Seemingly illogically, now is the time to look at Norwood Junction and its future importance in enabling an upgrade to the Brighton Main Line. Initially this station was not intended to form part of this series. Only now have we fully understood the critical role played by upgrading Norwood Junction as an essential early part of upgrading the Brighton Main Line.

What’s this got to do with Sussex?

The title of this series seemed to be a sensible one at the time of inception as we looked at how the Brighton Main Line could be upgraded. We would look at the London suburbs but we would also report on significant developments further south that create extra capacity on this vital route. That capacity is generally recognised as being needed as currently this main route is inadequate, capacity-wise, for the needs of the next generation.

It was expected that a lot of the focus would be within the London suburbs. In particular, it was realised that East Croydon and the junctions north of the stations would form the key to the entire line upgrade. What is becoming more apparent is how East Croydon impacts on even more than was initially appreciated by us. This is why, out of sequence, we need to include a look at Norwood Junction.


Norwood Junction track layout today

Norwood Junction not initially seen as a problem

Norwood Junction would not initially seem by us to need to be a vital part of the Brighton Main Line upgrade in its own right and we saw it as something minor that was part of the East Croydon scheme that was happening on the fringe of the area covered. In retrospect, and with more details emerging of Network Rail’s scheme, it seems to make more sense to consider it a separate portion of the entire Brighton Main Line upgrade in its own right.

Norwood Junction has five through lines – basically up and down, slow and fast with an additional down slow line to West Croydon. The station seemed basically fit for purpose as far as providing a train service is concerned and it certainly wasn’t seen as a particular constraint on the whole line.

It is true that there was a long-term desire to provide terminating capacity for local all-stations trains from the north but this didn’t really impact on plans for the Brighton Main Line and wasn’t considered urgent or vital. There was also a slight capacity issue that was caused by some trains on the fast lines making a station stop at Norwood Junction but capacity restrictions elsewhere meant that this was not of particularly great concern.

East Croydon design developments and consequences

The design process for redeveloping the layout at East Croydon appears to be going well. The work for this preliminary phase has been, for the most part, completed and the planning team has gone on to look at the proposed construction sequence.

It is now our understanding that the early reconstruction phase at East Croydon station looks likely to involve a period when there would only be four platforms available there instead of the current six and eventual eight. This is largely because of the need to relocate the westernmost two platforms further north than existing, due to the constrained geometry of the site.

Whilst closing two platforms and doing a London-Bridge-style slew to make best use of the remaining four platforms still open may seem to be an onerous restriction it is not as bad as it first seems. In the up slow direction you effectively currently only have platform 4. On the fast lines platform 2 has only limited capacity due to it being used bi-directionally and in any case there will be some limited opportunities to spread services more evenly between the fast and slow platforms. When a lot of analysis was done it was discovered that the station could work with only four platforms providing there were no terminating trains at East Croydon.

It follows from the last paragraph that before work can start at East Croydon, hopefully around 2020, there must be adequate terminating facilities elsewhere to replace those lost at East Croydon. This can easily be arranged on the Victoria side of the Brighton Main Line by terminating some trains at Selhurst and this can be achieved by installing a crossover on the double track spur leading into the depot. On the London Bridge side there may be some opportunity to terminate additional trains at West Croydon but realistically you need some decent facilities at Norwood Junction. From the south there are trains from Uckfield and elsewhere that currently terminate in service at East Croydon before continuing out of service to Selhurst Depot. Ideally you want to be able to run them to Norwood Junction in service and have sufficient capacity there to terminate the trains without delaying other services.

The implication of all the above is that, if you want to start work at East Croydon in 2020, you need to have an improved layout at Norwood Junction by that date. In fact, this all fits in rather well as the work at East Croydon will probably be subjected to some form of planning process (not necessarily a public inquiry) whereas it it believed that any work necessary at Norwood Junction could be done under permitted development rules and so it could commence much sooner.

Brief Station History


A uninteresting-looking public subway in front of Norwood Junction station entrance

Norwood Junction is a station with a much more interesting history than one might expect. It started off, not on the current site, with the name of “Jolly Sailor” – named after a local pub as there was not much else around. Originally its importance as a passenger station was almost entirely due to it being a multi-junction station. In different directions the next station at various times would be Selhurst, West Croydon, East Croydon, Beckenham Junction, Anerley and Crystal Palace. Nowadays the direct link to Beckenham Junction is gone and the link to Selhurst cannot be done with passengers on the train as it goes through Selhurst Depot.

Various railway books and others write about the Croydon Atmospheric Railway and the world’s first railway flyover at Norwood Junction. Possibly more relevant to today is the rather less famous (probable) world’s first use of reinforced concrete as a tunnel lining. The subway involved – a public one – is still there under the tracks and helps reduce the division of the local community that the railway created.


Depot and former Freight Yard

Apart from the station, there is important Selhurst Train Depot which could have equally well have been called Norwood Junction Train Depot as it is served by spurs from both stations. Lovers of trivia would be interested to know that it is one of very few places on the National Rail network south of the Thames and east of Maidenhead that currently has overhead electrification. This is in the form of a test track for Thameslink stock. The depot also gives away some of the early electrification history of the northern end of the Brighton Line by having a set of siding referred to as the “AC” sidings – which are in fact nowadays DC and have been for the past eighty seven years.

AC DC sidings

The AC DC sidings

There also used to be a large freight yard which even survived into the TOPS era. Its demise enabled the depot to be expanded but quite a lot of land has disappeared on the other side of the tracks in recent years for housing. Fortunately some railway land has been retained to the east of the running lines.

Station site constraints

To the north the station site is considerably constrained by a low height underbridge notorious for being hit by overheight vehicles including once, embarrassingly, a rail replacement bus. At the other end the restrictive multi-arched Tennison Road overbridge has been replaced in the past few years by a huge structure with plenty of clearance below and little in the way of obstructions to hinder any future layout. In the western direction expansion is hindered by the station and other buildings and in the east by Clifford Road.

Electric bus on route 312

Bus on fully electric route 312 at the Norwood Junction terminus in Clifford Road

A multitude of platforms

The station currently has seven platform faces (more than East Croydon!) and five through tracks. One platform face (platform 7) is clearly signed and maintained but the track is very rusty and overgrown. Platform 2 adjoins a track which is served by platform 1 and is no longer used. Platform island 2/3 is particularly narrow whilst islands 4/5 and 6/7 are not so constricted but they are not generously wide either.

Platform 1 plus former bay

A very wide platform 1 looking south from the northern end with evidence of a former terminating platform on the right

An inspection of the station will give strong clues as to how things were. There was clearly a terminating platform to the north of the station building adjacent to platform 1. Even if it could be reinstated, it is almost certainly too short to fulfil any useful purpose nowadays.

Station subway entrance on platform 1

The very wide platform 1 and the set back subway entrance

It is also clear that there were originally two tracks between platforms 1 and 2 . The large platform area on platform 1 with the subway entrance set well back is one major clue. Another telling, but not definitive, feature is a very short stub of what is presumably the old line and the presumed original platform face still present. One might think one could safely assume this was never used for passengers in its current state but this is not quite true as on one occasion many years ago, admittedly when the stub was longer than it is today, a DEMU service ran from this stub platform to Selhurst depot as part of an open day there.

Looking at the southern end of platform 1

Southern end of platform 1


Norwood Junction is nowadays a surprisingly busy station with over 6m passenger entries and exits per annum. Its practical catchment area is made bigger by the presence of a second entrance on the east in Clifford Road (generally known as the Woodside entrance) serving a large area that is otherwise generally devoid of a heavy rail service. One might not think this but, it is also a very important station when Crystal Palace play at home with many fans using it in preference to Selhurst.

Since the introduction of the East London Line service to West Croydon and the taking over of the running of the station by London Overground it appears to have surprising number of passengers using the London Overground trains to Canada Water and beyond – despite the trains being all-stations ones. The station is also served by Southern trains. In fact the majority of trains are Southern ones but despite this the station is run by London Overground. This leads to prominent notices about the temporary closure of the Gospel Oak – Barking line whilst more useful information to the average user at this station may be harder to find.

Most Southern trains are stoppers but there are generally 4tph semi-fast trains direct to London Bridge with some calling at New Cross Gate. As these semi-fast trains are run on the fast line they could potentially restrict capacity but it is probably fairer to describe these services as being a timetabling constraint rather than a capacity restraint.

A flawed infrastructure?

Infrastructurewise, Norwood Junction is not the most passenger-friendly station. Since the ending of slam door stock, platform 2 serves no useful operational purpose as trains stopping on the corresponding track have been banned from having doors opened on that side. Nevertheless the passengers on the platform are still exposed to the risks of the platform edge and passing trains. It is rather surprising that no kind of protective barrier has been placed on platform 2 for the safety and convenience of passengers using the adjacent platform 3.

north end of platform 2

The north end of platform 2 serves no useful purpose

Another quirk of Norwood Junction station are its yellow lines. These have clearly been relocated, sometimes more than once, and have in the past been quite impossible to comply with in places. The disused platform 7 has the obligatory yellow line as does an extension to platform 2 which serves no useful purpose.

Platform 3

Platform 3 showing former yellow line that would have been impossible to comply with and the extremely narrow stairs

Platform extensions at Norwood Junction also defy logic to the layman as the whole approach in recent years appears to be half-hearted. Platform 3 (up fast) in particular seems to be devoid of rationality. The extension at the north end has a very temporary feel and fails to continue as far as it could. Meanwhile, serving no purpose at present is an older extension to platform 2 that appears as if it could be widened to give platform 3 some of the extra length it needs when 12-car trains stop there. At the southern end platform 3 looks as it could be easily extended but this has not been done. Of course, one possible explanation for this is that Network Rail is reluctant to do the work knowing that the station is due to be redeveloped in the next few years.

temporary platform extension

Temporary-looking platform extension at the northern end of platform 3. It does not extend far enough north so SDO is needed. Note disjoint in yellow line.

The station subway

station subway

The station subway

The station has no passenger footbridge. Vital to the functioning of the station is the single internal station subway. One end comes out at platform 1 and the other at the Clifford Road entrance. The only platform that can be reached without using steps down to the subway is plaform 1 – the up slow.

The ambience of the subway has improved considerably since London Overground took over. In fact prior to that the subway and staircases were pretty awful. Any improvements to the width or the height of the subway would be extremely challenging. The design of the subway also effectively prevents any reasonably priced modification of the platform locations or their width. There are already stairs on either side of the subway to each platform island but the restricted width of the platforms prevents the stairs being wider with platforms 2/3 having very constricted access.

Access for All

The biggest criticism generally made about Norwood Junction station is the lack of any disabled access. Apart from anything else this means the station is also not at all buggy-friendly and the stairs are not pleasant to use if they are busy and you have luggage.

East London Line to West Croydon diagram

Penge West and Norwood Junction are the only stations on a large section of the East London Line that do not have disabled access

The lack of disabled access is particularly concerning as most other stations served by the East London Line are nowadays fully accessible. The other exceptions are generally below-ground-level stations where it would be extremely difficult and costly to provide what is generally known as Access For All. Like at Clapham Junction, at Norwood Junction this is probably more important for the people who are preventing from changing trains here rather than for originating passengers. It is not an easy problem to solve and, as one Network Rail manager said, “the politicians can say what they want but if the platforms are too narrow to accommodate lifts then your options are pretty limited”.

The saga of platform 7

Readers may have been somewhat surprised that platform 7 is in such good condition, clearly potentially useful yet still unused. The history has been a long one and, as is often the case, it is the cost of signalling and the difficulty of finding resources to do the job that lie at the heart of the saga.

platform 7

Platform 7 complete with yellow line in the wrong place all ready for the next train to stop there – many years from now

The idea of terminating trains at platform 7 really developed when it looked quite likely that Tramlink would be extended to Crystal Palace. The view was held that the trams might take over the remaining heavy rail line from Crystal Palace to Beckenham Junction. Platform 7 at Norwood Junction seemed to be an ideal place to divert the service that terminated at Beckenham Junction. The real advantage was that, by also reinstating a long disused track to the north of the station, the trains could terminate at Norwood Junction and make their way back to Crystal Palace without using any of the existing four tracks of the Brighton Main Line.

platform 7 looking north

The view looking north from the end of platform 7

Tramlink wasn’t the only reason for the enthusiasm for reinstating platform 7 and it could also be seen that this would allow various other useful options e.g. a train could run in service slow from London Bridge to Norwood Junction then run out of service via Crystal Palace to get back to central London and complete another trip in the evening peak.

platform 7 looking south

Another view of platform 7 – this time looking south

The problem with the Norwood Junction turnback scheme was that the signalling was prohibitively expensive. It was recognised that the only way the scheme would be viable would be if it happened when the area was being resignalled anyway – as it will be prior to the resignalling of East Croydon. And so it was that what was originally seen as an easy-to-do cheap upgrade option has been around for ten years waiting for a suitable opportunity to implement it.

A new layout


Potential future layout at Norwood Junction immediately prior to works at East Croydon

As can be seen, a new track layout is proposed for Norwood Junction. The idea seems simple and eminently sensible. Lengthen platforms as necessary. Reinstate platform 7. That gives six through tracks which can be split into pairs of fast, semi-fast and slow trains with some flexibility added for situations such as terminating trains, trains needing to access Selhurst Depot or new stabling sidings on the east side of the tracks or near simultaneous arrival of a slow train to West Croydon and East Croydon.

An alternative to the above plan is to actually use platform 1 as the terminating platform but, given the usefulness of this platform being the one platform that can be accessed without using the subway, this would seem less satisfactory as it would lead to increased use of the narrow platform 3.

Onward and Southward

With things dynamically changing it is hard to be sure what will be covered next in this series but the intention is to look at the importance of Stoats Nest Junction just south of Purley and how Network Rail proposes to resolve issues there.

Thanks to ngh for creating the diagrams and other assistance. Thanks also to Network Rail for helping us with our understanding of development to date.

jump to the end
There are 277 comments on this article
  1. KRW says:

    I know Goldenballs is famous but I think Beckham Junction is actually Beckenham Junction

  2. Timbeau says:

    Reference to “Beckham Junction” in the history section. Also Anerley is missing from the list of next stations.

    [Corrected Beckenhan Junction. Can’t believe I missed the really obvious inclusion of Anerley. I travel past that station and Norwood Junction often enough. PoP]

  3. Korot says:

    “In different directions … Crystal Palace” The connection towards London bridge is missing from this summation, it would appear. Otherwise, nice article. I wonder if, with a closure of platform 2, the narrow stairs to platform 3 can also be widened, using the width of platfrom 2?

  4. Tim says:

    In other junction name business, I notice from looking at this area on map that the junction near Birkbeck station is labelled “Birbeck Junction”. Is this a typo there, or does it actually have that slightly different name?

    [Typo. A few errors like this have cropped up. Considering the guy who creates the map is French it is surprising there aren’t more. PoP]

  5. Lemmo says:

    Interesting article, thanks PoP.

    I see from the 1895 OS map that there was a platform on the east side, and the remains of the steps access to this and the platform buildings can be seen in the pictures. Any plans for, or advantage to, rebuilding this?

    Also, when they rebuilt Tennison Rd bridge, did they widen the cutting on the west side to allow for six tracks under the bridge? That would have allowed Up Crystal Palace trains to be segregated from the mainline tracks. As services intensify I can see benefits to this.

  6. Vince says:

    In a Pedantic of South Croydon mode – ‘Tennison Road’ not ‘Tennyson Road’ – named after an Archbishop of Canterbury (like a few other roads in Croydon) rather than the poet

    [Funnily enough I was correcting this and then got your comment. PoP]

  7. ngh says:

    Re Lemmo

    No advantage to rebuilding the east platform as it would be more expensive and worse for cross platform interchange as P6/7 needs minimal work.

    Tennison Road Bridge was widened for 6 running lines and 4 tracks for depot access with full 5.7m OHLE clearance and 2 at reduced clearance to the east.

  8. MiaM says:

    If a few more switches were added the five track section could have two modes. One of the modes would be the proposed usage in the afternoon peak with separate down slow lines from London Bridge and Crystal Palace. The other mode would divert traffic so the track usage from west to east would be up Crystal Palace, up slow London Bridge, up fast, down fast and down slow Crystal Palace plus London Bridge. That mode would probably be useful in the morning peak when you want maximum capacity for northbound trains.

    It would cost a lot more switches but it would give more flexibility.

    Technically the best would be what Lemmo hints at, six track all the way from the station up to the actual junction.

  9. Briantist (in Gigabit internet heaven) says:

    Thanks Pedantic of Purley for a very interesting article.

    Can I ask a question about the station subway?

    What are the problems with creating a new, wider subway somewhere else within the station? Can’t it be done at all for some reason, or is it just a question of cost?

    And if it can’t, why is it so impracticable to create a passenger bridge over the station with new lifts and stairs?


  10. Briantist,

    Adding a second station subway with access to the platforms would not be easy. It would not be just about cost. To make it decently wide and not too deep I would have thought you would need some kind of concrete box construction and closure of the tracks above whilst in progress. Remember this is a very busy route. It might be worth reading the link I gave to the public subway to appreciate the difficulties. Also bear in mind one wants to get the work done quickly before progressing to East Croydon.

    No reason in principle why you can’t have foot overbridge. There would be considerable problems in accessing the narrow platforms 2/3. So you are probably talking about stairs or lift.

    I didn’t state it in the article but, from observation, one of the biggest problems seems to be in the evening peak when a lot of people alight from a semi-fast train on platform 3 that previously stopped at East Croydon. They try to funnel down two narrow sets of stairs but, inevitably, people are amassing on the platform as the stairs are unable to cope with the sudden demand. It wouldn’t be fun if a second train arrived in that time. Meanwhile, sometimes, some of the people on that island platform can literally touch the train they want to catch next but are unable to board. By the time they get to platform 1 the train is gone.

  11. Geoff Smith says:

    The ‘stub’ at the end of Platform 1 used to be rather longer, extending the length of the Up Gullet to Up Local crossover. It had indeed been part of the former platform. In 1872 there were 10 lines through the station – listed as between platforms :
    Local Up , Pimlico Up.
    Up and Down Main
    Pimlico Down , Local Down
    ‘Spur Siding’ , LCDR
    Two outside the LCDR platform.
    At least the first six dated from the opening of the resited station on 1/6/1859.

  12. Briantist (in Gigabit internet heaven) says:


    Thanks. Of course I’m wondering how deep (or what other provisions you might need to make) to be able to tunnel (and presumably use Spraycrete) a new subway without having to stop the trains.

    Or, alternatively if there is a suitable technology that could build a 40 meter long subway “overnight”, if the budget was there. I’m thinking that Crossrail did their “eye of the needle” thing next to a working tube line here.

    I’m asking out of general interest – this is also an issue for East Croydon.

  13. Briantist (in Gigabit internet heaven) says:

    … and/or Clapham Junction.

  14. Greg Tingey says:

    The whole idea seems sensible & do-able at (given today’s inflated standards) reasonable cost.

    Two historical notes – about 20 years back I actually used pf 7 & went up the now-disused, or even oou route to join the Dn slow towards E Croydon …
    Even further back, there used to be a steam freight shed in between the Birkbeck loop line & the main, which was easy of access.
    One could usually guarantee to see Maunsell moguls, ex-LBSC “K” class 2-6-0’s, Maunsell freight tanks (Class”W”) & the inevitable Q-1
    I have the pictures to prove it, too …..

  15. Malcolm says:

    Briantist: Technologies for building a tunnel without affecting trains above do exist: box jacking and pipe jacking are two possibilities. A 25m wide road tunnel was recently built in East Kent under such a constraint: see Atkins description here. The report says nothing about the cost, which would probably be the biggest constraint at Norwood Junction. And of course ground conditions are different, but the message is probably that engineers can build (almost) anything if there is a suitable budget.

  16. Herned says:

    @ Briantist

    Bridges and subways are regularly pushed into place either through embankments or where embankments have been temporarily removed. There are plenty of time lapse videos of that sort of thing on the internet. So assuming there was space to build once just offsite then that is what could be done. But it’s a lot more complication and effort than putting in a new bridge over the tracks, as the foundations can usually be put in without closing the lines, and then the bridge can be built in a factory and delivered complete and craned in. I would assume that unless the wind changes whilst the crane is in place there is a lot less likelihood of things going wrong, and also it is a lot cheaper

  17. JayKay says:

    If the platform 2/3 island is unduly narrow, and platform 1 is unduly wide, then why not reduce the width of platform 1, move the track across, and widen the island, thus giving space for stairs and a lift up to an overbridge? Or am I missing something…

    Shouldn’t be too difficult or expensive I would have thought – you’d have to close platform 1 while you did it though.

  18. marckee says:

    PoP: Having worked on the Norwood Junction station upgrade project when London Overground took it over, from memory the platform lengths and their use (and the bloody yellow lines!) are a legacy of numerous decisions, some of them intended to be permanent, and some of them intended to be temporary.

    These included things like allowing for short-turning of longer than average trains, especially in emergencies (a similar capacity was retained at Bushey too, on the WCML), and allowing for an additional stop in the event that engineering works had closed platforms, or entire stations, further up the line (this idea was floated again recently, with the Thameslink works).

    At the time (this was around 2010-ish), understandably, retaining the platforms made sense, as it allowed for the disruption of other upgrade work to be mitigated. It was also unclear exactly what services would be running through the station when these were completed and therefore a long-term strategy and reconfiguration was put on the back burner.

    Oh, and thanks for the compliment regarding the improved ambience of the station!

  19. Pedantic of Purley says:


    Another great improvement in my opinion is the glazed roof over platform 1 which makes the whole area are much lighter and more welcoming (clearly visible in picture of steps down to station subway). Also it is nice not to have the rain pouring through as used to be the case further down the platform. I don’t know if this improvement was due to TfL or Network Rail (or even Southern Railway).

    Also, must mention the increased staffing levels which make a difference. Noticeable that, whilst Southern have plans to close many ticket offices nearby, as far as I am aware, TfL have no plans to do anything similar at Norwood Junction or any other London Overground station.

    Finally, lighting is much better but that is a general improvement we see nowadays with white LEDs rather than the awful orange sodium predecessor.

  20. ngh says:

    re MiaM @ 04:42

    If a few more switches were added the five track section could have two modes. One of the modes would be the proposed usage in the afternoon peak with separate down slow lines from London Bridge and Crystal Palace. The other mode would divert traffic so the track usage from west to east would be up Crystal Palace, up slow London Bridge, up fast, down fast and down slow Crystal Palace plus London Bridge. That mode would probably be useful in the morning peak when you want maximum capacity for northbound trains.

    It would cost a lot more switches but it would give more flexibility.

    Technically the best would be what Lemmo hints at, six track all the way from the station up to the actual junction.

    1. The actual junction after which Norwood Junction is named is Norwood Fork Junction to the south as the junctions to the North (and the connections to Crystal Palace and Birkbeck) were added 3 years after the station was named Norwood Junction. As noted in the text and drawing this section to the south will be 6 track as this is were it is needed. (also see Sussex part 7 drawings with 6 tracks south of Norwood Junction)

    2a The 2 up tracks on the five track section aren’t the limiting factor even after rebuild East Croydon and possible enhancements at West Croydon and the stopping services can be regulated in the platforms.
    2b There isn’t room for the extra switches
    2c The aim is to have line speeds for everything at 60 (slow) /70 (fasts) mph to maximise overall capacity.
    2d The cheaper solution would be to rebuild Penge Road overbridge.
    2e Sorting Up trains pre arrival into Crystal Palace and London Bridge would reduce capacity as they will be crossings on the flat south of the station the train are far better off having clear run into the station then swapping afterwards with wait in the platform if needed. The arrangement is set up for segregating arrivals from the south for maximum efficiency in the peak.
    (Note similarities of proposal with final up Charing Cross track arrangement at London Bridge which will achieve 28tph!)
    2f Most Up services will be through P1 stopping or P4 not stopping with P3 much less used for slow arrivals from East Croydon that coincide with arrivals from West Croydon or stopping Thameslink services with just few tph will make big difference overall. (A sensible compromise might be to run some fast service through P3 to make use of the wider P4 for stopping Thameslink services)
    2g Maximising efficiency of the Windmill – Gloucester Road – Selhurst – Cottage – Norwood Fork Junction complex post rebuild will see more services from Palace to East Croydon than at present hence the 2 track up section helps optimise this as they will arrive at P3 from East Croydon. (Restoring what happened more than decade ago). A cheaper alternative to grade separating Gloucester Road in the rebuild and makes 6tph LO to West Croydon far easier and improves the frequency between Norwood Junction and East Croydon.

  21. ngh says:

    re JayKay @10:08,

    You aren’t missing anything and prefectly possible to do it in the later part of rebuilding and resignalling (it would be very hard to do now).

    Re PoP,

    TfL apparently have plans to “underground” overground ticket offices, stations still staffed though.

  22. Sykobee says:

    I use the station daily, for the semi-fast to London Bridge.

    The stairs on Platform 3 are highly inadequate. The station is quite unfriendly is you want a platform other than Platform 1.

    I think the best solution is an overbridge at the south end of the station, and fencing off Platform 2 (and renumbering them subsequently) so that you can use its width for the overbridge lift and stairs.

    I am presuming that Platform 4/5, which are the through-fasts in the new plans, will be pretty much unused except for semi-fast from London Bridge (but I suspect these will use P6). Unless … Platform 4 (nice and wide) becomes the semi-fast platform, and Platform 3 is the through-fast. That would pretty much eradicate the Platform 3 issue. However you can’t reverse a train on P3 in that case (as the diagram indicates).

    In the end the most telling thing about the plans is that they aren’t moving the platforms into a more suitable position, even though the temporary nature of previous works indicates they were considering it. This means a wide platform 4/5 with decent passenger facilities will be practically unused whilst the narrow P3 may have more traffic (semi-fast and terminating). Of course this cuts the cost and disruption significantly, the works are reduced to new track and points installations.

    Google maps shows a 13 (?!) carriage train conveniently to demonstrate the inadequate platform lengths: https:[email protected],-0.0742598,370m/data=!3m1!1e3

  23. marckee says:

    PoP: Re. the lighting: There was a fair bit more to it than just a change of luminaire. That round of upgrades included a new corporate identity, yes, but also some detailed, and genuine improvements to ‘passenger experience’, including the installation of waiting shelters, new PA system, new information signage and real time displays etc.

    The lighting was remodelled and redesigned from the ground up to improved RfL standards, with new columns and column positions on the platforms, in the booking halls and, where possible, on the station approaches too.

  24. Sykobee says:

    Or the more worrying conclusion, no semi-fast services (Southern or Thameslink) in the future… but they do have the Fast->Slow points south for P3/4 so hopefully that isn’t the case.

  25. Sykobee says:

    Ah, I also see the Fast Down to Fast Up points to the north of the station will allow termination of London Bridge -> Norwood Junction Fasts on Platform 3 for reversing back to London Bridge.

  26. Geoff Smith says:

    The “terminating platform to the north of the station building” is shown as “platform siding” on 19th century diagrams, but had no connection from any of the Down lines, so couldn’t have been used to terminate a train. It was signaled for departures, and most likely it was used for milk traffic &c..

  27. ngh says:

    Re Sykobee,

    As noted above I suspect mixed mode compromise for P3 would achieve reduced passenger use of P3.

    – most fasts including all stopping,
    – some fasts (not stopping but with no speed dis-benefit) e.g. those following something stopping in P4
    – slows from East Croydon that arrive at roughly the same time as a service from West Croydon into P1
    – services in the up direction terminating and turning back into the depot
    – down services passing through from the up fast to the depot
    – most most slows including all from West Croydon and a good number from East Croydon

    I suspect the down London Bridge services that stop will use P6 as this gives cross platform interchange with P7 and the access to the slow lines at East Croydon and West Croydon which is where those services go

  28. Henning Makholm says:

    I can see how platform 7 in the proposed layout (and the turnback siding to the south) will work well for terminating trains from/to Crystal Palace, but could somebody explain which trains platform 3 would be able to terminate?

    The only “down” line it seems to connect to in the diagram is the down fast from New Cross Gate, but there’s not really a demand for a mostly non-stop service just between London Bridge and Norwood Junction, right? Unless I’m missing something, the only southbound movement platform 3 could possibly be used for with this layout would be ECS arriving from LBG on the down fast and continuing to the depot, and ECS from the south reversing into the depot.

    The beginning of the article hints at terminating local all-stations trains from the north, as well as trains from the south (Uckfield), but neither of these seem to be possible with they proposed layout. Even taking a train out of the depot and running it empty to East Croydon to begin an Uckfield service there seems like it would be possible only by blocking the up slow line by a backwards movement down to East Croydon. (But I suppose such trains could reverse at Selhurst instead).

    Are additional points at Windmill Bridge or Norwood Fork junctions envisaged before this layout is useful?

  29. Sykobee says:

    Yeah, if they can use P4 for semi-fasts that would be great, there doesn’t seem to be anything stopping that with the layout. P3 is as you say, through-fast for overtaking semi-fast, double-slow in station, [double semi-fast which currently uses P1 and P3 but hopefully this should be timetabled out with the work being done], or terminating semi-fast for reversal/depot.

    As long as they allow fasts to overtake the stopped semi-fast which currently means the semi-fast is delayed before NJ and passengers on P3 just see a succession of through-fasts as the indicator board shows their train getting later and later, then some of the massive morning delays for these services should be prevented.

  30. ngh says:

    Re Sykobee,

    “double semi-fast which currently uses P1 and P3 but hopefully this should be timetabled out with the work being done”

    This should disappear with completion of London Bridge works (never underestimate GTR’s timetabling ability though!)

    Re Henning,

    P3: Terminating anything heading back into Selhurst depot more likely to be the case for services coming from the south as they have to change direction. The services coming from the north are likely to run empty and fast from London Bridge straight through P3 without stopping then into the depot arrivals road.

    Note the reinstatement of the Norwood Down sidings to the east of the line to East Croydon to provide more stabling hence there is less need for down slow to main depot site moves. The other down slow alternative is to run the services to West Croydon and terminate there the run back towards Norwood Junction but enter the depot on the way as this is far less disruptive overall.
    ” (But I suppose such trains could reverse at Selhurst instead).” That looks like the plan and happens for some moves at the moment, reversing in Norwood Junction at the moment can really screw things up in the morning.

  31. Geoff Smith says:

    Complaints about the subway go back at least 120 years, to a time when problems were often solved in months rather than years.
    Col Addison reported to the BoT in 1895, the response to which, from the LBSCR, was a proposal to erect a footbridge and give the subway over to public traffic. An alternative, suggested by the BoT, was a wider footbridge, divided for public and passenger traffic, with the complete closure of the subway. Correspondence went on until 1903, but the sticking point seems to have been the refusal of Croydon Borough Council to contribute any funds, even though they were the main complainants. The BoT reported that they had “no statutory powers under which they could call upon the Railway Company to improve the subway.
    According to the Report, at that time there was no subway access from the Up Local, so passengers needing to go from the Booking Office to the island platforms had to go outside the station to access the subway. It was no more than 6 feet wide and, at the top of the stairs to each platform there were large double doors which were kept locked until a train came in.

  32. Briantist (in Gigabit internet heaven) says:


    Thanks for the link to the Cliffsend Underpass document. It is interesting that the tunnel needed for pedestrians at Beckenham Junction would only be as lonh as the Cliffsend is wide!

    @ Herned

    Thanks too. It would be interesting to know how you might price up the various options. Would the low price of steel today provide a different solution to one where the price was “normal”?

    I’m also wondering about the depth you would need to make it, and how that would work for wheelchair-safe access slopes. This would be a better solution perhaps than a bridge that has be built high enough to go over (future-proofing) overhead wires.

  33. Saintsman says:

    It is fairly outrageous that Norwood Junction will not provide step free access after the capacity work.

    I presume that all platforms will be extended to allow (flexible) 12-car operation which will then remove the requirement for SDO.

    I don’t understand why platform 2 is being retained. Providing protection from this edge allows platform 3 to gain sufficient width to safely deploy boarding ramps. Removing much of the platform clutter would also help (e.g. Move shelter against the old platform 2 edge).

    The Clifford Road entrance is far from ideal for Access for all. A new building connecting directly to a new footbridge would improve the gate line. Such a bridge, needs both stairs and lift access and preferably joined onto the canopies. Which makes the southern end more likely, but this makes platform 4/5s waiting room in the “wrong place” (it will have little use). Platform 3 remains a challenge to provide 2.5m clearance and a sufficiently wide stair structure (even with platform 2 face closed). Potentially the buffer stop at the end of platform 1 would also be lost to allow this bridge to be constructed.

    The proposed Platform 6/7 arrangement seems a sensible.

  34. ngh says:

    Re Saintsman,

    Ideal time for Croydon Council (and local MP?) to press DfT to include Norwood Junction in the Access for All scheme next time round?
    Getting Norwood Junction done by the end of 2019 will need a good rummage down the back of the sofa for funds anyway!

  35. tjw says:

    Have found an interesting map, OS Five feet to the mile, London, 1893-6, the link should focus in on the Station

  36. Walthamstow Writer says:

    Having been in an Overground train on P1 when another train stopped at P2/3 and the alighting passengers were pressing the door buttons, banging on the doors and looking extremely exasperated can I ask why door opening on P2 has been banned?

    I don’t really know Norwood Junction as an area but recently tried to fathom how bus stops related to the station and I came away thoroughly confused as to how the place works! Perhaps I need to have a visit on foot to try to orientate myself?

  37. FLHerne says:

    P1 is much better for most passengers, because of the direct exit and wider platform/stairs.

    Opening on both sides would need someone to watch all those doors; GTR are having enough trouble over dispatch procedures as it is!

  38. ngh says:

    Re WW,

    The rule is doors only open on 1 side…

  39. Walthamstow Writer says:

    @ Ngh – and yet LUL manages this with no undue issues at Barking e/b, White City and Stratford Central Line w/b [1]? Given LU has issues with massive passenger flows, small platform sizes and high service frequency requiring controlled headways it strikes me as bizarre that the national railway can’t do something similar with 5 car trains at one set of platforms. I have yet to see anything that suggests that dual side door opening is an inherent safety risk. If it was then LU would surely have been prevented from adding platform 3A at Stratford.

    [1] I’m not suggesting this is a comprehensive list in case anyone goes into pedantic mode.

  40. Geoff Smith says:

    The extension to platform 2 was brought into use on Mon 9/12/85 – “doors [of 455 units] must be released on both sides of the train.”
    Presumably the possibility of a passenger stepping out of a SUB or EPB into thin air was not previously regarded as a danger.

  41. Graham Feakins says:

    In the days when Platform 2 was in use and doors were openable on both sides of a train at Platform 1, there was a habit for nifty passengers alighting from a down service on 3 to get on a train on 2 and alight the other side onto 1! Of course, the opposite happened as well. This avoided having to use the stairs and subway.

  42. @Graham Feakins

    Health & Safety probably have a rule against this tactic…

  43. Pedantic of Purley says:

    The reason that platform 2 cannot be used is, I believe, (wait for it) because you cannot have a platform in use where there is a live conductor rail adjacent – except on LU where you can for some reason. So it would not be permitted to have the doors open on platform 2 under any circumstances with the live rail where it is.

    I presume the theory is that there is a danger of waiting passengers falling on the platform and being electrocuted. Obviously there is a bit of a flaw in this argument when applied to Norwood Junction. At Horsted Keynes, admittedly unelectrified and served by slam door stock, passengers are permitted to board from either platform.

    Note that at the centre platform all platforms at Canary Wharf on the DLR the doors are permitted to be open on both sides. There is no question of grandfather rights here, the reason it is permitted is because falling on the bottom-contact conductor rail would not be fatal.

  44. Geoff Smith,

    I think the legal nicety of the argument is that if you open the door it is (or at least was) up to you to check it is safe to exit. If the door is opened for you then you have a reasonable expectation that this would only be done if it was safe – it is, in effect, an invitation for you to walk through the opening made.

  45. Timbeau says:

    As it was said up thread (Geoff Smith) that the doors did open on both sides on 455s, there must be more to it than PoP’s comment. Is it actually possible to open the doors on both sides of a 378?

  46. Pedantic of Purley says:


    Not at all. It is my understanding that the rule came in after the 455s were introduced. Obviously there was no point in having the rule prior to that as it would have been unenforceable.

    I could be wrong, and was working from memory and what I understood at the time. I also suspect that there may be places up norf where they only allow one side to open despite no or overhead electrification but that could be for all sorts of reason related to monitors. Canary Wharf DLR still seems to me to show definitively that it is not an outright ban.

  47. Anonymous says:

    It may not happen now, but I’ve been on a SWT service at Ascot when the doors on both sides were opened. It was on a 458 and it was noticeable that the open door chimes on one side were a different tone to the other side, which I assume is to assist partially sighted people.

  48. John Elliott says:

    Since the trains at Horsted Keynes have slam doors, I think it would be quite hard to stop passengers opening them on both sides.

    I have been one of the porters dispatching trains from the double-sided platform at Horsted (it takes at least two, one on each side). Of course, that’s a rather less passenger-rich environment than Norwood Junction.

  49. AlisonW says:

    Somehow I feel that platforms 1 & 2 would be a fine example of the Spanish option …

  50. Pedantic of Purley says:


    Nice idea but not really.

    Problem 1: For most trains it doesn’t solve anything. People tend to be getting on rather than getting off. Not a lot of people catch a train from West Croydon to Norwood Junction.

    Problem 2: Absolutely unenforceable. Why struggle to use platform 2 when you could simply use platform 1 which is much more convenient?

  51. ngh says:

    When introduced the 455s had guards for the first 6 years on the central division so that might have aligned with that change.
    DLR has safer bottom contract 3rd rail.
    Bombardier units on Southern and LO have correct side door opening so it is probably impossible without manual override.

  52. Geoff Smith says:

    @ Pedantic of Purley
    “The reason that platform 2 cannot be used is, I believe, (wait for it) because you cannot have a platform in use where there is a live conductor rail adjacent “.
    Clearly that wasn’t the case in 1985, so I wonder when that rule appeared, and when opening the doors for platform 2 ceased – I note ngh’s comment.

    There are two tracks at Morden with platforms both sides, one being used for arrivals the other for departures.

  53. Ian J says:

    @PoP: I also suspect that there may be places up norf where they only allow one side to open despite no or overhead electrification

    From recollection Crosscountry services in the west-facing bay platform at Reading station pre-rebuild started opening doors only on one side despite there being a platform face on either side of the bay. This on a non-DOO route of course so no monitor issues. There are probably other examples on non-electrified lines.

    Bombardier units on Southern and LO have correct side door opening so it is probably impossible without manual override.

    On the other hand LUL use correct side door opening too – presumably LUL’s system is programmed with “left”, “right”, and “both” options. The conductor rail is much closer on the tube lines too.

  54. Timbeau says:

    @Geoff Smith

    WW at 1242 provided a non-exhaustive list of other LUL double sided locations. but there seem to be no LO locations.

  55. Fandroid says:

    Whatever happens on LUL, it’s a sensible rule to ban door openings where there is a live rail adjacent to the platform. There are not that many deaths to passengers on the railway, and to add a bit more (serious) risk would being going the wrong way in terms of Safety First. The only question is – why isn’t that platform face protected by a barrier?

  56. Briantist (in Gigabit internet heaven) says:

    @AlisonW: It is interesting that if you travel from Loughton to Cutty Sark DLR (say), you might think that the Spanish solution is normal, as you would see it at Loughton, Stratford and Canary Wharf DLR .

  57. Just to live up to my name, the Spanish solution is generally used to refer to the idea of separate entry and exit platforms. Merely having platforms on both sides isn’t the same and doesn’t have the same benefits.

    As far as I am aware, on railways, the only place this happens in the UK is Tower Gateway. Of course it is common in other modes of transport( e.g. cable cars) and some buildings (theatres, cinemas).

  58. Malcolm says:

    Ian J says “The conductor rail is much closer on the tube lines too.”

    (I assume this relates to the platforms being lower).

    Since the danger in question presumably involves falling from the platform onto the conductor rail, I do not see that a fall of three feet from a mainline platform is any more or less likely than a fall of two feet from a tube platform. (Both distances approximate).

    I suppose the shorter distance could be relevant if the mishap being imagined is dropping a metal-handled umbrella into the gap between train and platform, and being electrocuted by the umbrella handle. This umbrella scenario could also be a (rather far-fetched) explanation for the Norwood Junction refusal to open doors on platform 2 (which, as PoP indicates in the article, appears illogical, as the falling danger there exists regardless of train door opening rules).

  59. Pedantic of Purley says:


    I suspect the point is that on Network Rail you could come in contact with a nominal 750V adjacent rail which is likely to be fatal and on LU you come into contact with a nominal 210V adjacent rail which will probably mean that it will ruin your day but you are quite likely to still be alive. And of course on the DLR you come in contact with a protective plastic coating on top of the rail.

    I ought to add that I am fairly sure any live rail adjacent to a platform was not normally permitted on National Rail but if it was unavoidable then it always was the case that it had to be sandwiched between wooden boards of greater height.

  60. marckee says:

    The only time I’ve seen it in recent years on NR track is at Sheffield. From looking at the station layout, I think it was platform 2c or platform 7, and the train was an old Class 142.

  61. hilltopper says:

    Given how well connected Norwood Junction is to services which don’t call at East Croydon (the Overground, the Wandle Valley line via West Croydon, the Outer South London Line via West Norwood etc.), would it not make sense for the Thameslink Fasts to call there in any case? As that would seem to improve connectivity across a wide area of South London, with relatively little disbenefit to Sussex-to-City commuters.

  62. Malcolm says:

    PoP: Is this another Wikipedia fail? According to the said source, the side rail on LU is nominal +420V, and the centre rail -210V.

    I suppose that even if WP is right, your point would still stand that the voltage and therefore shock potential is lower on LU.

    The ruling about “unavoidable” shows the weakness of such terms. A conductor rail adjacent to a platform can always be “avoided”, if necessary by demolishing the platform when the line is electrified. But typically this form of avoidance would be very expensive and/or inconvenient.

  63. Maybe Wikipedia is right. I thought it was lowest potential difference on the outside because that would be nearest the metal tunnel linings and so create the least corrosion (which was the main point in having the fourth rail).

    If it is 420V it is hard to explain why it is acceptable on LU and not on NR. I can’t even argue grandfather rights as that wouldn’t explain Stratford platform 3 (Central line).

  64. hilltopper says:

    @ngh “2g Maximising efficiency of the Windmill – Gloucester Road – Selhurst – Cottage – Norwood Fork Junction complex post rebuild will see more services from Palace to East Croydon than at present”

    Currently there are no such services – at least, none stopping. Some nominally “fast” Thameslinks run very slowly without stopping through Herne Hill, Tulse Hill, Gipsy Hill & Crystal Palace to East Croydon, but I assume this is due to the London Bridge works. Do you think it likely that this stopping service (discontinued prior to the ELL Overground, IIRC) will be reinstated?

  65. timbeau says:

    I understand the potential difference between the positive and negative rails is 630V, but the actual voltage relative to Earth in each rail can be varied – for example on joint LU/NR lines the negative rail is held at about 0V. So it would be possible at double-sided platforms for the positive rail on the platform side to be at 0V and the negative at -630V.

    This is clearly not possible in 3-rail territory.

    @PoP 0904
    Spanish solution; According to Geoff Smith (0718) it is also used at Morden.

  66. ngh says:

    Re Hilltopper,

    Yes, expect to see X – East Croydon – Palace – Tulse Hill – London Bridge making a comeback in the 2018 timetable consultation very soon.
    Much easier from Jan 2018 when you have removed 4tph from the Norwood Junction – Tulse Hill section in the peaks that have been there for 28 years and the 4tph off peak which joined them for the LBG rebuild.

    It would be good to see more fasts stopping at Norwood Junction in the future post rebuild as it vastly improves connectivity for South Londoners but it looks like the 2 prime candidate services to do that, Caterham and Tattenham Corner won’t have Thameslink services.

  67. Pedantic of Purley says:


    Use of particular platforms for boarding and alighting at Morden may be encouraged or happen anyway but, as far as I am aware, there is no requirement or restrictive signage requiring one to do so. See this photo from Wikipedia with clearly visible “Way Out” signs on platforms 2, 3 and 4.

  68. hilltopper says:

    Thanks ngh, useful to know.

    Would it make sense, longer term, for Caterham and Tattenham to be run as semi-fast TfL services, maybe with a stop at New Cross Gate? All-stops trains would mean impractically long journey times for those at the end of the line – I’m thinking similar model to the Metropolitan, where it’s basically fast out to Harrow or thereabouts but stops in a couple of places with particularly useful connections.

    The down side being that, to avoid holding up Thameslink fasts, they’d need to switch to the slow line before NXG – meaning the slow up platform would have to handle ~10 Overgrounds, plus Caterham/Tattenham, plus any remaining West Croydon or VIC-LBG services.

    Still reckon the Sussex fasts should stop at Norwood Junction – especially off-peak, where there’s a much greater variety of journeys being made than the predominant commuter-dormitory-to-Zone-1 pattern during rush hour.

  69. Henning Makholm says:


    So when the article says,

    It is true that there was a long-term desire to provide terminating capacity for local all-stations trains from the north but this didn’t really impact on plans for the Brighton Main Line and wasn’t considered urgent or vital.

    then apparently this is still not urgent or vital, because the proposed layout would not provide this capability, right?

    And when it says,

    From the south there are trains from Uckfield and elsewhere that currently terminate in service at East Croydon before continuing out of service to Selhurst Depot. Ideally you want to be able to run them to Norwood Junction in service and have sufficient capacity there to terminate the trains without delaying other services.

    then the trains in question would run in service to Norwood Junction P3, then run empty through the depot and reverse again at Selhurst P1, and enter service either there or at East Croydon for another trip to Uckfield?

    (By the way, in the current timetable it looks like there is exactly one train a week from Uckfield that terminates in service at East Croydon, namely the last train on Sundays — and that doesn’t sound like something that will throw East Croydon into a chaos. Perhaps PoP had a different southern origin in mind here? There are a few weekday evening trains from East Grinstead that end at East Croydon.)

  70. Nameless says:

    Thanks for the link.

    I observe that at the time it was drawn, the actual platforms were numbered 1 – 4. This contrasts with the current convention where the numbers are given to platform faces.

    Also interesting to note that housing land near the station was still being developed.

  71. Henning Makholm,

    Sorry, I was being very sloppy.

    There are trains that terminate at East Croydon, prior to going to the depot. Some of these have come from further south. Others (in particular the “Uckfield” (class 171) trains) have actually come from London Bridge and have terminated at East Croydon.

    I should have written something like”there are class 171 (Uckfield) trains that have terminated at East Croydon on their final leg from London Bridge and other services from the south that terminate at East Croydon.”

    If I get time I will reread and reword it tonight.

    Combined, there are quite a few of these and it would be helpful if the ones from the south could continue in service to Norwood Junction. Even quite late at night there can be a potential for conflict at East Croydon platforms. With the plaform rebuilding you don’t want to be tipping passengers out if you can avoid it so having the ones from the south contining to Norwood Junction in service may be helpful.

  72. ngh says:

    Re Henning,

    On the first point there are 4 sensible alternatives:
    1. Run to West Croydon then back to the depot (several options on route back from West Croydon)
    2. Run ECS from LBG on the fasts then through P3
    3. Use reinstated Norwood Down sidings
    4. Turn back in P7 and run the train to Streatham Hill.

    And the option to add points in later for down slow to fast provided the interlocking is created at resignalling.

    Uckfield (and East Grinstead, Caterham, Tattenham Corner) northbound into P3 and reverse into depot for several hours off peak / over night out of service for maintenance cleaning etc.
    Most trains are timetabled to enter service Northbound from Selhurst so no particular need to change that and the Norwood Down sidings provide a way of introducing units south bound on the slow side of East Croydon easily.

  73. RayL says:

    @Henning Makholm 22 June 12.22
    ” but there’s not really a demand for a mostly non-stop service just between London Bridge and Norwood Junction, right? ”

    There is and it came about when the Overground services from West Croydon took over the all-stations to New Cross Gate (and then went onward to Highbury & Islington) a few years ago. At that point, the Sutton – West Croydon line lost its trains to London Bridge. All trains (apart from a few in the peaks) now go to Victoria.

    This means that passengers from Sutton, Carshalton Beeches, Wallington, Waddon and West Croydon have to either a) take one of the two trains an hour that take the Crystal Palace route to Victoria and get off at Norwood Junction then change from P1 to P3 and wait for a fast or semi-fast train to London Bridge , or, b) take any train to West Croydon and either c) take the Overground one stop to Norwood Junction and again change from P1 to P3 and wait for a fast or semi-fast train to London Bridge, or, d) take a two-trains-an-hour all-stations London Bridge train from West Croydon P1 that trundles slowly through eleven stations in a big loop through the South London suburbs (Streatham Common, North Dulwich, Queens Road Peckham, etc).

    A fast service between London Bridge and Norwood Junction is very necessary.

  74. Jim Adlam says:

    Re: Morden, I use this station every weekday and can confirm that trains in the centre track open doors to both platforms 3 and 4, but there is no encouragement to passengers to use a particular side for boarding or alighting. No reference to a one-way system is made in announcements on trains or on the platform, or on signs. The only persuasive factor is that the describer signs above the gateline and on the platforms refer only to platform 3 and ignore platform 4.
    Also, people alighting on platform 5 often take a short-cut through a train in the centre track to use the less-crowded exit stairs from platform 2/3.
    Trains in the easternmost track at Morden also stop between two platforms (1 and 2) but doors only open on one side as platform 1 has been disused for many years.

  75. Littlejohn says:

    It is 10 years or more since I lived in Uxbridge but from memory trains in the centre track at Uxbridge station only opened their doors on one side. Is my memory correct and is this still the case?

  76. Henning Makholm says:

    @PoP, ngh:

    Okay, I must have misunderstood the first quote — somehow the immediate impression I got was of a regular stopping service throughout the day between London Bridge and Norwood Junction, so a train that had just arrived from the north would eventually need to return towards Anerley.

    With the comments, I can now see it must all have been about taking trains out of service at the end of the day.

    The Norwood Down sidings would then mainly be for trains that just need to be put somewhere for the night and don’t need depot-based maintenance, and can enter service southbound (or northbound via Crystal Palace) the next day?

  77. Philip Ayres says:

    Ascot: I’ve walked through a 458 at Ascot but you haven’t been able to do this on a regular basis for some time, certainly considerably prior to SWT replacing them with 450s on the Reading line.

    A similar arrangement can be found at Guildford which I believe has some warning that it’s due to the 3rd rail.

  78. Purley Dweller says:

    There is a regular fast service front Norwood to London Bridge off peak – the Tattenham trains non stop and the Horsham trains stopping at New Cross Gate. The Reigate and Tonbridge trains used to as well but the Three Bridges to Bedford replacements don’t. There is a half hourly slow service from Norwood junction too. It continues to and from Caterham. All of these are off peak.

  79. Unravelled says:

    If the platform ends could be better aligned, (maybe even if not), would it be possible to have a lift accessed bridge at one end of the station, with the later possibility of a conventional footbridge at the other if the subway is deemed inadequate. Placing them at the ends would mean that there wouldn’t need to be room for passenger access around the structures, which could almost fill the platform width.

  80. Anonymous says:

    If you look at 2nd Edition of Trackmaps ‘England South and L.T.’
    page 14, one sees the track passing through Norwood junction platform 7 as non- electrified and OOU as in the photos above. However, further towards East Croydon its marked as ‘NIRU’, becomes electrified, and is noted as being a ‘Perturbation Siding’. Can anyone tell me its function please.

  81. Phil says:

    Re Malcolm @ 23 June 2016 at 11:30

    The point about the 4 rail electrical system used on the tube is that the running rails do not form part of the current path returning the traction current back to the substation once it has been used to power at train.

    Thus in normal conditions they are at ‘earth’ potential and are deliberately kept isolated from both power rails. As such if an object ends up linking the +420 outer rail or – 210V with either of the running rails, then minimal current is drawn through the item and no harm arises. Indeed it is quite possible to continue to run the service as normal because the traction supply is unaffected (assuming the linking of a running rail to a traction rail doesn’t cause harm to the type of track circuit in use), though engineers will be alerted to this situation as it is not ideal.

    In a 3 rail system however both running rails (except for in the vicinity of points where only one of the two can be used) are utilized to get used to return the traction current from the train back to the substation (instead of the center rail on the tube) ANYTHING conductive (be it a beer can, spanner, shovel, umbrella, tin foil or human flesh will rapidly find 1000s of amps going through it / them (a low voltage of 750V means the currents involved are massive compared to what you get flowing in 25KV Overhead systems) with catastrophic results.

    Hence the risk of a serious injury occurring with a person falling between an Underground train and the platform where the +420V power rail is present is much less than on NR lines where the 3 rail system is employed*

    *Note:- On sections where tube and NR services operate on the same tracks, the center power rail is bonded to the running rails and the outer power rail voltage increased to compensate. On such systems the running rails thus form an integral part of the traction return system and must be treated in the same way as 3rd rail lines when it comes to considering the outcome of a person falling between the train and the platform.

  82. Phil says:

    24 June 2016 at 17:15

    The Perturbation Siding’ is used when the service is disrupted – or as the name suggests at times of “Perturbation”. (

    For example it might be decided by control to turn round a late running Uckfield at East Croydon and have the return start back from there but keeping the unit in the platform between arrival and departure would cause congestion. It also could be used as somewhere to dump a defective unit, rather than risk weaving it across all lines on the flat to get it into Selhurst depot during the peaks. Finally (as with the disrupted services) it provides a useful refuge to put a unit whose journey has been curtailed at East Croydon due to engineering work.

  83. Malcolm says:

    Phil: thanks for this. Just to clarify, are you saying that although the potential difference between the outer rail and the running rail is normally 420 Volts, the circuitry which causes this is effectively current-limited. So if a 1 ohm umbrella connects the two, the current through the umbrella will not be 420 Amps, but only a few Amps, as in this fault condition the potential difference will be reduced to a few Volts? And the potential difference between the centre rail and earth will rise to nearly 630 Volts? Or have I misunderstood?

  84. Fandroid says:

    I am unsure about this. Although high currents run through 3rd rail systems to provide the same power as the low current & high voltage AC systems in the UK, the rule still applies: ie I=V/R, so there can be just as much current, if not more, due to contact with OHLE. However, DC and AC have different effects on the human body.

  85. Malcolm says:

    Fandroid: Yes, DC and AC do have different effects on the human body, and which is more dangerous (other things being equal) is still subject to some doubt and dispute. However, other things are not usually equal.

    The fact that the trains take more current on the low voltage systems has no bearing on how much current will pass through a body who comes into contact with the supply. The main driver of that is the voltage, which is one reason why contact with a live 25 KV conductor is almost always fatal, whereas a lower voltage (whether 120V, 750 or anything in between) may not be. (Also, as Phil reminded us, the current through the body will also depend on the internal resistance of the source, making certain incidents on 4-rail systems much less dangerous than the voltage would suggest).

    The other point, though, is that the risk of accidental contact with a conductor is greatly influenced by the position and configuration of the conductor. In that respect, a top-contact conductor rail is far more dangerous than its relatively low voltage would suggest, just because it is comparatively easy to step or fall on it. That is the main safety benefit of the one-day-envisaged conversion from third rail to overhead.

  86. Greg Tingey says:

    1 Amp @ 25kV = 33.3 Amp @750V

  87. Malcolm says:

    Greg: I would quibble with the equals sign there. But granted that the two combinations you cite will be able to deliver the same power, if each is supplying suitable equipment. This may be what you meant.

  88. ngh says:

    Re Malcolm,

    Essentially yes apart from the ” as in this fault condition the potential difference will be reduced to a few Volts? And the potential difference between the centre rail and earth will rise to nearly 630 Volts?”.

    You frame of reference /measurement isn’t quite right.
    The insulation at the substation and between the conductor rails and sleepers and running rails is very high impedance so you essentially have a bridge divider circuit with the insulation between the conductor rail supply and earth dominating to reduce the potential across well conducting items (including umbrellas or the human body (@ 500-1500 Ohms) to as low as fractions of a volt. Depending on the precise supply characteristics it will have essentially have zero difference on the other conductor rail relative potential vs earth if nothing is also touching it.

    Vobject = Robject /( RObject + Rsupply-earth)*Vsupply

    Rsupply-earth being huge dominates what happens with 4th rail systems.

    I was always taught lethal current would be a few microamps in the tens of KVs. Given the short circuit protection on OHLE is typically set to 12KA this is slightly academic ( a train will only draw in the hundreds of amps but there are many per section).

    With 3rd (or 4th) rail supplies the rectifiers only supply typically 4KA so the “DC” fault protection is fairly close to that level to avoid frying the rectifiers. (The sections are far shorter and contain fewer trains)

  89. JohnM says:


    You are correct, if you earth the positive rail the negative rail goes to – 630 V and if you earth the negative the positive goes to + 630V. The circuit is quite simple a pair of resistor in the ratio 2 to 1. As far as I remember there is a 100 Ohm resistor from the Negative to the Running Rail and a 200 Ohm resistor from the positive to Earth. The voltage on lines that have regenerative braking now varies up to 790V.

    The Wimbledon, Richmond, North of Queens Park (except for Stonebridge Park Depot) and the Waterloo and City lines all have earthed negative rails and are effectivly a 3 rail system.

    The reason for having the resistors is that by monitoring the Voltage on the rails both positive and negative Earth Faults can be detected. The track is divided into a number of sections so train earth faults can be detected as they move sections. Track faults of course stay static.

  90. JohnM says:


    The 13 car train is very odd – it appears to be an 8 Car and a 5 Car. I can’t see the unit ends on the ‘8 car’. My suspicion is that it is two different images and it is just a coincidence there were trains in both images that make it seem like a 13 car. Car 4 looks a little longer than the others and has a slight (join ?) line across it. The train approaching platform 3 also appears to have lost 1/2 a car at about the same position.

  91. Malcolm says:

    I note differences between JohnM and ngh about the resistance between a conductor rail and earth on a genuine 4-rail system. For myself I do not mind which is correct, as the main thing I have learnt in this exchange is how much safer a 4-rail system is than a 3-rail one. I have always read, and believed, that the key benefit from the 4th rail is avoidance of stray currents in tunnel linings. Clearly that benefit exists, but it seems to me that the safety benefit, and the added fault detection capability which JohnM mentions, are also big pluses. (To set against the extra capital and maintenance costs of the fourth rail). Apparently London is almost the only place in the world with such a system.

  92. Graham H says:

    @Philip Ares

    Certainly doesn’t happen at Guildford, where the twofaced track to platforms 6 and 7 is the preserve of FGW Dmus passing through.

  93. Phil says:


    Its not ‘circuitry’ that limits the current is the laws Physics!

    Unless a electrical circuit forms a complete circle no current can flow. For example a 12V car battery has +12V constantly sitting on one terminal and -12V on another. If you touch the +12V terminal or the -12v terminal in isolation and your driveway (earth potential) then nothing will happen because the circuit is not complete. If you then touch the chassis or -12V terminal AND the +12V terminal your body completes the circuit and carries the current between the battery terminals.

    On 4 rail electrification systems the running rails are not connected to the substation and ALL the traction current goes out from the substation, though the positive rail, up through the train motors and back to the substation via negative current rail. There is simply no way for the current to get back to the substation via the running rails. The only way something can complete the loop is connecting to BOTH the +240V and -210V rails simultaneously – which is precisely what the motors on the train do.

    When BOTH power rails are connected the loop from substation via the rails and motors (or other conductive item like a human body!) the loop is complete and large currents will flow.

    Connecting the +420V OR the -210V individually to the running rails thus will not complete the circuit and no / minimal current will flow through whatever is connecting the two rails.

    The only exception will be if say there are two separate objects where both objects and the running rail between them complete the power circuit (e.g.e one item dropped at one end of the platform across one running rail and the +420V power rail, PLUS a second object dropped across the same running rail and the -210V power rail at the other end of the platform)

    The key advantage the 4 rail system has is that because of where the two power rails are located, the chances of someone (or an umbrella say) falling off a platform and managing to be in contact with BOTH power rails is minimal.

    On a 3 rail system where we do away with the dedicated negative rail and instead use the running rails, said the running rails form an integral part of the loop back to the substation. To take an extreme example you could* quite happily remove the running rails from a 4 rail system and drive a electric train into the resulting gap because the power circuit is completely unaffected. On a 3 rail system by contrast – you remove the running rails and the circuit from the substation to the train and back is broken.

    Thus the danger is if someone falls off a platform onto a +750V conductor rail in a 3 rail system, it is extremely likely they will also come into contact with a running rail which forms the other leg of the power circuit! The current from the substation will therefore will travel from the conductor rail, through the person, into the running rail and go back to the substation – exactly the same way it does with respect to the motors on a train resulting in serious injuries.

    * (Ignoring signalling or other constraints for the purposes of keeping things simple)

  94. Phil says:

    * in my previous post

    (Ignoring signalling or other constraints for the purposes of keeping things simple)

  95. Pedantic of Purley says:


    Trains like that are surprisingly common and are as a result of two overlapping images being stitched by Google[1]. Didn’t you know that there is photographic evidence that 12-car trains run on the Hayes branch? Mind you one of the carriages is a bit short.

    [1] other aerial views may be available

  96. 100andthirty says:

    Malcolm….The other benefit is that the insulated return, 4 rail system can allow trains to continue to run even when there’s an earth fault. This is useful to avoid stranding trains underground whilst the fault ins investigated.

    By the way, the value of the 4 rail system in managing stray currents is great. People who worked on upgrading the North London Line reported all sorts of electrolytic corrosion in structures. I also recall the slightly panicky phone call from a senior Thames Water engineer shortly after LU had announced it was planning to increase the voltage on sub-surface lines to 750V (regeneration voltage, approx. 900V). He equated 750V with the main line 750V third rail system and was very worried about the effect of stray currents on TW’s assets. He was very happy when reassured that the 4 rails would continue.

    LU’s Engineering Directors usually ask why “LU is different from every other railway” in this respect, so LU generally dusts off and updates a 3rd rail study roughly once a generation. The laws of physics don’t change much over the years so the outcome is usually the same.

    East London Line was converted the third rail, but a very considerable earthing installation was incorporated but was only practicable whilst the line was closed for modernisation

  97. ngh says:

    Re Phil,

    “Connecting the +420V OR the -210V individually to the running rails thus will not complete the circuit and no / minimal current will flow through whatever is connecting the two rails.”

    There are no such things as perfect insulators especially with conductor rail chairs covered in water / graphite dust / brake dust so there will always be minimal rather than zero current flow from the conductors rails to earth in reality.

    If you get too much water with lots of crud in it then this happens as seen on Wednesday (not my pictures): West Norwood Surrey Quays

  98. Malcolm says:

    Phil: I accept all that. Except that if both conductor rails were entirely isolated from earth, it would be meaningless to claim that one was at +120 and the other -240V relative to earth. So to make such a claim meaningful, there must be some connection. But for that connection, I am quite happy to believe JohnM when he says that there is a simple pair of resistors (presumably at the substation). In fact I think they must be of a somewhat higher value than the 100 and 200 Ohm that he quotes, because using those values would give a leakage current of about 2 Amps through them, which would produce 1.2 kW of heat – enough to drive a camping kettle. Such low resistances would also greatly diminish the safety you so cogently describe, so I think they may be more like 100 and 200 kOhm. But they must be there.

  99. Greg Tingey says:

    W = V*I
    Note the huge increase in current at lower-voltage & it’s current through the body that kills.
    Which is why the usually 12kV of an old-fashioned car HT makes you twitch, but does no real harm, because the current is tiny.

  100. timbeau says:

    “one was at +120 and the other -240V relative to earth.”

    -210 and +420 are the actual figures

    “I think they may be more like 100 and 200 kOhm. But they must be there.”

    The resistances in the potential divider are about 10 kOhm according to this link.

    The actual values are less important than the ratio between them, but they have to be very much bigger than the resistance of the conductor rails and the actual electrical load (motors, etc) of the trains in the section, so that the current they draw is negligible. (630V across 10000 Ohm gives a current of 63 milliamps, and losses of about 18Watts – negligible when trains are drawing power by the megawatt)

  101. RogerB says:

    If we agree on 10k ohms then falling across a conductor rail and running rail/earth is still fairly serious as the resistance of your body is around 100k. You would therefore experience 90percent of the 210V or 420V, which would be very unpleasant.
    In the light of Thursday can we stop worrying about passenger growth in London?

  102. Malcolm says:

    Thanks timbeau. Although I am always reluctant to say this when I have been contributing to a digression, perhaps we can return now to Norwood Junction. But I have found the digression very informative.

  103. Malcolm says:

    RogerB: I fear you are right, so the safety benefit of 4-rail over 3-rail is perhaps less than we may have thought, though a reduction from 630V to 180 or 360 may still save quite a few lives.

    Speculation about effects of Brexit on transport in London is off topic, on this or any other existing threads.

  104. ngh says:

    Re Roger and Malcolm,

    As noted in my comment above, I’ve seen several scientific papers suggesting 500 to 1500 ohms for body resistance.

  105. Malcolm says:

    ngh: Yes, obviously it varies. The 100k may well be a maximum. But it may be all academic, as whatever the body resistance (within this range), the current through the body will be enough to kill – and as Greg points out, it’s the Volts what jolts, it’s the mills what kills (mills = milliAmps).

  106. Briantist (in Gigabit internet heaven) says:

    @Greg Tingey
    “Which is why the usually 12kV of an old-fashioned car HT makes you twitch, but does no real harm, because the current is tiny.”

    Indeed. When I used to work in places like the BT Tower, we had to take great care in rooms that were still supplied with “buzz bars” that were used for running the Strowger telephone exhanges.

    Even though these were nominally low voltages (between 9 and 50 Volts) they were connected to enormous batteries and could supply almost unlimited power at high wattage as they were used to drive both the moving mechanical parts and long-distance copper wires to businesses, homes and other exchanges.

    To describe “buzz bars” as utterly lethal would be an understatement.

    And yet I still marvelled at watching a magpie fly on and off all of the rails at the sidings Woodford station the other day, totally oblivious to the potential electron flow!

  107. Timbeau says:

    @ roger b.

    Wouldn’t your 100 kohm be connected in parallel with the 10 kohm, not in series, so would experience the full voltage, but only one tenth of the current, through the potential divider.

  108. Henning Makholm says:

    @Timbeau: Your 100 kohm is in parallel with one of the 10 kohms, producing an aggregate resistance of about 9 kohm. Those 9 kohms are then in series with the 20 kohm at the other side of the supply circuit, meaning that you the potential difference between earth and the power rail you’re touching would be about 93% of its usual value. (It is true that there’s a resistor somewhere that takes ten times as much current as you do, but that’s scant consolation).

    One then wonders why the grounding resistances have to be as small as 10 kohm, of course. I surmise they have to be small enough that their conductance dominates over the routine unpredictable creep currents between power rails and earth, such that an actual earth fault will reliably show up as the power rails being at the wrong potentials.

  109. 100andthirty says:

    At the risk of further extending this digression, everyone has mentioned that the risk on DLR is lower because the conductor rail has the contact surface facing down and the top surface is insulated. This is true. However the equal and opposite effect is that the conductor shoes on the vehicles face up and are live at all times. Also at least half the doors at any platform are in the vicinity of a shoe. Thus DLR fit little insulating shelves above where the shoes come to a stand. This is all well and good whilst all the trains have the shoes in the same position. For the next fleet, this is unlikely to be the case, so more shelves will be needed.

  110. IslandDweller says:

    @100andthirty. Amazing what you learn here. I’d often wondered what those little shelves on the DLR were for!

  111. Andrew L-A says:

    A reference to Cliifford could probably sell a vowel.
    [Sold! PoP]

  112. Greg Tingey says:

    A few years back, I was just about to get on an Up train at Harlow (don’t ask) … a pigeon was flying towards the approaching train,, altered it’s path upwards & managed to be exactly half-way between the contact wire, the carriage, & the oncoming pantograph.
    There was a sudden *BANG* & & an oily puff of feathers …
    Driver had to re-set the contact-breakers before we could move off …..

  113. Phil says:

    Re Malcoln @ 25 June 2016 at 00:08

    I was attempting to keep things simple so as to highlight the principle difference between 3 & 4 rail systems – obviously things like track circuits , ballast conditions, number of trains in section all need to be considered when we get into the finer details.

    Also, as you say, naturally the power supply must be referenced to earth at some point, but on a 4 rail system, the way that is done is such that the running rails should play no part in the return of large traction currents and the risks associated with a fall from a platform are much reduced over 3 rail systems.

    Ultimately it boils down to this:- the risks of having an exposed conductor rail alongside the platform face at Norwood Junction does result in a much higher risk factor if someone fell between the train and the platform, compared to say Uxbridge. Hence it is understandable that the decision on whether to open the doors varies accordingly.

    Re ngh @ 25 June 2016 at 00:00

    The pictures you link to show a 3 rail system – BOTH running rails will form part of the traction current circuit, so yes if floodwater ends up linking the conductor rail and either (or both) of the running rails it will start conducting the traction current short circuiting the traction supply and causing the steam production as the rails heat up and boil the water.

    A similar effect will occur if the conductor rail and a single running rail are linked by floodwater.

    On a 4 rail system however if one conductor rail is linked to one or both running rails by floodwater then you don’t get the same results – the running rails do not carry traction current so there is no short circuit in the traction supply. Its only if BOTH conductor rails become linked by floodwater will you get steam production.

    Also your statement about crud in the water is actually completely opposite to the truth (again due to the laws of physics). The more crud in the water the less easy it is for the electrons to move about. Had the water been crystal clear then it would have had the potential to carry far more current.

    This is why conductor rail systems can still operate when surrounded by ice and snow – because the conductive liquid( water) is in a solid state (Frozen) the electrons cannot move across and current (the movement of electrons) cannot flow. Similarly a conductor rail touching wet ballast or wet mud is still fairly well insulated by virtue of the fact that the electrons have a hard time moving through such mediums compared to liquid water.

  114. timbeau says:

    Depends on the nature of the crud of course. If it includes a decent electrolyte like salt it will be a much better conductor than rainwater is.

    This is why Voyagers, whose roof-mounted electrical gear was supposed to be down-pour-proof, conked out when faced with the Dawlish sea wall in a gale.

  115. Malcolm says:

    Phil: I do accept your argument that 4-rail is less susceptible to various problems than 3-rail (and also safer). But floodwater may not be a very good example to choose, as a flood which manages to get to one conductor rail is rather likely to also reach the other one, what with them being at the same level. Also I suspect that the physics of your “crud obstructs electrons” may be perhaps a simplification too far. As timbeau says, some impurities definitely make water more conductive. The resistivity of wet ballast doubtless varies a lot, but the idea that the stones obstruct the electrons strikes me as a bit mixed-up.

  116. Nathanael says:

    I’m surprised that it’s even legal for them to consider not providing Access for All during construction. The DDA and Equality Act are quite weak, aren’t they?

    The US ADA requires that any new construction include full access, unless it exceeds 20% of the cost of the project, and if it does, they have to spend 20% of the cost of the project on access. Ontario’s AODA actually sets a hard deadline of 2025 by which *every* rail station has to have full access.

  117. Phil says:

    you say as regards salt.

    There was an incidence many years ago on the central division of BR when some well meaning sole decided that prevent wet ballast (following a heavy rail) freezing solid in the wagons during a mid-winter p-way job they would spray the wagon contents with salt water. This was fine and dandy until the signalling techs couldn’t get the track circuits to work after the new track had been laid. The true cause was allegedly only found when someone physically licked the ballast and discovered it tasted salty. I believe that the problem never entirely went away until all the ballast was removed as although rain tended to wash the salt away from the surface, every time the track was tamped it bought salt contaminated ballast back to the surface.

    Re Malcolm

    My “crud obstructs electrons” is indeed a very crude attempt to explain the phenomenon – but to do the subject justice requires rather more detail than would perhaps be appropriate in this page.

    The point is that while water is a conductive medium in a lot of instances, the application of heat or cold (freezing it or turning it to steam) makes a big difference to its conductive properties – as does the quantity / type of other elements like sodium (salt) that may or may not be dissolved in it.

    In general the elements that make up soil and organic mater are not conductive so dirty floodwater should in theory conduct less electricity than water leaking directly onto the track (e.g. from a from a burst pipe crossing the track on a pipe bridge).

  118. Malcolm says:

    Phil: agreed that further excursions into the physics of dirty water conductivity would not be appropriate here.

    My guess, and it is a guess, is that short circuits “caused” by floodwater do not always relate just to the conductivity of the water, but also to how much debris the water leaves behind, and exactly where it leaves it. (And of course to the nature of the debris).

    The salty ballast anecdote is fascinating. Keeping ballast clean and in good condition is of course a complex subject with its own specialists.

  119. Nameless says:


    I didn’t know that BR used to employ fish, benevolent or otherwise.

  120. Greg Tingey says:

    Once upon a day, BR’s shipping fleet was, IIRC, the largest ferry operation [ Both number of ships & number of passengers ] in the world.

  121. To my surprise, sole (pronounced Solay) is not a reference to the fish, but is water that has been fully saturated with a natural salt, to around 26%. Hence the BR ballast lickers…

  122. timbeau says:

    What is the difference between that and brine?

  123. marckee says:

    @Nathanael “I’m surprised that it’s even legal for them to consider not providing Access for All during construction. The DDA and Equality Act are quite weak, aren’t they?”

    It’s illegal for them to not consider it. It is legal for it to be considered and for it to be deemed ‘unreasonable’ – that wonderfully woolly legal term we have.

    When a new building is constructed or an existing one has work carried out on it, the legal precedent is that the building will generally be DDA compliant if you follow the provisions set out in the Building Regulations Approved Document M, or BS8300.

    You’d struggle to argue in court that not implementing those requirements for a new building is in line with the DDA. However there are many reasons why it may be deemed reasonable for existing, modified buildings to not be fully compliant with Part M of the building regs and yet still be DDA compliant. Listed buildings, buildings on tight, hilly sites etc can be very difficult to make fully Part M compliant.

    The courts tend to apply a higher threshold for what is deemed ‘unreasonable’ in the case of public buildings, especially those run by local authorities and service providers like transport authorities, mitigating measures will almost certainly have to be put in place. These can include things like bus or taxi services to more accessible stations.

  124. Old Buccaneer says:

    Re: fish, weren’t there a whole series of goods wagons with fish-related names under BR? & the railways ran fast fish trains when there was a domestic fishing industry I believe. Probably not via Norwood though.

  125. Graham Feakins says:

    @Malcolm – Just a small comment upon your thought that “as a flood which manages to get to one conductor rail is rather likely to also reach the other one, what with them being at the same level.” They are not but the difference in height between their bases is unlikely to affect what you say much. The 4th (centre) rail is lower. This explains all (lots of lovely illustrations):

    Incidentally, I was taught at school (a long time ago) that for most intents and purposes, distilled (pure) water was considered to be a non-conductor, or at least of very low conductivity. It’s the stuff added to it (e.g. in a town supply) or impurities it otherwise contains that makes it noticeably conductive.

  126. Fandroid says:

    Graham Feakins. The only thing regularly added to mains water is Chlorine. However, mains water is definitely not distilled water. It comes through our taps with all of the dissolved mineral content it had in its natural state.

  127. ngh says:

    Re Graham, Fandroid et al.

    Conductivity of sea water is about 6 orders of magnitude higher than distilled water with mains water about 3.5 orders of magnitude higher than distilled and the dissolved ionic constituents of “limescale” are far more effective at increasing the conductivity than sodium or chlorine for example.

  128. Timbeau says:

    @old buccaneer

    For the sake of brevity in telegraph messages, different wagon types were given code names. For some reason, engineering wagons were given names with an icthyous theme. These remained in use long after most other types had been phased out, which is why they are the most likely to be remebered

  129. Graham Feakins says:

    @Fandroid – Yes that’s what I meant but I didn’t express it very well. Apologies.

  130. Graham Feakins says:

    P.S. I was also told at school that much of the drinking water consumed by those at the downstream end of London and extracted from the Thames had already been in and out of said river seven or eight times. It certainly wasn’t distilled!

  131. Slugabed says:

    I think that’s a bit of an urban myth,as a large proportion of London’s (and the towns either side of it) water comes from other sources than the Thames.In fact,no drinking water is abstracted from the tidal region for drinking purposes.Teddington is the tidal limit,as you will know,which is well upstream from what most would consider as “London”.
    The worst you can say is that some of the water we have drunk may have previously been drunk by the good citizens of Oxford and Reading.

  132. timbeau says:

    Teddington is most certainly in Greater London. Most of the water taken from the Thames for drinking water is a little further upstream, between Staines, Sunbury and Hampton – hence the many reservoirs south and west of Heathrow.

    Don’t you mean fluoride?

    @Old Bucaneer
    Great Western telegraph codes

    and the BR engineers’ wagon types

    In the days before refrigeration, trains carrying actual real fresh fish were often given express passenger priority, for obvious reasons. (Immortalised in the Rev Awdry’s story of Henry the Green Engine and the “Flying Kipper”)

  133. Malcolm says:

    … and the trains carrying fresh fish had their dedicated wagon fleet, for further obvious reasons.

  134. ngh says:

    Re Timbeau,

    Plenty of Chlorine (or Hyprochoride) is added to kill most of the “bugs” in it and has been for the last century or so, (though not at any where near swimming pool levels).

    The levels of extraction at Hampton (and further west) are responsible for the in and out stats.

  135. John B says:

    My Thames Water friend points out that the water their sewage treatment puts back in the Thames is cleaner than the water they extract…

  136. Mike says:

    Malcolm – and with a blue spot on the XP-rated wagons’ sides, if I remember aright.

    John B – many years ago I went to an open day at Beddington sewage works, and a staff member at the end of the process was drinking glasses of discharged water to demonstrate that point. He mist have discharged a bit himself that day….

  137. ngh says:

    Re John B,

    and what about the water that bypasses the sewage plants to stop them becoming overloaded and goes straight into the Thames instead (the reason why the super sewer is being built…)

  138. Graham Feakins says:

    @ Slugabed – I fully intended to convey that my comment included the water extraction locations from (and much a bit later returned back to) the Thames from somewhere north of Swindon eastwards. There are 57 sewage (‘purified’) discharges into the Thames, many of those being to the west of London to start with.

    This Q&A site may be of interest to some:

    BTW, hasn’t Thames Water installed a desalination plant in the tidal region for use during drought conditions?

    Of course, should we be closer to Norwood Junction on this thread, we might be discussing the Croydon Canal instead….

  139. Ian J says:

    Connection between London’s drinking water and Norwood Junction: the Chelsea Waterworks Company was one of the major water suppliers in 18th and 19th century London, despite the fact that it took its water supply from the polluted tidal Thames. It was eventually forced by legislation in 1852 to relocate, making its reservoirs (and associated Grosvenor canal and basin) a useful site for the central London terminus of the LBSCR, opening as Victoria Station in 1860.

  140. Greg Tingey says:

    Croydon Canal is very relevant to a discussion of the NJ rail approaches, since this map shows the canal passing through the junctions to the N of the station & then the station-site itself.
    Try this map, as well – probably clearer.
    See also Diamond Geezer’s article

  141. Slugabed says:

    I have it from friend of a friend (inevitably) that,just South of Forest Hill station is a part of the Brighton line which,when the track was being re-laid and ballast replaced,was found to have lock walls from the Croydon Canal still intact below.

  142. Vince says:

    Slugabed – sorry to disappoint but the top lock (apart from two at Selhurst) on the Croydon Canal was at Honor Oak Park – although the line of the canal and the railway did coincide south of Forest Hill station.

  143. Fandroid says:

    I don’t know any detail about the Croydon canal but canals in general could have walled sections for various reasons other than locks. There could have been a swing bridge or lifting bridge. There could be a section where stop boards could be inserted to isolate a section. There might have been a wharf (usually with a wall on just one side of the canal)

  144. @Fandroid

    Is there a canalist/canalerist in the house?

  145. RayK says:

    In my experience it is far from rare for a canal to be walled for mile after mile.

  146. Putters says:

    Re the comment about DC and AC affecting the body differently.

    The main difference is that, as the voltage / current reverses with AC, the muscles in the human body will be twitched (for want of a better word for such an extreme reaction) one way or the other – ie with a 50Hz supply touching it with a limb will mean you will probably be thrown off by your own muscle contraction within 1/100 of a second. This is certainly in line with my own empirical observations with the household mains :O(

    With DC you have a 50/50 chance, and if it twitches the muscles to push on as opposed to throw off, you’re stuck.

    Anecdotally (the anecdote concerned coming from someone who was reputedly on the receiving end of a belt from an incorrectly isolated traction fed lift controller), catching a 630 DC busbar with the back of a hand will result in you being thrown across the room, palm on will be for you to stick and (if completing the circuit) smoulder.

  147. Vince says:

    Having gone back to Brian Salter’s Retracing Canals to Croydon and Camberwell I think there may be two possibilities for any brickwork under the railway. The Croydon Canal had a reservoir on Sydenham Common. In connection with this it was widened on the west side south of the present Forest Hill station to form an extra reservoir whilst the east (towpath) side formed a dam. There is also a footbridge across the railway at Sydenham Park which might indicate an pre-existing route which bridged the canal – although neither Brian Salter nor J Howard Turner in his history of the LB&SCR mention such a bridge in their list of crossings over the canal.

  148. NickF says:

    For what it is worth (DC versus AC) I have seen a chap get a belt from 110 volts DC in an old film studio years ago that used carbon arc lamps. He stuck to the supply until (very quickly thank God) he was pushed away with a broom (wooden handle). Bad burning on the hand, but otherwise OK.

  149. BCW says:

    Just to clarify/expand, the canal didn’t just coincide with the railway, but was actually bought by the railway as the basis for its route between New Cross and Croydon.

    The canal closed in 1836, less than 30 years after it was built – being one of the least successful canals ever built and the first substantial one to be abandoned.

  150. Fandroid says:

    Vince. If you look at the first of Greg’s links to maps of the canal, it clearly shows the widening of the canal as you describe it, and just to the south of the widened part it shows a short very narrow section, adjacent to the railway. This may indicate a bridge which was the ancestor of the footbridge you mention and be the location of the walls mentioned by slugabed’ s ‘friend of a friend’.

  151. Damo says:

    With regard to Putters anecdote. I have had a nasty DC jolt and I can confirm it seems to make things close and grasp, so agree with your back of the hand experience. However I can also say that while a 240v AC is unpleasant (my worst one coming from household wiring, in uneathered metal conduit, where the conduit had eaten through to the live but had not touched the earth in the t&e) 600v DC is bloody nasty (from an incorrectly isolated DC motor) including burns, locked joints and a smell not unlike fried bacon. On the plus side I did jump across the room. So could have been worse.

  152. Walthamstow Writer says:

    @ Damo – I do love the way that people who deal with electricity and who’ve had “shocks” are so matter of fact about the incidents they’ve had. You’re not the first person whose gone (and I paraphrase) “well nearly zapped to death, bit singed round the edges, jolly fun really”. Electricity and the risk of being zapped frightens the pants off me – probably because I really don’t understand it at all.

  153. Greg Tingey says:

    From the Kidner-revised history of the Southern, originally by Dendy Marshall:

    … the first 3 or so, i.e. from (W) Croydon to Anerley, being laid on the bed of the Croydon Canal, which had been purchased for £40 250 for the property & 1/- in respect of profits ……
    The course was fairly level, except for one formidable gradient of 1 in 100 from New Cross up to Dartmouth Arms ( Forest Hill), a distance of 2.75 miles. Here there had been no fewer than 28 locks on the canal.

    I can’t find it now, but I ‘ve seen a photo, taken in late “Brighton” or early “Southern” days, of old Atmospheric Railway pipe, unearthed in the course of heavy track renewals, too.

  154. John Elliott says:

    I’m not sure whether this comment is best placed on this article or on one of the East Croydon ones, but it relates to a paragraph in this article:

    It is now our understanding that the early reconstruction phase at East Croydon station looks likely to involve a period when there would only be four platforms available there instead of the current six and eventual eight. This is largely because of the need to relocate the westernmost two platforms further north than existing, due to the constrained geometry of the site.

    If there’s still a long-term aspiration to put the fast lines in a tunnel from Purley to central London, is there useful provision that could be made for this while the western half of the station is being reconstructed? (I suspect digging out a full station box immediately adjacent to the working half might not be considered practical).

  155. Pedantic of Purley says:

    John Elliot,

    I suspect that plan is dead or at least so far into the future that making provision for it doesn’t make sense.

    The authors of the Route Utilisation Study (RUS) that suggested this clearly thought that there was no -on-the-surface solution possible at East Croydon. The RUS teams are surprisingly small so it only takes one or two people with an entrenched position on this to produce a distorted report.

    Fortunately, other people have subsequently come along and realised that there was a lot more that could be done. I think a sticking point was the original reluctance to consider a scheme that required land take.

    The current scheme includes passive provision for more interventions if necessary (not necessarily in the East Croydon area but impacting on it) should the current scheme not cater for the most optimistic (or pessimistic depending on your point of view) traffic projections.

    Any future major sticking point is going to be capacity at either London termini or routes through central London (including Docklands if necessary).

  156. Paul says:

    Surely a theoretical fast lines tunnel from Purley to ‘London’ wouldn’t necessarily need to pass under East Croydon at all, hence a station box built in advance wouldn’t be a requirement?

    Just because East Croydon sees calls in all services other than Gatwick Express now doesn’t mean it always will.

  157. ngh says:

    Re Paul,

    But after a Windmill Bridge Junction / East Croydon rebuild the capacity limitation is then the London terminals and the immediate approaches not getting the trains to the terminals.

  158. Paul,

    The only reason Gatwick Express doesn’t call at East Croydon is because the current station layout means one cannot maintain the current throughput if they did stop. Also the class 442s have crippling long dwell times.

    With a general desire to make better use of the Gatwick Express service, nowadays they would stop at East Croydon if they could. However, a stop at Clapham Junction is a higher priority (but equally unlikely for the same reason).

    I cannot conceive of any situation where, through choice, a train was not scheduled to stop there once the station is rebuilt.

  159. Timbeau says:

    The USP of both GatEx and HEx is that they are non-stop to central London, exclusively to airport passengers. That’s what the airports pay for – an Express service in the original sense, and is seen as an important selling point in attracting airlines to serve them.

    No airport will willingly replace its express rail link with just another suburban commuter service. Noye the opposition there was just to extending Gat Ex services to Brighton.

  160. Fandroid says:

    If HEx survives that long, it’s not impossible to envisage that it might stop at Old Oak for HS2. I’m not sure that any intermediate stop in south London would be considered as worthwhile for GATEx though.

  161. Walthamstow Writer says:

    @ Timbeau – putting HEX to one side as it’s a “private” operation do the airport owners at any NR served airport station provide specific hypothecated funding for non stop services? Surely it is GatEx that is the glorious exception given many (all?) other airport stations are served by local stopping services with some also having inter urban or fast services also calling. I have been criticised for saying this before but in a rational world the Gatwick Express would simply not exist on such a congested and constrained rail network. It is an inefficient use of platform and track capacity and its new trains are just standard suburban stock with a different livery. Nowt special about that at all.

  162. Pedantic of Purley says:


    No airport will willingly replace its express rail link with just another suburban commuter service

    Yes. I think we have grasped that.

    A few years ago there was a big debate when extending Gatwick Express was first mooted. You are correct in that Gatwick Airport put up considerable opposition to the change. However we had the absurd situation of a very congested and busy East Croydon station in peak hours with nearly empty short formation Gatwick Express trains passing through. It was very emphatically agreed by all concerned except the airport (TOC, NR, DfT) that the way forward was to dilute exclusivity of the airport service. At the end of the day Gatwick Airport weren’t putting money, or at least serious money, into the Airport Express operation so their views only counted as far as trying to keep up percentage of users arriving at the airport by public transport.

    I am led to understand that a stop at Clapham Junction would make a lot of sense because Gatwick is used by people from all over South London. It just doesn’t happen because it can’t be done.

    As pointed out by Walthamstow Writer it is not the case of what the airport pays for. The airport basically pays for sod all.

    I think you have also misunderstood what is probably going to be the unique selling point of Gatwick Express in future. It is not dedicated non-stop trains to and from London but departure and arrival at Gatwick Airport on regular interval trains at platforms dedicated to the service and operated with airport passengers in mind.

  163. Old Buccaneer says:

    PoP re changing at Clapham Junction (CLJ): Typical journey times from Putney to Gatwick Airport via CLJ: 41-46 mins, 4 tph. For example. Allows 10 min interchange time.

  164. Purley Dweller says:

    It would be difficult to keep up the daft premium pricing if they stopped at Clapham and East Croydon

  165. Anonymously says:

    ‘No airport will willingly replace its express rail link with just another suburban commuter service.’

    But hasn’t something similar happened to Stansted Express? Once upon a time it only had one intermediate stop (@Totty Hale, for the Vic Line), but now trains seem to alternate adding additional stops at any one or more of Harlow Town/Bishop’s Stortford/Stansted Mountfitchet en route. Did the airport kick up any fuss at the time these extra stops were added, especially since this it has the longest journey time by far of any of the airport ‘expresses’?

    Re DC electrocution…..this discussion has reminded me of the railway safety film we watched in primary school (‘Robbie’- can be found on YouTube). Since we were in ‘third rail’ land, the version we watched showed the title character electrocuting himself on the third rail (the version on YouTube shown him being run over by the train on a non-electrified line; I believe a version for AC OHLE lines also exists), complete with smoke and sizzling 😮. I still vividly remember that image of his leg lying across the rail…it’s the type of thing you never easily forget at that age, so as to dissuade us from trespassing on the line. We were also warned by the safety officer that we would ‘stick’ to the third rail if we touched it, as others have alluded to, and that the shock would kill us (ironically, the character in the film survives the accident, but is left unable to walk).

    Just to be clear on this, am I correct in thinking that one can only be electrocuted if you are in contact with the third rail *and* one of the running rails at the same time? Would anything happen if you stepped onto the third rail with one foot whilst your other foot was in contact with the ground only?

  166. Timbeau says:

    Would anything happen? Yes, it would. As the running rails are at earth potential, the voltage across your body would be the same whether you are touching a rail or anything else.

    As for Stansted Express, my recollection was that, for line capacity reasons, it was never possible to be entirely dedicated. I’m fairly sure that, on the one occasion I did use it, very shortly after it opened (for no other reason than “track bashing”), there was a stop somewhere in Essex- Bishops Stortford I think.

  167. Greg Tingey says:

    Stansted Express services have always called at one intermediate stop besides Tottie Hale – it used to alternate between Harlow & Stortford.
    For some years now, they actually admit to stopping @ Tottenham, which they didn’t at first …

  168. Fandroid says:

    Thinking rather more deeply about it, Clapham Junction provides rail links to a huge area of south west London, the Thames Valley and southern England. It has also been seriously upgraded in terms of access for all. That means it’s ready to handle passengers with large bags. East Croydon doesn’t really offer much extra in terms of rail connectivity and is anyway well served by Thameslink which already provides good access to the City and is due to offer connections to a greater area north of London than it does now.

    While accepting that Clapham Junction cannot physically accept GatEx stops at present, it does have a potentially massive role in providing rail connectivity south of London. However, I guess we will all have to wait for CR2 to realise that.

  169. Greg Tingey says:

    The trouble with CJ is that, for that to work, even apart from CR2 it would need 4 platforms for each of the Brighton & LSW fasts – i.e. an extra 3 or 4 platforms, on the surface.
    Um, err …..

  170. ngh says:

    Re Greg,

    They could do a lot with just 1 extra platform in peak flow direction on the Fasts.
    SWML could be sorted with getting P7 properly usable and having P8 bi directional (porposal have been kicking arround for least least 12 years.
    See Sussex part 5 for sorting the BML fasts in similar way.

  171. c says:

    Surely everything can stop at CJ – the GatEx skipping is what messes up the pathing. I don’t agree with the non-stop airport thing, the paths are too precious. Most Amsterdam services stop once or twice in between Schipol and Central… Barcelona the same. In Australia, Brisbane and Sydney’s airport trains are integrated with the suburban network and not dedicated. As for Paris’ RER…

    If GatEx was eliminated, but then maybe 3tph equivalents were put into the timetable and everything out of Victoria budged a bit, would it hit 10tph Vic-CJ-EC-LGW ?

    Surely that would be enough with the standard stopping pattern.

  172. ngh says:

    Re C,

    See extensive discussion in “Sussex Part 5 – Up the Junction” article for why everything can’t stop at Clapham Junction (unless you remove some services so there are fewer and then they can all stop! but you have massively reduced capacity) it is about the dwell time and a GatEx crawling through at 10mph is normal and has a far lower platform occupation time than anything that stops. ETCS, AC electrification for improved acceleration and metro style stock for quick unloading and loading might enable another 2tph to stop.

  173. Old Buccaneer says:

    C, ngh, Fandroid: CLJ provides a service to Gatwick today. Which seems to work on the occasions I’ve used it. For those approaching Gatwick on the Victoria and District lines, Gatwick Express is an option; semi fast trains ex Victoria are also available. Thameslink also offers a wide range of journey opportunities via Gatwick.

    As I understand it, Gatwick has more ‘value’ than ‘premium’ users compared to Heathrow. I would expect this to be a factor in planning public transport access.

    On premium fares, I suppose Brighton folk don’t need to pay extra to travel on Gatwick Express branded trains?

    I have looked for – & failed to find – capacity utilisation numbers for Gatwick Express. I think the question of whether the new 12-carriage trains are a good use of capacity* is what we need to debate.

    Both track capacity and financial capacity.

  174. Ian J says:

    @c: In Australia, Brisbane and Sydney’s airport trains are integrated with the suburban network and not dedicated. As for Paris’ RER…

    Brisbane and Sydney’s airport trains both failed to reach patronage expectations and lost money for the private companies that built them – the Sydney company went bankrupt, the Brisbane one was sold for half the construction cost to a British universities pension fund – while Paris is about to build a dedicated airport express to CDG (the public enquiry is happening right now).

    Dedicated airport expresses seem be able to attract passengers that regular services do not and so increase public transport mode share to the airport (whether those passengers and their premium fares are worth the capacity they take up is a different question).

  175. IslandDweller says:

    @Timbeau on 4 July “The USP of both GatEx and HEx is that they are non-stop to central London, exclusively to airport passengers”
    That USP doesn’t exist for GatEx – hasn’t been for a while now. In the “normal” timetable (ie – the one before yesterday when a huge swathe of services including those branded “express” were dropped), at least 50% of trains branded “express” run through to Brighton. Trains branded “express” are not exclusive to Gatwick airport users.

  176. Alan Griffiths says:

    But Gatwick Express goes to Victoria (London, not Manchester or Southend).

  177. Timbeau says:


    The Gatwick Express services, whatever their point of origin, have always been non-stop from Gatwick to Victoria.

  178. IanJ,

    Dedicated airport expresses seem be able to attract passengers that regular services do not

    Possibly. But is the fact that services are advertised as direct airport expresses and promoted by the airlines that is important rather than the fact that they don’t have intermediate stops?

    Although not typical, I would suggest the DLR to City Airport is a counter-example where there isn’t much market share left to grab. More typically, lots of people are predicting low passenger numbers on Heathrow Express once Crossrail fully opens. I expect the reality will be that lots of people will be enticed to use it once and when they realise the alternative that is available they won’t bother in future unless they really wanted to start or end their journey at Paddington.

    As stated above, the stance with Gatwick Express seems to be to offer a product that means at Gatwick and Victoria you have dedicated platforms and can generally board the train straightaway. In normal circumstances you should be able to rely on a regular interval service with no intermediate stops outside the London area without having to confront a mass of confusing customer information screens. More relevant to a different thread, the idea appears to be to normally have an on-board host available to answer questions and help if necessary. For this you pay a small premium.

  179. ngh says:

    Re PoP,

    You mean the on-board hosts on GatEx they got rid of several years ago?

    Seperate GatEx gate lines at Gatwick in a months time will mean even less need for on-board ticket checks or similar.

  180. Graham H says:

    @PoP/Ian J – another factor in the equation is that about 1/3 of all GATEX tickets are sold by airlines as add-ons. This can be done, of course, with other operators but one can see it lacks the cachet of exclusivity.

  181. ngh,

    I mean the on-board host Charles Horton mentioned (and presumably committed to) in the appearance at the select committee as watched using the the link you provided in another thread.

    This maybe suggests that the second member of the train crew will be re-allocated to routes where they can do genuinely useful customer interaction – i.e. first time passengers not daily commuters.

  182. ngh says:

    Re PoP,

    But the OBS (as opposed to retained Guards) will based at the wrong depots for doing that on GatEx. and OBS only exist on southern not the other GTR “brands”.

    I think it is all about on-board ticket revenue collection which won’t be needed on GatEx with double gating in 1 months time but plenty elsewhere on Southern that could do with it.

  183. timbeau says:

    It is so drummed into people that “you must have a ticket before you board a train” * that custom for on board ticket sales on HEx and its ilk is very light. I recall using the “Heathrow Fast Train” in 1998 with a colleague who was very uneasy at our ticketless state as we boarded the train at Paddington. (Incidentally, the connecting road coach from Heathrow Junction delivered us straight to the terminal door, meaning a Paddington-to-check-in-desk time faster than I have ever achieved using the full HEx service!)

    (* the message is played on the trains, when it is presumably too late to comply. In any case it is not always true)

  184. Pedantic of Purley says:

    Of course the Paddington-to-check-in-desk time was faster still when the check-in desks were actually at Paddington!

  185. Malcolm says:

    PoP: Time from London terminus to check-in is now frequently negative, sometimes measured in negative days. Perhaps time to join the queue for security check should be the new metric.

  186. Malcolm,

    I nearly commented in like manner but noticed that timbeau had very clearly specifically stated check-in-desk.

    On an off-topic pedantic note I would like to point out that the real metric you tend to nowadays need to apply is “time to departure gate” as that is the cut-off that really matters and the one you need to make sure you have allowed enough time for.

  187. IslandDweller says:

    @PoP. On that pedantic note – the crucial time is actually “time to security check point”. BA already enforce a rule (and easyjet are about to introduce the same) that if you’re not through security by a certain time (easyjet will be T-30) then you’ll be blocked at the security check, you won’t be able to go ‘airside’, and you’re booking automatically cancelled.

  188. Pedantic of Purley says:

    Island Dweller,

    On a pedantic note as far as EasyJet is concerned, I am pretty sure you have got that wrong.

    See and click the “What departure gate will I be flying from?” option.

    Widely reported in the press at the time the rule started to be enforced e.g.

    No mention whatsoever of security check point but, self-evidently, if your haven’t passed security by T-30 then there is no chance whatsoever of being at the gate at T-30.

  189. James Bunting says:

    PoP, Island Dweller

    The media are obviously all working from the same press release but not all are being equally accurate in their reporting.

    Once a passenger has entered the security area and gone “airside” there is no simple way of stopping their progress. Because, at Gatwick, entry to airside is by gates that respond to the code on a boarding pass, in whatever form, it is the only way that late travellers can be stopped from going airside. Very similar in effect to the days when stations, particularly London Termini, had folding gates at the entry to the platform that a ticket collector could close to ensure an on time departure, or, I believe, that Eurostar programme their check-in gates to refuse to those attempting entry after their deadline has passed.

  190. Pedantic of Purley says:

    James Bunting, Island Dweller

    I have no doubt James is correct about bad reporting. Having travelled by EasyJet recently I am absolutely convinced that it is time at the departure gate.

    I think we have covered this all we can and it is off-topic anyway.

  191. Anonymously says:

    @PoP/Ian J….And yet the Picadilly line is still well patronised by airport users, despite taking an eternity to reach Central London. I speculated on the reasons why some rail/metro links are well used and other aren’t on a recent post on the Heathrow thread: Any thoughts?

  192. Ian J says:

    @Anonymously: And yet the Picadilly line is still well patronised by airport users

    The fact that the Piccadilly Line wasn’t emptied by the opening of Heathrow Express shows that Heathrow Express increased the total number travelling to the airport by rail and that the Heathrow Express passengers would be in taxis or cars if they weren’t on the train.

    On your previous point, I don’t think that you can assume from the handful of examples you give that the time between an airport opening and the opening of rail links has a permanent effect on rail mode share. A more obvious argument would be that cities with lower public transport mode share overall are also likely to have low airport public transport mode share (see any number of North American cities for examples, Toronto most recently).

  193. Fandroid says:

    There is a massive price difference between Piccadilly access to Heathrow and HEx. Very persuasive to a large proportion of users. Even the Connect fare from Hayes and Harlington to the airport is greater than the Oyster fare from central London. GaTEx keeps the modal transfer up from road to rail. It has a (sort of) social function. Ie reducing road congestion and the need for road investment.

  194. IslandDweller says:

    “And yet the Picadilly line is still well patronised by airport users, despite taking an eternity to reach Central London”
    But at least the Piccadilly line actually goes to central London. HEx goes to Paddington – which is not the final destination for most travellers.
    When you look at end to end journey times, for lots of central London, any time saving from using HEx is minimal, and no way worth the massive cost difference.

  195. RogerB says:

    re. ngh 5 July 2016 at 16:06
    Sorry for the delay, but thought I should point out that it’s a myth that 25kV AC electrification would be needed to provide improved performance.
    The limiting factor on the existing third rail electrification is the power available at the rail, which is down to the capacity and frequency of substations. The 4-REPs were rated at 3,200 HP and could have out-performed any 25kV sets. However, because of the power supply limitations a 12-car train, with 9,600 HP, could not be run.
    Should it prove worthwhile to improve performance on the third rail network it would be cheaper and less disruptive to uprate the power supplies rather than convert to 25kV AC.

  196. Old Buccaneer says:

    @RogerB I think we’ve discussed the uprating fiasco in relation to Siemens “lardbutts” (h/t Captain Deltic) on South Western lines? ngh may recall.

    But your point is well made, given the shocking (sorry!) state of affairs on GW electrification.

  197. Pedantic of Purley says:

    Roger B,

    Surely the point is for short bursts of acceleration a motor can handle all the power that you can supply? I understood that the rating quoted is generally the sustainable rating i.e. at what power it can continuously run at without overheating or burning out. In reality DC at 750V is never going to provide the same level of power for “burst mode” as AC.

    One could also argue that the limiting factor on the existing third rail electrification is the power available at the rail, which is partially down to the capacity and frequency of substations but is also dependant on how good a contact is made between the shoe and the third rail.

    Not relevant to this particular circumstance is the fact that AC overhead will generally outperform DC rail pickup because rail pickup becomes very problematic above 90-100 mph.

  198. Timbeau says:

    @Roger B

    Of course REPs worked in 12car trains all the time, but only one of the three units was powered. 3200hp for a 12car train was little better than a 12 VEP formation, with 3000 hp. (Three power cars, each with four 250hp motors)
    REPs were only allowed to work in multiple with TC units, and 2car powered units – although I am not aware of it ever happening.

  199. RogerB says:

    @Pedantic of Purley
    Yes, probably the existing shoegear would not be good for more than 100mph, but then it is substantially unchanged since the third rail electrification was introduced. I suspect it could be redesigned to be good for at least 125mph.
    If the power were available (more, and more powerful, substations) you can increase power to the train by having more shoes.
    Obviously this only applies to multiple units, if you want a loco then, yes, 25kV would provide more power.

  200. Anonymously says:

    @RogerB….Except there’s no real need (or, more to the point, commercial demand) for any 3rd rail system that enables speeds above 100mph. Worldwide, third-rail systems are typically commuter railways with speeds limited to 75 mph or 120 km/h, and so there is no viable case to be made for developing a new third-rail system that enables high-speed running. Plus the ex-SR lines are in a relatively geographically compact area (the SWML Waterloo to Weymouth is the longest route, I think, at around 120 miles?), and wouldn’t really benefit from speeds that high.

    I do agree though that there might be scope to upgrade and further raise speeds on existing stretches of line to 90-100mph wherever possible in the third-rail area, although that would depend on NR’s commitment to maintaining the third-rail system instead of converting to 25kV AC OHLE (FWIW, I consider far more likely that the existing equipment will just be renewed, even if only to avoid future OHLE cost blowouts as we’ve seen with the GWML).

  201. Graham Feakins says:

    @PoP & Anonymously – Remember that parts of the BML and Bournemouth main line were cleared for 100mph running (with speeds recorded in excess of 110mph on the Bournemouth route), whilst Eurostar was cleared for 100mph on the DC third rail stretches beyond Tonbridge.

    Re. PoP “In reality DC at 750V is never going to provide the same level of power for “burst mode” as AC”, I tend to disagree because the traditional DC motor running off a DC supply, as well as asynchronous motors, will achieve rates of acceleration from lower speeds that traditionally have been higher than the AC version and will therefore provide that burst of speed. Just look at trams accelerating up steep hills.

    One normally needs that burst of speed at the lower end of the speed range to accelerate sharply away from stops and caution signals clearing to green and similar. The “Southern Electric” network used to be like that but unfortunately timings have been made so lax that a stopping service journey that used to take 14 minutes now takes (sorry, “allowed”) 19-21 minutes, for example.

  202. Pedantic of Purley says:

    Graham Feakins,

    Nowadays train motors are the same whether the power supply is AC or DC. The motor will be an AC motor. The traditional motor running off DC either does not exist or is dying out on the railway. So any suggestion that DC is somehow better at low speeds would be invalid today.

    The commonality of motors is one reason why it is no big deal to have dual voltage train. It is also very well established that if there is a situation where both DC and AC are available for use (as certainly was the case in parts of the North London Line) then drivers were instructed to use AC.

    Historical evidence, not completely reliable due to other factors, suggests the DC motors have not exactly excelled themselves on London Underground. Look at the appallingly underpowered A60 stock and its modern replacement or alternatively the then-state-of-the-art 1967 Victoria line stock and the improvements that could be made once the better specced (and better accelerating) 2009 stock replaced it.

  203. Greg Tingey says:

    Coming back to Norwood Junction (!)
    There’s This article suggesting Tfl have “No Plans” to improve disabled access there.
    But, if the station needs a partial rebuild-&-improve (As this thread is nominally about) then surely “DA” would be included in the package?
    Or does the money come out of a different box, as per CR1 stations?

  204. ngh says:

    1. Running above 100mph you start running into lots of Physics problems rather than Engineering ones (that could be whole article in itself) the former you can’t design around (current rather than voltage related).

    2. (We have gone through this before numerous times on LR and I don’t have the docs in front of me today) Either the losses in the system till the point of contact with the 3rd rail will be stupidly huge if you want to increase traffic levels and performance or you end up installing more and bigger substations which becomes more expensive see the new 20MW one at Blackfriars to provide high performance on the TL core (at roughly the same cost as all the supply and switch gear that would be needed for AC OHLE from Stoats Nest to Brighton!) or mix of the 2.
    NR’s modelling suggested that it would be possible to shave 4-7minutes of SWML metro services with AC conversion… (ignoring padding and on that subject DfT have inserted some padding removal into the new SW franchise spec which helps explain some of the reported added unhappiness of the bidders)
    Doing the equivalent with 3rd rail would need a major redesign of the Electricity grid and distribution networks that NR is fed from South of the River with a minimum cost in 9 digits.

    3. Fault protection of very large 3rd rail suppliers is a pain in the posterior because you it becomes very difficult to differentiate between genuine high load and a genuine fault within the physics limitations of protection devices (Another reason transmission and distribution of electricity uses HV to reduce the current to bypass the fault protection and switch gear issues which are mostly current related)

    Regards, Dr ngh MEng

  205. timbeau says:

    “The traditional motor running off DC either does not exist or is dying out on the railway.”

    Certainly does still exist. The first trains on BR to use ac motors were, I think, the 465s (Networkers), and there are plenty of older units than them (e.g classes 313-322) – not to mention diesel-electric locos – still in service.

    SWT has a project underway to retrofit its 455s with ac motors . After a very long gestation, this project now seems to be going ahead at a very brisk rate, and I expect that by the end of the year the only remaining dc motors on SWT will be under its 456s. (Southern’s fleet of 455s are not part of this programme)

  206. c says:

    From IanC: “@c: In Australia, Brisbane and Sydney’s airport trains are integrated with the suburban network and not dedicated. As for Paris’ RER…
    Brisbane and Sydney’s airport trains both failed to reach patronage expectations and lost money for the private companies that built them – the Sydney company went bankrupt, the Brisbane one was sold for half the construction cost to a British universities pension fund – while Paris is about to build a dedicated airport express to CDG (the public enquiry is happening right now).”

    Yes but we aren’t talking construction here. The BML exists, this is how we use it.

    Everything stopping at CJ and EC is not a hardship and I don’t think makes the service too much less premium or stopping – when balanced with the benefits it offers.

    Add in Crossrail 2, more Overground trains, Battersea Power Station and the Northern Line (which might one day hit CJ) – and the usefulness will only ever increase.

    City Thameslink people would still need to ‘endure’ Blackfriars, LB and EC for instance, and that’s seen as a great journey – not an all shacks service. Cheap too!

  207. Anonymously says:


    ‘….see the new 20MW one at Blackfriars to provide high performance on the TL core (at roughly the same cost as all the supply and switch gear that would be needed for AC OHLE from Stoats Nest to Brighton!)’

    Does that include installation, labour, bridge reconstruction, trackbed lowering and the costs of extended line possessions and closures (which I assume would be much longer compared with the equivalent work required for 3rd rail renewal)?

    I accept as a non-engineer your arguments for AC OHLE; I’m just being pragmatic. If the ex-SR were only being electrified now, then there is little justification for choosing DC third-rail over AC OHLE, for the reasons you point out. But (to quote the latest political buzz phrase), We Are Where We Are…..

  208. Sad Fat Dad says:

    Re: Anonymously.

    Dr ngh MEng (and bar) was referring to the cost of the substations and switchgear only, i.e. the electrical distribution equipment that gets the juice from the grid to the contact system (OLE / 3rd rail), and not the contact system itself.

    And he’s right – the substation at Ludgate Cellars, and all its associated accoutrements, is a rather large and specialised (and thus expensive) piece of equipment. Whereas 25kV substations are industry standard and can be delivered somewhat more cheaply.

  209. RogerB says:

    It would be interesting to know why the Ludgate / Blackfriars substation was so expensive. It’s only transformers, switchgear and rectifiers after all. All kit that is routinely fitted within AC electric sets. Could it be something similar to the GW Electrification cost over-run – i.e. the job was being run by ‘Project Managers’ rather than Engineers?
    Disgruntled (engineer) of Bath.

  210. Paul says:

    As I understand it, the Ludgate Cellars building also controls changing the nature of the dual voltage section of route between City Thameslink and Farringdon. Basically the earthing systems and traction return systems are set up differently for DC or AC modes, and this is interfaced with the signalling system to the extent that the boundary can effectively move along behind individual trains. Of course up and down lines are also controlled separately.

  211. ngh says:

    And the lock gates of Forest Hill striketh back?

    (or is it the other alternatives at that location return of the old favourite the collapsed sewer of old or run off from the hill at Crystal Palace which eventually make it to the Ravensbourne?

  212. If it is down to Thames Water then they seem rather good at disrupting this route. Leak between East Croydon and South Croydon a few years back, flooding Thameslink tunnel (blame for it disputed) and now this.

    Would they still be liable for a collapsed sewer if disused (which I presume it is)? If so then probably big claims being put in.

  213. ngh says:

    Re PoP,

    and the previous collapsed sewer at Forest Hill in 2013! I think there may also have been a collapsed storm drain in early 2015 at Forest Hill that didn’t get the headlines because of the London Bridge rebuild issues at the time (just added to them).

    Plenty of formation issues there over the last few years.

  214. Walthamstow Writer says:

    @ PoP – Thames Water are supremely good at disrupting just about anything and then denying all liability.

    @ Ngh – according to BBC London News the “hole” is to be fixed by pouring 50 tons of ballast down it overnight. Shurely shome mistake? I can’t imagine you fix a sink hole by doing that.

  215. ngh says:

    Re WW,

    They omitted the second part: leaving someone to watch the filled-in hole for when it starts to grow again!

    At least granite won’t get disolved etc.

    The geology within a hundred m or so of the station look bad given issues with utilities under the roads in the area too. Also see Graham H’s comments about road rail conversion in the area not being go-er due to poor geology.

    The condition of the ballast in the area from the photos looks pretty clogged with fines along with plenty of DOW-MAC sleepers (which suggests they should be at least 27 years old) so might be due for some re-laying.

  216. Walthamstow Writer says:

    @ Ngh – Fines? (Track / P-Way was never an asset area I really understood).

  217. ngh says:

    A TW sewer is the culprit and they are helping to assess if ballast can be dumped on top (presumably with some kind of relining) or whether a bigger repair is needed!

    Re WW,

    Fines – silt etc. the equivalent of the stuff left in the bottom of the corn flakes packet that in the case of ballast binds togehter to form a paste that prevents the ballast working properly e.g free draining and moving

  218. ngh says:

    The sewer will need to be repaired before the hole can be filled which won’t be completed over night so NO up services through Forest Hill tomorrow.

    Just as well they have so many extra spare paths through Tulse Hill available!!!

    Southern Twitter feed says all ok tomorrow but their website says up the creek (left hand meet right hand and they wonder why users get frustrated…)

  219. Briantist (in Gigabit internet heaven) says:


    It seems that London Overground’s East London line is only running services from Highbury&Islington to New Cross Gate and Clapham Junction today.

  220. ngh says:

    Forest Hill:
    Foam concrete lorry on site, the hole is conveniently 25m south of the RRV access point roughly where there was a bend in the former canal route to the west in the station area. (About 10m south of the end of up platform). Because the RRV access point is to the east, the down lines are closed as well.
    Sewer presumably used the former canal swing bridge road rights which became a level crossing that long since closed (in early-mid Victorian times)…

  221. 100andthirty says:

    Re Forest Hill – and other media reports that the hole is much bigger than first estimated

  222. RayK says:

    I presume that fast setting foam concrete is available. Even so! How long does it take to reach adequate strength?

  223. ngh says:

    Re Ray K,

    Back filling carefully with ballast to ensure minimal voids and good packing will take a while then sending a tamper over and probably topping up again. Then temporary speed restriction and careful observation.

    Re 130,

    Just repetition of early inaccurate reports from yesterday a bit of foot – metre confusion.

  224. Graham H says:

    Moderators – a small “management ” point – this thread has effectively now merged into the same hole as the Southern thread. Would it be possible for the sake of an orderly diaogue if the relevant bits could be consolidated in one place? Otherwise there will be duplication and difficulty in replying.

  225. Pedantic of Purley says:

    Graham H,

    Good point. I think this (the hole) is more relevant to Southern’s woes at the current time than to the future of Norwood Junction so will delete any further comment here about the hole at Forest Hill.

  226. Walthamstow Writer says:

    A Mayor’s Q and A relating to step free access at Norwood Junction. Note the references to safety issues which reinforces comments made earlier.

    Norwood Junction Station
    Question No: 2016/2807
    Steve O’Connell

    In 2014, Geoff Hobbs at TfL advised that his goal was for Network Rail to complete an updated feasibility report into making Norwood Junction station step-free, as this would form a strong case for bidding for any available DfT monies for the provision of step-free stations. In view of Norwood Junction being the third busiest station in Croydon (with 3.8 million passengers entering and exiting the station in the last year) & it now being the only TfL London Overground Station in the network’s East London Line without disabled access, will you commit to doing all in your power to persuade Network Rail to introduce disabled access here?

    The Mayor

    While TfL’s contractor LOROL manages Norwood Junction station, Network Rail retains ownership. Norwood Junction station would benefit from being made step-free, and TfL nominated the station for DfT Access for All funding in 2007. While it scored highly against some of the criteria set by the DfT, Network Rail removed the station from consideration for funding as it could not identify a safe solution to deliver step-free access.

    As part of Network Rail’s development work into possible track layout changes at Norwood Junction, itself part of a potential Brighton Main Line upgrade programme, Network Rail is undertaking a further review to explore whether it is technically possible to provide step-free access safely at this location. If the answer is yes, then stakeholders and Network Rail can explore funding opportunities with the DfT. I would like to see the station made step-free to further increase accessibility on the East London line, and I will ask my officers to pursue all available options for achieving this as this process continues.

  227. Anonymous says:

    Apologies for reviving an almost extinct thread, but it occurs to me that a big opportunity has been missed in this discussion about future plans for Norwood Junction

    I have to travel from East Croydon to Bromley South on Saturdays, a journey that’s not at all easy. Tram to Beckenham Jumction was my first thought, but the bus onward to Bronley South takes around 50 minutes, The Tram never seems to match up well with the rail connection at Beckenham Junction, if it did the journey would only take around 10 minutes.

    Travel planners indicated that the 119 bus from East Croydon was a better option, even though it takes around 50 minutes. I wasn’t surprised to see that travelling all the way to |London Blackfriars and back to Bromley South was only slightly longer!

    Looking at Norwood Junction I see that there used to be a rail link to Birkbeck and onward to Bromley South. Today it has been converted into a road with housing on it.

    If there was any way of reinstating the link, it would make travel from East Croydon to Bromley immensely easier. Yes there is the issue of competing train companies using each other’s tracks, but maybe Thameslink could take it on as they run in both areas

    Any thoughts on this long lost connection?

  228. Old Buccaneer says:

    @ Anonymous at 0140: look, it’s not my patch but:
    1 Google Maps & StreetView suggest that it would be quite difficult & expensive to rehouse those affected;
    2 I have no idea what the demand for such a service would be (other than your individual Saturdays-only need)
    3 as this article (& others in the series) bears witness, there are lots of problems with conflicting movements & it is quite difficult to run the existing services reliably (even when there are no industrial relations problems)
    4 I have ignored the civil engineering costs which are likely to be significant
    5 all of which suggests reinstatement is highly unlikely if not impossible.

    For your journey, may I suggest changing from Penge West to Penge East? It’s about a 10-minute walk and the trains are about 16 minutes apart in the E Croydon to Bromley S direction. It’s also cheaper than going in to London & out again.

    There are two trains an hour (tph) from E Croydon to Penge W and 4 tph from Penge E to Bromley S.

    But that route doesn’t seem to work so well in the opposite direction.

  229. Jonathan Roberts says:

    @Anonymous 01:40

    There is indeed a metaphorical ‘Berlin Wall’ – lack of routes or interchanges – hindering easy access in middle and outer South London between the South Central and South Eastern networks. With 40% housing growth expected in some outer boroughs by the 2040s (though probably not LB Bromley), so much higher population and some jobs growth (eg Sutton Royal Marsden), there is a case to see which options might be possible.

    The recent TSLO report reviewed options for links.
    See also a map:

    If just for local travel, then a Penge East/Penge West pedestrian link alongside the railway tracks was a possibility, and TfL also assessed this recently. The main problems there are cost, and the length of the link which will deter usage numbers, therefore not a great BCR. TSLO looked at achieving greater usage and a better BCR might be achieved by relocating Penge West station northwards so much closer to the SE tracks (although that would be at considerable cost, so perhaps not good enough).

    If as a strategic link enabling attraction to rail of more and longer cross-outer South London journeys, then a tunnelled route from north of Norwood Junction to join the SE tracks at Kent House appears feasible. That was in the TSLO context of a limited-stop ‘R25’ service between Surrey, outer South London and Kent, aimed at fast direct travel to and from intermediate towns. A tunnel was estimated at ca. £1.2bn incl optimism bias. TSLO’s specification was a high quality 4 tph limited stop service, it would call at West Croydon, not East Croydon. A business case would be required.

    A Thameslink 2 type service was also mooted for the longer term, linking the BML and East London via East Croydon and this link, thence onto the Mid Kent Line and a direct Lewisham-Canary tunnel relieving the ELL and DLR cross-river lines which will be full in future years. Other SC-SE line permutations might arise, as it is understood that Canary Wharf Group is looking strategically at relieving cross-river accessibility and capacity pressures, and with better access to Gatwick Airport also in mind.

    But the big ticket items are for the longer term as possibilities. Penge East/West remains about the only short term intervention which might be possible. Given that the London Bridge rebuild will improve the quality of interchange there between SC and SE, and London Bridge is only 12 minutes on a non-stop train from East Croydon, heading via Central London is probably still the most efficient way for many people of getting across the ‘Berlin Wall’ – though it doesn’t directly address your desire to reach Bromley South – it would be Bromley North instead.

  230. Malcolm says:

    Jonathan: Interesting collection of options. But this desire line (Croydon – Bromley) is just one of an enormous number of orbital journeys with currently unsatisfactory provision. Surely the strategic issue is whether improving any of these is likely to be a better use of money than spending on a radial journey, or spending on some access for an unserved area. Both Croydon and Bromley are well-served by rail to a bigger-than-usual range of destinations (which just do not happen to include each other).

    (And East Croydon to Bromley South is achievable via Victoria or Clapham Junction, though whether that makes as much sense as via London Bridge I couldn’t say).

  231. 100andthirty says:

    Malcolm……I thought something didn’t compute when you mentioned East Croydon to Bromley South via Clapham Junction. As far as I can see from National Rail Enquiries and from my increasingly dim recollection of South London routes, you can’t get directly to Bromley South from Clapham Junction. You have to go via Victoria. I was sceptical about NRE, though as it showed East Croydon to Beckenham Junction as costing £12.10 by Tramlink!

    I would say that East Croydon to Bromley South via Beckenham Junction is the easiest journey. There are six trams an hour and four trains, but NRE shows four journeys an hour – the quickest being 31 minutes and the longest 41 mins – taking the best tram time to integrate with the Beckenham Junction – Bromley trains.

    I would suggest that a main line connection between south central and south eastern in the borough of Croydon is going to use up valuable capacity that would be better used serving the larger community who want to make radial journeys.

  232. timbeau says:

    @Jontahan Roberts
    “a better BCR might be achieved by relocating Penge West station northwards”

    There are frequent proposals for relocating certain south London stations to improve interchange (Tulse Hill is a favourite). Does the convenience of a relatively small number of potential interchange passengers trump the probably much larger number of users who find the station’s existing location on Penge High Street very convenient, thank you very much.

  233. timbeau says:

    Clapham Junction to Bromley South can be done without going via Victoria, by going via Crystal Palace and Beckenham Junction (the original route of the LCDR) – but if coming from Croydon there is no need to backtrack between Crystal Palace and Clapham Junction.

  234. 100andthirty says:

    Timbeau. Understood – but it’s really slow!

  235. ngh says:

    West Croydon has far more Norwood Jn services if starting out in the Croydon area (rather than may be changing at East Croydon or having that as starting point)

  236. Pedantic of Purley says:

    Or one could just get a 119 bus direct from East Croydon to Bromley South.

  237. Jonathan Roberts says:

    Agreed that Penge W station location is fine for existing users (partly a self-defining usage, of course, driven by its very location!). Relocation of platforms to the north side of the High Street rather than the south side would however make no real change to current local accessibility, but would allow an interchange link to start at the furthest north part of the new platforms so say 250-320m closer in walking distance to the SE tracks, after allowing for the access distances within the present Penge West station.

    At roughly 80m per minute for walking, that’s say 3-4 minutes shorter interchange, valued in perceived walking and waiting terms at 2½ x actual = 7½-10 minutes effective time saving. As observed above, whether such time saving on the interchange could make it sufficiently attractive to more than repay the cost of shifting the station, is a moot point, but it looks like the way to maximise the interchange potential at that location without inconveniencing existing users, if that were your priority.

  238. Pedantic of Purley says:

    Jonathan Roberts,

    I know Penge West is on the target list for many Accessible Access groups. Apart from Norwood Junction it is just about the only restricted access London Overground station in the area. If relocating were possible and desirable then disabled access might make the whole scheme more attractive.

  239. Jonathan Roberts says:

    @Malcolm 8:12
    Yes quite. The Croydon-Bromley desire line is one of many. In the context of current housing and economic activity, the existing railway infrastructure is quite unforgiving and doesn’t really permit many permutations that might be worthwhile, or feasible or operable, or show much sign of a worthwhile business case.

    The context considered by TSLO was large scale housing (hence population and some jobs) growth in outer London over the next three-plus decades. Outer London uses cars more because of lower densities and weaker public transport in aggregate (various specific lines and Tramlink excepted). There is a serious risk of greater surface transport congestion and slower journeys if the status quo were maintained on the public transport front.

    Hence the review to see whether something might be possible. This followed on from the London 2050 Infrastructure Plan thinking of an Outer Orbital which however had poor BCR because of the scale of new railway links with high capital costs. The ‘R25’ looked more selectively at trying to see if a join-up of major towns was feasible – it was – along an urban corridor within south London, which might best repay the rail investment with decent usage volume, and also a series of significant towns in the neighbourhood of London, paralleling the M25.

    But anything on that scale, as said above, would be for the long term, not least because you’d need the population growth to start justifying any such scale of investment! Most transport planners would probably rather see the eggs emerging in the housing nest before the chicken starting laying infrastructure… However Canary Wharf’s future needs might help justify some of the latter.

  240. Southern Heights (Light Railway) says:

    timbeau @ 09:26: Does the convenience of a relatively small number of potential interchange passengers…

    These are probably the “known numbers”, but not the “suppressed demand” that is also there…

    Brockley is another place where such an interchange would be possible and also has the potential to open whole new possibilities for journeys. Especially those avoiding Zone 1, something everyone can agree on is to be encouraged…

  241. Anomnibus says:

    @Jonathan Roberts:

    You’re not wrong about the “Berlin Wall” effect (ironically created mainly by the growth of the rail network). There is, in fact, no “South London” postcode: only “SW” and “SE”. (“S” is used for Sheffield, I think.) Compare with the alphabet soup of postcodes allocated north of the river, which spread out far more quickly both before and during the Victorian era. But I digress…

    Tramlink is quicker if travelling from Beckenham to Croydon. If orbital rail links are to be improved, I suspect Tramlink extensions would offer the best value over heavy rail, and would also be better than trying to use it to provide additional capacity on radial routes. Building billion-pound tunnels for a mere 4 tph is never going to wash its face. Building an extension from near New Addington to Bromley via Hayes, on the other hand, would be trivial given that it can mostly skirt the edge of the Green Belt.

  242. Anomnibus says:

    The “Penge Interchange” concept seems to crop up every couple of years or so, but there are some technical issues that make it difficult to justify…

    Despite its name, Penge West was mainly built to serve the nearby entrance to the Crystal Palace’s park and its surroundings so moving the station too far away from there won’t be popular. Also, don’t let the road’s name fool you: Penge’s shopping centre is some distance away down the hill. Penge East is the more convenient station, despite its location.

    Penge East also doesn’t butt right up against the Penge Tunnel: the nearest platform is around 140 metres away from the Croydon tracks. Relocating these platforms would mean moving it away from Penge’s shopping centre. 140 metres doesn’t sound like much, but Penge East’s proximity is already borderline: For much of the shopping centre, Kent House would become the more convenient station!

    So, neither station can be moved without invoking the wrath of the locals.

    Finally, even if relocating both sets of platforms were viable, the flying junction with the Crystal Palace branch would make it very difficult to build viable footbridges between the two sets of platforms. (Immediately south of the Penge Tunnel mouth, you have: the Up line from Crystal Palace, the four main tracks from Croydon at a lower level, then the Down line towards Crystal Palace at a higher level again.) You’d likely need to punch holes into the hillside to provide necessary subways instead. This isn’t going to be cheap, and the only benefit of doing this is an interchange; the locals would actually lose out.

    Which leaves us with the “really long footpath” option. Few would use it, it would be expensive to build and maintain, and it wouldn’t be noticeably better than the existing route via the local road network. A convenient interchange it ain’t.

    Building platforms on the Lewisham-Nunhead route above Brockley makes more sense to me. It would require replacing the bridge, but it would create a much more compact and easier interchange. It also opens up a lot more potential journey opportunities than the Penge option. Being one or two stops from Lewisham and Peckham Rye makes an interchange here far more useful than one in Penge.

  243. Southern Heights (Light Railway) says:

    @Anomnibus: Any such extension would encounter the following degrees of resistance:

    1. A rather severe gradient on the A232 Eastwards from Addington Road
    2. A huge number of local objections.
    3. If going past Hayes station, you would require street running along Addington Road, Bourne Way and Ridgeway at least, then how? Squeal around the backyards of houses?

  244. ngh says:

    Re Anomnibus,

    Largely agreed but rebuilding Brockley lane wouldn’t provide a link to Bromley South or other destinations on the LCDR Victoria – Kent routes but would improve via Lewisham connectivity improvements and that segment of SE London but is probably much more likely to happen than Penge.

    Penge is very close (just a gap of 550m) from Anerley so moving closer to Penge East make sense.

    The quality of the railway earthworks in the Penge “interchange” area is dubious so I doubt anyone would want to touch them.

  245. Malcolm says:

    Tramlink extension to Bromley: Let’s all please leave our crayons in the box on this one. “Trivial” seems most unlikely, however close the green belt might be; it’s certainly not going to happen any time soon; but we should probably never say never.

    The general principle, though, of using light rail to make orbital links, is a perfectly valid notion to advocate in an abstract way. Paris springs to mind.

  246. Jonathan Roberts says:

    No-one so far I’m aware has proposed relocating the Penge E station, that’s in your imagination, only PW has been considered, and discussed here. You can avoid the CP Down Line conveniently by staying on the down-from-LB side until you need to cross over the 4-tracks LBSC to reach any newly relocated NB platform. Any prospective interchange doesn’t need a mini-mountain to replace an already discouraging molehill!

  247. Anomnibus says:

    @Jonathan Roberts:

    If you leave both stations more or less as they are, then there’s no need to bother with a dedicated (and expensive) footpath that must perforce include a dogleg to get around the existing properties. You might be able to justify a footbridge over Penge High Street to take the bulk of the traffic out of the equation, but the existing walking route is otherwise perfectly sufficient.

    Engineering issues aside: what, exactly, is the proposed benefit here? Croydon to Bromley via Penge is quite the roundabout route, and would be slow and inconvenient. The long walk between the two sites effectively rules this interchange out for shoppers. This is where buses and trams excel: they may take longer, but you’re not schlepping multiple bags around from platform to platform, and you get picked up and dropped off right on the street. It’s not always about speed.

    I also don’t think there is much latent demand for people living in Bromley wanting to commute to Croydon, or vice-versa. Bromley has nowhere near the size, usage mix, or density of Croydon. I’d argue that Lewisham — particularly since its regeneration — is a much better choice of ‘anchor’ destination for an orbital service than Bromley.

    [No more from anyone about improving the Penge interchange (or not) please. There is no sign of opinion convergence. Malcolm]

  248. Jonathan Roberts says:

    Happy to agree that Lewisham is a key node for orbital usage!

  249. Walthamstow Writer says:

    Alternatively someone could exert some pressure on Bromley Council and TfL to work together to provide some better bus priority and then run a decent quality express orbital bus linking Bromley and Croydon. We know TfL really dislike express buses but, unfortunately for them, the public like them based on the usage of the X26, X68 and 607. Of course, eons ago there was a longstanding orbital route doing as I suggest (no history lessons please!). This would save anyone moving stations, building footpaths or tunnels without a business case. One day there may very well be a case for a substantial rail investment in these links but the short term solution is easy (relatively) and far, far more affordable. It can’t be beyond the wit of the relevant organisations to get past their institutional prejudices and do what is right for those wishing to travel.

  250. Graham Feakins says:

    Already mentioned elsewhere was the long-term official plan to run the Beckenham Junction branch of Tramlink to Bromley (North) via Shortlands but Bromley Council put the mockers on that at an early stage. The whole point was that Croydon Council/London Transport recognised that there was a need for a good radial link between Croydon and Bromley, contrary to the comment made by Anomnibus and hence the push in that direction to Elmers End/Beckenham Junction. The 119 bus route doesn’t really cut the mustard with its convoluted routeing.

  251. timbeau says:

    “There is, in fact, no “South London” postcode: only “SW” and “SE”. (“S” is used for Sheffield, I think.) Compare with the alphabet soup of postcodes allocated north of the river, which spread out far more quickly both before and during the Victorian era. ”
    On a point of fact, there is no real difference between north and south. The cardinal points (except NE and S) were used within the old county of London. Beyond that the alphabet soup (DA, BR, CR, SM, KT / TW / UB, HA, EN, IG, RM) is distributed equally north and south of the river within Greater London .

  252. Graham H says:

    @timbeau – just as a bye-comment on postal districts (and rather off topic), the London postal districts originated with the GPO’s horsedrawn van delivery circuits setting off from the GPO in the 1820s, at a time when S London was hardly developed. They were simply divided into E, W, and so on. Numbers came later (associated with the devolution of sorting to area offices? Penny post? Rail transport?) and extensions beyond the first round of numbering came towards the end of the C19. This took them well beyond what had been the old LCC boundary, although,of course, the division into districts long predated the LCC. Within each geographical area, the principal sorting office took 1, followed by the subsorting areas in alphabetical order (the assumed suburb name not always the most obvious however). The later expansion was also in alphabetical order.

  253. Greg Tingey says:

    the public like them based on the usage of the X26, X68 and 607. …
    And call them “Green Line Coaches” ??

    Graham H
    IIRC, the full development of the earlier postcodes (as we would now call them) came during the first international unpleasantness with Germany, to help the newly-recruited people drafted in to replace “normal” postmen away at the fighting ….
    And included areas that were certainly not within the LCC area, such as E17 (Walthamstow) f’rinstance

  254. Graham H says:

    @Greg T – a quick look at Wiki (ho hum!)suggests that there was only one round of numbering of the subdistricts, and that in 1917. (I was wrong to suggest two rounds). It also suggests that when the geographical system was first formalised for public use in 1856 there were indeed S and NE areas;these were abolished in 1866/8.

    One of the oddities of the system is that the boundaries introduced at the time went far beyond the built up area. Ealing (and to keep on thread!) Norwood weren’t reached by the continuous built up area until very much later. Think of all those Impressionist paintings of rural Penge…

  255. Pedantic of Purley says:

    Of course, the NE postcode still exists as any Geordie will tell you.

  256. Graham H says:

    Not to mention the S postcode…

  257. Alan Griffiths says:

    Graham H 31 August 2016 at 11:37

    “Not to mention the S postcode…” which extends well beyond Sheffield to cover Rotherham, Barnsley and deep into Derbyshire.

    On my first visit to Clapton, in 1973, I was surprised to see one street name sign that read NE rather than E5.

  258. Graham H says:

    @Alan Griffiths – not to mention the way that GU seems to cover much of east Hampshire. (It’s an interesting question, but probably not for this thread, as to whether the present postcode system – introduced c1970 – has actually fossilised the PO’s delivery arrangements as they were when the MLO programme was introduced or whether these days the present generation of MLOs use an aggregation of convenient but unrelated codes.)

  259. MLO is the Post Office Mechanised Letter Office, which automated letter sorting under their Letter Post Plan.

  260. Anonymous says:

    I’d really appreciate it if anyone on here could answer a couple of local railway history trivia questions about the Norwood Junction area:

    1) was the line from Corbetts Lane to East Croydon the first significant length of 4 track railway in the world? It was quadrupled in 1844. I note that London Bridge to Corbetts Lane junction was quadrupled in 1841, but this was only 1.5 miles.

    2) Was Norwood Fork the first flying junction in the world? As I understand it, it was built as part of the quadrupling. Wikipedia suggests Weaver Junction was the first, in 1881.


  261. ngh says:

    Re Anonymous,

    1) Yes though very dependent on definitions around termini. Also worth noting it included the first closures of level crossings (and replacement with bridges) due to traffic levels on railways!

    2) Yes, The Wikipedia Weaver jn claim was debunked under one of the previous Sussex articles (part 7?). Wikipedia also points out Norwood Fork was far earlier…
    Someone really should edit the Weaver Jn wiki.

  262. Slugabed says:

    Re Norwood Fork Junction
    I have been led to believe that the flying Junction at Norwood Fork was introduced as part of the Atmospheric Railway experiment,as crossing the pipe on the level presented difficulties.
    I’m prepared to be corrected on this,though.

  263. ngh says:

    Re Slugabed,

    Correct. It was also originally paired by use rather than direction though from memory that went in circa 1848.

  264. Anonymous says:

    Thanks for the quick responses. You’ve confirmed what I thought, but is there any evidence for it?

    As I understood it, the lines were always paired by direction as the slow lines were built by London & Brighton railway with two fingers up at the SER who had running rights over the centre tracks, limiting their future expansion?

  265. RayK says:

    Looking at the Wiki articles it would appear that they are regarding atmospheric railway use as something different and strange and therefore not railway in the standard sense. This means that it was ‘something other’ until the atmospheric railway ceased to exist so that the flyover was not strictly standard railway. It seems to me to be a rather over pedantic way of looking at it. I note that the Weaver Junction article quotes the Guinness Book of Rail Fact and Feats as it’s source. This suggests that that is the body that needs to be approached by anyone wishing to change the record. (Pun intended)

  266. ngh says:

    Re Anonymous,

    I’ll have to have a detailed check in the books (last looked at for research pre part 7 starting from the canal history) but logically the pneumatic experiment had to be paid by use. Also possibly worth having a quick look at the Christmas Quiz answers from 2 years ago where there was an extensive discussion of Penge, the 4 tracking and pneumatic then. By the end of the 1840s the SER had it own preferred direct route to Kent so was less worried by the via Croydon and Redhill one. Swapping to paired by direction was also more useful for the LBSCR local sidings along the route.

  267. Kit Green says:

    To not be recognised as a conventional railway (itself odd as the traction type is not the defining factor) we would have to accept that only atmospheric trains used those tracks for the whole time the system was in place.

  268. Anonymous says:

    I suppose you would also have to prove a junction between the atmospheric and non-atmospheric railways – otherwise it’s just an intersection bridge, and unlikely to be the first by 1844?

  269. ngh says:

    Re Anon,

    The atmospheric railway was converted to conventional by May 1847 upon the opening of the Epsom – West Croydon Branch with increased through traffic through West Croydon to London Bridge at which point it was definitely operating in the same way as today. So 1847 instead of 1844 but still the first.
    All the earlier intersection but no connection bridges were as close to perpendicular as possible for engineering reasons but this wasn’t build like that.

    [After the formation in 1846 of the L&B and L&C and the smaller south coast railways into LBSCR following the first major falling out with SER in 1846 and the failure of atmospheric experiment in 1846 there was a big infrastructure sort out to resolve lots of issues.]

    Pre-late 1850s iron railway bridges would have been prohibitively expensive so would have been avoided unless really necessary so it is very likely to have been the first flyover. (Another technological –> price break through in mid 1870s led to wrought iron bridges being commoditised after that point and usually chosen over brick).

  270. Purley Dweller says:


    Presumably that is why the widened line South of East Croydon has girder bridges next to the original brick ones.

  271. ngh says:

    Re Purley Dweller,

    Yes indeed! I did think about using that as an example with the recent Old Lodge Lane replacements (south of Purley) in lots of people minds – a good example of cheap 1890s but poor quality and design that needed replacement like lots of others nationally of pre WW1 vintage. Certain designers and contractors where far better (e.g. excellent Dorman Long etc. ones with welded steel from 1902 onwards) so there are some good pre WW1 examples but also far more bad ones.
    30+ metal ones are being repaired or replaced around to London Bridge during the works and a huge number on GOBLin are getting attention too (responsible for plenty of 20 and 30mph line speeds) including after re-opening. Also lots on the Southern WCML from the problem vintages being replaced in recent years.

  272. Purley Dweller,

    When ngh and I are in full nerd mode in the pub this is the sort of thing that gets discussed. The analogue with concrete is that around the 1860s (which very fortunately and coincidentally marked the start of the building of the Underground – Thames Tunnel excepted) the importance of getting a good consistent, quality-control-checked mortar was understood by, amongst others, Bazelgette.

    There is a difference though. Metal bridges are usually at their strongest on the day they are built whereas a well-built masonry bridge should get stronger with age due to exceptional properties of cement-based compounds.

    It has been calculated that the cost of rust to the economy amounts to about 3% of GDP. Not many people known that as a famous actor supposedly said.

    Apologies to self and other moderators for going off-topic.

  273. Anonymous of Croydon says:

    Ngh. Wasn’t it by the end of the 1860s (not 1840s) that the SER had its own route to Kent, i.e. the route via Sevenoaks?

  274. Timbeau says:


    It was, but they were not best buddies for a long time before that.

  275. Southern Heights (Light Railway) says:

    @PoP: You might see rust as costing 3% of GDP, others might see it as an important source of job creation and income! 😉

    @Anonymous of Croydon: “In May 1862, authorisation was obtained to construct a new railway from St Johns, London to Tunbridge” (from Wikipedia). It opened in 1868…

  276. ngh says:

    Re Anon of Croydon,

    Oops a coffee free 2017 not off to good start must stay awake.

    Re SH(LR) and PoP,

    The last estimate I saw (circa a decade old) was that corrosion not just rust had fallen to just 2% of GDP as result of efforts to reduce corrosion (see average length of car lifespan in the UK almost doubling as an example as rust isn’t taking many off the road). As it is lagging index it could well be lower in practice on new goods currently (viewed at the point in the future they are replaced) which leads to an increased up front cost for corrosion prevention thus providing a small GDP boost now in return for a larger GDP reduction later.

    [Sarcasm] ONS will fudge it with an imputed corrosion measure as GDP can’t be allowed to fall and then the BoE will ponder which productivity has fallen when no real revenue generating activity rather than imputed activity has actually taken place despite the obvious staring them in the face. (See imputed rent and imputed household maintenance fudges in at circa 10% of nominal GDP currently) [/Sarcasm]

Leave a Comment

In order to make LR a pleasant place for discussion, please try to keep comments polite and, importantly, on topic! Comments that we feel do not meet these criteria, or that contain language that could cause some people trouble at work, may be moderated or deleted.

acceptable tags

* (This won't be shown, but you can link it to an avatar if you like)

Recent Articles

In Pictures: The Northern Line TBMs Arrive


TfL have released a number of images showing the Tunnel Boring Machines (TBMs) for the Northern Line Extension. These were lifted into place at Ferrovial Agroman Laing O’Rourke’s TBM portal at Battersea earlier this month. The images highlight that whilst

Read more ›