In Part 1, we introduced the many components that need to be integrated into the grand plan at Old Oak Common (OOC), for a new interchange with HS2, Crossrail, the Great Western Mainline (GWML) and Overground services. Here in Part 2 we take a step back to look at the implications and explore some nagging questions, which we’ll conclude in Part 3 along with a tentative stab at how the decision-making might unfold in the near future.

OOC has grown into a very large project indeed. The vision is for a ‘superhub’ and a gateway to the broader economic transformation planned for the Royal Oak Opportunity Area, which will transform this forgotten part of west London.

OOC has already thrown up some surprising issues and problems, the most pressing of which is the Crossrail depot. This sits in the middle of the proposed redevelopment, where it thwarts the creation of a harmonious design overall, depresses land values and substantially reduces the number of jobs that can be created.

The Mayor’s Old Oak: A Vision for the Future paints a bold picture, but it will need strident leadership to drive this project through. It’s not simply that OOC is a complex project, but that the fragmentation of the rail industry makes joining the dots nigh on impossible.

OOC is likely to require a more uncluttered approach, and the signs are pointing towards a Mayoral Development Corporation. Following the footsteps of Docklands and Olympic Park gives a sense of what OOC might become. But even with this, driving this project through will require strong will, holistic design, tenacious negotiation and lobbying, some luck, and the funding to get started. Let’s see what might come of it.


Another era: Old Oak Common (OOC) in 1938, looking west with the new Park Royal industrial estate behind. The GWML is bottom left, the GWR Birmingham mainline to Birmingham (the Banbury route) crosses the left to right, joined by the Central Line. The North London Line (NLL) traverses the middle, with Willesden Junction just to the right. Photo courtesy of Britain From Above

The Vision for OOC and a ‘superhub’

It’s worth reminding ourselves why TfL are so keen to make OOC a success. Their analysis of the demand implications of HS2 was already underway four years ago, and gathered pace with the consultation for the Mayor’s Transport Strategy. A proposal for an interchange at OOC had already been put forward in a submission by Parsons Brinkerhoff to HS2. Perhaps this was the seed of the idea ― who knows? But it all blossomed in 2011, when we excitedly reported on the fanfare of the Farrells Vision produced for the London Borough of Hammersmith & Fulham (LBHF).

Then all went quiet as HS2 and DfT took time to catch up, and it was well into 2012 before TfL and Network Rail were asked to do a study on what the new interchange might look like and how the various routes might be drawn in. Over thirty options were quickly whittled down to ten, and three of these made it to a more detailed analysis.

In the meantime, the Mayor set four conditions for his support for HS2, one of which is an interchange at OOC. The primary objective is to relieve pressure on Euston, where the Underground lines are already close to capacity. Even with the additional capacity provided through the upgrade programme, Euston will be overwhelmed by the second phase of HS2: the ‘Y network’ of further routes north.

OOC is designed to siphon off demand, and the projected passenger numbers are substantial. OOC is expected to serve 250,000 passengers a day, which is on a par with Waterloo. Indeed the planners expect around one third of HS2 passengers to interchange at OOC for Crossrail. Adding an Overground station will take another 10%, and will also reduce the HS2 passengers switching to Crossrail by 15%.

Yet OOC is not just being envisaged as a “super hub” but as a destination in itself, with 19,000 new homes and 90,000 jobs. This is a very different proposition, and is why the Mayor’s Old Oak Vision is challenging but achievable, laying down the ground rules for the HS2 station design:

1: Support the creation of a new destination of exceptional quality
2: Be of an outstanding architectural quality
3: Locate station accesses to facilitate regeneration
4: Build in the potential for development over the station
5: Link to the Strategic Road Network
6: Provide a state of the art intermodal interchange
7: Provide additional rail connectivity

Note that the Old Oak Vision says nothing about the design of the Crossrail or Overground stations. This is a positioning document targeted squarely at HS2, with a view to influencing the negotiations over the HS2 parliamentary process and, ideally, embedding into the HS2 legislation as much of the design and cost as possible.

The Old Oak Vision acknowledges that the Central and Bakerloo will not be connected to OOC, and Willesden Junction on the WCML lies just out of comfortable reach to the north. The strategy is to develop a “Green Cross” of attractive pedestrian routes N-S and E-W. The station would become the hub of these, with the aim that the design of the thoroughfares might encourage further interchange with Willesden Junction and North Acton.

Green Cross and hubs

Old Oak: A Vision for the Future, Fig 4.5

Attractive though this might appear on paper, in reality the distances of around 750m will deter all but the most avid of interchangers. Might there be other options to link OOC with these other stations, and beyond? For example, could a local tram network or similar focus on the OOC to Willesden Junction corridor provide a frequent service between the two that might make interchange more feasible? We explore this below.

The new Overground station

In Part 1, we outlined the three TfL options for a new Overground station at OOC. These provide a neat solution to a complex problem while still offering great flexibility in terms of modularity and future route options. The full scheme is shown again below.

TfL 2013 option 1

The complete scheme. Source: TfL presentation, Golden Mile Transport Group, January 2013

The process of reaching the preferred scheme was through a vetting process, starting with eleven options and quickly narrowing down to three, then assessing these in more detail. The criteria included the cost, engineering feasibility and implications in terms of construction phasing, operational impacts and the overall business case. Options that had an unreasonably adverse effect on the delivery and operation of Crossrail and HS2 were rejected.

The preference for a south-side Overground station versus a west-side station is not just down to cost and feasibility, but also depends on the combinations of routes that are supported and the viability of each of these. TfL analysis using their Railplan Model indicates that the WLL will take 30% of HS2 passengers alighting at OOC, therefore this is the priority Overground route. Yet, as we’ve seen on the Overground network as a whole, OOC might unlock a substantial latent demand on other routes that is difficult to predict or model accurately.

The planners have also been analysing the business case, claiming an attractive Benefit Cost Ratio (BCR) of 2.6:1 for the core scheme with the WLL and NLL, and a 2.7:1 BCR for an enhanced scheme which brings in the Hounslow and Brent routes. With similar outcomes in terms of BCR, the preferred option is likely to depend on what is achievable in terms of overall funding and engineering feasibility.

More detailed modelling should also show the effect on passenger volumes of the interchange to a west side, and would additionally factor in the benefits a west side station would bring in opening up the western development. The photo of Acton Wells Junction further below shows where the west-side station would be.

All rails lead to OOC

It’s worth stepping back to see how OOC redraws the west London rail map. OOC is now at the centre of a web of new routes, which TfL estimates will put 400,000 additional people within an hour of the new hub.

OOC routes

All rails lead to… Source: TfL presentation, Golden Mile Transport Group, January 2013

Not shown on the map above is the proposed Crossrail branch to the WCML, which we explore below. The map is also ambiguous about Southern’s service from Croydon to Milton Keynes. Will this go via Willesden junction or OOC, or indeed will it be part of the network a decade away?

The proposed Overground station at OOC will allow the Southern service from the WLL to weave its way onto the Dudding Hill route and from there onto the West Coast Mainline (WCML) via a freight link at Harlesden (see the CartoMetro map). However this is a sinuous single track and significant work would be required to improve line speeds, as well as ease connection onto the WCML Slow lines around Wembley Central. Nevertheless, this is one of the two route options being proposed for the Crossrail WCML branch and, if adopted, could also be used by Southern or by a new Overground route.

Given the multiple route options at OOC, which are TfL likely to prioritise? The WLL route to Clapham Junction and beyond is likely to generate the most demand, but which route north from OOC makes most sense in business and operational terms?

Consider, for example, the full Overground scheme proposed through a south-side station. If one took an aspirational look ahead 20 years, at a 6tph frequency on all routes and with GOBLIN extended, then the OOC Overground platforms could be handing 18tph in each direction:

  • GOBLIN-Hounslow
  • NLL-Richmond
  • WLL-Brent and the Midland Mainline (MML)

Extending the ELL through Primrose Hill may also be considered worthwhile, as the broader Park Royal development will also make Willesden Junction a significant interchange.

What of the orbital routes?

The design raises several questions about the connections and alignments beyond OOC, and also the impact the increased traffic will have on the orbitals.

The proposed connection from the south-side Overground platforms to the WLL is grade-separated at North Pole Junction, which brings benefits in terms of capacity and operational reliability. However, as we explored in our articles on rail freight, the capacity of the WLL is restricted by the dearth of paths and the mixed traffic with long slow-moving freight trains. Plus, the redevelopment at Earls Court will cocoon the WLL into two tracks which curtails any option to maximise capacity on a wider alignment.

At the west end of OOC, crossing over the GWML is one of the sleepy backwaters of the London rail network: Acton Wells Junction on the NLL. Yet this hub of the various orbital routes could become more like Borough Market Junction, and the new Overground design pushes every train through here. Alongside the growing freight traffic, Acton Wells Junction may become a pinch-point which would severely limit the extent to which Overground services can intensify.

The photo below shows Acton Wells Jn in 1988, looking north, and the location of the proposed west-side OOC Overground station. The Dudding Hill line heads left to the MML, and the lines to the right go to Willesden Jn. The right pair of these is the NLL for the high-level platforms, and the left pair is the freight link to the low-level and from there to the WCML or looping back onto the WLL.

Acton Wells Jn sm

Sleepy Acton Wells Jn in 1988, photo courtesy of David A Wingham on Flickr.

The potential route south-west from OOC to the Hounslow Loop is an interesting proposition. Indeed, as we note below, the Hounslow Loop was one of the options assessed for a Crossrail branch. The junctions would have been relatively easy to provide at OOC, and still would be as part of the current Vision.

However this route will require significant investment including electrification, and capacity is again limited by the need to share with freight. As the WLL comes under pressure, options may be explored to shift freight onto the Kew to Acton Wells route. Combined with the existing NLL service to Richmond, this will limit the potential to provide an intensive Overground service.

Worse, the route has level crossings at Acton Central and Bollo Lane just shy of South Acton and, short of tunnelling the entire route, it is difficult to see how these can be alleviated. More than anything else it is this that prevents the Hounslow Loop becoming part of Crossrail.

Clearly, if OOC becomes the visionary success the promoters are trumpeting, it will put pressure on the orbital routes surrounding. Sadly, the lack of an integrated investment strategy for these might limit the full potential of OOC as an interchange, and perhaps weaken the business case.

Crossrail to the WCML

We posed a question in Part 1 – what would Old Oak Common be like without a Crossrail depot? This would transform the phasing and business case of the development, on which the viability of the interchange depends.

If Crossrail was extended west and onto the WCML, new depots could be provided and much of the land at OOC would be freed up. An alternative option is to redesign the depot so development above is possible, but this would still require a significant change to the Crossrail specification, programme and budget. Alas it is very late in the day for this and, in the absence of a clear new specification, Crossrail proceeds relentlessly as planned.

The Crossrail and GWML station will have eight platform faces. Crossrail will have two central platforms for reversing trains, with the outer lines continuing west on to the GWML Relief lines. This design allows for this central pair of lines to be extended westwards into two reversing roads, which can be discerned at close view on the annotated HS2 map below.

The aim is that passive provision will be made for these reversing roads to become the Crossrail WCML branch, at an estimated cost of £25 million. It is hoped that this will be integrated into the overall design presented in the HS2 hybrid bill later this year.

Eagle-eyed LR readers will no doubt observe the challenges in threading the WCML Crossrail branch through the grade-separated junction, above Old Oak Common Lane and then dropping beneath the NLL bridge at Acton Wells Junction. The gradients will be steep, especially as the WCML route then needs to drop further down into tunnel.

The HS2 plans for a new overbridge for Old Oak Common Lane are necessary to accommodate the double-decker buses needed to provide capacity on new routes into the rail interchange. We wonder whether a preferable solution is for a new road flyover rising over the entire GWML and North Pole depot, which would allow the Crossrail WCML branch to drop beneath the Up Crossrail GWML track, and then easily into tunnel.

As an aside, such safeguarding is not entirely unknown on the modern railway. The Thameslink tunnels from St Pancras up to the East Coast Mainline (ECML) at Copenhagen Junction are only now being populated with rails, with passenger services due to use them from 2018.


OOC west end showing the Crossrail WCML branch, and a potential link to the WLL discussed below. Derived from HS2 map C221-MMD-CV-DPP-010-201701.

It’s worth reminding ourselves that the Crossrail WCML option was subjected to an economic appraisal in the 2011 London & South East RUS, which found it had a favourable BCR of between 1.8 and 1.6 as a standalone scheme, rising to 2.6 to 2.2 if combined with HS2. The RUS recommended that the scheme go forward, and if not then that passive provision be made in the OOC area.

In terms of the route to the WCML, NR scoped over thirty options in pre-OOC days. These included alignments across the OOC site and the Kensal Cemetery just to the east, which would have rumbled unnervingly beneath Brunel’s grave. These alignments do not settle happily in our new and rapidly evolving ‘super-hub’ world, so we’re now down to four options. What Brunel might have made of all this is perhaps for another post, but the options are essentially variations on two broad themes:

  1. A sharp turn north to tunnel beneath the Dudding Hill alignment, and then weaving onto the WCML around Harlesden
  2. Constructing a longer route to the WCML at a point closer to Wembley Central, through the middle of the Park Royal Opportunity Area. This might offer broader benefits to the Park Royal redevelopment, e.g. the potential to provide a new station serving the putative “neighbourhood centre” around the Central Middlesex Hospital.

The first option entails a connection to the existing freight lines west of Willesden Junction which then burrow from the south side of the WCML to the north west of Harlesden, before traversing the flat junctions onto the WCML Slow lines at Wembley Central (again, see the CartoMetro map). Passengers on Southern’s Croydon to Milton Keynes service will be familiar with the frustrating crawl along this section.

A major upgrade will be required to serve an intensive metro service. Grade-separated connection to the WCML would also be preferable, to maximise capacity and improve reliability. So, although shorter, this option may not be significantly cheaper and is less attractive in terms of journey times or in serving new markets. The longer route option may be preferable, especially if it can be built as part of the Park Royal redevelopment, and a new station provided.

A Crossrail branch onto the WCML would require new depot facilities and, given the intractable problem of the depot at OOC, this could become a strong driver to prioritise the new branch. Network Rail initially estimated the investment cost of the WCML branch at £198m to £248m (2002 prices, from London & SE RUS Option K1), but clearly much more work is required before this project can be put forward for investment. If passive provision is made, that grade-separated junction at OOC might remain unused for some time.

Note that the Crossrail WCML will serve only 8tph, and there are no plans to maximise the return from the new route by sharing with other services. This relative under-utilisation of capacity may weaken the business case, especially for the longer and more expensive route option.

Returning to the map above, we have sketched in another option which surfaced in the discussion following Part 1. Our thanks go to Mark Townend, who has also drawn up the map below (see PDF here).

Townend OOC sm

Some OOC options. For simplicity, the TfL scheme in the Mayor’s OOC Vision is not shown, although this could be integrated. Map courtesy of Mark Townend.

This integrates the WCML branch with an Overground route from the WLL. WLL services would still use the new platforms on the south side at OOC, but these would be at surface level rather than on a viaduct. The route would then descend to pass under the GWML and adopt the abandoned ex-Great Western goods route from the WLL. This little-known link ran alongside the Central Line through East Acton and North Acton. The underpass beneath the GWML would be disruptive to construct, but the GWML is going to be widened here anyway for Crossrail. Economies of scale could be achieved by combining with the works for the new Old Oak Common Lane overbridge.

It’s a design that comes from considering Crossrail and Overground together, and perhaps indicates the synergies that are possible when the institutional context drives an integrated design.

Also included on Mark Townend’s map is a relatively straightforward option for an Overground route from the WLL via new low-level platforms at Willesden Junction, to terminate at the new OOC west-end station. The photo of Acton Wells Junction above is looking directly at this location and shows that there is plentiful space for a bay platform alongside the freight lines.

Clearly this is not an ideal solution, as the journey time from the WLL would be longer and the interchange at OOC less convenient. But there are other benefits, not least that it would be relatively inexpensive, and achievable if funding for the grand scheme faltered. As part of a long-term plan it could, intriguingly, be extended on to Acton Mainline and perhaps Ealing Broadway.


North Acton with the ex-GWR mainline to Birmingham on the left. The Crossrail WCML branch would drop into tunnel here, potentially joined by an Overground route from the WLL. The NLL and Acton Wells Jn lie just beyond the road bridge. Photo courtesy of David Hawgood.

The key advantages of the TfL scheme are that it offers multiple route options and that the Overground construction would be largely segregated from the works on the mainline, HS2 and the development itself. It is a neat solution in itself, but it focuses on one component rather than the whole.

Combining Crossrail and Overground will derive additional benefit from the considerable investment in the new WCML branch. Indeed, the option above comes from considering Crossrail and Overground together, and perhaps indicates the synergies that are possible when the institutional context drives an integrated design.

The key benefits are that it provides a more direct route and avoids Acton Wells Junction, which will be advantageous as orbital and freight traffic increases. It can also be combined with the current Overground design, which would still provide links up to the NLL in both directions and with the Dudding Hill route.

Alternative Crossrail routes west

With all the other demands on the TfL purse, it may be that a Crossrail branch to the WCML takes decades to materialise. Are there alternatives, especially if these could be delivered quickly and cheaply? A significant driver for this might be to identify a site for a new depot, which would free up the space at OOC and give a significant boost to the entire redevelopment.

It is hard not to be drawn to the GW Banbury route through Park Royal, or be tempted to revive the proposal for a branch to Hounslow. Both offer ease of access at OOC and a ready-made route, and it’s likely that both could offer sites further west for a depot.

The Banbury route was initially discounted because projected passenger demand would be significantly less than the WCML, plus the Central Line runs alongside the route to Ruislip, which essentially duplicates the route through central London. But although it might provide a solution to the projected capacity problems at Marylebone, the lack of electrification ultimately scuppers this as a viable option.

The Hounslow option would help relieve overcrowding on the Windsor Lines, and improve north-south connections in west London. But, as noted above, the main drawbacks with this route are the level crossings around Acton and the need to share with freight and the existing NLL service to Richmond. Arguably, TfL will need to resolve these problems anyway for an Overground service, and perhaps the investment could be shared could be shared across a number of beneficiaries. But a Hounslow option would be neither straightforward nor cheap.

Crossrail and the GWML east to Royal Oak

Looking east from OOC, Crossrail takes up the northern half of the alignment to the Royal Oak portals, leaving the remainder for the GWML into Paddington. Given the intensity of service on both Crossrail and the GWML, it’s likely that the track layouts east of OOC will need to be redesigned.

This redesign is likely to add to the cost and complexity of extending Crossrail services westwards from Royal Oak. Intriguingly, it might also provide an opportunity for the Royal Borough of Kensington & Chelsea (RBKC) to revive its failed bid for a Crossrail station.

Western at Mitre Bridge

Looking east in 1975 from OOC beneath the WLL bridge with the canal penned in above, behind the retaining wall. The alignment east from here will be (from left to right): Crossrail depot tracks, Crossrail and GWML. Photo courtesy of Martin Addision.

The key issue is that the Crossrail service beneath London is headway-driven, whereas the service west of the Royal Oak portal is timetable-driven. The section from Royal Oak to OOC forms the boundary between the two, where services are regulated. A more intensive service over this two-track section will make the whole route prone to disruption. It will almost certainly need to be quadrupled. Four-tracking from Kensal will also allow WCML and GWML services to be sorted, and this provides a buffer against disruptions on the alternative routes west.

The track layout below is derived from the Context Reports produced in 2009 by Crossrail for each local authority, and then checked where possible against the HS2 plans. Among other features this shows a new ramp joining the existing GWML empty stock flyover just east of OOC. This new ramp gives a clear route to the Crossrail depot independent of the GWML depot lines, and the plans show it being built across the finger of land at the west end of the ex-Kensal Green Gas Works development site. It will be double track, and the alignment on the incline back down towards Old Oak Common will be widened so that the GWML empty stock has a segregated track down from the flyover.

Ramp small

The new ramp east of OOC. Map taken from Crossrail 2009 Context Reports for LBHF and RBKC

This raises several questions, one of which is whether the existing flyover will have a future role. Once the new IEP depot opens at North Pole on the south side at OOC, and the First Great Western and Heathrow Express depots are swallowed up by HS2, the flyover presumably becomes redundant. Perhaps it could be used by a local light rail scheme or similar, but why is provision being made for a link from the Crossrail depot tracks onto the GWML south side towards Paddington?

Another question is whether the junctions on the critical section east from OOC will need to be grade-separated. If so, the new Crossrail depot line will need to be redesigned with a flyover instead of the proposed ramp, almost certainly with the depot tracks dropping down between the Up and Down Crossrail lines.

The new layout will also have to reconsider the flat junctions around Westbourne Park to the aggregates facility below the Westway, and the connections into the terminal platforms at Paddington. The diagram below is our interpretation of how it might look.

OOC layout east final

The new layout

Crossrail has advised that it will still need to build emergency reversing facilities close to Paddington to safeguard operations through the core in the event of service disruption. However these would not be used regularly as it sees no difficulty in running a full service west as far as OOC.

The plan above also retains the connection into the Paddington terminal platforms. We assume the purpose was to allow additional peak services from the GWML Relief lines to run into Paddington terminal, but with all Relief line services now using Crossrail this no longer applies.

Will a four-track Crossrail formation allow the RBKC to revive its dashed dream for a Portobello Crossrail station? Alas no, and we explore why in Part 3. However a Crossrail station at Westbourne Park would be more straightforward, involving the building of platforms on the mainlines for interchange with the Hammersmith & City.

In Part 3 we look first in more detail at the options for RBKC, the IEP depot and other potential rail connections into the OOC interchange, and then present some thoughts on how the decision-making might unfold.

And, chomping at the bit though we’re sure you are, you’ll need to head over to Part 3 to comment: we’ve disabled comments here so we can focus the rich discussion on the one post.

Written by Lemmo