Facebook Google Menu Linkedin lock Pinterest Search Twitter


Public transport

Mar 8, 2016

Melbourne Metro: what do you get for $10 Billion?

The Business Case for Melbourne Metro was finally released publicly last month. It confirms the need and the solution. Now the Commonwealth needs to come to the party with matching funding


Metropolitan rail network at the completion of Melbourne Metro program
Metropolitan rail network at the completion of Melbourne Metro program

It assesses a range of strategic land use, regulatory and capital options, but the Business Case for the Melbourne Metro predictably comes down strongly in favor of building the 9 km rail tunnel under Melbourne’s CBD promised by Premier Daniel Andrews at the last election.

It makes the case that Metro will provide many benefits for the estimated $10 Billion outlay, including:

  • Extra peak hour rail capacity across the network.
  • Improved reliability from separating lines.
  • Reduced pressure on existing strained CBD rail stations by siphoning off some pedestrian traffic to two new CBD stations.
  • Fewer trams clogging Swanston St.
  • Enhanced transport options at Parkville, North Melbourne and Domain through the construction of new stations.
  • Reduced car use by shifting some car trips to public transport.
  • A foundation for further investments to increase capacity.

The most publicly visible part of the Business Case is the 1.1 Benefit to Cost Ratio (BCR). Some are disappointed at how modest it seems but it’s not surprising; it’s in line with what is expected from this sort of project. Most mega transport projects in Australia don’t produce large BCRs (see Is investment in transport a game-changer?).

There’s a number of reasons for that but a key one is projects in mature networks don’t tend to produce the spectacular BCRs that come with small, early stage networks where connectivity is still poor e.g. like bicycle paths (see Does this freeway make any sense?).

The 1.1 BCR is also based on the conventional methodology preferred by Infrastructure Australia. In this narrow interpretation most benefits are time savings and increased comfort for train and car users.

If the so-called Wider Economic Benefits (WEBs) that Infrastructure Australia downplays (WEBs include things like agglomeration economies) are also included as in my view they should be, the BCR increases to a more impressive 1.5.

The Government has also chosen to rely on the most conservative of the multiple models it used in the evaluation, the Victorian Integrated Transport Model. However, Veitch Lister’s Zenith model, which has given superior results on road projects, gives a base BCR of 1.4, increasing to 1.9 with WEBs (see Who got the facts wrong on traffic forecasts?).

On top of that, the Government has used the more demanding 7% discount rate in its public utterances rather than 4%. Moreover it has also resisted the temptation to include future works “enabled” by Metro – like Melton electrification – in the calculation (in contrast with a certain notorious road project).

So it’s refreshing to see the Business Plan hasn’t sought to gild the lily; in fact, it’s surprisingly modest.

Even the publicly quoted cost of $11 Billion does the project no favours. That’s in nominal terms yet expenditure will be spread out over ten years. The P90 discounted figure is $10 Billion; that’s what I’ll cite from now on.

One doubt though is the Government has included the current upgrade to the Dandenong line (e.g. level crossing removals) in the base case. It seems to make sense because the work will be completed well before the Metro opens in 2026.

But the full suite of benefits from the Metro can’t be realised without the upgrade works. The problem here is the Government hasn’t released the business case for the Dandenong upgrade; we don’t know if the benefits exceed the cost.

A particularly interesting insight the Business Case provides is that getting a BCR that’s acceptable (i.e. equals or exceeds unity) doesn’t depend on generating a huge increase in patronage or in mode shift.

When it opens in 2026, the Business Case says Melbourne Metro will increase network capacity in the morning peak period by 39,000 “passengers”, equivalent to an 8% increase in trips. That might not sound like much of an increase for $10 Billion but it’s equivalent to the number of travellers nine freeway lanes carries in the morning peak. (1)

The Business Plan insists it’s enough headroom for future growth on the Sunshine-Dandenong corridor as well as demand spurred by possible future projects, including electrification to Melton and construction of new lines to the airport and Rowville.

The project also confirms a point I made recently; even huge investments in public transport don’t have a big impact on mode share (see Will simply building more public transport seriously suppress car use?).

By 2031, total public transport trips in Melbourne in the morning peak are forecast to increase by 2% as a consequence of building Melbourne Metro and car trips to decline by half of one percent. The net effect is there’s no substantial change in peak period motorised mode share. (2)

Consistent with another point I’ve made before, this project isn’t going to save the planet either. The social benefits (externalities) only account for 10% of all benefits. These are the sorts of things that are usually invoked popularly to justify investing in public transport e.g. reduced emissions and pollution, fewer road casualties, and better health from more exercise.

The rest of the benefits accrue to public transport travellers and motorists in the form of time savings and greater comfort, especially reduced crowding. These are private benefits; they could be paid for by the beneficiaries.

The impact on job location is also pretty modest. The Business Case forecasts Melbourne will have 3.9 million jobs in 2046. Metro won’t change the total number of jobs but it will have some effect on how they’re distributed. Its forecast to increase the number of jobs in the CBD by 2046 from 455,000 to 483,000. Without Metro, these future jobs would locate in the inner city and major suburban centres.

The Business Case gives a lot of prominence to what it calls the Extended Program – a series of future upgrades that would extract more value from the initial investment in Metro. The principal ones are electrification of the Melton Line, quadruplication between Sunshine and Deer Park West, and use of 10 car trainsets between Sunshine and Dandenong.

The Extended Program isn’t counted in the $10 Billion (real) cost announced by the Premier or in the 1.1 BCR. If implemented it would cost a further $2 – 3 Billion, but it would increase the combined BCR significantly, from 1.1 to 1.5 (2.1 with WEBs). It would increase peak period capacity by an additional 41,000 “passengers” and increase the number of jobs in the CBD by a further 19,000 to bring the total to 502,000.

What $10 Billion doesn’t buy, though, is anything better than a 20 minute off-peak frequency for Pakenham, Cranbourne and Sunbury residents. They might be disappointed because the new tunnel is part of their “metro-style” rail line.

Overall though, the Business Plan substantiates the need and the solution. Now the Commonwealth needs to match the State’s financial contribution so the project can be confirmed.


  1. “Passengers” is the term used in the Business Case and it’s unclear what it means exactly. I’ve treated it as synonymous with trips; if however it means boardings, then the increase is only 5%.
  2. Public transport’s share goes from 22.8% to 23.2%.


We recommend

From around the web

Powered by Taboola


Leave a comment

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

7 thoughts on “Melbourne Metro: what do you get for $10 Billion?

  1. Peter Vella

    @Adam Ford makes an excellent point.

    What is preventing 10-minute frequent service from being rolled out on all other lines in the off peak? Clearly, the necessary trains are there from peak hour.

  2. Smith John

    The consideration of alternatives in the Business Case (p93ff) rejects the Northern-Caulfield Loop Connection (now called ‘City Loop Split’) prematurely for inadequate reasons. [note 1]

    The City Loop Split provides the same capacity increase as the Melbourne Metro tunnel at what would certainly be hugely lower cost. However, critically, the study rejects all the non-favoured options *before* making any cost estimates, however rudimentary, so it’s impossible to compare the value for money (benefit/cost ratio) of the various options.

    The MM tunnel requires 18 track kilometres of new tunnel and five new stations to provide one new track pair from the west. The City Loop Split requires up to 4 track kilometres of new tunnel and related surface realignments, and no new stations, to provide the same capacity increase—one new track pair from the west. [note 2]

    Based on these proportions, let’s guess that the City Loop Split would cost $3 billion, compared with $10 billion for the tunnel. A saving of $7 billion is a LOT of money that could go to other worthwhile infrastructure projects.

    Of course the City Loop Split doesn’t give you the three new stations. To make up for that, some of your $7 billion saving could go to improving the tram service on the Parkville-Domain corridor (and if you wish, related projects such as Melbourne City Council’s proposed northern orbital tram line from North Melbourne to Victoria Parade, or a tram extension into the Arden development area).

    Also, the City Loop Split doesn’t give you longer trains to Dandenong. Some more of your $7 billion saving could solve that problem by instituting 9-car trains from Werribee to Dandenong via Flinders St direct (platform lengthening works in the inner area would be complicated but are certainly achievable).

    Let’s suppose all that costs another billion dollars. We’ve still saved $6 billion, which is a HUGE amount that could go to other worthwhile projects.

    I think part of the problem with assessing these megaprojects is that on the benefit side many people are attracted by the visionary aspect, but on the cost side they simply don’t understand very big numbers. Remember the old joke about the congressman discussing the defence budget at a committee hearing: ‘Why Admiral—a billion here, a billion there, and before long you’re talking real money!’

    Think of $6 billion as *sixty* $100 million level crossing grade separations.

    The tunnel is certainly neat, and if I found it in my Christmas stocking I would gladly take it. But in a world of many demands and limited budgets we need to think more about costs as well as benefits, and alternatives and priorities. If option B gives you 80 per cent of the benefits for 20 per of the cost, when such huge sums are at stake, it’s worth more serious consideration.

    Note 1. The ‘City Loop Split’ builds new chords between North Melbourne and Flagstaff and between Parliament and Richmond so that one track pair would run from Craigieburn to Frankston via Melbourne Central. Like the MM tunnel, this creates a new through-city track pair bypassing the present chokepoint at North Melbourne where the Craigieburn and Sunbury lines merge to the Northern Loop.

    Note 2. Other capacity uplift works such as longer trains or upgraded signalling can be done in either case, so are not related to this comparison.

  3. Waffler

    Adam @ #3

    The focus on enabling access to the CBD is because the job and population forecasts used to justify this work are based on assumptions of massive increases in jobs and population in the CBD in inner Melbourne. If these forecasts are realised the CBD just will not function without continuing to invest in massive PT projects to get suburban workers into the CBD.

    Of course I am not sure anybody has asked whether this really makes sense or whether there is a better answer!

  4. mook schanker

    Tony – The Business Case has a capacity uplift diagram and shows in the West, the Cross City will improve patronage quite a bit and also for the Northern lines, very roughly half the capacity improvement will be attributed to these areas.

  5. Adam Ford

    Great post, Alan.

    There is actually a major issue buried in the final stanza here.

    What kind of a network are we building and for what purpose? Because it seems to me that there is still a mindset within PTV that its headline task is enabling the maximal number of peak hour commuters.

    In which case you’re not building a Metro network. Because people in London and Paris would laugh at our twenty minute off-peak frequencies. Off peak frequencies are as crucial a benchmark for an effective metro as your headline carrying capacity.

    And given that we know that the real cost in transport is in CAPITAL, and that variable costs are ridiculously low – you’re paying some small incremental extra wages, extra power, extra maintenance to deliver a network that’s incrementally MUCH more capable of being integrated into people’s daily transport patterns.

    And Alan, I know you want a cost-benefit approach to network additions and not much actually stacks up on that basis. What about a cost-benefit on forgetting about network additions and spending the money to get service frequencies on the existing network up to third world standards?

    I honestly believe the inadequacy of service frequencies outside peak is THE major psychological barrier to many people who would otherwise be inclined to look at changing their transport patterns.

  6. James

    Good summary of the business case! Now if Malcolm could just make a “captain’s call” on this one. Although that might involve standing up to the loonies on the extreme right of his party.

  7. Tony Morton

    Of course, off-peak services every 10 minutes all the way to Pakenham, Cranbourne and Sunbury (hence every 5 minutes between the city and Dandenong) is something that can already be done now, with existing infrastructure. That’s because we already run more than this number of trains in peak hour.

    The Dandenong line upgrade with a renewed signalling system will also – if it’s improved to the same standard as the Ringwood line was in the 1960s – allow trains to Rowville every 10 minutes in addition to the existing peak Dandenong service.

    In other words, big improvements in train service for the south-east are feasible even before we consider what’s being added by the Metro tunnel. It’s in the west that we really need that extra capacity. The RRL has provided more space on the Werribee line (which is still not being utilised, to the disgust of Altona residents), but the Sunbury, Craigieburn and Upfield lines are still squeezed for capacity. With the Metro tunnel we can at least double the number of train services both in and outside peak hour, to match the doubling in population that comes with urban growth in the region.

    The wider point being that the new travel that becomes possible by public transport will drive mode shift, even if no existing motorists give up their cars.

    If we supplement expanded rail service with good frequent bus networks throughout the suburbs, serving the local travel that represents the real ‘heavy lifting’ of our transport task, then there is a good deal of low hanging fruit for getting further mode shift there.


https://www.crikey.com.au/2016/03/08/melbourne-metro-what-do-you-get-for-10-billion/ == https://www.crikey.com.au/free-trial/==https://www.crikey.com.au/subscribe/

Show popup

Telling you what the others don't. FREE for 21 days.