HS2 could cost up to £102.7bn and trains won't start until at least 2036

[Sorcerer]

The natural corollary of that is that the fares for other means of travel between Birmingham need to be increased as well. Arguably, the best outcome for DfT is HS2 providing the cheapest fare, but that cheapest fare not being cheap. (We also have to remember that this isn't until 2037 by which time all sorts of changes to fare policy could have been applied.)

I'd be more inclined to gauge the opposite will happen where HS2 will provide the faster but more expensive London to Birmingham/Manchester traffic while the WCML will have a few select slower but cheaper services for the more budget-conscious traveller. I think the sparks effect will be very real with HS2 even in it's scaled back form simply due to the uplift in capacity, and I think this will just highlight how short-sighted it was to scaled it back to the extent it has been.

 

[HSTEd]

I'd be more inclined to gauge the opposite will happen where HS2 will provide the faster but more expensive London to Birmingham/Manchester traffic while the WCML will have a few select slower but cheaper services for the more budget-conscious traveller. I think the sparks effect will be very real with HS2 even in it's scaled back form simply due to the uplift in capacity, and I think this will just highlight how short-sighted it was to scaled it back to the extent it has been.

The problem here is that slower but cheaper services will probably cost more than HS2 trains to operate, and will probably lose money hand over fist.

Market segmentation has serious downsides that, in my view, have long been hidden by the byzantine structure of the industry.

Every vehicle used on a classic service will haul fewer people for less kilometres than one on HS2, and whilst slow EMU carriages might be somewhat cheaper to operate, they won't be able to overcome HS2's enormous productivity advantage. A single HS2 crew member with ~16 vehicles can move (I believe) over a thousand seats on a London Birmingham round trip in slightly under two hours!

I think the financially optimum outcome is likely a classic WCML and Chiltern Main Line timetable that meets intermediate station travel demands with the minimum number of vehicles and crew, and everyone who can use HS2 does so.

 

[Bletchleyite]

The problem here is that slower but cheaper services will probably cost more than HS2 trains to operate, and will probably lose money hand over fist.

That carries the assumption that those services would be like Lumo, GoVolta, Flixtrain etc, i.e. dedicated to that purpose. They won't. They'll be the former WMT services which exist primarily to serve intermediate markets, but shoving an extra few coaches on them to provide market segmentation style fare offerings comes at a very low extra cost.

Market segmentation has serious downsides that, in my view, have long been hidden by the byzantine structure of the industry.

Every vehicle used on a classic service will haul fewer people for less kilometres than one on HS2, and whilst slow EMU carriages might be somewhat cheaper to operate, they won't be able to overcome HS2's enormous productivity advantage.

You're missing the point. There would be a half hourly Euston-Brum semifast service whether people were priced onto it or not, because you need to serve the intermediate markets. Making it a bit longer comes at a very low cost overall.

The financially optimum outcome is likely a classic WCML and Chiltern Main Line timetable that meets intermediate station travel demands with the minimum number of vehicles and crew, and everyone who can use HS2 does so.

The financially optimum argument ignores that if travel by rail becomes too expensive people will switch to coach and car. That cannot be ignored, or rather is ignored at the railway's peril.

With London-Manchester the railway has almost succeeded at that (there are coaches, but only really a Pendolino's worth per day) because it has such a time advantage - two hours as against five (though the WMT through Liverpool to London service did get decent custom despite taking five hours - that was killed simply because it wrecked the timetable by propagating delays across New St, not because the actual idea was flawed). But Birmingham is much closer to London and a coach journey isn't prohibitively slow, nor is driving to a Tube station or somewhere like Luton Airport Parkway.

 

[HSTEd]

You're missing the point. There would be a half hourly Euston-Brum semifast service whether people were priced onto it or not, because you need to serve the intermediate markets. Making it a bit longer comes at a very low cost overall.

Still a substantially greater cost per unit capacity than making the HS2 services longer though!
HS2 has DOO, much higher average speeds and no meaningful limit on track capacity.

Even if we set all capacity costs to marginal and ignore crewing, a HS2 vehicle will carry many times more people than a vehicle on the semi fast ever can.

 

[Sorcerer]

The problem here is that slower but cheaper services will probably cost more than HS2 trains to operate, and will probably lose money hand over fist.

Market segmentation has serious downsides that, in my view, have long been hidden by the byzantine structure of the industry.

Every vehicle used on a classic service will haul fewer people for less kilometres than one on HS2, and whilst slow EMU carriages might be somewhat cheaper to operate, they won't be able to overcome HS2's enormous productivity advantage. A single HS2 crew member with ~16 vehicles can move (I believe) over a thousand seats on a London Birmingham round trip in slightly under two hours!

I think the financially optimum outcome is likely a classic WCML and Chiltern Main Line timetable that meets intermediate station travel demands with the minimum number of vehicles and crew, and everyone who can use HS2 does so.

I wouldn't have expected a lot of options, mostly just one or two as was originally planned. I would've guessed it to be similar to Paris-Brussels services with Eurostar Red providing high-speed services while OuiGo does slower but cheaper trains on the classic line serving more intermediates. If running a London-Manchester service on the classic line I would expect a semi-fast pattern similar to the current London Northwestern route to Crewe.

HS2 has DOO, much higher average speeds and no meaningful limit on track capacity.

Now that you mentioned Driver Only Operation, I'm lead to believe the project even when finished will face the same delays as East West Rail for that very reason. Not a subject for this thread but just something being mentioned for posterity when the time comes.

 

[Bletchleyite]

Still a substantially greater cost per unit capacity than making the HS2 services longer though!

Is that actually so? HS2 vehicles are much costlier and are only available (as things stand) in 200m lengths. WMT can do 80m, 120m, 160m or 240m. This means the added capacity required may actually be cheaper overall. It might even be free, as keeping 8 car sets formed all day is cheaper than splitting and joining so they are generally left as 8s even when it's quiet.

DOO is irrelevant as lengthening a WMT service doesn't add any more staff.

 

[Palmerston]

Particularly given that Chiltern is wedded to 5 car trains for the foreseeable.

I think Chiltern will become more of a commuter operation post HS2 with a higher frequency of trains but more stops in them all. I wouldn't even rule out the idea that it could get lopped back to Banbury as a pure play commuter operation with an electric West Midlands service closing the gap at some point - the classic WCML will provide the budget option.

Direct Chiltern services could still be preferable for people as far as Warwick and Leamington. Solihull traffic to Marylebone likely to reduce, as it's very easy to get from there to Moor Street/Curzon Street.

 

[Bletchleyite]

Direct Chiltern services could still be preferable for people as far as Warwick and Leamington. Solihull traffic to Marylebone likely to reduce, as it's very easy to get from there to Moor Street/Curzon Street.

I don't to be fair think lopping at Banbury is that likely, other than temporarily if electrification starts. I do think it's likely Chiltern services will be slowed down a bit as they will lose their Birmingham-London focus, and probably take on a more consistent stopping pattern, perhaps something like 4tph Marylebone-High Wycombe all stations bar the innersuburbans (or maybe one of the 4 doing each innersuburban station) and 4tph Banbury/Moor St fast to Wycombe then all stations to Banbury, Leamington, Solihull and Moor St (other stations being served by WMT at that end). That should be possible to flight out, particularly if you wired to Wycombe and so sped that bit up slightly.

 

[JonathanH]

Still a substantially greater cost per unit capacity than making the HS2 services longer though!
HS2 has DOO, much higher average speeds and no meaningful limit on track capacity.

You are forgetting that well over £100bn will have been spent on the line and that higher revenue to the exchequer means a larger contribution towards the UK's crippling debt burden. There will be an optimum fare to maximise revenue which isn't directly related to which service is the cheapest to provide.

Would HS2 be the cheapest route to operate if it alone had to service the costs of its construction?

I also note that when HS1 opened fares across the Southeastern network (ie including the existing routes) increased faster than elsewhere to reflect the investment.

 

[Watershed]

I'd argue the better example of revenue maximisation is the Elizabeth line, where opening the core produced a revenue increase of something like £500m based essentially on massive volume.
The Elizabeth line is now something like 5.8% of the entire railway farebox.

I think its also far from clear if the Avanti revenue maximisation strategy actually minimises net cost of the railway to the taxpayer, it makes the most out of the supply of seats on Avanti, sure. But it means peak trains haul air whilst the government is forced to spend more money on LNWR operations to carry the people priced off.
Making £1 net revenue each from 100 people is much better than making £10 each from 5 people.

It is in the railway industry's (and thus the government's) financial interest that the only people that use the classic line are people who physically cannot be carried on HS2. It's probably in its interest to the point of encourage doubling back on Chiltern for places like Solihull etc that loadings on Chiltern line trains are flattened as much as possible.

When you look at the railway as a means of transporting people, absolutely that is the sensible approach. When you look at the railway as a subsidy sinkhole to be minimised - which is clearly how the government views it - I cannot see them adopting such a holistic approach.

In the government's eyes, HS2 is a long-distance journey for intercity passengers and therefore needs to be revenue maximised with compulsory reservations no doubt. The fact that a journey from Curzon Street to OOC takes no longer than a journey from Slough to Tottenham Court Road seems to completely pass them by.

 

[HSTEd]

You are forgetting that well over £100bn will have been spent on the line and that higher revenue to the exchequer means a larger contribution towards the UK's crippling debt burden. There will be an optimum fare to maximise revenue which isn't directly related to which service is the cheapest to provide.
Would HS2 be the cheapest route to operate if it alone had to service the costs of its construction?

Probably not, but unless you have a way to dismantle HS2 and recover the full costs of its construction, the servicing costs on this debt are to a large degree irrelevant.
That money is gone, the optimum strategy is the one that minimises the net cost to the taxpayer of the railway industry as we are today (or more correctly, at the opening of HS2).

The objective should not be to maximise net revenue from any particular subset of operations, but to do so for the industry as a whole (or at least render net revenue less negative).

== Doublepost prevention - post automatically merged: 23 Sep 2026 ==

When you look at the railway as a means of transporting people, absolutely that is the sensible approach. When you look at the railway as a subsidy sinkhole to be minimised - which is clearly how the government views it - I cannot see them adopting such a holistic approach.

In the government's eyes, HS2 is a long-distance journey for intercity passengers and therefore needs to be revenue maximised with compulsory reservations no doubt. The fact that a journey from Curzon Street to OOC takes no longer than a journey from Slough to Tottenham Court Road seems to completely pass them by.

The historic Avanti strategy is to maximise margin on a strictly limited number of seats, which cannot be meaningfully increased.
This might make sense in the context of a crowded WCML (although I have my doubts), but in the context of HS2, which has functionally infinite capacity, it does not seem likely to be the optimum result.

== Doublepost prevention - post automatically merged: 23 Sep 2026 ==

Is that actually so? HS2 vehicles are much costlier and are only available (as things stand) in 200m lengths. WMT can do 80m, 120m, 160m or 240m. This means the added capacity required may actually be cheaper overall. It might even be free, as keeping 8 car sets formed all day is cheaper than splitting and joining so they are generally left as 8s even when it's quiet.

DOO is irrelevant as lengthening a WMT service doesn't add any more staff.

The Class 895 is certainly very expensive, but in the context of where we are now, I'm not sure that's the yardstick that should be used.
If we want additional rolling stock to add capacity from London to Birmingham beyond the capability of the currently ordered fleet, we are more or less free to select any trains we want. HS2 is an almost TSI compliant railway after all. As an example, TGV M is reportedly around €27-35m per trainset, considerably less than the 895s per seat (Ouigo configuration will be 740 seats per 200m).

I think my central argument is that we are now no longer in the environment in which any of the previous HS2 planning took place. Previously the classic timetable for the WCML was derived on the assumption that Phase 2 of HS2 would be following soon after Phase 1 and that HS2 would be full of long distance services from across the north of England in relatively short order.

We are now in an environment where HS2 has an enormous capacity surplus for the foreseeable future, quite possibly forever.

That, in my view, fundamentally changes the calculation.
We should consider our position from first principles rather than cleaving to plans drawn up a decade ago in a different world.

After all, its still over a decade until HS2 starts operating at least.

I don't know if anyone remembers, but a few years ago a report was (briefly) published by the government with an annexe of the anticipated pre and post HS2 fare takings for various flows. I'm hesistant to attach it here because it was withdrawn shortly after I downloaded the relevant section and reuploaded without it, and I have no idea what its copyright status is.

They showed (as expected) the demolition of the classic fare revenue for flows like London-Birmingham, but also the collapse of London-Coventry by around half, from £34m to £17m.

I think by far the largest remaining flow on the WCML south of Birmingham was Milton Keynes to London, which I suppose is not surprising!

But if half of the Coventry traffic to London was expected to be lost to the baseline Birmingham International-London service, what would it be like if we filled the spare capacity on HS2 with an total 6-8 trains per hour London-Birmingham International-Birmingham? What would be the change in dynamics for the rest of the railway in the area?

 

[SynthD]

Unfortunately, I don't think this is a safe assumption at all. Experience shows us that the government is desperate to increase fares revenue by any means possible. So I think we will see revenue-maximised pricing on HS2

Except for the specific government announcement that it wouldn’t be operated this way.

You are forgetting that well over £100bn will have been spent on the line and that higher revenue to the exchequer means a larger contribution towards the UK's crippling debt burden

The debt is paid off by ticket revenue and doesn’t count as that kind of government debt apparently.

I have no idea what its copyright status is.

Still ok to share it for discussion. Does it say anything about reductions in coaches and personal driving?

 

[HSTEd]

Still ok to share it for discussion. Does it say anything about reductions in coaches and personal driving?

No, the section I have is simply before and after on ticket revenue using their modelled HS2 abstraction rates, using pre coronavirus revenue as a baseline. It's from 2023.

EDIT: On second thoughts I've decided to remove it because I'm not sure if its obtainable elsewhere.

EDIT #2: The wayback machine to the rescue. This link is to an archive copy of the original version of the "West Coast Strategic Advice" report that was mistakenly published with its appendices attached and then taken down. I can't really quote it because its a primarily a graphical report. The fragment I have is from Appendix B.


The internet knows all and sees all!