High Speed Buffers? HS2 in London (Part 2)


In this part of our series on HS2, we look at options for serving Heathrow Airport, how this might be affected by franchising, the London terminus and decisions made concerning a potential HS1-HS2 link.

Access to Heathrow Airport and implications for HS2 service options

Midlands and Northern stakeholders expressed a desire to be able to reach Heathrow via HS2 in order to open up fast access to this major international hub. The Heathrow Hub campaign has argued, so far without success, for the main HS2 line to be routed via the airport on its way to London. Heathrow Hub has also argued that use of the HS2 main line within London could allow some European rail expresses to start and terminate at Heathrow rather than at St Pancras. It has alternatively argued for widening the GWML to six tracks as far as Iver, with people movers serving a new hub station there and with European rail expresses reaching the airport that way.

HS2 links to Heathrow

Proposed direct HS2 link to Heathrow

The Government, HS2 and Heathrow

The Labour Government commissioned a report in March 2010 from Lord Mawhinney (a former Conservative Transport Secretary). He covered four topics:

  • High Speed to Heathrow
  • A London High Speed terminus
  • Airline slot allocation
  • A link to HS1 and mainland Europe

This was rather more than he was asked to do.

Lord Mawhinney’s mandate was renewed by the Coalition Government in May 2010, despite this change of government and a change of policy at that time concerning the third runway at Heathrow. He reported in July 2010 and, in relation to Heathrow, recommended that:

  • In the early stages of a high speed rail network, there is no compelling case for a direct high speed rail link to Heathrow. An Old Oak Common interchange is adequate.
  • Changing the route of the main high speed line to run via Heathrow, at an extra £2-4bn, should not be taken forward.
  • Rail/air through ticketing should be an integral part of any new high speed rail link to Heathrow
  • With the high speed line from Old Oak Common to Birmingham, appropriate junction engineering works should be included to allow an eventual airport link. A direct line would only be in prospect after the high speed network had been extended at least to Manchester and Leeds
  • A station at Heathrow Central Terminal Area be favoured for maximum connectivity

The report’s conclusions were adopted by the Coalition Government, who also said that Heathrow Airport should make a funding contribution towards the spur railway.

HS2 Ltd proposed a spur from near Denham, generally alongside the M25, to a terminal near Heathrow T5 (see above map). In one future option, that could have been extended towards the South West Main Line as part of a Southern Rail Access for Heathrow. Such a scheme would offer high speed cross-country intercity trains from places such as Southampton. The high speed spur would have had two junctions with the HS2 main line, one towards the Midlands and North, and one towards London. The latter would clearly have had an onward link in mind with HS1, discussed below.

Plan to get 4tph to Heathrow using only 2tph slots

As we shall see, the HS2-HS1 link has been cancelled, making the London-Heathrow spur partly irrelevant for the next couple of decades. The Heathrow-northbound spur was then questionable, though both still had potential relevance for any decision on an extra South East airport runway. On Government instructions in January 2013, planning for the spurs and Heathrow rail link was deferred until after the Airport Commission’s announcement which was due in 2015. High speed passengers to and from Heathrow will have to travel via Old Oak Common – as proposed by Lord Mawhinney.

In terms of service planning, the HS2 main line is seen as having slots for 18tph each way, running at consistent speeds to maximise hourly capacity. 18tph is equivalent to a train every 3.3 minutes, or every 3 minutes with a ‘white space’ every half-hour to allow a recovery and punctuality margin. Of those, HS2 Ltd had foreseen 16tph between London and the Midlands/North/Scotland, and slots for 2tph between Heathrow and the Midlands/North/Scotland. That in turn would have allowed up to 2tph on the Heathrow-London curve, with a Heathrow arrival from the Continent using the same HS2 main line slot as a Heathrow departure for the North, and vice versa.

No future provision for Heathrow-London link

The forecast passenger volume to and from Heathrow was itself quite low. This was in contrast to any high speed cross-country linkage which had been backed in 2010 in a report by the pro High Speed Greengauge21 organisation. Only 2tph were considered justifiable, making the whole High Speed spur poor value for money. In May 2013, after the shelving of the HS2-Heathrow link, Greengauge 21 suggested using NNML and the HS2 spur from the London direction, for other rail services to reach Heathrow. However, that is no longer an option. Whilst the alignment might be useful to note as a possible future corridor for orbital London & Home Counties travel and/or for London-avoiding trains, the spurs are not being safeguarded. The Commons Select Committee has explicitly directed the Promoter not to use the Bill powers to implement passive provision for the Heathrow spurs. The Government accepted this. The decision releases two additional slots per hour for internal UK high speed travel to and from London, which may be much more worthwhile commercially.

It is unlikely that any High Speed spurs will be revived in the foreseeable future unless the Scottish Government insist on a link to allow Scotland-Heathrow High Speed trains as an environmental alternative to planes. These are not in the High Speed Scotland and Northern England Broad Options recently published (March 2016) by the DfT.

Franchising and open access

As a corollary to the abandonment of the slots into Heathrow, might those two (or more) HS2 slots per hour become available for open access trains? No one has so far mentioned the availability of HS2 to accommodate open access services, but legally the possibility must exist. High speed line capacity will not be used fully in Phase 1. Indeed no one has so far indicated how the HS2 services might be structured within or separately from the existing WCML franchise as a commercial operating package, or as a ‘super-express’ element of the established West Coast and East Coast franchises.

Nor has pricing for route access and train paths been debated much, neither for ‘captive’ services between the main city-centre hubs nor for through expresses continuing onto the classic network. These are rather fundamental questions which are still open for large-scale debate. HS2 Ltd has proposed a draft group of HS2 Phase 1 services, along with indicative WCML replacement services, while the Government has proposed HS pricing should have parity with WCML, but that is as far as it goes at present.

The latest DfT demand case analyses for HS2 were published in November 2015, and show the proposed Phase 1 service patterns. These do not include the Watford-Euston DC trains, which TfL aims to run at 4tph in future years.

Indicative services after phase 1

Indicative plan of the services on the WCML and HS2 following phase 1, as presented by HS2 Ltd to the Commons Select Committee

Future Phase 1 services

Broadly the current WCML intercity frequencies are retained and transferred to HS2 (with Liverpool rising to 2tph, at present in peaks only), except for the hourly North Wales diesel service retained on WCML (of course, North Wales is arguing for line electrification). With an approximate 50 minute run time between London and Birmingham and 3tph, an HS2 ‘captive’ train ought to be able to ‘cycle’ every 140 minutes, 160 minutes at worst. HS2 is planning on a standard 25 minute turnround at Euston except for the longest distance trains. This layover seems excessive for a Birmingham shuttle, but would point to Train 1 resuming at Euston as Train 9 160 minutes later, so 8 trains on that service.

As HS2 is proposing a Phase 1 order for 16 ‘captives’, this also points to the bulk of Birmingham trains being planned with the ability for expansion to double-length (400m) for higher capacity, with higher running costs but no more train slots and more seats per train. Consequently HS2 train operating costs per seat mile might not be so startlingly different to now, though energy consumption will be significantly higher. Of course, you then have to fill those seats…

It is the re-growth of WCML services where the revised train frequencies arise: 2tph retained ‘intershire’ services (our phrase) from Manchester, another two retained from Birmingham, an extra train from Crewe plus two hourly spare paths. The outer and inner commuter services from Northampton and from Milton Keynes or closer in would maintain their frequency (6tph Northampton, 8tph Milton Keynes and closer) and these would move to 12-car trains over time.

Plus ça change…

The regrowth of services on existing lines amounts to a ‘reborn’ WCML passenger service structure remarkably similar to now, with just a few more intermediate stops and lavish provision at Milton Keynes. The Watford outbound offer is unclear, while the Phase 2 Luton/Bedford and Stevenage equivalents aren’t discussed. There is nothing to be seen in the way of new WCML service origins outside London (Super Voyager or bi-mode from Hereford/Shrewsbury, anyone? Yet bi-modes are now being considered for HS1, for St Pancras-Hastings).

There is no variation offered for the WCML London destination, no stop at Willesden Junction for the Old Oak Common development lands from intermediate commuting towns in the absence of WCML-Crossrail (similar to the GE stop at Stratford for Canary Wharf), let alone any Javelin-style commuter trains to Stratford via Camden Junction and HS1. The more one looks, the less the WCML capacity to/from London amounts to ‘new’ connectivity. There are however additional regional passenger services and extra freight paths that can then be operated, which are discussed in the DfT November 2015 documents.

Status quo, or more Midlands commuting?

Many intermediate places will, of course, still expect their direct services to London and other places. It’s the need to maintain these that eats up the available paths. The understanding is that Network Rail’s WCML development team currently consider HS2 will only free up passenger train slots on WCML for, at best, an additional 2tph, though there will be more capacity on trains at the intermediate stops.

It is worth noting that, if we had expected the ‘classic’ London commuter frequencies to increase and/or do more diverse things in London, this doesn’t appear in HS2 Ltd’s thinking to date. The Northamptons and inners are shown as static in trains per hour (though with greater capacity) – so we didn’t need HS2 for those. The implication of the service pattern is that HS2 fundamentally enables priority for future passenger growth on WCML ‘intershire’ services – which could be translated as more longer distance commuting to London, and even from London in a contra-flow direction (except such contra-flow isn’t obvious from the Watford stops). At this stage, the DfT has not intervened and proposed any changes to HS2’s ideas, and has gone as far as republishing them in its November 2015 update to the strategic case for HS2.

New costs

How much such services will reward the combined capital and operational expenditure for a revised WCML plus HS2 Phase 1 is an unanswered question. To require roundly £20bn for capital expenditure for the Phase 1 line and trains at 2011 prices leads to an annual capital interest charge of £600m (equivalent at, say, a public sector 3% pa borrowing) before any capital payback. As with Eurotunnel, the capitalised interest before the Phase 1 line opened would also be large, at crudely £2¾-3bn, with a further £80-90m interest chargeable annually. If, as Treasury seems willing to contemplate, the private sector were invited to finance HS2, consider a (low) 12% annual charge with capital interest around £2.4bn just for HS2 Phase 1, before any repayment of the original capital. The figures would be higher too when expressed in 2026 prices.

Taking additional WCML operating costs as a nominal £25 per train mile (also 2011 prices) for an amalgam of continuing 11-car Pendolino and 12-car 110mph Desiro equivalents, and some high-level mileage estimates adds a further £300m to operating expenditure. This leads to an additional £1bn of revenue required annually just to break even on HS2 Phase 1 capital interest and extra WCML operational expenditure on a wholly public sector basis, and over £3bn p.a. or more if private sector were financing the HS2 element. This also assumes that HS2 pays its way on operating expenditure, though it might not do so initially.

To achieve a repayment of £20bn HS2 capital investment in say 20-30 years on a mortgage-type basis, one would be looking to an excess of half a billion pounds in revenue per year as a static repayment over 27 years. This is for HS2 Phase 1 only, again on a wholly public sector basis.

If you were looking for economic growth as your payback, then sums can be different. They are ultimately a decision for HM Treasury about the balance of financial merit. Here we return to the paradox discussed in Part 1, that a scheme may enable much new housing and commercial development, with such projects increasingly seen as being justified by unlocking new economic growth rather than just about transport links (see the National Infrastructure Commission arguments about HS3, for example), but the DfT transport assessment rules don’t incorporate that directly. Most recently, it has been reported that Transport Minister Robert Goodwill has said that HS2 will remain in public ownership until there is “certainty” on operation and passenger numbers.

What is the foreseeable demand in the London area?

The above leads to the obvious question, about what the volume of demand will really be. An important supplementary question, particularly as it affects rail planning in the London area, is what sort of demand is it? When this is established one needs to ask, what should HS2 and a revised WCML really be offering in the nature of future service structures to and possibly through the London area?

You end up with a circular argument if you aren’t careful, with rail schemes which project present demand trends self-defining their solutions even though we are looking at several decades of dynamic demand, population and jobs growth, where there may be more than one way to define the optimum end state in say 2040 or 2050 or later. Why a London terminus at all is a question at one end of the spectrum, why not a new terminus is another. Meanwhile the rail industry has already considered, and sidelined, one through running option for some WCML commuter services (Crossrail 1 – WCML link), and others may be feasible over several decades. Let’s start with why a terminus at all.

Why the likelihood of an HS2 terminus?

This is best explained by consideration of the alternatives and their present state of maturity, including the former HS2-HS1 scheme. A distinction should be drawn between HS ‘captive’ services, ‘classic compatible’ services with the capability for sharing tracks with conventional UK trains, and the foreseen WCML service offerings, where conventional inner and outer commuter services and new ‘intershire’ services will be the main groupings.

Travel volume points to the London urban area as by far and away the biggest single source of intercity rail travel demand in the south of Britain, though there is also much that is generated by the Home Counties, not least south and east of London. The scope for through links with European mainland railways is also material, even if ‘Brexit’ type attitudes, UK border policies and the Channel Tunnel safety rules are administrative and political obstacles to the business case for through continental services extending further within Britain than the London area (where a line to Heathrow at least has some rationale).

It needn’t have been like that. The original London-Folkestone ‘King’s Cross Bill’ promoted in the late 1980s allowed through running with a tunnel via South London arriving just east of King’s Cross terminus with a through station offering links to existing main lines. This ‘HS0’ would have been able to accommodate through intercity and London & Home Counties regional trains, and also with direct City and Victoria trains rejoining the Southern network near Peckham Rye. It would have been relatively easy to reach North and NW London, though line capacity problems would have arisen in due course with the growth in commuting volumes.

Pantomime season

This part of the Bill was quashed in October 1991 in an announcement at the Conservative Party conference at Blackpool, even though it was navigating the House of Commons’ procedures at that stage. The announcement followed strong lobbying from East London interests and Arup who had designed a route via Thames Gateway. There was also electoral nervousness by the Major government about the Bill’s impact on marginal Conservative constituencies in SE London ahead of the 1992 General Election. The changes led to BR Chairman Sir Bob Reid 2’s famous “Pantomime” comment.

Pantomime Cartoon

Willy Rushton’s ‘Pantomime’ cartoon Original owned by the author

A revised Channel Tunnel Rail Link

The revised Channel Tunnel Rail Link scheme – still at that time called the ‘Union Railway’ project – used the original BR route through most of Kent, but then diverted via Ebbsfleet and the route we now know. Most of the argued-for local regeneration stations in the Thames Gateway were not built. So only in its margins was it initially a regeneration railway.

Even Stratford International was an afterthought given powers under a Transport & Works Order after further strong local lobbying. Reputedly, the Environment Secretary Michael Heseltine failed to ensure that continental trains were required to stop there. They never have called. It has also been claimed that Heseltine wanted HS1 to terminate at Stratford.

The biggest change in principle of all was that the line ended not at a through station but at buffers, at St Pancras rather than Stratford. The story is that Union Railways, led by John Prideaux, had designed a new through station in the then vacant railway lands north of Kings Cross/St Pancras, pointing towards the WCML. But the BR Board decided to adopt a link into St Pancras, and to allocate St Pancras’s refurbishment costs to Union Railways. (Shades of Euston, as we shall see.) The then Transport Secretary, John MacGregor, announced on 22 March 1993 that the Government preferred a lower cost option terminating at St Pancras (with the line then expected to be on the surface from west of Dalston), rather than continuing underground from Stratford towards the former BR proposed interchange station at King’s Cross low level.

HS2-HS1 – the poor value link that died

It was almost inevitable that any HS2 scheme would be expected to provide a link to HS1. In north London the two routes were tantalisingly close and the two London termini closer still.

HS2 Ltd had no intention of offering expresses to serve domestic travellers from north of London to Stratford and Ebbsfleet. This was even though East and South East London, and East Anglia and Kent, are fast growing catchments with 8.6m people, which is 16% of England’s population and comparable to Greater London’s catchment as a whole.

This left the desire by Midlands and Northern cities for direct European rail expresses. A policy requirement for a link to HS1 was therefore willed upon HS2 Ltd, on political instructions. There are shades here of the Section 40 requirement for Regional and Night Eurostar services imposed on British Rail with the original Channel Tunnel Act 1987.

With the available railway geography at Old Oak Common, this mandatory new link required a new through railway if journey times were not to be glacial. This emerged in the original HS2 scheme as a compromised £900m HS2-HS1 link: a new single-track tunnel between Old Oak Common and Camden Roundhouse, mostly paralleling the main HS2 Euston tunnels, then via a slightly upgraded existing North London link (between the WCML and the North London Line) whose current running speeds are 15-20mph in places.

The mothballed HS1 link

The final version of the St Pancras HS1 line, the one that actually got built, created a low speed single track chord (capable of being upgraded to double track) to join the HS1 London tunnel portal with the North London Line (NLL) east of Camden Road station. This has remained mothballed since its creation.

This mothballed line connecting HS1 to the NLL was not the ideal connector for HS2 to connect to HS1 as major works would be required if it were to become a corridor with frequent services in addition to local passenger services on the NLL and through freight trains.

The proposed HS2-HS1 route

European-bound expresses would leave HS2 and emerge at the eastern portal of the HS2-HS1 tunnel. They would then share tracks both with through freight bound for UK ports and with the North London Line which by then would have an even more frequent service than today. All this traffic would pass through Camden Road station and finally diverge via the mothballed link to reach HS1 at its tunnel portal. The overall alignment is shown below and beneath that is the detailed corridor on the surface section through Camden.


HS2-HS1 outline of proposed link in 2010

Camden Local Map showing HS2-HS1 surface section

Camden Local Map showing HS2-HS1 surface section
Base local map reproduced from HS2 material under Open Government scheme

HS2-HS1 operational issues

HS2-HS1 was not a minor scheme. If it was desired to allow full ‘GC’ European gauge trains, the NLL would also need rebuilding for these larger loading gauge trains or with an additional track alongside the NLL. TfL was directly concerned that the combined impact of even infrequent European passenger trains would knock out additional NLL Overground capacity, then already growing fast in volume.

The fact that the service on the HS2-HS1 link would be infrequent did little to placate TfL. Even if a slot were used only once every few hours, it would have to be protected on an hourly basis in each direction. TfL also estimated that only one NLL path an hour was realistic to be allocated to HS2, even if it was attempted to ‘flight’ European trains in groups to match the single track’s capacity.

The same problem of reduced capacity throughout the day (regardless of the level of actual use by high speed trains) arose for UK freight flows. Freight had a further potential problem because not only was this a section of the main UK railfreight network, as it is currently structured, the Camden Road-Roundhouse-Camden Junction section is available as a supplementary holding loop to manage the punctuality of en-route freight trains. Although the main location for a holding area for cross London freight routes is Wembley yard the additional Camden facility would be missed if lost.

Alternative alignments and longer tunnels were investigated in order to avoid the operationally problematic NLL section. These offered little in the form of operational benefit for HS2 Ltd. Also absent was any net environmental benefit for the local community.

The financial case against HS2-HS1

Capacity, gauge widening and environmental mitigation works were pointing to well over £1bn costs for HS2-HS1, for possibly only a few trains a day in early decades of operation, though HS2 thought costs would be lower. The business case for direct European expresses was very poor, as assessed by HS2 Ltd, even on their costs.

Low cost jet fares could be contrasted against long journey times by rail, with poor train and train crew utilisation, border delays into the UK, and trains which might need 500-600 paying passengers to justify their commercial existence. This wasn’t a good marketplace to be in, equivalent to one train having to compete against a frequency of three jets at 150-200 passengers per flight, and also (UK border rules) no intra-UK flows permitted on such trains.

Eventually the Government bit the bullet following Sir David Higgins’ HS2 Plus review (March 2014), and accepted that spending such scale of money wasn’t worth it. The link has been deleted from the HS2 Phase 1 Bill, through Additional Provision 3 (AP3). The Commons Select Committee has said (paras 252-254) that:

“252… The House’s instructions to the Committee included a specific instruction not to consider petitions on whether there should be such a link.
253. The economics of cross-continental rail travel and modal shift from aircraft use are complex. The question of a continuous fixed link between HS1 and HS2 was outside our remit. We do not comment on it save to express a view that the success of and need for a national high-speed network is not necessarily contingent on a fixed link to the international network. Journey patterns are complicated.
254. Quick and comfortable ways to get between HS1 and HS2 will nevertheless be needed. Euston and St Pancras are some 800m apart. A tunnel between them could run under roads parallel with Euston Road, arriving in the northern part of St Pancras. The coherent design plan we have suggested as an imperative for Euston should include convenient ways to get between HS1 and HS2.”

So the immediately available and lowest cost HS2-HS1 link may be a heated, covered footpath, with porters assisting, between Euston and St Pancras International. The Government has ruled out powered links such as a travolator-style tunnel because of concerns over vibration damage to instrumentation at the Francis Crick Institute. Crossrail 2 would provide the first opportunity for such a powered link (although it has to circumnavigate that damage risk as well). At least train passengers could connect tolerably well, between all European and Midlands, Northern and Scottish intercity trains, at close-by stations.

HS2-HS1 not quite dead yet

HS2-HS1 isn’t absolutely dead as a long term possibility, in some form, but it’s clear that any future link beyond Old Oak Common will have to enable London & Home Counties travel in order to start to make a decent business case, with any European rail expresses as a bolt-on extra where selected slots are preserved for their use.

A JRC report entitled East and South East London Partnership report on HS2-HS1 and Stratford International written and published by the author in October 2012 looked at opportunities for improved London & South East commuter corridors. It highlighted that there could be a viable case for the HS2-HS1 link in the future. Greengauge 21 looked at the issue from a rather different context, including joining UK high speed services rather than standard commuter trains in order to try to justify a link between HS2 and HS1 in their May 2013 Travel market demand and the HS1 – HS2 link report. Again it was joined-up near-London domestic services which had a potentially worthwhile case – not through continental trains.

So, why have a London terminus?

So we can now answer the first question, about why a London terminus at all? The answer is, because decisions taken in the aftermath of the King’s Cross Bill took a narrow view about:

  • The possible scope for onwards through trains, north and west of London, as opposed to interchange at London’s Northern main line termini
  • The underlying business case requirement set out in several recent studies, for London & Home Counties travel to be the predominant infrastructure user for a new cross-London railway of benefit for through high speed trains (whether those trains were domestic or international in purpose)

This is clearly demonstrated by looking at the financial reality after construction of HS1. The Javelin domestic services underpin the infrastructure costs of HS1. Southeastern, in effect, is being subsidised by the UK Government to avoid what would otherwise be state-aid on the Continental main line. What the Javelin services might imply also for HS2 usage, is discussed in Part 3.

Written by Jonathan Roberts