Monday 16 November 2015

Crossrail 2 Journey time modelling

TfL have not made public their modelling for Crossrail 2. As such, I've done some journey time modelling myself.

Journey time modelling

Journey time modelling is a hard problem in general, as there are many possible routes to consider on a network like that in London. However, if the problem is simplified to only include a subset of routes, it is possible to obtain an answer.

The simplifications used are:

  • Only journeys from SWML destinations affected by Crossrail 2 are considered
  • Only journeys to central London destinations are considered
  • Interchange times are estimated
  • A maximum of three trains is allowed to get from start point to destination
  • Buses are ignored
  • Trams are ignored
  • Walking is ignored
  • Time to enter/exit stations is ignored

Even with these simplifications, it is possible to get some reasonable figures out, and take a look at two key questions:

  • Will Northern Line passengers change onto Crossrail 2?
  • Will Crossrail 2 passengers change onto the Northern Line?

The results are here:

Each page contains a set of possible options for each of a long list of possible journeys. A key of station codes is given at the bottom of each page.

The first finding matches that of my previous article. Journey times from Crossrail 2 suburbs to London Bridge and Bank are quickest by changing onto the Northern Line.

The second finding was unexpected however. The figures for Balham indicate that many passengers travelling from Morden to Victoria and Euston will not change to Crossrail 2! This is because the time saved by the new route is not enough of a saving over the existing route (changing to the Victoria line at Stockwell).

Finally, it is important to remember that the time to enter and exit the station is not included. This can have a big impact together with the walking distance to an individual's actual start and end point.

Summary

This quick post provides links to journey time analysis for Crossrail 2 in South West London. The analysis is imperfect, but enough to raise further concerns about TfL's Balham option.

6 comments:

  1. Hi Stephen,

    thanks for doing this - it makes it so much easier to visualise the impacts. One very minor point - how did you arrive at the 24 minutes 6tph for Surbiton fast trains? Would Surbiton's semi-fasts really replace fasts (18 minutes) - hopefully not, as there'd be a bit of riot if they did!

    ReplyDelete
    Replies
    1. Network Rail are clear that they want to remove those trains that join the fast lines at Surbiton. Implicit in that is that no fast line trains would stop at Surbiton (since the goal is to maximise the paths from Woking to London). Given that Surbiton would only be served by the slow lines, journey times will necessarily lengthen.

      I'm sure NR will do their best to keep trains fast, but if TfL gets Balham for CR2, then the slow lines will be responsible for Earlsfield. The knock on effect will be even more of a slowdown for those Surbiton trains on the slow lines (each faster slow line train will follow a stopping one through Earlsfield). This is why Surbiton residents should support the Swirl plan!

      So, in my opinion, the 18 minute Surbiton expresses are dead under any plan. Surbiton trains will almost certainly all stop at Clapham Junction, and probably Vauxhall. Stopping at Wimbledon also makes sense IMO.

      See also the full source code for details of where I modelled the trains as stopping, and the intermediate journey times:
      https://github.com/jodastephen/railmodel/blob/master/src/main/java/org/joda/railmodel/Crossrail2BalhamSWLondonModel.java#L123

      Delete
  2. Stephen,

    That's a bit of a shocker...Crossrail 2 bringing slower journeys to Surbiton, the busiest SWT-only station on their network. That would not be popular to say the least...people pay a large premium to live beside the fast trains to the City / Waterloo.

    I did pick up on the possibility of the Wessex Route study bringing service changes from Surbiton, but thought (perhaps naively!) that the 2 extra trains per hour would stop at Surbiton. On the the not insignificant basis that
    1. Surbiton was the busiest SWT-only station on the mainline
    2. Surbiton passenger numbers had grown significantly.
    3. SWT / Netwrok Rail had agreed on the swift need to expand passenger capacity at the station on safety grounds.

    If fasts are removed, Hampton Crt/Thames Ditton/Berrylands to City journey times would be adversely affected too - decent numbers commute up/back to Surbiton to catch fasts.

    I guess I need to follow up with Network Rail about the exact implications of the following passage:

    --

    http://www.networkrail.co.uk/long-term-planning-process/wessex-route-study/

    " CP6 Priorities 0.6.16

    Greater use of the Slow Lines from Surbiton inwards. If some of the trains which currently switch from the Slow Line to the Fast Line at Surbiton were instead to remain on the Slow Line, then their paths could be used by longer distance services. Analysis suggests that this could potentially release two paths in the high-peak hour.

    However, the disbenefits would be
    a) longer journey times to London Waterloo for those whose trains remain on the Slow Line, and
    b) a performance risk with two additional trains per hour using the Slow Line."



    ReplyDelete
  3. Hi Stephen,

    it looks like you could be barking up the right tree. At today's Crossrail 2 consultation in Surbiton, the local Lib Dem councillor (and opposition council leader) was told by Crossrail 2 that they will refuse to guarantee that fast Surbiton trains would continue if Crossrail 2 was implemented.

    ReplyDelete
  4. how did you calculate the time? was it total of all train users? what sample size? i am confused on how you got that big of a number for the time.

    ReplyDelete
    Replies
    1. The text says "The 'total effective time' is the sum of all effective times modelled. It is a reasonable proxy for the total enhancement provided by the scheme." ie. it is just a sum of all the individual journey times calculated below.

      Delete