Airports & aviation

Nov 16, 2015

Is it high time Melbourne got a rail line to the airport?

64 of the world's busiest airports are served by heavy or light rail but Melbourne Airport, which ranks 50th, isn't. Wisely, the Victorian Government has no plans to build a rail line yet

Alan Davies — Editor of The Urbanist

Alan Davies

Editor of The Urbanist

An airport train like this (driverless) one in Dubai would be nice to have in Melbourne…but there are higher priorities for scarce public funds


Last month the Public Transport Users Association (PTUA) revived calls for the Victorian Government to get behind the idea of an airport rail link.

Melbourne Airport is going to keep growing – it’s time to get moving on providing a mass transit heavy rail connection…It’s high time our world city had trains to its airports.

The key new piece of evidence is a spreadsheet compiled by the PTUA showing that 83 of the world’s busiest 100 airports are expected to have heavy or light rail services by 2021. (1)

The irony is if the PTUA had run this survey at the same time last year, Melbourne would’ve been on the list as one of the “going to” cities because the Government of the day went to the 2014 election with a specific plan to build an airport rail link.

But like the promised East West Link motorway, the idea went down the plug hole when the electorate threw the Napthine Government out of office.

The new Premier, Daniel Andrews, made it clear soon after taking office that airport rail is not a priority for his Government. (2)

No one for a moment says that an airport rail link is not a worthy project. But be very clear: the services that people use every single day are my priority.

I’ve discussed the issues with the airport train idea before and it’s worth reiterating the main points (see also Is it time Melbourne got a rail line to the airport? for more detail).

I think Mr Andrews is right. Most Melburnians travel by air only once or twice a year at best and get there by car or taxi. That’s not surprising given over 90% live more than 5 km from the CBD and 90% live in households with at least one car.

Residents who use the airport frequently tend to be business travellers who take taxis or drive to or from home; they don’t pay for parking or taxis out of their own pocket.

The main public transport options (particularly important for visitors) are taxis or the privately operated SkyBus at $18 one-way between the airport and Southern Cross rail station.

SkyBus offers a 24 hour, seven days a week service, mostly at 10 minute frequencies, with interchange to the metropolitan rail system at Southern Cross.

While a mass transit system will be required one day if airport passenger numbers continue to grow, SkyBus has ample scope to scale up capacity in the near to medium term by increasing frequency.

There are other much higher public transport priorities at this time for scarce public funds; they include the Melbourne Metro tunnel, elimination of level crossings, extension of rail to Mernda, duplications/quadruplications of existing track, improved maintenance, and upgraded signalling.

A train would offer a more comfortable ride than the existing SkyBus service, but it would cost circa $2-3 billion for the lowest-cost route recommended in the former government’s feasibility study.

Something else would have to be foregone or delayed so the money could be found.

A train would replace a public transport service that costs Victorian taxpayers nothing (i.e. SkyBus) with one that wouldn’t cover any of its capital costs.

A train wouldn’t reduce congestion on the Tullamarine freeway any more than building new roads would; the road space liberated by any motorists who switch to rail would quickly be filled by other drivers.

Nor would it reduce the cost of parking at Melbourne Airport or the airport’s revenue from this source; that’s the result of anti-competitive practices, not the absence of alternatives to driving.

The environmental benefits from building a train would be very small because most passengers would come from other forms of public transport. Put another way, it would be an enormously expensive way of reducing emissions and pollution.

The two existing Australian airport lines – Brisbane and Sydney – were privately financed but both struggled financially in their early years despite costing much less than a rail line to Melbourne Airport would i.e. just $0.4 billion and $1.2 billion respectively (3)

Tickets wouldn’t be cheaper than SkyBus either. It’s virtually certain the government would insist on charging circa $18 one-way as is the case with the Sydney and Brisbane airport rail services.

Building airport rail because other cities do is a poor argument. Many cities built them for reasons of prestige and politics, not because mass transit was justified by demand or operational considerations.

It’s the sort of logic that partly explains why so many cities built inner city freeways in the 1960s and 1970s.

The key advantage of a rail line at this time is that it would provide a faster trip than SkyBus during Melbourne’s (landside) peak periods i.e. 30 minutes on average vs up to 40 minutes for SkyBus. (4)

But because it would follow an indirect route, a train would be slower in the (landside) off-peak i.e. 30 minutes on average vs 20 minutes for SkyBus. (5)

SkyBus could be made faster in the peak by giving it a dedicated road lane at considerably lower capital and operating cost to taxpayers than building an airport train.

In fact, the current widening of the Tullamarine-City Link motorway offers an opportunity to provide a High Occupancy and Toll (HOT) lane from the airport to the city centre.

A HOT lane gives vehicles with (say) three or more passengers, as well as motorists willing to pay a higher toll (the amount varies with demand), the right to a dedicated lane so they enjoy a faster and more predictable journey time. (6)

If the (private) owner of the airport wants to build an airport rail line now, it should pay for it. Meanwhile, the Government should investigate how it can improve the operations of SkyBus (no, it’s not perfect).

The rail projects already on the Government’s plate will collectively account for a generation of funding even on optimistic assumptions. On top of that the Government needs to deal with a host of operational problems as well as increase the quality of bus services, particularly in the outer suburbs.

Costly rail projects that largely replace one form of public transport with another – and I include proposals for rail lines to Doncaster and Rowville in that category – should be given a low priority.

The appropriate action for the Government to take at this time is to reserve the alignment for a future airport rail line (already done a decade ago) and get on with implementing projects that really make a difference.


  1. Includes 7 airports with bus shuttles that connect to nearby rail lines.
  2. Mr Andrews also argues that there won’t be sufficient capacity to accommodate an airport rail line until the Melbourne Metro line is up and running, expected to be circa 2026 at the earliest.
  3. Costs inflated to 2012 $$. The initial Sydney operator went broke. Not surprisingly, the current operator is doing very well; it purchased the asset from the receiver for circa $300 million.
  4. Although note Skybus claims the average journey time in (landside) peak periods is only 30 minutes on average
  5. Peak landside congestion periods don’t necessarily coincide perfectly with peak flight periods (airplane passenger arrivals and departures)
  6. HOT lanes are also a way of introducing motorists to the idea of congestion charging.


Leave a comment

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

15 thoughts on “Is it high time Melbourne got a rail line to the airport?

  1. hx76

    If the Airport train starts at Southern Cross – No.
    If the Airport train starts at Caulfield – Yes.

    Stopping: Caulfield – Richmond – CBD – Footscray – Sunshine (pick up Bendigo, Ballarat & Geelong V/Line passengers) – Airport.

    Service usable by all passengers (not just Airport passengers) for no extra charge so that all Melbournians can enjoy rapid cross-city train journeys. Airport passengers to pay a surcharge getting on or off at the Airport station itself.

  2. Tony Morton

    mook schanker #13: Yes, examples abound of airport railways that clearly weren’t designed with mode share as an objective. Step forward, Heathrow Express – and the Sydney and Brisbane links at least in their original PPP incarnations.

    But the fact that things often get built for the wrong reasons doesn’t mean we can’t plan and build something to achieve an objective most of us agree is desirable, as the Dutch have done with the Amsterdam-Schiphol link. The advantage for us is we can see what’s wrong with all those sub-par examples and avoid their mistakes.

    Meanwhile, the reason we reserve easements for transport corridors well in advance of construction, as we’ve done for Melbourne Airport rail, is precisely to avoid the high capital costs associated with land acquisition, road works and service relocation. If these costs weren’t avoidable we wouldn’t bother reserving easements in the first place.

    We have a problem when our planning mindset so readily dismisses rail projects as too costly despite their popularity, while churning out destructive follies such as the East West Link (which is still being sold to eastern suburbs residents as an airport travel solution despite being far more wasteful than even the most pessimistic scenario for airport rail).

    Skybus leaves people behind at the terminals even today. I don’t think an aspiring world city can afford to let the car/taxi share of travel to the airport continue at the 80%-plus level for the foreseeable future.

  3. mook schanker

    Tony – I’m not suggested detailed business cases, just a compelling well thought out view. I understand PTUA are volunteers so fair enough.

    I’ve been to plenty of airport heavy rail lines that are a waste; terrible patronage, no relationships to any integrated transport network, incredibly high capex & opex and were mainly built for political prestige. Melbourne has more pressing infrastructure projects and hence clear prioritisation of projects is they key. If money was infinite, yeah sure, go build a high capacity Tulla line….

  4. Oz (Horst) Kayak

    The PT service between the Airport and Central City edge (Southern Cross or North Melbourne Station) should be based on a maximum reliable travel time of 20 minutes. There are several options possible to achieve this

  5. Alan Davies

    Aenveigh #9:

    Where did the boost for trains in Zurich come from i.e. at the expense of which other mode? Were there other changes made at the same time? Growth numbers are for marketers; we need to look at the change in mode share.

    Those reasons for rejecting a HOT lane on the Tulla sound like: “we can’t do it because the Minister thinks it’ll be politically unpopular”.

  6. Jacob HSR


    Ted Baillieu was elected for promising to build a railway or 2.

    The Doncaster one or Rowville or the Airport one.

    The Tories sacked Baillieu and hoisted the East-West money-pit upon the voters.

    Then for the 2014 election they re-promised to build a railway!

    The voters promptly kicked out the Tories.

    Vic spends $500m every 10 years to host an F1 race to attract more tourists. Now, how are the tourists meant to get to the city from the airport?

    The airport railway is for tourists that are coming in increasing numbers, not for locals who go to the airport once per year.

  7. Aenveigh

    re #6: I think Zurich’s experience was a 50% boost replacing buses with trams at the same frequency. So mode can be relevant, even though ‘in theory’ it should be subordinate to other factors such as frequency and price.
    Of course, capital/recurrent costs relevant here, light rail cheaper to run as volumes escalate but larger upfront cost (which is even higher for heavy rail).
    I also thought I read in the paper that a dedicated T3/Skybus lane wasn’t implemented not because of the Transurban concession, but because VicRoads claimed the ‘lane-weaving’ to get in and out of it would reduce overall freeway throughput (not stated if the T3/bus lane would be an overall faster trip for those vehicles though, which may obviously represent a much greater percentage of *people* on the freeway than vehicle numbers).

    (found article: relevant text below with link to full story)

    The state government also investigated tolling the Tullamarine Freeway all the way to Melbourne Airport from 2018 and putting in “high-occupancy lanes” for buses between the airport and Bulla Road, but rejected both proposals.

    The investigation was done at the insistence of Infrastructure Australia, the advisory body of the federal government, which is contributing $200 million to the project.

    But Mr Williams said VicRoads found that tolling motorists would be unfair, and putting high-occupancy lanes on the freeway would create excessive weaving and merging near entry and exit ramps.

    Read more:

  8. Alan Davies

    Tony Morton #7:

    If you kept their fare at the same premium level but made it inclusive of a day’s unlimited travel within the metro area (equivalent to what you get with a day ticket on the London Underground) I dare say mode share would expand rapidly

    That is an excellent idea now that integration with myki is apparently going to be implemented. It’ll mainly appeal to leisure visitors but should be popular; it could add a percentage point or two to SkyBus’s mode share (at the expense of taxis).

    *SkyBus gets an increase in patronage at the same revenue per ticket.
    *PTV gets an increase in patronage.
    *PTV incurs low marginal costs, esp since a lot of use would be off-peak, but presumably foregoes some revenue.

    Travellers get a trip between airport and SX, a myki card, and (say) a day’s unlimited travel on train, tram and bus for their $18 one-way (could be made to work in both directions, but likely to appeal mostly to inbound travellers).

    Has to be potential for SkyBus to share some of its ticket revenue with PTV.

  9. Tony Morton

    Alan at #6: I suspect the real reason so many are driving or (especially) being driven to the airport is that they’re deterred by the idea of getting a train into the city only to negotiate the maze at Southern Cross between the suburban platforms and the bus terminal, then pay $18 for a bus ride that’s going to be less comfortable than a taxi and always carries the risk of getting caught in traffic anyway.

    They’re so deterred by this many are prepared to pay upwards of $50 in taxi or parking fees to avoid it, even when they’re the only one travelling. So they already have a price signal; a realistic congestion charge isn’t going to tip the balance.

    The problem with Skybus isn’t with the service considered in isolation – it’s the lack of integration with the rest of the public transport network and the failure to offer a quality of service comparable to what’s on offer when travelling from the city to most other suburban centres. Sydney and Brisbane do appear to provide this, but their problem is the high fare isn’t commensurate with the degree of service provided in return – the system charges a premium fare but then puts its hand out again as soon as you want to go anywhere but the CBD. If you kept their fare at the same premium level but made it inclusive of a day’s unlimited travel within the metro area (equivalent to what you get with a day ticket on the London Underground) I dare say mode share would expand rapidly.

  10. Alan Davies

    Tony Morton #5:

    You can aim for any number you like but simply replacing the bus with a train isn’t likely to double mode share c.f. Brisbane’s airport train has a mode share of around 10%. Reducing the fare significantly would win some mode share but suppressing driving somehow is the only way to win a really big increase. Of course, those actions would also improve SkyBus’s mode share without the need to build a rail line.

  11. Tony Morton

    Marcus W #4: 23% is quite a bit better than 10%, but I’d think Melbourne could aim at least for Schiphol’s 35% (as of a decade ago), given the latter has a rail link operated as part of a broader suburban/regional network with cars and taxis as the main competing mode. This is more directly comparable with the context for Melbourne Airport rail than Hong Kong is likely to be.

  12. Marcus W

    What proportion of mode share would an airport rail link end up receiving? In Hong Kong their ‘gold plated’ rail service only carries 23% of travellers, with public buses being preferred by 47%.

  13. Tony Morton

    mook schanker #2 – Developing detailed plans and business cases should be the job of the government through PTV: the PTUA is a group of community advocates, mostly volunteers. We ultimately rely on others’ expertise to ensure that Melbourne has the chance to emulate international best practice when it comes to public transport network planning.

    The point really is that so many other cities, many of them smaller than Melbourne, have managed to make rail to their major airports work. We’re increasingly looking like an outlier, particularly given our size which ought to translate into superior financial capacity.

    Skybus does a very good job of carrying 10% of airport travellers, but it’s not likely to scale up much further, particularly since the State Government can’t even talk Transurban into giving it dedicated lanes on CityLink. And even with the service offering it provides, it leaves 84% of airport travellers using private cars or taxis. That’s got nothing to do with innate preferences or geography and everything to do with the failure to provide a high-capacity link to the largest single destination outside Melbourne CBD that’s integrated with the rest of the network.

    We even have a planning easement set aside for Melbourne airport rail that ought to mean the cost is well below that of similar projects elsewhere, which usually require huge outlays for land acquisition, roadworks and service relocations. Network constraints in the short to medium term before the Metro tunnel is in place are largely overcome by terminating trains at Southern Cross platform 8 and extending the dedicated RRL tracks an extra 2km from Sunshine Junction to Albion Junction.

    It’s worth it in energy/carbon terms too even if the trains run on brown coal or diesel (though it’d better justify the outlay on fine new rolling stock to purchase renewable electricity for it). The PTUA crunched the numbers on energy and emissions by various transport modes some years ago:

  14. mook schanker

    It’s a shame the PTUA are in the spruiking business rather than getting stuck into the analysis and coming up with a compelling case.

    In regards to emissions Sky bus vs airport heavy rail, it would be interesting to see how the figures stack up. I wouldn’t be surprised if a brown coal powered train with inherent network feed losses would produce more emissions or at least be fairly close…

  15. Jason Murphy

    It’s like being at the toy store with children.

    *If you have this toy you really want, you can’t have this other thing you really really want. Now choose carefully.*

Share this article with a friend

Just fill out the fields below and we'll send your friend a link to this article along with a message from you.

Your details

Your friend's details