Euston vs Old Oak
One controversial project, two contenders for the title of its London terminus. WILLIAM BARTER cuts through the politics and compares their technical merits
Independent consultant WILLIAM BARTER considers the merits of Euston and Old Oak Common as London’s HS2 terminus.
Aconstant theme of opponents of
HS2 is that the terminus of the new line should be at Old Oak Common rather than Euston. On the other hand, some supporters looking for an economy suggest Old Oak could be dropped completely. No doubt the last few miles into Euston will be expensive, so why are we planning two London stations anyway?
What is Old Oak Common for? The key reason for having a station there at all is the interchange with Crossrail, which will take passengers off HS2 onwards to the City and Docklands - now slow and crowded journeys from Euston. Going west from Old Oak to Heathrow is an opportunity that may suit the briefcase brigade, but probably not those with heavy luggage. And from a catchment area roughly east of Reading, changing at Old Oak to go North will be an improvement over CrossCountry services. By contrast, Euston caters for the immediate central London area, and locations on a north-south axis.
In terms of track layout, there will be six platforms for HS2, and eight for the Great Western Main Line - two each way for each of the Fast and relief (Crossrail) lines. So, for each pair of tracks, and assuming all trains stop, a continuous flow of trains is possible, alternating between platform faces, without extending the plain line headway.
On HS2, the track layout allows for turnbacks towards the country from all platforms. This is intended for emergencies - if the tunnel to Euston is blocked, for example, but all proposed trackwork is on the flat, and for the two outside platforms reversal means a long wrong-direction run to crossovers just outside the portal of the Northolt Tunnel. As the normal service could run on two platforms each way, the centre platforms once intended for international trains are available for contingencies - such as working round a train with problems, or to turn one back if running seriously late.
For Euston, it is possible to get some idea of the proposed station layout from ‘Plans and Profiles’, published during the Phase 1 hybrid Bill. Figure 1 (see page 60) is my interpretation of the track layout from that source.
The key feature is the 11 platforms, laid out essentially as two half-stations of six platforms each, with one platform common to both. The two half-stations are approached by way of a grade-separation, so that arriving trains can alternate between the two, taking different routes at the first opportunity, minimising the risk that one train slowing potentially delays the next, especially if at least a 200-metre train can wait for its platform without standing foul of the ‘King’ points that split trains between half-stations.
The common platform can be entered and exited from either side of the throat, so it looks like a very handy thing when trains are running out of course. Note that, together with its grade-separated access, it falls into the Phase 1/2A station, and having one platform semi-independent of the other five would allow some departure from rigid platforming patterns, as well as improving robustness and offering some choice of handover times for paths extending onto Network Rail.
For planned operation, some platforms in each half-station share a common lead-in, thus ruling out simultaneous arrivals and departures (‘parallel moves’) at some adjacent platforms. However, the grade-separation means that arrivals and departures that would otherwise conflict can be allocated to different half-stations. And for maintenance of pointwork, a six-platform service can be operated whichever throat is under possession.
It is only the Phase 1/2A station that takes land outside the existing station footprint. The other five of the 11 HS2 platforms can be built within the existing footprint, as the conventional station will be dealing mainly with suburban and outer-suburban trains taking short turnrounds and needing fewer platforms for the same frequency.
All in all, I’d call HS2’s plan for Euston a little miracle of flexibility and robustness in the space, probably - and quite rightly - inspired by terminals south of the Thames operating intensive services, such as Charing
Cross or Cannon Street. But apart from the pure joy of watching trains performing this railway ballet, why have Euston at all?
In essence, it’s where most people want to go. The expectation is that about one-third of passengers will choose to interchange at Old Oak, while two-thirds will prefer to travel on to Euston. Tottenham Court Road is served pretty much as well either way - personally, I would hop off at Old Oak in the morning to use Crossrail, but return via the Northern Line and Euston in the evening so as to have a rather more leisurely boarding experience.
While the Euston tunnels are expensive, it is reckoned officially that the lost benefits of not providing them would outweigh their extra cost. Remember that throughout the 19th century, railways built to stop short of central London had to be expensively extended, for instance to Waterloo from Nine Elms, to Liverpool Street from Bishopsgate, to Charing Cross and Cannon Street from London Bridge.
Euston also has a much wider range of options for onward travel - two Northern
Line branches, the Victoria Line, and the sub-surface lines at Euston Square which will become integrated with the new Euston station and multiple bus routes - as well as taxis, which will be important for those with heavy luggage or mobility issues, whose ride from Old Oak would be slow and expensive. While Euston is separated from international services and Thameslink at St Pancras, it is possible that the Euston/St Pancras Crossrail 2 station will provide a direct walkway shorter than those found for interchange at most airports.
Old Oak has some other rail connections within a kilometre or so, and ideas to improve on this exist but, without Euston, HS2 would be essentially dependent on Crossrail, to the extent that a blockage on Crossrail would probably mean HS2 being suspended.
As a through station, Old Oak uses relatively little space, adequate for 18 through trains per hour each way and taking short dwell times - this sounds ambitious, but it boils down to about ten passengers per doorway within whatever portion of a two-minute dwell is left for passengers after the door operation itself. That shouldn’t be a problem - so long as the seat allocation system plays ball and doesn’t book everyone for Old Oak into the same coach!
But as a terminus for the full service, an
Old Oak terminus would have to replicate the functionality of Euston - it could be no smaller than Euston is planned to be, and equally in need of the grade-separated throat. Not only would this demand greater land-take than
Why have Euston at all? In essence, it’s where most people want to go.
has been allocated in the Phase 1 Act, but greater land-take than at Euston as the whole station would be on new land. So only the tunnels would be avoided - but as the spoil for the Euston tunnels is to be brought out at the Old Oak end, if they are ever to be built, they must be built before Old Oak becomes a working station.
Given all that, why have Old Oak at all?
The reason is largely to shift demand onto Crossrail and relieve Underground lines at Euston. This becomes critical with Phase 2, when passengers who currently arrive at King’s Cross from Edinburgh, Newcastle and Leeds will be diverted to Euston. Also, the
Old Oak area will grow as a destination in its own right as it is developed. And even if there were no station for the planned service, the risk of blockage of the tunnels, or nonavailability of Euston in an emergency, means that you would probably want an emergency de-training facility there, so you might as well have a proper station anyway.
But… since Oakervee, the phrase of the month has been ‘use Old Oak Common as a temporary terminus’, so as to allow time to ‘get Euston right’. This is starting to worry me, especially in view of the apparent lack of any serious operating input to the Oakervee review (apart, that is, for some seriously misleading advice from Lord Berkeley!).
First, how temporary is temporary? The
Evening Standard suggests ten years, pretty much the whole duration of the Phase 1/2A service before Phase 2B (now ‘HS2 North’) kicks in.
Let’s take it as read that Old Oak could not be the terminus for the full service of 17 trains per hour plus one path left unallocated for future growth. Apart from the commercial madness of tipping travellers out of their trains in Zone 2, TfL says Crossrail could not carry the full passenger load. Oakervee has talked about planning for just 14 tph - this does not allow for any significant saving at the terminus, unless one accepts that at such a crucial location, as the trunk route is capable of presenting 18 tph to the station, a constraint will be built in that will hamstring the capability of the railway forever. I really hope that no-one in any position of responsibility accepts that! Oakervee then recommends ‘passive provision’ for 16 tph, but what can that mean other than providing the gradeseparation?
How about just the Phase 1/2A service of 10 tph? Six platforms is six platforms, right? Wrong. Because a terminus isn’t just platforms, but also a throat, and the Old Oak throat cannot be as flexible as Euston, as Euston has that grade-separated access to one platform. Choice of timings would be more limited, so HS2 trains would be handed over to Network Rail when it suits HS2 and not when it suits the multitude of constraints on the rest of the railway. An example arises at Crewe where, by a happy coincidence of service interval, running times and turnround times, it looks as if Up and Down HS2 Manchester trains will naturally fall in parallel at the crucial Crewe North Junction - a train planner’s dream! But if starting closer to Manchester from Old Oak, this happy coincidence vanishes.
In addition, critical pointwork for some platforms is further from the station, increasing the ‘reoccupation time’ - that is, the minimum interval between departure of one train and arrival of the next. If this does not actually rule out running 10 tph, it would certainly jeopardise their reliability.
A further issue is that with Old Oak as well as Euston, if needs be, a seriously late Up train can be terminated at Old Oak and sent back north from there to make up time - but if
Old Oak is the only terminus, any such train must either be accepted there regardless of the impact on others, or cancelled out completely.
Altering the Old Oak layout now, to make up for these deficiencies as a terminus means either taking extra land, thus needing new parliamentary powers, or reconfiguring it more appropriately as a terminus at the expense of its ultimate function of feeding trains through to Euston as rapidly as possible.
There’s a further issue, as to just how acceptable HS2 trains terminating at Old
Oak would be as a substitute for West Coast trains terminating at Euston. HS2 would still offer some advantage in pure journey time, but putting in an unnecessary interchange for most passengers, in ‘perception’ terms, wipes this out. This has two implications:
Until most existing West Coast passengers into Euston - and thus their trains - are
transferred to HS2, the capacity they release cannot be reallocated to local and commuter services - something that is a key, albeit probably underestimated, benefit of HS2 (see diagram opposite).
And, until Euston is turned into an outer suburban sort of terminus with frequent trains on short turnrounds, it can’t be trimmed back to operate off 13 platforms and thus release the site of the present platforms 14 to 18 for building the other five HS2 platforms.
So, either a pretty non-commercial operation has to be accepted until Euston is ready, or delay in getting Euston into running order knocks back completion of work for Phase 2B.
What if it’s less than ten years, though?
What if Euston becomes available for the
Phase 1/2A service after only, say, three years of operation at Old Oak? Still not good. Timetables based on Euston and Old Oak may have similar numbers of trains, but in their structure and detail they will be completely different, especially on Network Rail. So that means two major timetable changes within three years - apart from the commercial disadvantage of repeated radical upheaval of the service on the UK’s prime trunk route, the magnitude of the task would stretch the national pool of train planning resources, and risk making May 2018 look like a walk in the park.
As I see it, the only reasonable ‘temporary’ use of Old Oak would be a self-contained shuttle to and from Curzon Street, and nothing should be allowed to distract from getting Euston up and running for a proper Phase 1/2A service. Stations exist first and foremost to run trains and get passengers on and off them, so if anything to do with over-site development or rebuilding of the conventional station risks delaying the first six platforms coming into use as soon as they and the throat can be built, that would be very wrong indeed!
Are there sensible economies that could be made, however? The grade-separation at Euston is expensive - could it all be done on the flat? After all, some stations manage to run 18 tph on a flat layout (and in the case of metros, many more than that!) But on inspection, all precedents have some feature than one could not replicate at Euston for
HS2, or would not want to. The Slow lines into Charing Cross, pre-Thameslink, used to carry 18 tph on three platforms, but with short turnrounds and ten-coach trains the reoccupation time was short - HS2 is quite a different kettle of fish, as longer turnrounds means more platforms means more pointwork means a longer throat, which means longer reoccupation times, especially with 400m trains. Paddington has the benefit of multiple track to Ladbroke Grove, where a second chance to sort trains out presents itself; Fenchurch Street with quadruple track from Christian St Junction into the station is similar, although a glance at the Working Timetable shows that nearly every train has to take pathing time between the two. Even Oakervee accepts that the maximum a flat layout for HS2 could handle is 14 tph; but even that seems optimistic, and, quite apart from the madness of building in a capacity constraint that can never be eased, I doubt if it would operate as reliably as 18 tph on the grade-separated layout.
One might query, however, the scale of Old Oak Common station on the GWML. The
Fast line platforms are there primarily to allow Heathrow Express to call. But once Crossrail, albeit slower, takes people right into the West End, City and Docklands, what is the future of Heathrow Express? And if it has one, does calling at Old Oak Common justify the four Fast line platform faces? Maybe this is where ‘passive provision’ actually makes some sense.
One loose end that does strike me is the conventional station. Having said that it will be dealing mainly with suburban and outer suburban services with the longest distances run being Chester and Manchester, 13 platforms is OK. But once again, there is more to a station than platforms, and a throat specified for a (relatively) infrequent service of (relatively) long-distance trains may not support the intensified usage that comes from running nearly the same number of trains on fewer platforms. I remain to be convinced that some additional pointwork need not be added to support the new service.
‘Getting Euston right’ sounds good. But Euston is the right terminus for HS2, and in its operating functionality, Euston as planned for HS2 is ‘right’ already. Let’s make it reality.
The only reasonable ‘temporary’ use of Old Oak would be a selfcontained shuttle to and from Curzon Street