Kent Route Study (Part 3): Victoria Metro Services


As part of our series on the recent Kent Route Study, we look at what it refers to as the ‘Victoria Metro Services.’ It is nearly impossible to do so, however, without also looking at services via the Catford loop that either terminate at Blackfriars or continue through central London via the Thameslink core. As a result, those services are covered here as well.

Included within the Route Study in the Victoria Metro Services is a slightly anomalous service from Victoria to Dartford via Nunhead and Bexleyheath. This is really quite separate from the other services and we hope will look at this in isolation in a separate article. It will not be covered here.

A slightly modified SouthEastern diagram showing the routes covered by this article

Back to pre-Channel Tunnel days

To understand the main routes we are talking about and the potential options available, we have to go back a long way in time to the days of the 1970s, 1980s and early 1990s – way before the Thameslink Programme was being implemented. In fact we need to go back and look at services prior to the opening of the Channel Tunnel.

Before Eurostar, in the days when getting a train to the continent meant the boat train, many of the services in the South Eastern Division of the Southern Region of British Rail had remained largely unchanged for a long time. This was certainly true of the group of services that generally started from either Orpington or Sevenoaks. The Sevenoaks services would be routed via the delightfully named Bat & Ball station. These services would continue to Victoria or Blackfriars either on the main route via Kent House and Herne Hill or via the Catford loop line (via Bellingham).

Bat & Ball station in 1962. Photo by Ben Brooksbank

Of the two routes to Victoria from Shortlands available to fast trains, the direct one via Kent House and Herne Hill was (and still is) generally used in preference to the alternative via Bellingham and Catford. One reason for this is a small time saving, but another consideration is that passing loops at Kent House meant there was more capacity for a mix of fast and slow trains.

Trains start and terminate at Kent House

Current day layout at Kent House (as taken from Carto Metro)

As an alternative to using the four tracks through Kent House station to enable fast trains to overtake slow trains, platform 2 at Kent House could be used to terminate short workings from Victoria or Blackfriars. Whilst terminating a train at Kent House potentially prevented the opportunity for a fast up train to overtake a slow up train, it also meant that there was a unoccupied train path between Shortlands Junction and Kent House, which reduced the likelihood of a fast train being slowed down in the first place.

The service in the good old days

In addition to the facilities at Kent House, there were also options available for the same manoeuvres to be carried out at Herne Hill – provided the relevant outermost platform was not occupied by a train travelling via Tulse Hill and/or Elephant & Castle.

The services in question had been well established for many years and can be thought to have consisted of:

  • An all stations service all day from Orpington to Victoria. This was really the backbone of the services provided.
  • An all stations service all day from Sevenoaks to Blackfriars via Bat & Ball. This provided an alternative destination for those joining east of Beckenham Junction and ensured that stations on the Catford loop line had a service.
  • Various supplementary short distance peak period services via Herne Hill to provide extra capacity. Typically these might start from Kent House in the morning or terminate at Kent House or Beckenham Junction in the evening.
  • Various supplementary trains from Orpington that ran fast from Beckenham Junction to Victoria in the morning and vice versa in the evening. Whilst these provided a faster service for those further out in the suburbs, the primary operational benefit was that these trains didn’t further stifle capacity on the two track section between Shortlands and Brixton since they did not call at any stations on this section of line. It was therefore possible to regard them as fast trains when on the two track section and so they fitted in very well with the succession of other fast trains originating from outside London.

Off-peak it wasn’t so great

For much of the 1970s and 1980s the off-peak services was much sparser. The Catford loop only had a half-hourly service. More surprisingly, for many years, the quite busy Orpington – Victoria service was also only half-hourly. Even in those cash-restricted days, it was said that British Rail wanted to run every 15 minutes on the Orpington – Victoria service during the working week. The option of a 20 minute frequency was not really possible on a two track line with so many fast services based on a 30 minute interval service.

The reason for not having a 15 minute frequency on the Orpington-Victoria service was stated to be that the signalling did not allow a sustained 15 minute service throughout the day to be mixed in with the fast services from the coast. As if to verify this, the 15 minute off-peak service was provided some time after the line was resignalled and control centralised and moved to Clapham Junction.

Eurostar needs train paths

With the opening of the Eurostar route from Waterloo to the Channel Tunnel in 1994, the regular pattern had to be altered to fit in the Eurostar trains. This required a complete rewrite of the timetable and the only saving grace was that in the morning peak the Eurostar trains were predominately going against the local peak flow.

Things could have got much worse in 2003 when section 1 of what was then called the Channel Tunnel Rail Link was opened. This meant that Eurostar trains destined for the Channel Tunnel could not stay on the southernmost pair of tracks from Shortlands to Bickley in order to travel via Orpington and the Sevenoaks tunnel to Channel Tunnel portal. Instead there would be a need to travel via Swanley and join the Channel Tunnel route near Ebbsfleet. To do this they would need to cross tracks on a flat junction at either Shortlands or, less likely, at Bickley or Swanley. Having a train nearly 400m long crossing vital junctions, on the flat, during peak hours using the railway infrastructure that existed at the time would have probably wrecked any semblance of the existing metro timetable.

To avoid a Eurostar conflict at the critical junctions, there was the alternative option of sending Eurostar services via the Catford loop line. This would have added only 3 minutes to the journey time, but would have meant traversing more junctions (albeit less critical ones). More problematically, as we shall see, the Catford loop line, despite the relative paucity of trains, does not have spare capacity due to its awkward mix of stopping and non-stopping trains.

The issue with Eurostar trains was not one that could be fixed by mere re-timetabling. To resolve this major issue, the sort of thing that can cause diplomatic disquiet if handled badly, the decision was taken to provide a grade separated junction, in place of the flat junction, at Shortlands. This would minimise conflicting routes. The cost at the time (£80 million), seemed enormous, for something intended to resolve a problem that would only last for four years – by which time the full Channel Tunnel Rail Route to St Pancras would be open.

The Kent Route Study acknowledges the usefulness of the grade separated junction when referring to capacity restraints:

Shortlands Junction, where services via Herne Hill and the Catford Loop diverge, where a previous grade separation scheme has removed many of the clashes

The Route Study follows the mention of Shortands Junction as a constraint with a mention of other significant constraints. These were identified as Swanley Junction (where trains to Sevenoaks via Bat & Ball diverge from the main line), the mix of fast and slow trains and signal spacing.

Eurostar trains no longer take up train paths

In November 2007 the Channel Tunnel Rail Link was opened all the way through to St Pancras. This put Victoria and Blackfriars South Eastern metro service possibilities back to where they were before, but now with less of an issue at Shortlands junction. In fact, not only had Shortlands Junction been upgraded, but enhancements had also earlier been made to the Chislehurst junction complex as a result of needing to introduce the Eurostar services. In addition to that, Orpington now had eight platforms (four terminating) instead of the six platforms (two terminating) that it had in the days prior to the Channel Tunnel. These additional platforms were also created as a consequence of Eurostar. The previous option of terminating on the through tracks at platforms 4 and 5 at Orpington was simply no longer tenable.

Trains no longer start and terminate at Kent House

With the Channel Tunnel related enhancements still in place, but no longer any Channel Tunnel trains, the benefits of running the limited peak service of trains starting from or terminating at Kent House made much less sense as the conflicts caused by running to the east of Beckenham Junction had largely disappeared.

Then Thameslink comes along

The benefits of being “back to normal” were not to last for long. With the final approval of the Thameslink Programme came the need for the bay platforms at Blackfriars to be taken out of use for a number of years (and the eventual replacement of only two out of the original three). Any service using the bay platform needed to be incorporated into the temporary Thameslink timetable in operation for the duration of the works – or removed from the timetable.

And so it was that potential for a stable pattern of services came to an end in March 2009 when the terminal platforms at Blackfriars were closed for major reconstruction of the station. This meant that any service to Blackfriars had to be a Thameslink through service.

The timetable was amended to try to ameliorate the situation by running various peak hour Thameslink services that really did not fit any pattern. So today we have, for example, the 07:28 Orpington to Bedford service via Herne Hill, the 07:47 Orpington to Luton service via Bellingham and Catford not calling at Ravensbourne or Beckenham Hill and the 08:02 Beckenham Junction to Bedford service.

The baseline timetable

The “baseline timetable” for Victoria and Blackfriars in the Kent Route Study for Victoria metro services is far from obvious. Indeed the Kent Route Study does quite a good job of referencing the baseline timetable without actually telling you what it is. A big help is provided by means of an extraordinary spider diagram of trains arriving at a London terminal between 08:00 and 08:59. This is figure 2.7 in the Route Study. Unfortunately it does not show stopping patterns, so is not quite as informative as it could be. One cannot distinguish between a train passing through a station and one stopping there.

The extraordinary spider diagram in the Route Study

The first real surprise in the baseline spider diagram is that it is extraordinary how many peak hour trains will be nowhere near the maximum length permitted. Getting very slightly off-topic, of six trains from Hayes arriving in central London in the morning peak hour only two are 10-car and the other four are 8-car. Needless to say none are 12-car. This is on a line which Network Rail shows to be the most overcrowded.

The second surprise, highly relevant to the current topic, is that the baseline timetable shows just four trains arriving in the morning peak hour at the two terminal platforms at Blackfriars.

To find what these are you have to be good at following lines to see where they lead. The spider diagram seems to show that two of the trains terminating at Blackfriars are from Maidstone East via the Catford loop line and the other two are from Orpington via Herne Hill. In fact the latter are the 07:28 and the 07:47 services we referred to earlier. The 08:02 from Beckenham Junction, also referred to earlier, seems to have disappeared but there will be a new 08:21 Beckenham Junction to Victoria service consisting of just six carriages.

Also of possible interest are the 07:34 and 08:04 Bromley South – Victoria. Assuming these are all-stations services, these almost seem to be a replacement for the Kent House – Victoria trains of a previous generation. It is not apparent how this would work in the timetable. Do they terminate at platform 2 at Bromley South in the peak period or do they start empty from further down the line and first pick up passengers at platform 1 at Bromley South?

The Catford loop service

In the Route Study the Catford loop is referenced in the Victoria metro services and the Blackfriars metro services which includes Thameslink.

Something that will be very different pre- and post-Thameslink Programme is the Catford loop service. The Catford loop line is a bit of an operational nightmare. The basic problem is that there is nowhere for one train to overtake another anywhere between Shortlands Junction and west of Peckham Rye. More realistically, it needs to be treated as double track without passing places between Brixton Junction and Shortlands junction. Not only does it have a mix of fast and slow traffic, it is also a freight route – though no freight is scheduled during peak hours.

Difficulties operating the Catford Loop

To make matters worse, the short section between Peckham Rye and Nunhead has to also accommodate the half-hourly Victoria – Dartford services. A further consideration is that the nature of the stations along the line is very different. The southernmost two Beckenham Hill and Ravensbourne are very quiet with Ravensbourne having a surprisingly rural feel about it. Bellingham is somewhat busier with Catford being the busiest station on the loop – but it is nowhere near as busy as the adjacent Catford Bridge station. As a group these stations are probably half as busy, if that, as the stations on the direct route to Victoria via Herne Hill.

Future prospects for the Catford loop service

Attitudes change over time and it is now thought that a substantial reason for the lack of patronage on the Catford loop is due to the poor service. There now seems to be a determination to provide a 4tph (train per hour service) in the off-peak. The Route Study does not directly discuss this so we will have to see what, if anything, happens about this.

A further issue for the loop is the need to provide something more than 2tph in the peak. With the line needed for fast services (eg Thameslink to Maidstone East), having a 4tph regular all stations service is not a practical option. The current solution is to resort to having fast trains stop selectively at individual stations. This leads to very uneven intervals at some stations. All the baseline scenario commits to in the Route Study is:

Additional service on the Catford Loop to retain current service levels (seven trains over the 3-hour peak period (07:00-09:59)

which actually seems to be considerably worse than what is provided today.

On a slightly different issue, in preparation for the eventual Thameslink timetable for the recast services, whatever they may be, the stations on the Catford loop line are now managed by Thameslink. This may have been done because it was envisaged that the bulk of trains on the loop would be Thameslink trains to Sevenoaks via Bat & Ball. Shortlands station and stations east thereof continue to be managed by Southeastern. This even applies to Bat & Ball station, which, as explained by Geoff Marshall in one of his many delightful station videos, is managed by Southeastern even though the service is almost exclusively provided by Thameslink services.

The takeover by Thameslink of the Catford loop services and stations does make a lot of sense when you consider that 8-car Thameslink trains are stabled at Bellingham sidings. Whilst sidings can be reallocated, the Thameslink trains have used these exclusively for a long while now.

What the Route Study offers

Finally, getting to the Route Study and its utterly predictable anti-climax, anyone looking for infrastructure improvements will be severely disappointed. We discover that there are virtually no plans for infrastructure improvements along the relevant routes. In fact the only real infrastructure improvements proposed are those to stations.

5.15.4. At Denmark Hill, the main issues identified are congestion on the platforms, stairs and interchange footbridge, and at station entrance / exit gatelines, both in the morning and evening peaks. By implementing the proposed interventions, it is anticipated that there will be reduced queuing at the bottom of the platform access staircases and decongestion at the main gate lines, with improved passenger safety and reduced passenger walk times.

5.15.5. At Peckham Rye, crowding and congestion have been identified in both the morning and evening peaks on platforms, access stairs and at the main station entrance/exit gateline. The options identified will inform choices for funders in the short/ medium term, and input into wider regeneration schemes being master-planned for the area.

5.15.6. At Bromley South, passenger crowding in both the morning and evening peaks has been identified on the station platforms and interchange bridge. Removal of buildings on Platforms 3 & 4 will aid in short/medium term decongestion in these areas, but consideration of longer-term congestion relief options should be made to resolve future capacity concerns.

5.15.7. At Brixton, passenger crowding occurs with passengers leaving the station from Platform 1. The only exit is a metal staircase to the ground level. There is no access to Platform 2 either. Passengers queue back from the staircase onto the platform making train dispatch difficult as passengers wait on the wrong side of the yellow line or struggle to alight the train.

In the case of Denmark Hill one can’t help thinking that really there should have been a comprehensive integrated upgrade done when the Access for All facilities such as the new footbridge were put in a few years ago. For Bromley South one has to go back further, but multiple opportunities have been missed with the building of the relief road above the tracks at the country end of the station and the shift of the epicentre of the town away from the southern part of the High Street outside the current entrance.

On the line capacity issue, we have the unimaginative proposal frequently trotted out by Network Rail that increasing train length where needed to the maximum allowed is all that is necessary – and if it isn’t then run trains that have a higher density rolling stock capacity (generally interpreted as doublespeak for fewer seats per carriage).

3.9.2. The peak services on these routes operate in 4-, 6- or 8-car formation. The maximum length that can operate on metro services currently is 8-cars due to platform length restrictions on the route and at Victoria. The analysis indicates that six additional vehicles will be required by 2024, which could be accommodated through lengthening existing services to 8-cars or switching to higher density rolling stock. All the platforms on the route can accommodate the services lengthened to 8-cars or some alternative higher density rolling stock types.

As with the London Bridge metro services, this conclusion seems hard to believe given the expected population increase and the belief that, even if the population remains stable, the number of journeys increases as we become a more mobile society.

At least the Route Study does discuss what restricts capacity in case it is envisaged that more is necessary. It does briefly mention extending the length of metro trains via Herne Hill but quickly dismisses this because it has a BCR of 0.4 – 0.6. One may well question the methodology.

On the basis of extra generated income in the short term, it is fairly obvious that train lengthening schemes in suburbia are going to struggle to make a good business case as the extra income derived is going to be relatively small and the costs of the project in a crowded city will probably be large. Yet, intuitively, many people would argue that being able to catch a train to work in the morning is just something that is necessary for a city to function. Possibly the basic problem is that Network Rail are looking for something to increase capacity by 10-15% in the short term and are not looking too hard at the longer term. A longer maximum train length potentially gives you 25% extra capacity at least and this is seen as overpriced overkill.

Potential infrastructure improvements ignored

Other options looked at are addressing the age old problem of junction conflict at Herne Hill. This does get a very brief mention

In addition to the platform extension scheme at Herne Hill station, a flyover scheme has been looked at, which would unblock this bottleneck.

This doesn’t really provide enough information to enable stakeholders an opportunity to make an informed choice. It does not even mention the bolder Turn South London Orange proposal of tunnelling under Herne Hill on the Victoria – Kent House route to enable fast trains to avoid the station altogether and at least remove the conflict between fast and stopping trains.

Also getting a very brief mention is the possibility of work in the Kent House area:

Towards Kent House, the option for some four tracking has been examined.

This, almost certainly, should have been a reference to a prolongation of the existing four tracking rather than worded to suggest that currently the line was only double track at this location.

Absent in the Route Study is any kind of explanation of why longer four tracking at Kent House would be helpful, although one can hazard a guess. Network Rail’s lukewarm-at-best support for this extension of the four track section at Kent House is probably pipped by TfL’s enthusiasm – it is one of the very few track infrastructure proposals they made in their submission for taking over Southeastern services. However, even TfL, in their devolution of rail services in South East London submission, talk of “small scale investments that could be pursued” rather than outright positive enthusiasm.

Hard done by or treated reasonably?

There are many areas of the country where commuters feel that they get a raw deal. Those from the South Eastern suburbs of London would probably have more good cause than most for feeling that it has been many decades since there have been significant improvements made (especially with regard to capacity) with their interests primary at heart. They have lost direct services from Bromley North and on the Greenwich line they have lost the direct service to Charing Cross. Future benefits that are and will happen are pretty much limited to the rebuilding of London Bridge and Crossrail.

The rebuilding of London Bridge station with consequential benefits for London Bridge Metro commuters was more of a side-effect of the Thameslink Programme, rather than a specific scheme to improve services on the South Eastern side of the station. The benefit of Crossrail to South East London is limited to just two new stations south of the river Thames – and one of those had to be fought for by local MPs and the borough where it will located.

It seems that the Kent Route Study is offering most commuters in South East London more of what they have been used to – virtually no improvement at all and, in places, some regression compared to what was formerly offered 20 or so years ago.

Written by Pedantic of Purley