Canberra-Eden rail needs another look

Eden, a small town on the far south coast of New South Wales, could become one of the most important places in Australia. Its natural harbour, Twofold Bay, is one of the finest anchorages in the country, yet its rugged hinterland has kept it isolated from the national freight network. Meanwhile, NSW’s only significant ports (Botany, Wollongong and Newcastle) are at capacity, and so are the freight corridors that feed them. For this reason, in 2018 I was commissioned by the Cooma and Monaro Progress Association (CMPA) to prepare a concept plan to extend the existing but closed Queanbeyan-Bombala railway to Eden. The CMPA plan proposed a railway constructed by a 50-50 public-private partnership, costing $2.5-3.0 billion and having a Benefit-Cost Ratio (BCR) of 2.4, i.e., $2.40 in external economic benefits for every dollar of public expenditure.

The plan’s release was met with substantial enthusiasm, and within a month it was adopted by the NSW coalition government as a major infrastructure initiative. At an August 2018 press conference at Cooma Railway Station, the Premier (Gladys Berejiklian), Deputy Premier (John Barilaro) and Transport Minister (Andrew Constance) presented it as a keystone transportation project, pledging $1 million towards a full business case. “This is not a project for a feasibility study,” said Constance, “it’s a project for a railway to begin with a feasibility study”.

However, when the Government study’s executive summary was finally released in October 2020, the results weren’t just unfavourable, they were dismal. The study’s cost estimate was almost double the CMPA plan’s, but even worse were the projected economic benefits – a present value of just $256 million over 30 years, with a BCR of 0.04 (which was rather unhelpfully rounded down to 0.0 for the report’s executive summary). The full report was confidential (“commercial-in-confidence”, said the department), so it was not possible to review their methodology, but it hardly mattered – the result was so bad that even if there were some errors in the report, they could hardly be of such a magnitude to reverse the project’s economic viability – or so it appeared.

The report was an embarrassment to the government, and to the individual ministers who had championed the project; they quietly dropped the policy without further mention. For me, as the author of the initial plan which had given a very positive result, it suggested a blight on my professional credibility.

The CMPA were perplexed by the disparity in the results of the Feasibility Study versus the original Concept Plan, and requested permission to review the full study. After discussions with the Snowy Monaro Regional Council, who had helped fund the concept plan and were stakeholders in the government’s feasibility study, they managed to negotiate supervised access to the complete study for myself and a handful of other reviewers.

It was immediately apparent that the study was severely flawed. I was astonished at the number of major errors, every single one of which was a net detriment to the project’s viability. 18 errors were over $100 million in magnitude, and of these, four were over $500 million. Some of the largest and/or most serious errors are listed below:

  • Several very large cost items should have been excluded; these alone would reduce the capital cost by $1.6 billion:
    • An extension to Canberra Airport (which the study concluded there was insufficient demand to justify, but included its cost anyway),
    • An expansion of the Port of Eden (which should properly be excluded as an external business project, or if included, should account for the additional benefits of that project),
    • A major intermodal container facility at Hume (which, aside from being wildly overspecified for the container volumes anticipated, is entirely unnecessary – there are existing plans to build a suitable facility at Fyshwick for 1% of the cost, privately funded).
  • Freight demand was estimated by analysing the existing freight flows between origin-destination pairs within the study area, and then estimating what percentage of these existing flows would switch to rail. This is ridiculous, it’s akin to estimating the demand for the Western Sydney Airport by analysing pre-existing traffic flows to Badgery’s Creek. The study seems to have ignored several major classes of potential freight (notably quarry products, paper and agricultural produce), including some that were identified by their own survey results.
  • Passenger farebox income is estimated at $1.2m/annum on ridership of 2,404 passengers per weekday, which implies a one-way ticket price of just $1 – on the study’s own numbers, this should be some 15-20x higher.
  • The discount method employed is patently erroneous – the $256 million net present benefit was supposedly calculated by applying a 7% discount rate on real (undiscounted) benefits of $1,823 million earned over 30 years. That’s mathematically impossible – under any reasonable set of assumptions, the amount should be around triple what has been presented.
  • The study shows no income from freight operations, either in the form of access charges or haulage, despite government funded expenditure on six freight train consists over 30 years – yes, six! – each of 4 locomotives and 86 wagons. It’s not evident whether the proposed operator is selling access to the railway (like the ARTC), or actually operating the trains (like Pacific National). It seems to incur the costs of both business structures, but generate the benefits of neither.

The aggregate effect of correcting these errors, in present value terms, is:

  • Construction costs reduced from $5.0 billion to $2.6 billion
  • Lifetime net income increased from a loss of $381 million to a gain of $962 million
  • Net present economic benefits increased from $256 million to $2.4 billion
  • Benefit-Cost Ratio increased from 0.04 to 1.44

In other words, far from being a repudiation of the original concept plan, the government’s feasibility study, properly corrected, largely validates it.

The four other independent reviewers have also found significant flaws in the study. While these flaws do not necessarily mean that the railway is either viable or unviable, all reviewers concluded that the study’s flaws are sufficient to invalidate it:

  • Dale Budd (a senior adviser on the Inland Rail project and former director of the ARTC) said “There is a great deal of material in the report whose validity can be questioned or challenged.” He singled out the lack of a clearly identified business structure for the proposed railway operator, and the lack of any direct consultation with potential freight customers, as the most severe errors.
  • Nicholas Kilpatrick (logistics manager for Manildra Group for 20 years) said “It appears like a report being fleshed out to appear more in depth than what has actually occurred… personally I cannot see any return in value for the expenditure of $1 million.” He noted the unexplained omission of freight income, low passenger revenue for the expected ridership (“this does not compute”), and the fact that costs had been calculated over 50 years, while benefits had been calculated over 30 years.
  • Bob Nairn (independent civil engineer and author of several books on transport planning and economics) said “The modelling and economic conclusions in the report are not convincing, and the report… has not proven [that the proposed railway line] is not feasible”. He criticised the report’s decision to ignore potential freight from the Riverina region, additional passenger demand (education, retail trading and skiing), and alternative funding models such as public-private-partnership. Also, if port upgrade costs are to be included, then the benefits of that project to the port operator, non-rail freight operators, and the wider economy should be included also.
  • Colin Mellor (transport economist) said “In summary, there seems to have been a combination of careless errors in the Feasibility Study.” In his opinion, both passenger and freight demand/revenue appear to be considerably underestimated. He highlighted the inconsistent study periods, and the study’s apparent failure to distinguish between economic and financial assessments (eg, elimination of taxes and other transfer payments from economic analysis, as is standard practice).

The CMPA has reached out to the state government and relevant departments on numerous occasions for comment, but they apparently have no interest in addressing the flaws of the study. Each time, their reply has been to simply repeat, without reasons, the study’s conclusion that the project is not economically viable, and as such, the study has now been closed. Furthermore, the department has pressured reviewers to sign confidentiality agreements regarding their findings, again relying on the implausible rationale of “commercial-in-confidence”. There is nothing in the study that satisfies such a description; it is being used merely as a convenient excuse to avoid scrutiny of a matter of public importance. The same questionable tactic has been used to suppress other rail feasibility studies in the state – notably Narrandera-Tocumwal, Blayney-Demondrille and Cootamundra-Tumut. If these studies are as obviously flawed as Canberra-Eden, then this suppression is doing a grave disservice to the nation.

How can the NSW government be satisfied with their (our) money being so egregiously wasted? The reviews conducted by myself and others demonstrate beyond reasonable objection the Feasibility Study is fatally flawed, and therefore cannot be used to justify the rejection of this critical national infrastructure project. The NSW government should immediately instruct the Department of Transport to withdraw and declassify the study, and refer it to Infrastructure Australia for further review. If they fail to do this, either the state opposition, or the federal member for Eden-Monaro, should hold the state government to account.

List of Errors

The following table lists the errors identified in the Canberra-Eden Railway feasibility study, ordered from largest to smallest. Each error is in Net Present Value terms, using a discount rate of 7%. For readers who desire a more technical discussion, more detail will be published in an upcoming post.

($million, NPV @7%)
Canberra Airport Link753Capital Cost
Port Eden Upgrade640Capital Cost
Land Sales596Lifetime Net Income
Affordable Housing500Economic Benefits
Discount Method390Economic Benefits
Freight Benefits339Economic Benefits
Soft Costs326Capital Cost
Rollingstock Capex272Lifetime Net Income
Intermodal Terminals216Capital Cost
Passenger Revenue214Economic Benefits
Below Rail O&M200Lifetime Net Income
Earthworks189Capital Cost
Freight benefits on higher demand178Demand Estimate
Track Costs177Capital Cost
Delay need to upgrade Port Botany175Economic Benefits
Passenger benefits on higher demand159Demand Estimate
Above Rail O&M136Lifetime Net Income
Passenger Vehicle Operating Costs115Economic Benefits
Viaduct specification93Capital Cost
Decentralization policy67Economic Benefits
Freight Access Charge on higher demand56Demand Estimate
Freight Access Charge43Lifetime Net Income
Residual Value23Economic Benefits
Concrete Culverts23Capital Cost
Passenger Access Charge22Lifetime Net Income
Passenger Access Charge on higher demand19Demand Estimate
Congestion Reduction16Economic Benefits
Passive Level Crossings8Capital Cost
Corridor Width7Capital Cost
Colinton Tunnel6Capital Cost
TOTAL$5,958 million
(increase in NPV)

Comments are closed.