NSW Premier Nathan Rees has his hand out for billions in federal funding for Sydney public transport projects, but the gurus at Infrastructure Australia should be asking some hard questions because something seems terribly wrong with the cost of rail construction in NSW.

I was in Perth recently, so I took a ride down to Mandurah on the city’s new southern rail line. It’s a 72 kilometre, 45 minute trip to the coastal growth centre where the trains are met by specially fitted shuttle buses that whisk you to the waterfront precinct in five minutes. It was all very nicely done and the engineering standards are high. As a longstanding and very frustrated Sydney public transport campaigner I could only envy Perth’s achievement.

A few days later I had a couple of chardonnays with Alannah MacTiernan, WA’s shadow minister for strategic infrastructure who was, while minister for planning and infrastructure in the last Labor government, responsible for pushing the project through in 48 months.

“Congratulations Alannah! Spiffy line,” I said. “What did it set you back?”

“About $1.3 billion,” she replied.

Either I’d misheard her, or she was losing it. Conditioned by the NSW experience, I was certain nobody could build rail that cheap.

Later, I got Alannah to dig out the final cost figures, and it turns out that the Mandurah line actually came in for $1.22 billion. The Sandgropers got a double track rail line for $17 million a kilometre.

This was a revelation. I was reminded of a remark by former NSW Transport Minister John Watkins when EcoTransit Sydney met with him in August last year: “One thing I’ve learned in this job is that everything to do with rail is a lot more expensive than you think it’s going to be.”

Indeed. When the go-ahead for NSW RailCorp’s proposed 13 kilometre line to Sydney’s south-west growth centre was announced in June last year (characteristically, it’s since been indefinitely deferred) the estimated cost was $1.36 billion — more than the price of the Mandurah line. How could a simple 13 kilometre project over unchallenging greenfield terrain with just two new stations, a stabling yard, reconstruction of Glenfield station and connections to the Southern Line possibly set the taxpayer back $106 million per kilometre?

In Perth they got 72 kilometres with 11 stations (two underground, and most with well-integrated bus interchanges and hundreds of park-and-ride spaces), a stabling yard, two major water-crossings plus more than a kilometre of tunnel under the city and the project came in for just $17 million a kilometre.

It wasn’t as if it was all plain sailing for the WA government. During the course of the project the cost of steel nearly doubled, the mining boom sucked technical skills out of Perth, a savage industrial dispute delayed construction. Not surprisingly, initial costs blew out. On the evidence of this comparison, if Sydney’s proposed South-West Rail Link had been constructed for the WA Government, it would have cost under $250 million.

It is also instructive to note that the South-West Rail Link estimate is only slightly below the $1.5 billion paid for Sydney’s M7 orbital motorway, which is 40 km long, with many bridges and off-ramps and the enormous Light Horse Interchange. This project traversed the same geological environment as the South West Rail Link and involved at least as much earth-moving per kilometre.

In NSW, rail construction projects fall under the Transport Infrastructure Development Corporation (TIDC), a government instrumentality charged with designing and delivering public transport projects. TIDC works for the Minister of Transport and for some years, rail bureaucrats and even politicians have joked that the acronym really stands for Triple Initial Development Cost.

When an addition to the NSW rail network is contemplated, TIDC works up a basic design for the project and figures out a cost for construction by the private sector based, presumably, on recent experience. To that they add a very generous allowance for “contingency”, a margin for ‘risk and financial risk’ and their own fee for managing the project.

At this point the government triumphantly announces how much it’s going to spend on the project and things get murky. The inflated estimate skews benefit-cost analysis so that the project appears uncompetitive with motorways — an outcome neatly coinciding with the dominant philosophy of state governments from Greiner onwards. The forces within government traditionally opposed to rail argue vociferously against it. In the uncommon event that the project survives this assault, the tenderers structure their bids to reflect TIDC’s inflated estimates. Even if the project is cancelled, its estimate becomes the baseline for the next estimate. It takes little imagination to see that this process, continuously repeated, would put costs on an escalator.

Government secrecy has concealed the problem. To understand the extent of the apparent overestimations, those without access to information confidential to TIDC, the Ministry of Transport and the NSW Treasury can only compare rail projects with similar characteristics in other states and with similar NSW projects from the recent past.

Of course, there are variations related to the technical differences between, for example, WA’s narrow-gauge heavy rail (which we have used for comparison here) and NSW’s standard gauge systems. Geology and topography can also be relevant to costs. Nevertheless, these differences could account for only a small proportion of the huge discrepancies.

Here are a couple more examples of apparent egregious overcosting:

The Parramatta-Epping link

The Parramatta-Epping link (an integral section of the Parramatta-Chatswood Line indefinitely deferred by NSW Transport Minister Michael Costa in 2003) was to have been slightly over 11 kilometres long. Five kilometres were on the surface, along the route of the existing single-track Carlingford Line, and this section required some earth works to accommodate double track. The rest of the project consisted of a total of 6 km of tunneling in two sections. It was last costed at $2.2 billion.

What should it really have cost? It is possible to arrive at a rough (and generous) estimate. Assume that the existing surface track, with its wiring, signals and bridges was completely replaced with new double track and all necessary fittings. On the evidence of the Mandurah Line’s $17 million per km (which included significant engineering challenges and incurred a 25 per cent cost over-run) reworking the existing section of track should cost no more than $100 million.

The cost of tunneled sections can be adequately estimated by looking at two comparable Sydney projects and by assuming that the costs for these were not excessive. The Airport Line came in at around $800 million 10 years ago. There were five stations on this line and it was a state-of-the-art engineering through wet sand most of the way. This project was 8 km long, so a cost of $100 million per km would seem robust.

More recently, the now indefinitely deferred North-West Rail Link, Stage 1, from Epping to Castle Hill, was costed at $660 million in November 2006. This was to have been a conventional heavy rail link (not to be confused with the North-West Metro proposal, also indefinitely deferred). Curiously, TIDC’s estimate for North-West Link seems reasonable. The project was to have been 9 km long, all in tunnel, with three stations. At $73 million per kilometre it would have been rather cheaper than the earlier (and more technically difficult) Airport Line, but to be conservative let’s use the Airport Line’s $100 million per km in calculating a reasonable cost for the tunneled sections of the Parramatta-Epping connection.

A two-level underground rail-bus interchange at Parramatta Station is a feature of the Parramatta-Epping link’s design, however the Airport Line, which we have specifically used as a cost comparison, featured five stations — four of which were underground, including the Domestic Terminal station whose construction was especially challenging and involved a major cost over-run — so an allowance for Parramatta interchange and underground platforms at Epping is built into the analogy.

On this basis, total cost for tunneling for the Parramatta-Epping link would be $600 million. Add $100 million for the surface section and you end up with a total of $700 million for the whole project. The TIDC estimate is over three times this figure.

Metro West proposal

The current “Metro West” proposal (Parramatta to Central) has been submitted for federal funding with an estimated total cost of $8.1 billion. The rationale for this project is that it would increase capacity between Parramatta and the City, which is what the now-truncated Parramatta-Chatswood line was intended to accomplish. Metro West is only in the very early stages of planning so it is unclear how a precise estimate of $8.1 billion could possibly have been arrived at. Nevertheless, sufficient details are in the public domain for its basic outline to be clear. All in tunnel, it would be about 23 km long with, probably, 11 stations. On the evidence of the Airport line costs and the North-West Link estimates this should cost a maximum of around $2.3 billion (not including rolling stock), but the official estimate is three and a half times higher. Even if rolling stock is included in the TIDC estimate, it should total no more than $3 billion.

Had the estimates for these three projects been merely 130 per cent of comparable projects — either actually built or properly estimated — the excess might be legitimately explainable by regional differences in materials, geology, labour costs, land acquisition (where applicable) but at 200 per cent such an anomaly requires detailed public explanation. Three hundred per cent and above seems inexplicable and possibly scandalous.

Absurdly high estimates skew the benefit-cost ratios for rail projects with the result that they are indefinitely deferred, typically in favour of motorways. This seriously corrupts the planning process and inhibits the introduction of necessary public transport infrastructure at a time when the global decline of oil production looms, car use is falling dramatically, there’s enormous pressure on existing transit, and vast areas of Sydney are effectively without any public transport at all. On the face of it, if these projects are in fact funded, the Australian taxpayer will be forking out at least three times as much as they should for much-needed infrastructure.

Gavin Gatenby is co-convenor of public transport advocacy group EcoTransit Sydney.