The Premier of Western Australia, Colin Barnett, announced on the weekend that his Government will start construction in 2016 of the 8.5 km rail line to Perth airport he promised at last year’s State election.
The Forrestfield-Airport Link will branch off the existing Midlands line at Bayswater. It requires two 8 km tunnels; a new underground station to serve both the domestic and international terminals; and two new ‘park and ride’ commuter stations.
The train will operate at 10 minute frequencies in peak hour and provide a 20 minute journey between the CBD and the airport.
Airport rail is the sort of infrastructure that goes down well politically (the Victorian Government has promised one too if it wins this year’s election!) but there are a number of issues to consider with this project.
One is that in order to find funding for the airport rail line, the Barnett Government has deferred construction of the 22 km MAX light rail project until 2022. Professor Peter Newman from Curtin University says light rail should be built first and the Opposition claims it would carry four times as many passengers as the airport line.
The airport line is classified in Infrastructure Australia’s latest report as Early Stage (i.e. at the first of four stages of proving up) whereas the light rail project is at the more advanced second stage, Real Potential. (1) It’s likely Infrastructure Australia hasn’t seen completed business plans for either one yet. (2).
Another issue is cost. The line was estimated to cost $1.9 billion at the election, but that has increased by $300 million to $2.2 billion.
At $260 million per km, the airport line is much more expensive than the 72 km Perth-Mandurah line completed in 2007. The latter is often cited by advocates as evidence that governments in other State’s deliberately overstate the cost of providing new rail lines.
The Mandurah line is almost entirely at grade and cost $19 million per km (2012 dollars). It includes a 2.2 km underground section into Perth CBD that cost around $136 million per km including two underground stations.
However while the airport line is considerably more expensive than Perth-Mandurah, it’s a lot cheaper than many other tunnels. For example, the (now abandoned) 9 km Melbourne Metro rail tunnel is estimated to cost $9-11 billion. (3)
Its replacement, Melbourne Rail Link, involves constructing an at-grade rail line from the airport to the CBD and a tunnel from Southern Cross station to South Yarra station. The estimated cost is $11 billion.
The Sydney Morning Herald reported on Monday that the cost of tunnelling is high for roads too. It calculates the cost per kilometre of Brisbane’s Clem7 and Airport Link tunnels at $441 million/km and $747 million/km respectively. Melbourne’s proposed East West Link tunnel is estimated to cost $1,000 million/km.
Why it is that Western Australia seems able to build a tunnel for a lot less than other jurisdictions isn’t clear. One likely answer is that the estimate announced by the Premier on the weekend hasn’t been fully proven yet.
WA’s Roads Minister, Dean Nalder, told Perth Now that “until this goes out to tender we won’t know the accurate, specific cost. This ($2.2 billion) is our best estimate at this point”.
Early estimates are invariably overly optimistic, especially when they’re driven by politics. For example, the most recent costings for Melbourne Metro are multiples of the early estimates, but they were arrived at only after around $50 million was spent on detailed analysis and design.
There are a number of other possible explanations (see Why do subways cost so much more here than elsewhere?). A key reason though might be that much of Perth is built on permeable, sandy soils that are relatively easy to excavate.
The ABC reported last week that the new line was initially planned to have only 4 km of the line in tunnel. However the WA Government decided on tunnelling the entire alignment rather than using surface construction because the additional cost isn’t large.
The ABC understands the Government’s modelling on different routes has revealed there is little cost difference between sinking the entire rail link and sinking only part of it, and building the rest above ground.
The most important issue though is whether spending $2.2 billion on an airport rail line is worth it. It might seem to cost a lot less than it would in other States, but $2.2 billion is still a lot of money. As noted, the Government has to defer the MAX light rail project in order to find the money.
The experience with airport rail lines has not been good. Brisbane’s 15.9 km Airtrain only cost $372 million to build (2012 dollars) but has struggled for viability and for many years offered low frequencies and restricted operating hours.
Sydney’s 10 km Airport Link tunnel cost $1,251 million (2012 dollars) but fared even worse; the original proponent went broke. The current owner has a profitable business but only after buying it for around a third of the cost of building it.
Mr Barnett’s expectation of 20,000 passengers per day warrants further investigation too. It equates to around 8 million passengers a year. That’s way in excess of Brisbane’s Airtrain, which carries about 2.5 million p.a. on the back of sub 10% mode share.
It’s not that far behind Sydney’s Airport Link train which carries 12 million passenger movements a year (about half from airports, the rest from Mascot and Green Square). It has an exceptionally high mode share (around 18%) in large measure because of its proximity to the CBD and high levels of traffic congestion in the surrounding area.
According to the airport rail feasibility study done by Parsons Brinckerhoff for the Victorian Government, airport rail isn’t usually required on engineering grounds until airport passenger movements hit around 60 million per annum.
At present, Perth international and domestic terminals combined handle around 14 million passengers a year (Sydney passenger traffic is circa 39 million p.a.).
Politicians and voters love the idea of airport rail, but it’s premature at Perth Airport. It shouldn’t be a priority; there are many other transport projects that offer a higher economic and social return and should have a higher priority for scarce funding.
I originally wrote that both projects were classified as Early Stage as per Infrastructure Australia’s latest report; however I was subsequently advised that the light rail project was elevated to Real Potential by Infrastructure Australia in December 2013.
Although the deferral of MAX was announced previously, the Government’s Public Transport Plan 2031 still puts light rail in Stage 1 (before 2020) and airport rail in Stage 2 (before 2031).
The Victorian Opposition says it will proceed with Melbourne Metro if it wins this year’s election instead of the Government’s Metro Rail Capacity Project.