Melbourne is currently doing a lot of work on their rail network in preparation for the future Melbourne Metro, so when it opens they can maximise its potential. In fact, it goes further than future proofing with much of the works being done now so the transformative effect of the Melbourne Metro can be unlocked from day one. This includes:

  1. Longer Platforms for New High Capacity Metro Trains;
  2. Removing Level Crossings so more trains can run safer without any localised congestion effects;
  3. High Capacity Signalling allowing more of these High Capacity Metro trains to be run per hour;

So what lessons should the CRL learn from Melbourne?

Level Crossings

Across the electrified network in Auckland there are currently 45 level crossings, 31 are road/pedestrian crossings while a further 14 a pedestrian only crossings. The majority of level crossings are on the Western line with the rest are primarily along the short Onehunga Line with another cluster around Takanini.

These level crossings result in

  1. Safety risk where many incidents happen across the network yearly affecting drivers and families forever.
  2. Service delays due to incidents affecting users of the network.
  3. Slower average speeds on the network for safety reasons due to them.
  4. Capacity constraints to the rail network post-CRL.
  5. Congestion to local traffic which have to wait for trains to clear before they can move.

As the number of trains on our rail network will increase post-CRL from 20 to 36 and eventually 48 trains an hour, the problems above will become ever more apparent. We will see some change, the Sarawia St level crossing is finally being removed at the end of this year, while the Normanby Rd and Porters Ave are being removed as part of the CRL. More should also be removed with ATAP including 585.3m in the indicative budget for the level crossings in Auckland between 2018 – 2038.

We were able to obtain information about what AT/KR potentially plan for level crossings which was written about by Matt here. But no concrete announcements have been made to date, except that the southern level crossings may be dealt with as part of Transport for Future Urban Growth.

In contrast over in Melbourne, they have recognised the key importance that removing level crossings has on both the rail/road network and in May 2015 set up the Level Crossing Removal Authority. The website is also really good so do recommend taking a look.

The Authority is charged with removing 50 level crossings throughout Melbourne with 20 crossings removed by 2018 and the remaining 30 expected to be completed by 2022. The upgrades go beyond simply upgrading/closing/grade separating level crossings and at the same time they are upgrading stations making them more accessible, including better interchanges with bus/tram services as well as active modes, adding new stations, extending lines & increasing capacity through preparing for better signalling/more tracks/longer platforms for higher capacity trains.

One of the interesting parts is the Authorities preference for Rail-Over solutions where possible which they can explain in this fact sheet here, but can be summarised as;

  1. Takes less time due to no disruption to utilities, reduces the number of trucks needed and works with the water tables in certain areas;
  2. With modern building techniques, the rail-over solutions they can say can be delivered with minimal effect to services and the road network as well as retail;
  3. Preserves trees and creates more open space;
  4. Provides more space for Tram/Bus interchanges;
  5. They say with modern design in this report here the elevated structures can be designed to have less of an amenity impact as well as significantly reduce noise compared to older elevated rail systems.
Elevated Rail Noise Mitigation

The full academic report from RMIT University/The University of Melbourne which goes into the history of Melbourne Level Crossing Removals and why they believe in general Rail-Over is the best option which I recommend as an informative read can be found here. This, of course, is of interest to Auckland because the Aurecon report for level crossings in Auckland didn’t consider Rail-Over options only Road-Over, Rail-Under or a Hybrid of the two.

High Capacity Signalling and Trains

These upgrades all link well with the major Melbourne Metro project. The removal of the level crossings on the Cranbourne and Pakenham lines, the new stations along with old stations upgraded to allow longer High Capacity Metro Trains while working in partnership in delivering High Capacity Signalling (Increasing capacity of these High Capacity Trains from 20 trains per hour per direction to 30) will help tap truly into the full potential of the transformational Melbourne Metro project.

This type of integrated network thinking is needed for our own transformational City Rail Link with Matt writing about making sure we take full advantage of the CRL while building it recently here, here and here (Please give these a read if you have not had a chance). We need to be thinking like Melbourne is if we want to maximise our City Rail Link not limiting ourselves to models we know constantly under predict PT usage.

It would also be interesting to see if any new reports have been done around Auckland level crossings.

So the clear lessons from Melbourne are:

  1. The importance of future proofing and building an integrated network to maximise the transformative investment that is the CRL;
  2. Future proof the CRL for longer trains so in the future we can have our own “High Capacity Trains” project;
  3. Ensuring we have enough trains ready when the CRL opens we don’t want to be “caught short” as AT put it in the recent IPEMU debate;
  4. Consider addressing level crossings sooner rather than later. It makes sense to do as many as possible pre-CRL while there are fewer services to impact and patronage is lower. Perhaps even doing it at the same time as the disruption at Mt Eden for the CRL.
  5. More work around implementing improved signalling in the CRL so it is easier to implement higher frequencies sooner.
  6. At least consider the option of Rail-Over for level crossing removal in Auckland even if it turns out not the right solution for us in the end, there is no harm in considering.
Share this


  1. At some stage you have to consider how much of the PT budget you spend on 3 train lines. The overwhelming majority of PT users have to use a bus, but with almost all of the PT budget going to trains, the bus system is pretty much just as bad as it was 20 years ago.

    1. Not sure that is true

      1. 40km of priority lanes being currently being rolled out
      2. Double Decker Roll Out
      3. PTOM Bus Quality Standards
      4. New Network

      Still more things we can do.

        1. more routes, not just the spokes, better frequency and reliability out of peaks and late at night

        2. We need many more queue jumps at lights, more clarity on locations of bus stops (one integrated stop as opposed to 3-4 spread out over half a k). We need standards that are actually enforced, stopping drivers from blasting crap music or having the bus air conditioning set to ridiculous levels. Most importantly though, bus users need to be treated as equals to other PT users.

          Many bus stops are crap, graffitied, leaking and cold, and could easily be improved. Bus build outs and accelerated building and replacement of shelters would go a long way towards improving this. The CRL and other related infrastructure are absolutely necessary, but buses are the poor, and neglected backbone of our transport system.

          A massive improvement in level of service for bus users is doable, and for a low cost. Let’s get it done – and soon

        3. Melbournes issue was many rail lines crossed some very busy main roads or were in the centre of shopping strips, where the station was next to the crossing.
          As well they have some high frequency train lines. The end result was the gates were down for road users for long periods in the morning and afternoon peak

          Auckland may have one or two crossings that fit those criteria. My own street has a rail crossing. Its no different to having a set of lights where roads met. You get delayed for a small time which doesnt matter in the overall travel time. The trains of course are not delayed and nor should they be.

      1. Extending the Northern Busway and building AMETI are almost a billion dollars on their own. Then you can add in the NW Busway (or light rail way, which would still be a bus upgrade).

        1. Oh, and the bus network is improved by rail improvements. For example electrification, double tracking, and the CRL allow metro frequencies which allow frequent cross town bus routes to connect to the rail trunks. Auckland is at a scale where cars no longer scale, but also where buses on roads no longer scale for trunk routes. Surface buses will continue to move into distribution roles while rail and busways move into the trunk roles.

        2. Looking at London (which we are a long way from being)

          * London Underground (LU) – a record of over 1.3 billion passenger journeys;
          * Buses – a record of nearly 2.4 billion passenger journeys – over half of all bus journeys made in England;
          * Docklands Light Railway (LR) and the London Overground – on the DLR a record 110 million passenger journeys

          So surely it makes sense to spend a reasonable percentage of the total budget improving buses?

        3. Regarding pax on Tube vs Bus in London:

          – 1.2billion National Rail trips also taken between London and the South-East
          – 184million London Overground trips taken in London (that’s separate from the 110million noted for DLR)

          So overall the rail (Tube, National Rail, Overground, DLR) mode share is higher than that for buses. Which isn’t to disagree with the general idea that buses should (but don’t!) get closer to half of the PT investment budget…

    2. Once the CRL is in place you’ll have about half of the passenger-kilometres on the network south of the harbour being on the rail network. It’s the backbone of the system.

        1. There’s little point building up your extremities if your core is weak. So feel the (cash) burn

        2. For capex sure, until we start spending it on something else. After CRL (and the Ameti busway) it will be NW busway and LRT that gets all the capex.

          But day to day, it will be the bus network will still get lions share of the opex funding.

  2. What would be nice to see in the future is increasing platform also so that we can make do of a flat train and allow more people to fit in the train without the steps being a safety issue

    1. That works in principle however, its not just raising the platforms but moving the tracks so KiwiRail rolling stock can pass through

      1. Consequences of moving the tracks to avoid freight wagons hitting raised platforms are both the likely need to rebuild stations with side platforms but also increasing the horizontal gap between the platform and passenger train doors. Both can be overcome but at probable considerable expense

  3. Harriet – I think you nailed it in the first paragraph. If you ask yourself “What can we learn from Melbourne?” the answers are surely :

    1 – Longer Platforms for New High Capacity Metro Trains;
    2 – Removing Level Crossings so more trains can run safer without any localised congestion effects;
    3 – High Capacity Signalling allowing more of these High Capacity Metro trains to be run per hour;

    Keep on at AT ad infinitum, and it may sink in one day.

    1. Agree, and if you look at those three – number 2 and 3 can be done at anytime, whereas number 1 needs to be sorted during the building of the CRL or it is gone forever.

      1. Also agree. It just makes sense to remove the level crossings improve signalling and speed up services. It make sense commuter trains need to be fast. Given high speed commuter lines the entire city changes: it would change where intensification occurs, new colleges are built, what are considered up and down market suburbs. And as Jezza says much of it can be done starting now and in a somewhat piecemeal manner which is ideal fill in work between major projects (Waterview/Harbour bridge/CRL/etc) preventing loss of skilled staff and good contractors.

  4. Sure we should consider rail-over solutions to improve existing level crossings, but I’d be interested to know what the budgets are in Melbourne for this sort of work. I imagine its very expensive to put the whole thing on a new viaduct. I know from my geotech experience on some of these Auckland transport infrastructure projects that fill embankments almost always trump viaduct and bridge on cost grounds.

    For the south Auckland level crossings I think it would be quite difficult from a geotech perspective to grade separate many of the level crossings due the deep peat swamp that most of the Takanini / north Papakura area is built on. Bridges would need to be piled some 30 – 60m deep through soft organic deposits – very difficult and expensive to construct. The alternative is to have the railway on a raised fill embankment with short bridges over the roads, but this would be a major visual imposition on the area, and would cause huge ground settlement/subsidence issues for properties adjacent to the rail line – so high cost once again.

    Looking at somewhere like Morningside Drive (where the headline picture is from), previous AT studies have shown that a rail-over-road grade separation would involve raising the track from some point east of Kingsland station right through west of the New North Road over-bridge (some 2km of bridge) – again an extremely expensive project when compared to what you actually get for it.

    I’m not saying it isn’t possible – you can do anything for a price. But in several cases isn’t the cost prohibitively expensive?

    1. Sure wont work for every occasion however we should at least consider it as an option when doing options analysis for separation.

    2. Some good points Joshua. Isn’t this roughly the area where the Mill road project will go? I know that there is a flurry of building in the Takanini Ardmore Papakura area but there is also plenty of empty space – can it all be planned now with road and rail and new housing areas in a single plan?
      I was a child in East Anglia living in a railway family and my memory is of the flat drained marshland and the railways seeming to go hand in hand – maybe the Victorian builders were bolder and better than todays.

  5. Absolutely. At the very least any new platform should be built minimum 200m long but ideally 215m long to fit the entire train inside (9-car EMU) – 200m would be ok just as the ends of the EMU don’t have passenger doors so they would just fit the doors within 200m.
    If we do this then I have no doubt whatsoever that Auckland will end up with 9-car EMU within 20 years as the network frequency gets close to being maxed out. Once you get to that point then it isn’t essential to have even higher frequency and better returns can be made by increasing capacity through a 9-car EMU (which would have minimal additional running costs and still only need 1 driver etc). You get a 50% increase in capacity for next hardly any cost besides the purchase and maintenance of the additional EMU cars.
    As it is I think we will see the future express trains to Pukekohe being configured as 9-car EMU once the 3rd main and electrification to Pukekohe goes in.

  6. Should really remove all those crossing’s on the Western Line while the work on the CRL – Mt Eden station is being done. Would be expensive to do them all at once, but one big hit of pain and then one fantastic solution without prolonged chaos in the future to make changes.

    1. +1 Yes as Harriet says. Also think the cheap ones on the Onehunga line should be done by simply closing the roads (Captain Springs, Alfred, Victoria & Galway). Could easily pick up the frequency to 3 tph which is a 50% improvement on now.

  7. Melbourne may be removing 50 of the worst, but there are around 170 in the metro area so will still be many many remaining

  8. What a disappointment it is to see a golden opportunity ignored to do something about the Glen Eden station Glenview Rd level crossing . The large section where the Salvation Army shop and several other business were has been cleared and now there is frenzied building actiivity to get a couple of 10 storey blocks of 1 bed and 2 bed apartments constructed.
    With road gradient differences there could have been a road underpass or between the large empty sections either side of the rails a road overbridge that could have tied in nicely to the mall area.

    1. The problem with the Melbourne Airport rail link is it’s a bit like the Puhinui idea. There are rail lines that are quite close, but not close enough to divert an existing line. So you end up building a branch and running a special pattern just to serve one station at the terminals. The older plans to branch off the Broadmeadows line are hampered by the fact the broadmeadows line still needs to go to Broadmeadows, a bit like how you wouldn’t pull trains off the southern line to run via Puhinui. The current proposals are for a dedicated service running once the Melbourne Metro tunnel is open. Before then there isn’t network capacity to add a new line, but even then you end up running a whole service just for the one station.

  9. What about looking at an inner city ‘isthmus’ metro, and outer city rail?

    Run metro style (plastic seats, 70%+ standing, flat floor) units Mt Albert to Otahuhu via the CRL.
    20tph Mt Albert to Britomart, continuing 10tph via Newmarket, 10tph via Panmure to Otahuhu.
    Swanson to Mt Albert all stops then express to Mt Eden and terminate Grafton (use existing trains.)
    Papakura (Pukekohe) / Manukau to Otahuhu all stops then express to Newmarket and terminate Grafton (use existing trains.)
    Onehunga to Penrose all stops then express to Newmarket and terminate Grafton or Britomart (low freq service, use existing trains.)

    Metro criteria –
    Remove all car level crossings and subsequently pedestrian crossings.
    Extend the services to New Lynn and south as outer city rail lengthens.
    Gate all metro stations so it has no guards and fast automatic door opening / closing.
    Personnel at all stations and on the platforms as required.
    Distinctive metro livery for clear wayfinding and purpose.

    Mt Albert, Mt Eden, Newmarket and Otahuhu become transfer / feeder stations linking fast, high freq (3-6min) inner metro and slower, normal outer rail.
    Grafton becomes a double ended terminus, linking West with South and keeping outer city rail turnarounds separate from the high freq inner city metro services.

    1. TBH this sounds incredibly complex to operate, and confusing to use. What is the point of making everyone get off at Mt Albert and Newmarket, etc, to shuffle on to another train to go a few stations further. Why express people a couple of stops only to lose the time making them change?

      Also the network couldn’t handle 20tph from Mt Albert without a huge amount of work, and certainly couldn’t handle 20tph plus the whole western line running express on the same tracks. You’d have to quad track half the network.

  10. Interesting that you choose Melbourne as your comparison model. Melbourne probably already has one of the better public transport networks in this part of the world. And it appears that they are not satisfied yet, planning ahead for their future Metro. And yet they do not consider rail to their airport a priority. Is there something that we can learn from that?

    1. Yes and that is building lines just to the Airport is a bad idea.

      Building a line connecting 10000s of residents and jobs in the Southwest which also stops at the Airport much smarter idea.

      1. True. The public transport options in the vicinity of Tullamarine (where the airport is in Melbourne) are about as crap as they are to the industrial area in Mangere, though. For that reason, I’m fairly certain a PT line to Tullamarine airport and beyond wouldn’t just serve the airport.

Leave a Reply

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