The first of the trains is in the country and undergoing testing with the good news being that the testing is going well with the train performing well. This was the update to the AT board yesterday:

Completion by KiwiRail of the first phase of electrification and permanent energisation of the overhead 25kV traction system was achieved in September between Westfield and Wiri
including the Wiri Train Depot.

The first EMU operated successfully under its own traction power within the depot on 30 September following the first weeks of static testing and commissioning in preparation for dynamic testing from the weekend of 5 October.

Progress to date has been good, with the control system stability checks going well. Testing progressing according to the programme.

The EMU testing and commissioning programme has been finalised and issued to KiwiRail.

The second train is in transit to New Zealand, due to dock around 4 November, with the following trains due to be shipped at 2 weekly intervals.

CAF are now well established in production mode after the summer shutdown. Vehicles for the first 10 trains are now in production, with trains 3, 4 and 5 in test. The supply of materials to the production line is working well.

EMU at Wiri

Our new electric trains are going to be wonderful. They will be bigger, more frequent, quieter and faster than the trains we have today. However there has been one benefit that hasn’t been talked about much and that is how much it is going to cost to run them. We all suspect that because the trains will be running on electricity that they will be cheaper but the question is just how much cheaper. The good news is that thanks to Auckland Transport I now have some information that can help to answer that question. The information that has been provided compares the fuel/electricity costs and the maintenance costs which are admittedly just one part of the total cost equation. The cost per kilometre for running our new electric trains compared to what we have now is below.

EMU OPEX costs per km

So all up the electric trains are 54% cheaper to run on a per kilometre basis than our existing trains which is a massive difference. Some quick calculations suggest that on the current rail network the trains run about 3 million kilometres per year and so based on that figure it suggests fuel and maintenance costs are currently about $24 million per year. By comparison running the same frequencies with electric trains would save about $13 million per year. That’s not too bad however the real benefit comes in the future when we want to run more services. We could easily see the number of service kilometres double from what they are now and so the good news is that if that were to happen, the train operating costs would still be less than what they are now.

As mentioned the costs above only represent one part of the cost of running the rail network and many of those costs will not change with the introduction of the EMU’s. Some of the other costs include:

  • The Transdev contract which includes Staffing costs e.g. drivers, on-board staff, ticketing staff as well as the people behind the scenes that keeps the network going.
  • Track access charges that we pay to Kiwirail to run and maintain the rail network.
  • The running and maintenance of the various stations on the network.
  • Insurance.

The big thing that would impact on the operational costs from running more trains is the need for additional staff.  We would obviously need more drivers to drive the trains as well as more on-board staff (something I understand the unions have demanded be retained despite the trains being designed for driver only operation). At this stage I’m not sure just how many extra staff would be needed or how much that would cost but I suspect that with the savings that will come from the electric trains we should be able to make a vast improvement to the number of services that a run each day for no additional cost above what we’re already paying. That should mean we can improve both peak and off peak services for little or no additional overall cost and boosting frequencies can have a really positive effect on patronage.

Being able to run more services for the same amount of money is the kind of story that Auckland Transport need to be shouting from the roof tops. This is especially important with so many people concerned about rates and council finances. So good for passengers, good for the city, good for the environment and good for the  accountants. What’s not to love about these trains?

Share this


  1. Do these costs include track access fees or any other costs related to the tracks/signals?
    It would be interesting to compare these operating costs with other transport options: ferry, bus, LRT.
    What exactly would the non-driver staff be doing onboard? There are supposed to be security cameras throughout the trains and information displays. Will they be paid a living wage?

      1. Depreciation costs on the new trains will be significant compared to the assets they replace. Assuming 25 year life on $400M asset, depreciation will be $16M PA, (more when we include other infrastructure) so $13M operational savings mentioned above are great but not the whole picture.

        The business end of running the new trains must be about delivering an improved EBIT line from passenger and revenue growth as much as driving savings.

    1. BTW I think track access charges cover (in part) infrastructure costs, such as signalling? It’s seem strange if KR charged TAC’s and then invoiced separately for signalling etc.

  2. it would be a shame if train staff were retained solely because of Union demands.

    Efforts to protect labour from technological progress seem misguided. In the long run they seem more likely to reduce productivity and thereby lower overall employment in the wider economy.

    For example, having more onboard staff will increase the cost of running trains. By extension this will reduce the number of services we run, thereby reducing the number of train drivers that are needed. Hence inflating the demand for labour for onboard staff depresses the demand for labour elsewhere.

    And that’s before one considers the socio-economic benefits from simply being able to run more and/or cheaper services, which ultimately will lead to job creation elsewhere. Not to mention lower fares, which benefits a large number of (often low income) people.

    Unions definitely have a role to play, but resisting the benefits of technology is not one of them, IMO. Is there anyway we can get some clarity on this issue? Be a shame to talk up a storm in a tea-cup without having some more information what is being argued by either side.

    1. If we lose train managers, during the peaks the labour could be reallocated to providing more drivers and more services as you say, and off/peak and into the evenings it could provide both on-board and at-station security. This would be a better allocation of the same labour resource as it would provide more peak services. Some former guards would become drivers, others would do security.

      1. yes although Unions could argue that there’s no guarantee the same people who are currently working as train guards are suited to being security guards and/or train drivers.

    2. From the perspective of the union members though who are TMs though, you’d be pretty annoyed if you had been paying for union membership and the union not advocating for your job to be retained.

      1. I think people have been reading too much US and Australian literature on this subject. So called ‘featherbedding’ is not an issue here at all and hasn’t been for decades. Our employment law very weak, so no need to place all the blame on unions.
        I suspect unionisation low for TM’s with lots of new staff in past few years. Anyway if we are looking are future big increases in rail frequency all this will mean is not hiring new staff, and reallocating existing staff to services than need them.

        1. Plus, if you provide for security via other means (such as a transit police – something which I don’t support, hire extra normal police if you need them!) then you also have to pay for that. So abolishing train managers / Maroi Wardens / security staff could lead to extra costs down the line again. Even more so if we only ramp up security after we have run down the safety reputation of our system, and lost patronage because people are afraid to get mugged on board. So I say keep human presence on board, and/or station staff.

          Staff-less systems are for robot cities designed by engineers*!

          *[I am an engineer]

        2. I take it you’re pretty fit from not taking all those staff less elevators about the place. Bring back the bell hop! 😉

  3. Any ideas what is the “idling” cost of the current Diesel hauled trains?

    Looks to me that each trip ending or starting at Britomart they sit there in Britomart idling away for a good 10 minutes or more (especially the loco hauled units, but also the ADL/ADK sets), so technically “the cost per km” of this is $0 as they don’t go anywhere (well the fuel cost is built into the per km rate somehow).
    By idling they consume fuel and also require that the ventilation fans at Britomart be running at a higher rate than they would otherwise need to be.

    So surely, the reduced costs of Britomarts ventilation bill should be included as well by AT? (not that it will be a huge part of the running costs – or is it?)

    1. Technically, when idling, the immediate cost per km would be $∞ ($infinite) (not $0) as it is using fuel while not moving anywhere. I suppose a better figure would be average cost per km from the end of the previous trip (to include terminus idling) to the end of the current trip.

    1. Unlikely. Savings not reinvested in more services are more likely to be retained so that Auckland’s PT moves closer to the Government’s 50% farebox recovery target.

      1. However given the opex savings on the improved service delivery and assuming more passengers and higher farebox, it’s likely there won’t be so much pressure to increase fares in the future.

  4. Matt, did they give you a figure on which trains the diesel figures were based on, or was it averaged across the whole fleet?

    The reason I ask is that the EMUs are both bigger and faster, meaning they can move a lot more people at a given number of service kilometres. In other words the EMUs are even more cheaperer!

    1. I think what Nic means is that if you compared the cost *per seat-kilometre* then the cost advantage of the EMUs would be even greater than the 54% differential quoted in the post.

      1. Per-seat Km’s are useful numbers when you’re an airline and I think when the trains are mostly full.

        Basically, the running costs of an empty EMU is under half that of the same empty Diesel unit.

        So you can reliably say that you can deliver the same good/bad/indifferent service for under half the running costs with EMUs.
        But thats not why we’re doing this [and I hope AT understand this point].

        When the EMUs are full (e.g. at peak) then the percentage/amount of the passengers rail fare that is “consumed” by running costs of the train they are on will be way way less because the trains are cheaper to run and/or they also can carry a lot more people over the same distance and/or time.
        Which in theory means with EMUs the farebox recovery percentage automatically goes up – even if everything stays the same passenger number wise?

        So, what we have here in a nutshell is that the EMUs are the brand new Airbus 380’s of the PT world, compared to the old DC-8 and DC-10 diesels we have in use now.

    2. and also that the total in-service hours required to operate said EMUs for said kilometres would reduce, which in turn would deliver some additional savings (NB: That’s assuming that the above rate is not a composite rate that includes time-based costs).

    1. It’s more impressive if you don’t compare the savings to the existing trains and timetable, but rather compare it to what the costs would be to run the old clunkers (or new diesels) to what the EMUs will run to. So the savings are probably closer to double once we start running them every ten minutes all day all week.

  5. Do you know or can confirm if the new electric trains put power back into the grid under braking similar to the system set up in hybrid cars when the batteries are charged during braking?

  6. Are the figures above for the EMUs based on a 3 car train or 6 car? If 3 car, then if it’s coupled to make a 6 car train, does that double those per km electricity costs or only slightly increase it? And what were the diesel figures based on?

    1. Yes, a 6-car set would have less than twice the running costs of a 3-car set, but presumably not much less (slightly better wind resistance than running two 3-car trains separately). However, maintenance costs would remain the same, so in combination to with only a slight reduction in comparative air drag, I would be surprised if the “per seat” costs of a 6-car train would be less than 80-90% than the 3-car train.

      It’s different with driver costs. Drivers are expensive OPEX (even though they aren’t paid massive bucks) so having one driver less for twice the passenger capacity does make a difference (but that cost isn’t included in the above figures).

      1. You’re kidding right? The second train will be using half as much electricity as the first at minimal cost to the first train, the difference in windage is immense as the second train encounters no headwind, only friction on the sides.

  7. I know it would still be cheaper, but just curious to know how a 6 car EMU would compare in running costs to a 6 car SA set

  8. Certainly the direct OPEX of an Electric EMU is substantially cheaper than a Diesel one,

    But Track Access Charges,will be higher (possibly considerably) for an electric Unit

    There has been several hundred million dollars spent for the electrification, and that expense needs to be depreciated and maintained, thus the cost for track access for EMUs ( which are the only equipment that uses the overhead in Auckland) should be higher than an equivalent diesel vehicle..

      1. I think +1. What staff do Vancouver use on the skytrain? It’s driverless but do they have staff to help passengers?

        1. They have roving “ambassadors” who tend to be based at the stations rather than on the vehicles. They also have CCTV and help points at the stations, and CCTV with two way communication on the trains, and staff dedicated to monitoring and resolving that. Most of all they have their own transit police whose sole purpose is to be dispatched where needed on the network.

  9. Here is a question someone may know to answer to here.

    For the new overhead lines I have noticed that between some of the supports there is an additional one or two bright steel cables. These are not to be confused with the dark catenary cables for the power.

    These bright cables appear to have no insulation from the supports and therefore not live however every so often there appears to be an extra PVC covered cable connected to it.

    What are these cables for and why does their application vary so much?

    1. I think what you are referring to are structural supports. The contact and cantenary wires are under tension so they pull against the masts and other anchor points, often at funny angles where they are strung around corners and what not. Those other cables are there to keep the posts upright, same way as power poles are often anchored with extra cables. They won’t be live but sounds like some are insulated where necessary to prevent arcing from the contact wire.

      1. I thought about that however the cables seem rather slack and a pole would need to lean over about 5m before it would come into tension. They other thing is that they tend to be only on one side so the northbound track may have them but not the southbound. Of the ones I’ve seen they have all been on straights.

    2. Could you provide a bit more detail SFL, such as a specific location or a photo? At first I thought you were referring to the return wire (or earth wire), but that is aluminium and continuous so probably not. In some locations (eg Newmarket Station) it’s covered because it slightly breaches clearance requirements. The return wire is also bonded to rail at regular intervals. However, Nick’s answer might be the right one.

  10. On the topic of cabin staff I rather like the idea. There is nothing wrong with a little onboard security and aid to keep things all in order.

    1. They only really operate the doors, which the Driver can do and as for security I have never really seen them do anything, which is why you have transit police to enforce the rules. I had to laugh the other day when the ticket inspectors were checking tickets out of Britomart (Why do they do this when you can’t get on a Train at Britomart without a valid ticket) and there was a signal failure, which caused heaps of delays etc. You could just see people getting ready to explode!
      Wouldn’t you make the call hey we won’t check tickets this afternoon, man they walk into these things!

      1. You can get on without a valid ticket at britomart. Let’s say you are going from New Lynn to Glen Innes. You take the train to Britomart from gateless New Lynn to Britomart and transfer onto any /GI train there and then leave onllat gateless Glen Innes. This is why Ticket checks also have to take place in Britomart/new market despite the gates.

      1. I think onboard train staff are like bank tellers, technological advancements have taken away a lot of their role but not eliminated it completely. You need some for certain things, but we don’t need one to process every customer.

        I think we’d be better with a team of TM/security/ticket inspectors that roam the system and can focus on particular hot spots (say Friday night heading into town for security, events and sports fixtures for assistance, certain stations for evasion), but aren’t necessarily based on every single train running at any time. Do we really need anyone on peak commuter runs?

        1. I don’t recall seeing a single solitary security guard on the Seoul Metro system in the two years I lived in South Korea – at any hour of the night. It’s certainly a very low-crime city, but it shows it’s possible to operate a (massive) rail network with minimal staffing costs.

  11. Do we really need anyone on peak commuter runs?

    If we cut the number of staff on any given train run, we could run more frequent services off-peak for the same staffing costs. That’s a huge potential benefit.

  12. I like the armed PSO at train stations they have in Victoria which patrol stations at nights or after big events. Kinda off topic tho

  13. Are these figures per carriage/km or per train/km, and if the latter what is the train length assumed, and is it the same for diesel and electric?

  14. The interesting thing about these per train kilometre figures is how low they are generally.

    It means that more frequent offpeak services could well be profitable on a marginal costs basis. This would reduce the total subsidy needed for the rail service. It’s important to promote that goal to win political support for more frequent service.

    Let’s estimate the marginal cost of running an extra offpeak service on a 30km route at an all up average speed of say 30kph (say 40kph in service = 45 minutes running time, plus 15 minutes layover).

    In very round figures:

    energy: 30km @$2/km: $60
    train maintenance: 30km @$2/km: $60
    driver: say 1 hour @$60/hour: $60 [fn1]
    variable track maintenance: ?? [fn2]
    servicing capital cost of the train: $0 [fn3]
    admin, station, signalling staff etc: $0 to a good enough approximation [fn4]

    Total: $180 plus ?? variable track maintenance.

    To turn a profit, on a marginal costs basis, at an average fare of say $4, the new service needs about 45 new boardings (that’s new public transport trips, not just people transferring from other services). Plus something to cover the variable track maintenance cost which I have not guessed.

    This number seems surprisingly low.

    To put it another way: because of the very high fixed costs of the rail network, it makes sense to run the most intensive service you can to spread the cost, PROVIDING marginal costs can be kept low enough (this is the principle of Vancouver’s driverless Skytrain, which runs very frequent services at very low per train/km cost).

    To put it a third way: the way to improve the financial results of your public transport service is to focus on growing all day patronage, not just the journey to work. This focus should flow through to a lot of things about how you manage the service:
    – less obsession with peak hour seats to the central area;
    – more priority to things that are important to your less frequent, less expert, less CBD-focussed offpeak riders: a complete, legible network allowing anywhere to anywhere travel with transfers; good facilities at interchange points; clockface and memory timetables; initiatives to improve punctuality; real time information systems; integrated fares and ticketing.

    footnote 1: wild guess – maybe others can advise? I exclude ‘train managers’, as urban trains all over the world have had one person operation successfully for many years. If you want a frequent cost-effective train service, that is the way you have to go.
    footnote 2: costs varying with the tonnage carried. Not fixed track or signal costs.
    footnote 3: as you bought the train anyway to use for the peak period service.
    footnote 4: as they are already there to facilitate your existing service. Some increase may be needed in key places as patronage grows.

Leave a Reply

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