Could the Crossrail 2 route be improved?

[pitdiver]

If anybody knows New Southgate, I was born there and lived there for 35 years. They would have noticed a large development has taken place at the bottom of station rd/High Rd. I am wondering whether this was done on the understanding that CR2 would eventually get there. Let's hope the people who purchased the properties aren't disappointed and end up with just a view of an old gasworks site

 

[Magdalia]

The reason to go by Finsbury Park rather than Dalston is simple: cost. This allows us to re-use the formerly six-track alignment north of the Copenhagen Tunnel. Rebuilding Finsbury Park (which has a disused platform on the western side) will obviously be disruptive and time-consuming, but it should be cheaper than building an underground station.

There is a misunderstanding here. There is no formerly six track alignment to re-use, apart from the eastern bore of Copenhagen Tunnel itself. The ECML is already 6 tracks all the way from the top of Holloway Bank to Alexandra Palace.

Furthermore, because the lines are grouped by direction not by use, there already exists a lot of grade separation that takes up additional space, with flyovers at Holloway, Hornsey and Alexandra Palace, and double diveunders south of Finsbury Park.

Another important factor with this part of the ECML is the location of the existing carriage sidings/depots at Hornsey and Bounds Green. These need to have continuous access and out, particularly to and from Kings Cross.

There is an existing example of trying to do something like this, which is the dedicated Thameslink tracks through London Bridge. This included new tracks at Borough Market, a complete rebuild of the station, and the Bermondsey flyover. It required 4 years of disruption, with lots of truncation and diversion of services. Trying to drive Crossrail through Finsbury Park in the same way would be much more difficult, disruptive, time consuming and expensive.

 

[NorthLondoner]

If anybody knows New Southgate, I was born there and lived there for 35 years. They would have noticed a large development has taken place at the bottom of station rd/High Rd. I am wondering whether this was done on the understanding that CR2 would eventually get there. Let's hope the people who purchased the properties aren't disappointed and end up with just a view of an old gasworks site

Much of New Southgate already benefits from existing Great Northern services as well the Piccadilly Line within walking distance. I don't think Crossrail 2 would be necessary with upgrades that can already be conducted on both the Great Northern metro services as well the Piccadilly Line. Something I'd imagine will be alot cheaper than building a tunnel all the way from Stoke Newington.

 

[Mgameing123]

The first thing I would do to Crossrail 2 which they should have added for the Elizabeth Line is make sure every core station has 4 platforms. 2 for each direction. It would increase capacity as one train could arrive one platforms while another departs on the other platform.

 

[HSTEd]

The first thing I would do to Crossrail 2 which they should have added for the Elizabeth Line is make sure every core station has 4 platforms. 2 for each direction. It would increase capacity as one train could arrive one platforms while another departs on the other platform.

It might allow you to push traffic somewhat higher, but given that two platform stations already achieve ~30+ trains per hour in Paris and on the Underground, I am not sure how much capacity you would get.

Four platform stations are far more difficult and expensive to construct, however.

 

[Mgameing123]

It might allow you to push traffic somewhat higher, but given that two platform stations already achieve ~30+ trains per hour in Paris and on the Underground, I am not sure how much capacity you would get.

Four platform stations are far more difficult and expensive to construct, however.

I also want to mention that it will bring more timetable resilience and more flexibility for maintenance works so if ever night services come we could close one of the tunnels and just run the core as a single track line.

 

[SynthD]

The Elizabeth line was designed to run at 32tph later. The stability of the service already heavily depends on the non-exclusive lines it runs onto (which also have nighttime maintenance). 32tph can be achieved with flying junctions (Stepney Green), terminating platforms on the inside (Old Oak Common) or short trains and fast turn arounds (Brixton).

 

[Crithylum]

The first thing I would do to Crossrail 2 which they should have added for the Elizabeth Line is make sure every core station has 4 platforms. 2 for each direction. It would increase capacity as one train could arrive one platforms while another departs on the other platform.

Building twice as much station costs twice as much. It would be far better to build Crossrail 3 and accept some weekend closures

 

[Mgameing123]

Building twice as much station costs twice as much. It would be far better to build Crossrail 3 and accept some weekend closures

I agree but would it be worth it considering the day when Crossrail 2 hits it’s maximum capacity?

 

[Crithylum]

I agree but would it be worth it considering the day when Crossrail 2 hits it’s maximum capacity?

Building a parallel line to similar places is probably better imo. Provides relief as well as making new convenient connections. This means some people may be able to take a single direct train instead of multiple. Instead of 4 tracking the Central line XR1 was built, which for people traveling from Farringdon to Tottenham Court Road to Bond Street is a massively more efficient solution than more Central line capacity.

If capacity becomes a problem in the future, it can be resolved in the future. Getting something built soon should be the priority over making sure it is never capacity constrained. 4 tracking isn't like HS2 where not building 20% strangles the other 80%. Not building 50% means the other 50% is fully utilised.

 

[Petitoiseau]

How many ”cross-city” rail projects around Europe and Australia are taking this approach? It’s not cheap to build new infrastructure alongside a working mainline.

For example, Munich is constructing a new pair of tunnels for their ”Second S-Bahn”, even where it parallels the existing alignment between Hauptbahnhof and Laim.

The RER E extension comes to mind. Also S21 in Berlin and the Citytunnel in Leipzig. As for Munich, the new portal is only 1,5-2km farther out than for the 1st Stammstrecke.

 

[Bald Rick]

I also want to mention that it will bring more timetable resilience and more flexibility for maintenance works so if ever night services come we could close one of the tunnels and just run the core as a single track line.

It also means lots more kit to maintain (and fail) for very little capacity gain over a railway with ATO.

Building twice as much station costs twice as much.

More than twice as much.

 

[Sad Sprinter]

About 10 years ago, the team said that the main reason to go to New Southgate was to put a depot there. I am surprised that land to build a depot can't be found somewhere between Northumberland Park and Enfield Lock.


I went to a public meeting in Wimbledon about a decade ago in which I suggested that there was no point in demolishing Wimbledon town centre to build new tunnel platforms when there are surface platforms on the fast lines which are never used except during tennis. I suggested that a new straighter mainline could be tunnelled from Clapham Junction to Hersham with optional platforms at Kingston to take all of the fast trains well away from Wimbledon, allowing Crossrail 2 to put twice as many stopping services through Wimbledon without the demolition of a single blade of grass in the Wimbledon area. The previously subdued crowd suddenly became very enthusiastic, but the panel refused to discuss it and just said it's too late to change the plan now.

Yes I went to the one in Balham which had a similar theme. Panic suddenly spread amongst the previously subdued crowd when it was revealed the local Waitrose would have to be demolished to make way for a new station entrance.

I remember an armchair rail planner by the name of Stephen Colbourne proposed the tunnel you mention. In fact he was at the Balham meeting. He claimed he was speaking to people in TfL about it but haven’t heard from him in a good few years.

Whilst the current Crossrail 2 route is the only one that seems technically feasible, I’m yet to be convinced it’s the most ideal pathway. I see no point in spending billions on a new railway to north London and leaving the great northern electrics services as a half-hourly service to Moorgate

 

[HSTEd]

Whilst the current Crossrail 2 route is the only one that seems technically feasible, I’m yet to be convinced it’s the most ideal pathway. I see no point in spending billions on a new railway to north London and leaving the great northern electrics services as a half-hourly service to Moorgate

Forcing the abandonment of the ECML Moorgate line is hardly going to be popular with the people who use it.
It'd probably be much politically palatable to just permanently assign it the Hertford Loop

 

[Sad Sprinter]

Forcing the abandonment of the ECML Moorgate line is hardly going to be popular with the people who use it.
It'd probably be much politically palatable to just permanently assign it the Hertford Loop

Yes I think that would be better. Shame the NCL can’t be extended down to meet the line out of Waterloo

 

[Mgameing123]

It also means lots more kit to maintain (and fail) for very little capacity gain over a railway with ATO.



More than twice as much.

How much would it gain for the cost? Because having 1 train arrive while the others still on the platform could boost frequencies up to every minute.

 

[SynthD]

How much would it gain for the cost? Because having 1 train arrive while the others still on the platform could boost frequencies up to every minute.

It depends on the demand. Imagine if you were genuinely able to double the number of passengers the Elizabeth line could carry. Would they all appear, would the investment be paid off? Not likely. You also run into other limits, like the people from the street reaching the platform, and out in an emergency. Has what you propose happened anywhere?

 

[Bald Rick]

How much would it gain for the cost? Because having 1 train arrive while the others still on the platform could boost frequencies up to every minute.

Could it? Is that done anywhere in the world?

Where would these trains go every minute? How many escalators would you need at each station to deal with that many people? Where would you build these escalators in the space constrained areas of London? How much bigger would the ticket halls need to be? where would they go? How many additional trains would be needed? Where would they be maintained? How many addtional emergency escape shafts would be needed to comply with the rule that you can only ever have one train in each emergency escape section? How much extra power supply would be needed for 5MW of train every minute?

etc.

 

[Petitoiseau]

Could it? Is that done anywhere in the world?

Where would these trains go every minute? How many escalators would you need at each station to deal with that many people? Where would you build these escalators in the space constrained areas of London? How much bigger would the ticket halls need to be? where would they go? How many additional trains would be needed? Where would they be maintained? How many addtional emergency escape shafts would be needed to comply with the rule that you can only ever have one train in each emergency escape section? How much extra power supply would be needed for 5MW of train every minute?

etc.

I don’t think it’s done at that frequency anywhere. However that’s what makes the difference between 24tph RER A and 32tph RER B/D in Paris (often scheduled 1 minute apart!). RER B and RER D have separate platforms but share a tunnel. However that is just one interstation, and not a whole line.

 

[HSTEd]

I don’t think it’s done at that frequency anywhere. However that’s what makes the difference between 24tph RER A and 32tph RER B/D in Paris (often scheduled 1 minute apart!). RER B and RER D have separate platforms but share a tunnel. However that is just one interstation, and not a whole line.

Doesn't RER A peak at a train every 90 seconds?
Indeed they sometimes have two trains in a platform at once (one arriving as one leaves).

Given our ability to have signal sections much shorter than the train, I'm not sure platforming is going to be a major issue in capacity terms in the modern era. We can have a train in the platform, doors open, most of the time if we need to.

 

[Petitoiseau]

Doesn't RER A peak at a train every 90 seconds?
Indeed they sometimes have two trains in a platform at once (one arriving as one leaves).

Given our ability to have signal sections much shorter than the train, I'm not sure platforming is going to be a major issue in capacity terms in the modern era. We can have a train in the platform, doors open, most of the time if we need to.

I checked, and the peak hour seems to have 25/26 trains depending on how you count. That would be a train about every 140 seconds.

https://rera-leblog.fr/wp-content/uploads/2024/08/horaires_est_ouest_dimanche-1-au-dimanche-8-septembre_800x1200.pdf

 

[The Ham]

Given that Crossrail 2 in any form now seems unlikely to be built in the commuting lifetime of anyone currently working, a Clapham Junction rebuild must be even less likely. The level of technological and other change in that time means what may seem essential now isn't by that time.

I can't see government policy moving towards new build railway projects anytime soon. It is really hard to see it getting onto the agenda for spending priorities for a good while.

It all depends on what the future rail use is (see the summary below for TL;DR):

Whilst that's true, the other factor to consider is that future rail use also isn't what it currently is and whilst rail use on the SWT network has been generally falling between 2016 and 2019 there is an argument that the network was at the limit of use at that time (at least in the peak hours, but with any significant failure taking a long time to recover).

Future use is always going to be hard to predict, however with the strikes looking like they are going to be stopping as part deals get sorted there's a fair chance that in the sort term we continue to see growth.

What that looks like is harder to predict, however based on the rolling average 12 month total increasing by between 1.3% and 4% over the last 6 quarters (with around 2.7% being fairly typical when you average across those 6) if we assume the next 6 quarters the rolling annual total grows at 2% then we'll see quarterly use for SWR of between 40 and 44 million each quarter.

If that trend continues longer term (which is unlikely, but given I've assumed a lot than average from the last 6 quarters, it could continue for longer than the next 6) then we could see the equivalent of 2016 use in around 2029.

Even if the average growth fails to 1%, then we still reach 2016 in about 2035. Although that is still far from certain, although there were several quarters in 2013-2016 when there was this 1% per quarter in the rolling annual growth, so with higher than 1% in the early years it's not impossible to see some fluctuations and still reach those sorts of numbers over that timeframe.

The problem is, that by the time we can be sure that those numbers are happening it's almost too late to start building to provide the capacity as it'll take time to deliver the scheme.

That's not to say that we should start building it now, however there should be a point (which isn't now now, however it also isn't when the annual use for SWT is 230 million) when we should start the process towards building it and another pint in time when we actually start building it.

That time may never happen (or even the first may happen but the point to build it not ever happen), but likewise both points could happen within the next 15 years.

For example if the get on with work short of starting building works when we hit 170 million and growing (current annual use is 153) and start building when we reach 200 and growing (as 2019 was 203 million and falling) then on the 1% growth numbers we'd be doing so in 2026 and 2030, with the trend showing that by 2042 (with enough capacity, which there wouldn't likely to be without Crossrail 2) that rail use could exceed 300 million a year.

Even being in a position to start building as we reach the similar numbers seen in 2019 (if passenger use is still growing) could, arguably, be starting too late.

Based on 2% growth then these are the next 6 quarters of rail use for SWT:
April - June 2024 41.2 million
July - September 2024 40.1 million
October - December 2024 42.6 million
January - March 2025 42.0 million
April - June 2025 44.5 million
July - September 2025 43.5 million

Likewise at 2.5% growth then it's:
April - June 2024 42.0 million
July - September 2024 40.9 million
October - December 2024 43.4 million
January - March 2025 42.8 million
April - June 2025 46.2 million
July - September 2025 45.3 million

It'll be interesting to see what the actual numbers are, higher than those (and not just one but all of them higher, so no getting too excited if we see 42.1 million for the next quarter) and we'll likely need to look at Crossrail 2 sooner than some think, much lower than the 1% numbers (see below) and it may not be on the cards for some time (if ever) because percentage growth in the next 6 quarters is likely to be higher than the 6 quarters of growth starting in (say) 2030 (even if the extra number of people each quarter is higher at that point).

Based on 1% growth then these are the next 6 quarters of rail use for SWT:
April - June 2024 39.7 million
July - September 2024 37.0 million
October - December 2024 37.8 million
January - March 2025 35.6 million
April - June 2025 36.4 million
July - September 2025 33.6 million

It's worth noting that in 2019 there were 52 million to 55 million per quarter, so whilst those would still be a significant way short of that, the point is we probably don't want to wait until we are actually seeing those numbers if rail use on the SWT network is growing (if takes 20 years to get to that point then there's not too much of a rush to do much as it could take another decade to get close to the 2016 numbers). What we need to see is what is the trend showing and even then we'll probably actually need to see data from 2027 (at least) to be sure that it's not just flushing through the impact of Covid and the strikes and even then it maybe at least 5 more years of data to be sure that we want to start building.

Obviously, SWT use is only one factor, however it's something which is easy to monitor.

Summary:

If annual growth rises at 1% per quarter then we may want to start building Crossrail 2 at 2019 passenger numbers which could happen in 2030.

Obviously if growth is lower than you're right it may be a long time coming, however if growth is higher (which it had been for the last 6 quarters) we may want to get a move on sooner.