Back to the Future: (Re)lengthening and Shortening at Waterloo


With the current three week blockade at Waterloo to lengthen platforms 1-4 coming to a close, we take a look at works, and at how the detail doesn’t necessarily always match the headlines.

The rebuild is part of the Waterloo and South West upgrade plan, with the aim of increasing peak time capacity into Waterloo by 30%. The three week blockade was marketed as bringing “more platforms, longer trains and better journeys down the line”. There is far more to the works, however, than just that. Indeed, Londoners with long memories will remember those platforms used to be 10-car once upon a time.

The blockade also marks the end of an era, with a change of South West Franchisee from Stagecoach (who have run it for the last 21 years) to a First /MTR joint venture who took over at 02:00 on Sunday 20th August.

A short international diversion – for a limited time only

The blockade sees the completion of the first stage of returning Waterloo International to regular intensive passenger use, albeit for a short time only. Initially, this is to reduce the impact of the works during the three week blockade at the other side of Waterloo station, then for a fourth week to reduce the impact of the nine day blockade at London Bridge – with a limited number of SouthEastern services on the working days (the nearest stop on the services being Sevenoaks).

The five platforms (20 – 24) in the International terminus were originally designed to accommodate 387m Eurostar trains operating at six trains per hour (tph), with long dwell times in curved platforms. Passengers passed through a long convoluted route through passport control and security, with indirect connections to the London Underground stations below ground.

Eurostar’s use of a station named after one of Britain’s most famous victories over the French was thankfully (at least from a Foreign Office perspective) only ever intended to be temporary. Eurostar’s departure to St Pancras in November 2007 thus left five platforms that would, superficially, appear to be very useful to Network Rail and the Train Operating Company (SWT). In theory, they should have allowed both to deal with rising passengers numbers at Waterloo, which have doubled in the 20 years since privatisation. In reality, though, the platforms were completely unsuited to commuter rail use and required an extensive rebuild to make them fit for purpose.

This wasn’t the only problem. To make things more complicated, the Waterloo International terminus belongs to British Rail Residuary (via London and Continental Railways) not Network Rail. As a result, the chances of anything happening quickly were always slim, due to the added complications and costs this inevitably brought with it.

Nonetheless, in 2014 platform 20 became available for use after some quick and inexpensive works. Openings in the wall between platforms 19 and 20 were created, which opened up access. Stairs and ramps were also added to address a height gap, as the International station platforms are all higher above the ground level than the domestic ones. Platforms 21 and 22 were also available for more limited use after the construction of a temporary scaffold access bridge over the former Eurostar ‘orchestra pit’.

The first phase of the International works has now been on-going for 15 months whilst the second phase will be complete in another 16 months. The current work is intended to provide a more optimal, longer-term solution for using the International platforms. The primary aim is to allow the majority of the South West Franchise’s Windsor line services (those serving Putney, Richmond, Twickenham, Hounslow, Feltham, Kingston loop, Windsor and Reading) to move into the International Terminal. This will then open up more platforms at Waterloo for all service groups, albeit after some alterations to the station throat.

The first stage at International

The main task in phase one has been rebuilding the track in the station throat to allow an increase to the number of trains that can pass through into the ex-International platforms without disrupting traffic elsewhere. To put some context on the scale of changes required – only 6tph were timetabled into the platforms during the days of Eurostar. The aim now is 18tph, based on 12-car, 240m trains (though at the moment the maximum train length on the Windsor lines is 10-car).

As one might imagine, this has required some serious work. Not least, is shortening the platforms at the country end and infilling bridge decking in the areas formerly covered by the platforms over Westminster Bridge Road. This enabled a new track layout capable of handling more train movements (especially more parallel moves). Just as importantly, it enabled the moves to take place far closer to the starting signal at the new platform end, which then improves train throughput in the throat.

This work also had the useful side effect of removing some of the most curved parts of the platform – which are at the country end. This will improve sight lines for train dispatch – especially for drivers of trains of fewer than 12 cars, who can now see the signals at the platform ends. It also means fewer platform staff required for dispatch. Shortening the platforms at the concourse end (the other area of high curvature) has also helped with this, whilst also helpfully creating a new concourse for Windsor line services at the same level as the International platforms. This should allow it to cope well with the higher volumes of commuters now passing through Waterloo and also provide a much more direct route to the platforms.

We hope to cover the other part of the Waterloo and South West Route Upgrade in a future article.

Ultimately, the reduction in platform length is equivalent to seven cars. The works are virtually complete at platform level and the platforms are now being used during the blockade for all the Windsor line services (albeit at a reduced service level).

The new International track plan.

The next phase at International will involve completing the works to cope better with passengers below platform level, with the addition of access from platform to Tube without needing to going via the main concourse (a feature of many Waterloo platforms). It will also involve the creation of a second Windsor line concourse in the ‘orchestra pit’ that was the entrance to the former International Terminal at near-street level. This second stage is due to complete in late 2018, with the first trains running after the December 2018 timetable change. The total cost of the International rebuild is around £400m, half of the current total capacity scheme.

Lengthening the suburban platforms

The purpose of the works during the current, disruptive three week blockade is to enable the South West Main Line (SWML) suburban routes (via Wimbledon) to operate with all 10-car services, rather than just the small number currently able to do so (thanks to platform lengthening works at other stations over the last four years). Before the works, most SWML suburban trains used platforms 1 to 4. These were only eight cars long and optimally connected to the SWML slow lines with some shared use of the longer platforms 5 to 7 (platform 5 being 10-car and platforms 6 and 7 being 11-car). This brought with it the added problem that platforms 5 to 7 were less optimally connected to the slow lines, adding plenty of operational difficulties to running even the small number of 10-car suburban services that the available stock would allow. As a case in point, SWT leasing and refurbishing the 24×2-car Class 456 EMUs to enable 10 car services by extending doubled-up 4-car 455s would theoretically allow just over half the services to run at 10-car lengths; however, the operational issues at Waterloo limited them to less than half the theoretical level.

Pre-Works Track plan, slow lines

Those with longer memories will probably question the use of the term ‘lengthening’ to describe the works being done at platforms 1 to 4. Indeed perhaps the term “re-lengthening” is actually far more correct.

The present format of the domestic platforms and the concourse at Waterloo date from the early 20th century rebuild between 1903 and 1925 (the official opening was actually in 1922, pre-completion). This was carried out by London and South Western Railway and later Southern Railway, initially under the direction of Chief Engineer J.W. Jacomb-Hood. Jacomb-Hood had travelled extensively to America and Europe in an effort to understand what constituted best practice, which he then incorporated in the rebuild. The SWML suburban platforms were designed to take eight cars with a steam locomotive at either end – equivalent in length to a modern 10-car EMU.

Like much of that Waterloo rebuild, the suburban platforms required extensive use of non-standard (often unique) switches and crossings in order to maximise the number and length of platforms. This was almost entirely due to the unusually narrow station throat, which is still a major issue today.

These non-standard switches and crossings proved consistently difficult to maintain over the following nearly 80 years. The suburban platforms were electrified in 1916, two years after Jacomb-Hood’s death, but they retained their 10-car equivalent length even after resignalling to 4-aspect colour light signalling, whilst more switch and crossing complexity was added in 1936.

Changes in the eighties

With the 1936 vintage signalling along the SWML and Windsor Lines becoming increasingly unreliable and dangerous, in the early 1980s the decision was taken to resignal the many existing signal boxes along the southwestern routes in Greater London and replace them by two new main signalling centres at Wimbledon and Feltham as part of the Waterloo Area Resignalling Scheme (WARS). As part of WARS, the decision was also taken to rationalise and standardise switches and crossings where possible, the goal being to improve reliability and reduce maintenance requirements. As part of this work, platforms 1 to 4 were shortened to eight-car length in 1984. This deliberate reduction of terminus capacity may seem perverse now, but it should be remembered that 1982 was the low point in passenger numbers (and railway finances) after a decline over the preceding 25 years. To a certain extent, this made the decision sensible at the time it was taken.

Spectacular growth

Fast-forward 35 years – from that low point to the present day – and one finds spectacular growth rather than continued decline. This is especially true since privatisation, with passenger numbers at Waterloo alone increasing by an incredible 69% during that period, and by 116% franchise-wide in the same time.

On the SWML suburban side, increasing capacity isn’t just about moving to an all 10-car operation by increasing train length and hence passenger capacity. It’s also about increasing the number of trains that can potentially run on the SWML slow lines in the future. With the current signalling, track layout and rolling stock dwell times, the maximum number of trains on the SWML slow lines is limited in practice to about 18/19tph by several pinch points – with platform dwell time at Vauxhall in the peak flow direction usually being the ultimate limiting factor. With all 10-car operation, new rolling stock with wider and quicker opening doors, larger vestibule areas, more metro style seating layouts, and walk-through gangways and better acceleration/deceleration, the possibility of increasing the train frequency is worth factoring into the overall picture.

This is why a less publicised part of the blockade works involves taking the opportunity afforded by the work to add an extra (third) staircase to the slow line platform island (platforms 7 and 8) at Vauxhall. This is a classic example of how making non-obvious changes at one station can help speed up services across the whole line. This extra staircase, positioned at the country (Clapham Junction) end, should encourage passengers to distribute themselves more evenly along the trains (especially as they get longer). The more evenly distributed the pattern of boarding and alighting becomes, the more the required dwell time at Vauxhall drops.

Resignalling… and shortening

This leaves a final limitation on maximising slow line train frequency – signalling. It is understood that resignalling with a modern, more advanced signalling system such as ETCS (European Train Control System) could allow an increase in frequency to 22tph (allowing for all the other possible limitations on the slow lines). This explains the specification, when designing the current platform lengthening on the suburban platforms at Waterloo, that the throat layout and number of platforms should be able to cope with 22tph and 10-car train lengths. Hence, after rebuild the suburban services will operate from platforms 1 to 6 with a more uniform platform use, as platforms 5 and 6 will be more accessible – especially as regards parallel moves of services into and out of platforms.

Somewhat counter-intuitively to achieve this, given all the other limitations, platforms 5 and 6 are actually being shortened to 10-car length and the ends of platforms 5 to 7 are being realigned too. The Underground station’s access has always been a bit poor for platforms 1 to 4 compared with many of the other SWML platforms, so the work to improve the access with additional stairs will also be completed during the blockade.

Starting work on the hole for the new stairs to the Underground (Christmas 2016)

Post Works Track plan, slow lines

The impact of the works

The works were originally envisaged as requiring the closure of platforms 1 to 9, as the track the end of platforms 7 and 8 would also need to be altered. This would also require the closure of platform 9, and much of the publicity material still advises that platforms 1 to 9 would be closed. In the end, however, the additional closure of platform 10 was required, due to the need for a barrier train at the end of platform 10 and across the Up Main Fast. This extra requirement was always going to make operating the reduced service during the blockade even harder, especially as there was no longer room to allow for anything to go wrong.

The reduction in the available approach tracks, however, and the available ways of getting to them from the normal tracks used by the services further out actually has ultimately had a bigger impact on the number of trains that could be run. The scale of the likely disruption, the lack of a margin for error and previous negative experience at London Bridge is likely why SWT and Network Rail embarked on a 14-month publicity programme, which included encouraging passengers to avoid using Waterloo during this blockade altogether.

The new signalling gantry (for P1-8) across the end at the new ends of P1-6 which happens to be directly above the old water pipe network used to refill the steam locomotives before electrification just over a century ago.


The lack of resilience available turned out (unsurprisingly) not to be good for Network Rail, SWT or passengers. There were seven points or signalling failures in the first two weeks of the blockade, causing significant disruption. There was also a well-publicised incident in which an operational train departing platform 11 hit the barrier train. This caused major disruption, with just five platforms and two approach tracks available at one point.

Many of the failures occurred at sets of points operating in very different ways to their normal usage pattern, a lot of which could not be effectively simulated beforehand. It is not particularly surprising therefore that, despite extensive pre-blockade maintenance, some of those points did not take the change well. Indeed Network Rail had spent a lot of weekends in winter 2016 and early spring 2017 replacing the points in the Waterloo throat ahead of the works to either help reduce the scope of the works or to improve resilience during them but this hasn’t always helped.

In the end, SWT decided to declare the ten working weekdays of the blockade (before the end of the franchise) as void days for season ticket holders – essentially ‘free’ days where they are eligible for refunds or extensions. Positive as this is for travellers, it shouldn’t be seen as an entirely magnanimous gesture – it also puts pressure on the incoming franchisee to be similarly ‘generous’ despite the delay/cancellation/disruption repay regime being completely different in the new franchise. Most (or possibly all) of SWT’s generosity will also be recouped from Network Rail.

It’s not over until the fat controller sings

The disruption for passengers, new TOC and Network Rail isn’t necessarily over when the blockade finishes, as the train drivers’ union ASLEF is unhappy with the way platforms 7 and 8 have been resignalled. The main signals for those platforms are on the new gantry (see above photo) at the 10-car mark, but those platforms are actually 12-car length. As a result there are small co-acting signals (effectively repeaters for the main signals) to enable 12-car trains to use the platforms, but most peak time trains will be longer than 10-car so those services will represent the rule here, not the exception. ASLEF have already balloted in favour of action, so life post-blockade may not immediately be so smooth as hoped.

There are also no significant services changes due yet, so commuters would do well to moderate their expectations. The first real changes will come with the December 2017 timetable change. To a certain extent, this work thus doesn’t so much represent an early Christmas present for travellers. More a present for Christmas yet-to-come.

We hope to cover the other part of the Waterloo and South West Route Upgrade in a future article.

Written by ngh