London 2050 (Part 3): Tracks to the Future

How might we shape the pattern of London’s growth and development to help bring about a more sustainable outcome? In this part (and the next) of our continuing series we’ll look at the ‘quantity and quality’ schemes arriving at this electronic platform now for rail (above ground and below), surface transport and integration and interchange.

Because there’s a lot to say, we’ll focus first on the overall shape of what’s offered, and particularly on the National Rail element within London and the Home Counties.

The story so far

Part 1 set out the context for the proposed London Infrastructure Plan 2050 (‘London 2050’). It is founded on foreseen growth in population and jobs, and has the goal of ensuring that London maintains a pre-eminent status at a World City level in the decades to come with better capacity, better reach and better quality of lives. We saw that the consultation document has three outputs: a policy case, a manifesto and a funding bid. The requirements for large scale expansion of funding were described, including the transport networks. It was noted that London has been offered less freedom of planning choice than one might have desired, because of how the modelling has been undertaken.

Part 2 expanded the topic of forecasting, and showed how this influenced the spatial and transport choices for London. We discussed some key topics:

• agglomeration theory and practice for the principal London jobs locations
• population densification and dispersal options, including beyond Greater London
• the politics of a growing London
• impact of current London airport planning
• future options for outer London, and the merits or not of the proposed Outer Orbital railway
• some quality and equality issues.

Limited timescale to deliver new capacity

We can now review, also in context, the transport schemes which individually and jointly are brought forwards in London 2050 – and those which appear absent from such advocacy.

London 2050’s Transport Support Paper (TSP) has set out a long list of possible projects. These are costed at a nominal level of detail, and are set out in TSP pages 129-164 for rail schemes (we’ll cover road and other transport schemes in due course). It is interesting that these rail-based projects are sorted by mode, not by project development phase, overall timescale, affordability, wider benefits, nor relative capacity impacts.

So this is not a prioritised list in the sequence in which it is published, although such measures are set out in the TSP tables. The bulk of existing thinking already takes us to the 2030s or beyond. To look as far as 2050, there has been much behind-the-scenes analysis to consider a mix of options that could help to deliver London’s capacity requirements. As we shall see, even that may not have gone far enough, though the forecasting points towards more of the same.

While there will be a need for some new lines, the larger task will be to expand, operate and maintain the rail capacity on existing lines, in order to accommodate the demand volumes. The funding needs, engineering and operational challenges, and pressures on project management, available workforce and skills, are far greater than ‘more of the same’.

Two simple numbers make this point: according to London 2050: “given the limited scope for London’s roads to provide additional capacity, demand on the Underground and rail is likely to go up by 60 and 80 per cent respectively” (London 2050 consultation document, page 9). That’s over the next 35 years, compared to the 183 years it has taken until now to develop the network since planning began for the London and Greenwich Railway in 1831. It will take strong policy, political and funding prioritisation, to deal with this tall order!

In volume, the actual numbers that might be foreseen are suggested below, on a route by route basis. These are not set out within London 2050 itself, but from available transport industry planning documents. We start with the National Rail services. As we shall see, these are already largely forecast through to 2050 and beyond, though whether there is adequate capacity is another matter, and whether London 2050 anticipates those requirements appropriately, with outline funding identified, another matter still.

You can’t buck the system

The Hunting of the Snark is sometimes a useful aide-memoire. “Other maps are such shapes, with their islands and capes! But we’ve got our brave Captain to thank: (So the crew would protest) “that he’s bought us the best– A perfect and absolute blank!”

That’s great if you want to maximise your options. The problem is that a blank is the one thing London’s spatial and transport geography isn’t, and doesn’t permit for future projects. Its past will shape its future. Already detailed Network Rail planning is under way on the main lines for the 2020s, up to the fourth quinquennial period set out in London 2050’s costings, towards 2030.

Control Period 7 (CP7), if it exists in the future under Treasury control of Network Rail, will end in 2029. That’s roundly the date in a nice world when:

• Crossrail 2 and HS2 Phase 2 are being tested pre-opening
• the first of the New Tubes for London – the Piccadilly Line – has been working perfectly since 2027 (according to the latest TfL re-spec)
• all of Network Rail’s 2011 London & South East Route Utilisation Strategy proposals, and some more recent scheme thinking, have been implemented to achieve adequate capacity by 2031.

How hopeful are we about all that?

As for Control Periods 8 to 11, on to 2049 (only one year short of 2050), Network Rail last year in its Long Term Planning Process (LTPP) market analyses set out expectations about foreseeable 30 year changes to demand volumes for the period 2013 to 2043, on the main rail corridors. There is a specific strategy for the London & South East catchment (actually London and Home Counties).

So it isn’t a huge jump for the main line rail sector to have prognostications about 2050, compared to 2043, even if the rate of demand will have to be revised for jobs and population. Just one more Control Period and a bit. So it can be 2043 plus or minus some variation – providing that a similar economic geography is anticipated.

Network Rail already has some 2050+ forecasts!

We’ve shown that, whether you like it or not, the forecasting processes are largely geared towards more of the same, at least for jobs – additional radial-centric commuting. This is driven by a location fix for the bulk of additional jobs, to be ‘agglomerated’ densely in versions of the existing Central Activity Zone, plus options for one or more of: (a) higher CAZ density; (b) an expanded CAZ into the present City Fringe; (c) support for the CAZ plus Satellite Activity Zones, with SAZ indicated at Canary Wharf, Stratford and Old Oak Common.

The 2011 London and South East Route Utilisation Strategy (LSE RUS) forecast growth from 2011 to 2031. Some route-specific changes were also included, having their main impact before 2023, such as Crossrail 1 and Thameslink.

The LSE RUS used a similar growth factor from 2011 to 2023, 0.86% commuting growth per annum, as is used by the 2013 to 2043 LTPP, for the same period. This is noted in the LTPP document:

For all four [LTPP] scenarios the level of growth in central and inner London employment over the first 10 years of the forecasts was kept at the rate used in the London and South East RUS. This growth rate of 0.86 per cent per annum was taken from the London Plan 2011.

Like the LSE RUS, the LTPP adopted the GLA’s 2031 forecast in the 2011 London Plan, of 1.3 million additional population from 2007 to 2031, and 750,000 additional jobs over the same period. Where the LTPP diverged from the RUS, was to vary the forecast beyond 2023, with variable 20 year population and jobs projections to 2043.

There were four scenarios taken forward in the LTPP, varying from ‘Prospering in Global Stability’ (PGS) to ‘Prospering in Isolation’ (PII), ‘Struggling in Global Turmoil’ (SGT), and ‘Struggling in Isolation’ (SII). Only PGS and PII compare closely with the general London 2050 scenario, while PGS is closer in function and style. [See LTPP Chapter 6, pages 25-43]

In PGS, London is the first and principal point of interaction between a prospering UK economy and foreign economies. London and Home Counties area employment growth is forecast there at 0.88% per annum beyond 2023 to 2043, a compound 32.0% increase on 2011. This is lower in PII, where strong UK economic growth is driven by burgeoning domestic production located outside of large city centres, and London and Home Counties employment growth after 2023 is forecast at 0.67% per annum, compound 26.7% from 2011. SGT is 15.8%, SII 14.7%.

In contrast, London 2050 looks at a compound 28.6% jobs increase in its central projection, an average 0.65% per annum overall increase in employment from 2011.

What this all means for London 2050 planning, is that Network Rail’s existing 2043 maximum demand forecasts (which exclude the effects of HS2) are broadly the equivalent of London 2050’s rail requirements projected to 2047-48 (PII) or 2054 (PGS). So we already have some London 2050 route-by-route estimates, for London’s main lines. We shall assume in this discussion that PGS applies, as these are the maximum capacity demand figures already published for 2043 and we are looking at known data.

The distribution of additional jobs is foreseen in London 2050 as follows:

If trends continue, the professional, real estate, scientific and technical activities sectors are expected to see the largest increase in employment over this period (nearly doubling to 1.4 million). Two thirds of the total increase in jobs are expected to occur in the Inner London boroughs.

This points once again to increased, heavy reliance on rail travel to and within Inner and Central London, including radial and orbital flows and large-scale passenger distribution from termini.

Sequence of discussion

The sequence adopted for the rest of this article is therefore:

• Tabulation of what the 2011 LSE RUS saw in terms of capacity gaps by 2031 if nothing new was done other than existing committed schemes.
• New tabulation including the 2043 LTPP forecasting for PGS volumes, and contrasting this against the same 2031 capacity gaps as a starting point – so this represents a nominal commute-to-jobs equivalent of a ‘London 2050+’ forecast demand versus a baseline 2031 capacity supply.
• A summary of LSE RUS comments about how that report sought to fill the pre-2031 capacity gaps (generally with positive results), plus a summary by this author about what the foreseeable additional gaps between 2031 and 2050+ mean, and possible ways of addressing those.
• Review what extra schemes London 2050 proposes to bring to the National Rail party, and whether they alleviate the foreseeable 2050+ capacity gaps.
• Round-up of National Rail and London 2050, and related topics otherwise left hanging.

What the 2011 LSE RUS estimated, for 2031

Merging the LSE RUS and LTPP estimates of projected demand flows (and taking PGS as the upper limit), shows below, in the next two tables, the following gaps in demand and supply on London’s National Rail corridors in 2031, and in 2050+. We should bear in mind that these are forecasts made a long way ahead of any out-turn results.

The table below shows peak hour commuting capacity gaps on London’s main lines in 2031, as set out in 2011 LSE RUS, without the RUS’s subsequent proposals for intervention and relief (nb: This is for the busiest single hour, not for the more usual 3 hour AM peak period)

2031 morning peak busiest hour, from LSE RUS 2011

Demand, capacity, route utilisation and gap forecasts (do-minimum)

The 2011 LSE RUS projected a 43,000 passenger capacity gap overall, on the London’s main line routes in 2031, starting with a 2010 demand baseline. This assumed that no major investments were undertaken beyond those committed then (which included Crossrail and Thameslink). Every route needed some lesser or greater intervention to deliver the expected capacity requirement in 2031.

This was for the busiest morning peak hour (taken by the RUS as 08:00 to 08:59, though 08:15 to 09:14 can be busier). The RUS ignored 07:00-07:59, and 09:00-09:59, as infrastructure investments required for the busiest hour should accommodate demands in the other hours. This may be true for infrastructure but might not be so for train numbers, as extra trains are likely to be required for additional inbound travel in those periods as well. An estimate of a full 3 hour AM peak period in 2031 might be over 0.9 million inbound passengers on National Rail, if taking the busiest hour as 45% of 3 hours.

The LSE RUS thought that delivering this additional capacity was possible in most cases, for example with investment in Bow-Stratford capacity to assist the Great Eastern Main Line. Other solutions are required elsewhere. West Coast main line commuting relies on HS2 and Crossrail-WCML, in order to free up enough train slots. The Great Western main line needs Crossrail extension to Reading and diversion of Heathrow Express onto the relief lines as a Crossrail Express, to free up enough capacity by 2031 between London and Reading on the fast lines. The latter opens up a complex commercial debate between Heathrow Airport, Network Rail, Department for Transport and the Office of Rail Regulation (ORR), because of track ownership and slot allocation issues.

There is still no complete solution offered in the LSE RUS for the South Western mainline long distance and outer suburban commuting growth. Options there are a fifth track into Waterloo from Surbiton (far from easy) or a revised Crossrail 2 coming the other way from Central London, to take away some inner suburban flows and allow some slots to be used by longer distance trains. Even so full relief is not guaranteed. Crossrail 2 was also likely to be useful for West Anglia in due course. Longer trains and platforms were a common medicine, and the main solution for London Overground.

What the 2013 LTPP estimated, for 2043

In planning for London 2050 main line capacity, it might be theoretically possible to regard 2031 as giving the ability to restart planning with a zero capacity gap (but not on SWML). However there will be a busier baseline railway, having little or no extra capacity available using conventional methods of capacity relief, because getting to 2031 will have used up most of the margins. It also assumes that all supplementary LSE RUS investment requirements have been authorised and completed, which might be optimistic.

Worse, Network Rail’s LTPP assumes a wider spread of population than London 2050 is discussing, with comparable growth across the whole London & Home Counties area, but is similar to London 2050 on jobs. London 2050 has a high population within the GLA area. So pressures on inner suburban services with this maximum growth may be greater than projected, already by 2031 and much more so by 2050.

Whilst in theory those extra inner passengers might require fewer extra trains than outer commuters, because of permitted inner standee levels, this may not correlate with available inner train slots on allocated tracks. Meanwhile Network Rail’s outer commuter forecasting might still hold true, for example if higher value jobs exceed expectations, so resulting in a worse net capacity gap than shown below!

Alternatively, as London 2050 shows in its modelling options, move a million of the forecast population outside London, and outer suburban commuting pressures then grow bigger, faster, sooner than Network Rail’s PGS (Prospering in Global Stability). You also need more train paths for outer suburban commuters, because of the standee proportions mandated on inners but not outers.

Overall, Network Rail’s forecast growth in outer suburban commuting gives no comfort that in any scenario there will be fast tracks with spare capacity for relieving inner area crowding.

The London 2050 (mainline capacity) table is derived directly from the Network Rail LSE RUS and LTPP data. An 85% route capacity utilisation is used, as in 2031 estimates, as a ‘safe maximum’ for workable flows. The LTPP 2043 maximum growth scenario for jobs is taken as equivalent to London 2050+, on the jobs estimates described above, though without potential extra inner suburban volume caused there by greater population).

Below are the peak hour commuting capacity gaps on London’s main lines in 2050+, based on Network Rail 2043 projections – similar to London 2050 jobs growth taken to 2054. No interventions beyond LSE RUS 2031 baseline, so adds to the 2031 gaps:

2060 morning peak busiest hour, from LSE RUS 2011 and LTPP 2013, using PGS (Prospering in Global Stability) as the growth rate

Demand, capacity, route utilisation and gap forecasts (do-minimum)

Commentary on 2031 and 2050+ forecast capacity gaps

We now see over a half-million commuters wanting to use rail in the busiest inbound peak hour. A 3 hour inbound peak volume would then be perhaps 1.15 million passengers, a throughput over 70% more than in 2010. These figures also ignore reverse-flow commuting, which is a growing phenomenon in London and the Home Counties.

Multiply that by roundly 1,100 to 1,250, and you see the annualised volumes rising to 1.26-1.43 billion, so that London’s National Rail network would on its own be handling nearly as many passengers as the whole of Britain’s main line railways do currently.

About half of the inbound passengers would be standing, including outer commuters, if they could get on the trains (and would be less likely to find space, any space, on the inners), in the unfortunate event of no new capacity being provided in the intervening decades.

Even with solutions to capacity gaps, the standee London commuter will be a more typical experience on National Rail, in future decades.

That’s enough numbers. What does this mean on the ground? The following 2031 commentary on capacity gaps and solutions is summarised from the LSE RUS. The 2050+ commentary on the further gaps foreseen in the LTPP is by this author.

2031 and 2050+ commentary on London main line capacity gaps and solutions

Route into	Potential interventions, other project comments. RUS comments for 2031 solutions. JRC comments for 2050+.
London Paddington	2031: 5,800 long distance & outer suburban gap, HEX onto relief lines as Crossrail Express, allows 20 tph on fast lines. 2050+: Further 3,400 gap, numbers are pre-Heathrow Hub. Are the main options 6-tracking to Hayes, a new line, or a high capacity ‘digital railway’? (Also see SWML) Long distance passenger growth pressures may be under-estimated with attractiveness of commuting area, M4 congestion and GW electrification ‘sparks’ effect.
London Marylebone	2031: 1,200 gap = new tt, resignalling. 2050+: Further 3,600 gap, trains longer or other solution? There is potential for Chiltern-OOC and Chiltern-Heathrow. London 2050 mentions electrification.
London Euston	2031: 2,400 suburban gap, 8 tph via WCML-Crossrail. 2050+: Minor long distance gap (All 11-car WCML solves, plus several extra WCML slots). Further 4,500 suburban gap requires use of WCML fast slots. HS2 allows relief of these forecast WCML commuting pressures.
London St Pancras	2031: HS1 all 12-car. IEP, HS2 help with MML interCity capacity. 2050+: HS1 more paths, and to Hastings. MML gaps = 5-6 tph if average 450-500 pax/I’City train, feasible on MML. HS2 less likely to assist with 2060 capacity needs, if this is London outer commuting from Kettering etc. Is there a ‘digital railway’ solution for Thameslink outers standing?
London King’s Cross	2031: HS2 relieves ECML, high capacity trains for T’link ECML + Anglia. IEP for Kings Lynn [train spec changed now]. Higher tph on Moorgate line with 6-tracks Alexandra Palace-Finsbury Park. 2050+: Further 5,400 gap. The case for Crossrail 2 (CR2) starts here. Should CR2 trains run on GN? – might be good VfM and benefits, more line capacity if CR2 to Hertford. Welwyn viaduct/tunnel widening might be required for outer commuting, or ‘digital railway’ capacity gains.
London Liverpool Street	2031: WAML: +2tph began Dec.2011, 12-car services, Stratford shuttle by 2019. GEML: Crossrail 1, Bow-Stratford works, 12-car 26 tph early 2020s, 28 tph 2031. 2050+: Further 3,800 gap West Anglia, 9,500 GEML. New line(s) needed. WAML: Lea Valley 4-track + CR2. GEML: capacity gap if 800/900 per 12-cars = 11-12 tph. 8-track Liv.St. New GEML main line? Or WA 4-track via Stansted to Colchester? Crossrail 1 Gt.Eastern full. CR1 Abbey Wood full if to Ebbsfleet, see Fenchurch Street comments. More real tracks needed, not just ‘digital tracks’.
London Fenchurch Street	2031: Fenchurch St. full at 85% level, if all trains 12-car. 2050+: Further 6,800 gap. Another 8-9 tph. Points to new line also relieving Crossrail 1 branches.
to/via London Bridge	2031: Main capacity gain with Thameslink services (Kent now excluded), & all SE 12-car. [With no Thameslink Kent, will 2031 capacity be achieved?] Lengthening on Uckfield and Vic-Kent, and 12-car on Southern suburban. Clapham stops on Gatwick Express are also considered. 2050+: Forecast gap 27,300 extra. This is 30-34 tph, at 800/900 pax per train. Fewer tph if some 12-car inners and standees. Large-scale capacity shortfall from Kent and Sussex corridors, incl outer commuting. Overloads possible on SAZ links to Canary Wharf, Stratford, Old Oak Common. [Doesn’t help that Croydon isn’t included as a SAZ, to reduce inbound rail demand within London.] Can HS1 greater use and ATO etc on SE main lines solve this scale of problem? Might require at least one new line from outer London, possibly towards London Bridge and Canary or Stratford, plus potential for through 12-car services BML to Old Oak Common to connect with HS2, with re-organisation of Clapham Junction and East Croydon junctions.
London Victoria
London Blackfriars	2031: 900 gap causes RUS concern as frequency and platform capacities limit available solutions. RUS considers Overground-style trains. 2050+: Gap worsens, either high-density trains or platform lengthening or ‘digital railway’.
London Waterloo	2031: Re-use of Waterloo International & 10-car suburbans, addresses inner problems. SWML needs 5th track Waterloo-Surbiton, or Crossrail 2 relief, with full relief not guaranteed. 2050+: Further 5,000 gap, mainly SWML. 12-car Windsor lines, inners. The first 16-car commuter train, on SWML? SWML + GWML overload might point to new SW/GW main line, via Heathrow?
Main Orbital Routes	2031: 6,800 pax gap = 14 tph overall (JRC estimate). Longer trains needed. [Initial TfL project under way for 5-cars.] 8-car Southern trains on WLL also supported. 2050+: JRC demand projection is LSE RUS 2031 back to 2023 then x 20 years with PGS growth. Further gap 2,200, points to 6-8-car operation on all orbitals.

These tables shows some stark consequences. Unless all London main line schemes proposed to 2031 are adopted in one form or other, the starting capacity in 2031 will be worse than expected, causing greater difficulties in succeeding years. That assumes HS2 and Crossrail 2 are built, along with a large range of 10-car and 12-car commuter operations. London 2050 population growth may make the 2031 gap a moving target. The forward gap to 2050+ is a call to action on many routes.

These PGS forecasts are not London 2050’s, but are in some ways easier to understand and more conventional than London 2050, because they are a straightforward Network Rail extrapolation of where it is thinks the shape of London and Home Counties travel to work is heading. London 2050 has largely restated the known Network Rail schemes, and then added further schemes to suit the potential spatial formats. Network Rail has also been busy thinking about other projects, so the London 2050 list includes some of those as well.