HS2: What if it had followed the incremental German/Swiss "Neubaustrecke" concept?

[Sorcerer]

It's a reasonable proposition, though the Skinkansen and LGV 1 were a lifetime away now. Frankly the UK domestic high speed network is also much better than either Japan or France's were at the same time so there is less of a motivation to do it that way. Also the Chinese and to a lesser extent the Spanish also avoid or at least started to avoid doing some of the more difficult city centre stations as part of the endevour.

But the circumstances surrounding the construction of the Tokyo to Osaka Shinkansen were still very much the same as the circumstances surrounding the purpose of HS2. Both lines were intended to serve as new trunk railways to relieve pressure on existing infrastructure, and in our case we had the luxury of being able to replicate different high speed rail models around the world including the Shinkansen when coming up with HS2 and draw upon their own experiences. Arguably the biggest reason we never built on high speed railway infrastructure in the past, and most certainly a big reason why we still struggle with it with HS2 right now, is the lack of political will to see through such a project combined with the gross mismanagement.

Finally doing it incrementally starting with an easier task is essentially the way that innovative organisation do things today, SpaceX didn't try to build Starship as their first project. It's also the case that if you build capability incrementally and organically instead of buying it in you are frequently capable of doing things that nobody at the start of the endeavor could have possibly imagined, see Starship.

I doubt SpaceX would have gotten as far as building Starship without the government subsidies it has received in the past. The issue at hand with high speed rail isn't innovation anyway so I don't think the two really compare.

 

[Technologist]

The Ribblehead Viaduct is hardly universally praised by all, although that might be hard to appreciate from inside the railway circle.

Additionally, that viaduct is made of stone, if you want them built at reasonable cost modern ones will not be, they will be made of precast concrete with a far less rosy look. Although you might be able to clad them with factory assembled brick slips.

You are taking that a little too literally, I should have put the Harry Potter viaduct up everyone likes that!

From research I've been involved with in other sectors the public don't actually wany things to pretend that they are something else, if we are building a modern fast railway line the viaduct should be modern and fast looking. What the public want is something which looks like its meant to be there, has some design flair and some graceful proportions. People go to visit and drive across the Millau Viaduct it's something to be proud of.

The people who live right under it will never like it whatever you do, but the public at large won't mind ploughing beautiful architecture across stunning nature particularly if its also the cheapest and most effective way to deliver something of benefit to them.

 

[HSTEd]

You are taking that a little too literally, I should have put the Harry Potter viaduct up everyone likes that!

It is, at least, made of concrete!

From research I've been involved with in other sectors the public don't actually wany things to pretend that they are something else, if we are building a modern fast railway line the viaduct should be modern and fast looking. What the public want is something which looks like its meant to be there, has some design flair and some graceful proportions. People go to visit and drive across the Millau Viaduct it's something to be proud of.

Design flair gets very expensive though, given it substantially increases the work that goes into the design and such of the viaduct and its supports.
If you take a look at the very long viaducts in China they are a hundred miles of almost identical bridge spans, occasionally with one shorter or longer because of a road or similar piece of infrastructure in the way.
Apparently there are only two standard span lengths, 24m and 32m.

Rather hard to make that appear to have "graceful proportions".
If you try to give them the level of design polish of the Millau Viaduct it certainly won't be cheap or fast to build!

 

[Technologist]

It is, at least, made of concrete!



Design flair gets very expensive though, given it substantially increases the work that goes into the design and such of the viaduct and its supports.
If you take a look at the very long viaducts in China they are a hundred miles of almost identical bridge spans, occasionally with one shorter or longer because of a road or similar piece of infrastructure in the way.
Apparently there are only two standard span lengths, 24m and 32m.

Rather hard to make that appear to have "graceful proportions".
If you try to give them the level of design polish of the Millau Viaduct it certainly won't be cheap or fast to build!

There are plenty of products which we expect to be both affordable and also aesthetically pleasing. It is clear that aesthetics played zero roll in the Chinese viaducts, humans like order and complexity, the Chinese viaducts provide only one of those things. However that doesn't mean that pretty minor changes couldn't make them more attractive.

Fortunately the UK landscape in most places is more interesting than those flat plains in China, ergo actual viaducts would probably never be more than a few km long and many would have a curve on them. Rather than two spans I suspect that we'd be looking at working with more like a Brio set of standardised parts. We'd be looking at about 5000 spans on a viaduct heavy HS2 phase 1, I'd be surprised if there was significant economies of volume between having 5000 pieces in 2 designs vs 5000 in 10 designs especially as we are just talking different tooling in the same factory.

But the circumstances surrounding the construction of the Tokyo to Osaka Shinkansen were still very much the same as the circumstances surrounding the purpose of HS2. Both lines were intended to serve as new trunk railways to relieve pressure on existing infrastructure, and in our case we had the luxury of being able to replicate different high speed rail models around the world including the Shinkansen when coming up with HS2 and draw upon their own experiences. Arguably the biggest reason we never built on high speed railway infrastructure in the past, and most certainly a big reason why we still struggle with it with HS2 right now, is the lack of political will to see through such a project combined with the gross mismanagement.


I doubt SpaceX would have gotten as far as building Starship without the government subsidies it has received in the past. The issue at hand with high speed rail isn't innovation anyway so I don't think the two really compare.

SpaceX got government contracts to provide services that the government wanted, they offered to do these things at about 1/2 or less than what the competition offered to do it for. They developed a rocket for about 1/5 to 1/10 what NASAs should cost process thought it would.

This is an example of government procurement supporting innovation not an example of a subsidy.

Secondly SpaceX is more an example of operational excellence rather than innovation, if you look at the stages of development, Falcon 1 was simplified rather than innovative, that was funded directly by Elon Musk, venture capital and government venture capital from DARPA (small minority). Falcon 9 was simplification (one engine type) and improved manufacturing techniques, followed by rapid iterative innovation. It's engines are for example less advanced in their operating regime than the ex Soviet engines their competitors use.

The innovation with reusability was that they designed a bigger rocket than they needed because it doesn't actually cost much more to oversize your rocket, they then started testing re-use in flight with the discarded boosters that their customers had already paid for. Again this is incremental innovation rather than new technology.

Its only with Starship do we start seeing genuine new technology but by this point everyone is basically happy to hand SpaceX blank cheques because Starlink is worth ~$400 billion or so when at scale.

There is plenty of opportunities to improve railway construction and much of it is iterative innovation/operational excellence.

 

[HSTEd]

Fortunately the UK landscape in most places is more interesting than those flat plains in China, ergo actual viaducts would probably never be more than a few km long and many would have a curve on them. Rather than two spans I suspect that we'd be looking at working with more like a Brio set of standardised parts. We'd be looking at about 5000 spans on a viaduct heavy HS2 phase 1, I'd be surprised if there was significant economies of volume between having 5000 pieces in 2 designs vs 5000 in 10 designs especially as we are just talking different tooling in the same factory.

The viaduct sections are short enough they probably don't have to be curved.
The simple way to build viaducts in that situation is to build them out of straight sections that are slightly wider than the railway requires. That way the track on top can slew across the viaduct as required to maintain its smooth curve, its not as if high speed lines have particularly sharp curves, so that is not too much of a problem.

You'd have to make sure all your span designs had lifting points in largely the same places so that you can use standard lifting and construction equipment.
Also you'd probably build your viaduct so it was continuous, since the whole point of viaducts is to reduce interaction with the surface to a minimum.

 

[Technologist]

The viaduct sections are short enough they probably don't have to be curved.
The simple way to build viaducts in that situation is to build them out of straight sections that are slightly wider than the railway requires. That way the track on top can slew across the viaduct as required to maintain its smooth curve, its not as if high speed lines have particularly sharp curves, so that is not too much of a problem.

But I want a nice gentle curve on the longer ones, we can do that technique to cope with the transition to curves.

 

[HSTEd]

But I want a nice gentle curve on the longer ones, we can do that technique to cope with the transition to curves.

If you want smooth curves your costs are going to go through the roof since at that point almost every single section is going to have to be designed and manufactured individually. The design and tooling work on even a single new section design will stretch into six or probably seven figures.
There will be several different structural analyses to run, a production analysis, the seismic and other safety analyses etc etc etc etc.

If you want an affordable high speed line you have to stay away from stuff like that!

EDIT:
30m spans on a 4000m radius curve are going to curve less than half a degree.... would you even notice if it was curved or not?

 

[Technologist]

If you want smooth curves your costs are going to go through the roof since at that point almost every single section is going to have to be designed and manufactured individually. The design and tooling work on even a single new section design will stretch into six or probably seven figures.
There will be several different structural analyses to run, a production analysis, the seismic and other safety analyses etc etc etc etc.

If you want an affordable high speed line you have to stay away from stuff like that!

EDIT:
30m spans on a 4000m radius curve are going to curve less than half a degree.... would you even notice if it was curved or not?

What I meant was manufacture a standard 4000m radius curve and then used a slightly wider track bed to smooth the track out.

Also there is nothing to stop us using a whole load of standard straight sections and then putting the smooth curve on a non-structural finish.

The point is that if we are building thousands of bridge sections per year we can invest in standardised customisation which is still very cheap.

 

[Austriantrain]

It's a reasonable proposition, though the Skinkansen and LGV 1 were a lifetime away now. Frankly the UK domestic high speed network is also much better than either Japan or France's were at the same time

I think that is a very important point. Japans railways obviously were quite slow at that time, and while the French were not so slow, the very large distances betwwen most major cities meant that only true High-Speed made journey times competitive with air achievable (although one shouldn't forget that the LGV Sud-Est was - like HS2 phase 1 - primarily a capacity project).

In the UK, OTOH, journey times from London to almost every major city in England are very competitive, both compared to the car and the plane. So I doubt that the average Mancunian or Brummy will say "wow, so great, I'll be in London even faster", and the average Liverpudlian would surely love a second hourly train, but that's probably it. Most might actually think "that is all nice and well, but I'd rather get to New Street/Piccadilly faster than having difficulties getting there, even if I then speed into London". While that doesn't preclude the fact that those faster journey times *will* be good for economic development, that's an abstract concept to most people.

I don't actually think there is anything wrong with HS2 as originally planned per se. It's just the integration into cities that bother me - when I read in the most recent Modern Railways that the West Midlands strategy now proposes four-tracking Coventry - Birmingham, massive investment to integrate the New Street/Moor Street/Curzon Street interchange and new undergound railways in B'ham, it seems to me that an integrated approach from the start would have yielded better results. I still can't see any argument for a Shinkansen in the UK.

 

[adamedwards]

One thing to bare in mind with the French LGV lines is the projects really started with the diversion into the RER lines of much of the suburban traffic at the Gare de Lyon and other Paris stations to free up space in the surface termini. It also helped in France that french trains were so much longer than ours anyway, so their stations are already long enough for 2 x TGV sets. Imagine if the 1960s Euston had been built with 20 coach platforms to every track.

If the London Northwestern and Overground trains in to Euston had all been sent into a 1960s "Crossrail A" route to London Victoria, Euston could then have been rebuilt for more longer distance services. Assuming the tunnel entrance was around Willesden, you then free up the tracks from there to Euston for our LGV Britanique to Birmingham. None of this happened, so we have a Euston which is still a suburban terminus as well as long distance, hence the expansion of platforms needed for HS2.

 

[AdamWW]

You are taking that a little too literally, I should have put the Harry Potter viaduct up everyone likes that

And unlike Ribblehead, the Glenfinnan viaduct is concrete.

 

[Austriantrain]

One thing to bare in mind with the French LGV lines is the projects really started with the diversion into the RER lines of much of the suburban traffic at the Gare de Lyon and other Paris stations to free up space in the surface termini. It also helped in France that french trains were so much longer than ours anyway, so their stations are already long enough for 2 x TGV sets. Imagine if the 1960s Euston had been built with 20 coach platforms to every track.

If the London Northwestern and Overground trains in to Euston had all been sent into a 1960s "Crossrail A" route to London Victoria, Euston could then have been rebuilt for more longer distance services. Assuming the tunnel entrance was around Willesden, you then free up the tracks from there to Euston for our LGV Britanique to Birmingham. None of this happened, so we have a Euston which is still a suburban terminus as well as long distance, hence the expansion of platforms needed for HS2.

An extremely good point, with the only drawback that to do that in the UK, you would need to accept a maximum train length of ca 300 meters.

If that was acceptable, it would have been interesting to see how the BCR would have turned out with Crossrail 4 in London and Crossrails in Birmingham and Manchester - only building HS lines out in the open and using existing city approaches and stations (more expensive on Headline costs for sure, but possibly much more useful).