How would you build a hypothetical 21st century railway network for Britain completely from scratch?

[PTR 444]

So if all of Britain’s railways somehow closed down and some years later there was a plan to completely rebuild a brand new rail network fit for the 21st century, how would you do it differently to the Victorians in the 19th century?

For the main Intercity corridors, I would have 186-200mph high speed track all the way with slower sections penetrating city centres, German style. All routes would run next to existing motorway routes between city centres while nearly all motorways would have an Intercity line running next to them. As a consequence, the main line from London to the East Midlands and Sheffield would run via Milton Keynes rather than Kettering.

I would completely segregate Intercity and regional/commuter trains, therefore all routes would need to be 4-tracked where there is demand for both types of service. This means that cities like Bristol, Southampton and Nottingham could support their own S-Bahn style service at metro-level frequencies, while also possessing frequent intercity trains to different parts of the country.

If you were briefed with such a momentous task, how would you go about it?

 

[HSTEd]

Build everything like a Shinkansen, stops at 20-30km intervals, ATO throughout and the top speed of all trains being 260km/h or higher. Ticket gates throughout and run high intensity clockface timetables on all routes, using a small number of standard stopping frequencies.

Urban networks would be built as fully automated light metros.

 

[Thirteen]

I think a 21st century railway would need to be future proofed and even then it still would need to be upgraded every so often by the time we get to the 22nd century.

 

[JonathanH]

It would probably need to be entirely underground to avoid issues with buildings and other issues in the way, and to actually ensure it could be built.

 

[The Planner]

Why follow motorways? You go the way it needs to go between A-B.

 

[fandroid]

Continental gauge routes throughout, allowing double-decker trains where they'd work, without physical barriers to their use. But more importantly to transform railfreight capacity. I'd even build strategic freight-only lines to create the opportunity for massive modal transfer off roads for all long-haul freight. I've always felt that a super-freight role would be a suitable alternative use for HS2 if the naysayers were proved right about high-speed!

No big city terminus stations, every one to be through-running and they could be underground as they'd need far fewer platforms on the main line and far fewer stations than a conventional underground local system.

If you doubt this notion, it's exactly what Germany is doing at Stuttgart and contemplating for Frankfurt and Munich.

New Civil Engineer magazine points out that tunnelling is becoming more and more competitive for transport schemes.

 

[Purple Orange]

If the rail network was being built from scratch, after the existing one had closed, we’d have a country where the towns & cities would have been shaped by the old network with track beds & viaducts running throughout. The legacy of the Victorian era railway is that it is rather comprehensive, providing connectivity to the majority of the country, but it doesn’t have the capacity we need and has too much mixed use traffic. It is roughly 40% electrified, but 70% of services are electric which tells us how concentrated our railway is, yet there is too much diesel running under wires. So I’d do the following:

• Build a high speed intercity network based upon:
  • The full Y-shaped HS2 network
  • The current NPR plans plus extensions in to Liverpool, Leeds, a junction connecting NPR to the HS2 eastern branch to York & Sheffield.
  • Build a high speed line to Scotland.
  • Build the Great Western mainline to High Speed standards
• Build underground networks in all our major cities.
• Rebuild the remaining network to facilitate overground metro & semi-fast networks around our cities.
• Ensure that no area that had a railway line before is without a railway line after.
• All lines to be electrified

 

[edwin_m]

Why follow motorways? You go the way it needs to go between A-B.

The motorway network was centrally planned - sort of - to provide a nationwide transport network based around the perceived needs of the UK in the 1960s and 1970s, which probably haven't changed much since then. A centrally planned rail network would look much more like that than our current one, which was defined by expected passenger and freight movements 100+ years earlier and by the commercial decisions (many of which turned out to be totally wrong!) of competing private companies.

 

[The Planner]

The motorway network was centrally planned - sort of - to provide a nationwide transport network based around the perceived needs of the UK in the 1960s and 1970s, which probably haven't changed much since then. A centrally planned rail network would look much more like that than our current one, which was defined by expected passenger and freight movements 100+ years earlier and by the commercial decisions (many of which turned out to be totally wrong!) of competing private companies.

In which case it would end up the same way with many links not getting built. M64, Solihull Strensham, M23, Ringways etc

 

[edwin_m]

In which case it would end up the same way with many links not getting built. M64, Solihull Strensham, M23, Ringways etc

The ringways and things like Solihull Strensham are less important for rail, because they bypass all or part of a major destination where most passenger trains would stop. Even freight and non-stopping trains could usually pass through the centre without causing anywhere near the impact of through road traffic on the local network.

I should have made clear I was thinking more of the major links. In particular, the way the M6 joins the M1 instead of continuing separately to London suggests that the MML and WCML might have been similarly combined.

 

[NoRoute]

I should have made clear I was thinking more of the major links. In particular, the way the M6 joins the M1 instead of continuing separately to London suggests that the MML and WCML might have been similarly combined.

Going back in history originally they were like that, the early Midland route went to Rugby and joined what is now the WCML down to London. The separate Midland route down to London was a later addition. Some of the Beaching plan maps considered going back to that arrangement.

 

[PTR 444]

In which case it would end up the same way with many links not getting built. M64, Solihull Strensham, M23, Ringways etc

I certainly don’t think every motorway would need a parallel HSL. Maybe just the key London radial motorways as well as the M5, M42, M62 and M25. The latter would form an LGV Interconnexion style route encircling London, with trains between Southampton and Birmingham for example going that way rather than via Oxford as they do now.

 

[zwk500]

It's worth remembering that the population map would look radically different depending on when the railways had closed, but for the purposes of this I will assume they closed in the 70s as Motorways took over so the main intercity routes would look broadly similar.

As general principles I would have the following:
UIC gauge, Speed Signalling, ATP (or equivalent) as standard. 25KV AC OLE only.
All lines to be fully grade-separated. ATO to be used with Conductor-Only Operation.
All stations to be ticket gated, e-tickets available as standard.
Entire network to work only in metric, all documentation to be held on cloud-based platforms with dynamic updating and proper error raising/fixing protocols.

Streamlined IC routes: Anglia as is, ECML Main line via Leeds, MML 4-track throughout and straighter alignment London-Leicester-Derby-Sheffield (Nottingham on Branch), WCML direct London-Birmingham-Manchester-Preston, GWML direct Reading-Bristol etc. All These lines at 140mph alignment, mostly 4-tracked for local services. All major stations Hourly to London minimum, some 1tp2h extensions off route.

Commuter routes 100mph max, Clockface taktplan.
Regional Routes 90-100mph or bust, 1tph minimum viable service.
Freight would have ECP brakes, capable of 60mph minimum.

 

[Nottingham59]

Starting from scratch gives too many possiblities for me. But developing the existing network I'd start with standardised national train fleets:

• Pointy-nosed: There is some sort of regulatory threshold at 250kph, so I'd standardise the main line fleet to 249kph (=155mph), which seems a sensible aspiration for ECML, GWML, MML, XC, HS1 UK etc. A standard length of around 260m (=11car Pendolino) would fit most main stations. Lengthen St Pancras and other city stations as needed. (Personally, I'd have selected this standard for HS2, so 520m platforms at Euston, Birmingham and Manchester, but much cheaper tunnels and earthworks. But it's too late for that now.)
• Mainline BEDMU: 110mph, which is the fastest feasible design with end-corridor connections. Units of 3 or 4 cars, which can be linked together as needed. For the bendy bits of the WCML and all other mainlines. With interchangable under-floor battery modules and HVO diesel modules as needed, and in-line power cars for the longest routes off the wires.
• Regional chuggers: 80mph designed for lowest possible cost of operation.
• Rural branch lines: 50-80mph trams or tram-trains, with track brakes etc. for safe operation with pedestrians and road traffic. So rural stations don't need overbridges and lifts, and road/foot crossings can use just traffic lights.

All trains to use in-cab signalling and eventually Autonomous Vehicle technology to be capable of automatic/driverless and even unattended operation.

 

[yorksrob]

I'd build everything like the network covered by the South Eastern and Chatham Railway Companies Managing Committee.

Two of everything to everywhere so there's always an alternative during engineering works, disruption etc.

 

[GRALISTAIR]

No London Station would be a terminus

 

[zwk500]

No London Station would be a terminus

I would have terminus platforms with through lines. Trains from York do not need to continue to Brighton, for instance.
But I would concentrate on 4 or 5 terminals, all with crossrail type links between each other, for The Southern area, Western, North West, ECML and East Anglia.

 

[30907]

I'd build everything like the network covered by the South Eastern and Chatham Railway Companies Managing Committee.

Two of everything to everywhere so there's always an alternative during engineering works, disruption etc.

...but hopefully considerably better aligned and definitely better constructed!

Actually your point is worth pursuing - while the early motorway model is attractive and makes some sense for high speed lines, in a relatively densely populated country you might do well with more than one 110mph Intercity route between (say) London and the West Midlands (or between Leeds and Manchester, as Leeds-Bradford-Halifax-Huddersfield-Manchester is rather sinuous!).

Germany, for example, has two routes E-W through Nordrhein-Westfalen, two between Frankfurt and Berlin and the Rhine and Munich....

No London Station would be a terminus

I think that needs qualification - a city the size of London is not comparable with Brussels or Stuttgart, and there will be an imbalance of (sensible) HS routes north and south of the Thames. I agree that from scratch you would build RER and Intercity routes across the centre, but not HS2, HS3 and HS4. (3 to the West and 4 via Cambridge to the East Coast).

 

[stuu]

By starting with objectives, like population size and travelling times, and basing the network around that. Start with everywhere with 10k people getting an hourly service and build up from there. Or whatever the objective should be to best match capital with benefit.

 

[HSTEd]

I would have terminus platforms with through lines. Trains from York do not need to continue to Brighton, for instance.
But I would concentrate on 4 or 5 terminals, all with crossrail type links between each other, for The Southern area, Western, North West, ECML and East Anglia.

I don't see why would have multiple terminals if we are building from scratch?

With ATO (or driverless trains) and the like the platform reoccupation times go through the floor.

Surely we would build something like the Tokyo megastation complexes? Shinjuku station complex had something like 1.3bn entries and exits in a pre-coronavirus year. That would be something like half all the entries and exist on the entire extant National Rail system.

Journey times would be much shorter and the whole thing would trend into an overgrown Metro/RER - even with 30km stop spacings Shinkansen trains manage 110mph+ service average speeds. At 20km its still 80+mph.

EDIT:

Sure passenger dispersal from a megastation through metros etc would be challenging, but it would offer many advantages on a national scale by making transfers through London far easier and thus simplifying passenger routings. Lines could have secondary London stations to aid passenger dispersal, but all trains would terminate or at least call at one London super-interchange like Shinjuku. I don't think termination makes sense in such a system to be honest, not that it really matter when you have computer operated trains regardless.

 

[zwk500]

I don't see why would have multiple terminals if we are building from scratch?

In essence because you still need some slack in the system. Even with ATO you need to clean and service long distance sets and there will be delays from passengers and faults.
London is also massive and in my Alternate reality the tube developed as is so there's not the space for an underground HbF nor is there the land for an above ground one.
However as I mentioned I'd have far fewer terminating services and lots more XR/Rer type crossing services formed of ibeer suburban lines.

 

[Purple Orange]

In London, would it not be best to have one long distance terminal station, for the type of journey that the full Y-shaped HS2 network would have delivered in to Euston, plus Eurostar and the long distance GWML services? That would require a station with 20-24 platforms at 400m in length.

Then have several Crossrail and Thameslink operations for the type of services that flow in to Waterloo, Victoria & London Bridge, paired up with the services that run in to Liverpool Street, the southern WCML, Chiltern mainline, southern ECML etc etc?

 

[jayt22]

Maglev or - after some research to confirm passenger service is possible - Hyperloop for long-distance services. All services should be automated.

Without expensive and unreliable human drivers, costs could reduce, and service reliability increase. Existing drivers could be redeployed as customer service staff, either onboard trains or in stations. People would still be needed to staff the railways, at least for a while. (And customer service staff would have guards' training, on board in case of very rare emergencies. )

I don't expect the RMT would like this, not one bit, but the DLR and other advanced rail transit systems have no need of drivers.

Many airplane flights are now entirely automated, with the pilots only there to manage the machines - and very occasionally take over in an emergency, or for a bit of practice. EASA has in fact proposed a new category of single-pilot commercial flights, so that new generations of planes able to deal with emergencies better than experienced pilots, will only have a human in the cockpit for very rare events.

If the European Aviation Safety Agency can propose such a system, surely it should be possible to build a new, very safe and much more efficient railway.

 

[Nottingham59]

Welcome to the forum!

Without expensive and unreliable human drivers, costs could reduce, and service reliability increase.

I tend to agree with you on this point. The technology that is bringing self-driving taxis to the roads would readily be converted to allow self-driving trains. The cognitive load on a train driver is, shall we say, certainly no greater than that needed to drive a bus.

But making it happen will be tricky. The railway tends to lag about 40 years behind the aviation industry. They've only recently started using aluminium alloys instead of steel for train bodies, and I've not heard of any plans to use structural composites, as aircraft manufacturers have been doing for twenty years.

Remember that the railway is an integrated system, with inbuilt monopoly power or some travel flows, and the pace of change is glacial. There is almost no scope for a South West Airlines to enter the market with a much lower cost base and disrupt the established cozy system.

 

[Bletchleyite]

Rural branch lines: 50-80mph trams or tram-trains, with track brakes etc. for safe operation with pedestrians and road traffic. So rural stations don't need overbridges and lifts, and road/foot crossings can use just traffic lights.

I'm not sure we would really build those now - probably just buses - but I did create a thread a while back suggesting light rail style operation to be the way to go for quieter lines. I think "interurban tramways" of the style of Blackpool towards Fleetwood or the Isle of Man Electric Railway without the old trams have a lot of potential for very low cost operation of rural rail. Level crossings are a big cost, so allowing them to be operated on sight with barriers (and the train stopping before proceeding across them at a low speed) would be a big money saver.

I should probably say - if I was starting the railway up now it would have the powers to tender or indeed operate itself fully integrated bus services!

 

[Nottingham59]

I should probably say - if I was starting the railway up now it would have the powers to tender or indeed operate itself fully integrated bus services!

Yes, I agree totally with that.

I'm not sure we would really build those now - probably just buses

Well at the speed that Autonomous Vehicle technology is progressing, I think buses and rural tram-trains will be functionally indistinguisable within 20 years. Keep the rails and run battery AV trams with catenary recharging at stations; or pave over the trackbed and run autonomous taxis or trolleybuses on a segregated busway. The shape of the wheels becomes irrelevant.

 

[zwk500]

Yes, I agree totally with that.


Well at the speed that Autonomous Vehicle technology is progressing, I think buses and rural tram-trains will be functionally indistinguisable within 20 years. Keep the rails and run battery AV trams with catenary recharging at stations; or pave over the trackbed and run autonomous taxis or trolleybuses on a segregated busway. The shape of the wheels becomes irrelevant.

The shape of the wheels is very relevant, although it's certainly possible you could get the Unimog type arrangement where the rail wheels are lowered in only for guidance and all drive and traction is from the rubber tyred road wheels.
Would only be suitable for fairly low speeds though, probably 30/40 kph max

 

[jayt22]

Welcome to the forum!

I tend to agree with you on this point. The technology that is bringing self-driving taxis to the roads would readily be converted to allow self-driving trains. The cognitive load on a train driver is, shall we say, certainly no greater than that needed to drive a bus.

But making it happen will be tricky. The railway tends to lag about 40 years behind the aviation industry. They've only recently started using aluminium alloys instead of steel for train bodies, and I've not heard of any plans to use structural composites, as aircraft manufacturers have been doing for twenty years.

Remember that the railway is an integrated system, with inbuilt monopoly power or some travel flows, and the pace of change is glacial. There is almost no scope for a South West Airlines to enter the market with a much lower cost base and disrupt the established cozy system.

You're right about the glacial pace.

Replacing my local crossing some years ago, at one point they (Network Rail) decided that the only effective way to replace the single barriers was to take out a farmer's field, dynamite under the railway and create a huge ramp system to enable disabled users to cross between platforms. A ramp that would almost certainly never be used... all at a reported cost of £18 million.

When it was pointed out at the last minute that this would have a catastrophic safety impact on local roads and probably lead to the closure of the farm - because the field was critical to the farm - someone thankfully worked out that LIDAR had been in use on German railways for thirty years and therefore might be acceptable to stop trains if there was an object on the level crossing.

The single barriers were replaced with double-barriers and the excess budget presumably spent on significant safety improvements elsewhere.

In the unlikely event my suggestion ever went anywhere, in my imagination it would not be built by Network Rail under the auspices of the Railways Act. It would be a completely new system, regulated by the bodies created under a "21st Century Mass Transit" Act of Parliament. If it was built, they could turn existing lines over to freight traffic in the main. There could be no mixing of traffic with known, fully automated "trains" on a purpose-built network. Would this cost more than HS2? Possibly not!

 

[HSTEd]

In essence because you still need some slack in the system. Even with ATO you need to clean and service long distance sets and there will be delays from passengers and faults.

I'd argue you'd run trains "through" the London station complex, both to disperse the passengers over probably three London stops (one central and two peripheral), but also to allow the cleaning and such to be done at a cheaper location that is not the centre of London.

Given that London is in the corner of the country it wouldn't take long for a train from London to reach a stabling point in Brighton, Ramsgate or the like anyway - and it might as well stay in service.

You would want London to have the smallest possible station footprint - and massive arrays of platforms for terminating trains doesn't really qualify

 

[telstarbox]

It's a bit chicken and egg though. If Central London wasn't served by thousands of daily NR services as it is today, the land values would be much lower. The City and West End might be more residential and building heights would probably be lower.