Rail replacement - will GBR enable a greater use of diversionary routes to avoid it?

Bletchleyite · 7 Sep 2026

tutacre said:
There isn't a push in the way described - there is an understanding though.

The understanding/ general consensus is that passengers prefer to complete their journey by rail, even if it means a lengthy diversion, as per GWR into Euston (and later, via Milton Keynes) and Avanti via Settle. However, the main examples are FirstGroup in the private sector. I'm not sure that the nationalised operators have actually done anything genuinely innovative yet when it comes to diversions - not to say that they won't of course. I suppose you could reasonably make the case for LNER to Bedford, but extremely limited track record, so far, and it was very much a trial/test to see whether it was viable.

I think it's potentially contractually easier post nationalisation. The initial set of DfTO members maybe wasn't that useful in that sense - you aren't going to get many cases where Southeastern drivers could conduct LNER ones as there's no even vague overlap in their operations, for instance, and I would imagine there is significant overlap in routes signed between TPE and Northern drivers anyway - but if it's all under one roof the idea of sending an 810 down through Bletchley to Euston (say) starts sounding a lot more plausible of a thing to do. (I doubt they are cleared now, but the near complete rebuild that bringing that route into EWR will result in could have that taken into account).

What do others think there?

RailUK Forums

Harpo · 7 Sep 2026

GWR have led the way on conducted diversions and perhaps this is the way forward for others?

I don’t see GBR changing much operationally and I especially doubt that diversionary route knowledge will increase as it would add back a lot of cost that’s been eliminated previously despite it creating operational handicaps.

Bletchleyite · 7 Sep 2026

Harpo said:
I don’t see GBR changing much operationally and I especially doubt that diversionary route knowledge will increase as it would add back a lot of cost that’s been eliminated previously despite it creating operational handicaps.

I'm not suggesting that MML drivers would end up signing the Marston Vale, but if it's all GBR (even if it's in divisions of some sort) arranging route conductors is probably easier than it being entirely separate TOCs.

If the MV was wired that use case might be better with WMT drivers in 10 car 730s doing hourly Bedford-London shuttles, but of course it's not and not planned to be even with EWR.

Tazi Hupefi · 7 Sep 2026

Route (and in some cases, traction) knowledge is still being lost / withdrawn from depots, or concentrated at other depots, nationally as things lapse even for what could be classified as primary routes, let alone diversions - with only small increases in the other direction. I think Transport for Wales within England is probably the most problematic for this if you were ranking TOCs - they really should sign Middlewich, Bescot/Aston, Aston/Stechford, Soho/Bescot, Altrincham etc, I think they also lost Stafford, Stoke and possibly Winsford now. At one point they also signed the Oxley curve too.

Some TOCs still also have insane policies where even if you sign the route, you cannot route conduct someone else unless you also sign the traction (!) which defeats the point of a route conductor - they may as well drive it.

Harpers Tate · 7 Sep 2026

Standardisation (within context) of traction would help. As in (historically) if you could drive a DMU, you could drive a DMU.

Bletchleyite · 7 Sep 2026

Harpers Tate said:
Standardisation (within context) of traction would help. As in (historically) if you could drive a DMU, you could drive a DMU.

The thing that would probably simplify it most would be moving away from route knowledge onto a system more like Germany, where they have a ring bound book (or iPad) on the desk which tells them the information British drivers are required to learn like track layouts and speed limits - the Buchfahrplan (or elektronische Buchfahrplan). I can't see that happening here though, it's a big cultural shift.

zwk500 · 7 Sep 2026

Bletchleyite said:
The thing that would probably simplify it most would be moving away from route knowledge onto a system more like Germany, where they have a ring bound book (or iPad) on the desk which tells them the information British drivers are required to learn like track layouts and speed limits - the Buchfahrplan (or elektronische Buchfahrplan). I can't see that happening here though, it's a big cultural shift.

That won't (probably can't) happen until we get a Speed-signalling or ATP system, which for now means ETCS. And route knowledge covers so much more than just speeds, so I'd be surprised if we got to a stage where relying on the Sectional Appendix alone would be acceptable except for emergencies like trapped/stranded trains.

One big thing against an 'alternative London terminus' level of diversion is that most of the railway is now full or very nearly full. So in most places you'd need to start pulling trains out of the timetable to give diverted trains space to get into, which either means backing up the freight until the job is sorted or backing up passengers onto fewer trains, which will double up with the diverted passengers.

I think GBR will make planned diversions easier to manage - and potentially the LNER Bedford trains may be an early sighter of that. It may also make it easier to keep diversionary knowledge available, and manage mid-route diversions where the train ultimately does get back to booked route, but I'd be surprised if there was a meaningful improvement in ad-hoc diversions to alternate termini.
What might be easier to arrange is ad-hoc strengthening or more relief extras on a connecting branch but not running through - Marston Vale would be a classic option if there was a problem on either the WCML or MML. Although again, at some point there'll be Bedford-Oxford services and the argument will be just to change twice and use existing services.

Bletchleyite · 7 Sep 2026

zwk500 said:
One big thing against an 'alternative London terminus' level of diversion is that most of the railway is now full or very nearly full. So in most places you'd need to start pulling trains out of the timetable to give diverted trains space to get into, which either means backing up the freight until the job is sorted or backing up passengers onto fewer trains, which will double up with the diverted passengers.

There is that, but equally it does seem possible to find space. The GWR services into Euston being an example - despite having lost P17 and P18 to HS2 development the station isn't really full. But by contrast running WCML services into St Pancras wouldn't really be workable (though now so much has moved onto the Liz, running them into Paddington actually might be, at least a couple an hour, and at least in theory on a Sunday it shouldn't be entirely infeasible to run via Kensington Olympia into Victoria though again only a couple an hour taking the Southern paths).

zwk500 said:
What might be easier to arrange is ad-hoc strengthening or more relief extras on a connecting branch but not running through - Marston Vale would be a classic option if there was a problem on either the WCML or MML. Although again, at some point there'll be Bedford-Oxford services and the argument will be just to change twice and use existing services.

Though at least in the latter sense the ticketing aspect is easy - GBR can just say "use them then" with no need to faff around agreeing acceptance and what's going to need to be paid for that. No doubt there would be internal accounting as with BR, but that's a theoretical exercise and not one that should get in the way of actually doing it.

Of course the MV doesn't work well for diverted passengers at present due to its low capacity of 2 coaches an hour, so it isn't usually promoted (though there is sometimes acceptance that way in disruption). The same will probably be true of EWR initially when that's just a 2 car. Full EWR with five an hour and dedicated stock (possibly designed for standees due to the high demand from the new theme park) would be quite different.

For planned works, that said, advertised acceptance is probably less important now than it was. Most tickets are sold via a site (Trainline) that does splitting, so when you get people being sent by planners from the MML across via Tamworth (as I once saw - hundreds joining WMT services there) they probably just have split Advances for the trains they are actually using, not an off peak return from the MML that has acceptance that way. Indeed in part their choice of diversionary route may have been influenced heavily by price in a way it wasn't in the past.

zwk500 · 7 Sep 2026

Bletchleyite said:
There is that, but equally it does seem possible to find space. The GWR services into Euston being an example - despite having lost P17 and P18 to HS2 development the station isn't really full.

Platforms at Euston aren't the only constraint on the WCML. The GWR diversions are reversing at MK across both Platform 2 and 5, so there's going to be lost of crossing between fast and slow. Somebody's found a plan that notionally works, but then again presumably the same thing was thought about Finsbury Park all those years ago for Christmas...

Bletchleyite said:
But by contrast running WCML services into St Pancras wouldn't really be workable (though now so much has moved onto the Liz, running them into Paddington actually might be, at least a couple an hour, and at least in theory on a Sunday it shouldn't be entirely infeasible to run via Kensington Olympia into Victoria though again only a couple an hour taking the Southern paths).

The problem is that all those trains that have moved from Paddington platforms to the CCOS are still running on the Relief Lines. So you are either needing to fit between them or to cross to the mains somewhere along the route. You might get the odd train an hour to do it, but then are you just giving yourself more trouble than, e.g. if you amended the timetable to run 165s on a Reading-Oxford all stop and operated a MK-Didcot 5-car DMU more often?

Bletchleyite said:
Though at least in the latter sense the ticketing aspect is easy - GBR can just say "use them then" with no need to faff around agreeing acceptance and what's going to need to be paid for that. No doubt there would be internal accounting as with BR, but that's a theoretical exercise and not one that should get in the way of actually doing it.

Yes, ticket acceptance is an area where things will hopefully get better - although there's always the risk that the same battles just go internal as Cost Centre KPIs rather than different companies. See the various BR issues where e.g. Locos were modified to remove compatibility with passenger stock to stop them being nicked off a lower priority sector.

Bletchleyite · 7 Sep 2026

zwk500 said:
Yes, ticket acceptance is an area where things will hopefully get better - although there's always the risk that the same battles just go internal as Cost Centre KPIs rather than different companies. See the various BR issues where e.g. Locos were modified to remove compatibility with passenger stock to stop them being nicked off a lower priority sector.

Good senior management is needed to minimise this. You get that sort of insularity and empire building in all large companies, but the extent of it varies depending on how good the senior management are at making people feel an important part of the whole (and getting rid of managers who don't act that way), as if this fails people just end up running their own little empire as they curse the people at the top and this sort of thing is the usual result.

zwk500 · 7 Sep 2026

Bletchleyite said:
Good senior management is needed to minimise this.

Well then thank god we'll have the DfT keeping a close watch. o_O

The Planner · 7 Sep 2026

It will end up as route conducting, the overall cost is probably less.

== Doublepost prevention - post automatically merged: 7 Sep 2026 ==

zwk500 said:
Platforms at Euston aren't the only constraint on the WCML. The GWR diversions are reversing at MK across both Platform 2 and 5, so there's going to be lost of crossing between fast and slow. Somebody's found a plan that notionally works, but then again presumably the same thing was thought about Finsbury Park all those years ago for Christmas...

Thats only 1tph, and a slot exists to do that in 5. P2 makes little difference.

HSTEd · 7 Sep 2026

If you could replace all extant DMU vehicles (~2,868) with a giant order of technically identical trainsets (multiple unit compatible and the same driving setup), then I think you could probably manage a substantial reduction in rail replacement by allowing more efficient movement of stock between routes during disruption. This does seem unlikely for economic reasons though!

If all trainsets come under the same driver qualification you could heavily reinforce whatever line was not closed.

For an example of what that could look like, see the old 15-coach Voyager diversions off the WCML.
In cases where platform lengths and such prevent such reinforcement, you could maintain some units with Class 378 style interiors which would allow many more people to be carried in disruption.

Sure, it might not be great having to stand on the train, but it will get you home much quicker than waiting for a bus in most cases.

rmHawk765 · 7 Sep 2026

Bletchleyite said:
The thing that would probably simplify it most would be moving away from route knowledge onto a system more like Germany, where they have a ring bound book (or iPad) on the desk which tells them the information British drivers are required to learn like track layouts and speed limits - the Buchfahrplan (or elektronische Buchfahrplan). I can't see that happening here though, it's a big cultural shift.

Crucially this does not remove the requirement for route learning (source: a friend working at DB, though admittedly this may not apply to all railway operations in Germany). It only allows the driver to proceed at a max speed of 100km/h, and it is worth noting that signage on many German routes is a lot more coherent than in the UK. They also use speed signalling so a lot of the knowledge requirement simply doesn't exist.

However we will have to see what changes in regards to route knowledge ETCS Level 2 cab signalling brings as it is rolled out onto more and more lines. Any complete cab signalling system makes remembering speed limits among other operational considerations redundant.

Hadders · 7 Sep 2026

I’m no traction expert but standardised traction comes up regularly on these forums, but I’m not convinced it’s thought about properly.

Met’s assume rolling stock lasts 40 years, you’re not going to buy 40 years worth of trains in one go and replace them all again in 40 years time. Trains will always get replaced in batches over a period of time.

So what train do you order in 20 years time? Technology moves on, how is a train built today going to be compatible with trains built in 10, 20 or 30 years time?

This didn’t happen in the past and I simply don’t see it happening in the future. You’re never going to get everything compatible with everything else.

Perhaps some of our traction experts can advise otherwise.

Harpers Tate · 7 Sep 2026

Bletchleyite said:
The thing that would probably simplify it most would be moving away from route knowledge .....

Agreed. No doubt I'll be shot down for even mentioning this, but I wonder whether a method of working (for use in emergency/contingency/diversion scenarios only) which might be termed "extreme caution line-of -sight" could be implemented. With a throughout absolute top speed over the "alien" part of route of, say, 20mph and lower speeds determined by the driver such that s/he can to stop the train before encountering (for example) a junction or crossing if sight deems it neccessary. So - just as a road vehicle driver should be doing (adjusted for the very different braking characteristics of a train). I'm thinking about what happens sometimes when trespassers and/or animals are observed on the line, and a driver is instructed to proceed at extreme caution; or when one train is being "rescued" by another. Extending, if you like, the scope of uses to which that methodology is currently put.

HSTEd · 7 Sep 2026

Hadders said:
Met’s assume rolling stock lasts 40 years, you’re not going to buy 40 years worth of trains in one go and replace them all again in 40 years time. Trains will always get replaced in batches over a period of time.

I don't see any particularly reason we couldn't do that if we chose to.
Other railways abroad have conducted total or near total traction replacements.

The ~16,000 passenger rail vehicles in the UK do not represent a large portion of the total global fleet of rail vehicles. It would likely be possible to place a big order with the likes of Nippon Sharyu or CRRC for a train fleet if that was desirable. (EDIT: N700 Shinkansen production, excluding the later N700S derivatives totals 3,000 vehicles alone!)

However I do believe we would be better just to order 500 vehicles to a standard design every year, and maybe change the design every decade or so.

Hadders said:
So what train do you order in 20 years time? Technology moves on, how is a train built today going to be compatible with trains built in 10, 20 or 30 years time?

The Association of American Railroads has maintained multiple unit control equipment compatibility between locomotives built circa 1950 (with limited modifications, circa 1940) and those built today.
The vast majority of US railroad locomotives are fully interchangeable in control gear terms.

Various BR multiple working systems allowed compatibility between units built a great many years apart.

I think the current zoo of incompatible trains in the UK is more a function of manufacturers wanting to secure lock-in than any real technical need.

Harpers Tate · 7 Sep 2026

Hadders said:
I’m no traction expert but standardised traction comes up regularly on these forums, but I’m not convinced it’s thought about properly.

Met’s assume rolling stock lasts 40 years, you’re not going to buy 40 years worth of trains in one go and replace them all again in 40 years time. Trains will always get replaced in batches over a period of time.

So what train do you order in 20 years time? Technology moves on, how is a train built today going to be compatible with trains built in 10, 20 or 30 years time?

This didn’t happen in the past and I simply don’t see it happening in the future. You’re never going to get everything compatible with everything else.

Perhaps some of our traction experts can advise otherwise.

My non-expert way to go about this would be to start NOW with a standard. It may be something already in place on some part of the (newer) fleet. Everything ordered from here on out conforms. A new standard is only established when there is valid reason to update; not just because we can and not just because it's old. It needs to be "broken" before it's fixed. That may be 20 years, I don't know. At that point, you start over with a fresh standard which then applies for the next (20?) years. And so on. At any one time, starting after the last of the old oddments have gone, there would thereby be only perhaps two standards in play, and crews should, I think, be able to "sign" both.

JamesT · 7 Sep 2026

Hadders said:
I’m no traction expert but standardised traction comes up regularly on these forums, but I’m not convinced it’s thought about properly.

Met’s assume rolling stock lasts 40 years, you’re not going to buy 40 years worth of trains in one go and replace them all again in 40 years time. Trains will always get replaced in batches over a period of time.

So what train do you order in 20 years time? Technology moves on, how is a train built today going to be compatible with trains built in 10, 20 or 30 years time?

This didn’t happen in the past and I simply don’t see it happening in the future. You’re never going to get everything compatible with everything else.

Perhaps some of our traction experts can advise otherwise.

I suspect many people will be thinking of how classes 14X, 15X, and 170 can interoperate despite there being 20ish years between the first and last of those. It may be that only worked as those were relatively simple units, but it doesn’t seem impossible to design a suitable extensible standard and require all new rolling stock to meet it. (Cue XKCD 927).

eldomtom2 · 7 Sep 2026

Theoretically I suppose with today's modern computer-equipped trains all that would need to be transmitted between trains would be a target acceleration/deceleration rate, with the train's computers then calculating what level of throttle/brake to apply to match it.

The Planner · 7 Sep 2026

Harpers Tate said:
Agreed. No doubt I'll be shot down for even mentioning this, but I wonder whether a method of working (for use in emergency/contingency/diversion scenarios only) which might be termed "extreme caution line-of -sight" could be implemented. With a throughout absolute top speed over the "alien" part of route of, say, 20mph and lower speeds determined by the driver such that s/he can to stop the train before encountering (for example) a junction or crossing if sight deems it neccessary. So - just as a road vehicle driver should be doing (adjusted for the very different braking characteristics of a train). I'm thinking about what happens sometimes when trespassers and/or animals are observed on the line, and a driver is instructed to proceed at extreme caution; or when one train is being "rescued" by another. Extending, if you like, the scope of uses to which that methodology is currently put.

That wouldn't work, for example if the diversion was 20 miles and the driver and TM didn't sign it, that is an hour which just holds up trains that use the diversionary route in normal circumstances.

Hadders · 7 Sep 2026

JamesT said:
I suspect many people will be thinking of how classes 14X, 15X, and 170 can interoperate despite there being 20ish years between the first and last of those. It may be that only worked as those were relatively simple units, but it doesn’t seem impossible to design a suitable extensible standard and require all new rolling stock to meet it. (Cue XKCD 927).

Ok, so to use an extreme example would you have made a Pendolino compatible with a 14X

HSTEd · 7 Sep 2026

Hadders said:
Ok, so to use an extreme example would you have made a Pendolino compatible with a 14X

Well I can certainly make a case for coupling a Pendolino to a Class 350, a case for a Class 350 to a Sprinter and a case for a Sprinter to a Pacer.

The easiest way to do all of those things is to make as much as possible of the fleet (preferably all) control and coupling compatible.

By the end of the BR era the railway was rapidly trending towards a small number of multiple unit standards:
1. Time Division Multiplexing (TDM) on loco hauled stock
2. The PEP-through-Networker Electric Multiple unit system,.
3. The Pacer/Sprinter/Networker Turbo system

There were some holdouts, most notably in freight and the Southern Region's legacy Mark 1-derived specification, but the overall direction of travel was pretty clear.

GBR is an opportunity to move in this direction again, which would enable better utilisation of rolling stock and routings during diversions.

Bald Rick · 7 Sep 2026

zwk500 said:
One big thing against an 'alternative London terminus' level of diversion is that most of the railway is now full or very nearly full. So in most places you'd need to start pulling trains out of the timetable to give diverted trains space to get into, which either means backing up the freight until the job is sorted or backing up passengers onto fewer trains, which will double up with the diverted passengers.

Bletchleyite said:
The GWR services into Euston being an example - despite having lost P17 and P18 to HS2 development the station isn't really full.

GWR are fitting into Euston over Christmas because New St is shut and there are fewer services from the West Mids into Euston.

17 and 18 were barely used anyway.

mmh · 7 Sep 2026

Hadders said:
I’m no traction expert but standardised traction comes up regularly on these forums, but I’m not convinced it’s thought about properly.

Met’s assume rolling stock lasts 40 years, you’re not going to buy 40 years worth of trains in one go and replace them all again in 40 years time. Trains will always get replaced in batches over a period of time.

So what train do you order in 20 years time? Technology moves on, how is a train built today going to be compatible with trains built in 10, 20 or 30 years time?

This didn’t happen in the past and I simply don’t see it happening in the future. You’re never going to get everything compatible with everything else.

Perhaps some of our traction experts can advise otherwise.

The topic here wasn't whether different trains should be compatible with each other, but whether different trains should be able to be driven without drivers needing specific training on them.

Falcon1200 · 8 Sep 2026

Harpers Tate said:
No doubt I'll be shot down for even mentioning this, but I wonder whether a method of working (for use in emergency/contingency/diversion scenarios only) which might be termed "extreme caution line-of -sight" could be implemented. With a throughout absolute top speed over the "alien" part of route of, say, 20mph and lower speeds determined by the driver such that s/he can to stop the train before encountering (for example) a junction or crossing if sight deems it neccessary.

The Planner said:
That wouldn't work, for example if the diversion was 20 miles and the driver and TM didn't sign it, that is an hour which just holds up trains that use the diversionary route in normal circumstances.

IMHO @Harpers Tate's suggestion does have merit, but only in a limited number of circumstances: The diversion should be specified in the Sectional Appendix, of a maximum length, the route should be set throughout from start to end, and the move should be made at a speed of 10mph, or even 5, depending on circumstances. Locations where this would be helpful, from my experience in Scotland, are at Newton (Lanarkshire), where not all Drivers sign the route through the station platforms, and between Law Junction and Shieldmuir Junction via Wishaw. As @The Planner says however a slow speed move over any great distance would cause its own issues!

Far better of course is Traincrew trained for reasonable diversionary routes in the first place, with sufficient trains booked over the routes to ensure that knowledge is retained. This does happen, for example in Scotland;
1S37 0531 Euston-Glasgow Central is booked via Wishaw and Bellshill
1M73 0745 Glasgow Central-Liverpool Lime St is booked via Carmyle, Mossend Yard and Wishaw

tutacre said:
Some TOCs still also have insane policies where even if you sign the route, you cannot route conduct someone else unless you also sign the traction (!) which defeats the point of a route conductor - they may as well drive it.

That seems over the top, and without that condition standardisation of rolling stock is simply not necessary!

The Planner · 8 Sep 2026

Bald Rick said:
GWR are fitting into Euston over Christmas because New St is shut and there are fewer services from the West Mids into Euston.

17 and 18 were barely used anyway.

Only 1 less.

Adrian1980uk · 19 Sep 2026

I think diversionary routes will become more used, not read it yet but there's an article in rail on this exact subject, but as trains get fuller at weekends buses become less and less practical to move sufficient numbers of people.
Full 10 to 12 coach trains taking 700 passengers, probably 10 to 12 buses and if that's every 30 mins then it becomes impossible to source that quantity.

Challenge as mentioned above, most tracks are full to capacity so little opportunity.

The Planner · 20 Sep 2026

Adrian1980uk said:
I think diversionary routes will become more used, not read it yet but there's an article in rail on this exact subject, but as trains get fuller at weekends buses become less and less practical to move sufficient numbers of people.
Full 10 to 12 coach trains taking 700 passengers, probably 10 to 12 buses and if that's every 30 mins then it becomes impossible to source that quantity.

Challenge as mentioned above, most tracks are full to capacity so little opportunity.

They will get used, but not with significant frequency and I very much doubt they will actually be learnt. The use of pilots will be significant. The cost will be interesting as depending on length under the current regime a TOC can claim that money off NR, under GBR thats money that may have to be just absorbed.

JohnofWessex · 20 Sep 2026

What about SWR the Yeovil - Salisbury via Castle Cary service that as far as I understand it is partly to do with maintaining and retaining competencies in case diversions are needed

Rail replacement - will GBR enable a greater use of diversionary routes to avoid it?

Veteran Member

RailUK Forums

Established Member

Veteran Member

On Moderation

Established Member

Veteran Member

Veteran Member

Veteran Member

Veteran Member

Veteran Member

Veteran Member

Veteran Member

Veteran Member

Member

Veteran Member

Established Member

Veteran Member

Established Member

Established Member

On Moderation

Veteran Member

Veteran Member

Veteran Member

Veteran Member

Established Member

Established Member

Veteran Member

Member

Veteran Member

Member