This post by Matt L was first published in September 2013. It feels like we’ve had electric trains forever but this is a good reminder that they were only fairly recently introduced.

At the transport committee meeting on Wednesday Kiwirail provided an update on electrification progress while Auckland Transport had an update on what is happening with the EMUs.


The presentation was broken down into three areas looking at the key pieces of work remaining to be done. These are signalling, traction (wires) and operational preparedness.


The new signalling and train control systems have now been completely installed and are operational. They said about 18 months ago there were an average of over 100 faults a month to either the signalling system, points or other systems associated with controlling the trains. This is down to about 50 a month and is expected to go lower still. My understanding is that this number includes any small issue and often they may not even affect passenger trains but even so it still seems like a lot. The one piece of work left to complete is the automatic train protection system which will help prevent accidents as it will stop trains from going through red signals. We don’t currently have this system on the rail network although I believe that AT are retrofitting our existing trains with it even though they will only be used for another few years. The new EMUs will come with it already installed.


The wires are the most visible part of electrification. Kiwirail say they their contractors have now completed all of the mast foundations and have over 70% of the masts in place (2785 out of 3825). If you look at a part of the rail network that has already been electrified you will see that there are three wires that make up the system. One is an earth wire, one is a catenary wire and the other one is the wire the trains connect to get power. All up there will eventually be 550km of wire on the network and so far 355km have been installed (~65%). The parts left to do are the western line from Mt Albert to Newmarket, Newmarket to Britomart, the eastern line from Westfield to Britomart and a bit around Papakura.

Electrification Progress - Sep 13

Associated with the wires are various pieces of work around making the network safe. Kiwirail say that 90% the earthing and bonding of elements in and around the rail network is complete while they are 60% of the way through installing the screens (like the one below) on bridges to help prevent access to the wires.

Bridge Screens

The work will continue over the rest of the year and into another extensive rail shutdown over the Christmas period. They said the rail network would be closed from 26 December through to 19 January which is almost four weeks. I sometimes wonder if these large shut downs will ever end and one of my greatest fears is that Kiwirail will get so used to doing all of their normal maintenance tasks in these shut downs that they will become the norm even after electrification is finished.

In addition to the weekend and Christmas shut downs, there have also been shutdowns on Mondays to Thursdays in the evenings. I have heard that these will now be extended to Friday and Saturday nights however there has been nothing official from AT or any of the other agencies involved.

It’s worth noting that there is a shut down this weekend which will see the wires from Westfield to Wiri completed and livened up. That will mean all AT will be able to test the new train all the way from Wiri to Newmarket and along the Onehunga Branch line.

Operational Preparedness

The only part I found interesting in this section was that Kiwirail have decided that they won’t be maintained the wires themselves and as such will be contracting out the work to another company who they are currently under negotiation with. While they will be working on behalf of Kiwirail I do worry about the fact there will be yet another player being involved in keeping the network running.

Electric Trains

As most readers will have known by now, the first of our new electric trains are now in the country and it is tucked away at Wiri going through initial testing. AT say the next two trains will arrive together in November then we will start getting two a month until November 2014. From December 2014 things ramp up even further and there will be four per month arriving until all 57 are here which should be around July 2015 (which is earlier than originally planned due to CAF opening an additional production line)

The first three trains will undergo testing for 3-4 months to ensure all bugs are ironed out. After that trains 4-9 are expected to take about two months to test while after that they will take about 4-5 weeks before being ready to be used in service or for training. Testing will take place at night or during some weekend shutdowns when there are no other services on the tracks so don’t expect to see one while waiting for a train.

The first passenger services are still on track for April and will start on the Onehunga Line. The key reason for waiting till then is that they need a about 7 trains in the country first. Three are needed for Onehunga services (two operational and one spare), another 2-3 are needed for training purposes and another few will be going through testing. Trains will be rolled out on a line by line bases and will only happen when there are enough to replace all of the services on a line in one go to avoid mixed running a mixed fleet where possible. Following Onehunga the Manukau services should start running with electrics a few months later in mid 2014. Both Onehunga and Manukau are shorter runs with less services so not as many trains are needed to be operational. The Southern line will be the next to go electric at the end of 2014 and the Western line will be the last to change over.

 EMU Introduction timeframe

You will notice on the diagram above that after the services on a line are replaced, that a few months later there is a timetable upgrade. This is to get the lines eventually to what is proposed below. AT said that the shuttle services to Papakura would begin in the end of 2014 once the southern line services were replaced.

Post Electrification frequencies

However while those of us out west in particular will have a while to wait before we see the new trains, AT also said that there may be the opportunity to at least replace all weekend services with EMUs in September next year. I’m also hoping that as the other lines start converting to electrics, that a few trains will be freed up to boost western line frequencies in the interim.

AT also say that modelling has taken place to look at timetable impacts of the new trains and they expect they will be able to deliver time savings of around 12% on a run. That means a trip from Papakura to Britomart would drop from 53 minutes to 45-49 minutes (I’m guessing depending on if stations like Westfield and Te Mahia close as proposed). Of course AT won’t know the exact time savings until they can actually test the trains in real life.

We are really starting to get into the business end of this project but unfortunately that also means a lot more disruption. Looking at current plans there is a weekend shutdown every 2-3 weeks from now up until Christmas and that is something that will not be doing any favours for patronage and people’s views on the service. If there is one advantage to the pain, at least now we are starting to see the light at the end of the tunnel.

Share this


  1. This is a pertinent post at the moment.

    The issue of dwell times rose almost immediately upon the implementation of the electric trains. This single issue has been the largest impediment to Auckland realising the full benefit of an electric train service.

    After two years of waiting for this urgent issue to be solved whilst AT spent it’s time focusing on fantasy projects we finally have a “solution”. Between 6am and 7pm train doors are now going to open automatically. Only problem is the doors take just as long as they did previously to open and close!! I’m sure one of our regular writers will tackle this issue next week.

    Congratulations AT you’ve done it again.

    1. I have timed a few quicker dwell times recently, around 40 sec rather than 50 sec, still poor but at least better than it was, they have obviously found some fat in the system.

    2. The slowest part of all dwell times is when people are trying to get off the trains the passengers on the platforms do not stand back so the ones on the carriages can get off without having to fight their way through the ones wanting to get on . It seems to me they have in their heads that tthe train will leave without them . The worst place I have seen is Britomart as you try to get off the train the ones on the platform waiting for the next service stand so cloose to the doors the passengers trying to get off have to push their way through the mob causng all sorts of problems . So why don’t AT teach people just to stand back for say 15sec’s untill the train empties then they can get on and the Dwell time could then be reduced by half ?

      1. In SG and Japan, many platforms have markers indicating where passengers should wait to board, usually on either side of the train doors. Perhaps that could be a workaround? Although it assumes the train will always stop in the same position each time.

      2. All these sorts of etiquette things could be displayed as posters at stations and on the vehicles themselves etc to improve our education on these things. Of course they have done this with other things already.

  2. “Only problem is the doors take just as long as they did previously to open and close”

    Well seriously WTF did you expect it would do?

    Its been well documented on this blog that the doors cannot start the opening cycle until the train is verified as completely stopped by the on board software, only then can the trains or LE release the doors.
    Its how CAF has programmed the software. Presumably because thats what they were told to do by ever cautious safety boffins.

    I’ve been on trains where the wheels skidded on a wet patch or something on the rails when the train was stopping at a station, and the train refused to let the doors open as it didn’t believe it had stopped, when clearly it had, it took several minutes for the LE to convince the train otherwise and get the doors open.

    So at best, the “all doors open” change is going to have the doors opened 1-2 seconds quicker than previously when the LE punched the door release button when the train said it had stopped and everyone then opened their own door themselves.

    And don’t forget at the end of the stop the train won’t go anywhere until it verifies all the doors are shut. So the smarter way to run the show is to encourage faster door closing over faster door opening. And thats easily done by not having the TM keep one door open while he closes and then verifies by eye all the other doors are shut before he shuts his one – FFS – the train *knows* the doors are shut and the TM acting as door manager clearly doesn’t add anything but delay to the “get the train out of here” process and the end of each stop.

    But anyway I think a related and probably, bigger problem than that is clearly the ETCS implementation which is clearly being way too conservative with how fast trains can approach/depart the station – that has the bigger impact on overall journey times than if the doors open and close a second or so faster at each stop.

    To my mind if you’re busy handing at bouquets of shit to AT, the single biggest (financial and operational) mistake AT made with EMUs since this post was not buying more EMUs near the end of the first order when they were offered a good price to buy more – as it would have enabled CAF to keep the production lines running, helping CAF out and saving AT a good wad of cash – way more in fact than the borrowing costs of buying those EMUs sooner than planned and ensuring we had EMUs on hand to run 6 car EMUs on all peak services instead of this silly mis of 3 and 6 car EMUs they run now.

    And yes that would have impacted the BMUs purchase proposed by AT and about to be approved by AC [if not already signed off]. But in that case we could have retrofitted a new trailer car to a few of the EMUs that needed to run to Pukekohe for a total cost which would probably be a fraction of the cost we now face of buying these additional EMUs from CAF after such a delay as we’re doing now.

    1. It’s easy to blame AT for not buying more trains, but IIRC it was central government that made the call to purchase 57 sets (up from the 38 sets they first proposed) – though I assume AT ultimately became the owners. Which brings me to the other point: AT can only spend money that it has been given by central government via NZTA or by AC. If they’ve only been funded to buy 57 sets, then that’s all they can do. Easy to dump on AT, but in so many ways they’re not in control of their own destiny.

    2. I believe driver door operation is on it’s way along with the introduction of Transport Officers. However, they will have to tackle the lag at the beginning and end at some point as every second will count for trains running through the CRL.

      It’s obviously solvable as most other systems around the world are run safely without a 3 – 4 second check at the beginning and end.

    3. Are the train managers also checking for general safety issues before closing though so that no one is jammed between train & platform or perhaps look like they might, or some other possible likely issue? Of course they then go in so it’s only of some use, due to the delay before actually starting. I guess the driver will check screens of CCTV before hitting go, but these screens are pretty small etc aren’t they? I think they doors should be just left open for less time…could be dangerous to start doing that if people are not used to it and start trying to open doors again etc.

      1. My bag got stuck in the doors of an old train once. I think it was a trainee manager who started closing the door too soon/was too nice. Of course, if the morons in front of me hadn’t blocked the entire escalator and then stood as well, none of us would have even been in a position to need the niceness of the train staff.

        I think the current doors work like lift doors so they open when people get in the way, but in principle you can see why having someone watch the doors is useful. And why it is useful to have someone have a sharp word with idiot college pupils who bounce the doors open.

        1. They are more like a garage door than a lift door. A lift door will have a sensor that stops the doors closing if anyone is standing in the doorway, whereas a train door will only stop closing if it comes into contact with someone.

  3. Only way to improve dwell times is to set the process up like the desils used to have and which the current Wellington trains use. No need for short cuts in safety then like butchering the door close process. Currently no matter who opens and shuts the doors, it’s slow to very slow. Would love to see the ETCS to be slightly less conservative on approach to speed reduction areas (approaching reds at end of platforms) have more in fill belise at the slow start platforms to allow the trains to move at more than glacial speed until it reaches the signal. Anything else would just be a safety issue or cost cutting measure.

Leave a Reply

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