Skip to content

Should we be building new rail lines up in the air?

Crikey, 3 March 2016
It would be a pity if the “Sky Rail” brouhaha in Melbourne over removal of level crossings were to damage the potential use of elevated rail for totally new rail lines in all Australian cities.

Whether or not it’s the right solution for grade separation in Melbourne, elevation has considerable potential for building new rail lines in situations where the only feasible alternative is building underground.

Level crossings are a special case because the rail line already exists. In fact Melbourne is unusual; most other cities don’t have the problem of circa 170 level crossings holding back the maximum frequencies rail lines can run at.

In other cities, the focus of interest is mostly in building brand new heavy or light rail lines (or busways) in areas that aren’t currently served by high capacity public transport operating on a dedicated alignment.

More often than not the most prospective projects are in heavily built-up areas that lack reserved land for new transport infrastructure.

In these situations conventional surface rail is too expensive and politically difficult. Land is very expensive to acquire, there are many roads to cross requiring grade separation, and the cost of disrupting other activities – especially traffic – is high. For similar reasons, open trenching is rarely an option either.

Tunnels are typically used instead. The CBD is the key example (e.g. Brisbane Cross River Rail, Melbourne Metro), but it would be difficult to build a major new rail line at grade or in an open trench anywhere within 10 – 20 kms of the centre in Australia’s major cities.

Elevated rail is a plausible alternative because it requires a relatively small footprint for columns and stairs. It is widely used around the world. Sydney has elevated rail in Edgecliff and even managed to cram a monorail into the main streets of the CBD (although it was grossly inappropriate and mercifully is now gone).

The key advantage of elevation relative to going underground is cost. Research by Flyvberg et al on the comparative costs of urban rail found that the cost of tunnelling, including cut and cover, is on average at least twice as much as elevation.

It can be a lot more though; there’s enormous variation between cities depending on factors like the local geology and topography.

Given the stratospheric cost of infrastructure in most of Australia (e.g. $11 Billion for the 9 km Melbourne Metro tunnel), the cost advantage of elevation means rail lines that mightn’t otherwise get built would have a much better chance of proceeding.

That’s a huge benefit. Residents served by a new “el” would get a significant improvement in accessibility. It would also make dedicated rail more cost competitive with on-street tram/light rail systems that get caught up in traffic.

Like everything, though, there are drawbacks; trade-offs are inevitably involved. The key downsides of elevation relative to tunnelling are visual changes to the streetscape, noise, and disruption during construction.

There’s no way it can ever be as discreet as undergrounding and it won’t work everywhere e.g. elevating the entire Paris Metro would be disastrous. Notwithstanding Sydney’s monorail folly (or perhaps in part because of it), elevated rail would be unthinkable today in the CBDs of Australia’s major cities.

But the downsides shouldn’t be over-stated. As elevated lines would in most cases run down the median of wide, heavily trafficked roadways, both visual intrusion and noise should be manageable problems. In any event, designers can improve the appearance of viaducts and engineers insist noise can be contained to acceptable levels with baffles.

Moreover, the sorts of areas that warrant provision of new rail lines are also key locations for higher density housing; in fact that’s a key justification for getting a rail line in the first place. Viaducts are typically no higher than the 3 – 4 storey medium density housing that new rail lines are built to encourage.

Perhaps the most important point, though, is that surface rail isn’t anywhere near as inconspicuous as a tunnel either. Undergrounding is done because it’s the lowest cost option available in heavily developed areas; it’s not done because the impacts of surface rail are all of a sudden no longer deemed acceptable.

There’s a good case for assessing impacts associated with elevated rail – like noise – against the accepted standards for at-grade systems, not underground ones.

Above all else though, elevation means bang for the buck compared to tunnelling; every dollar of capital investment is effectively doubled.