The preferred option for a Doncaster rail line

The 2008 Eddington study concluded it would cost $1.8-$2 billion to construct a rail line connecting Doncaster Hill in suburban Melbourne to the existing Hurstbridge/Sth Morang rail line in the vicinity of Collingwood.

Last year, a study for a group of eastern suburbs municipalities led by Professor Peter Newman from Curtin University contradicted Eddington’s finding. It asserted the line would only cost $0.84 billion.

Then last week the government released the stage one results of the in-depth feasibility study on Doncaster rail it promised during the last election campaign.

The consortium commissioned to undertake the study comprised SKM, AECOM, Aurecon and URS. It looked at a range of options, including various tunnel configurations.

It concluded a 12 km line running within the median of the Eastern Freeway from the Doncaster Park & Ride facility to join with the Hurstbridge/Sth Morang rail line in Collingwood is the best option.

Its preliminary estimate is it would cost $3-$5 billion. However there are two important caveats.

First, the proposed line stops at Doncaster Park & Ride. The team says another $1 billion would be required to tunnel to the activity centre at Doncaster Hill.

It’s a steep grade and would require a deep and expensive tunnel/station. The increase in patronage would be relatively small compared to the additional cost.

Second, the team says there is insufficient spare capacity on the Hurstbridge/South Morang line to take Doncaster trains.

The Hurstbridge and Sth Morang lines would need to be separated, or “decoupled”.

A new dedicated tunnel from Clifton Hill to carry South Morang services direct to the CBD would therefore be a necessary part of the project. The estimated cost of the tunnel is $4-$6 billion.

So the total cost of the study team’s preferred option (excluding the extension to Doncaster Hill) would be $7-$11 billion.

Finding funds on that scale would be extraordinarily difficult (it’s comparable to the cost of the proposed Melbourne Metro), but what matters most is how that outlay compares to the expected benefits.

The team hasn’t undertaken a benefit-cost analysis yet (that’ll be done in Stage 2, apparently), but there are some telling conclusions we can draw from the work done to date.

One is that virtually all the patronage on the new line would come from the Doncaster Area Rapid Transit (DART) bus service and, to a lesser extent, from other nearby train lines. In other words it would mostly replace existing forms of public transport with another.

The reduction in car use attributable to the line would be minor. The reduction in traffic on the Eastern Freeway would include buses no longer needed but would also be minor (and would in any case be replaced by induced traffic).

Two new intermediate stations would be built at Bulleen and Kew, but they’d be located in the median of the Eastern Freeway.

They’d have little potential for walk-up patronage or for transit-oriented development. They’d primarily rely on park & ride and feeder bus services.

The travel time for Doncaster residents would reduce on average from 35 minutes to 25 minutes. However those who’d use DART as a feeder service to Doncaster Park & Ride (i.e. most peak-hour passengers) would incur transfer penalties in changing over to train.

Even without a Doncaster rail line, the modelling done by the study team indicates public transport patronage will increase significantly in the region.

For example, transit’s mode share for trips to the City of Melbourne (all purposes) from Manningham municipality is expected to increase from 45% in 2011 to 60% in 2031; from Maroondah and Whitehorse municipalities it’s expected to exceed 70% by 2031.

Moreover, car use (kms of travel) in the base case is expected to increase by only 5% in the region over the period, compared to 50% for the metropolitan area as a whole.

Thus the benefits of a new rail line seem modest relative to the cost. The Public Transport Users Association (PTUA) argues, however, that the costs are over-stated.

It says there’s no need to build a new tunnel for Sth Morang services, since the capacity of the existing Hurstbridge/Sth Morang line could be increased by at least 50% by upgraded signalling.

A failing of the study team’s report is that it’s very short on detail and so it’s hard to independently evaluate all the claims (unlike the excellent documentation released on the Rowville rail study).

Even so, the study team looked at this argument and rejected it. The team says its costings were done from bottom-up engineering fundamentals by professional cost planners.

When it comes to technical issues, I’m more disposed to accept the view of a team of expert engineering and transport consultants than that of an advocacy group.

But even if the PTUA were right, the portion of the line from Doncaster to Collingwood still requires an investment of $3-$5 billion (or $4-$6 billion if it’s extended to Doncaster Hill as the PTUA seems to want).

That’s still a massive expenditure relative to what appears to be a modest pay-off.

The PTUA has an answer to that too. It reckons a shorter line from Bulleen to Collingwood could be constructed within two years, as an immediate first step, for $100 million. Yes, for $0.1 billion.

That’s about 75% of the length of what the study team is recommending yet it would cost only 3% of the latter’s estimate! That’s extraordinary.

It even makes the $0.84 billion estimate of Peter Newman’s team look excessive. It suggests the PTUA’s number is manufactured for political purposes.

The Danish economist Bent Flyvbjerg has pointed out the tendency of transit agencies to under-estimate the cost of new infrastructure. I suspect transit advocacy groups might take that predisposition to a whole new level.

There are better alternative investment opportunities for public transport than building a new Doncaster rail line.

The funds might be spent, for example, on eliminating level crossings, upgrading system performance, or providing better bus services in the suburbs to connect to the existing rail network.

Or the Sth Morang rail line might be extended to Mernda in the northern growth area. It’s further from the CBD than Doncaster, is growing much faster, has fewer local employment opportunities, and may ultimately have a larger population.

Residents of the Doncaster area deserve good public transport. Fortunately, they’ve got the new DART system, which is experiencing strong patronage growth. There’s scope to give DART buses further priority over cars on the road system at considerably lower cost than building a new rail line.

The study team will now move on to preparing the stage two report, with more detailed costings and economic analyses. On the basis of the stage one report, Doncaster rail does not appear to be a wise investment.

Our cities need investment in public transport. It’s vitally important though that it’s spent on sensible projects that really deliver, not devalued by populism or wasted on political pork-barrelling. Just being rail doesn’t necessarily make a project a wise investment.