Public transport

Mar 13, 2017

A proposal for a Melbourne Ring Metro

Melbourne needs an orbital or "ring" light metro linking major suburban centres to take radial trips off the road system, argues guest writer Dr Garry Glazebrook

Alan Davies — Editor of The Urbanist

Alan Davies

Editor of The Urbanist

Metropolitan and major activity centres (source: Plan Melbourne)

In the first of a series elaborating on his proposed $80 billion Building Australia Fund, guest writer and transport consultant, Dr Garry Glazebrook puts the case for a Melbourne “ring” rail line using light metro technology: (1)

Melbourne has well-developed radial rail and tram networks focusing on its CBD, a legacy of rapid growth after the Gold Rush when rail technology was the dominant form of urban transport in cities.

The CBD is also relatively centrally located within the metropolitan region, with new development spreading north and west to balance the East and South-Eastern suburbs. It’s also relatively unconstrained compared to cities such as Sydney and Brisbane, and the addition of Melbourne Docklands and Southbank in recent decades has provided ample space for residential as well as commercial development.

However Melbourne’s rapid growth is generating major congestion on both road and rail networks. The heavy rail network is close to capacity, while the tram network suffers from partly sharing road-space with traffic.  There is a growing gap between opportunities available to people living in the inner city suburbs and the those in the outer suburbs, who are facing increasing commute times to reach the job-rich inner suburbs.

Accordingly Plan Melbourne identifies a number of centres outside the CBD where activity needs to be focused. However, making the plan viable will require a significant upgrade to the accessibility of these centres by public transport. This suggests a need for one or more  “ring metros” which can complement Melbourne’s strong radial public transport networks. This will allow public transport to take some of the radial trips currently concentrated on the road system.

London has long had its “Circle Line”, which is a form of inner ring; a few cities, such as Berlin, Tokyo and Moscow, have had ring metros some distance out from the city centre for many years. Others such as Shanghai and Paris have recently built or are building such circumferential systems – in the case of Paris, two rings, using both metros and light rail. (2) (3)

However for such a strategy to be effective, such a “ring metro” needs to have certain characteristics:

  • It needs to operate at high frequencies throughout the day – typically headways of 3-4 minutes – since many people using it will be transferring from the radial network.
  • It needs to have well-integrated interchanges with the radial train and tram networks, as well as with major bus routes.
  • It needs to operate in its own right of way and be grade separated from roads and existing public transport systems so as not to cause delays to its own patrons or to the rest of the transport system.
  • It needs to have reasonable capacity (e.g. 3,000 – 10,000 passengers per direction per hour in peaks) to accommodate current travel patterns and potential growth, especially as the sub-centres themselves grow in importance.
  • It needs to be relatively low cost to build and operate, and to be capable of staged construction.

In terms of potential rights of way, several possibilities exist:

  • Alongside or in the median strip of existing circumferential freeways.
  • Underground (probably in cut-and cover tunnels), especially as they approach the key centres.
  • Above ground where the environmental impacts are acceptable and the corridor space is available.

Taking these factors into consideration suggests the ideal system would be some form of  automated light metro rather than the traditional train or tram systems used in Melbourne.  It is likely that the vehicles used in such a system will need to be relatively small in length and cross section, able to climb significant gradients and negotiate tight curves, and operate automatically on short headways (down to 2 minutes).


Examples of such systems overseas include the VAL system developed in France and used in a number of countries; SkyTrain in Vancouver; and the Docklands light railway (DLR) in London. There over 50 cities world wide with “light metro” systems, some of which would have the desired characteristics.

In Vancouver the Skytrain system operates underground in the city centre like a traditional metro, but is largely elevated elsewhere, with stations integrated into high-density activity centres. In London the DLR was able to utilize some existing underground and above ground rights of way near the Central Business District, but again was largely built on new above ground structures through the Docklands area, where it was also integrated into major new development.

In the case of Melbourne, the existing ring freeway system offers some potential for minimizing the impacts of any new ring metro. However unlike freeways, which ideally skirt major centres, the ring metro must penetrate to the heart of such centres and interchange with existing stations if it is to be useful. Thus some underground sections and stations will be required.

One of the advantages of a light metro with automated trains is that a network can be built up over time, providing high frequency services across a variety of routes, minimizing interchange times. This is how both the Skytrain system in Vancouver and the DLR system in London have developed.


The cost of any “Ring Metro” for Melbourne will depend on its extent and on the technology used. Light Metros typically have lower axle loads, shorter trains and stations, and driverless trains, which make them significantly less expensive per kilometre than conventional rail like Melbourne’s new 9 km Metro tunnel or the new Sydney metro – the latter is costing approximately $200 m – $300 m per kilometre, and is mostly underground.

Assuming a mix in the order of 50% surface in freeway reserves, 25% in other surface or above ground locations, and 25% underground, a typical average cost of $150 million per kilometer suggests a likely capital cost of the order of $10 billion for a 66 km ring light metro for Melbourne. Such a system could connect centres such as Frankstown, Dandenong, Monash, Box Hill, Latrobe, Epping, Broadmeadows, Melbourne Airport and Sunshine.

In this context, the lower operating costs and relatively high speeds (compared with street-running light rail) of an automated light metro should mean that farebox revenue should be capable of covering 100% of operating costs, with the potential to fund a component of the capital costs. In addition, application of land value capture mechanisms at key centres on the route would provide further opportunities for recovering part of the capital costs.

Subject to more detailed assessment, it is assumed that the project could return 50% of the capital costs, with the remainder covered by the proposed Building Australia Fund.

Thus it is assumed that the Building Australia Fund would invest of the order of $4 billion on this project over a fifteen-year timeframe, commencing in 2020 to allow time for necessary planning and design of the route, selection of the most appropriate technology, etc.


As noted, construction of any “Ring metro” would need to be staged. It is suggested that the first stage could include a number of the key centres identified earlier, as well as Melbourne Airport. In addition to interchanges with radial rail and tram routes which cross the ring route, additional park and ride stations could be built on the ring metro for people accessing the system by car or bus. This could allow many people to access Melbourne airport without adding to road congestion on Tullamarine Freeway. Assuming typical spacing between stations on the ring metro of 1 – 1.2 km the system would be able to average 40 kph or more, making it much faster than Melbourne’s trams and competitive with car-based travel for many circumferential trips.

In addition to integrating the ring metro with other transport modes, the opportunity to add medium density housing along the ring metro route, as well as higher density nodes at key centres, should also be taken. Close integration with land use is a key to the success of any public transport system achieving its potential to transform the urban fabric of the city and to maximize the urban accessibility benefits public transport can provide.

See also Building rail connections across the suburbs and What should we do about the airport?


  1. For a copy of the full “Building Australia Fund” report and Appendix (containing details of key projects) contact Dr Glazebrook at [email protected]
  2. A ring metro system for Perth was proposed by the ALP in Western Australia, and is now a centrepiece of the incoming State Government’s transport plan.
  3. See Metro rings and loops for a more complete discussion of circumferential public transport routes
(Visited 78 times, 2 visits today)


Leave a comment

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

18 thoughts on “A proposal for a Melbourne Ring Metro

  1. Socrates

    As with the Sydney post this is an interesting idea and such a plan is badly needed. Melbourne is already the same land area as London or Paris so the comparisons are not far fetched. In fact Paris T2 is a good model for what you propose, and very popular. I have only two caveats. First costs could be greatly reduced (halved) by using elevated rather than underground sections. Second, it would need to be accompanied by land use controls to make it effective. A containment boundary is needed, and incentives to build new retail and commercial space near such lines.

  2. Apocalyptoid

    Great idea, governments across Australia should get on with building smart infrastructure like this instead of stupid freeways as usual.

  3. James

    It would be interesting to see what the Melbourne transport model, the VITM, forecasts the patronage of an orbital line would be. For all it’s faults it would give an indication of the likely demand for this sort of travel.
    That said, I’m not a fan of polycentric city ideas. They are always talked about, but never really work. I prefer the monocentric city because trips are encouraged to be radial and public transport is a real viable alternative when you have massive congestion heading into one mono-centre like our inner city. I imagine driving would be very much competitively advantaged for orbital travel. I’ve love to be wrong though..

  4. Saugoof

    I am totally in favour of the idea. It’s always really annoyed me that practically all public transport in Melbourne is only heading in and out of the CBD. Busses don’t really help here because bus schedules and routes are practically impossible to follow unless this is a regular route you do. So for a casual user busses are pretty much useless. As an example, I took a public transport trip from Highett to Oakleigh the other day. By straight line or road this is a distance of a mere 5km or so. Taking the train required a trip all the way to South Yarra and then back out again. All up it took about 90 minutes to make the trip by train, I could have made it in half the time if I’d walked instead.

    That said, Melbourne used to have a circle line of sorts going through Carlton and Fitzroy. That route probably wouldn’t make much sense nowadays.

    1. Tony Morton

      Saugoof – your Highett to Oakleigh example is one that could be done quite easily and directly by public transport, but for the poor standard of our bus services. There’s already a bus that runs direct along North Road from Ormond station: in principle it’s an easy change, but it’s let down by poor frequency, lousy interchange facilities and no traffic priority. If this route and others like it were to run every 10 minutes with traffic priority and decent wayfinding, it would serve a lot more of this kind of suburb to suburb travel as part of a comprehensive network.

  5. Guest

    This is too expensive. Money that would go into this should be spent on renumbering and reorganising Melbourne’s bus network. Auckland has done something similar to this. More cross-city buses can be used on existing roads or with priority running ways. Larger capacity vehicles could also feature.

    If there eventually is a ring rapid transit service, both buses and trams should be looked at. The current tram system does very well indeed (203 million boardings) and improvements to the tram network are long overdue (thinning out tram stops for speed, upgrades to ensure DDA compliance, extensions and new lines).

    1. garry glazebrook

      Hi, thanks for your comments.
      The problem with buses (and trams) is that they are too slow, unless given exclusive rights of way, because they get caught up in traffic. This will be particularly the case approaching centres, where traffic congestion can get quite severe. this not only makes them inefficient, but ineffective in attracting people out of cars. The trams work best in the inner suburbs and for short trips, but in the middle suburbs distances can be quite long. Have a look at what other cities are doing…

      1. Guest

        Hi Garry,
        Thanks for taking the time to respond to this. Service speed is determined by the level of priority (roads, dedicated lane or exclusive guideway) and the distances between stops, not vehicle mode.

        Buses and trams are inexpensive and tried and tested in Melbourne. A metro is not and is very expensive, meaning that coverage will be very limited. That shortness will cap passenger numbers.

        In any case, the route choice should be done first and then various modes looked at to see which fits best, rather than the other way around.

  6. Jacob HSR

    What about the high voltage electricity pylon network? There is basically nothing but grass under the high voltage transmission lines – occasionally they go over a road or a carpark.

    There are transmission lines parallel to the ring road and obviously across town. It would probably be cheaper to bury the transmission lines and use the land to build a public transport system than to build ugly tunnels.

    A very long tram that goes along the land currently occupied by high voltage transmission lines with the VL tram stations being 2 km apart.

    1. garry glazebrook

      Interesting suggestion, and could be relevant in certain parts of the corridor, but would need close examination. Wont work for when the route needs to get into centres or interchanges with rail stations though, which will be critical.

      1. Jacob HSR

        What centre? Lalor Station is not next to the shops nor the library – it is far away and next to a bloody park and houses!

        Jolimont station is not next to the MCG – one has to walk far to get to the MCG.

        The Hotel at KUL airport is a “5 minute walk” away from the terminal.

        The idiots should build a cycleway from Jolimont station to Richmond station. And another cycleway from North Melbourne to Southern Cross. Then people can rent a Baillieu bike to go safely between stations.

        As for the proposed orbital tram line – you could have short cycleways to go from a tram stop to a train station.

  7. Roger Clifton

    A driverless train could give remarkably short transit times if it never stopped. Pickups could be done by joining a just-started carriage onto the front, while the last carriage breaks off to deliver alighting passengers. Passengers shuffle forward until they reach the freshly-arrived carriage that will stop at their destination. It then offers an express service between any two stops, with most of a passenger’s journey having travelled at the maximum speed of the line. Thus a journey of 20 km through a traffic-jammed city need only take a quarter hour.

    1. garry glazebrook

      This sort of approach has been tried with express trains in the UK a century ago. Sounds attractive, but in practice a ring metro will tend to have fluctuating patronage with multiple on- and off-loading points as it circumnavigates the city. Unlike traditional radial routes where patronage tends to fall off from the centre. Hence the system would need to have full capacity for most of its route. Thus apart from the practical and safety issues, this is not really desirable for a ring system.

  8. Tony Morton

    But while we ponder fancy technology, can we at least get Melbourne’s existing orbital bus routes to run every 10 minutes 7 days a week, and give them some real priority at traffic lights?

    The thing with new infrastructure is it doesn’t seem to matter whether it’s buses or trains or the current flavour of the month – the cost is dominated by the need to engineer a dedicated right-of-way through an existing urban environment with no land reservation. The technology is secondary, which is why advocates like PTUA will generally opt for whichever of our current established technologies provides the appropriate capacity given the new line’s role in the network. (It’s also why we are so protective of what remains of the Doncaster rail reservation.)

    Fixing up the orbital links in our network is vital, but shouldn’t detract from the fact that 50% of Melbourne’s jobs are within 15km of the city centre and likely to remain so for the foreseeable future. This means a high priority will always need to be given to high-capacity radial links.

    1. garry glazebrook

      Yes but to date almost all of the effort and expenditure has been on radial routes. The new metro under Melbourne plus additional trams should provide sufficient radial capacity across the network for another 25 years or so (unless all future job growth occurs in the CBD). But without much more effective circumferential public transport travel options, cars will continue to totally dominate travel patterns in middle and outer suburbs and the potential secondary nodes will never achieve their potential…With regard to the specific modes, rail has proven world wide to be far more attractive than buses, notwithstanding the attempts to improve the latter with busways. For example Ottawa is currently replacing part of its busway system by light rail. Curritiba is looking at metro, and Brisbane City Council is considering the option of replacing its busway with light rail as well. Data from the UK, US and Australia shows almost no growth over the last thirty years in overall bus patronage, while rail patronage is growing rapidly. Automated mini metros also offer low operating costs.

      1. Guest

        Brisbane City Council has ruled out a rail metro and will now use BRT superbuses to do the metro function. This will work out to be around $500 million cheaper than rail it estimates.

        It is true that the automation is an advantage. But it is not a fixed thing. Vehicle technology is improving markedly an in the future an autonomous bus could be on the market. The automation advantage of rail vs bus in terms of labour costs to operate would thus be eliminated under such a scenario.

        1. Tony Morton

          Brisbane unfortunately still operates in a silo mentality, with BCC in particular favouring buses because that’s their bailiwick. Recent busway projects in Brisbane have wound up costing about the same per kilometre as rail projects in Perth – the reason I think being the right-of-way as the dominant cost component rather than the specific technology. It’s a shame Brisbane didn’t make some different decisions around 20 years ago that might have led to a true multimodal network, rather than duplicating large parts of the rail system with BRT infrastructure.

      2. Tony Morton

        Thanks Garry. I think when it comes to suburban travel needs in Melbourne it’s as Paul Mees used to say: the priority is to cater for long-distance trips in the radial direction, and to do what we can to compete with car travel for short-distance local trips. If you can do that, you’ve already addressed at least 80 per cent of suburban travel in Melbourne.

        Long-distance orbital travel across the suburbs receives a lot of attention but is less significant as a proportion of total travel. It should be catered for as part of an overall frequent transport network – including consideration of light metro and other technologies building on what currently exists – but having the comprehensive network that caters for diverse local trips is the key.

        (It is different in Sydney, which has evolved into more of a polycentric pattern – perhaps due to the location of the CBD within the overall metro area – and so probably does need more effort put into dedicated orbital corridors focussed on suburban centres.)

Share this article with a friend

Just fill out the fields below and we'll send your friend a link to this article along with a message from you.

Your details

Your friend's details