A few times every year we’re unfortunately reminded of the lack of action there has been on removing rail level crossings across Auckland. Removing these crossings have numerous benefits such as increased safety, reduced delays for drivers and allowing for signalling improvements which will trains to run faster. Removing them from the western line in particular will become even more important after the completion of the City Rail Link when frequencies will be able to be further increased and trains could be running in each direction every few minutes. In that situation, crossings will probably be closed more than they’re open.

Across the electrified network 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.

Yet despite the need to remove level crossings, the last ones to be removed were about seven years ago as part of the New Lynn trench construction, but there has been nothing since then. Looking to the future, Auckland Transport are hoping to remove the Sarawia St crossing and their latest board report suggests they’ve come to an agreement with those who appealed the consent so hopefully work will start on that soon. We also know that the Normanby Rd and Porters Ave crossings will be removed as part of the City Rail Link works with the latter replaced by only a pedestrian and cycling bridge. Other than those crossings, AT have previously told us some crossings will need to be dealt with as part of packages and the packages with the highest priority are: (not in any particular order)

  • Southern NIMT – Walters Road, Manuroa Road, Taka Street, Spartan Road
  • Western Line – Morningside Drive
  • Western Line – Woodward Road
  • Western Line – St Jude Street, Chalmers Street, St Georges Road
  • Western Line – Glenview Road
  • Western Line – Bruce McLaren

One of the challenges with the level crossings across Auckland is just how they might be done. Some crossings, such as St Jude St, appear to present significant technical challenges to grade separation.

Harriet has been doing a great job requesting up a storm of information recently and one of the items she received from AT was a feasibility study of grade separating the level crossings that involve roads. The report says of the 31 crossings, AT first identified the crossings that, from primarily a road operations perspective, it might be feasible to just close the crossing. They found ten could potentially be closed and a further five which could possibly be closed leaving 16 crossings to look at, although more work would likely be needed to confirm if crossings could be closed. The report doesn’t separate out what the feasible and possible closures are but includes crossings such as Fruitvale Rd, both Rossgrove Tce and Asquith Ave, and George St – which would likely be removed when the New North Rd interchange is torn down.

For the remaining 16 crossings, the study looked at three options for each one:

  1. Road bridge over rail on existing road alignment with the railway retained at its current level (Road Over)
  2. Rail trench under road with the road retained at its current level (Rail Under)
  3. Hybrid of 1 and 2 consisting of partial raising of road and lowering of rail to achieve required train clearance beneath road bridge

The study is only really a high level look at the options so doesn’t state which of the three options is preferred or even rule any options out, although based on the results, options at some locations would almost certainly be ruled out. The potential costs for each option a have been blacked out so we can’t see those. I’ve only looked at the Western and Southern Line crossings so if you want to see the ones on the Onehunga Line, or more detail about all of them, take a look at the report.

Morningside Dr – This is one of the most common level crossings that gets discussed as it’s also the one that probably appears the most frequently in the news. Option 2 of rail in a trench really seems like a no starter as they say it would require significant regrading of the rail line including having to lower Kingsland Station as well as the New North Rd bridge and the road underneath it.

Woodward Rd – Option 3 seems the most likely here as Option 1 would require raising the New North Rd intersection by 2m and Jersey Rd by 6.5m although they say it could be closed too. Option 2 here is also effectively ruled out here as it would be restricted by the Mt Albert station

St Jude St – St Jude St is one of the busiest crossings for road traffic in the country with almost 20,000 vehicles per day passing over it. On top of that, the crossing is effectively on the side of a steep hill. They say that for a road bridge to get back to ground level by Gt North Rd, it would need to have a gradient of a very steep 12.5%. It’s also worth noting that during double tracking the rail line (Project DART) was already lowered once, a missed opportunity to do things properly as part of that project?

St Georges Rd – As rail is already at the maximum gradient for freight an changes to the rail line would require option 2 or 3 for St Jude St.

Portage Rd – Any options for lowering the rail line have already been ruled out due to the close proximity of the New Lynn trench and Whau river crossing, both of which would otherwise have to be lowered. Again this appears to be something that could have, and should have been tied in with DART.

Glenview Rd – AT seem to be stuck with this crossing as no option is considered feasible. Option 1 would require raising the West Coast/Glenview Rd intersection by a whopping 8.5m, above the height of most of the buildings in Glen Eden. Meanwhile Option 2 would require regrading 1.2km of track and Option 3 would need 1km of track regraded. AT will have to look for other options here but once again, I can’t help but think this could have been done as part of DART when the whole line was being dug up.

Bruce McLaren Rd – The close proximity of the intersection to Railside Ave to the east of the crossing and access to industrial properties to the west makes Option 1 difficult while Options 2 or 3 would interfere with plans to add additional rail access to the stabling yard also next to the crossing.

Metcalfe Rd – Given all the other road connections to Metcalfe Rd on either side of the crossing, it makes Option 1 difficult while options 2 or 3 would require redevelopment of the Ranui station and have potential impacts on the ponds to the east.

Walters Rd – This appears to be one of the least difficult of all the crossings.

Taka St – Option 1 would likely require closing access to Takanini Rd. Option 2 would require Takanini Station to be rebuilt but given it’s never been upgraded, that’s probably not a bad thing. It would also require lowering the tracks at Manuroa Rd

Manuroa Rd – This is similar to Taka St

As mentioned earlier, other than a few crossings, most have no time frame for removal. ATAP identified level crossing removals as an important item and in their costs suggested spending $203m in the first decade and $385.3m in the second decade on addressing them. I’d certainly much rather we focused on these kinds of projects rather than mega projects like the East-West Link.

Share this


  1. Actually, if AT should focus on anything it should be these persistant ‘tack faults” at Britomart that occur at least monthly (yesterday being the most recent). Does anyone know what these “track faults” consist of, and why they keep happening, and how they can be fixed?

    1. AT can’t focus on track faults as the tracks belong to Kiwirail. The fault yesterday was points but they were not the points that normally fail at Britomart.

    2. Agree, although as Bigted rightly points out it is Kiwirail’s responsibility. However, AT are a significant customer of Kiwirail so should have a reasonable amount of influence here.

      1. Exactly, the implication of Ted’s pedantry about direct responsibility is that nothing can be done about anything by anybody because no one body has complete control of everything. Nonsense, while it may be inefficient, or a hassle, the way the responsibilities are structured do not absolve any party from playing their part in making improvements. And AT as Kiwi Rail’s track customer has every right and chance to insist on fixes. The fact that KR will actually do the work is a fact, but so not the issue.

        1. This is where the current rail model falls down, pre 2008 Ontrack maintained the rails but now that falls on Kiwirail and their core business is moving freight that goes no where near Britomart. The other issue is there is not really a lot of time available to Kiwirail to maintain this area even when they are willing, the main issues around points at Britomart are at the top of the tunnel near the Vector curve (allegedly caused by track slip from all the trains on the curve).

  2. I live near to the Asquith and Rossgrove crossings. In morning traffic these are both essentially used as routes to St Lukes Rd city bound. Traffic coming down Asquith Ave splits into two streams, one taking Rossgrove & Linwood while the other continues down Asquith to St Lukes Rd. The two stream of traffic converge again at St Lukes Rd & Linwood intersection. If these two adjacent rail crossings were analysed as one (which is essentially what they are), the combined traffic flows would move them up the list to one of the busiest level crossings on the whole network – and therefore a high priority for grade separation.

    1. The issue doesn’t appear to be volumes buy network function. So it seems at are saying that moment is one that could be handled by New North Rd – which makes sense

  3. The southern line looks rather straight forward.
    Church st east doesn’t cause many issues and short of supplying those handful of businesses with an alternative access route it will stay as it is.
    Spartan, Manuroa and Taka could be done as one by lowering the tracks then crossing at Taka and Manuroa but I would expect Spartan would close. Closing Spartan would require a major upgrade to the Spartan GSR intersection due to the high vehicle movements to/from the two businesses that will be caught between the NIMT and GSR.

    1. Probably a combination of Kiwirail (government) and Auckland Transport (AT). The long term savings in doing so (improved safety and transport flow) would outweight the costs.

    2. The main benefit of replacing the level crossing vs just closing it seems to be the car drivers who use the crossing, so why don’t we just toll every crossing – user pays seems like a great solution here.

    1. The trains reach 110km/have on the southern line which is faster than the snails pace that even hypetcars are allowed to drive on the nearby motorway.

      Worth thinking about who would win out of an EMU vs concrete truck at 110km/h compared to a DC or DFT vs a truck at 80kmh.

    1. Hi Brutus,
      Not sure if you’ll see this post.
      What do you think of my ideas for possible future improvements to Sylvia Park/Carbine Road area around the Eastern line?

      I came up with some doodles of what I think could help in the future.

      Carbine Road – A PT/Pedestrian only bridge linking from Panama Rd to Highbrook, providing PT another alternative route to connect Sylvia Park & Eastern suburb residents alongside Waipuna & Lagoon Drive.


      And also another Sylvia Park bus station running along the SE highway to take future pressure off AMETI running along Lagoon Dr.
      It would use the existing highway as is. Seeing as a new office tower is being built alongside the highway I reckon in future something like this could be considered?
      eg. Joining up the new office tower, SE bus station, Eastern line station, and upper level behind H&M/Kathmandu


      Keen to hear feedback/scrutiny – Good and bad

  4. Although the report you reference says that about half of the crossings could be considered for closure that is not the same as saying that all 16 can simply be closed. There are a few such as Spartan Road which have simple alternative routes and could be just closed tomorrow but others need alternative routes to be created – e.g. George Street can be closed but only when the nearby Dominion Road/New North Road intersection gets sorted (currently out for business case) or Sherrybrooke Place which needs an alternative road link to connect with Woodbank Drive (not cheap as it will require a bridge or culvert across the Waikumete Stream, but a hell of a lot more economical than a railway overpass – and all for the sake of accessing less than 20 houses – similar to Sarawia Street in that respect)

    1. Spartan Rd is not as simple as it sounds, sure there are plenty of alternative routes for the traffic on the east side of the MIMT but it is used by a lot of vehicles (many of the HMV) from both VTNZ and Halls refrigerated transport to avoid the intersection with GSR that is too dangerous to try moving that number of HMVs that would be required daily. The rumored Rangi Rd to Mahia rd link also does nothing to alleviate these vehicle movements.

      Please note I’m not advocating to leave Spartan rd open but it could be best to leave it as it is and close the Spartan Rd GSR intersection.

      1. Hangon. If the crossing serves only 20 houses, why not just leave it be. Those 20 households will just be a bit more inconvenienced as train frequency increases. If you are looking to avoid inconvenience for the travelling public, sure you would get more bang spending your bucks elsewhere?

      1. That’s what I was thinking (especially if the owners of adjacent land won’t allow a strip to be bought to allow alternative access – although maybe it would be cheaper to compulsory purchase those strips of land than by whole sections with houses on them).

  5. Your post includes a page of detailed analysis for Morningside Drive (which was included in the Aurecon report to AT which was published in 2014 and took many months of pleading to finally get a copy) but you will notice that all the financial information is redacted. While I do appreciate that there may be some commercial sensitivity, this lack of information makes it very difficult to lobby for funding. In talking with the Mayor and key councillors I have had to resort to vague numbers like “half a billion dollars” to give some scope to the kind of funding required over the first decade. Fortunately, we hear that some real numbers for a package of works are now on the table – hopefully they will make their way into the first draft of the next 10 Ten Year (2018-2028) Plan which should emerge later this year. But with a tight budget and a government unwilling to help with extra funding sources (think fuel tax) there will be a lot of hard arguing required to keep the money in the budget when there are so many alternative projects competing for a place in the queue. ~~~~~~~~~~~~~~~~~~~~
    For those puzzling over the chart: Option 1 is road over rail, Option 2 is road under rail, Option 3 is a hybrid requiring a significant length of rail to be lowered and the station completely rebuilt

  6. There have been earlier studies which gave a more definitive priority list, (1) St Jude St, (2) Manuroa Road, (3) Morningside Drive, Woodward Road, Glenview Road http://www.aucklandcouncil.govt.nz/SiteCollectionDocuments/aboutcouncil/committees/transportcommittee/meetings/transportcomagp287-36220130416.pdf
    With the number of crossings requiring separation, I think they should be done at the rate of at least one per year starting with the highest priority. Clearly the council cannot keep delaying forever, certainly post CRL there is the potential for a significant increase in train numbers, with rapid decrease in level crossing open time. This at the same time as increasing traffic from infill housing.

    The studies have not considered lowering St Jude St, to the point of grade separating through traffic from Blockhouse Bay Rd, i.e. New North Rd also drops away from the intersection, so going under Blockhouse Bay Rd and under the tracks would eliminate the grade problem. Traffic turning between St Jude & Blockhouse Bay Rd could be closed.

  7. Morningside drive is such a diabolically dangerous crossing – it seems to incorporate every possible hazard – I think a strong case be made for closing it to road trafic.

      1. I personally would close it today. Diversion through Sandringham Rd to new north is wholly plausible, and would improve overall flow. Few people taking that route are travelling to the Kingsland shops; they are transiting (sorry informal but based on observing very very few turn offs in the am before the sandringham intersection)

        It is worth saying though that there is heavy traffic through the morningside intersection especially in the am. I venture to predict that if it was closed there would be heavy lobbying to leave it open. People don’t perceive safety or people movements, they will see an out of the blue requirement to travel differently, sorry.

  8. I live near the St Jude crossing which is also less than 100m from the platform at Avondale Station.
    Getting rid of this crossing is essential but it will be no easy job. Building a bridge over it just isn’t possible because of gradients. I can see only two options:
    1. Trench the railway lines, this is not going to be an easy accomplishment as it would require a large amount of service downtime, dropping the elevation of the train station, overhead wires and the tracks for a good few hundred meters on each end. Then there is the added problem of the Chalmers Street crossing not too far away, which I’m guessing would also need to be grade separated for this to work.
    2. Close the crossing permanently.

    1. St Jude – The patron saint of lost causes, kinda ironic name for a level crossing really. It would be likely those 3 would be dealt with as a package with St Jude’s closed or separated with the other two just closed.

      1. Wont improve the through traffic as you are limited above and below by heavily congested intersections. Doesnt seem to have any real safety concerns. Went for a look on AT Traffic count data and St Jude 5 day ADT is 14,800, so 15,000 is the figure not 20K.

        And ranking all the road counts above 15k we have about 1100 locations throughout the city. I would saySt Judes Rd isnt all that busy,

        1. Hat-tip to you for highlighting AT’s published traffic counts. The 20 k appears to be St Jude St + Chalmers St + Donegal St which provide alternate routes to reach the top of the Blockhouse Bay Rd ridge (no right turns at the top of St Jude St). I guess that if you grade separate St Jude then the steeper Chalmers St (~12% grade) would not cost much more to include.
          I had a look through the priority list vs the latest traffic counts: that places St Jude’s 3rd (14.8 k, 5d ADT) after Walters Rd (17 k) and Railside Ave/Bruce McLaren (15.6 k). In your opinion, are there any priority grade separations?

        2. I dont know every crossing well but safety has to rate some places higher than others, Morningside is the clear winner here. There may be others

    1. With the Onehunga ones that can be just closed, which I notice is three in a row straight out of the station. As I understand it, the driver has to manually activate the Galway St signals before pulling out of the station as its so close. Some easy wins there to be had.

      1. Drivers don’t and can’t manually operate any signals, they are set by TC in Wellington when the proceed signal is given to the driver.

        1. Poorly worded on my part, what I meant was there is an additional delay on leaving Onehunga as the train can’t depart until the Galway St crossing is activated and clear.

        2. Like anywhere else on the network where drivers must get a signal from TC before proceeding, it is the only staring point on the network where early signals (before departure time) can’t be given though.

  9. Some lateral thinking required. Close St Judes St crossing – New North Rd traffic can turn right into Blockhouse Bay Rd then use Rosebank Rd to reach Great North Rd, or turn left into Blockhouse Bay Rd and then use Wolverton Rd to reach New Lynn.

    1. Those alternative routes you have listed are already heavily used in the peaks with a lot of right-hand-turn congestion. After the Waterview Connection is open the local traffic situation there will become a lot clearer.
      In defense of lowering the tracks: It would allow a very good PT bus-railway interchange to be built by linking Crayford St East and Layard St Nth for Blockhouse Bay, Rosebank and New North Rd bus routes. All of these were proposed to be FTN routes by 2023.

      1. As a frequent driver of St Judes Rd, its not a priority. Really the congestion at the New North Rd intersection above and that at the Avondale roundabout below means you probably at capacity in the local network for the peak periods now.

  10. Sherrybrooke Place could potentially be extended to the east to connect with Waikaukau Rd. The land below the railway line here has been cleared over recent weeks, presumably the developer is going to replace the footbridge over the stream with a road bridge, would be ideal to acquire a strip of the “Fat Cat Backpackers” land between there and Sherrybrooke and connect the two roads.

  11. Removal of the crossings would cost shit loads ,my Question is where is the money going to come from to do this. There are other ways of highlighting the crossings . I have promoted a concept which got brushed aside which in comparison to the cost of complete removal of the crossings is bloody cheap. The concept was take on board in the USA late last year and they are now making the rail crossings with the concept that was put to Kiwi Rail .

  12. Well this is interesting. I commented on Transport Blog over a year ago to the effect that the level crossings in Auckland were a problem and should be got rid of, to enable a far smoother operation / timetable / lessening of crashes etc. But (from memory) I got shot down by some of the commenters here who maintained that there was nothing wrong with the level crossings and it wasn’t a problem and that i should pull my head in and not make such outlandish comments.

    Glad to see that i was right all along and that you now all want to do exactly what I was suggesting. Hooray!

    Seriously: cars and humans crossing railway tracks in busy cities really don’t mix well.

  13. The Glen Eden situation is crazy. A main road for Glen Eden, Oratia and parts of Henderson which drives right through the Glen Eden shops and frequently gets backed up in the morning waiting trains to come across. As the timetable stands the trains are perfectly off-sequence (when on time) meaning the barriers come down every 5 minutes.

    Technically it wouldn’t be difficult to lower the existing track but like New Lynn it would be a long trench so big $$ and big disruption.

    A bit of foresight when double tracking would have been nice.

    1. The road which has the barriers is just a local road , Glenview Rd. The main rd you talk about Gt North Rd doesnt cross the rails here. and can continue to flow while the trains are passing. Like all suburban shopping strips they arent designed for large amounts of traffic and thats not going to change, nor should it.

      1. I suggest you:
        A) look closer at the role of Glenview in the network – Waikumete Cemetery means Glenview is a key link from a large catchment south of it towards Kelston, Te Atatu, Henderson and Sh16 westbound. Local status is irrelevant to its actual role in this case.
        B) observe reality before pronouncing. West Coast Road does indeed get affected by the crossing, and carries a reasonable high volume in peaks.

        What I find interesting about this study is the lack of looking at alternative crossing options. While there are none practical for many crossings, Glenview is one where other options are possible in the blocks to the east and west, due to the existing street grid running parralel to the rail line with short stubs and other intersections on WCR suggesting options to hook in.

        1. Every intersection is slowed when you meet another road, have a road running through a shopping strip slows it down again. Its not all about the single passenger driver getting to his work quicker and saving 45 secs.
          Glenview Rd may well serve a large area but matthew couldnt understand why he was held up along Gt North Rd. Perhaps he needs to join a support group for those who cant live with the idea that congestion cant be solved, it just moves.

        2. Matthew didn’t mention GNR – his reference to Oratia is clearly about WCR.

          I’m not sure you’re on the right lines here. The crossing at Glenview has multiple impacts, many of them on PT as well as vehicles. Plenty of buses through GE now, and they all get caught up in the traffic. As GE is an effective interchange for quite a wide area, the train and bus impacts merit close attention to this intersection and rail crossing.

    2. A possible Glenview Road solution is a road overbridge. With the old park&ride carpark not yet redevolped and on the opposite side the vacant commercial buildings (ex sallies on corner etc) between Glenview Rd and the new Park&Ride (isn’t that KR owned land?) a regraded up approach from just where the firestation is and curving west and up the the required overbridge height over the tracks look possibles. The actual overbridge being about 100m west of the current crossing. Then the old P&R land plus whatever is needed eastwards in front of the old station building as the ramp ip to the bridge.
      Opportuniy too for lots of parking next to station on north side of Glenview to present crossing.
      Also either a pedestrian tunnel like that at Kingsland as the ped crossing next to the level crossing is a deathtrap just waiting for a victim

      1. …pedestrain tunnel at present crossing location or move that completely unused ped bridge from east end of station to west end with high fences to stop track runners..

        Fruitvale rd level crossing just gets closed and a pedestrian overbridge installed

      2. At Glen Eden I would expect an even underpass to be the better solution, perhaps to the west forming a intersection opposite Glendale Rd, thus connecting with the bypass already in place on the south of the shops. The park and ride should over of course be developed and not wasted as car parking so close to the station, as the whole strip would improve massively with the reduction in traffic from the removal of both the through traffic and carparks. Main drag could then by narrowed, traffic slowed, incentivising through traffic to us the bypass and to improve place experience for retail, new developments and people.

        1. Another option would be to extend Waikumete Rd westward, take it under the railway line between the Caltex petrol station and Singer Park, and connect it to West Coast Rd, then close the Glenview Rd level crossing.

  14. It seems conceivable that new cars will be able to actively prevent crossing in front of a train relatively soon. Not self driving, just another assist like Mercedes and the like already build in for a lot of scenarios. Some more substantial and better designed barrier arms would help in the meantime. Is this not the ‘operation lifesaver’ solution?

  15. Sarawia St: What is the reasoning for building an expensive bridge to service just a few houses, when it would be waaaay cheaper to open up the walkway between Laxon Tce and Furneaux Way?

    1. As it turned out there are some people in Furneaux Way that have too much money and successfully fought the council over that option.

      1. iirc to get the required width to open up the walkway, required purchasing one of the apartment buildings. With 2014 CV of $1.4m-$1.7m per apartment, would cost over $10m to buy and demolish.

  16. Looking at the map, there are a lot of crossing that are close to each other. It would be very inefficient for the train to wait on every crossing.

    It is possible to setup the crossing gate sequence so that the gate opens and closes in sync when the train is approaching, so the train does not need to slow down much?

    1. They already do but the problem is that so many of the crossings have platforms near them so the train has to stop anyway and people look at the stationary train before driving/walking around the barriers not checking for the train from the other way. Morningside is a prime example of where trains cross during the peak so when a Britomart bound train is stopped at the platform you can bet there will be another coming from the other way before it moves.

  17. I’ve often wonder that where rail crossing (have to) exist if a change to the type of signalling wouldn’t help improve their safety.
    If, rather than the red falling light, they replaced them with conventional traffic lights.
    Most people understand and react to conventional traffic lights where the flashing red rail crossing lights are less recognised, especially in an urban setting.
    Sure use crossing arms in conjunction with the lights, but change the lights to conventional lights.

Leave a Reply

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