Doncaster Bus Reform

As has been quite rightly noted, I haven’t written all that many posts on buses. So I thought I might have a look at buses for a change – more specifically Doncaster buses. As I’ve blogged on numerous times before, I favour the provision of heavy rail to Doncaster Hill (running at near metro frequencies) combined with the extension of the 48 tram and a complete reorganisation of bus services.

Unfortunately, these capital works projects seem unrealistic in the current political climate – Doncaster will receive only buses for the foreseeable future. So the question then becomes as follows: how can we make buses work better using current resources and what sort of investment is required to make them competitive against cars? This question fits in nicely with a hypothesis I’ve been looking to test – namely that bus frequencies could be substantially increased by axing unnecessary and circuitous routes and replacing them with fewer higher frequency routes running along main roads.

Testing the hypothesis for Doncaster

To see whether I’ve been on the right track with my suspicion about buses in Melbourne, we need to work out total route kilometres currently run in our sample area (in this case Doncaster and surrounds) and compare that against total route kilometres run on the hypothetical high frequency-fewer routes system. My methodology was thus: determine the length of the route and services run each way per weekday, then multiply these together to find out total route kilometres travelled per direction per weekday. This was a real bastard of a job and explains in part the long time between this post and the previous one – but I got it done and the results are presented below.


Length (km)

Services each direction per weekday

Total km travelled per direction per weekday


































15 *one way /2








































44(34.5)(28 )

















































Non M’ham routes



















As you can see, each weekday sees buses in Manningham (excluding) travelling 12611km each way. Including non-Manningham routes affected by my study, that figure increases to 13525.5km.

The Alternative

To test my theory, I came up with an alternative bus arrangement for Doncaster and surrounds, which is shown below.

This new system of routes probably has various errors, omissions and irregularities, and may not be the best way to run buses in Doncaster, it merely sets out the broader idea I’m trying to advocate, and I’m not going to die in a ditch over specific route locations. It’s certainly more simple than the status quo – have a look at the Manningham local area map on Metlink – it’s madness! The service standard I’m advocating is something I’m far more certain about – 100 services a day for most routes (with more on the main trunks). That works out to a bus every 15 minutes from 6 to 7am, every 10 minutes 7am to 10pm and every 15 minutes 10pm to midnight. That pretty much has to be the minimum to compete with cars. The total route kilometres from these routes and service standard is set out below.


Length (km)

Services each direction per weekday

Total km travelled per direction per weekday




































































Well the data didn’t really support my hypothesis – my new routes would see 22915km travelled per weekday, compared to 13525.5 under the status quo. The proposed extension of the green, yellow and red SmartBus into Doncaster, combined with Eddington’s $250-300m investment in DART (if it ever goes ahead) would tip the scales back in my favour somewhat, but I suspect not enough – especially given that I haven’t even looked at the even more woeful weekend services. I still stand by higher frequencies and fewer routes – but looking through this makes it clear that new investment is indeed needed on top of route reform. On the plus side, many of the service improvements I’m advocating are in off-peak, which means employing more drivers without the need for many more buses, so the new investment shouldn’t break the bank.


31 Responses

  1. What an excellent job – I enjoy reading modelled PT reform.

    One question you could pose is whether the DART is really needed at $300m. What is the annual stream of payments that represents? If that was put into building your system, including an excellent link to Box Hill and maybe a bus lane down Tram Rd, what it be more use than the DART.

    After all, an express train can do Box Hill to Richmond in 15 minutes and FSS or Parliament in 18, if buses came down Tram Rd into a purpose built interchange (like they did during the Laburnum works) could they not swap pax directly onto the trains, with a turn up and go of ten minutes, an average wait time of 5, a transit time BH to D of say 10 minutes, totaling 33 minutes to Parliament from D, and potentially saving the difference from the existing transit time AND the longer wait time for the current infrequent buses?

  2. I should add my system assumes from Shoppingtown or the park and ride – it would be less effective from Donc Nth or Templestowe

  3. Frightening Phin

    Go to

    If the map directions don’t work, enter Box Hill train station and Hotel Shoppingtown as the end points, and get map and directions from that.

    4.8 clicks and 7 minutes (and the directions selected look reasonable, including the side roads you need to use)

    Then look at the BH-Shoppingtown segment of the bus timetable say Donvale-BH via Shoppingtown.

    18 minutes!

  4. Thanks Ricc. I’ve wondered about DART as well – it seems Eddington is advocating higher frequency radial bus services – especially on weekends. $300m seems like a lot to pay for what we’re getting – but I don’t try to understand how Eddington gets his costs anymore.

    Agree that the emphasis should be on improving crosstown routes – especially for the southern part of Manningham. Shoppingtown – Box Hill is very important, and I can’t see why a bus should take more than twice as long as a car point to point. Routes to Blackburn and Nunawading could do well as well, but – as you rightly point out – the emphasis needs to be on low median wait times (5 minutes or less) and good quality interchange facilities. With smartbus considered by government to be a benchmark for crosstown buses (route 401 excepted), I can’t see them doing it though…

  5. The general frequency from Westfield to Box Hill on weekdays is presently quite decent for something in Melbourne – 6 buses per hour but an uneven service, usual offpeak wait 7-10mins. Weekends get useless with a 30min (lol) sunday frequency & the interchange from bus to train is poor. Takes roughly 10 mins from my experience.

    I like the way you summed up the kilometres – reminds me of the system in NSW where operators are paid for a set number of km’s by the govvy and have to design the times around that. Need some more peak services for a busy route – just cut back some local route in the off-peak!

    Only other suggestion is that 293 would be better going to Greensborough than Eltham (as it does now minus back street diversions).

  6. Sorry – correction to what I said, Westfield to BOX is 8 an hour (4 291, 2 293, 2 295) not 6. Sat is two 291, one 293 & one 295, Sun one 291, half a 295, half a 293 (lol – 120 frequency). Night = two 291.

  7. Yep. With the Hurstbridge line doing twists and turns, the closest point from Doncaster to that being Heidelberg you would want a fast bus (with it’s own lane) going there and interchanging quickly.

    I’m not against the freeway but just don’t know what $300m is supposed to achieve. Even a special flyover at Victoria Park and all Hursties stopping at VP before running express would be cheaper than this.

  8. I once did a similar (albeit significantly less detailed) exercise for the bus routes in Canberra, (which are tortuous verging on torturous). In my little model , I kept the numbers of buses constant, and was able to increase the frequency about 30-40 percent by straightening out the routes.

    I found, to my great surprise, that the greater walk times to get to the buses (since they were now running a lot further from people’s houses) undid a lot of the benefit i obtained by decreasing wait times and travel times. It was still better on average, but for people with high negative value on walking (e.g. the old and infirm) it would have been a bad move.

  9. Jason, I suspect that you had difficulties with Canberra due to their terrible street layout; you either run infrequent buses along local streets or frequent buses along major roads that are at least 20 min walk from most people. Canberra has few middle-level roads that are straight, direct, lined by shops, and have frequent junctions to side streets which buses need to get their pedesheds.

    My guess is that this inefficient road layout means that the ACT requires twice as much buses to provide a given service frequency within 10 minutes walk of most residents than a city with a walkable grid and through streets (eg inner northern and eastern suburbs of Melbourne).

    Manningham has many demographic and geographic simlarities to the ACT. And some parts have the Canberra cul-de-sac problem, as well as extremely low densities in sections Templestowe. But I suspect that its street layout is still generally more bus-friendly than most Canberra suburbs.

  10. Jason, I applaud your efforts to try and make sense of Canberra’s road network – your attempt to upgrade the only available public transport ‘software’ (bus route) changes were defeated by the defective ‘hardware’ (the road system). This is where the urban planning and transport planning decisions made decades ago (from Burley Griffin’s original monumental structure plans made in the shadow of Ebenezer Howard’s ‘Garden City’ and Daniel Burnham’s ‘The City Beautiful’, to the ‘neighbourhood units’ with the strictly defined urban units defined by a hierarchy of roads) affect the best (or worst) efforts to provide efficient public transport to Canberrans.

    Peter has hit the nail on the head of how the ‘neighbourhood units’ subvert the provision of efficient public transport through the road layout by discouraging through traffic through its cul-de-sac and windy road layouts, which also give the lengthy ped-sheds he describes.

    Parts of Manningham do share this layout, as their architects and planners learned from the same inspirations (Howard, Geddes, Mumford, et al) and were inspired (mostly) by the post-war British new towns, the American ‘Radburn’ and ‘Levittown’ planning concepts as well as the importation of planners and their planning experience and ideas that forcefully injected themselves into Australian urban planning in the 1950s (MMBW Plan, County of Cumberland Plan, National Capital Development Plan).

    All of which make it hellishly difficult to provide an efficient, regular, attractive bus service to Canberra and similar, low density parts of Melbourne.

    Good to see you spreading your wings Phin. Well done.

  11. I have wondered about those post-war new towns as well as the 70s street calming and cul-de-sacing.

    I suspect some trends that came from the ACTUAL market, not planning, ran against these plans.

    First, the plans assumed Cadbury-style company town decentralisation, so that jobs and so on could be created locally. The ‘man’ the breadwinner would be the only one working so the road network only had to deal with him.

    The Cumberland plan assumed electric trains to Penrith would manage well, because they would be equivaelnt to today’s interurbans and would only be carrying a moderate number of genuine interurban/intercity commuters.

    The death of manufacturing and 2-income families has definitely put more cars on the road.

    Second, the nonsense about telecommuting has got ahead of reality and you still see some of the Mcmansion advocates saying “Go and buy a big house, then lock yourself in it all day and work”

    Which defies not just the current capabilities of technology, but people’s desires for socialisation.

    And I don’t think a road heirarchy has yet been designed that can handle traffic beyond 2000 lanes an hour. Freeways are limited by their exists. Conventional roads are limited by their intersections. Curvy roads might slow the traffic down but then that just takes more cars and longer journey times to cover.

  12. LS: I believe Burley Griffin intended a city based on trams. And the older parts of Canberra (North Canberra and South Canberra) aren’t too badly planned for public transport.

    The rot really set in from about the 1960s when Woden, Belconnen (and later Tuggeranong) were developed.

    One of the paradoxical things about the later versions of Garden City & (especially) Radburn ideas is that they were based on ‘cities for people’ and taming the car. But in doing so they killed prospects of efficient, direct and fast public transport.

    Some of the late Garden City ideas were a backlash against the then univesal ‘Sydney Roads’, and aimed to calm traffic by making local roads windy.

    This left the stage open for the freeway builders who said ‘let us take your through traffic and put it on our new controlled-access highways and freeways’. Without roadside shops and frequent intersections, we can even plant trees along the roads and call them ‘parkways’, as they are in Canberra.

    For a while it looked like win-win, with the local suburbs freed from through-traffic and faster car travel between suburbs on the new highways. The Canberra housewife would still walk to her local shops, which was planned around suburban units of about 3000 with its own primary school. Buses got her to a nearby town centre, but cross-suburban travel was (and is) woeful.

    About the only grids were provided by major roads 2-4km apart. While these are potentially good public transport routes, they have no shop frontages, and as Jason found only a small fraction of residents are within walking distance of one due to the paucity of intersecting streets. Like freeway buses, they allow frequent service that serves almost no one.

    Then there’s Radburn. Again it’s a reaction against busy multi-purpose streets like Sydney Rd which perform a variety of functions thought conflicting (eg transport corridor, shopping, trams, pedestrians etc).

    While such roads (especially if they’re in fashionable areas) like Glenferrie Rd, Chapel St, Glenhuntly Rd are today trendy and sought after they weren’t years ago. Arthur Boyd described them as boring tram-lined streets with incessant intersections or similar.

    Radburn basically tried to seperate car and pedestrian traffic by providing a seperate system of walkways. This was sold as a triumph of walkable/human suburbs against car travel. It proved nothing of the sort and only entrenched the latter.

    Yes the cars might have been calmed on their culs-de-sac. The cyclist might have benefited from their road underpass.

    But without the middle hierachy of straight, direct roads that allow a suburb to be served by one direct frequent bus route instead of three circuitous infrequent ones that take 20 minutes to traverse one suburb, buses drew the shortest straw of all.

    Especially where Radburn developments were done as part of social housing projects or low-income suburbs, seperate pedestrian paths were considered unsafe for pedestrians (who had no passive surveillance from homes or even cars). Chalk up another trip for Mum’s taxi instead of the kids walking home!

    It turnes out that Radburn was misguided, as ‘new urbanism’, ‘designing out crime’ principles and now environmental conciousness and resource resilience is now proving.

    Plus a Radburn echo in the shopping mall resulted in a pedestrian space surrounded by acres of parking; turning its back on its surrounding potentially walkable neighbourhood.

    But what man can mess up man can fix.

    In a way we are lucky we have population growth since we can channel this to redevelop our existing suburbs as well as doing a better job with our outer areas. This need not involve massive increases in residential densities or wholesale street changes as is often claimed.

    Just like water tanks or solar hot water are considered desirable points for new housing, walkable suburbs with economical to serve public transport corridors should be promoted.

    Buyers should be taught to look at a proposed development and recognise the sort of street patterns not conducive to public transport, and that a wrong choice could lead them to live in an unserved ghetto like Sandhurst Estate.

    Imagine if Gowanbrae had a road link to Glenroy or Broadmeadows, for example; it would not have had to wait until 2008 for public transport, and a high-frequency bus from Airport West would have been well-used, economical to run and provide a useful cross-suburban link for tens of thousands.

    In the meantime we should not passively accept that 1960s – 1990s suburbs like Rowville, Hoppers Crossing, Wantirna, Lysterfield or Gowanbrae are condemned to be cul-de-saced forever.

    It would not be futile for transport departments to be planning how these suburbs could be opened up to provide workable pedestrian and public transport grids, building new walkways, one-way bus lanes and selective new roads.

    Google Earth provides opportunities for the initial research to be done from the office. In some cases houses at the end of culs de sacs may need to be purchased, resumed and demolished. In others it may be possible to resume only part of a block leaving the house intact or a block sufficient for two units or a small park. Other cases no resumption of houses may be required.

    At worst the purchase of two outer suburban houses might cost $500 – 800k. However if it allows two inefficient local routes to be reduced to a better used, faster and more direct route then this would be money well spent. As a bonus, land resumption should be cheaper in the poorer suburbs, where the social need for improved public transport is greatest.

    These are the sorts of things we should be looking at closely to make public transport work. Vicroads certainly does with long-term freeway plans even shown in the Melways. While resumptions appear to be beyond the scope fo the existing bus service reviews they should be on the table where clear benefits of improve services are apparent.

  13. With regard to Canberra – the NCDC based Canberra on being a collection of neighbourhood units segregated by the arterial and sub-arterial roads.

    The size of the neighbourhood unit was entirely dictated by the private vehicle – the limit was about 4,000 to 5,000 people iirc before the traffic noise on the collector roads was deemed to be too inconvenient for those living on them.

    Note that any rat-running was also discouraged by not provided direct access across the arterial roads – rather the intersections are staggered in the worst possible way.

    Griffin’s Canberra did incorporate tram lines but any blog reader who has seen the plans would have to agree with me they would have been a dismal failure today – they were an operational nightmare (trunk and branch inhibiting frequency improvements on the branches) and the branches neither had the critical mass or the directness to make them what we would now call viable. I should give the guy a break though – he did plan it in the 1910s!

    Very interesting experiment Phin – this is probably my favourite of your posts thus far. An idea for a future post might be to compare providing the minimum service frequency (to compete with cars, not the DOT’s pathetic service standards) on the current network structure and provide on the streamlined network. Given the streamlined network would win this pants down – it might be prudent to introduce Jason’s walking distance factor into the equation. You might then find that the token and community transport type services would still need to be run to benefit the less able, without screwing the rest of us over with the inefficient use of limited PT money.

  14. Yes, please do some more bus-related postings.

  15. Great job (probably better than I could do).
    A few suggestions:
    Run all the 300 services to Mitcham via Shoppingtown unless you plan to extend the A route.
    548 is just a route that passes along the 72 extension. Until then it should run to Camberwell, and after the extension it should be disbanded to fund other routes.

  16. Riccardo, I like your idea of having Doncaster buses run to Box Hill – it gives the necessary push for off-peak and weekend express services. 😉 Which is something our planners can’t seem to grasp (along with electric interurban services to Geelong, third tracks without third platforms, etc etc). The trouble is fitting all those services down three tracks – probably the only way would be to have express services change direction at noon like the city loop does.

    Of course a heavy rail line will be needed sooner or later (Box Hill trains will get full before we can achieve a significant modal shift away from private cars).

  17. The idea of having off-peak express services only operating in peak direction sounds useless and adds to the present notion that “everybody wants go to to the CBD” – yuck.

    It already works okay in the afternoon peak when there is both local trains from Blackburn & express trains from further out heading back to the CBD with only one up track.

    I think it could work but if you tried to run anything more than a 10min express/15min local (or v.v) frequency it could get a tad congested. Could reverse the direction of the middle track at noon, but you have to keep both express & locals in both directions!

  18. I think the trains can operate on an “everybody wants to go the CBD” notion for a while yet – a lot of people still do, even offpeak. And more services (esp. offpeak) encourages more usage, which in turn may rise to a point where demand for orbital services leads to new investment. Until then though we should be trying to do more with what we have. (I’m not saying cross-town services should never be on the cards, just that there is a tipping point effect here – need PT to be a 1st choice for general travel =>crosstown demand, whereas now offpeak PT is definitely not a 1st choice).

  19. “A lot of people” still do, but only 15% of Melbourne residents go into the CBD each weekday (according to teh PTUA).

  20. The new Laburnum has space for a third track between the other two without space for another platform.

  21. Phin,

    bad news!

    I overlooked a woman on the train this morning who was working on stuff she appears to have taken home from work – maps, plans and timetables for the Queenscliffe and Ocean Grove GTS routes.

    from what I could see, she wasn’t applying any of our learning to her methods, her scribbles seem to indicate they make it up as they go.

    Some of the maps were for extremely convoluted routes

    What can we do about this?

    Should I have spoken up, offered to give her a couple of days off work and got the team in here working on it?

  22. I know I should be minding my own business but this is too important 🙂

  23. Should have a business card with or and dropped it in her lap…. or else the website of the ‘consolidated’ postings! : )

  24. The “every house must have a bus within 400m” rule is what kills bus planning in Melbourne.

    It’s treated as a welfare service to provide a basic service for those who require transport and not an essential part of a city. Hence the result of many infrequent & indirect routes instead of a network based on providing an efficient service. For a laugh look at this map:

    Riccardo, if you see her again you should butt in and refer her to this blog even if it seems a tad rude.

  25. That 682 route is either really funny or really depressing… I’m trying to make sure I see the funny side!

  26. Riccardo, my guess is that the person you saw has a task of documenting ‘what is’ rather than planning ‘what should be’.

    Route reform will probably have to wait until that area’s local bus service review.

  27. Hope so Peter, but what can you do to get DoT people thinknig the way we do?

  28. Ricc, my recipe for this is as follows:

    1. Write submissions for 1 (or preferably all) bus reviews. Start as soon as the review is announced. Include an analysis of the current situation, identify network-wide issues, make specific recommendations on routes (including maps).

    2. Find out where and when first round meetings are being held. Have your submission about 90% complete by then and bring it along for reference. Representatives from DOT, bus companies, consultants doing the review and local government will be present so there’s the chance to talk to them, especially during breaks.

    3. Seek to use the meeting to encourage others with good ideas to put these up and win support for your ideas.

    4. Finish and lodge submission, modified to include good ideas heard from 3 above. Make sure you include contact details!

    5. Because you put in a submission you’ll be invited to the second round meeting where the consultants & DOT come back with their draft proposals. Another opportunity to talk to DOT people and others!

    6. Repeat until all areas have been done and attended.

    Generally there is widespread support for simpler routes, longer hours and better frequencies. It appears that the 90%/400 metre rule and MOTC span & frequencies (hourly until 9pm 7 days/week) are the firmest rules there are.

    As blogged before, this pattern optimises the basic access (ie social obligation) but not overall service levels or patronage compared to a model of more frequent but somewhat sparser routes.

    A key message could be to convey the view that people are willing to walk further to a better service and that 15/30 min is far better than 30/60 min. Or, more radically turn-up-and-go. I’ve found members of the public at these things more informed and less parochial than feared and you’d be pleasantly suprised at the number willing to accept trade-offs for a better service.

  29. Peter,
    Your ‘how to’ on getting input into policy with the metro bus reviews is really positive stuff and as constructive and evidence based as always. When your contract at the other place is up, you should come back to the other, other place – you are missed.

    The people who have put together the concept planning for the metro (and now regional) bus reviews have managed to do a number of things right and do them better than they have been done for a long time, namely:
    1) Involve the citizenry in the plan-making stage of planning rather than the plan-implementation stage
    2) Listen to the concerns of the citizenry at the consultation stage and pick the gold from the dross and invite the constructive citizens back (as opposed to the monomaniacs, foamers and axe-grinders)

    Of course, it helps at the moment that the mood of the times has swung around toward a greater level of public consultation in transport planning (as in the bus reviews) and a greater willingness of citizens to publicly engage with the government (I’ve been to a couple of the Garnaut session in Sydney and Melbourne – great turnouts and lots of intelligent questions, as well as seeing the high turnout of ordinary citizens for public consultation for my LGA’s local structure plan).


  30. I have noticed you don’t monetize your website, don’t waste your traffic, you can earn additional
    bucks every month because you’ve got high quality
    content. If you want to know how to make extra money,
    search for: best adsense alternative Wrastain’s tools

Leave a Reply

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

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

Google+ photo

You are commenting using your Google+ 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

%d bloggers like this: