This is a guest post by reader Malcolm McCaskill based in Hamilton, Victoria.

©Patrick Reynolds
©Patrick Reynolds

When visiting Auckland earlier this year, I was surprised how slow the trains were compared with those here in Australia.  To see what travel times should be achievable when the dwell time issue is resolved, I assembled timetable and distance data from Auckland, Wellington, Melbourne and Perth, using publicly available timetables and distances between stations from Google Earth.  I then used linear regression to develop a relationship for each line  between travel time for an in-bound morning peak service and distance between stations.  The best relationship was for Wellington’s Kapiti line, which explained 97% of variation, and had an intercept of 1.0 (minutes per station) and slope coefficient of 0.85 (minutes per km). This implies that for each station the combination of deceleration, dwell time and acceleration adds one minute to travel time, and that without stations the average speed would be 71 km/hr. This regression approach was much less successful for other lines, because there was less variation in station spacing. For example on the Eastern Line it only explained 11% of variation. So I used the Kapiti relationship as a benchmark to compare the other lines, calculating travel times using these coefficients from the distance between stations.  For all lines except Kapiti, timetabled travel times were unusually long when approaching the terminal station, so I did the calculations for the station before the terminus, because it is not a dwell time issue, and instead one of platform availability. How did Auckland perform ?

Malcolm McCaskill train times

Auckland’s Eastern Line is currently timetabled to be 4.2 minutes slower (13%)  than the Kapiti benchmark, while the Western Line was 8.8 minutes slower (20%).  Melbourne’s Pakenham line was marginally faster (2%) than the benchmark, while Perth’s Fremantle line was 21% faster.

So we would expect that if the dwell time project looks to Wellington for optimal practices, Auckland should achieve travel time improvements of 13-20%. However if it looks to Perth there should be further travel time improvements in the order of 21%. Achieving times similar to Perth should be quite feasible, because like Auckland, its trains are powered by a 25 kV AC electrical system, which enables both regenerative braking and much faster acceleration than the 1.5 kV DC system used in Wellington and Melbourne.

Faster travel times are not simply an issue of passengers travelling faster. It is the equivalent of increasing the fleet size without having to purchase more trains, while also increasing the productivity of train crews, leading to a reduction in subsidies.

This benchmark approach is also useful to estimate travel times post-CRL travel times. From New Lynn to Britomart travel times would decline from the current 34 minutes to 19, or a saving of 15 minutes, while the travel time to Aotea would be 17 minutes. This is equivalent to halving its distance from the CBD.


[editors note] when the new Western line timetable goes live in May we may see some improvements in travel times while AT have say the next timetable change (likely next year) will see them incorporate many of track, signalling and operational improvements being implemented.

Share this


  1. It would be interesting to see if this could be applied from Pukehoke to Newmarket as well.
    Can the transfer between trains which can add up to an extra 15 minutes at Papakura be interrupted as a long dwell time?

  2. I imagine you could explain it all with three factors:
    -Number of stops per kilometre
    -‘Wonkiness’ of alignment (curves, grades, switches etc)
    -Dwell times

    1. And:
      4. Number of level crossings and the interactions these have with/impacts on ETCS.

      This fact might be why the Eastern Line is the “best” worst performer of them all, given it has the fewest number of level crossings, with no level crossings north of Westfield.

      And another cause may be that by stopping 1 station short of the terminus, you don’t see the long trip time from Orakei to Britomart for what is essentially a fairly straight flat run designed for top speed running.

      When the EMU’s should be flat out over running across Hobson Bay, they are instead coasting in “neutral” for half that crossing due to the existence of a “zone border” in the overhead electrification.

      Even if the driver gives the EMUs some welly before the cut-off [assuming ETCS allows them to that is] – it still takes ages to cross the “neutral zone” – which is not helping with improved journey times one bit.

      Also note those Wellington and Australian lines mentioned don’t have ETCS.

      And it seems our version of it is [as usual] a half-arsed, very conservative implementation with too few track side communication devices. So long stretches of track are tarred with the same “wonkiness” by ETCS – even when they are not.

      And good ol’ KR doesn’t implement the same level of ETCS as the EMU’s so the EMUs are forced to work with their State of the art ETCS “downgraded” to the dumbest ETCS mode possible to work safely with KR’s loco locos.

      Its telling that the time from Panmure station to Britomart is really no quicker timewise now than when the Eastern line first opened – 85 years ago – when they were running steam passenger trains along it.

      And we have 1 less station on it now [Tamaki] compared to then. So you’d think it would be at least 1 minute quicker right – if each station adds 1 minute to the journey time.

      1. “When the EMU’s should be flat out over running across Hobson Bay, they are instead coasting in “neutral” for half that crossing due to the existence of a “zone border” in the overhead electrification.

        Even if the driver gives the EMUs some welly before the cut-off [assuming ETCS allows them to that is] – it still takes ages to cross the “neutral zone” – which is not helping with improved journey times one bit.”

        The neutral section is at the Westen end of the causeway, but the slow speed through it is more a reflection of the 70(?) speed restriction around the curve rather than the neutral section.

      2. Ahh, that would explain why the air con turns off briefly each time the train passes through there. So are you saying the trains are not powered for that bit? Seems like they have more momentum than I would have imagined, I’ll have to pay closer attention next time!

        1. The aircon turns off early, and comes back on last, so you can’t judge their lack of noise/running for the actual length of the neutral section.

          The actual “unpowered” bit is about as long as a 6 car EMU I’ve read (because you don’t want the EMUs straddling two *powered* track sections at once because supposedly all sorts of electrical nasties could occur) so it has to be a bit longer than your longest EMU.

          Having said that, if the trains get some welly on when leaving Orakei heading to Britomart, the coasting section can be fairly quick, but it is noticeable when the driver turns off the acceleration and coasts.
          They do need to slow for the curve near Tamaki Drive though so quite often the train [driver] will coast for quite some time when heading to BM even once over the unpowered section.

          When leaving BM, sometimes it seems the ETCS won’t allow the driver to pick up a lot of speed as it runs beside Tamaki Drive, prior to the cut-off so the train will move a lot more slowly over the neutral section then and pretty much ooze across the bay until halfway, pick up some speed, before it then slows down into Orakei.

          From Orakei to Meadowbank is similarly slowish being only 0.8 of a km distance – but from Meadowbank to GI its full on, the track allows is good and so the run into GI is normally pretty fast.
          Then the next stop after GI is Panmure.

          Which is therefore a real surprise that the 110km/hr rated Eastern line line/trains can’t go from Panmure to Britomart (in either direction) faster than the 17 or so minutes they used to take in steam days.

          1. The actual unpowered bit is only about 2m long. The notional neutral section – space enough to give safe operating conditions at all speeds and configurations – starts about 200m before the actual dead spot and ends wherever the train is when the resetting process is complete, punctuated by the closing of the VCB, which is the “bang” you can hear from the vicinity of the pantograph. The faster you go and the longer the train, the further down the track that happens. The Eastern line speed is 100km/h, not 110. Purewa-Meadowbank is the only sustained 100km/h stretch. Most of the curves are 85km/h and two of the turnouts have been restricted to 60km/h for a while. That’s not just ETCS, it applies to all trains and the speeds for non-ETCS trains are often tighter. The only way ETCS is more restrictive in this case is in how far before the curve/turnout you have to be under the speed limit. The rumour mill says that could be cut soon. The curve by Parnell Baths is 60km/h, followed by a 50 and a 40km/h section. In hot weather 40km/h all the way. Approaching the neutral section from Orakei, you get to the 100km/h maximum, coast until the neutral section warning sign without losing speed and then apply brakes for the curve and/or heat restriction. The neutral section has absolutely no effect on transit time in that direction. The braking pattern would be the same regardless. In the other direction you can’t accelerate until clear of the 60km/h curve. A 3-car train will get to about 85km/h and a 6-car will get to about 80km/h before coasting for about 600m. Once back under power you can get to about 95 before having to brake for Orakei. Without the neutral section, those accelerations would be joined up and you’d get maybe 700m on the causeway at 100km/h instead of at around 80km/h, but have to brake slightly earlier – a saving of just 5 seconds. If a driver is taking a more conservative approach than that, it is usually for good reason – weather or track conditions, signals indicating a speed or stop condition earlier than usual, running ahead of schedule, minimising the inevitable wait at 125 signal etc.

          2. Greg/James – thanks for the info, very interesting. I thought the issue of speed restrictions in hot weather had been dealt with a number years ago, is that not the case?

          3. Every year the track has to be assessed and adjusted as the temperatures rise. Heat restrictions aren’t permanent anymore, but they do still occur and usually in different places each year.

    2. also, have temporary speed restrictions for track faults been taken into account in the comparison between the various networks.

      they can have a huge impact on running times.

  3. ‘Faster travel times are not simply an issue of passengers travelling faster. It is the equivalent of increasing the fleet size without having to purchase more trains, while also increasing the productivity of train crews, leading to a reduction in subsidies.’

    This is a hugely important point as it is clear the AKL network is already short of units because of the rapid increase in ridership.

    ‘This benchmark approach is also useful to estimate travel times post-CRL travel times. From New Lynn to Britomart travel times would decline from the current 34 minutes to 19, or a saving of 15 minutes, while the travel time to Aotea would be 17 minutes. This is equivalent to halving its distance from the CBD.’

    Given the multi billions effortlessly spent on almost any road project justified on the basis of ‘travel time savings’ even if only some of these minutes are recoverable in practice they are apparently extremely valuable in terms of NZTA’s Economic Evaluation Manual. Daily there are now 60k journey on the network, and rising, so every second multiplied adds up.

  4. This is exactly the issue I had been complaining.

    This is huge opportunity for improvement to be made from AT.

    If the rail commute time is halfed, it would double the frequency as well.

    So in theory without adding more train and drivers, western line peak frequency will be 7.5 minutes instead of 15; and 5 min peak frequency for eastern and southern line.

    In order to achieve this AT can look at:
    -Improve signal efficiency around britomart tunnel.
    -Review safety speed limit on corners, bridges and tunnels
    -Increase acceleration and deceleration around station
    -shorten unnecessary station dwell time and procedures.

    1. Halving many be extremely difficult in practice, but there surely can be little doubt that major time gains can’t be found, especially with dwell times. It does look like there is a lot of redundancy built into timetables currently, so the dreamy dwell time may be a result of keeping to unambitious timetabling.

      So inefficient.

      1. If the train frequency is short enough for a turn-up and go, I doubt we need a timetable at all.

        For example, they could replace the time table for peak time to ‘Every x minutes’.

        In that case there is no need for the train to slow down to match the unambitious timetable.

        1. Whatever the frequency there will still be a need for a timetable for operational reasons, to ensure things like trains don’t hold others up at junctions, that crews and trains are at the same place at the same time at the start of shifts, etc.

          So trains will still have to run to a timetable, however unambitious (or otherwise) it is.

          1. AT could still have:
            -Ambitious internal timetable for optimal operation estimate
            -un-ambitious public timetable just saying ‘up to every x minutes’.

            In that way if the train is faster or slower than the internal timetable, it would still well within the punctuality requirement of the public timetable.

          2. of course there would be an operating schedule, and has to be. But once frequencies get high enough and are consistent through the span of operations the users soon stop bothering to check. But I would say that that point is some way off, and certainly not till post CRL; need to get down to sub 5 minutes or so…?

      2. I have noticed (Eastern line anyway) when the train is running behind schedule a bit, the dwell time seems to pick up quite a bit.

  5. What we need is a campaign like Stalin used to do, a “War on Delays!”
    We need signs, officers etc. trying to get people on faster, we need better engineering, we need everything working together because for a few dollars the savings are immense

    We need to roll it out onto our streets. Drivers who don’t get going on a green for 3 seconds will be shamed publicly. Suddenly we’ll see that our “inadequate” roads are actually more than enough for our cars once the drivers start driving efficiently. Think of it as 6 Sigma or Kaizen for the road network.

    Once we have drivers, bus passengers, train passengers actually HURRYING (and I don’t mean speeding) we’ll see just how good our network is, and whether we need more.

    First rule of asset replacement… as any NAMS person will tell you: don’t replace until you are using all your existing capacity.

    1. Well done, that was your most bizarre post ever.
      If it’s all the same to you, I won’t take any suggestions from Stalin. He had some shortcomings.

      1. I preferred the one about the consequences of detonating a 100 kiloton nuclear device over the Auckland CBD, being the rationale to enforce compulsory population dispersal to the regions.

      2. Stalin turned a backwards agricultural wasteland into the world’s 2nd most powerful country in 30 years. Pretty good.

        And do you think our network is already at maximum capacity, we couldn’t eke out 5-10% more without expensive asset building?

        My gut feel is that every light change is at least 2-cars-inefficient. Multiply that across Auckland

        1. Umm 2.9million dead due to executions and Gulag deaths, at least 10 million due to the famines he caused and no one actually knows but the guess is 15million dead in the conquest states. So around 20 million to 30 million dead. Maybe you need to find a different role model.

          1. I don’t always agree with mfwic, but I fully agree with him that famine and mass murder are bad. Also, Stalin’s Russia was hardly the epitome of efficient infrastructure development.

            But that being said, this is possibly Early Commuter’s best troll comment yet. I am in awe.

          2. Hi Peter,
            Can you name any other country that moved from an agricultural, peasant-based economy into one that put the first ever man in space (or an equivalent achievement) in less than 40 years?

            In 1920, Russia was the laughing stock of the technological world. By 1940, they had the most advanced tank in the world (T-34). By 1960, they were ahead of the Americans in the space race. That’s impressive. And yes, a few eggs were cracked. Omelettes don’t come any other way.

            One might compare them to say the United Kingdom; the most technologically advanced nation in the world in say 1880, and yet by 1960, a laughing stock

          3. Sorry Patrick one last comment. If anyone hasn’t read ‘A day in the life of Ivan Denisovich’ I strongly recommend it. It covers the cruelty of the Gulag but the plot is surprisingly uplifting. It is not a long book and a week on the train or bus will probably be more than you need.

          4. Dear Patrick

            My central point is this:
            – We want a transport system that moves people more quickly, efficiently, and safely
            – There are ASSET and NON-ASSET solutions
            – busways, cycleways, and motorways are all expensive ASSET solutions
            – there are cheaper NON ASSET solutions (e.g. congestion charging, but also improving driver efficiency through behavioural education)

            Here’s an analogy you might understand
            – You want to put some jam on your toast, but
            – There’s hardly any left in the jar
            – You can
            — Buy a new jar, or
            — Scrape the existing jar clean

            What I am saying is that we need more NON-ASSET solutions to our transport problems. Not just “build this” vs. “build that”.

            Would you rather we (a) built more diabetes clinics (asset) or (b) educated people to eat less sugar (non-asset)?
            Would you rather we (a) built more motorways or (b) improved the throughput of our existing roads?

          5. EC I think Mussolini was the tyrant who got the trains running. Perhaps you should be promoting him as your example.

          6. What about Pol Pot? We haven’t had him in here yet. He certainly tidied up Phnom Penh, go rid of all those pesky layabouts that weren’t contributing to the throughput of their roads.

            Seriously EC, your childish trolling by extolling the virtues of dictators is pathetic and offensive. Please grow up.
            I used to enjoy the comments section of TB but now find I’m reading it less and less as your total dominance of every. single. conversation. is getting really boring.

  6. The EMU’S are far quicker than the diesels they replaced but for all that acceleration and greater allowable top end speed (110 km/hr max vs 90) they lost it by the fitting the safety override ETCS that hugely restricts the trains approaches to curves, end of line (Britomart) and red lights and the door system, namely plug doors rather than sliding and to eliminate train managers eventually, the system controlling them.
    The Western line is very restrictive on speed anyway, curves and level crossings that dictate red lights and low line speed in many places that is not an easy fix and Onehunga has a similar slow track speed albeit it is straighter but plagued by red lights. And hence on the Onehunga line the EMU’s are a lot slower per trip than the DMU’s they replaced.

    Then there are the many junctions that cause delays.

    What are these track, signalling and operational improvements?

    1. What level of ETCS did they fit and why is it slowing thing down?
      I would have naively assumed that line speeds are primarily limited by curves. The purpose of signalling is to prevent trains from colliding with other trains, not to slow them down when in fact the line is clear.
      Similarly with level crossings: there is no reason why a properly managed level crossing should in any way slow trains.

    2. Raised Onehunga and Newmarket line speeds and a fresh tranche of ETCS updates have recently been implemented as per the existing programme of changes. The network upgrade didn’t finish with the last EMU hitting the rails. There is still at least a year of incremental tweaks to come.

      Level crossings affect speeds when their proximity to a red signal demands a conservative approach to prevent minor errors or faults resulting in death. In the past the risk was simply accepted, now it’s not. When approaching a red signal, the ETCS doesn’t know what’s happening in the next section, so applies a restrictive approach speed that leaves it or the driver enough time to act before the next high risk location (level crossings etc). If a level or pedestrian crossing is immediately after the signal, as at many platforms, the allowed speed is very low, up until the train reaches the next signal, where ETCS gets its next movement authority. Now that the first stage implementation is complete and proven, more ETCS hardware can be installed in the “gaps” to reduce the effect of these restrictive conditions.

      1. And level crossings can’t just change state in an instant, there is a minimum time required to raise or lower the barrier arms and a minimum time is required to allow vehicles and people to clear the crossing. Like a pedetrian crossing at the lights, you can’t just flick from green to red and have the people halfway across dissappear.

  7. So its just western line timetable being improved in May?, that is disappointing, I was thinking they were going to improve all the timetables, especially with the new network coming to south in October trains need to run later at night to compensate for the removal of distance services e.g. 471/472 and others which ran to Midnight. Without any improvement to the Southern timetables operating time, the new network will make PT unusable after 10pm for most of South Auckland (at least on Sun-Thurs). Also many of the new network buses which run beyond 10 will be running for no reason as there will be no trains to feed to them.

  8. So we are slower than Australian operations, huh? Send this to AT management. Safety is paramount but the station dwell times are too slow. The doors may be a bit slow but it’s the train managers to me. Have a look outside, get going again.Do the Australian networks use TM’s?

    1. Brisbane definitely has train managers, I’m not completely for Sydney/Melbourne but I don’t recall seeing one when I’ve used the systems there. No idea about Adelaide or Perth.

    2. Perth and Melbourne don’t, Sydney and Brisbane do but they are in one of the unused cabs so don’t add a second door closing phase to slow things down.

  9. doors should open and close automatically at each stop.
    the current system has people waiting until well after the train has stopped before pressing the open door button.
    doors should start opening just prior to when the train is stopped

    1. I still see people who are new to the trains freaking out trying to get on-board the train and end up getting left behind, next train 30 minutes… Surely they can setup sensors that detect when the train passes and trigger a message on the PA saying your train to “Britomart” (or other destination) is arriving on X platform press the green button to open the doors. Much like the automated – stay behind the yellow line during the movement of trains warning.

      I don’t necessarily agree with all the doors opening all the time, maybe at major stations but that is about it. The cold wind coming in unnecessarily is annoying.

    2. I find that wait for the green light to come on most frustrating and annoying. On the Paris Metro that handle on the older cars is flicked up and people are stepping off as the train is coming to a halt.
      Ok, so maybe we don’t need to go that far, but……

    3. The doors can’t open until the driver releases them, and the driver can’t release them until the train has come to a complete halt. That’s how the trains were designed. Because of this, I don’t think there would be any major time reductions to be had if the drivers opened all the doors at every stop. The doors on the middle cars, for instance, would still take an age to open due to the ramps first needing to extend and retract.

      1. Why can’t the button remember that it’s been pushed (like a “BUS STOPPING” button except “DOOR OPENING” and then when the driver releases the doors, the ones that have been pushed, open? That way there’s no delay 🙂

        1. Who knows why they don’t do that by default. Those things were made to AT’s specs, so perhaps they cheaped out. Hopefully that functionality can be added at some point.

  10. Interesting methodology, Malcolm! Following Nick R, it would be interesting to see what would happen if you (or someone…) included a measure of “curviness” into the analysis to control for the impact of alignments on speed.

    1. Also, on a technical note, it would be useful to add in additional explanatory variables (e.g. presence of ETCS, “curviness” of lines, number of level crossings) and then “pool” the analysis across all seven rail lines. You’d gain a better sense of the underlying drivers of rail line speed by doing that, and an analysis of the model residuals would still give you an understanding of the “unexplained” over- or under-performance of individual lines.

      1. This is a “high level analysis” that can be done from publicly available timetables and station spacing distances. Most lines I chose had plenty of curves to reduce this as a source of difference. In Perth the Fremantle line probably has more curves than any other of their lines. A more detailed analysis would need to account for curve constraints, braking and acceleration, but relies on data not generally in the public domain, and are more difficult to use. For lobbying and benchmarking, this level of analysis should be sufficient.

        I attempted a more detailed analysis allowing different slopes and intercepts for each line. This was not useful, because lines with a relatively even station spacing tended to have a high intercept, low slope and low R2. Analysis of residuals indicated the problem of terminal stations. Had I used travel time data from afternoon peak timetables, I could probably have avoided the problem. The timetable data are only to the nearest minute, which limits the precision of such relationships.

  11. In the light of the new health and safety legislation, there is no hope of the doors being able to be opened while the train is still moved even if it is only the teeniest bit. Saw the new rules in operation today at a job I was on. I very self important person came around dressed up to the nines in every piece of safety gear that his employer could lay their hands on, and proceeded to try and shut down the work site based on some theoretical accident involving someone who should not have been there in the first place. Got the foreman really riled up and provided us with some good entertainment.

    1. The new legislation doesn’t say to overreact to ridiculously unlikely health and safety risks. In fact it’s not really any different to the old legislation in that regard. It’s a rather sensible piece of legislation that puts a greater onus on the people in charge to manage risks appropriately in ways that are reasonably practicable. If your employer has some crazed health and safety goon making unreasonable requests then that’s simply because they don’t understand the legislation or have recruited the wrong person.

  12. It would appear you’ve ignored track geometry. Perth’s lines enable line speed between stations. In Auckland it’s a case of negotiating curves with speeds much lower than line speed. Line speeds themselves are also lower here. 70-90km/h out west, vs 130km/h Perth. That’s a big difference.

  13. Geoff, Perth’s Fremantle line has quite a lot of curves, and certainly none of the 130 km/hr running of the Joondalup or Mandurah lines. The Kapiti “benchmark” should be the most relevant, because much of it is limited to 70 km/hr, and operates with TM’s and under the same legal framework as Auckland’s system.

    On Perth’s Joondalup line, the time from when a the train begins to decelerate before a station, until it resumes line speed is 1 minute (I timed it some years ago). Dwell time was 20 seconds. Compare that with the Western line where dwell time is 1 minute, of which doors are open 20 seconds.

    Auckland is the only Australasian city to go fully to ETCS, and would be a warning to other cities not to go with a down-scaled or incomplete technology, otherwise travel times would slow down and fleet requirements increase.

    1. @ Malcolm “much of it [the Kapiti Line] is limited to 70Km/hr”. .

      Not really. Of the 55Km journey from Wellington to Waikanae, the stretch 2.5Km-23Km is predominantly 85-100Km/hr and from 36Km-55Km is predominantly 95-100Km/hr.

      75Km/hr or lower only predominates between 0.6Km-2.5Km and 23Km-36Km – i.e. less than 30% of the total route.

      Average timetabled station-station speeds calculated from inter-station spacing and timetabled timings (assuming 1 min deducted at each intermediate stop for decel/dwell/accel) are as follows:

      Waikanae . . . . . . . . . . . . . .
      Paraparaumu . .7.2Km. . . .72.0Km/h
      Paekakariki. . . .9.4Km. . . .80.6Km/h
      Pukerua Bay. . .8.5Km. . . .63.8Km/h
      Plimmerton. . . .5.9Km. . . .59.0Km/h
      Mana . . . . . . . .1.3Km. . . .78.0Km/h
      Paremata. . . . .1.3Km. . . .78.0Km/h
      Porirua. . . . . . .4.1Km. . . .82.0Km/h
      Kenepuru. . . . .1.6Km. . . .96.0Km/h
      Linden . . . . . . .1.2Km. . . .72.0Km/h
      Tawa . . . . . . . .1.2Km. . . .72.0Km/h
      Redwood. . . . .0.8Km. . . .48.0Km/h
      Takapu Rd. . . .1.1Km. . . .66.0Km/h
      Wellington . . .11.8Km. . . .64.4Km/h

      The timings are no different for the Matangis than for the old trains they replaced (in fact the Johnsonville Line timetable has been slowed down!), and two stations on the Kapiti route have been deleted (Muri and Kaiwharawhara) with no improvement in end-end timings.
      A fair amount of slack is built into these timings to make on-time performance look good, but the reality of no journey-time-improvement despite higher-performance new trains is disappointing.
      Also, bear in mind that Wellington runs a number of express services during the peak period (skipping 5 stops) which cut the Waikanae-Wellington journey-time from 60min to 57min.

      Even though it performs better than the Auckland lines in the analysis, I don’t know that the Kapiti Line is a good bench-mark to judge what could be or what ought to be aimed-for.

    2. Auckland’s ETCS is full Level 1, not “down-scaled or incomplete technology”, so it cannot be a “warning to other cities” in that respect.

  14. To estimate attainable average speed, wouldn’t you just work directly from distance, line speed and train acceleration and braking characteristics? The relevant equations are: speed (metres per second) = acceleration (m/s/s) times time (seconds). Distance (metres) = (t^2 x a)/2

    For example, a train accelerating/braking at 1 m/s/s (said to be the specs of Auckland’s trains, as noted in the post of 10 October 2011) to a line speed of 20m/s (72kph), with dwells of 30 seconds, loses 50 seconds for each stop.

    For Newmarket-Papakura (27.6km with 14 stops including the terminus), that would give 35 minutes


  15. … A train accelerating/braking at 1 m/s/s to a line speed of 25m/s (90kph), with dwells of 30 seconds, loses 55 seconds for each stop. That would give Newmarket-Papakura in about 32 minutes.

    You would have to allow for any speed-restricted curves. I don’t have access to that information, but a glance at the map suggests that there would be few if any curves that might be restricted to below 72kph.

    If you want to take acceleration and braking as a very conservative 2/3 m/s/s, that would add about 2 minutes to each of the figures above. Any way you slice it, the present timetable of 45 minutes is pathetically slow.

    1. If the present timetable, from Papakura, is 45 minutes, I’m a kangaroo.

      The trains, on journey planner, are listed from 54 minutes to 58 minutes (actually, I think, it’s an either or situation) and, in marked contrast to the diesels prior to November 2014, push 65 minutes a lot more than they do 50 minutes. The trains, at night, going from Britomart are typically faster but if anyone thinks the trains are doing 45 minutes from Papakura they are not using 60 second minutes.

      It feels as though the big issue is Wiri. Regardless of whether or not there is train manager and/or driver exchange, the trains are generally forced to stop there to wait for the Eastern line (and apparently there is no signal either which slows things down too).

      Getting rid of train managers would be a mistake. In our 6 car today, for instance, there was a situation where the train manager was desired but was not available. While this is fairly extreme, the EMUs are also vandalised much more than the diesels were or, at least, the vandals are better at making this noticeable.

      If you are really worried about the wind, harden up because the doors, when the open automatically (as sometimes happens) do, at least, make things feel faster (which probably suggests they are).

      1. I was discussing *Newmarket* to Papakura, like the top poster, to avoid discussing train movements at the terminus.
        Push button door opening does not need to be any slower than automatic door opening, providing passengers can push the button to recorde the request before the train has stopped, which should obviously be arranged in an efficient operation.

        1. And the retractable footsteps should not need to be deployed unless there is a wheel-chair user present. . .

          1. If I get on at the middle car (and I usually try not to), and there’s nobody else behind me, I press the red button as I pass. Get those ramps in before the TM takes over.

          2. You know what really destroys dwell times? When the doors have to go local and only one is in use. I don’t see any need to gate all stations (solely for the expense), but some sure need it.

        2. Oh, whoops, my mistake. I guess not so much a kangaroo as an illiterate. However, it should be noted that in the same way that trains often stop in the tunnel or by Vector Arena, trains approaching Papakura often rest near the former Mitre Ten site or, even, opposite McDonalds. This only happens, in my experience, on the to Papakura journeys and, indeed, the trains seem to rocket along between Papakura and Takanini these days (I used to consider this quite a long trip but I never timed it so maybe I’m just noticing the acceleration and mistaking it for increased speed).

          As a final note, there really needs to be a 3:08 from Britomart to Papakura because of the crowding that occurs at Britomart and then is exacerbated enormously by the St Peters boys and RI pupils. I think in 2014 this was less noticeable because a lot of the Britomart people would end up going via GI but now everyone is on the same 3 carriage trains. In some ways, it is worse than 5.

          With the queuing of door opening, it is my understanding that this is not currently possible so I didn’t suggest it as an immediate solution.

        3. Opening all doors makes negligible difference in most cases and worse in others. It’s usually done as a matter of policy at certain stations and courtesy at others, for example where there are a lot of passengers or passengers on wheels. Not opening the middle doors with ramps is the fastest, but passengers seem to gravitate towards those doors and open them anyway. The ramps are the biggest problem and I don’t know why they (the manufacturers) can’t or won’t improve it, but there are several imaginable reasons. The trains are capable of under 30 second dwell times when the ramped doors aren’t opened, but that only happens when it’s very quiet and you end up waiting at a timed station for the scheduled departure time or at a junction for another service.

      2. The worst, worst part of the stop at Wiri (it happened in the diesel days too, to be fair), is if you’re going to Manukau, nothing feels more frustrating to have the Eastern Line have the doors shutting as you turn up at Puhinui.

Leave a Reply