How fast should our trains be

In just under a month’s time, on 12 March, Auckland Transport will introduce a new rail timetable. There are a couple of significant changes that we know will come with this timetable:

• Westfield Station will close
• Parnell Station will open, served initially by Southern Line trains with Western Line trains only stopping in evenings and weekends.
• Onehunga line trains will skip Greenlane and Remuera stations. This is said to be to speed up services enough that it frees up a 3-car train.
• Speeds on the other lines will also be improved, enough that it frees up an additional two 3-car trains.
• In total the changes free up three 3-car trains, which will be used to boost capacity on some peak services.

From what we hear, what won’t be included in this timetable change is any improvement to frequencies, which is desperately needed off-peak and weekends, especially in the South where the new bus network has already been implemented.

The issue of train speed is something we’ve talked about a lot recently, such as Stu’s excellent post on the economic benefits of speeding up our trains, and a guest post on how our trains are slower than even steam trains used to be. I want to focus on it again in this post.

If you recall, when we moved to electric trains Auckland Transport actually slowed down the timetable. They  padded it out to make the punctuality results look better, especially as the old diesels started to break down with increasing frequency. Following the initial services that were converted to electrics, there also turned out to be a number of issues with the signalling system, other infrastructure and operations (such as dwell times), which prevented trains from being as fast as they should.

In July 2015, Auckland Transport released a report outlining all of the work they intended to do to improve train reliability and speed. From what I understand, not all of it was eventually implemented, for a variety of reasons. However, much of the work they did do has contributed to the record highs we currently experience for punctuality and reliability. In the 12 months to the end of December, 98.6% of all trains arrived at their final destination and 96.5% of those did so within five minutes of the schedule. That’s up from lows in mid-2015 of round 96% and 83% respectively, and some individual months were considerably lower again.

So what can we expect with this new timetable? In their board reports, AT have said

Train Run times on the Southern and Eastern Lines with electric trains will be shorter than previous pre-electronic train control system (ETCS) signalling with diesel trains, with equivalent times on the Western Line due to large number of rail level and pedestrian crossings and speed limits at these crossings.

We’ll have to wait to see just how big the speed improvements will be, but we’re hearing they aren’t significant – perhaps a 3-4 minute improvement on the current timetable at best. If correct, then in our view that’s simply not good enough. And if that is the case, AT’s lack of action mean serious questions need to be asked by those charged with overseeing and managing the organisation, especially seeing as the train network has been all-electric for over 18 months now. Also remember that speeding up services was something the council recently outlined as a key expectation of Auckland Transport.

• ensuring full value is obtained from council’s very large investment in rail electrification by reducing journey times, particularly through shorter dwell times at stations and more efficient rail operations

But just how fast should our trains be? We’ve made plenty of our own estimates in the past, so with this post I’m going to look at some official estimates of what should be achievable.

The data for this comes from a document I was given by the EMU project team in early 2013. The document includes a lot of detailed technical information from train manufacturer CAF, such as acceleration and braking profiles. Importantly for this post, it also includes modelling just how fast the trains should be. In fact, it goes further, showing how CAF’s proposal compares to the original requirement from Kiwirail (the government initially put Kiwirail in charge of the EMU tender but it and the team were later transferred to AT).

The modelling shows exactly how fast both Kiwirail and CAF say the trains would be between stations, including dwell times. For the Western and Southern lines, the modelling includes trains stopping at Parnell, while for the Southern and Eastern lines it includes stopping at Westfield. Onehunga times weren’t included in the part of the document they provided me. For the purposes of this post, I’m only going to show the times listed as being Kiwirail’s Requirement, but in all cases, CAF estimated their trains would be up to about 30 seconds faster over the entire journey. The modelling is also done for each direction of each route, but for the purposes of this post I’ll just list the times to Britomart (the differences in direction are minor).

First up, I’ve replicated the tables from the documents. One of the things that immediately stands out is that with the exception of Newmarket, all stations are expected to only have a 30 second dwell time. Even off peak at a quiet station, AT can’t achieve that kind of time. Newmarket is expected to have a 1-minute dwell for Southern line trains and a 2-minute dwell for Western line trains.

That’s fairly detailed, so next I cut the results down to compare them to our current timetables, remembering that current timetables are rounded to whole minutes. It’s also worth noting that on the current timetables, the Southern and Eastern lines have a two minute variation in travel times. In both cases I used the lower of the two.

Now I appreciate there’s always going to be a difference between some modelled results and what can be achieved in real life, but the difference between the current timetable and what was required of tenderers is, as Trump says, yuuuge.

In fact, the savings are so big they’re CRL sized – in the info published by AT about the CRL, they list New Lynn to Britomart as improving from 35 minutes to 27 minutes. Under the required times from Kiwirail above, which don’t include the CRL and include stopping at Parnell as well as Westfield, the trip is supposed to take ….. 27 minutes. Or course, that doesn’t negate the time saving benefits of the CRL. These time savings would then stack on top, so New Lynn to Britomart should be as low as 19 minutes post-CRL. Think about what impact that would have on how people choose to get around.

A really important advantage of getting speeds close to these required levels is that it would free up more trains, which could then be used to further increase capacity on the network. Or course, with times so much faster, that capacity will certainly be needed for all the extra people that will want to use the services.

It’s perhaps worth pointing out that we don’t know what happened along the way of buying the trains, and it’s possible that for budget reasons that AT had to make some compromises in design – but I’d certainly hope not.

I imagine there are still a number of improvements that could be made to improve travel times – such as eliminating level crossings on the Western Line – but in my mind, the biggest opportunity is improving our horrifically slow dwell times. That will likely require a mix of both physical and operational changes.

Coming back to the question of how fast our trains should be, I think that if we could get within 2-3 minutes of what was originally expected (and remember CAF said their trains would exceed those times), then I suspect we’ll be doing well.

Share this