Fantasy: If you could re-design HS2 ...

[Bletchleyite]

It’s not quite the same. The area served by OOC is far, far more populated than the area served by the Strathclyde suburban network, and the Elizabeth Line is far faster (and nicer!) too. But I wouldn’t personally be deterred by having to travel via Motherwell.

There is a reasonable chance I may have to, as I'm planning a trip to Scotland at the end of the month, and it's too far to drive. But I won't like it, despite the fact that my actual destination is Exhibition Centre, which is directly served from Motherwell so it's one change either way. (Curiously you can't get an itinerary from Bletchley to Exhibition Centre at all, so I suspect it's technically a non permitted route, but that's a bit moot in the circumstances and I can always get a paper ticket on the day from a WMT TVM as they are not planner based, and I do intend to use an off peak return).

If I lived in the North West still, though, the car would be on the agenda, and travelling via Edinburgh isn't entirely out as an option either. Standing on an overcrowded EMU with luggage - which is what the Liz would be like at OOC, is not my idea of fun. And Lumo to Queen St would be an option too were it not for their luggage policy (I will have more than they permit).

 

[HSTEd]

Standing on an overcrowded EMU with luggage - which is what the Liz would be like at OOC, is not my idea of fun. And Lumo to Queen St would be an option too were it not for their luggage policy (I will have more than they permit).

Won't most of the Elizabeth line trains at OOC be starting at OOC?

 

[Bletchleyite]

Won't most of the Elizabeth line trains at OOC be starting at OOC?

By way of extending the Paddingtons back? Yes, but a 345 is still a standee train. 400m of HS2 train won't get everyone seats on the Liz.

 

[Nottingham59]

If you could re-design HS2 with hindsight, knowing everything you know now, would you do it differently?

I think there are several things which should have been done differently.

ORIGINAL PROJECT SPECIFICATION
The orignal brief was (more or less) "Design a high-speed line to run from London to the West Midlands, with provision for extensions to the North and for access to Heathrow Airport".

It would have been far better to have started with something like "Specify and design in outline a rail network fit for the twenty-first century, with scope for twice as many passenger-km as carried by the current network, including provision for high-speed running and with better access to Heathrow and other major airports". Within that framework, the design of HS2 would have been very different.

BETTER INTEGRATION WITH THE CLASSIC NETWORK
HS2 has been designed as a stand alone project, with minimal connection to the classic network and no opportunities to share capacity. This is what I think could have been done:

• HS2 trains standarised at 500-520m long (2x260m), to match standards on classic lines (i.e. 265m Pendolinos on the WCML; 260m 10-car IETs on the ECML). For the first 30 years, only order classic-compatible trains.
• Except for the non-stop 360kph through tracks, all HS2 lines from Interchange through Delta to Curzon St have a maximum speed of 230kph (140mph). These lines should be part of the classic network, with connections at Berkswell (or north of International) to allow WCML traffic to reach Curzon St; and at Ham's Hall to allow access to the 125mph Derby Line.
• Curzon St and Moor Street to act as Birmingham "Hauptbahnhof", being the main terminal for all long distance trains in Birmingham. 7-platform Curzon St can handle many more trains than the 7tph timetabled by HS2 Ltd. Capacity at New St and International released for more frequent local services.
• The HS2 station in Manchester to connect to the WCML at Ardwick, with additional platforms on the east side with access to Guide Bridge/Leeds and Marple/Sheffield via the Ancoats Viaduct. High level pedestrian/travellator link across the whole Piccadilly complex, giving access to all platforms.
• Longer trains on WCML. 540m platforms on the north WCML at Crewe, Preston, Warrington, Carlisle, Glasgow to handle 520m HS2 trains.
• HS2 Euston built on two levels. Lower 540m platforms connected to HS2 tunnels; upper level 540m platforms connected to WCML fast lines. Longer plaforms at Hemel Hempstead, MKC, Bletchley, Rugby to handle 260m units running in pairs.

INSTITUTIONAL LEARNING
To reduce costs and maximise capacity, the rail industry as a whole should learn how to do the following:

• Build new lines cheaper than now, including any legislative changes needed, better contract management, and phasing for minimum cost
• How to build new lines with less disruption to local communities. Probably involving more use of viaducts, with deep structural parapets to minimise noise. Or more tunnels.
• How to reduce platform re-occupancy times of long-distance trains at terminals to allow 4tph to be timetabled off one platform. Would requires Shinkansen-style cleaning teams; passenger mezzanine floors; better ticket inspection processes; uniformity of destination; better marshalling of waiting passengers; lots of lifts for passengers with luggage
• Develop procedures to allow trains to split and join when they are not entirely alongside a platform.
• Design of couplers to allow trains to split and join with 99.99% reliability and with passenger doors open and in use.

PROJECT PHASING

• Build the network in smaller phases, so that learning from one phase can be utilised in the next
• Reduce peak requirement of all inputs (from concrete to environmental consultants). In particular do not give authority to proceed to four different construction consortia all on the same day, where they will all be competing for the same inputs at the same time

Suggested phasing:

1. Build HS2 Euston first, fitting out four 540m high-level platforms connected to the WCML fast lines. Leave the 6-platform undercroft to be fitted out later
2. Order 260m long classic compatible train sets, to run initally on WCML
3. Longer platforms at Crewe, Preston, Stoke etc. so that longer WCML trains can split at their first scheduled stop
4. Run 6tph 520m trains on WCML to destinations in the north, splitting and joining as necessary. This solves the WCML capacity problem for a decade. Run 8tph into Manchester from the south using existing paths: 3tph Avanti, 2tph XC, 1tph terminating at Airport, 1tph terminating at Macclesfield, 1tph via Warrington and Chat Moss.
5. Build 520m platforms at MKC and other main stations on the southern WCML. Build up frequency to run 12tph 520m trains off three long platforms, always keeping one spare.
6. Build HS2 Piccadilly with connection to WCML at Ardwick, and suburban platforms connected to Ancoats viaduct
7. Run 520m trains on WCML from Euston to Piccadilly
8. Build Curzon St to Interchange/Berkswell and to Derby line at Hams Hall
9. Run 520m trains on WCML from Euston to Curzon St.
10. Terminate trains from Leeds at Curzon St, and trains from Bristol at Moor St, creating a new Birmingham Hauptbahnhof
11. Build HS2 phase 1 from Old Oak Common to Berkswell, including the non-stop 360kph through lines at Delta.
12. Run 4tph HS2 high-speed services from OOC to Curzon St, reversing 2tph to Derby, splitting at Derby to serve Sheffield and Leeds (2tph) and Nottingham (2tph)
13. Build HS2 phase 1 from Hams Hall to Handsacre
14. Run 10tph 520m high speed service from OOC: 6tph direct to Crewe and beyond; 2tph Crewe and beyond via Curzon St; 2tph Derby via Curzon St. So HS2 only uses 8tph through Colwich
15. Build HS2 phase 2a from Fradley junction to Crewe
16. Build the tunnels from OOC to Euston and fit out the 6 platform Euston undercroft
17. Build HS2 phase 2 to Manchester, including tunnel
18. Build HS2 phase 2 to Leeds

Phasing like this would maintain flexibility and postpone spending on the most expensive parts of the programme until really needed.

 

[SynthD]

• Longer trains on WCML. 540m platforms on the north WCML at Crewe, Preston, Warrington, Carlisle, Glasgow to handle 520m HS2 trains.
• HS2 Euston built on two levels. Lower 540m platforms connected to HS2 tunnels; upper level 540m platforms connected to WCML fast lines. Longer plaforms at Hemel Hempstead, MKC, Bletchley, Rugby to handle 260m units running in pairs.

INSTITUTIONAL LEARNING
To reduce costs and maximise capacity, the rail industry as a whole should learn how to do the following:

• Build new lines cheaper than now, including any legislative changes needed, better contract management, and phasing for minimum cost

Is that a very high level wish, or can we look into why certain platforms may need to demolish listed buildings or cut off rail lines to become 540m long?

If the third last step is add further capacity to Euston (long after paying for building the bare space in step 1), the benefit seems more delayed than the expense.

 

[Bletchleyite]

Over 500m is a bit long to be honest. I'd just have built it for 300m single units. Portion working really is not necessary, and European standards are really irrelevant for a railway where it makes no sense to run through trains to mainland Europe.

Another thing I would change is that it would be built for 915mm platform level boarding, what they have done has locked in inaccessibility forever (though multiple entry heights are possible per the SBB Giruno it results in a complex interior step/ramp arrangement). This would be achieved by dropping between the bogies and having a single wide set of double doors at roughly either the 1/3 or 2/3 (not both) position in that section.

 

[Nottingham59]

can we look into why certain platforms may need to demolish listed buildings or cut off rail lines to become 540m long?

Sure. There/s been extensive discussion about 420m platforms on the WCML. Might be worth asking the mods to reopen some of the threads below.

Remember the original plans for HS2 included specific designs for 420m platforms at Crewe, Preston and Carlisle, and Piccadilly. I think it may be feasible to extend to longer than that.

From Google maps, many of the main stations on the WCML have railway land around them and look like they could accommodate very long platforms, including Warrington BQ, MKC, Rugby, Bletchley, Stoke. Others have spare land, like Hemel Hempstead and Glasgow Central (extending across the river).

I don't think 520m long platforms would be feasible at Stockport, Macclesfield, Liverpool Lime St., Motherwell or Waverley.

If all else fails, you could move Crewe station to near Basford Hall and have as many platforms as you liked to split and join 18tph coming up the mainline from Euston.

See:

Longer trains on WCML - How much capacity would they have added, compared to HS2?

In an Alternative Universe, how much capacity could have been added by running longer trains on the WCML, instead of building HS2? To get the most capacity out of HS2 Phase One, it will be necessary to build 420m platforms at Crewe and Preston and Carlise, to allow more HS2 services to run...

www.railforums.co.uk

How to add 400m platforms to Manchester Piccadilly

If HS2 Phase 2B is not going to be built, then capacity to Manchester will be severely limited by the number of trains that can get through Colwich Juntion and Stafford. The only way to increase capacity to Manchester is to run longer trains. This is how I would add 3 x 400m platforms at...

www.railforums.co.uk

Where to put 400m platforms in Liverpool

Following on from the Manchester Piccadilly thread, where in Liverpool could take 400m platforms? These would be needed to handle full length HS2 train sets, even if the approach were on existing lines...

www.railforums.co.uk

400m platforms at Glasgow Central?

[Crayon alert!] Could space be found for 400m long platforms at Glasgow Central? When HS2 phase 1 starts, the fastest route from London to Edinburgh will be via HS2 phase 1 and Carlisle. So I'm expecting England-Scotland passenger traffic via Carlisle to grow strongly over the next twenty...

www.railforums.co.uk

Suggestions for new short-ish lines

Crossgates - Whinmoor. Halt near Stanks. Single track would do it. Whinmoor is a deprived area of Leeds. This is an excellent idea as the area available will become East Leeds new town of 10,000 houses to the east of Whinmoor but nobody with clout in West Yorkshire is interested from the County...

www.railforums.co.uk

 

[Sorcerer]

Over 500m is a bit long to be honest. I'd just have built it for 300m single units. Portion working really is not necessary, and European standards are really irrelevant for a railway where it makes no sense to run through trains to mainland Europe.

It may not be necessary but if ever the full route was actually built I think it would do good to allow two services on a single path such as the planned Liverpool/Lancaster splitting service along with full 400m Manchester services. It's only because of the cutbacks that we won't get the luxury unfortunately.

 

[HSTEd]

A 300m single unit train with a large loading gauge (see above for ~3.6m width and 4.95m) height would give HS2 so much capacity that I can't imagine it ever being filled.

Although I'd suggest the longest platforms we could get would be the best.

I think a Euston reconstruction should have removed everything on the entire site, put the platforms into the open air temporarily and then build a new concourse covering all platforms, with lifts and escalators down to them at intervals.

If you filled the entire site you could probably get enough capacity to allow diversion of trains out of Marylebone and even the midland Main Line side of St Pancras

 

[Nottingham59]

If the third last step is add further capacity to Euston (long after paying for building the bare space in step 1), the benefit seems more delayed than the expense.

Well, Old Oak Common is costing £1.7bn at 2019 prices, and that has six platforms in a box below ground level and eight platforms above ground.

The additional cost of building a bare underground box at Euston twenty years before it was needed would be around £1bn. A tiny fraction of the eventual cost of a whole new railway.

EDIT: HS2 stations are surprisingly cheap compared to the cost of any reasonable length of new line.

 

[SynthD]

They’re cheap because they are green/brownfield. Old Oak Common is not two stacked stations. A stacked Euston in its original width, with upstairs working while downstairs is dug out is not cheap.

 

[Bletchleyite]

It may not be necessary but if ever the full route was actually built I think it would do good to allow two services on a single path such as the planned Liverpool/Lancaster splitting service along with full 400m Manchester services. It's only because of the cutbacks that we won't get the luxury unfortunately.

That service is only necessary because of 400m units working to Scotland. 300m units could fit adequately everywhere with minimal work and/or ASDO and could mean a more useful pattern, including avoiding Oxenholme and Penrith losing their London service.

== Doublepost prevention - post automatically merged: 15 Mar 2026 ==

A 300m single unit train with a large loading gauge (see above for ~3.6m width and 4.95m) height would give HS2 so much capacity that I can't imagine it ever being filled.

3+2 and even 3+3 are grim. You can keep that (and go to Spain and ride those horrid Talgos as penance).

Although I'd suggest the longest platforms we could get would be the best.

I think 300m ish (some may require ASDO) is the most you could get at most WCML platforms without serious work, except...

I think a Euston reconstruction should have removed everything on the entire site, put the platforms into the open air temporarily and then build a new concourse covering all platforms, with lifts and escalators down to them at intervals.

If you filled the entire site you could probably get enough capacity to allow diversion of trains out of Marylebone and even the midland Main Line side of St Pancras

Yes, I think Euston should be fully rebuilt, ideally something like 20 300m* platforms all on the same level with the concourse over the platforms a la New St but with a higher ceiling, and with most of the concourse on the paid side to avoid the need for the A end nonsense. With diesel totally banned it would be fine, you can create a feeling of modernity and space with uplighting. To get the length the platforms should be pretty much all the way to the Euston Road.

* Actually perhaps 340m ish, to allow one platform to be occupied by two 8 car (160m) local services where expedient.

 

[Sorcerer]

That service is only necessary because of 400m units working to Scotland. 300m units could fit adequately everywhere with minimal work and/or ASDO and could mean a more useful pattern, including avoiding Oxenholme and Penrith losing their London service.

I think the main issue with a 300m unit is that at some stations it wouldn't be possible without extra work. Liverpool Lime Street's three long InterCity platforms (6, 9 and 10) barely fit a 265m long 11-coach Pendolino, so a 300m long train wouldn't fit without crossing over some points. This means you would unfortunately need a subfleet while 300m units would work Manchester and Scotland. Pity those opting to cancel phases didn't consider these limitations.

EDIT: I just realised you were probably specifically referring to Scotland services rather than WCML and HS2 services in general. However, Glasgow Central also seems like it'd need some work to fit a 300m unit in Platforms 1 and 2 which are the main London-bound platforms, so the point of a subfleet still stands.

 

[HSTEd]

3+2 and even 3+3 are grim. You can keep that (and go to Spain and ride those horrid Talgos as penance).

Who said anything about 3+2 seating. It would be 2+2 because all the extra capacity will be in the form of standees in the very wide gangways!

I will get my coat!

 

[BogiePicker]

Over 500m is a bit long to be honest. I'd just have built it for 300m single units. Portion working really is not necessary, and European standards are really irrelevant for a railway where it makes no sense to run through trains to mainland Europe.

Another thing I would change is that it would be built for 915mm platform level boarding, what they have done has locked in inaccessibility forever (though multiple entry heights are possible per the SBB Giruno it results in a complex interior step/ramp arrangement). This would be achieved by dropping between the bogies and having a single wide set of double doors at roughly either the 1/3 or 2/3 (not both) position in that section.

What have they gone for, 1015mm?

 

[Bletchleyite]

What have they gone for, 1015mm?

I believe it's something like that, yes, similar to the Liz core. Problem is it locks in inaccessibility at non captive stations basically forever. You might manage to raise a few platforms at key terminal stations like Manchester Pic and Liverpool Lime St, but anywhere freight and other units will need to go through are knackered, and it's intermediates where you get a big dwell improvement by not needing the ramp.

 

[Nottingham59]

What have they gone for, 1015mm?

From the HS2 Train Technical Specification:

The HS2 Platform will have the following dimensions:  Height: 1115mm above rail level; and  Offset: 1655mm from track centreline; platforms have a minimum curve radius of 1000m, but there is no adjustment of offset.

 

[HSTEd]

HS2 fouuht a long and probably unwinnable court legal battle with the EU over their choice of platform height until eventually being bailed out by Brexit.

And now we will be stuck with a handful of very intensively used, very akward platforms of the wrong height forever. Or at least without an enormously disruptive rebuild.

EDIT: Corrected 'court' to 'legal'

 

[Shrop]

... European standards are really irrelevant for a railway where it makes no sense to run through trains to mainland Europe.

Just to confirm my understanding, is this comment that "it makes no sense", because of loading gauge, or because of Schengen, or something else?

 

[Bletchleyite]

Just to confirm my understanding, is this comment that "it makes no sense", because of loading gauge, or because of Schengen, or something else?

A combination of Schengen, Tunnel security requirements and the lack of demand for long rail journeys to the mainland (which prevented Regional Eurostar getting anywhere).

It's all so niche that designing HS2 around it just makes very little sense.

 

[JamesT]

A combination of Schengen, Tunnel security requirements and the lack of demand for long rail journeys to the mainland (which prevented Regional Eurostar getting anywhere).

It's all so niche that designing HS2 around it just makes very little sense.

Wasn't the idea that building HS2 to European loading gauge meant we could buy cheap standard rolling stock, instead of needing something bespoke that fits on our legacy network?

 

[Bletchleyite]

Wasn't the idea that building HS2 to European loading gauge meant we could buy cheap standard rolling stock, instead of needing something bespoke that fits on our legacy network?

It was one of the ideas, but that obviously doesn't work unless you build all of it. Though I doubt the saving would be that great, you don't buy rolling stock literally off the shelf unless buying used which they wouldn't be, it is custom built for you.

 

[MarkyT]

It was one of the ideas, but that obviously doesn't work unless you build all of it. Though I doubt the saving would be that great, you don't buy rolling stock literally off the shelf unless buying used which they wouldn't be, it is custom built for you.

Even with a European bodyshell, trains are usually fairly highly customised for the particular application and climate, and livery, internal fitout of course. The vehicles built for Saudi Arabia recently have extensive special measures added for coping with the heat and sand, and no doubt the Moroccan trains are similar. The problem with the original mixed fleet concept was the HS2 contract still had to pay in full for development of a new UK classic compatible bodyshell, but once that approved design existed, series production of each car should be little more expensive than a more standard design for mainland continental operation. A bigger fleet of CCs spreads the development preliminaries wider so the premium for the major custom redesign is smaller per train. None of this stops future operators buying UIC trains for parts of HS2's network, but there will be considerable challenges to managing them with their captive loading gauge.

In addition, standard European bodyshells would not interface nicely with the ~1200mm platforms settled on, so there would still be custom work around the doors and steps, even though the floor height of many European HS trains is around the same as the HS2 platforms.

 

[osmarks]

Remember that over half of all passengers travelling from London to Doncaster on ECML are actually continuing further, whereas very few of those travelling from London to Sheffield go any further. This suggests that there is great value in developing a new rail route in some form from London to Sheffield, which of course could continue northwards from there, whilst also serving the population centres I have mentioned.

What northward continuation would you have? Scotland and Newcastle are served fine by the existing lines. The only place in the right direction is Leeds, based on eyeballing the map. I suppose with some upgrades you could move Leeds traffic to the MML or future MML-like line, but I'm not sure why you would want to.

 

[HSTEd]

What northward continuation would you have? Scotland and Newcastle are served fine by the existing lines. The only place in the right direction is Leeds, based on eyeballing the map. I suppose with some upgrades you could move Leeds traffic to the MML or future MML-like line, but I'm not sure why you would want to.

I think you can make an argument that the ECML is not particularly well routed to serve the broader north east.

It avoids many of the larger settlements in favour of older but smaller ones. Darlington instead of Middlesbrough, Durham instead of Sunderland etc.
I think if you were building a railway line to serve the North East now, you wouldn't select the ECML's route.

After all, the North East has more people further north than the North West, which is more or less a wilderness north of Preston and Lancaster. Only Carlisle until you reach the central belt.

 

[Krokodil]

Wasn't the idea that building HS2 to European loading gauge meant we could buy cheap standard rolling stock, instead of needing something bespoke that fits on our legacy network?

We were going to have to order a fleet of "classic compatible" sets anyway, to serve Liverpool, Scotland and Newcastle. Bearing in mind that a HS1-HS2 link was part of the original design I'd guess that building the line to European loading gauge was both to allow through running without compatibility issues and to allow for the possibility of double-decked stock in the future.

 

[MarkyT]

We were going to have to order a fleet of "classic compatible" sets anyway, to serve Liverpool, Scotland and Newcastle. Bearing in mind that a HS1-HS2 link was part of the original design I'd guess that building the line to European loading gauge was both to allow through running without compatibility issues and to allow for the possibility of double-decked stock in the future.

Although they would physically fit, there would still have been a compatibility issue with continental trains at the high platforms, unless a dedicated handful were made lower for those trains alone.

 

[HSTEd]

3+2 and even 3+3 are grim. You can keep that (and go to Spain and ride those horrid Talgos as penance).

Further to this, I actually got an opportunity to measure things or remember how first class in an 80x looks.

The 2-seat side of the first class on TPE's 802s takes up pretty much half the width of the train (which is 2.74m overall.)

That implies a 3.6m wide train could have 2+2 of TPE's first class seating, and still have a gangway ~860mm wide, which is almost enough to meet standards for wheelchairs!

It would certainly make loading and unloading faster. And if buffet vehicles stick around make them a lot more practical as the buffet space wouldn't end up so long and thin to preserve the gangway route around them.

 

[NCT]

I've just been brushing up on Rhine-Ruhr express progress

The full programme involves lots of 6-tracking and some very intricate station rebuilds. There are places that have had building permits for 4 years and still no spades in the ground. Dortmund has to be completely redesigned because of Deutschland takt.

HS2 had it right by going for as simple and as self contained as possible, and by going for a non-divisible phase 1 that's as small as possible and no smaller. If HS2 were to be a series of online upgrades and had more interfaces with the conventional network for clever through running, it would have ended up in a same sticky place as RRX.

 

[Shrop]

What northward continuation would you have? Scotland and Newcastle are served fine by the existing lines. The only place in the right direction is Leeds, based on eyeballing the map. I suppose with some upgrades you could move Leeds traffic to the MML or future MML-like line, but I'm not sure why you would want to.

My post that you were referencing here, was giving consideration to a new rail link after HS2 in its present form has been completed.
Since HS2 has failed the Sheffield and Leeds areas, and given the potential demand that I demonstrated, then yes I do think a new route from London to Scotland via Sheffield and Leeds could be worth further thought.
Also, when you say Scotland and Newcastle are served well enough by existing lines, I'm thinking of something like Paris to Marseille. This route is slightly longer than London to Edinburgh and has been achieved in just 3 hours for many years.
So I'm afraid I can't agree with your sentiment, I think we should be aspiring for a London-Edinburgh service taking 3 hours flat.
After all, if that could be achieved in 10 years from now (which of course is well beyond the UK's capability), we would then *only* be several decades behind the French!
So yes, I do think there is plenty of scope for a future route from London to Sheffield and Scotland to fill in one of the gaps in the UK network