Facebook Google Menu Linkedin lock Pinterest Search Twitter


Public transport

Dec 16, 2012

Does Sydney's new light rail plan make sense?

There are literally hundreds of existing light rail systems in the world. The value of some is questionable, but Sydney's proposed CBD and South East Light Rail line looks like it'll be among the best


Proposed light rail, George St Sydney

Cities are increasingly turning to buses because they’re relatively low cost and flexible. They come in small units that can be scaled right up to Bus Rapid Transit; they use existing road space; and they can go wherever they’re required, including around unanticipated obstacles.

But like all technologies, buses aren’t the right solution for all occasions. High and concentrated patronage stretches the technical limits of buses and demands other mass transit solutions.

That’s especially apparent in a dense location like Sydney’s narrow and constrained CBD. All up, more than 6,000 buses enter central Sydney each weekday.

In the morning peak, about 20% of commuters come to work by bus. That’s over 1,000 buses between 7.45am and 8.45am.

Around 180 of those buses use Sydney’s premier civic spine, George St, in the busiest hour, equivalent to an average of three per minute.

Despite technical advances, the sorts of buses used in Australian cities are noisy, polluting, lumbering vehicles. In many situations that’s an acceptable trade-off, but not when they operate at very high frequencies in very dense locations.

The quality of the environment at street level matters in the city centre where walking is the primary mode of transport. It’s especially important in Australia’s recognised ‘global city’ – the nation’s leading tourism and business centre.

So on the information available, the O’Farrell government’s announcement last week that it will spend $1.6 billion to replace all buses in George St with light rail and extend the line to the south-east suburbs makes good sense (see press report here, and TV reports here, here, here and here).

The CBD and South East Light Rail line will run from Circular Quay to Central Station and then on to Moore Park with lines branching to Kingsford and Randwick. It will service the sports stadia around Moore Park, Sydney Boys and Sydney Girls high schools, Randwick Racecourse and UNSW.

Within the CBD, the line will remove 180 buses in the busiest hour and, combined with other network changes, facilitate removal of 220 across the whole CBD.

The vehicles envisaged for the route will accommodate up to 300 passengers (100 seated, 200 standing). Even a large “bendy” bus only takes 100 passengers.

At maximum loadings and operating at two minute frequencies, the vehicles are theoretically capable of moving 9,000 people per hour in each direction.

The business case for the line still has to be tested before the government fully commits to it. It’s intended it will be funded in part by a PPP arrangement, although taxpayers will most likely still end up carrying most of the risk and paying for it.

The estimated $1.6 billion looks a little optimistic to my jaded eye. But so long as it’s within cooee of that figure, it ought to offer great economic and political value for the money (compare it to the est $10-$13 billion for the WestConnex project or the $7.5-$8.5 billion for the North West rail link).

Apart from cost, another risk is the need for some bus users on other routes to interchange to light rail at Central station, or thereabouts, in order to get to the northern end of the CBD.

Light rail will offer high frequencies on George St, but there’ll also need to be frequent services on the feeder bus routes to minimise transfer time. Interchanging is likely to be unpopular but is unavoidable if Sydney is to have effective public transport – it needs to be made as effortless as possible.

There is a host of issues that will need to be addressed. They include:

  • capitalising on the opportunity to reimagine George St as a street befitting a city of Sydney’s world status
  • minimising conflict between pedestrians and residents on the one hand, and what are essentially short trains operating on streets, on the other
  • ensuring light rail has reasonable priority over other traffic along the entire route
  • avoiding excessive dependence on a single route in the CBD c.f. Swanston Street in Melbourne

I’d also want to see the business case before I was convinced travellers will have significantly faster overall door-to-door journey times and better predictability compared to buses operating with a similar degree of priority.

The key point about this project, though, is it’s underpinned by a sensible rationale. When the hundreds of other light rail and tram projects throughout the world are examined, not all make financial or environmental sense.

The CBD and South East Light Rail line isn’t primarily about reducing emissions or oil consumption, although there’ll likely be some benefit on those scores. Rather, the key justification is the level of projected passenger demand appears to warrant a fixed-rail solution (patronage levels will be evaluated in the business case).

The overall cost of building and operating light rail appears to be lower than continuing to run buses. In particular, the proposed solution will address the severe bus congestion and associated disamenity inflicted on the CBD.

While it has its virtues, to be frank about it, Sydney’s CBD isn’t as appealing as it might be. It would benefit from a project to clear most buses and cars out of the centre.

Like Clover Moore, Mr O’Farrell should be looking ahead to what Sydney could be. Sometimes I wonder if NSW’s leaders really believe their own rhetoric about Sydney’s leading place in the world.

Some of the many new light rail projects completed, under construction or proposed throughout the world are of questionable value. We should be wary of any project to replace one form of public transport with another.

We should also be wary of opting for high capital and relatively inflexible rail-based solutions unless the circumstances are appropriate.

However subject to the business case being tested, this looks like one of those cases where light rail really is the right way to go. The O’Farrell government deserves applause for this aspect of its new Long Term Transport Master Plan (I’ll look at the rest of it another time).



We recommend

From around the web

Powered by Taboola


Leave a comment

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

16 thoughts on “Does Sydney’s new light rail plan make sense?

  1. Dudley Horscroft

    Australian and NZ tramway construction seems to come out at rather close to $30M/km rather than $133M/km. Comparable recent US figures are Norfolk at $28.6, Charlotte at $30, Dallas at $40 and Phoenix at $43.5, all in $M per km. At worst Sydney should be no more than $40M/km, giving a rough total of $480M for 12 km. Re times, 24 min from Kingsford to Central is an obvious misprint, when you consider the current bus timetable. The 0602 arrives at 0620 (18 min), the 0707 arrives 0730 (23 min), the 0800 arrives 0825 – (25 min). Consider: fewer stops, far better acceleration and braking, shorter dwell times (more doors per 100 passengers) and being largely segregated from other traffic – the timing should be closer to 14 min, not 24.

    Re Rachel612 at 7, the trams to buy are the Skoda 15T – can be 60 m long and seat around 192, plenty more standing room, very good acceleration and braking, and quiet. With loops at each terminus, they can be single sided so maximize seating (double sided loses 25% of seats) and they would have 12 wide doors so dwell times would be very short.

    Re Dylan at 14, thankfully that was a rumour – though no doubt it felt like it. Re Tom at 15, all the traffic lights in Melbourne at the Little Streets should go, they prevent a proper coordination of lights in the CBD – which used to be on a 80 second cycle, allowing 45 cycles per hour, and given that it was normal for three of the old W trams to pass each cycle, the theoretical capacity for Swanston Street was 135 trams per hour in the peak. Probably actually worked out at 120 as some spare slots would hve been allowed.

  2. Tom the first and best


    The lights for Fliders Lane and the little Streets should go as they are regularly ignored as they tell pedestrian to stop when there is no traffic. They should be replaced by zebra crossings.

    The lights at the Burke St Mall are a similar proposition. I would not be adverse to a Collins St Mall either.

    Work could be done at Latrobe St to provide station exits on other sides of the intersection to reduce crossing.

    The CBD South station in the proposed tunnel may well turn into a Degraves St subway for Swanston St.

  3. Dylan Nicholson

    I heard a rumour that some intersections in Canberra have a 10 *minute* cycle time…anyone know if this is true? I can’t imagine it doing anything but driving people nuts….

  4. Austin M

    11 fair point on the short cycles … as long as the cycles don’t mean more time % then needs to be dedicated to the east west direction for cars. Ideally 2 trams should go through on each cycle as the pedestrian phase will largely dictate the phase’s length.
    The main issue would be the effect it has on pedestrians (This is Swanston Street walk after all so pedestrians should be the number 1 priority). How short can you make the phases and still give reasonable time allocation for large pedestrian volumes and particularly allowing for pedestrians with limited mobility. I imagine in 90 second 2 phase operation (taking out 6 seconds for each orange and all red time on each phase) that we are operating say two 39 second crossing phases. How much can this be reasonably lowered? If we go for 60 second operation this would be about 24 seconds of crossing time. A 45 seconds cycle would be 16.5 seconds of crossing (less than half the current crossing time). All this and how the signals integrate with the neighbouring signals on what would now probably be different cycle times (abutting intersections may also have more than 2 phases probably then needing at least 90 seconds).
    Worth looking into though. I wonder what is the shortest cycle time implemented in a CBD environment with a similar length for crossing distance.

  5. Tom the first and best


    Putting trams in tunnels is a bad idea because their task is to get people to local destinations and tunnels make that harder.

  6. Tom the first and best


    Cars are banned from Swanston St and therefore there is no need to allow for them turning. I believe that cars will still be allowed in George St so provisions for turns will still need to be made.

    A lot of the time trams is Swanston street is at lights. With the new super stops, the trams stops are opposite each other so not all stops are before traffic light intersections. Trams all have to stop at the stops, red light or green, so only 1 or 2 or at the very most 3 trams get through in a cycle so it is best to have shorter cycles.

  7. Dudley Horscroft

    A good idea in theory, until you start looking at the details. The section from Moore Park to the SE is good, though the line should not terminate at Prince of Wales Hospital – it should continue the short distance to Coogee – and the other branch should use Doncaster Avenue to avoid the Dacey/Alison Road Junction.

    From Moore Park to Central is a bit of a worry – the map indicates the line will use Devonshire St, but this has problems connecting across to Moore Park. Further north routes are one way streets. The better route would be via Greens Road, Oxford St and Wentworth Avenue to Eddy Avenue – a bit longer but also services the Oxford St area and probably faster.

    The big problem is what to do with the buses in George St. As you note, about 180 buses per hour, three every minute. Add these to the 120 buses per hour in Elizabeth Street and you have a situation worse than George Street! A ludicrous proposal. The solution is to do both conversions at once, the George Street to Norton Street/USYD via Parramatta and City Roads, and SE Sydney to Central and Circular Quay. Routing the SE lines via Oxford Street means that only a small amount of tunnelling would be needed to get from Whitlam Square to the unused rail tunnels via St James to the Conservatorium of Music, and thence into Bridge Street and Loftus/Young to the Quay. A guaranteed traffic free route with a convenient stop at St James and another at Martin Place/Parliament House.

    Given the known cost of tracklaying on city streets, and cost of trams, depots, etc, plus the cost of the extra sections of tunnel and clearing out the blockages in the existing tunnel, I suggest that this is where the cost of $1600 million comes from. Someone has looked at the costs of doing what is necessary and put that into the figure for just the SE line only. The problem will come if later people think that is the intended cost of just the SE lines and gold-plate the SE lines to suit this ‘mythical’ figure.

  8. Austin M

    I don’t remember to many right turns on Swanston Street last time I was in the city. Also considering a tram needs to pull up (unload and reload) at nearly every stop a 90 second cycle time doesn’t seem that bad to me when you consider the actual signal delay outside these load unload times (and you should probably consider peak loading for timetable reliability) I would be surprised if it was extremely excessive.
    With all that said maybe the heavy pedestrian demand or abutting intersections are dictating the cycle time and a 45 second or 60 second cycle could be trailed. The faster the cycle time however the more time in the hour the lights will be orange or red and peak throughput/average delay may actually suffer (in the city the peak throughput consideration may actually be pedestrians or the signal phasing may be a balance for all modes). I suspect the signals are not optimised for cars but are focused more on PT and pedestrians as a much longer cycle time would likely be used for car priority (i.e. 120-240).

  9. Tom the first and best

    Sydney should never have got rid of its trams!


    The 90 second light cycles in Swanston St are too long and thus restrict tram flow and thus should be reduced.

    Right hand turns could switched to hook turns or banned to allow there to be no right hand turn cycles.

  10. rachel612

    Why can’t we just buy the same trams Melbourne or Berlin or someone else uses? Why must we spend 3 times the price to have special “Sydney” trams?

    The Melbourne ones work perfectly well!

  11. IkaInk

    @Alan – Apologies, I scanned the document too quickly and got things muddled up. $133million per km is far more reasonable, but still horribly expensive.

  12. Smith John

    …. – It’s unclear what happens to western (Broadway) buses. P16-17 of the publicity document at http://haveyoursay.nsw.gov.au/document/show/601
    implies they could be rerouted via Elizabeth St. Certainly it would be most undesirable, and there is no need, to force western bus riders to change to light rail, as some have assumed (particularly the people who want to talk down light rail, including Infrastructure NSW in its recent plan).

    – Running George St to southeast, and Elizabeth St to Broadway, does have merit in allowing more transfer opportunities (including with trains) where the routes cross on Eddy Avenue. But this would have to be carefully planned and should probably include restricting general traffic on Eddy Avenue.

    – If eastern (Park St/ William St) buses are to head directly west, forcing riders to change for Circular Quay, it implies the need for much more pedestrian priority at the interchange points on Park St (of course, all the through traffic is in the cross city tunnel, isn’t it….). And/or locating stops so that an east to north transfer (and vice versa) is just a walk around the corner.

    – Capacity is said to be 30 vehicles per hour. It should be 40 (as on Swanston St), providing traffic lights are adjusted to 2-part, 90 second cycles like most of those in Melbourne’s CBD. This requires a strong policy to de-prioritise traffic movements that conflict with the light rail (2-part cycles are harder for turning traffic)

    – The route between Central Station and Moore Park should consider a route further north on or near Foveaux St, to expedite possible future extension to Bondi Junction via Moore Park Rd.

    – 24 minutes for a 6km trip with 7 stops on a ‘dedicated route’ from Central to Kingsford seems unreasonably slow. A congestion-free run with 1 stop per km should be looking at double that speed.

    The publicity document refers to ‘extensive feasiibility studies and consultation…’ As usual, none of the supporting documents appear to be public. The culuture of secrecty in NSW about what should be a completely transparent process is most regrettable.

  13. Smith John

    The project should be supported, with a few cautions:

    – It’s unclear what happens to western (Broadway) buses. P16-17 of the publicity document at http://haveyoursay.nsw.gov.au/document/show/601

  14. Alan Davies

    IkaInk #2:

    Nobody will disagree with your concerns about the ridiculous costs of projects.

    I guess we’ll have to wait and see if the $1.6 billion estimate changes while the business case is being prepared. I’m not aware of many infrastructure projects where the initial scoping cost proved to be correct.

    The 5.6 km line you refer to is the Inner West Light Rail Extension from Lilyfield to Dulwich Hill, along the old Rozelle goods line. It’s already under construction. It’s a piece of cake.

    The one I’m talking about above (the CBD and South East Light Rail project) is around 12 km of track, on existing roads with bridges and resumptions. Construction is expected to take five to six years.

  15. IkaInk

    Yes, $1.6billion seems extraordinarily cheap for a 5.6km light rail route. Why that’s only $286million per km!


    In reality this sounds like a good project, but Australia needs to stop allowing projects to escalate to these absurd costs. Hobart is a very different city to Sydney and costs will be higher in Sydney, but Hobart estimated costs at roughly $30million per km for a light rail route only 3 years ago. The magnitude of cost difference simply isn’t justifiable. Costs for PT projects need to be reigned in so that government can actually afford to build them, especially in New South Wales and Victoria.

    I know that you, Alan, believe that the Mandurah line is irrelevant to any discussion about current PT costs in Australia, but there are two very relevant points that must be considered:

    * The Western Australian Public Transport Authority is actually competent. Something that can’t be said of people in charge of public transport planning in New South Wales or Victoria.

    * The Mandurah Line’s final route was selected after considerable and extensive public debate which led to the selection of a route that excellent economic sense. Compare this to something like the Regional Rail Link, which appeared in no previous public transport plan, then suddenly was touted as Melbourne’s most urgent public transport project, which even the politicians involved admitted involved costing done on the back of an envelope!

    Competent organisations don’t suddenly choose huge infrastructure projects and build them without proper consideration. They plan new routes well in advance and refine them over time and then build them when funds become available, that’s called planning, not throwing billions in a blind reaction to problems that should have been identified earlier.

  16. Dylan Nicholson

    But will they call them trams, and more importantly, will they go ‘ding’?


https://www.crikey.com.au/2012/12/16/does-sydneys-new-light-rail-plan-make-sense/ == 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.