This year the first of our new electric trains will arrive and one of big benefits of them will be that they have faster acceleration than the clunky diesel trains we have now but the question is just how much faster they will be. For some reason it is something that Auckland Transport have been pretty reluctant to actually talk much about which I am guessing is due to them not wanting to get peoples expectations up. We however are not AT and are free to talk and speculate all we like so with that in mind, some time ago I built a model to try and work things out. I actually blogged about it back then but at the time I had only showed the western line, with this post I thought I would look at the whole network. Before I go into the results, for those that are interested, here is an explanation as to how I have worked the times out:

First I have worked out the distance between each station and for each leg of the journey I have assigned a maximum speed that the trains can travel and I have kept most of the network at 80kph with the inner sections at 60kph. When the EMUs were announced AT said that they would be able accelerate and brake at 1m/s². Based on that I then worked out how long, both in time and distance it would take to reach the top speed and slow back down again for each section of track. For those interested it takes 22 seconds and 247 metres to reach, or slow down from 80kph based on that acceleration of 1m/s². The next step was to work out how long the train would travel at top speed. To do that I subtracted the acceleration and braking distance off the distance between stations and worked long it would take. For each station I then added in a dwell time to represent how long it would spend on the platform. To be conservative i generally used 45 second however for busier stations I used 1 minute. I then added the time spent accelerating, braking, at top speed and the dwell time together. Lastly to try and be conservative I added in a multiplier of an extra 20% to account for things like slow drivers, corners and junctions which that then gave me an overall result for each station.

Here are the results verses the current times for each line, I have left out Te Mahia and Westfield as based on the draft RPTP they are likely to close, I have however added in the Parnell station. For the Western line I have added in the time allocated for the driver to change ends at Newmarket together with the travel time. You may also notice some of the times look longer than current. That is because AT obviously round the times up or down to get to an exact minute.

Travel Speeds 2

As for how much time is saved, well the western line ends up being 3-4 minutes quicker for a trip to Swanson, the Southern a massive 7 minutes quicker out to Papakura, its 3-4 minutes quicker to both Manukau and Onehunga. The big reason the western line doesn’t gain much compared to the other lines is that it has a lot of really closely spaced stops which continue to slow things down, there also aren’t any plans to drop any stations along the route like is expected to happen down south.

Edit: Here is an image I was working on to show the cumulative time from Britomart which also highlights just how close many of the inner western line stations are to each other

Travel Speeds Map

Share this


  1. I suspect AT aren’t saying the EMUs will be faster because initially they won’t be – with the possible exception of Onehunga services, EMUs will have to stick to diesel speeds while running a transitional diesel plus electric fleet in order to keep to time – the time benefits won’t be realised until all diesel trains are off the line and the timetable can be updated to reflect that.

  2. I’m confused by the tables you have shown – what do the numbers in in each line mean? Is this the time it takes to get to each station from the previous station? I think it would be clearer to show cumulative times which would also help in instances in which stations have been removed/added as you can easily compare the total time taken to get to a station from Britomart.

    1. Yes it is time from station to station, I did have cumulative time in there but it started looking bit messy. I’m working on am image to hopefully show it better.

  3. Matt L,
    From this earlier post from our former Admin at this link

    The second picture in that post which is a image of the technical specs on the EMUs says (
    About the acceleration speed: 1m/s2 acceleration,
    Also says time to 60 km/hr = 24 seconds (as opposed yo 40 seconds now)
    And it also mentions that the EMUs are “capable of a 10 minutes faster trip Papakura to Britomart than existing trains”

    So, that is in line broadly with your comments, but which is the correct one? the EMUs top out at 110km/hr but I doubt they’ll run at that speed very often, if at all on the current lines.
    And not while the existing mix of Diesel and EMUs is in operation either as those timetable have to cater for the slowest train until the EMUs are fully deployed.

    I recall doing a similar post on the trip time to Papakura and finding improvements, but being largely dependant on dwell times, which will be improved with the EMUs for how long the trip takes.
    Be blowed though if I can find that post in the archives, but never mind.

    1. My thinking is that while the EMUs will be capable of 1m/s2 acceleration and braking, they likely won’t do those sorts of speeds in normal use which is part of the reason I built in the multiplier (the 22 seconds is before the multiplier).

    2. A focus on dwell times is very much needed. Hopefully HOP will help a little, with less focus on collecting tickets. Veolia really need to do some work here. I thought the point of having a multinational operator was to bring efficiencies through using international best practice to train staff. Here they have clearly failed. Could make a 5min difference in journey time.
      A ten second google search gives a whole bunch of studies, I’m sure Veolia could have a look if I did.

      1. Official Station Dwell Times (per requirements of AT) on all intermediary stations excluding Newmarket on Western Line services (3:00 mins) is 30 seconds. A note also that if a train pulls into a marked station (Newmarket, Otahuhu, Glen Innes, New Lynn, Papakura (on the way to or from Pukekohe) – basically any station with BOLD times in the current timetables early – it can not leave the station until the indicated time regardless of Dwell Time present. So bit surprised 45 seconds and 1 minute was used in the analysis (and yes I also know the reality of the situation as well).

        But as we know – those dwell times can swing from station to station, service by service

  4. For bonus points, consider adding the CRL stations and times and a parallel line above the Western line showing ex-Britomart via CRL timings.

    1. Current track “maintenance” by KR has much of the Auckland rail network under significant speed restrictions, I believe as low as 40km/h in some places, 60km/h in most others, and certainly no higher than 80km/h. 110km/h would require KR to actually be allowed by the Minister to spend money on really upgrading and maintaining the rails, and that ain’t going to happen with the current bunch of muppets. Plus we still don’t have a fully grade-separated network, and until that happens most sections of the network have road crossings that would make 110km/h distinctly unwise.

      1. If money was made available, how difficult would it be to bring those aspects of the network back up to good spec? Are we talking of more years of long summer closures, or could it be done as part of normal maintenance, if the funding was provided?

      2. This is a *** disgrace in a world where lines can be buiilt to 300km/h+ specs, hell some switches trains can traverse at 220km/h+. An absolute disgrace and embarrasment that 40km/h areas are considered acceptable by KR. May as well go back to horse-drawn wagons.

  5. I don’t think there is much we can do about the inner west closeness of stations. At the moment Morningside is the low outlier at 291 pax. However there is really great potential for intensification here as is a light industrial area, and will only be 10mins from CBD with CBDRL, so will grow hugely.
    Baldwin Ave is 529 which is quite low, but not due to street network rail best PT option for this area, and again with CBDRL will be very handy link to city so rail will make area very popular.
    Mt Eden will dissapear with CBDRL to be replaced by much more usefully located Newton station.

    1. Agree, there’s not much sense in removing stations JUST so the trains go faster. It’s a balance act where at the end, we should probably lean towards retaining stations unless they are really beyond hope of increase in any reasonable future time.

    2. I’ve always wondered about closing Balwin Ave and Morningside and consolidating those into one station at St Lukes Rd. Given that Baldwin is less than a km from Mt Albert and Morningside less than a km from Kingsland, a station in the middle might make sense and not lose any customers (most folks will still be within a few hundred metres of one of those three stations).

      It could also give better bus connections to St Lukes Rd crosstowns, and would probably make it a lot easier to grade separate the three level crossings between Kingsland and Mt Albert if each wasn’t right next to a station any more.

      1. Don’t close Baldwin Ave. House prices are already on the rise in the area surrounding it and will rise even further once the EMUs are on stream. I can already see patronage significantly increasing post-EMU fleet launch. In regard to Morningside, its in AT’s strategic interests not to close it, as Westfield St Lukes will actually wake up within the next 4 years to the need to run a shuttle bus service between the station and the mall due to increasing numbers of visitors choosing not to come by car.

        1. Baldwin Ave will hit a plateau though at some point, as no sites for intensification here, and no feeder bus opportunities. Therefore increase will be much slower than many others after initial CBDRL boost. So after looking again at Baldwin Ave, and playing with walking distances it could easily be shifted, with the maximum extra walk being something like 800m for existing users. The corridor further West seems wide enough to put a simple station between Asquith Ave and St Lukes Road. The St Lukes overbridge will need a bit of work though as road will need to be widened to put bus stops at the top of the bridge. Other issue is will need space under the overbridge for a pedestrian underpass next to rail line on each side of the railway. Top of bridge not good place for pedestrian crossing. Other issue is with disability access from St Lukes Road which could be quite complex.

          However Morningside should stay open, there is potential for thousands of extra people to live here once is redeveloped, and should be easier than many places as no heritage buildings and reasonably large sites, with less fragmented land ownership. As is large area should be able to have more stories than other places surrounded by single storey dwellings. Having a train station in the middle would really encourage this.

        2. Rob, AT have recently announced changes where every New North Rd bus will divert via St Lukes Rd and Morningside Dr to access St Lukes mall on the way between Mt Albert Station and Morningside Station (and Kingsland also). So there is no need for a rail shuttle, the regular frequent service will provide that same function.

          But I do agree there is a lot of potential around Morningside for intensification and development.

      2. Apart from St. Luke’s, would it not also be in AT’s interests to keep Morningside open as a second station for major events at Eden Park a la for the RWC?

        If the level crossings are an issue (which they may well be) would it not be possible to trench the line? (obviously cost dependent…)

        1. With six car EMUs that can hold a crush crowd of 1000 people each will we even need to do special event services? Can’t they just run the peak schedule of six big trains an hour each way on the western line?

          1. possible for smaller games, but why run 6000 per hour capacity for 50,000 crowd, when you can run 18,000 per hour? At the RWC the existing fleet was handling 16,000 post match, and quick crowd clearance important post match for safety.

          2. My main reason is not wanting to effectively shut down and sever the entire western line and bung up Newmarket any time a big game is on. We will get to the point where it is unacceptable to shut the rail system down to serve one stadium event. The trick is to simply boost the western line in both direction.

            Anyway, if I’m not mistaken the Western line can now handle a train every three minutes with the new signalling. That’s perhaps twenty thousand an hour each way as long as you can load a train within 90 seconds or so, and provided the other lines aren’t running at high frequency so they have somewhere to go to.

        2. The station could always be retained as a special event station one but closed the rest of the year although I do agree that the area around the station is one of the better locations for intensification, especially post CRL when trains would only take about 10 minutes to get to the centre of town.

          1. Agreed. Morningside is ripe for intensification. Personally, I’m loathe to move Baldwin Ave, as living in Japan, I’ve seen so many similarly positioned stations become very well patronised over time – even after the surrounding area has reached capacity as far as building intensification is concerned. Its because those less-intensified areas become havens for those who have money and are looking for housing closer to the city but on comparatively larger plots of land. Such areas are also popular with renters who want a house (can afford a bit more space), not a unit or an apartment.

          2. I tend to think that there are more pressing things to do with any available funds than to combine those two stations, let’s spread the station upgrades and get those EMUs functioning well and focus on rebuilding pax while building towards the CRL first….

            The CRL is the real way to improve travel times on the western…. Also lets invest in expansions let the Mt Roskill branch too…

        3. One advantage of building the Western interchange would be as a station for Eden Park. If it has four platforms (two terminal and two through running) you could use the two terminating lines for special services from whichever line doesn’t connect directly to the Western line. After all it will only be a five minute walk from that station along New North Road

  6. Do these times account for changes in grade?
    The 1m/s/s acceleration will be for flat ground.
    To calculate, train weight and track elevation (long section) are required.

    1. Hi Anthony – I don’t think they are “required” because this is just an estimate – and any extra delays in one direction will be made up in the other, going downhill. Of course one could go into more detail, but is that really a good use of time for a blog post on a subject that already has a number of other factors in it that have to be estimated? So unless the difference was massive (which I would not expect: after all, the trains are designed to climb CRL gradients, which are going to be massively higher than main line), I think this type of calculation is just fine…

  7. I hope they don’t close Westfield. It performs poorly probably due to its location. Perhaps they could look at a new station in The Westfield business park instead on the Eastern line with a Northern connection to the southern line. This area has a large workforce population and a rail line running right through the middle of it.

  8. I just want the EMUs to hurry up!
    I can’t wait to start catching it to work seeing as I am on the Onehunga line.
    Great seeing all this stuff in the article too by the way.

  9. Is there the possibility, once the patronage have built up thanks to bus feeding etc, to run express services that “jump” minor stations overtaking normal services while they’re waiting at the stations? Something like Britomart-Newmarket-Greenlane-Penrose-Onehunga. Overtaking at Penrose and Newmarket for example. With buses feeding to these stations there could be enough patronage to justify it.
    I have no clue if is feasible, I just remember my experience with European trains.

      1. It’s technically feasible as the lines are bi-directional. That doesn’t necessarily mean it’s practical of course; it depends on signalling and time-tabling.

    1. To do that third mainlines are required. AT has said they are not interested in establishing a third main between Westfield and Wiri, which would also be used by freight. I think there is now a third main between Westfield and Middlemore, funded fully by Kiwirail. It’s not currently electrified howver so would only be any use to something such as a Hamilton to Auckland service. Long term there is also talk of establishing a third mai along the eastern line for POA services.

        1. True, but that’s much less practical as it would require all trains to run pretty much perfectly to timetable so they are stopping/passing at the station at the same time. As jonno1 mentions it is also possible to overtake on the other main, but a) that would require frequent crosses, b) it wouldn’t work at high frequencies and c) it would require an even greater amount of running perfectly to time.

    2. Gian, you could run fast services now, with say, an Eastern fast service leaving Britomart after an all stops Southern service, and the fast getting to Westfield first. If that fast is timed to just catch up with a Manukau, it could again go fast south of Wiri once the Manukau has got out of the way. But opportunities to make that work are quite restricted, but if you had the units available and a desire to put a fast service between Pukekohe/Papakura and Britomart it could be done. However, once you increase the intensity of the all stops services then the benefits fall away and additional ‘fast’ lines are required.

  10. 2-3 minutes average travel time between stations on the Western Line – standard for urban Japan…welcome to the real world Auckland. Short distances like that dont affect dwell times and overall travel time between terminuses – line management in JP proves that. Such short distances between stations encorages patronage and station proximity retail. The sheer convenience of hopping on a train to go 1 or 2 stations to do shopping or 2-3 stations to go eat out somewhere should not be underestimated. Now if only could a supermarket / retail development could be properly integrated into the northern end of Newmarket Station…

  11. I’m surprised that Western line improvements aren’t higher – as the gains are made in acceleration and deceleration – of which there is a lot with lots of stations. CRL will really make the difference there.

    The perception (which is what matters to the public) will be that it is quicker and better. Look at London Overground as a comparison, with new, nippy EMUs. And those replaced EMUs, not DMUs!

    Papakura will benefit a lot, which is good considering the high patronage.

    A third main with some faster services, at least in peak, is definitely worthwhile. It should be wired/gantried and futureproofed today while there are works and closures going on. If signalling is right, it could weave in and out very easily. Happens in Japan…

      1. Makes sense. I look forward to some provisional timetables! Hope they don’t keep the savings as padding…

        Auckland dwell times are quite pitiful generally and 30 seconds should easily be enough with main directional flows. In some cases, much less should be doable.

  12. With a third main Westfield to Papakura (which is long term Kiwirail plan) could use this to run peak time expresses to Papakura and Pukekohe.
    Could run along the Eastern line, stopping at say Glen Innes and Panmure (to keep behind train in front as well) and then from there express to Papakura.
    Probably could get sub 30 minute journey times to Papakura. Would add extra peak capacity, as well as attract extra passengers who have extra pull factors keeping them in their car like free parking, or work a bit far from station (ie connecting bus required). Of course this shouldnt be key consideration of our network, but could be useful if built for freight purposes. Could use a few slots for these expresses, and one for a Hamilton service as well.

    1. Luke, this would certainly be true if you built a third and a fourth line, but by building only a third line for the purposes of getting freight off the other two tracks, the additional line would become in effect a single track freight railway with maybe a 20 minute transit time. This would probably end up with just enough capacity for the existing freight services, never mind putting on fast passenger services in additon..

  13. Matt, could you re-do your calculations but with faster line speeds as a comparison? Ley’s say inner at 80kph and outer at 100kph? I suspect that you would be looking at maybe a 20 to 30 second transit time saving per station, but for the long runs such as Britomart – Orakei and Meadowbank – Glen Innes well over a minute saving.
    And what about an emu doing 100kph between Papakura and Pukekohe compared with now?

Leave a Reply