Last week I wrote about the issue of dwell times on the new electric trains and how they combined with other factors mean the EMUs that are meant to be much faster are have instead actually ended up being slower. Rail users will also be acutely aware that the performance of trains has been appalling in recent months – on that it’s been reported that while the withdrawal of the old diesel trains will remove one of the causes of unreliability, the new trains are having teething problems and will need 6-12 months of bedding in.

That very same afternoon as my dwell time post the AT board papers were published online and an attachment at the end of the business report deals directly with the issue of train/rail performance by showing what is planned to improve it over the next year. The initiatives are broken down by month and in to four key categories – Operations, ETCS (the signalling system), Rail Infrastructure and EMU Reliability Plan.

All the initiatives are shown in the images below – although they can be hard to read so click to enlarge or open the AT report (1.03MB) and scroll to the last three pages – however here are the things that most stood out to me.

  • A range of fixes for the new trains over the next few months which should improve their reliability.
  • Improvements to signalling could provide significant savings including
    • potentially 30 seconds per train from addressing how the signalling system deals with level crossings (from August)
    • faster line speeds on the Onehunga line saving 15 seconds (September)
    • faster line speeds on Southern Line to Penrose saving 20-25 seconds (September)
    • faster speeds around the curves behind Vector Arena saving up to 30 seconds per train (October)
    • faster speeds on the curves on the rest of the network, this most affects the western line where savings could be 15-30 seconds (November)
    • additional balise to speed up dwell times at some stations and improve timekeeping between Britomart and Newmarket
  • Reviewing door opening and closing procedures to reduce dwell times (September), prior to that they are considering having doors opened and closed automatically at peak times.
  • Changes to rules for drivers which could save 20 seconds per train (October)
  • Hiring platform supervisors for Newmarket to improve punctuality (November) and extra drivers to reduce the need for some drivers to change ends on Western Line trains saving up to 2 minutes (December)
  • Closing Westfield Station saving 2 minutes per train
  • A new timetable in May next year that will provide a minimum of 15 minute frequencies between 7am and 7pm, 7 days per week. This will be to tie in with the roll out of the new bus network in South Auckland.
  • New platforms at Henderson and Otahuhu to be able to turn back trains running late – this may be good for train operations but not necessarily good for passengers if you’re travelling to a station past Henderson or Otahuhu.

EMU + Rail improvement action plan 1 - Jun - Sep

EMU + Rail improvement action plan 2 - Oct - Jan

EMU + Rail improvement action plan 3 - Feb - Jun

It’s nice to finally see laid out just what is planned to be done to improve services for customers both in terms of reliability and journey times as many of the issues listed are ones that have been raised numerous times already. It seems that for the Western line in particular the combined time savings could stack up to be a decent amount

While I understand many of the specifics won’t be known until the initiatives are rolled out, perhaps AT could try and pre-empt customer any frustration from slower trains by publishing some of the key points from the document above in a general public focused way on their website and at stations around the network.

Share this


  1. I’m surprised there’s no consideration of running express trains. Why is this? For the Southern line at least, every train doesn’t need to stop at every station. The stations at Westfield and Te Mania are usually empty and could be served by every second train. So either close those stations or stop at them less regularly during peak.

    1. According to the article they are shutting Westfield Station down. So at least part of your suggestion is definitely being considered.

      Can’t wait for the new timetable – in my opinion 15 minute intervals is much more important at this point than considering an express service. The distance covered by our trains simply isn’t great enough.

    2. Without a third (or forth) main to allow express trains to bypass those stopping any express train won’t be much faster as it will quickly catch the train ahead and not provide much of a time saving.

  2. AT have already had an attempt at closing Te Mahia but the Manurewa Local Board went complaining to the Mayor and the Mayor intervened. AT then decided to keep it open for about another two years then review its final fate.

    Still of the decision it and Takanini should be closed with a new station near Walters Road further south. Westfield I am a bit conflicted with as the station does service near by industry in the early morning and evening runs. However, once Otahuhu interchange is open next year buses feeding to Otahuhu Town Centre and the Great South Road (past Westfield) should compensate enough.

    As for turning back trains that are late, I hope AT decide to do what Melbourne does and publish publicly every month the amount of trains turned back or expressed followed by penalties if required.

    1. My understanding from a local Manurewa board member is that the Manurewa board paid for someone to count the number of people using the train at te Mahia and found that there was significantly more people using te Mahia than ticket sales/Hop card usage suggests. Also Fletcher Housing is building a number of houses (450+) on the so to be former Manukau Golf Course. I think this was the reasons for keeping te Mahia open.

      1. Those reasons came through when AT finally sent an update on the Southern Stations.

        The Weymouth housing development could have been dealt with by the two bus routes feeding into Manurewa.

        As for Fare data being different – well there was your fare evasion issue.

        Still say close it.

        1. I admit there does seem to be a big fare evasion issues at this station (gating Manurewa could help with this as I was told by the Board member many people using te Mahia were using to get to local schools or Manurewa Shops). Still the Manukau Golf course is just across the road from te Mahia and I am guessing with over 450 homes there it could mean that there is more than 1,000 people living in the te Mahia catchment. If 10% of the people use the train that would result in a 100% increase in the number of people using the station.

          I have read a AT report which says “Modelling suggests that there will be no patronage growth at Te Mahia station over the period to 2041 and little change to overall patronage on this section of the rail network with the station closed”.

          I think to see significant growth at te Mahia you would need to spend a lot of money to improve the facility as it is not the safest or nicest station to use.

          I guess I slightly favour improving te Mahia over getting rid of it but I acknowledge the case for keeping it is not strong.

        2. I believe patronage would be higher if fare evasion was stopped and safety at the station improved. The other day the train didn’t even stop there due to a “security issue”. You could see a bunch of what looked like the usual groups of hooligan youth.

  3. They should also increase the speed of the train. Although the train can in theory run at 120kmh, I found out in practise it runs quite slowly compared. For example city to new market it is like running at 30kmh. Fastest I saw it running is only 80kmh. Average is about 60kmh.

    1. Increasing curve speeds is a biggie here but we are talking straightening curves and quite how this is going to happen out West without a substantial budget is beyond me. When it was double tracked some curves at least should have been done. Most of the trip to New Lynn is governed by 50 km/hr curves, or less and there is a lot of them.

      It is not simply a matter of running trains faster over them as is or the risk of overturning skyrockets.

      1. There absolutely needs to be a straightening roadmap for the Western Line. The curves are an accident of history and geography, and need to be fixed.

      2. I read it as changing the speed limits on the curves. The EMUs are able to handle curves better than their predecessors so that allows room for improvement. I agree that straightening curves would be ideal but if they can get speeds up on the interim it would be fine. In fact with the close station spacing they can probably get speeds up to almost identical to what’s possible if they also sort the level crossings

        1. Has banking (cant) of the tracks been used to the maximum extent possible on curves ? On the Victorian Regional Fast Rail Project, train speeds were increased by using cant, whereby the outer rail of the curve was raised by up to 15 cm higher than the inner rail. Most railways have some cant on curves, but well below the 15 cm used here. By rejigging cant to speeds attainable by the EMU’s, it should be possible to remove many of the curve restrictions on the Western Line.

          There must be specialist consultants in cant who could advise on potential gains from optimizing cant. There is a down-side of excessive cant, that slower trains such as freight impose more wear on the lower rail if they can’t achieve the same speeds as passenger trains.

        2. Most if not all curves are to some extent canted however some areas could cope with a more “banked circuit” approach but there are some zones that won’t. Between St Georges Rd and Portage Rd level crossings New Lynn and the curve at Avondale (Blockhouse Bay Rd bridge) would be perfect, both currently at a leisurely 50 k’s. However Mt Eden Porters Rd level crossing is on the start/end of a curve so it has to be near level for cars as is George St LX. Similarly just down from that at Kingsland is a turn out diamond limited to 40 ks so that area could not be too canted otherwise trains using the turnout would potentially fall over.

          Straightening curves would do more good. And I don’t know if the EMU’s are actually more capable of cornering faster. They don’t have tilt technology. The SA’s certainly were and were much more low slung than the ADL’s or EMU’s, both of which are quite tall.

        3. It is a bit ironic after spending millions per EMU that this conversation is even been had. Who is to be held responsible for fitting them with such slow operating doors or such a restrictive control system that it has killed their speed advantage to a speed liability that is slower than the diesels they are replacing?

        4. It is more about the fact that the current speed limits allow for a large % overspeed misjudgement by the driver. With ETCS the speed limit is enforced with only a small overspeed margin allowed, therefore it is possible to increase the curve speed as long as the ETCS overspeed is well below the derail speed and the new curve speed is still within the window where customer comfort is maintained (i.e. cant still compensate enough) and rollingstock/track is not damaged over time. i.e. using ETCS for good rather than evil 😉

        5. Aren’t the EMUs longer per carriage than the current carriages? Wouldn’t this tend to make them worse on the corners? The diesel railcars certainly seemed to take the Vector corners faster.
          On the matter of cornering, I was on a 6-carriage EMU pair yesterday, in the front of the second set. Going around those corners behind Vector was the loudest of any train I’ve been on in Auckland as it seemed to be hauled about the track by the leading EMU, grinding wheels on the rails. Another double set observed heading into Britomart from the platform was similarly loud. Any ideas why the trailing set for 6-car pairs is so loud (mainly on the corners)/what could be done? I’d guess it is a bit of the lead pulling, so power control alteration for paired sets?

      3. Overturning isn’t a risk, you’ll have passengers banging across the cabin well before that becomes an issue.

  4. Your report does concern me with respect to the responsibility of the manufacturer of the new units. You state that ” the new trains are having teething problems and will need 6-12 months of bedding in.”. With modern manufacturing methods and Quality Control during build this should not happen and should not be required. This excuse is not acceptable from AT or the Council.
    Example; does it take Air NZ 6 -12 months to bed in a new Dreamliner? NO.
    Get a life guys, if your I Phone took 6 months to be bedded in your would be spitting tacks…….
    I think they have a manufacturing problem with the new units and that is not being publicly addressed.

    1. Despatch reliability of a new model of aeroplane is typically 96-97%. Over time as problems are worked out, that goes to 99-99.7%. But the airlines and manufacturers know this, and work very hard to get the numbers higher, and have a small amount of excess capacity to deal with delays and faults and failures.

      1. 787’s not a good example: years late and numerous structural, software, electrical bugs which had to be ironed out (it got the name 7LATE7 for a reason…) and resulted in a grounding for a month. Plus the fact Boeing is still struggling to get the first 20 early-builds out the door as no one really wants those sub-spec frames (selling one or two old ones at a time). There were years of testing and operation before Air NZ took their first, there were over 150 in service and hence plenty of experience for Boeing to iron out faults. Air NZ also does to to work-in new frames, especially first in service, with trans-Tasman hops for the widebodies, which is where the 787s started.
        There are not hundreds of these particular EMUs in service, so there is less experience at this point than was available to Air NZ. They are also fundamentally different forms of technology with different governing regulations and standards to meet, so are apples vs oranges

    2. You’re assuming it’s physical build faults and not software issues. Other than the loss-of-power issue, which has not appeared again, I’m not aware of any particular “hardware” faults with the EMUs. Door timing and the like is software-controlled, and some of that is down to client requirements that, once put into practice, appear to be a complete canine’s morning meal. The signalling system is not CAF’s, and their trains obeying the system’s settings is also not CAF’s fault. If the signalling system is set to be too conservative, how is that a fault with the trains? Likewise the questionable situation of TMs unlocking doors before they will open – AT’s decision, not CAF’s. I wouldn’t be surprised if it’s also AT’s decision that passengers cannot request a door release before the doors have been unlocked, either, but even if it was CAF’s decision it’s a simple software fix rather than a build fault.

      There is not a single thing on that list that points to construction issues on the trains.

        1. Definitely. But remind me, how many years behind schedule was the Dreamliner before it entered revenue service?

          Comparing trains and planes is just stupid. They’re completely different beasts.

          There is not anything obviously wrong with the actual software on the EMUs. They don’t just stop dead, or fail to open doors as requested, or not deploy ramps out of the middle cars. They’re doing precisely what’s programmed, at least on the face of it, and it’s those settings which are causing problems. It’s fixable, but demanding perfection from an entirely new system is beyond unreasonable; it’s actually stupidly naive.

    3. Don, the way to think about the adoption of new technology is not to jump to conspiracy theories about cover-ups, but instead consider the systems and processes the technology is embedded in. For example, following on from your Air NZ example- compare the airport and rail systems, keeping in mind some of the changes Auckland Transport is trying out to improve the trains. First, rail is part of the public transport network, unlike the airport, meaning that there are many more stakeholders involved who have the ability to voice their opinions, create regulations, request amendments to operating programmes, etc. This is reflected in the boards of Auckland Transport and Air NZ. Second, compare the physical context in which planes and trains operate: I am not a pilot or train operator, but I would argue there is more variability on land than the air, which means more time is needed to “bed in” new trains than new planes. Third, there are institutional differences in the organizing structure of the operators of planes & trains. It might be more difficult to change work processes or let go rail staff because of the level of unionisation, unlike airline staff. Also, AT works with TransDev/Veolia, where Air NZ does not sub-contract its flying operations. More layers in the organisation structure mean more time is taken for problems to be discussed and policies to be implemented, but also help with balancing oversight with efficiency. Fourth, here are some other ways rail and air operations differ significantly. All airline customers pay their fares, while some train riders do not, reducing revenues. Airplane routes are not structured to “stop-start” like train routes, which have multiple stations. Airplane routes and large parts of airports are largely protected from the public, unlike rail tracks and train stations. We should not be comparing using new iPhones and Dreamliners with new trains. The operating environments for each technology are substantially different, as are the “durable investments in complementary assets” (or patterns of use) that users develop in each environment. Hope this helps..

      1. Sorry Matt and Ha, there is no suggestion of a coverup. Just some bad purchasing decisions and management I am guessing.
        Jon Reeves is on to it above.
        The difference between Air NZ and AT’s management is that Air NZ Managers can get sacked for poor work decisions.
        Todays modern train manufacturers are also in the aviation and safety business, they know how to build good equipment.
        If the package purchased included software, it should have been tested and been ready for this environment. Or the scope of work had not been defined correctly.
        If AT is asking for new ideas now, they have let us the people who have paid for this train set down badly.
        You have a train set that can travel at 100kph averaging 60 kph. That is failure for a start.
        Unfortunately I see the changes in the AT list making stuff all difference at all, and these problems dragging out for sometime yet.
        Stop making excuses for AT…

    4. Don the equipment is working just fine. It does exactly what it was set up to do. The issue is it will take a while to decide how it should be set up now and make those changes. They could leave it as it is just like Apple does with your phone after you give them your money or they can make it better. That will take a while.

      1. That’s not a correct or even relevant statement, if anything it’s CAF that provides the product and then does very little, Apple, Google or whomever do significant amounts of ongoing work to the software on a device and are regularly pushing out software updates.

        Issues like not being able to pre-order a door opening, usual practice on many trains around the world, doesn’t occur in Auckland and I’m happy to assume until proven otherwise that that was a decision from AT. If anything most of these problems boil down to AT lacking the expertise to run an EMU fleet or even a modern PT system which Auckland is slowly developing, something they will in time develop but something Auckland has never bothered to fund in the past.

        1. Don is suggesting it was a problem caused by the manufacturer. It wasn’t as I understand it, they provided what was paid for. The issue is that the system now needs to be improved and that is up to the operator. Apple do significant improvements to their hardware but they charge you another $1k to benefit. Trains last a hell of a lot longer and it is up to the operator to see how they can best make use of what they bought. Dont get me wrong I was as surprised as anyone they would but moving mechanical equipment from a nation where cow stabbing is public entertainment but in this case the trains are working as ordered.

  5. Good to see an across the board go at this glaring problem BUT…..

    The ETCS is designed to prevent crashes. Because it really does seem like a square peg being forced into a round hole (word has it its an intercity system made for long spaces between signals being used in a tight signalled metro system) I can’t help but feel the approach speed increases will increase the risk of collision. Either use the old ETP system or start again but don’t bastardise this or their ain’t much point to it.

  6. Also good to see proposals at more cross overs between Papakura and Puhinui, This is so badly needed and it will mitigate delays when there is a break down etc. But I note it is dependant on funding.

    More loops?? Hopefully this means Onehunga especially where once one train goes down that’s it until it returns to the NAL north of Penrose. The current Te Papa loop was never wired in unbelievably.

    Positive to see shuttle drivers at Newmarket being proposed but really does this not show the stupidity of Newmarket’s design? In the interim connect up platform 4 to the Western line as it was meant to be for increased flexibility. And then reinstate the Kingdon St platform and cease west bound trains at Newmarket’s main platforms. Either that or run west trains south as they arrive at Newmarket and vice versa and have passengers transfer for Britomart.

    However nothing I have seen above will fix the log jam design that is Britomart apart from allowing the EMU’s to arrive on the platforms faster than the current ridiculous jogging pace.

  7. Westfield could easily be a “part time” station, with its use restricted to the peak hours for industry around the station – which would require some consultation, but lead to better outcomes for almost everyone. Say a select number of trains between 6:30-8:30am, and then again from 5-6:30.

    Now that the railway workshops that were the reason for that station have been supplanted by the Wiri Depot, there’s little reason to continue it. But it would be a pity for its current users to lose it completely.

    1. Westfield is a depot not a workshops, and it hasn’t been completely supplanted by Wiri – it continues to be KiwiRail’s Auckland loco depot.

      1. It shouldn’t really be a surprise there has been little growth at the station – there has also been very little investment in upgrades.

        A walking cycling bridge over the railway line from Bell Ave to the Waikaraka Cycleway and Westfield station not only would provide a useful addition to the cycling network it would also increase the potential walking catchment of the station.

        I would be disappointed to find out if the renewed proposal to close Westfield has anything to do with the recent overblown proposal for an East-West motorway.

  8. And if times don’t improve, who gets sacked? It sounds trite, but good performance management demands accountability.

  9. Not sure the idea to allow automatic opening of doors only at peak times is a good idea. This is a sure fire way to cause confusion to the public, as they will then also need to figure out when they need to open the doors themselves and when they don’t. Slowing things down at off-peak times. If you’re going to go down that path then it should be all or nothing. Chose one or the other, not half and half.

      1. Why? Serves no purpose to have that. It’s the slow TMs and the fact that doors can’t be pre-ordered to open until the TM activates doors that’s causing the issues.

        1. Most train systems around the world have automatic closing doors. If you miss the train you miss the train. Tough luck.

  10. I really like AT’s attitude to tardy rail performance. “Closing Westfield Station saving 2 minutes per train”. This seems like a one finger salut to passengers unless they will also have extremely regular bus services connecting with Southern trains at Otahuhu to Westfield.

    The issue is, just as we’ve seen with Waitakere, rather than truly fix problems AT just want to cut out stations.

    After a lot of studying of passengers at Swanson, the PTUA found the one main reason passengers have given up on Waitakere station is because of extremely poor reliablity of trains. About one in 10 trains Waitakere trains is cancelled! Pretty bad if it is the last train of the day, or if you truly need to get to an appointment/class/work at a specific time.

    1. Westfield is very, very lightly used. Lowest patronage on the network, I believe. Keeping a barely-utilised station open just for the sake of dozens of passengers a day is not actually the reasonable course of action; it’s many, many thousands of extra minutes a year for all of the other passengers on services that stop for the sake of the few who use Westfield, and the station is not likely to grow in passenger numbers because there’s nowhere nearby.

      1. I like the mix you use… “dozens of passengers per day” vs “many, many thousands of minutes per year”. Perhaps you could say, thousands of passengers per year and thousands of minutes per year? I’m not going to promote a station which has limited scope in usage. But I disagree that same arguement be used for closing Waitakere station which is a completely planned situation by AT running train reliability down over many years forcing passengers to quit it, and the fact the Huapai and Kumeu are expanding towns.

        However, if Westfield is to close, in the interests of current and future pt users AT should be providing good bus/rail connections to that area, and the new interchange at Otahuhu due to open in 15 months should be used for that. But don’t cut Westfield station until that option is in place.

        1. Fine. Thousands of passengers a year and hundreds-of-thousands of minutes a year. Happy now? If they’re talking about two minutes per train being saved by closure, multiplied by, (extremely) conservatively, 3k passengers per day for 300 days a year, that’s 180k minutes a year lost to a station that is used by an average of 100 people per day.

          Westfield is a dog. Its connectivity to the surrounding area is worse than useless, and even if footbridges are built to connect some of the surrounding area it’s still hundreds of metres into the yard which means walking catchment will forever be small because the station is forced to be significantly disconnected from its surrounds.

          No matter how you spin it, the station is a massive time sink for the vast, vast majority of passengers who pass through it for the benefit of a minuscule number; 0.3% of boardings across the network in March 2014 (last stat I could conveniently find) were at Westfield.

  11. It is good to see FINALLY AT train management recognising the issues CR MIKE LEE has been telling management for over 12 months. But I do wonder about the expertise of the AT management. Slow closing of doors etc… would have been in the specifications they agreed to 3 years ago so could have forseen this as an issue adding upto 5 minutes per journey. Why was nothing done about it earlier?

    AT rail management are really behind the eight ball and it also makes me wonder why we are allowing Transdev to clip the ratepayers ticket when it seems like they provided very little forward thinking on these train operating issues.

  12. “New platforms at Henderson and Otahuhu to be able to turn back trains running late”

    Unbelievable! I’ve been joking for several months now that when Waitakere trains are canned, they’ll probably start sacrificing Henderson-Swanson instead of Swanson-Waitakere every day so they can meet their KPI’s. Now we find they are not only actually planning to do this, but also spend up large to give themselves the ability to do this.

    If they are going to build a new platform, let it be at Swanson, so it can actually be used for something that grows patronage (Kumeu rail) rather than for the purpose of cancelling trains every day and inconveniencing passengers.

    1. Auckland Transport and their handling of rail look more incompetent as each day passes. Agreed Geoff, a platform at Swanson (as originally planned until that flawed 2012 report) makes much more sense.

      I’m just wondering, can AT cancel rail services without holding public consultation first? Given no public consultation over the Waitakere station being axed ever took place, one must wonder the validity of the decision?

      1. What would be the point of putting a turnback at a terminal station?
        And yes, they can and do cut/add services without public consultation.

        1. Cutting routes without public consultation?. Hmmm.. can you name some they cut without ANY public consultation apart from Waitakere rail services?

          Are they not meant to consult the public on the RLTP about cuts to the rail network? Auckland Transport never consulted the public over this plan.

        2. Yeah they dont typically cut routes without consultation, the only cut PT route I can think of recently is the waiatarua peak only bus which was consulted on.

          In the western consultation they specifically said they will not take any feedback regarding waitakere station into consideration, which I believe is a serious issue.

          AT just seem to pick and choose what they consult on and the people making the decisions dont even know the areas they are affecting. They are just basing it on possibly untelling numbers and some aerial maps, when in reality they should be going out there in person BY PT and seeing whats actually going on.

          I bet everyone on the AT board and most of the AT transport planners commute to their PT-friendly workplaces in their Audi’s.

  13. Extra platform at henderson? Great so next up is the run down of Sturges-Swanson every time they run late. Why not just implement a 3rd platform at Swanson which they supposedly need to run a Huapai shuttle ADL (note: I do contest this being a requirement however). Any extra rail coverage in Auckland is better than not.

    Also in regards to 7am-7pm 15 minute frequency it would be nice for them to elaborate on the range of lower frequency before/after 7, like… are trains finally going to run to midnight every night like the NEX?

  14. Unrelated to trains but contained in the report linked to above. Pedestrians and their inability to cross the road! – I wish AT would put this amount of effort into patronizing drivers about what a ‘Red flashing man’ means, and how drivers are required to give way to crossing pedestrians, rather than aggressively ram them off the crossing. Of the 50 deaths, how many were actually caused by a pedestrian rather than a careless driver.

    Why a simple change to a flashing green man, to reinforce a pedestrians right to be on a crossing can’t be made, I don;t know. Or Have a red traffic light until the crossing phase has entirely finished.

  15. Says the new timetable for may this year was meant to have frequent 15 min services 7am-7pm 7 days…

    Guess they have pushed that back to next year?

Leave a Reply

Your email address will not be published. Required fields are marked *