I must say I’m a tad surprised I haven’t dedicated a post to the topic of Airport Rail before, and it is certainly worth a bit of analysis and discussion. The Auckland Regional Transport Committee considers rail to the airport to be the equal sixth highest priority transport project in the region – which isn’t that high, but it’s certainly something that is likely to “go into the mix” some time in the near future. Furthermore, the RTC considers that ensuring there is a route protected for airport rail is a very high priority.

There has been some background work done into the feasibility of a railway line to the airport, with ARTA commissioning a study of the whole southwest rapid transit corridor in 2007. After much analysis, the preferred option was for heavy rail that would connect to both Onehunga and Puhinui. The corridor also included the construction of the Avondale-Southdown line, although it doesn’t require that line to be constructed to work (in fact, I would imagine most trains from the airport would go straight to the city, although it would be useful to run some Henderson to Manukau services via the airport). Here’s a map of the preferred option:option-2d In terms of whether or not it’s viable to the build the route, this study did make some investigations into the expected number of passengers who would use the line in a “2041” situation. Now I never quite understand how this kind of thing is figured out (and I’m generally highly suspicious that matters like peak oil are excluded from the calculations), but anyway, the level of demand over the Mangere Bridge during the two hour peak period was anticipated to be around 3,260 during the two hour peak period. At the northern access to the airport, anticipated demand was 2,850 and from the eastern access around 1,400. Now to me that doesn’t seem like a lot. Let’s say a train can carry around 500 passengers reasonably comfortable – that would only be around 6 trains (in both directions) worth of passengers within a two hour time period. In some respects I think to be viable, you would probably be looking at needing at least twice that number – enough to justify four trains per hour in each direction at peak times I think.

But here’s where those previous issues like peak oil and other unknowns come into play. The study talks about them in a bit more detail on page 28:

The demand modelling and externally estimated air passenger demand was estimated on the same basis as in the ART/APT models. The models assume the real cost differentials and other perceptions that influence transport choice behaviour remain as at present. Taking a long term view, there is significant uncertainty that cost structures will remain unchanged either because of resource scarcity and long term international market demand, foreign exchange movements, Government policy regarding the pricing and supply of transport and inputs such as fuel, changing land use intensity and feedbacks between transport infrastructure supply and land use, and changing personal preferences and behaviour. A current example of such uncertainties is current movements in global fuel prices and increasing
awareness of potential “peak oil” issues.

An estimate has been made of the sensitivity of the mode split calculated in the ART/APT model to changes in several influencing factors on overall travel demand, the concentration of demand in the RTC corridors and on the relative attractiveness of modes. The sensitivity has been carried out for typical 10 km car trip. This estimate has used modelled inter-zonal trip numbers, times, distances and fares for a random sample of OD pairs and using the ART mode split model to analyse sensitivity.

It’s interesting the effect that these matters end up having on predicted demand. After running a reasonably comprehensive analysis of what different factors might contribute to changing travel patterns in the future, the study makes the following conclusion:

The overall conclusion is that by 2050 there could easily be a doubling in PT mode split from a combination of factors that increase the price and constrain the demand for private transport.

Now that is a rather swooping guess, not that I necessarily disagree with it in any way. And this has a big effect on matters. By doubling the modeshare of public transport, we would end up with patronage forecasts of around 6,500 passengers over the two hour peak. While this is still less than the (approximately) 8,000 passengers that would ensure full trains (in both directions) running at 15 minute frequencies over that time, one much consider that this total is not altogether dissimilar to the number of passengers entering Britomart during the peak hour at the moment – and that’s certainly not considered to be ‘under-used’.

In terms of cost, the full option shown in the map above is certainly not cheap – as shown in the image below:

This cost does include the Avondale-Southdown section of railway line, which I think could well be built separately (and is quite possibly only really justifiable for freight purposes). So that knocks the price back to around $1.4 billion. Nevertheless, I can’t see $1.4 billion magically appearing for this line any time soon.

In my opinion, while the airport line is an important future part of our public transport network, we have other stuff that needs to be done first. An airport line is useless without the CBD rail tunnel, because you’d be feeding far too many trains into Britomart than it could cope with – so there’s no point even really considering building the airport line until we’ve at least started work on the CBD rail tunnel. However, that doesn’t mean we should write this off. At close to 7,000 passengers during the peak period in 2041 you’re certainly not going to running empty trains along the line – while often these passengers will be “high-value trips” (in that shifting them to the airport more reliably and more quickly could have a big economic benefit) particularly for time-constrained business-people. It might also lead to a heck of a lot of money not needing to be spent on taxi fares all the time. What we should be looking to do in the near future is to sort out a final alignment and get the designation process going – so the route can be protected and so that it is fairly simple to – once it is justifiable – build the line.

Share this


  1. Why not build the Airport to Puhinui line first? The route seems to go through fields mostly (from what I have seen from a plane) so relatively easy/cheap to construct. And it would hook up with the existing southern line to the city, until the other line north from the Airport through Mangere can be built.
    I still have good dreams about the airport line in Hong Kong zooming me to Hong Kong Island in 20 minutes.

  2. $75m per km? Why is everything so ridiculously expensive? Surely track, rock and sleepers aren’t that expensive. It never used to cost $75m per km to build rail.

  3. According to the back of my enevelope, a fully grade separated rapid transit line from Penrose to Manukau, (excluding the Avondale-Southdown) route would require:

    4km of duplicated and upgraded track in the Onehunga branch corridor
    8 level crossings closed or grade-separated on the Onehunga branch

    700m long rail bridge across the harbour

    14km of new electrified twin track in the motorway corridor or across open ground
    1.2km of tunnel or elevated track in the airport precinct

    2 new major rail interchanges, possibly grade separated (upgraded at Penrose and new at Wiri)

    5 rebuilt motorway interchanges
    4 more (at least) grade separated local road crossings

    4 new suburban stations
    1 new airport interchange, preferably integrated with the terminal itself
    3-5 expanded suburban stations (dual platforms on the OBL, plus maybe new platforms/track to Manukau station and Puhinui)

    At least one new stabling yard and maintenance facility

    Once you spell it all out, you can see the costs involved. I reckon building the bits of new track would be one of the cheapest parts of the whole exercise, its all the work around them that costs the big bucks.

  4. You don’t need a new bridge, they’re building one for that purpose 😉

    We’ll have to requisition it first, of course….

  5. Nick the new Mangere Bridge duplication has capacity to handle a rail track, IIRC you can largely thank the Greens for pushing Labour on that one… Might just be a single track though…

  6. What are the chief origins and destinations for trips to and from the airport? That’s what is needed to determine if this makes sense at all. I don’t know the answer, but I suspect it is highly diverse. The origins of morning business travellers are their homes, so you can figure out what proportion would find the generalised costs of catching one (or more) trains are compared to taxis. Tourists are heading for hotels, or just heading out on coaches or hired vehicles, so again the proportion of hotels within a reasonable catchment of walking or taxi from stations is needed.

    What do the Sydney and Brisbane airport rail schemes tell you about what not to do? Bearing in mind Sydney has a far bigger airport and 4x the population, and an already well connected and integrated rail system. In Brisbane, Airtrain gets 5% of trips. http://www.bne.com.au/files/pdf/2009%20Preliminary%20Draft%20Master%20Plan/Chapter12_SurfaceTransport.pdf

    Bear in mind that Airtrain was a PPP operation, with fares operating on a profit oriented basis, and while it nearly went bankrupt, it has since engaged in massive cost cuts and patronage has increased.

    Presumably it would make sense for a similar connection in Auckland to be funded privately, so the risks are born by the private sector and the high value business users actually pay the cost of running it (since they are saving $50 + taxi fares, rail fares of $10-$20 should surely be possible).

  7. I think a PPP would still require a designation and that is what needs to happen soon or Airport Rail goes the way of North Shore Rail or good PT in East Auckland or previous electrification or all the other missed opportunities…

  8. I’m not so worried about the designation after the start of SH20A apart from the new runway, it’s the way through Mangere (and second track through Onehunga) I’m more worried about…

    If it is done as a PPP, the airport could pay for the development on their land with a percentage of the ticket going to them, say an $10 ticket from the airport station with normal fare pricing stating at the Montgomery Rd and Southern line stations…

  9. I’ve just realised that the line shown above does not make sense. I’d be opposed to having a Sydney type pricing structure (imposed to protect the private monopoly), but one thing they do get right is having two stations. The railway should go past both the domestic and international terminals. It may have to go under the second runway. Unfortunately even with current delays in construction by AIAL, it may be too late to get a tunnel built before, rather than after, the runway is built, and save money. The second runway is being built to accommodate domestic traffic, and free the major runway for large aircraft.

  10. George, once the second runway is built the domestic terminal will be brought into an enlarged international terminal. Effectively there will no longer be a separation between the two – so one station will be fine I reckon.

Leave a Reply

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