London & Freight Part 2: The Freight Must Flow


The more we looked at rail freight in London the more we realised the significant challenges it poses, including to TfL’s aspirations for the Overground and its ‘strategic interchanges’. As the network in London reaches capacity it quickly becomes clear that choices will need to be made between providing paths for freight vs passenger, and that might actually require reductions in passenger services.

London’s growth has focused a burgeoning travel demand on same city core as in the 1860s, so it is to cross-city lines and the orbitals that we look to try and expand the core. But it is these orbital routes that are now in high demand for freight traffic. Something has got to give, but we are still not entirely convinced about the strategy for rail freight in London, and indeed who decides.

Our first article provided some context, looking at the various strategies, the national flows and the vexed issue of loading gauge. In this article we focus on intermodal (container) traffic, which is by far the largest freight growth sector and the most relevant to London. Then in our final article we will look at some of the options, and tease out some of the key issues around strategy and decision-making.

Given that most freight flows through London have distant origins and destinations, for the geographically-challenged reader there is no better than the marvellous Adlestrop atlas to help you trace the often tortuous freight routes – current, proposed and potential – beyond the London bounds. And within London, many of the routes and junctions are now included in the fabulous Carto Metro.

More and more boxes

The 2011 London & SE RUS crunched the numbers for 2030 projected intermodal demand and then analysed the route options.

Key Freight Growth Origin Points

Key Freight Growth Origin Points

In the tables above the estimates are given for 5-day and 6-day operation and for 640m intermodal trains. This is a step towards the aspirational 775m trains and a 7-day railway outlined in DfT’s Strategic Rail Freight Network: The Longer Term Vision (PDF), which we looked at in Part 1. Given this, the RUS has made the assumption that overall the railway needs to move to a 6-day pattern with 640m trains, and has used this as the basis to assess the train paths required. However, this has significant investment implications, funding for which is not currently committed. If this investment does not materialise then the routing options will have to be re-analysed based on the increased train paths required and, as we shall see, this is where the planners’ alchemical ambitions may come up against some physical limits.

The map below shows the preferred routes for the four main intermodal flows: Southampton, Essex Thameside ports (Tilbury and the new London Gateway at Shell Haven), the Haven Ports (Felixstowe and Harwich Bathside Bay) and Channel Tunnel.

Recommendations for Freight Routeings

Recommendations for Freight Routeings


Intermodal traffic from Southampton is projected to require 31 additional train paths by 2030, but only 10% of these trains are destined for London. The rest largely head north via Basingstoke, Reading and Oxford. This route is now cleared for W10 gauge, and further planned works cover the partial diversionary route to Basingstoke via Andover, i.e. avoiding Winchester.

Southampton Average Freight Trains 2030

Southampton Average Freight Trains 2030

The main bottleneck at Reading will be eased by the new underpass from the Newbury lines direct to the Didcot Slow GWML to the west of the station. This is part of Phase 2 of the Reading remodelling, but south of Oxford there are still significant capacity issues:

  • in the Eastleigh to Southampton area
  • at Basingstoke, where southbound freight services need to cross the entire layout on the flat to reach the South West Main Line (SWML) Down lines
  • on the 10-mile two-track Didcot to Oxford section, which is shared between freight and six passenger trains (four of which are fast) per hour

According to the RUS, none of these present insurmountable problems, but it does recommend the development of W10 diversionary routes, via Melksham or via Acton and the WCML. The map above gives an idea of the roundabout routes these imply, which understandably does not impress the freight TOCs. Also, going via Melksham only covers the southern section of the main N-S route, because freight will still need to travel on the crowded Didcot-Oxford section. Other than the Melksham route, the fallback diversionary route is via London, using the orbital lines such as the West London Line (WLL), which TfL covets for its Overground services.

All freight north of Oxford currently goes via Leamington Spa, where it hits bottlenecks which further limit the route capacity.

For example: the main freight route north from Leamington Spa runs via Coventry, joining the WCML at Nuneaton. This route is capacity constrained by single track sections between Coventry and Leamington Spa as well as a flat crossing move through Coventry station, so much so that the RUS states:

Capacity over the Nuneaton-Leamington Spa corridor therefore appears to be a potential barrier to future freight growth from Southampton 1

And herein lies the common problem: if the only alternatives are via the London orbitals then TfL clearly has a strong motive to resolve these pinch-point way beyond its turf.

One solution to the Coventry pinch-point may lie in the Kenilworth cut-off to Berkswell, offering a direct avoiding route towards Birmingham and beyond. This short section is intact, now a greenway, and a modest investment could provide a useful route option for freight. But sadly this will disappear under HS2 – rather than resolve a capacity problem, in this instance HS2 may be compounding it.

Closer to home but still well beyond the London boundaries, the newly-sanctioned East-West Rail route could take Southampton freight destined for Daventry, the East Midlands and the North-East. Indeed the 2011 London & SE RUS has already factored this in to its recommendations and should reduce Southampton freight movements through London. As we shall see in Part 3, perhaps this is the sort of regional orbital route that London rail planners need to be promoting.

The East - West Route (source: Wikipedia)

The East – West Route (source: Wikipedia)

Essex Thameside (Tilbury and London Gateway)

Mwmbwls was writing about the new London Gateway port at Shell Haven back in 2009, noting the haggling over the developer’s obligations to fund road links, before returning again in 2010 to raise the urgent issue of investing in the rail links and in particular the route north and west via GOBLIN. Alas, as the building work progresses apace and the December 2013 opening date looms, there’s been little progress on rail capacity through London.

The London Gateway

The London Gateway (courtesy London Gateway)

The estimated 50 freight trains per day from the Thameside ports in 2030 will have to travel west via Tilbury and Barking towards the GEML, where the flow combines with traffic from the Haven Ports.

Essex Thameside 2030 Freight Averages

Essex Thameside 2030 Freight Averages

The RUS puts on a brave face but the analysis underlines how meagre the options are, and the problems they present. The four least-bad routes are:

  • weaving across the Anglia mainline and Crossrail on the flat at Forest Gate/Stratford, onto the NLL and the ECML via the Canonbury Curve – however, as we note below, train paths are very limited
  • via GOBLIN to South Tottenham and then the West Anglia lines north to the ECML at Peterborough, though this involves significant extra distance and problematic interaction with passenger traffic north of Cheshunt
  • via GOBLIN to the MML, but this involves threading into the intensive Thameslink route through flat junctions west of Kentish Town
  • via GOBLIN to the WCML at Willesden

Given this, the preferred route is to the WCML via GOBLIN, which is all very reassuring until one remembers that the GOBLIN electrification still does not have the go-ahead. But Thameshaven trains bound for the ECML will still have to go via Forest Gate/Stratford and the NLL to Finsbury Park as there is no link from GOBLIN travelling west onto the ECML.

The paucity of viable options available prompted the RUS to look into some workarounds with timetable tweaks and new infrastructure, but found very limited potential to stretch time or space. For example, the 4-track alignment east of Stratford is hemmed in. Similarly on the MML the 4-track section from Carlton Road Junction through to Finchley Road is tunnelled or in deep cutting which rules out infrastructure options such as flying junctions. These bottlenecks effectively shift a more intensive freight pattern onto the 2-track GOBLIN through Gospel Oak, where it competes for space with existing freight services and TfL’s Overground.

On timetabling, the RUS identifies just 16 paths each day across the Anglia mainlines at Forest Gate. Of these, eleven each way are for the Haven Ports to London/the West/Daventry, which leaves five paths each way for the projected nine Thameside-ECML freights, therefore the remaining four will have to be at night. All other daytime Haven Ports freight traffic will head the other way on the ‘cross country’ route via Ely and Peterborough, and it is hoped that this will release the paths required for Thameside traffic. However, as we note below, there are limitations on the cross country route, and it is also not electrified. It is certainly cutting it fine, and arguably such a tight working timetable with relatively few route options will impact on network resilience.

The RUS also brushes aside concerns about the lack of capacity on the WCML:

…whilst it is recognised that freight paths on some parts of the WCML are currently scarce, construction of High Speed 2 (HS2) can be expected to alleviate this issue, with passenger demand from locations such as Northampton and Milton Keynes Central generally then being catered for on the fast lines, in turn freeing up slow line paths for freight traffic. (Sec 9.5.9 p172)

So the RUS is counting on HS2 and hoping that demand growth – freight and/or passenger – won’t push the orbitals to capacity for another 13 years. This is a concern because HS2 will not be completed until at least 2025, and has a few hoops to jump through yet. And as we’ve seen above, TfL are predicting severe overcrowding on the Overground by 2016, so there’s every chance we’ll need alternative infrastructure solutions in the near term.

Haven Ports (Felixstowe and Harwich Bathside Bay)

The Haven ports are projected to generate 58 intermodal freights per day in 2030, and the bulk of this traffic is expected to head on the cross-country route via Ely and Peterborough to terminals in the Midlands and the north. Indeed, as noted above, the RUS is counting on the Haven Ports cross-country route to free up sufficient freight paths to allow Thameside-ECML traffic to cross the ECML at Forest Gate. However the London route will still be needed for destinations not suited to the cross-country route as well as for diversionary purposes.

Felixstowe/Bathside Bay Freight Averages 2030

Felixstowe/Bathside Bay Freight Averages 2030

On-going schemes on the cross-country route include gauge clearance to W10 and the new Ipswich chord, and the Chancellor’s Autumn Statement confirmed investment to double track Ely-Soham. However there are still significant investment gaps, including:

  • the route is not electrified
  • capacity restrictions in the Leicester area, where passenger routes cross but where the MML has been shrunk to 2 or 3 tracks
  • capacity constraints at Ely
  • the single-lead junction at Haughley near Ipswich

Without this investment there are limits to the level of demand growth that the cross-country route can accommodate. Therefore the risk is that freight will still need to travel via the London orbitals through the same bottlenecks as with the Thameside traffic noted above, and this does not bode well for network resilience.

Channel Tunnel

Compared to the Haven Ports and Thameside, Channel Tunnel freight does not pose immediate problems. A minimum of 35 specified paths per day in each direction between the Channel Tunnel and Wembley Freight Operating Centre have been protected by Network Rail for the duration of the Channel Tunnel Railways Usage Contract up to 2052, although growth in traffic is expected to use all these by 2030. In addition the RUS expects significant growth from the Thames ports on the Kent side.

Channel Tunnel/Thames Gateway Average Freight 2030

Channel Tunnel/Thames Gateway Average Freight 2030

The main route for Channel Tunnel traffic is via Maidstone East and Catford, then the South London Line (SLL) through Denmark Hill onto WLL to join the WCML at Willesden. The other options are via Tonbridge and Redhill to the WLL, or using HS1 to the Dagenham area.

Channel Tunnel traffic via HS1 faces the same issues as with the Thameside traffic above: the lack of paths through to the mainlines north. Therefore HS1 is unlikely to become a direct replacement route, but rather will serve new markets such as time-sensitive goods or for the larger European gauge trains. HS1 was built to European GC loading gauge, but this traffic requires investment in new terminals in the London Riverside area. However, as we explore below, the success rate in creating new strategic rail freight interchanges in the London area has been very poor, despite this being a key TfL policy objective.

The HS1 route would become much more attractive if a new European gauge route was constructed north from London. Both of these are objectives of the Strategic Freight Network, but without the investment HS1 will remain on the sidelines for Channel Tunnel freight.

So, Channel Tunnel traffic will largely go via the SLL and WLL and north via the WCML. Along with other freight such as aggregates and the growing traffic from the southern Thames ports, this will clearly limit how TfL can expand its Overground services on the WLL in particular. The WLL is 2-track and few if any additional paths are available.

Channel Tunnel freight to the East Midlands and the Northeast via the ECML also fills paths on the NLL, i.e. it does most of the loop around London via the SLL, WLL and NLL – going north via GOBLIN and Haringey West Jn, and south via Finsbury Park and Maiden Lane. Channel Tunnel traffic to the MML does not use the WLL, but goes via Barnes and the Dudding Hill line, and so gets tangled up in the busy Windsor Lines. More intensive use of the Dudding Hill route is also limited because it is not electrified.

The RUS further points to the particular difficulty of freights awaiting paths onto the mainlines, and the lack of places where trains can be regulated:

there is only limited capability for southbound trains to be held whilst awaiting a path through Kent or northbound trains to be held whilst awaiting a path on the WCML. Freight trains must in general therefore be kept moving to avoid delaying the following passenger traffic (and vice versa). The planned commencement of London Overground services via the South London Line to Clapham Junction can be expected to increase this existing issue, given that these passenger trains will use sections of currently freight-only line

As passenger and freight traffic increases, this will significantly reduce operational resilience as delays rapidly knock-on. Clearly capacity needs to be provided, either as loops or 4-tracking key sections.

The RUS does not identify options, but perhaps this is an issue TfL should be more aware of, for instance in relation to the ongoing development plans around the WLL at Earls Court. This section of the route could be safeguarded for 4-tracking through West Brompton through to Kensington Olympia. However the proposed development is tight to the railway boundary and effectively limits future widening, and this will directly affect TfL’s ability to ramp up Overground services. When asked about this, TfL did not appear concerned:

TfL does not intend to safeguard a 4 track alignment in the Earls Court area, to provide freight passing loops on the West London Line. TfL considers that the predicted growth in passenger volumes on the route over the foreseeable future (the next 20 years) can be met by a combination of train lengthening and additional services that can be accommodated within the existing track layout on the WLL. Following these changes, the WLL will continue to be able to provide up to 35 freight paths to and from the Channel Tunnel per day, allowing for the predicted growth in Channel Tunnel freight traffic over the next 20 years. The route will also continue to be able to operate the current number of freight paths to and from the Kent Thameside ports. These are predicted to grow over the next 20 years but this growth is expected to be focused on the London area so it will not necessarily have to use the WLL.

It should also be noted that neither TfL nor Network Rail believes that the provision of freight loops is an effective way to deliver freight train services on the WLL. The use of such loops can have an adverse impact on the reliability of the intensive passenger service operating on the route because of the time freight trains take to accelerate from stand. Keeping freight trains moving over the length of the WLL is the preferred option because this poses less of a risk to the reliability of passenger services. Network Rail is investing in enhancements to Chelsea Bridge to enable the speed of freight trains to be increased on the route to ensure that they can operate reliably.

The comments made above are consistent with the findings of the London and South East Route Utilisation Strategy (LSE RUS), which sets out the rail industry’s view of the enhancements required to meet the predicted growth in freight and passenger demand up to 2031 on various routes including the WLL. The LSE RUS does not include in its recommendations any requirement for the provision of extra passing loops for freight.

Here at LR Towers we like to feel we are sufficiently au fait with the intricacies of rail operation to cogitate on how passenger and freight services might be efficiently interleaved on a 2-track line. But the more we look at freight on the crowded orbital lines the more uneasy we become. We’d be interested to hear back from the LR illuminati on this: do modern freight operations render freight loops a last resort, if not obsolete? And what is the practical limit in terms of freight and passenger trains per hour on a line like the WLL with stations along the length, modest line speeds and flat junctions at either end? Taking a step back, should we be looking at dedicated freight routes as the only practical solution?

Back to the story, and no doubt eagle-eyed readers will have spotted the diversionary Channel Tunnel route via Redhill and wondered why this doesn’t become a main route north, continuing west via the North Downs Line to Guildford, Reading and Oxford.

The North Downs Line

The North Downs Line

The major drawback with this route is that trains need to reverse at Redhill, as well as cross the Brighton lines. As we explored back in 2009, this could be resolved by building a flyover crossing the mainline south of Redhill station, which would allow this to become a key freight route avoiding London. However, apart from the expense, you’d then be up against that most mighty of retailing behemoths also vying for this space, in the shape of Tesco. No doubt a solution could be engineered that accommodated both uses, if there was the will, but the RUS declares that this is a track it does not wish to go down. It cites various reasons:

  • the longer distance – although a glance at the map shows the route to be very direct for freight to the West Midlands and the Northwest
  • the high cost of a Redhill flyover and new tunnels that would be required in the Guildford area
  • addressing level crossings, particularly the one at the west end of Reigate station
  • large sections of the route are not electrified
  • the lack of capacity beyond Reading, which is also the main freight route north from Southampton

So the RUS settles for Tonbridge-Redhill-Croydon as a diversionary route only, and a new freight route using the North Downs Line remains firmly locked in the ‘too hard’ box.

We are still puzzling over this and, given the recent decision by Reigate and Banstead Council to reject Tesco’s planning application, there may yet be an opportunity to secure this alignment as part of a strategic freight route west. But alas it appears that Network Rail and TfL are quietly walking away, which means that Channel Tunnel trains will continue to use the WLL and increasingly occupy the paths that TfL might consider it prudent to covet for it’s Overground services. Watch this space for Mwmbwls’ follow-up…

The rest, and the Strategic Rail Freight Interchanges (SRFI)

We have not gone into detail on the other freight flows in London as their projected growth is much less than intermodal. Some of these flows are significant nonetheless, including:

  • Aggregates and construction materials, with major flows to the aggregates hub at Acton and to depots such as Purley, St. Pancras, Marks Tey and Bow
  • Municipal waste, 29% of which left London by rail in 2005
  • Domestic intermodal to depots in London

The latter provides a fascinating story which shines a light on the practical difficulties in delivering strategic aspirations for rail freight. The 2011 London and SE lists the London freight terminals, mostly those in the pipeline, and confirms that

…most of the growth in domestic traffic is expected to be in trains to and from strategic rail freight interchanges – that is, terminals with modern intermodal rail facilities serving significant concentrations of distribution and logistics industries.

Buried in the archives is the SRA’s 2004 Strategic Rail Freight Interchange Policy (PDF), and this was only updated in a policy guidance note in November 2011 (PDF). The SRFIs are a key part of the plan outlined in the Strategic Freight Network (SFN) 2009 vision document (see Annex D p23). They are also embedded in the London Freight Plan, which is itself a component in the Mayor’s Transport Strategy, and TfL offers toolkits to help local planners and other stakeholders to establish such rail freight facilities. TfL wants to see a SRFI on the fringes of each of London’s quadrants, and a clear path through for developers to create smaller facilities throughout the capital.

Given this unusual policy alignment and government support for rail freight interchanges, surely we should see most if not all of the thirteen terminals on the RUS list up and running within a few years. Not so: the problem is that the SRFIs are proving very hard to get off the ground. Nevertheless TfL is bullish about the prospects, claiming that:

Network Rail forecasts that the rail freight market could grow by up to 800 per cent by 2015 if the planning system continues to support schemes for rail-linked distribution parks like Howbury Park

We’ll skirt over the minor issue that Network Rail is very unlikely to be able to provide the train paths for such a step change in rail freight, but what of Howbury Park? This SRFI near Slade Green has been granted planning permission and two freight paths per hour have been built into the new Thamelink timetable.

Howbury Park

Howbury Park

However there was considerable local opposition as, perhaps unsurprisingly, local people and their elected members raised concerns about issues such as noise, HGV traffic and loss of amenity on green field sites. The key problem is that it is hard to identify significant local benefits for a SRFI, but the disbenefits are tangible and proximate, and this appears to be a common problem facing new rail freight infrastructure in London.

Mwmbwls has covered this already in his posts on the proposed South Radlett SRFI near St Albans. But here the knock-out blow was not to be: developer Helioslough reread the fine print in the various strategies and the inspector’s report, and pushed on up the line. As a result in June 2011 a Judicial Review of the Secretary of State’s refusal to grant planning consent found in favour of the developer, and so the legal planning process is ongoing.

South Radlett

South Radlett

Given the experience with Howbury Park and South Radlett, what then of SIFE Colnbrook by the M4/M25 interchange, or TfL and the SFN’s aspiration for a European gauge terminal in the Barking area? Perhaps the protracted battles and inability to secure suitable sites for SRFIs should prompt a fundamental re-evaluation of the whole concept, and perhaps consider a network of smaller terminals?

More broadly, does the case of the SRFIs in the London area provide a litmus test of the ability for rail planners to deliver on strategic freight objectives overall?

We shall be exploring the issue of the SRFIs in more detail in a future post, but bringing the issue back to the nuts and bolts of timetabling and business planning, the uncertainties around planning decisions make it very hard to make robust projections of future rail freight demand, and then to plan rail paths for freight traffic that may not materialise. From this you may question the freight projections in the RUS: are they too optimistic, or do they underestimate demand based on a pessimistic view of growth starting from a low base point? More specifically this also creates a chicken-egg situation for the planners: do you create timetable paths to strengthen case for planning, or await planning decision before making timetable changes, and risk losing the train paths?

Next steps: from strategy to reality

In Part 3 we will explore some of the options and draw some uneasy conclusions that, while the RUS process has helped identify some of the problems, the investment strategy is far from clear. Clearly the London rail network faces significant challenges as a result of rail freight. Yet despite the careful alignment of strategies, it is not clear how these will allow the aspirations of TfL and the Strategic Freight Network to become a reality. And the clock is ticking…

1 London & SE RUS 2011 p165

Written by Lemmo