Another of the interesting items at the Auckland Transport board meeting yesterday was about the plans for removing the remaining four road level crossings on the Southern Line in Takanini at Spartan Rd, Manuroa Rd, Taka St and Walters Rd. The board were asked to approve a business case for route protection of their proposed solution for dealing with the crossings.

The level crossings see the highest number of train movements through them compared to others in the region, with AT’s passenger services passing through as well as regular freight and the odd intercity train too. They claimed that with all the rail traffic, by 2030, it is expected a train will pass through the level crossings every 2.5 minutes and longer term there are plans to add two additional tracks to the line through here.

The Walters Rd is also the busiest level crossing for vehicles, with recent traffic counts showing over 18,000 vehicles a day. That’s more than most state highways outside Auckland, and a significant increase in usage over the last decade or so thanks to all the greenfield development nearby.

In the consultation last year it was proposed to grade separate two of the crossings and close the other two but replace them with a third grade separated crossing.

Following the consultation, it appears the plan is now to also replace the two closed crossings with pedestrian and bike bridges. So five crossings in total but what’s particularly concerning about all this is just how much it’s expected to cost, likely coming in at a whopping $650 million. All crossings are expected to be bridges.

Three new multi-modal grade separated bridge crossings, two new active mode grade separated bridge crossings and two consequential road closures are recommended at a cost $436 million (P50) to $647 million (P95), of which $144 million is associated with property purchase.

There is a breakdown of costs per crossing too.

These costs are massive, especially when you consider that the Newmarket Crossing, which replaced the Sarawia St level crossing, was completed in 2018 and cost only around $8 million. Even the active mode bridges here will cost four times that, not to mention it’s hard to understand why you would also need $22 million in property acquisition for them.

One clue as to why they might be so expensive can be seen in the images showing what the Walters Rd crossing might look like – huge and spatially inefficient.

The paper to the board notes that the Papakura Local board support the proposal but also

There is opposition from parts of the community (i.e., Takanini Business Association and Mr Wallace – Takanini Village Limited) about the impacts of closing some level crossings to traffic and preference for an underpass rather than bridge option for grade separation due to visual impact and urban design reasons. Councillors and local Members of Parliament have also raised these concerns on behalf of Mr Wallace and sought deferment of the recommended option and further investigation of underpasses.

AT say that while and underpass is preferred visually, it is much more complex and disruptive to build due to the peat ground conditions and having to deal with the rail line so as a result, is estimated to cost at least twice as much.

Notably they do say that:

the final design, cost and timing of the solutions for each crossing will be determined via the Auckland Level Crossings Single Stage Business Case (SSBC) which is due in late 2023.

As well as the cost, another aspect I find hard to understand is how they claim that by making it easier, less disruptive and more reliable to drive that it will result in a significant increase PT mode and reduction in vehicle travel. It would be great if it happened but it sounds very much like green-washing and manipulating the model to get a desired outcome.

  1. Modelling identifies there will be an emissions reduction of 7,090 tonnes per year and 182,800 tonnes reduction over 40 years compared to the baseline network. Route protecting the project further contributes to emission reductions by enabling improved rail service, improved safety for all users including walking and cycling thus increasing the attractiveness of these modes leading to positive mode shift. The projects could result in a daily reduction of 134,900 vehicles kilometres travelled.
  2. The level crossing upgrades will have positive contributions to the following Transport Emissions Reduction Plan (TERP) outcomes through a reduction of reliance on cars and supporting people to walk, cycle and catch public transport; and facilitating work to begin to decarbonise heavy transport and freight.
  3. The details of the exact contribution to TERP will be determined with the final design as part of the Auckland Level Crossings SSBC.

I didn’t watch it but I understand the board agreed to proceed with route protection for four of the crossings with further review of the Walters Rd one.

Finally, a few other final thoughts:

  • Why is the pedestrian access to Takaanini Station not included in this programme?
  • Given the costs, should the Taka St crossing also be closed and replaced with an active mode bridge?
  • If costs like these are replicated elsewhere, it’s going to be hugely expensive to remove the other rail crossings on the network, such as the 15 on the Western Line. Perhaps we need to have a harder conversation about potentially just closing some – perhaps in association with a discussion about creating low traffic neighbourhoods in some of these areas.
Share this


  1. Infrastructure building in NZ is getting out of hand… I’m not sure exactly who but someone out there is seriously taking the Mickey with these outrageous build costs!
    Partly it is stupid and/or overbuild designs. Partly it is not keeping it simple and basic. Then there are the pointless design elements (nobody cares if a bridge or underpass has designs/art on them especially in industrial areas).
    Overseas many of these would simply be completed as follows:
    Precast large culvert sections ready to go, close half the railway line at a time, excavate it and the space below, install said culverts, replace rail above, repeat.
    Drainage can then be done and lights etc after the fact. Whole thing built within a week for under $5m.
    For bigger roads additional bridge type supports may be needed to support the rail above but that’s also east enough.
    Something is seriously wrong with this whole picture.
    Time to bring back the MoW.

    1. It is completely out of hand thanks to scope creep and overcomplicated design, you’re suggestions are great, and that type of thinking should be rolled out ASAP. The decision to have things done more simply relies on the client to allow it, the contractor will simply go ahead and build things if they are allowed to. The problem rests with the client and the designers, not the contracting model.

      However, a MoW with in-house project construction will have 0 commercial interest to finish on time and under budget like the status quo, so it is the worst idea out there if you want to decrease project cost.

      1. You have geotechnical problems regarding building on peat and issues around acid sulphate soils in this location which complicates investigations, design and build programmes. Acid sulphate soils may require special stormwater treatment and liming of soils before disposal. I suspect these issues make this site more expensive than other sites

      2. Captain Slow – “MoW with in-house project construction. . .is the worst idea out there if you want to decrease project cost”. I totally disagree.
        The flawed ideology that private companies are inherently more efficient and all government employees are lazy sods has cost us dear since it took root in the 1980s. Teams of experienced and dedicated public servants delivered incredibly successful construction-work in the past, while private sector involvement has more-recently led to some spectacular budget blowouts. Incompetence can exist anywhere. Believing that private companies are automatically free of it, and that their need for profit will somehow guarantee a cheaper and better outcome, is generalising ideology over evidence.

        1. So long as you only count the benefits and don’t count the costs then the MoW did a great job. Ignore for example the big pile of reinforcing steel they ordered for the Benmore dam which sat in a pile rusting because in the meantime they had redesigned it as an earth dam. Or maybe the cost of building the Newmarket viaduct which appears to have been designed by a railway engineer with no idea trucks and cars don’t need a flat carriageway. Or maybe the Grafton Road overbridge designed with a W support because that was the designers initial.

        2. Hi miffy. Note that I said, Incompetence can exist anywhere. Many MOW projects proceeded without the kind of bloopers that you cite. Meanwhile, take a look at the Transmission Gully if you need an example of a private sector cost-blowout.

      3. “a MoW with in-house project construction will have 0 commercial interest to finish on time and under budget”

        Commercial models are only one way to get performance. Dissing the public service is an ideological thing.

  2. The Board decision on this followed a messy pathway.

    Having heard the residents’ representative present a rather convincing piece of analysis for changing to an underpass at one location, they should have quizzed the two men who’d come to present the topic (presumably AT officers or consultants). There may have been an excellent response to the residents’ analysis. Perhaps they could give more details about why an underpass wouldn’t work. Had the residents downplayed the risk of flooding or the impact on underground utilities? Perhaps an underpass could work in this location but there were benefits to keeping a consistent approach for all the five crossings? Perhaps quizzing them would reveal they hadn’t been resourced to really look into it enough.

    But they dismissed the men without questioning them.

    So on the basis of hearing from the residents only, the Board first decided to have the item deferred until the June meeting while the subject was looked into.

    Then they thought that this was too short a time for a proper review, so they decided to make it July.

    Then they realised there wasn’t a July meeting so they decided to make it August.

    As the two officers(?) approached the door to leave, the Board suddenly thought to check whether this would suffice, with the reply being that no, an application had to be lodged before then.

    So the Board then decided to split the five, proceeding with four and discussing the fifth again in August. This meant that if the review showed up that it should indeed be an underpass – it would be too late to determine if the same thing applied to all five!

    At this point the WK CEO pointed out that funding was in a package for the five, and [some process / funding reason that put the funding at risk for the fifth one].

    My impression: the Board is not sufficiently focused on understanding issues before making decisions, and doesn’t put proper questions to the people present, which would allow more information to surface.

    And this is without getting into the subject of whether there was a far cheaper approach more in line with the strategic direction of mode shift and safety.

  3. This is disappointing. I have been campaigning for a committed programme for level crossing removal since 2004 but progress has been very slow and incremental. You cite the example of Sarawia Street but this was actually trivial compared with others since the railway was already in a pronounced defile leading to the Newmarket tunnel so construction involved a relatively simple bridge over the tracks – in almost all other cases lengthy approach ramps will be required plus land purchase to provide access to affected adjacent properties. The obvious solution was to simply close Sarawia Street and provide alternative access to the few dozen affected properties from the west off Basset Road but AT was legally challenged and eventually (over 6 years later) forced to build a bridge at Cowie Street and a connecting loop road which was completed in 2018 at a cost of about $8 million. A few other level crossings have been closed as part of station development (e.g. Kingdon Street and Porters Pass) or grade separated (such as those around New Lynn station or Normanby Road where the tracks were partly lowered as part of the new Maungawhau Station development.
    Real progress seemed to happen when the ATAP was refreshed in 2021, including an allocation of over $200 million for “CRL Day One” projects, including replacement of the four Takanini level crossings with a couple of grade-separated crossings. For the first time it looked like Auckland finally had assured funding to remove all level crossings on the Southern Line. But after several years we now hear that the budget is far too small and all they are doing at this stage is “route protection”.
    We can only look with envy at Melbourne which has been showered with billions from the Victoria State government to get on with the job (they started with about 170 level crossings) removing dozens of crossings every year. They have had the luxury of being able to lower or raise kilometres of track on some lines so that the actual crossings are not only much less expensive, they do not require lengthy approach ramps and so are also much less disruptive to the surrounding communities.
    So further delay with the Day One crossings is very disappointing, and begs the question of when a start might occur for the Western Line crossings.

    1. When you see the scope of the proposed bridges, it really does the beg the question of whether raising or lowering the track would be a cheaper option, removing the need for bridges or underpass, or reduce the cost and scope of the infrastructure required.

      1. The reason raising the track is discounted is because this is a combined passenger and freight corridor and will eventually need to be quad tracked making this option, especially with stations, hugely expensive. Unlike in Melbourne where twin suburban passenger only lines being elevated has proven to be hugely beneficial creating new public realm and removing severance.

        Trenching also has this problem plus the additional problem of swampy conditions making going under on either mode expensive and difficult. Essentially the water table means any underpass wants to effectively float back up, which is surmountable but only at considerable expense.

        So because this is the NIMT, and its busiest section, plus the flat and swampy geotechnical conditions road and or people over is really the only option.

        Do the designs need to be that crude, ugly and overbuilt. Surely not.

        On the basis of those visuals this programme looks like it needs more creativity and value for money focus.

        1. “Do the designs need to be that crude, ugly and overbuilt. Surely not.”

          If, for technical reasons, it has to be road over track at the current track elevation, then I think so.

          What you have to understand about this section of Walters Rd is that you have shops right up to the rail line on the northern side of the road and on the southern side, the western side of the NIMT has some kind of upskilling facility right up to the track and the eastern side has houses… right up to the track.

          If you look at the design, it has the access road loop going underneath to serve the commercial and light industrial premises on the western side, whilst they’re evidently just buying up the houses on the eastern side of the NIMT. The shops on that side do have a driveway entrance (only, so no exit) to their carpark which is being lost. Of course, you can still use the main entrances to that carpark which are off Walters Rd. This solution isn’t really viable on the western side of the NIMT, because the premises closest to the track on both sides of Walters Rd. all have their own individual access points rather than sharing a carpark.

          You can see that the access road loop would be built even if they did the underpass… the only meaningful differences are on the eastern side of the NIMT. The guy that’s mentioned as objecting is from Takanini Village Limited and that’s located on the eastern side of the track (see link at end). And from their perspective, the underpass is clearly better.

          So, the options here are basically buy out the businesses or build the bridge.

        2. I am familiar with the location, I think the images on your link is on the other side of the tracks. This double loop road is on the western side and looks far from optimal. Especially as it clearly involves buying land, better i think to buy whole sites design new access, ie re-masterplan the whole area, and sell the balance with better less land hungry access.

          Of course this would mean higher up front cost but much of that would essentially be bridging finance, being recouped later. So how is this done elsewhere? With a city owned development agency, ie Eke Panuku, but properly funded and actually understood by its owner.

          Having transport and landuse planning siloed from each other is one of the many suboptimal things we do here.

        3. Agree Chris. Looking at the images in the post they will have to buy enough land off the businesses that will have a significant impact on the way they operate. Why not buy the lot?

        4. Sure, like I said, either you build the bridge or you buy the sites.

          But I’m not sure this has a major effect on them… it looks entirely like they’re just losing their car parks, which as we know don’t generally mean that much to businesses.

        5. The Melbourne skyrail project has shown very little in the way of forward planning for additional track capacity. Regional trains doomed to follow frequent all stopping suburban services. Surely some of the trenches could have been dug a few metres wider. Having said that the amenity created underneath is fantastic.

        6. The Melbourne lines are not all passenger only, e.g. the Frankston line, with a variety of level crossing treatments including rail viaducts.

  4. Indeed, their proposal misses the fact that both Takaanini AND Te Mahia stations are only accessible by pedestrian level crossings. Apparently station access and safety is of no value at all. The whole thing is disappointing, and I didn’t see them mention that there is no funding in the next 10 years for this anyway (which is what I was told when I provided resident feedback).

  5. One wonders if, for these prices, it would be cheaper to elevate the track given that, according to KiwiRail, the track has to be dug and rebedded anyway.

    No, wait, it costs $250m to build a station on the ground and you might even have to rebuild Te Mahia into an elevated station, too… I’m not sure how much of a run up the trains would need.

    1. How to remove Takanini level crossings:

      1) Build a new twin-track elevated viaduct parallel to the existing line.
      2) Divert rail traffic onto new viaduct.
      3) Build a second twin-track viaduct next to the first.

      Result: Grade-separated four track. This wouldn’t be particularly cheap, but I’ll wager you could do it for $650m. It would address a number of issues with the existing railway (e.g. severance) and provide a long-term increase in rail capacity.

      1. Don’t forget you’d also have to rebuild Takaanini and maybe also Te Mahia train stations, so that they could function as elevated stations and my original point was that we can’t seem to build train stations for less than $250m.

        1. Yes. I agree that the cost of stations in NZ is ludicrous. There should be a design challenge for anyone who can come up with a model for above-ground stations thats fulfil all necessary requirements at a lower cost.

          This would provide the opportunity to insert the (long-delayed) station at Walters Road.

      2. That”s how they rebuilt the Newmarket Viaduct. But it won’t be cheap in the Takanini area because the bridge pilings would have to be dug down to bedrock.

  6. For that kind of money, could they build out a proper motorway junction at Alfriston Road?

    Takaanini Industrial Area traffic could then come and go via Porchester Road, while the current level crossings all get much cheaper active modes connections, probably underpass-in-a-box-culvert.

    I know, I know, severance, blah, blah. But if that’s a problem with the railway, why wasn’t it a problem for the motorway?

    1. In this area, the motorway severed farmland from housing… in fact, they’re still completing the sprawl on to the western side of SH1 as we speak. As far as I can tell, the oldest housing development was built contemporaneously with the motorway.

      What’s actually going on is that Papakura was systematically gutted of its shops and everything moved to Takanini 10-15 years ago. The situation is that you’d be taking a road that already gets quite congested (Porchester Road) and adding a lot of traffic from a road that gets horribly congested (the Great South Road) to it.

      There were plans for another train station as recently as, iirc, 2017 to go right in this area, incidentally.

        1. That sounds like a proposal to replace the current Takaanini station. I was talking about a proposal to build an additional station by Tironui Station Road:

          which is undated but is apparently based on this review from December 2014:

          I don’t know when they stopped talking about it.

        2. *that is, I couldn’t confirm it was recently as 2017. I don’t remember where I found the second link but I remembered it as being from 2017.

  7. Silly question – can’t pedestrian and cycle crossings be level crossings? Barriers to prevent people crossing when trains are there, but easy crossing when not.

    Then sort out a simpler option for two bridges:
    – Is it a stormwater mitigation requirement to have the planting/gardens?
    – Is it cheaper to purchase entire sites rather than take frontages and provide access to them?
    – Can the property access ‘roads’ be a one way narrow street or laneway?


    1. Pedestrian level crossings are an unnecessary safety hazard. They also force all trains to slow down when passing them which makes it a pain to get through a stretch of a couple kilometres with several level crossings.

  8. Also thinking ‘Hey, It’s a little under 5km between Myers Road bridge and Subway Road, what’s the big deal? Just drive around.’

    It feels like there was a shortage of forethought when they developed the Takanini area, or an assumption that the railway would be dead by the time it mattered.

    1. Shortage of forethought? In the development of Takanini? The ARC came up with a growth strategy based on building a new station and growing a new centre around that. From memory it included a new underpass at Glen Ora Road and closure of Walters Rd. ARTA was onside. The Papakura District Council got onside and rounded up land owners who worked together to come up with a structure plan. Then some doofus in the rail said their diesel trains wouldn’t be at full speed from Papakura to Takanini and that might be a nanosecond less efficient so they vetoed a new station. That left a stack of commercial and residential development designed around a non-existent station that can only now be reached by car. Forethought you say?

      1. Apologies, I had no idea!

        I can see how Glen Ora underpass makes sense, why didn’t they do that regardless of the station?

        In fact, why didn’t they look at bridging or digging all the available connections while surrounding landings were uncommitted? Not like there was a lack of precedent…

    1. Funny you should say that. The Board’s performance yesterday was so inadequate and negligent – on climate issues – that it filled me with a kind of hope.

      I’ll write about their extreme intransigence soon, but my conclusion is that the Board’s bold refusal to follow Council and Government direction has now reached the point that commissioners must be brought in.

      1. yes please – bring on the commissioners

        the numbers bandied about to get minorish infrastructure, suggest a massive disconnect from any reality i share. TERP is toast at those prices.

        ps – To ride a bike near Great South Road is to die.

      2. Hi Heidi if you can write and update us on the terp aspirations versus current performance including board would be valuable- how are we going on our aspiration to cut ghg 50% by 2030, increase W&C 700%, PT use 550% etc or whatever the targets were. At least EV’s, low emission cars are having higher uptake though noted this will not fix congestion.
        So! Where do we go from here, not cutting ghg 50% as per accepted ipcc science is ethically abhorrent and morally bankrupt if it is within our power to do so.
        Thanks for being an island of sanity in an ocean of….

        1. Hi Tim Tam, thanks for your comments. Progress on all those things has been hindered…

          Details in today’s post.

  9. Realist in the comments above – has got the right attitude, but sadly the wrong costs. $5m and a week’s worth of work? Absolutely not, but the methodology is in the right direction.

    There was a project in Lewisham in the UK when I was there in the 1990s, where they built a giant concrete culvert besides the tracks, and then closed the railway for one weekend only, and pushed / jacked the culvert across the rail intersection, so one line could cross over the other. You can read about it here:
    and watch a video about something similar here:

    Seems to me that this project could be delivered a lot quicker, safer, and for a lot less money than the $650 million they have budgeted for it. Could I get the job if I offer to do it for half the price?

    1. Yes, but what were the ground conditions. Most of Takanini was a peat swamp. NZ has major problems with soft swampy and/or slip prone ground associated with a rainfall double that of London.

      1. The article says many have been undertaken in very difficult ground conditions. This is from the summary: “The authors have presented a highly effective and well proven technique which enables engineers to construct underbridges beneath existing infrastructure. Undoubtedly the success of the technique is due to the ability to control ground movements, which is at the heart of the Ropkins System™. Table 1 illustrates the full range of projects constructed using this technique. Most are located in cities and all have avoided disruption and minimised environmental intrusion. Many have been undertaken in very difficult ground conditions necessitating sophisticated ground stabilisation measures, including grouting, ground water lowering and artificial ground freezing.”

  10. Insider Engineering knowledge (as my father is a retiring Structural Engineer) states that bridges are less expensive than tunnels. A true cost analysis would show that roads require much more expenditure for maintenance than a rail line (although we are fighting decades of neglect with our rail network so that would affect proving that rail with bike and pedestrian accessible stations via town centres is the solution from the past and for the future). So if we wish to address this ridiculous spend more to keep the capitalist growth based system spinning, we should look to the actual best solution for health, the environment, and therefore the people. That solution would be: reduce the number of crossings for private motor vehicles, perhaps to two in the interim. We must work towards the end of the private motor vehicle and little steps such as making it marginally less convenient to cross a rail line would be a more saleable solution. Infrastructure costs the natural world so we must be more selective about it in order to respect our natural environment. The automobile is reaching it’s end of acceptable life stage, if we wish to preserve our own fragile existence; so take a deep breath, near some trees, and stop making the private motor vehicle the first priority. We quite literally have enough buses and train carriages for Africa in this city and if we had an integrated public transport system like the serious cities of the world; this roundabout, back and forth, waste of breathe that is AT and it’s consultation process; we could just have some qualified town planners and public transport experts to design the greater city to justify those greenfield expansions that have reduced out arable land for feeding our future!!!

  11. $650 million!

    The question is how long will take to complete the project given our track record we seemed take twice as long to complete the project.

  12. A couple of pedestrian/cyclist underpasses beneath the railway were recently installed at Petone and Alicetown as part of the Petone-Melling cycleway. Also, another one went in at Trentham Station on the Hutt Valley Line. These were installed by the method outlined by Realist above, and required a block-of-line for a long-weekend only. See KiwiRail video here:

    1. The extreme costs of urban sprawl just go on and on. Why would any political party sign up to this because it is just such economic nonsense?
      I accept that a number of families want a 3-4 bdr house on 600m of land, but if that’s what you want then surely it is only reasonable that you pay for it? If someone was told that to live on the Hibiscus Coast you have to pay a $6 toll to use the Northern Motorway extension and another $8 for Penlink and $4 for the park and ride at Albany then suddenly living in a 3bdr apartment in Mangere might seem like an attractive option. Auckland might then be able to afford a rapid, connected, affordable PT system.

  13. It certainly begs the question of whether raising or lowering the track would be a more affordable option, eliminating the need for bridges or underpasses, or reducing the cost and scope of the infrastructure necessary when you see the extent of the proposed bridges.

  14. Why don’t we just live with the level crossings at least it preserves the road grid layout. Better access to the station for passengers don’t worry too much about trucks and cars. After all the trains have right of way.

  15. Silly suggestion to just leave level crossings as is. With current frequencies the barriers are down for about a third of the time (it varies from crossing to crossing.) After CRL frequencies will be able to increase (ultimately to double) so barriers will be down for well over half the time, creating grid-lock at peak times. This is quite apart from the safety issues.

    1. So all the traffic which could previously cross the railway line on the street it originated on now has to be funneled onto one or two roads which now have to be bigger to cross a bridge so a journey which was simple becomes a major expedition whether your on a car or bike or foot. Thousands of Tonnes of carbon intensive concrete needs to be poured and then we have major disruption for all modes. How many rail bus journeys must be endured to build these monstrosities’ which have being costed at $640 who knows what the final price will be. All for a bit of frequency and we are going to have 9 cars so we probably have the capacity without increasing the frequency .
      My suggestion is that AT should publish capacity figures for their trains. So we had x number of seats available and y passengers. It would be a useful statistic to have when making these sorts of decisions.

      1. I am wanting to see the plan in its entirety from Papakura to Takanini Motorway on ramps.
        I am not convinced directing more funnelled traffic from Walters Rd is the answer. In my opinion in observing ever increasing traffic on Walters and Great South Rd’s a good percentage of this traffic is heading to or from major arterial routes and at times is choked with static traffic.
        I agree there needs to be change, but lets have a cohesive plan between the upgrading of public transport and the need to move private transport efficiently, with the expected increase of both requirements planning & action for the future is necessary now.

  16. Level crossings also increase journey times. Trains literally crawl out of stations such as Morningside because of safety concerns – one fatal and 2 serious injury accidents at the adjacent level crossing in the last decade (2 of them pedestrians not cars). With 15 road level crossings (not to mention the pedestrian only ones) this adds minutes to the journey time on the Western Line which is a further disincentive for users. Some of the traffic delayed at level crossings includes buses.

Leave a Reply

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