A Study in Sussex Part 6: The Approaches to London Bridge


In part 4 of our Study in Sussex we looked at the restrictions on capacity at Victoria Station and came only to the rather vague conclusion that there was probably more capacity available in future. Other possible ultimate destinations for non-Thameslink trains from Sussex are somewhere on the West Coast Main Line via Clapham Junction and the Blackfriars bay platforms. In both cases the track layout means these are really more suited to local metro services. In principle Thameslink could operate at more than 24 trains per hour (tph) but Network Rail seem adamant that this will not happen, with providing 24tph a hard enough challenge for now.

The only other medium term possibility for handling an increased number of services from Sussex once they get to central London is an increase in capacity of the terminating platforms at London Bridge. On this occasion we will look not so much at the station in its reconstruction phase, but instead at the track layout after 2018 when the Thameslink Programme should be complete. It is, however, a convenient time to also mention other changes not directly related to London Bridge station that will occur before 2018 that will affect train service (or frequency).

Never Trust Any Source

restrictions on use

A Network Rail slide detailing capacity at London Bridge before, during and after building work there as a result of the Thameslink Programme. FCC is First Capital Connect (defunct but effectively Thameslink), SET is SouthEastern Trains.

In May 2014 Network Rail made available a presentation containing a slide (pictured above) showing the capacity of London Bridge in terms of trains per hour at various phases of the Thameslink Programme. Here it is worth bearing in mind London Reconnections’ Third Law of Trustworthiness of Sources™ . This states that one should never, ever assume that a slide within a presentation is up-to-date.

In this instance despite the date of the presentation being on the slide, it is clear that it is actually quite old. It refers to 18tph Thameslink services down to East Croydon when it is now settled that 16tph will operate. The commencement of non-stopping of trains to/from Charing Cross is also shown as May next year and this will now commence in January.

Did we mention this before?

As an aside, you will note that, according to this slide and other documentation, after 2018 there will be only 22tph to Cannon Street – down from 25tph. It is a rather little publicised feature of the Thameslink Programme that, as it stands, it will lead to a reduction in the number of paths available into Cannon Street in the morning peak. Network Rail had hoped that a clever plan to get around this would reinstate at least two of these trains per hour and talked optimistically to the GLA Transport Committee about it, but it seems that it is now accepted that the plan then conceived would actually cause more problems than it solves.

Next installment in the Cannon Street Empties saga

Despite the above setback, Network Rail and SouthEastern are quietly hopeful that they can actually manage 24tph to Cannon St in the high peak hour after the Thameslink Programme is complete. On occasions when the they have been unable to get Empty Coaching Stock (ECS) out of Cannon Street by the “back door” (the reversible curve on the western side of Borough Market Junction) and have had to run them through London Bridge the disruption was not as great as feared. Furthermore there is still the option of running trains from Cannon Street in the morning without stopping at London Bridge – either as ECS or in service. On top of that there is a lag between the maximum peak frequency inbound and outbound so, coupled with carefully tuned platform occupation times at Cannon Street, it may just be possible to eke 2 extra trains per hour out of the timetable at the busiest time in the morning peak.

Don’t blame everything on Thameslink

You will also note that the number of trains into Charing Cross per hour is marginally down by one. On this occasion the Thameslink Programme probably isn’t the cause. On the contrary, without the removal of the 1tph in the peak period Thameslink service that currently goes via London Bridge using SouthEastern lines, it may have been worse. The cause is almost certainly due to plans for 12-car trains which take longer to clear critical junctions, plus the need for extra resilience.

Extra trains from the south? – Not a lot

Until reconstruction of London Bridge began there were nine terminating platforms permitting a service of 30tph in the morning peak. On top of that there was a single Thameslink train each hour. After the Thameslink works these figures will be 20tph and 16tph respectively according to this and other published information. So in terms of number of trains the net benefit on the Southern side of the station is just 5tph, from 31tph to 36tph. Again we have stated many times that, in terms of the number of trains coming into London Bridge from the south, the Thameslink Programme offers much less benefit than most people would imagine – something just over 16%. Of course the figure continuing onward from London Bridge is dramatically different.

What is quite surprising is the difference in the overall number of trains arriving from the east and the south (basically the “country” end) at London Bridge before and after the Thameslink Programme. Taking into account that there will be 16tph Thameslink services through London Bridge and not 18tph as originally planned, we can add up the totals on the slide and find than the total overall improvement in capacity is 1tph. It is not entirely surprising that whilst the Thameslink Programme website extols the many virtues of the programme it is careful not to specifically proclaim “more trains”.

Platform Capacity

One other very significant thing present on the slide is how many trains per hour the six remaining terminating platforms at London Bridge handle at various stages of the Thameslink Programme. Again the slide is not to be entirely trusted as it is almost certainly quite old.

When the number of terminating platforms went down from nine to six at the start of the rebuilding of London Bridge station there was very little disruption to the timetable. Two trains per hour on the South London Lines service had already been removed and a further one diverted to Victoria – which despite Network Rail’s protestations seems quite good at handling extra traffic when this is required. So in fact the reduction of three platforms at London Bridge only led to a reduction of three trains per hour and it seems that six platforms could, if stretched, handle 27 trains per hour.

Only two tracks

From January 2015 the number of tracks between London Bridge and New Cross Gate goes down to two for a short distance.

What will happen from January 2015 is that the number of trains that can be handled at London Bridge will go down to around 23tph. This would appear to be nothing to do with platform capacity and will be entirely due fact that it will be necessary to temporarily reduce to two the number of tracks into London Bridge from New Cross Gate. This is to allow the Bermondsey Diveunder to be constructed.

In September the pain started to be felt

One thing that has remained unclear even as work has begun on the London Bridge rebuild is just how Network Rail intended to get down from 27tph to 24 or even 23tph – i.e. which services were going to be cut. We think we have at last discovered the reason for this. Network Rail themselves did not know and could see that none of the choices were at all palatable from a passenger perspective.

The most recent changes to the Southern timetable took place at the start of September. This involved the judicious removal of just a few services and diversion of one service to Victoria. To accommodate this diverted service it has been necessary to cancel one Gatwick Express train from Gatwick to Victoria but curiously no corresponding Gatwick Express train from Victoria has been cancelled – suggesting either some dividing of Gatwick trains at Victoria or the stock being used for a commuter service from Horsham.

In January 2015 the pain gets worse

Southern Railway’s January 2015 timetable has not officially been published but appears to be a bit more of the same with a few more trains withdrawn. There are certainly going to be further terminations/starts from South Bermondsey. It would seem that these get in the way of the intense service from New Cross Gate which will by then be reduced due to the Diveunder work. The significance of all this is that, like Victoria, it is not actually the number of platforms at London Bridge which is the ultimate restriction otherwise the service could continue as it does today.

London Overground saves the day again?

The extension of the London Overground to Clapham Junction was originally controversially brought forward to provide some sort of alternative service when the South London Line service was withdrawn in order to facilitate the Thameslink Programme. It now seems likely that London Overground will provide enhanced alternative services to further support temporary arrangements during building work taking places as part of the Thameslink Programme.

Although looking at possible London Overground enhancements does appear to be straying from the main purpose of this article’s topic, it could have some less obvious consequences that may have some relevance later on.

If there is anything good that is coming out of the January 2015 reduction of trains into London Bridge from the south, it is a rumour that London Overground will be running extra trains on the East London Line to Crystal Palace. It has to be said that, although the extra 2tph to Crystal Palace is a known aspiration of London Rail, this recent rumour didn’t come from London Rail or any other part of TfL.

It is not clear if the above rumour refers just to the peak service or if it is even just one or two services. What we do know is that the most recent Commissioner’s Report makes it clear that TfL have quietly increased their order of the next tranche of London Overground trains (the original order was destined for West Anglia in its entirety) by six trains. The report only mentions extra trains on the North London Line but it seems entirely plausible that the six (4-car) trains are to replace 5-car trains on the Euston – Watford Junction service and that two of them would be destined for the East London Line service to Crystal Palace.

There is already a known demand for such additional London Overground services to/from Crystal Palace to relieve places like Brockley – at least in the peaks. TfL have, in the past, referred to them many times as “pixie busters” – an industry term for trains run for the purpose of bring down excess overcrowding that gets flagged up to the DfT as Passengers In eXcess of Capacity (PIXC).

Welcome though these rumoured extra pixie busters would be, they would, of course, only be 5 cars long. Depending on which rumour you believe, these will either be additional trains taking the East London Line up to its design capacity of 18tph, or they will merely be the diversion of two of the 4tph New Cross trains. The latter option would enable the current ELL core frequency to be maintained and so not introduce potential issues there as the ELL core is pushed towards its limit – Canada Water platform re-occupation time being a particular issue.

If the rumour of extra London Overground trains is true it almost certainly won’t mean more capacity on the slow lines between Sydenham and New Cross Gate, as they will probably occupy the former slots of Southern trains that will from January 2015 terminate from the west at Crystal Palace during the London Bridge works (having originated from Victoria). Of course the replacement London Overground service to Dalston Junction will only be 5 cars long instead of the original Southern service of 8 or 10 cars to London Bridge.

Where this might start to get interesting is when one considers what pressures there will be after 2018. Will there be strong pressure to retain such London Overground services? Could they co-exist with reinstated Southern services to London Bridge? Possibly the main question will be where the higher unmet demand is after 2018 from stations such as Brockley. Would it be to London Bridge or the East London Line? Before we get too excited we need to bear in mind that the timetable planners in 2018 would probably be looking for 8tph London Overground from West Croydon/Crystal Palace (as now) and 8tph Southern. Nevertheless, if demand for travel from inner South London continued to drift towards Canary Wharf and away from the City, there could be circumstances where 2 extra terminating slots at London Bridge became available after 2018.

Bermondsey Diveunder

It is finally time to look at Bermondsey Diveunder now that details of it are emerging. In a sense Borough Market Viaduct and Bermondsey Diveunder are the only two parts of the Thameslink Programme that are fundamental. Everything else is just upgraded to fit in with these completely new structures. It is true that the Canal Tunnels connecting St Pancras Thameslink with Finsbury Park are critical to the actual proposed service, but the alternative of terminating at Kentish Town or elsewhere was always available. We will save a detailed look at the actual construction of the Bermondsey Diveunder for another time.

Bermondsey DiveUnder

A Network Rail view of the Diveunder.

The above diagram shows the track layout of the Bermondsey Diveunder. Note the provision of four tracks that actually diveunder. The southernmost one is especially important as it enables London Bridge terminating trains to reach the Thameslink down fast with the minimum of conflict and without affecting East London Line services. The northernmost track is reversible but seems to serve very little useful purpose as an up line for regular day to day operation as it rather unhelpfully joins a down line west of the diveunder.

The Bermondsey Diveunder basically takes the Charing Cross Lines through New Cross and puts them under the Thameslink Lines thus enabling the two separate busy services to cross without conflict. If that was all it did then it would only require two tracks beneath the two Thameslink tracks. It is in fact more useful than that.

Unravelled Bermondsey Diveunder

An alternative diagram drawn by our own Unravelled.

For a slightly different perception see the above diagram. Although this does not provide such a large overview it does show the construction in more detail and where it is located in relation to current structures. CHP refers to the Combined Heat and Power plant which is an incinerator that burns some of London’s waste and is located just off Surrey Canal Road.

On the northern side of the diveunder there is the previously mentioned further track that as well as forming the Charing Cross down line also connects with SouthEastern’s Cannon Street lines. Although this additional connection is not much use for normal day to day operation this enables various things to happen which might be useful during service disruption or engineering works – such as terminate trains that normally run to Cannon Street at London Bridge low level instead. This could be done without needing to cross Thameslink or Charing Cross services on the level.

On the southern side of the diveunder is a further track that enables access to the down slow from London Bridge to New Cross Gate without crossing the Thameslink tracks on the level. From the slow line there is a crossover that gives access to the fast line (which is now mainly but not exclusively for Thameslink). So there is a pretty much conflict free route from London Bridge terminating platforms to the down fast. This (and its simpler equivalent on the up side) means that, whilst there are only 16tph Thameslink trains, there will also be other fast trains from East Croydon. These trains will terminate at London Bridge and slot in between the Thameslink trains utilising four of the eight per hour five minute gaps between Thameslink trains. The trains in question will be the Uckfield diesel services and almost certainly a train that combines Tonbridge and Reigate portions at Redhill making 20tph and four spare slots.

Bermondsey Diveunder Construction

Thanks to Unravelled we have have this fairly recent photo of construction of the diveunder part of the diveunder. Clearly as much of this as possible will be built prior closing tracks and realigning them.

Capacity – but useable capacity?

Track Layout when works complete

A multi-coloured track diagram showing how the layout will be at the end of the works.

The above diagram of the works when completed shows that, thanks to the Bermondsey Diveunder and provision of the Brighton down slow to use it, trains can join or leave the fast Thameslink lines and terminate at London Bridge with the absolute minimum amount of disruption or conflicting moves.

We mentioned above that there will be four spare slots on the fast Brighton Line on the approaches to London Bridge at the end of the works. If there are four spare slots on the fast line then the obvious question is: why not use them? It seems that the answer is that it is too difficult at present to find a matching slot through East Croydon. The TSGN timetable south of the river will be built around maximising northbound capacity in the morning peak period at East Croydon. This effectively means there is no opportunity to “tweak” the East Croydon end to fit in with paths between between London Bridge and East Croydon. There are hopes that by 2018 an examination of the timetable may result in 22tph being possible. Beyond that it is hardly worth looking since Network Rail would be extremely reluctant to give away the last two paths due to their usefulness in assisting recovery from minor disruption.

Looking at the signalling

In fact there is a further opportunity to improve pathing between London Bridge and East Croydon on the fast lines. That is to introduce ETCS (European Train Control System). Even its most basic form (level 1) would help matters. Level 1 is merely cab signalling superimposed over the existing signalling which would be easy to do as the signalling would be upgraded anyway from January 2015 (and operated in its entirety on this stretch of line from the new Regional Operations Centre at Three Bridges). This would remove restrictions on signalling sighting distances for suitably equipped trains and offer more resilience as well as the possibility of taking a minute off the East Croydon – London Bridge schedule. Maybe this added resilience would be enough to enable Network Rail to agree to the final two paths being used.

In fact we could go further when looking at capacity and argue that the fast lines from north of New Cross Gate to south of Norwood Junction could quite easily handle 30tph if needed with Automatic Train Operation and trains 2 minutes apart. This wouldn’t fit in very well with Thameslink at London Bridge with its alternating gaps of 2½ and 5 minutes but could be done by padding the timing of some trains between suitable points that would not conflict with anything else. Of course this cannot be done without removing any issues currently at East Croydon, and, in addition, any fast train stopping at Norwood Junction will become extremely problematic. It must be emphasised that there are no proposals to do this. In Network Rail’s 30 year plan published last year, Network Rail was as confident as it can be that it can get to 2043 without resorting to such untested ideas. In any case such an idea would require either Thameslink increasing its core capacity (which Network Rail insists won’t happen) or terminal capacity at London Bridge which would be beyond the most optimistic projection.

Perhaps scheduling the fast lines south of New Cross Gate for 2 minute headways is a little unnecessary but the idea of extending ATO south of London Bridge as far as Norwood Junction or East Croydon may not be. A disadvantage would be that it would probably require certain services to be operated by a dedicated sub-fleet equipped with ATO capability. An advantage would be that it would give much more resilience to disruption, and of course by then ATO on Thameslink would be tried and tested and the majority of trains (the Thameslink ones) that use the fast lines would already have ATO capability.

ATO just for the fast lines out of London Bridge could be construed as overkill but we have already seen in our article on Clapham Junction that ATO could be beneficial elsewhere. Clearly a decision to install ATO (or not) is going to be dependant on looking at the bigger picture and not just one area in isolation.

Fast or Slow

We have established from the first diagram that only 20tph will terminate at London Bridge once Thameslink is complete. One wonders why the figure is so low. A simple explanation could be that Network Rail is struggling to find trains suitable to terminate there due to conflicts and capacity issues elsewhere. It would appear that the most likely explanation for these trains is:

  • 8 via South Bermondsey (up from 6 today)
  • 8 via Forest Hill (up from 6 today)
  • 2 diesel services from Uckfield
  • 2 services from Redhill with portions from elsewhere – at least Tonbridge and Reigate and possibly Three Bridges as well OR, more likely, 2 services from Epsom running fast from Norwood Junction to London Bridge

The 8tph via Forest Hill would very neatly fit in with the 8tph London Overground via Forest Hill giving an alternating service to each destination. There is also currently a single direct train from Guildford to London Bridge in the morning and an equivalent return journey in the evening, but it is presumed that this will no longer run.

One suspects that one would be struggling to fit in more services via South Bermondsey especially as these have to fit in with London Overground to Clapham Junction at Queen’s Road Peckham and Peckham Rye, and anything via Tulse Hill is also likely to be problematic. Nevertheless Network Rail are known to have looked at providing 2tph around the Wimbledon Loop in each direction terminating at London Bridge. These would have complemented the 2tph in each direction terminating at Blackfriars. Network Rail haven’t abandoned the plan as a long term aspiration but it has been made hugely more difficult to timetable due to the decision to continue to provide through services on Thameslink from the Wimbledon loop. The full ramifications of this decision continue to be felt in the planning process.

The services via Forest Hill would probably have to dovetail in with London Overground so going beyond 8tph would appear to be difficult. If it turns out that providing the extra services on the Wimbledon Loop or a suitable suburban alternative is too hard then there could be capacity available at London Bridge. It would seem that it may well be available for additional fast services to Sussex if there was the demand for these.

What is significant is that most of the services that will terminate at London Bridge in future are local services. One is inclined to ask whether these really need a full 15 minutes turn around time. What if this could be reduced to just 12 minutes? This is not really an unreasonable possibility in the future, especially with improved signalling. Even with the temporary arrangements 27tph was managed at one point during the Thameslink Programme. Whilst it would not be reasonable to suggest that the extra slots are there for the taking, there may well be some wriggle room in the future.


In simple terms, if East Croydon weren’t a problem, it looks like one should be able to easily find two additional slots per hour for trains to and from Sussex if they were needed. If more were needed then there are further potential possibilities. As with our look at Victoria, it appears that the problem with capacity lies not so much at the London terminal but elsewhere. Any issues north of East Croydon to London Bridge are solvable. On the London Bridge route of the Brighton Main Line at any rate, it seems to be becoming more and more obvious that the critical problem is East Croydon.

Needless to say Unravelled has been photographing progress on the Bermondsey Diveunder for a while. The work currently being done is preliminary work and it is still hard to have a clear idea on the ground where the exactly the diveunder will be and what it will look like.

Written by Pedantic of Purley