Has the very fast train run out of steam?

The idea of very fast trains on the Sydney-Canberra-Melbourne corridor seems to have run out of steam after an initial flurry earlier this year when the CRC for Rail Innovation and the Greens both called on the Federal Government to fund a major concept study.

There has been very little in the popular media since that opening blast. Hence it was interesting to see author, educator and consultant, John Legge, arguing in The Age on Saturday that the figures on very fast rail add up.

Note the extensive green areas near Canberra

He asserts that: (a) high speed rail works in Europe, China and Japan, (b) the US is now building high speed rail projects, (c) rail is faster, more comfortable and more convenient than flying, and (d) it can be built for a mere $15 billion. Ergo, it will be an unqualified success on the Sydney-Canberra-Melbourne corridor. But he provides little in the way of evidence or argument to support these claims.

Surprisingly, Legge makes no mention of what I think is the best argument for the project – constraints on Sydney Airport’s ability to expand capacity. Even more surprisingly, he doesn’t mention the environmental benefits of rail over air until the very last line and then just in passing. His key argument is that rail would out-compete air simply because a very fast train is a more attractive product.

I don’t want to deal with everything Legge says because I’ve covered most of the arguments for and against high speed rail here and here. But I do want to take issue with a couple of the points he makes.

First, what works in other countries won’t necessarily work in Australia. Most of the routes in Europe and Japan where high speed rail is strong are relatively short e.g. London-Paris is 340 km; London-Brussels is 198 km. London, with a population of 8 million and Paris (10 million) are very big cities compared to Sydney and Melbourne. They are world cities that have huge numbers of business travellers and tourists compared to Australia (London is an ‘Alpha’ world city). Congestion in the air and on the ground makes air less competitive in those cities, particularly for short trips.

Distances are much longer in Australia. The flying distance between Sydney and Canberra is around 750 km. The existing Sydney-Melbourne rail line is 950 km. A new high speed rail line would have to thread a path between the Kosciuszko, Snowy River and Alpine National Parks, so a distance of 850-900 km seems more realistic than the 750 km assumed by Legge.

Second, he claims without any evidence or supporting argument that it would cost $15 billion to build the line and acquire rolling stock. This computes to $17 million/km on a 900 km line. However construction estimates for Europe and the US calculated by the US Government Accountability Office range from a low of $14 million/km to $82 million/km without rolling stock. My (conservative) estimate is $27 billion to build the line plus $140 per one-way passenger trip for operating costs, including rolling stock!

Third, his assumption of a Sydney CBD to Melbourne CBD travel time of “less than three hours” is arguable. That’s an average speed of 300 km/hr or more, so the peak speed required to offset the slower travelling times in the suburbs of Sydney, Canberra and Melbourne plus any other stops on the route would be considerably higher. Some high speed trains have exceeded 400 km/hr in test track conditions (and mag levs have exceeded 500 km/hr) but the only line of significant length that has achieved an average speed in real world conditions of 300 km/hr is the Wuhan-Guangzhou line which opened six months ago. It has 468 km of rail on viaduct and 177 km in tunnel. Needless to say, China is a vastly different market to Australia.

There are also the usual unreflective assumptions: that all travellers will want to go CBD to CBD rather than leave from home in the morning or return direct to home in the evening; that time consuming anti-terrorist security precautions will continue to be a part of air travel but will never need to be applied to high speed trains; and that the monopoly train operator, having destroyed air as a competitor, would never cramp up seat space like we’re used to with planes.

But it’s not enough that a high speed train service can compete successfully with air for market share – there has to be some other advantage to compensate for the environmental impact of constructing 900 km of rail line across south-eastern Australia; for the extensive Government assistance with land assembly and tax breaks that would inevitably be “necessary”; and for substituting a monopoly operator for the four airlines that currently compete on the Sydney-Melbourne route.

That advantage isn’t going to be greenhouse gas savings – there are considerably cheaper and more effective ways to offset emissions associated with the planes on this corridor than building a very fast train. The quantity of greenhouse gas liberated during construction would in any event swamp the relatively modest operational greenhouse savings.

As I’ve mentioned before, the viability of a very fast train in the Sydney-Canberra-Melbourne corridor will depend primarily on whether a second Sydney airport and associated transport links is needed. If so, and if it were to cost around $15 billion and emit similar levels of greenhouses gas during construction, then a very fast train might be competitive. The cost of a second Sydney airport at Badgery’s Creek was estimatedin 1999 at between $6 to 8 billion dollars.

9 Comments on “Has the very fast train run out of steam?”

  1. TomD says:

    An interesting range of issues and consideration again raised by you Alan!

  2. Jim Wells says:


    Quite right to query Legge’s estimate of cost. NSW now says 11 km of SWRL will cost #2.1bn.

  3. Russ says:

    Alan, good post. Studies I’ve seen on European lines seem to indicate that the optimum length for high-speed rail is 400-1000km. Any shorter and their is no great speed advantage over other rail, any longer and flying is substantially faster. Melbourne-Sydney just fits into that category; being a longer line also means it will average a higher speed (provided it doesn’t stop very often).

    Another point that was missing is that any fast rail needs to be a “network”, not just a line. Redirecting all trains through Canberra would take over an hour, destroying any chance of the line being competitive with air. (It is worth noting that the last federal study on the project assumed around 5-12 stops between Melbourne and Sydney, conceptualising the project as a glorified country network).

    A network, although more expensive than a single line, would connect up a much larger number of potential destinations. Hence, it is probably better to think of a fast-rail as four (overlapping) lines: Albury-Melbourne, Albury-Sydney, Albury-Canberra, Canberra-Sydney. Such an arrangement also raises the possibility of building an airport specifically for long-haul international flights in Albury (or maybe Wagga) that could serve all three markets simultaneously.

  4. MO says:

    In response to Russ’s comments about the number of stops on the line. The Chinese now propose the tain does not stop but passengers join and depart from pod carriages that attach / detach to the roof of the train.
    The email doing the rounds with the concept movie states the Beijing to Ghanzhou has 30 stops in between. By not stopping at 30 stations @ 5 mins is a 2.5 hour saving.
    If we can develope this system we will have a fast service that also supports the reginal locations.

    • Alan Davies says:

      Yes, I’ve seen that too (video here).

      Hugely important idea but probably a long way from implementation yet I’d imagine. A key issue for more R&D will be how to avoid sudden acceleration of the pod from standstill to 350 km/hr or more (and vice versa) – what they show in the video seems way too short. The top rail might need to be a fair bit longer.

      And another point: that design only envisages one pod at a time. Also, if you get on via pod at Goulburn, say, you can only get off at one place. Could possibly work however for Sydney-Canberra and Canberra-Melbourne passengers, although I’d imagine you’d need something vastly bigger than a pod for that route (Sydney-Canberra is a very busy air route)

  5. […] bit of attention to the proposed Very Fast Train between Sydney-Canberra-Melbourne (here, here and here), wondering what warrant there is to replace one form of public transport with […]

  6. […] the former nevertheless has important lessons for us. I have previously discussed the VFT here, here, here and […]

  7. […] air’s higher speed (I’ve looked at this and other issues in greater detail here, here, here […]

  8. […] who hold out hope that it could be part of a Sydney–Melbourne HSR project ignore the fact that numerous studies have found such a project is not viable (the Commonwealth is undertaking another study at the […]

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s