight at the moment, so (as with the 05:20ex Oxford the other way), it was a case of 'why run empty stock when you could have passengers on board? The same applies with the 20:22ex Paddington which ends up forming that return working the next day.

I did try and slot another service in from Moreton to Didcot (stopping at the usual stops and Shipton and Ascott) around 10am which would be useful for shoppers and take a bit of pressure of the following turbo (09:25 ex Gt.Malvern). But in the end the pattern of trains through the Charlbury-Wovercote section meant that is would have been very tight in terms of getting a train through the other way to form it, and then back through between the hourly fast off-peak service. That would also be the case the rest of the day and would impact badly on the 90+PPM^‡(resolve) aspirations that have led to the money being spend in the first place, so I rejected it.

Still, the timetable planners in FGW^▸ (First Great Western) and NR^» (Network Rail - home page) might be able to do some wizardry - I'll be very interested to see what they come up with!

[Hafren]

A small point - if I'm reading everything right - the xx:12 slot might be a bit hard to path without a significant recast on other GWML^▸ (Great Western Main Line) routes! Especially with the stop at Slough, it would hold up the xx:15 Cardiffs. Ideally, as the trains that run on the main lines but stop before Reading, the Oxford/Cotswold trains will depart shortly after a non-stop Reading, and before the next Heathrow Express (as they provide a 'buffer' before the next non-stop HST^▸ (High Speed Train)).

With the 'primary' HSTs leaving at xx:00/15/30/45 (and a few others like Cheltenhams and Plymouth/Penzance a few minutes later)and HEx at xx:10/25/40/55, the ideal slot for Oxford/Cotswold trains is the current xx:22/52ish or something like xx:07/37.

[willc]

RE^▸ (Religious Education) timetable slots, as I understand the present situation, the Oxford/Cotswolds trains are pretty much the first to be written in when the timetable out of Paddington is drawn up, due to the problems with pathing trains through the single-line sections, so don't be surprised if there is a wider shake-up of FGW^▸ (First Great Western) services once that issue is resolved, or at least greatly eased, when the redoubling is complete.

Belated add and getting a bit off topic, but I think it illustrates the kind of tidying-up further afield that may become possible thanks to the redoubling - it is to be hoped that they can make adjustments to Oxford services in general, because at present the fasts from Paddington at 22 and 52 minutes past the hour are followed just five minutes later by the stoppers. By Reading, the following fast train has caught the stopper and they then depart almost simultaneously from Reading. Surely it would be desirable to arrange matters so that someone joining a stopper at Maidenhead and Twyford had the chance to switch to the fast service at Reading?

Same kind of bunching applies to departures from Oxford, with the 01 and 31 minutes past fasts followed by 07 and 37 stoppers and the same issue of departure times at Reading occurs eastbound, so you can't catch a fast from Oxford and join the preceding stopper if you want to go on to Twyford or Maidenhead, as there's only two or three minutes gap between them.

[Richard Fairhurst]

The redoubling will make the whole timetabling situation far more flexible between Worcester and Oxford, so I think near two hours is feasible

I think we're forgetting that the Cathedrals Express did use to run from London to Worcester in 1hr59. If I remember rightly, it was written into Great Western Trains' original service level commitment that there would be at least one such out/back service every day.

The stopping pattern at the time was London-Oxford-Charlbury-Kingham-Moreton-Evesham-Worcester Shrub Hill and onwards. One thing that's changed since then, of course, is the inclusion of a Reading stop, which makes a lot of sense given the town's rise as a commuter destination.

[willc]

The redoubling will make the whole timetabling situation far more flexible between Worcester and Oxford, so I think near two hours is feasible

I think we're forgetting that the Cathedrals Express did use to run from London to Worcester in 1hr59. If I remember rightly, it was written into Great Western Trains' original service level commitment that there would be at least one such out/back service every day.

The stopping pattern at the time was London-Oxford-Charlbury-Kingham-Moreton-Evesham-Worcester Shrub Hill and onwards. One thing that's changed since then, of course, is the inclusion of a Reading stop, which makes a lot of sense given the town's rise as a commuter destination.

No-one's forgetting what used to happen - indeed those from the west of the line wish it were still so. But when that timing applied, the Cotswold HSTs^▸ (High Speed Train) weren't locked into the standard pattern and timings for Oxford-London fast trains, which they are now, and the typical 55-minute time for that leg (as opposed to past timings in the 45 to 50-minute range) and the current constraints and padding on the Cotswold Line mean getting near to two hours is quite an ask. Industry Insider's efforts show that the redoubling will speed up operations west of Oxford and bring that mark back in range.

New dates for the redoubling roadshows/meet the manager are as follows:

Evesham - Monday 30th March 1830-2000hrs
Kingham - Tuesday 31st March 0700-0830hrs
Moreton - Tuesday 31st March 1900-2030hrs
Charlbury - Wednesday 1 April 0700-0830hrs

[James_H]

The whole situation with morning timings from Oxford to Hanborough is just plain daft.

What they obviously need to do is to get the Cotswold line train that leaves Oxford around 08:00 onto the single line earlier, which would obviously involve getting the train that stops off at every single station off the single line earlier.

Why can't they just delete a few stops from this service and get it off the line.  Does Hanborough really need a train into Oxford at 07:34 and another one at 08:01?  Why can't they just send the second of these through Hanborough non-stop and get the northbound train onto the line earlier?

Alternatively, what is the point of Combe and Finstock stations?  I often ride the 06:53 from Oxford as far as Charlbury, go to the Co-op and then get the 07:48 (I think) back to Hanborough - more often than not, no-one gets on at either station (and it's never once been more than a total of three people at both stations combined) - so in all honesty, what's the point in having them?  There are a lot more than 3 people who want to go from Oxford to Hanborough at that time in the morning...!

Alternatively, maybe they could just bring the whole service forward by 5 minutes or so (although I guess this may well be impossible due to the single track sections)... but it's really hard to believe they honestly can't do better than a 2 and a half hour gap.

Rant over...

[IndustryInsider]

Alternatively, what is the point of Combe and Finstock stations?  I often ride the 06:53 from Oxford as far as Charlbury, go to the Co-op and then get the 07:48 (I think) back to Hanborough - more often than not, no-one gets on at either station (and it's never once been more than a total of three people at both stations combined) - so in all honesty, what's the point in having them?  There are a lot more than 3 people who want to go from Oxford to Hanborough at that time in the morning...!

Well, the adverse publicity (and the sheer effort) that would have to go in to close them will probably prevent it from ever happening. The only chance would be if the remaining stretch of the line is redoubled and FGW^▸ (First Great Western) puts a case towards the Government that it would be financially prohibitive to provide the additional platforms that would be needed.

Passenger number are very low- I would say an average of 5 per-day between the two stations. I've never seen more that 10. But, I doubt that there's many more than a handful who would want to make the Oxford to Hanborough journey - certainly not enough to use that as an excuse to close Combe and Finstock.

Certain forum members, such as Lee Fletcher, oppose any attempt to close a station (unless an improved facility is provided literally right by it) and I admire his stance, though in the case of Combe and Finstock they ARE clearly a burden on FGW financially and operationally AND are very close to alternative stations -

* Finstock is under 3 miles away by road (B4022) from Charlbury station for both Finstock and Fawler villages that it serves.

* Combe village is 2.5 miles away from Hanborough station via road, but Combe station is 1 mile away from Combe village anyway, so it's only an additional 1.5 miles.

In either case, there certainly would be no case whatsoever for a new station, but as I said I doubt either will be closing any time soon!

[mjones]

[...

Certain forum members, such as Lee Fletcher, oppose any attempt to close a station (unless an improved facility is provided literally right by it) and I admire his stance, though in the case of Combe and Finstock they ARE clearly a burden on FGW^▸ (First Great Western) financially and operationally AND are very close to alternative stations -

...

I'd be interested to know what the true costs and impacts on the timetable of keeping stations at Appleford and Culham; particularly the former as it is so close to Didcot. e.g if Appleford and Culham were closed, would enough additional capacity be available to provide more stops at Radley, for example, or to stop more Reading-Oxford fast services (e.g XC^▸ (Cross Country Trains (franchise))) at Didcot as well?