What’s wrong with Crossrail 2
A lesson in British infrastructure planning
Crossrail 2 is London’s most important megaproject. It will go from Hertfordshire to Surrey via central London. It will involve ten or eleven new underground stations, two major additions to above-ground stations, and considerable works to pre-existing lines. It will be an engineering feat of scope bigger even than its sister, Crossrail 1, which now forms the Elizabeth Line.
But it is almost certainly not going to happen. This is mainly because it is staggeringly expensive. Even in 2015 the Treasury were unwilling to fund it. Since then, the delays and cost overruns on Crossrail 1, and the state of the public finances, have diminished the chances it will be built.
This is an enormous shame, because London is crying out for Crossrail 2. It will ‘unlock’ 200,000 homes, by giving areas suitable for development better public transport links. It will provide a much faster route across town than the Tube. It will have access to HS2 at Euston (if it ever arrives), and Tottenham Court Road will be a major interchange with the Elizabeth Line. The Tube lines that Crossrail 2 will relieve – the Victoria, Piccadilly and Northern Lines – are busy, and so are the services to Hertfordshire and Surrey that Crossrail 2 will take over.
In fact, of the ten most popular Tube journeys in 2023, shown below, four are on the Victoria Line and will be directly improved by Crossrail 2. Another two, on the Waterloo & City Line, will be indirectly improved when Crossrail 2 makes Waterloo station less busy.
There is no law of nature that states Crossrail 2 must cost as much as it does. Although infrastructure in Britain is much more expensive than in other countries, there are practical steps we can take to reduce costs. One of those is designing infrastructure better. Crossrail 2 is overbuilt: it involves unnecessary tunnelling and unnecessarily complex works to stations. It is also trying to solve too many problems at once, in ways that only serve to add to the price-tag.
This post explains what is wrong with Crossrail 2. The next post will discuss in detail a better way of achieving nearly the same aims. My scheme, which is designed explicitly to reduce costs, would give us more bang for our buck, increase the chance of Crossrail 2’s getting built, and leave money over to build other transport infrastructures into the bargain.
The purpose of Crossrail 2
To evaluate the current proposals for Crossrail 2, we need to know what its goals are. Then we can see how effectively the current proposals meet those goals, compared with the alternatives.
This doesn’t need to be an especially complicated exercise: you can work out what the goals of a railway project are just by looking at a map. Doing this, we get the following list of core objectives:
Relieving the Victoria Line and (to a lesser extent) the Charing Cross branch of the Northern Line, by digging a tunnel from Victoria to Euston/King’s Cross, and then on to Seven Sisters and Tottenham Hale.
Relieving the northern branch of the Piccadilly Line, by stopping at either Wood Green or Turnpike Lane, and by stopping at New Southgate, which is very close to Arnos Grove Tube station.
Improving inner suburban services on the lines out of Waterloo. These are the suburban trains that run through the London urban area, to places like Epsom and Hampton.
Improving services on the West Anglia Main Line from Tottenham Hale to Hertfordshire.
The route, however, has many more twists and turns than we might expect if that were all Crossrail 2 was doing. I think we can therefore add the following purposes to the list:
Relieving the southern branch of the Northern Line through the stop at Balham, which will also provide some relief to the inner-suburban Southern services into Victoria.
Relieving the Bank branch of the Northern Line through the stop at Angel, and thereby also directly connecting the Islington/Clerkenwell area to the West End.
Giving a rail connection to Chelsea.
Directly connecting Dalston with the centre of London.
The problems with Crossrail 2
The key to good infrastructure planning is to kill as many birds with as few stones as possible. Crossrail 2, in my view, is trying to kill too many birds with too many stones.
The reasons for this lie in Crossrail 2’s genesis. The history of the project is quite complex, but it seems to have come about by combining two proposals. First is the long-standing idea to build an underground line from Chelsea to Hackney, two districts of Zone 2 that are badly-served by the Tube. The other project is the natural successor to Crossrail 1, which runs across central London from east to west: a north – south crossrail line.
It can be seen from these two maps, both from the 1989 Central London Rail Study, that the current proposal for Crossrail 2 is a combination of these two schemes.
There is a conceptual distinction between a Tube line and a Crossrail line. A Crossrail line should run as fast as possible through Zones 1 to 3, only stopping at important destinations in the centre, and at the most important interchange stations. This speeds up the line, which allows it to go much further out, well beyond the M25, while keeping journey times competitive. The point is to provide an express tube service, and fast commuter trains.
The focus of a Tube line, on the other hand, is the London urban area. It stops roughly every kilometre within the centre of London, and every mile or so in the suburbs. This means the inner suburbs of London get a good service, but it restricts how far out the trains can go. Most Tube lines terminate around Zone 5; if they went much further, journey times would be intolerably long.
Tube and Crossrail lines should be seen as complementary to one another. There is a division of labour between the two types of line; when used together they provide a comprehensive network for the whole of the London city region.
But they should not be mixed and matched. For one, if a Crossrail line stops too often within Zones 2 or 3, it will slow down journey times. The line will be less powerful than it otherwise could be. Too many stops will also make the line much more expensive to build: Crossrail lines run long trains, which requires platforms are ~250 m long, compared with ~120 m for most Tube stations. As a result, the stations on Crossrail 1 were fabulously expensive: Bond Street came in at £660 mn, Paddington cost £685 mn and Whitechapel, which was originally budgeted to cost £110 mn, ended up costing £831 mn.
This is the heart of the issue with Crossrail 2. It is a hybrid between two schemes, each of which would be a good idea on its own terms, but when combined together they produce an overpriced, ineffective mash-up.
I put this down to what I call once-in-a-generation syndrome. British infrastructure planning is stop-start: there is no ‘pipeline’ of Tube lines that are going to be built. Each scheme is planned and funded in isolation, and very few of them end up getting built. This means every project that does get built is seen as a once-in-a-generation megaproject. We only get an opportunity every 20 years or so to build everything, which means each project must carry the burden of fixing more than its fair share of London’s transport problems.
This explains, more than anything else, why Crossrail 2 has so many bits tacked on: I don’t think the stops at Dalston, Angel and Chelsea wouldn’t be there if TfL could be confident that a Chelsea – Hackney Tube line were ever going to be built. Likewise, there would be no need to stop at Tooting/Balham if there were separate plans to relieve the Northern Line.
Crossrail 2 suffers from another defect: overbuilding. Good infrastructure should accomplish its purposes as effectively as possible, while pouring as little concrete as possible. From the perspective of designing the route, this means planning intelligently to re-use pre-existing or abandoned infrastructure where possible, preferring above-ground to underground, building smaller stations, and cutting out unnecessary stops.
A few examples of the overbuilding proposed for Crossrail 2. Wimbledon station, which has two spare platforms that could be used for Crossrail 2, is to be rebuilt with four new platforms east of the current station. There is unnecessary tunnelling from Wimbledon to Clapham Junction. The Dalston tail wags the dog, preventing Crossrail 2 from being routed via Finsbury Park, a more important interchange station. There are going to be new tunnelled stations in the Alexandra Palace / Wood Green area, even though there is an under-used railway line running nearby. At times, it feels as though Crossrail 2 took every opportunity they could to make the line cost more.
My proposal
The simplest way to improve Crossrail 2 is to reduce its scope to focus on its four core purposes: relieving the Victoria Line; relieving the Piccadilly Line; improving services in South-West London; and improving services on the West Anglia Mainline. I therefore propose the following.
This scheme will not try to do as many things as Crossrail 2. But it will be more effective at the things it does do, and, more importantly, it is much more likely to be built.
The next post explains in more detail what the route does, and how it does it:
