Friday, February 28, 2020

Useful Network No 39: May's new Endeavour Hills bus network

Late last year I proposed some ideas for a local bus network revamp in Endeavour Hills. Improvement is sorely needed. Local route run to limited hours and, particularly on Sundays, are infrequent with service typically only every two hours.

Several routes overlap one another yet there are no direct services to Dandenong's Hospital or TAFE precinct just a few kilometres down Heatherton Rd. And there are just three trips per weekday to Fountain Gate Shopping Centre. The map below highlights the issues.

My proposed network would have sorted all that with simpler routes that retained service to the vast bulk of existing stops and on nearby streets where routes were moved. And the upgraded services would have been well-used thanks to the favourable catchment; existing routes 843 and 849 get excellent usage on Sundays with over 30 passenger boardings per bus service hour. 

It turned out that the Department of Transport was making its own plans for Endeavour Hills buses. I first got wind that something was afoot when I contacted Luke Donnellan MP's office shortly after that post. They helpfully mentioned that changes were coming. 

News on these came out two weeks ago via Public Transport Minister Melissa Horne and Luke Donnellan (the local MP). This was on their Facebook pages but (unusually) not on the public transport minister's media release website (which provides an easier to find and more enduring record). No doubt better information will follow nearer the start date.

What can we learn from the above? Most notable is the new connection to Dandenong Hospital and Chisholm TAFE. That should mean a bus via a currently unserved part of Heatherton Rd (there's no other way). That will provide much better directness than now, eliminating a change in Dandenong for travel from parts of Endeavour Hills. Fountain Gate is also mentioned but that already has an Endeavour Hills connection via the infrequent 842. Plus there will be better frequency and new coverage of David Collins Dr. 

This Dandenong Journal article elaborates on this some more. Without a map not all details are clear. The map below is my best attempt at reverse engineering the network from the coverage seen. 

The above would be a 'smell of an oily rag' upgrade. Which is not a bad thing given the scope allowed by the large number of routes in the area. The four main routes from Dandenong (843, 845, 849 and 861) get reduced to three. Current service levels on each are every 40, 60 and 120 minutes for weekdays, Saturdays and Sundays respectively. 

A reduction to three routes should allow each route to operate approximately every 30, 45 and 90 minutes on those days assuming no extra funding. A little extra funding might allow every 40 and 60 minutes per route on Saturdays and Sundays respectively, along with improved hours. That would be a substantial improvement given how restrictive current timetables are. And an upgrade is needed to retain frequency on the Kidds Rd corridor in Doveton that goes down from 4 to 2 routes with the deletion of 849 and the rerouting of 861.

The above was written based on information publicly available at the time. Then two days ago Luke Donnellan's office sent me the media release which had more detail than the local paper picked up on. In particular frequencies on the 843, 845 and 861 will improve from 40 to 30 min weekday, 60 to 30 min Saturday and, biggest of all, 120 minute to 40 minute on Sunday. In other words better than speculated above on weekends. As well there will be two more off-peak trips return to and from Fountain Gate on Route 842. This should provide something like an hourly interpeak service rather than the two or three hour gaps as currently exist.

What are the implications of this change on the Useful Network (ie the availability of service every 20 minutes or better) in the area? Provided trips on routes 843 and 845 are evenly spaced the existing Useful Network in the area on Kidds Rd and Power Rd remains. While there is a small frequency drop on weekdays this should be more than compensated for by the improved Sunday service (20 min combined if 843 and 845 are evenly spaced) and (hopefully) extended operating hours. The simpler service (ie two more frequent routes instead of four less frequent routes) is particularly welcome since PTV does not excel at communicating frequencies on multi-route corridors. Also any longer hours will prove a godsend for those whose only other nearby service is the limited hours Route 844

How does this network score for local route coverage in Endeavour Hills? Removing the 849 would leave gaps if other routes aren't altered to compensate. The best clue that this will happen appears in the line below. 

The specifics though are not clear. My guess is that Daniel Solander Dr will be served by an 845 routed off Heatherton Rd (which would keep the 861 at an increased frequency).  And Scottsburn Way's 849 might instead be served by an 843 extended to Endeavour Hills Shopping Centre (removing the confusing mid-route dogleg there and providing a stronger terminus). Both seem sensible.  


Overall the changes proposed make sense especially if accompanied by needed operating hours extensions. These would improve buses in Endeavour Hills from below average to about average for a Melbourne outer suburb. 

Improved routes and extended operating hours will benefit many lives, expanding access to health, education, shopping and job opportunities. The minister would do well to implement similar network reforms in areas that need it including Dallas/Campbellfield, Hadfield, Craigieburn, TarneitSunshine, Frankston South, Karingal, the Mornington Peninsula, CaseyGreater Dandenong. Reservoir/Thomastown, Bayswater and Box Hill

Endeavour Hills itself could do with a second round. Good though it is, the above changes appear to leave 842 to Fountain Gate without weekend service or do anything about the new but poorly used Route 863. And the Useful Network remains sparse, with long waits if buses are just missed. Highest priority for a 'Stage 2' could be building on 861's rerouting to provide a new 'Useful Network' 20 minute frequency route along Heatherton Rd from Dandenong to Fountain Gate via Endeavour Hills along with an 844 operating hours upgrade and southern extension in Doveton. 

Tuesday, February 25, 2020

Timetable Tuesday #63: 706 the little shopper route that survived

One of Melbourne’s shortest bus routes is the 706 between Chelsea and Mordialloc. It is also one of the simplest, with about 90% of it running along Station St, parallel to the train line. The map is below.

The area map shows the 706 in relation to other routes and stations. It serves no major destinations the train doesn’t. It overlaps the much more frequent 902 SmartBus between Chelsea and Edithvale. Edithvale and Aspendale stations are already fairly close (about 1km) with the distance set to shrink when it is rebuilt and moved north as part of the level crossing removal.

However the 706 provides unique coverage in the residential pocket between the widely separated Aspendale and Mordialloc stations. If this did not exist then the 706 would have no rationale.

Official attitudes towards this have varied. At one time 706 was a much more comprehensive service. Service cuts in the ‘90s put an end to that. Then just a couple of years ago the (then) Transport for Victoria decided that the area between Mordialloc and Aspendale wasn’t worth serving at all. They proposed that the 706 be deleted with no other route to take its place.  

This was part of a wider Mordialloc area review that included an ill-advised splitting of a route popular with school children (708) and a dubious rerouting of another service (705) away from industrial area jobs. One of the stated benefits of this rerouting was to serve the Don Tatnell Leisure Centre (indoor pool) but it wouldn't have helped much as the peak-only 705 doesn't run when the pool would be busiest. Probably for the overall good the Mordialloc review sunk almost without trace. Except for some extra school time trips added to the 708's timetable. Not proceeding also meant a reprieve for the 706.

Could something else have been done? Unfortunately the area’s road network is a major constraint. Aspendale is the nearest station (as the crow flies) to the residential area of Aspendale Gardens and parts of Waterways. Unfortunately there is no direct road between them. If there was then it would have been possible to run a bus from Mordialloc to Waterways via Aspendale Station and Aspendale Gardens, providing good coverage with only minor duplication.


Route 706 is a very quiet route. It attracts six passengers per service hour. This is below what would be considered viable. On the other hand it runs during the off-peak weekday period only, meaning it would not increase the operator’s peak bus requirement. This greatly reduces its operating costs, especially if the bus used can form a service on another route. As you’ll see from the timetable discussed later it’s not much of a service but deleting it is unlikely to save much money. This is why I’m more tolerant of it remaining than in numerous other cases of duplicating and overlapping routes.


Route 706 is strictly a weekday shopper service with three trips per direction operating between the am and pm commuter peaks. There is a late morning, midday and early afternoon service. The timetable (in one direction) is below.

Mention should be made of the 706’s history. The route can trace its origins back to 1936. In its heyday through to the 1980s it operated many more trips, including during commuting times and Saturday mornings. Up to the 1990s its route was a little different, using Nepean Hwy instead of Station St at the Chelsea end. Cuts early that decade deleted Saturday service and lessened weekday operating hours. Then it disappeared, with two or three trips operating as occasional extensions of the Route 700. Hence at one time you could catch a bus from Chelsea to Box Hill and back again. Much later, when the 903 orbital replaced the 700 the Chelsea to Mordialloc portion got its old route number back but retained its reduced number of trips.


What would you do with the 706? Was Transport For Victoria right to propose its deletion given its low usage? Or would you have retained it given its likely small running cost? Is the Chelsea – Edithvale portion needed or would an amalgamation with the weakly terminated 858 be worthwhile? Please leave your thoughts below if you have any ideas.

Sunday, February 23, 2020

[FOI request] Inside Transdev's 302 and 304 bus upgrade business case

Last September I wrote about bus routes 302 and 304 from Doncaster/Box Hill to the City via the Eastern Freeway. These services were becoming increasingly crowded with frequent complaints from passengers inconvenienced by cancelled services and buses too full to pick them up.

Passengers got together to achieve the following:  

* A Facebook support group established to advocate for improvements. 

* Support for service upgrades expressed in Parliament and elsewhere by local MPs Tim Smith (Kew) and Paul Hamer (Box Hill)

*A Freedom of Information request from Philip Hodgkinson requesting documents in relation to issues with Routes 302 and 304

These efforts have had at least some success. PTV announced some extra peak services starting in November 2019. However this is only a stop-gap measure and more is needed.

Bus operator Transdev was very aware of issues facing the 302/304 during this time. They prepared a business case explaining why an upgrade was needed. This suggested how extra morning peak capacity could be easily and economically added. It was obtained through Freedom of Information a few days ago. 

Getting it proved quite a drama. Even though it's our money they're spending, the Department of Transport and its ministers can string FOI requests out for months. DoT's annual report (page 231) discloses how many requests were received but not how many were dispatched on time.

Network management in other cities

While I wouldn't swap Melbourne's extensive network for what most North American cities have, their transit agencies often have amazingly open attitudes towards administration and data. For example outsiders can read about what happens at transit agency board meetings or even watch proceedings on YouTube.  And cities like Portland, Oregon publish detailed quarterly cost and ridership reports for each route (below).

I suspect this is partly because there is more active management of transit resources elsewhere than here. Funding is more precarious, especially in the US. For example reductions in sales tax revenue can precipitate cuts in service. Not exactly conducive to having a stable service you can rely on.

US cities may seek federal funding for transit projects. They also use referendums to authorise increased funding on new lines or services. Winning such votes obligates transit agencies to publicly advocate for services and infrastructure in a way that our meek public servant types only do behind closed doors (if at all). US transit agencies rarely recover much of their costs from the farebox so receiving state revenue is essential. But they operate in a politically competitive environment with numerous other special interests pleading their slice of the tax revenue pie. If they don't advocate they don't get and may have to cut service (as has been common recently in the US and UK).

A different type of pressure applies where farebox revenue is more significant or you give bus operators more autonomy. Then it's commercial more than political pressure (although you can never entirely take the politics out). Imperatives to increase patronage are greater either because of increased direct revenue or contract incentives. Examples include bus operators up to the early 1970s (when public subsidies started) and the commercial rail franchises in the UK. Also to some extent services in New Zealand where a 50% farebox recovery ratio is mandated. In all these cases there is pressure to trim costs and adjust services in a direction that maximises profitability (often while also meeting politically decreed minimum service and reliability standards).

The Transdev bus franchise contract signed in 2013 is the nearest recent example Melbourne has had of this approach. This promised a new era for bus services. It has payment incentives for increased patronage, even mandating a 'greenfields network'. The contract hasn't changed but other critical things, such as the minister's willingness to veto a proposed network, were greater than imagined. Hence there has been only small changes since the minister rejected Transdev's 2015 greenfield network proposal (partly for sensible reasons).

Success here requires knowledge of a network's costs and awareness of opportunities to boost profitability or ridership. Plus a supportive bureaucracy and pro-reform public transport minister of the type we haven't had since 2014.

It's important though to be aware that maximum profitability does not always coincide with maximum ridership; a network that optimises profitability (ie overloaded buses) usually offers a poorer service than one that maximises patronage. However when wider community benefits of greater transit access are considered the latter is often preferable. Plus returns on capital are increased when your patronage increase is obtained by working your existing bus and train fleet harder.

An overall network view is also essential; having bus operators independently seek to maximise patronage may lead to wasteful duplication in some places and poor service elsewhere if they can  'cherry pick' profitable routes. Hence the need for a strong planning agency such as the original concept of PTV. Although even this was undermined by the separate (but ultimately ill-fated) planning Transdev was doing for its greenfields network.

Unlike in other cities, Melbourne's buses appear to be under little political, financial or commercial discipline to do better. Coming from consolidated revenue rather than being a hypothecated percentage of a particular tax, our funding has been more stable than in US cities. The main financial discipline appears to come every seven or so years when we renegotiate contracts with private bus operators (as done in 2018 after some argy bargy). Service is expensive to provide but interest in optimising resource use on both underproductive or overproductive routes is low, despite required data known to exist.

Our politics are also different. It lacks Americans' vigorous local campaign and referendum culture.  Modern bureaucrats are under the thumb of the minister with little independent professional authority unlike the VR or MMBW of yore. For a while it looked like we were returning to something like that with PTV Mk1 but this proved short-lived (although some great bus reform happened during this time).

Contracted operators tend to be fairly quiet (in public) and we haven't heard much lately from industry bodies such as BusVic. They had bus upgrade proposals before the 2018 election but these were not widely promoted or always well targeted. In a policy area like health industry or professional associations like the Pharmacy Guild, the AMA or the Royal Colleges would lobby for or against significant changes. Whereas in transport opinions from industry, engineering or planning professional associations carry little weight, especially when it comes to services (though you hear more from them when it comes to infrastructure).

The pressure on public transport ministers to perform seems less than in the troublesome 2007 to 2011 period but I suspect that it is building again following renewed issues with service delivery, outbreaks of industrial disputation and  over-runs on major projects.

With the possible exception of Transdev's contract (the oversight of which has had problems with a major fleet management crisis), we have had no enforced cuts that required savings to be found. Neither has there been a recent public-domain bus plan with ministerially endorsed service or patronage targets to encourage network reform. Without this urgency there has been little drive to pursue widespread cost-effective bus network improvements despite these being needed to meet the challenges of population growth. Instead our buses stagnate with barely half the patronage of trams despite them serving vastly more homes, jobs and destinations. And people miss out on the opportunities that a job-ready network could unlock.

Inside the business case

The first thing I'll say is you should read it. It gives an insight into planning and processes that's rarely seen. As well as background on other routes Transdev run. Also apparent is sensitivity over redacted financial information, especially when compared to what other jurisdictions allow to be published.

The business case starts by defining the scope. That's confined to the morning peak. The afternoon peak is not addressed. Neither are other issues concerning Route 302/304 such as the latter's short weekend operating hours.

A major reason for the 302/304 being worthy of a service upgrade is its strong patronage. It's more popular than most other Melbourne bus routes. Also noted is that overcrowding is leaving people behind at stops and affecting punctuality.

Rising complaint volumes are also cited. 8.6% of service delivery complaints relate to the 302/304 despite these comprising only 3% of Transdev services. That's a complaint rate of nearly three times the average. And it's rising as a ratio of all complaints Transdev receives.

The document then talks about how services can be quickly and economically added. There is criticism of PTV's convoluted service change process for adding timetabled trips. Transdev suggest bypassing this by adding 'ghost trips'. That is trips that run but don't appear in the public timetable. This can work on routes sufficiently frequent that they can be used as a 'turn up and go' service. Route 302 and 304 during the peaks would qualify for this.

One of the solutions suggested is to more intensely use buses used on school trips during school days. These could run short trips starting at East Kew in to the city and relieve some crowding.

It is here that we find our first redacted item. It is apparently a diagram of vehicle blocks. It is understood that information to do with bus scheduling and staff rostering is often considered commercially sensitive within the industry. This is because ingenuity here can give companies a cost saving edge when tendering for business. Bus operators regard this as part of their intellectual property that they would not wish disclosed. Cost information has also been removed.

A place where resources could be gleaned for additional 302/304 trips were the November changes to bus routes in the Elsternwick area (216/219/220). These routes were poorly used. The proposed network changes (leading to the creation of the new 603 and 604) would apparently save buses. I wrote about them here.

Route 273 was also mentioned. This operates every 10 minutes in the am peak and every 20 minutes in the pm peak. Usage is very low. Morning trips could be pulled off this route (saving two buses) to shift capacity where it is more urgently needed on the 302/304. This would have to go through PTV's service change process which the business case says requires a minimum of 9 months. That brings us to about the second quarter of 2020. So much for buses being a responsive and flexible form of transport!

To summarise the business case advocates for some weekday morning upgrades to Routes 302 and 304 to relieve crowding. It suggests 'ghost buses' to provide quick improvements. Later it suggests the use of resources from other underused routes to enable added timetabled trips. However these would take longer to add due to PTV's service change process.

You might enjoy these well-regarded books on transport topics

Better Buses, Better Cities: How to Plan, Run, and Win the Fight for Effective Transit Steven Higashide NEW!

The Public City: Essays in honour of Paul Mees Gleeson & Beza

A Political Economy of Access: Infrastructure, Networks, Cities, Institutions (Access Quintet Book 4) David Levinson

Human Transit: How Clearer Thinking about Public Transit Can Enrich Our Communities and Our Lives Jarrett Walker

Transport for Suburbia: Beyond the Automobile Age Paul Mees

(Sales links: I get a small commission if you buy via the above - no extra cost to you)

Friday, February 21, 2020

Pictures of the new Carrum Sation

The new Carrum Station entered passenger service on Monday. It's the latest milestone in a package of works that saw the construction of a traffic bridge for Station St across the Patterson River last year. Other parts of the project include moving train stabling south to near Kananook, the extension of McLeod Rd to Nepean Hwy (under the new rail bridge) and a new outdoor plaza for the area. 

It includes elevated rail, similar to the Dandenong line. This was a matter of some political controversy; some wanted the line sunk rather than elevated. Complications with the Frankston line included 'preservation of bay views' as an argument against an elevated structure and electoral politics since Frankston line stations served marginal seats like Carrum, Mordialloc and Bentleigh that were deciders on who formed government in 2010 and 2014. 

The result was that the Dandenong line got sections of elevated rail while Frankston line grade separations are mostly trench rail. Carrum though is an exception, with it getting elevated rail. Unlike neighbouring Bonbeach, Chelsea and Edithvale that are getting trench rail by 2022. 

Elevated rail has proved successful with some objectors becoming proponents. It's cheaper, quicker and less disruptive during construction than trench rail in many areas. It opened the ground area for other uses such as parks and recreation. And most of the political opposition vanished with grade separations being a major electoral plus for Labor in its large 2018 win. That result, despite anti-Skyrail campaigns, was so strong that the Frankston line seats are now held by Labor with large margins and several new seats in the traditionally Liberal east were won. 

Anyway below are some pictures of the new unfinished but in service Carrum station.

As the picture just above shows, Carrum is probably the station that has the best bay view from, taking that mantle from Brighton Beach. However its first week has seen some wild weather, with gales and heavy rain. Passengers scurrying to work or home might have appreciated it more if better and more continuous shelter coverage was provided along the platform, a design flaw that  dogs other recent stations such as Southland (Platform 2).

You might enjoy these well-regarded books on transport topics

Better Buses, Better Cities: How to Plan, Run, and Win the Fight for Effective Transit Steven Higashide NEW!

Breaking Point: The Future of Australian Cities Peter Seamer

The Public City: Essays in honour of Paul Mees Gleeson & Beza

A Political Economy of Access: Infrastructure, Networks, Cities, Institutions (Access Quintet Book 4) David Levinson

Human Transit: How Clearer Thinking about Public Transit Can Enrich Our Communities and Our Lives Jarrett Walker

Transport for Suburbia: Beyond the Automobile Age Paul Mees

(Sales links: I get a small commission if you buy via the above - no extra cost to you)

Tuesday, February 18, 2020

Timetable Tuesday #62: Frankston's loopy 770 and 771

Certain pockets of Melbourne have concentrations of confusing bus routes. Reservoir is one. Another is Frankston. A local bus review was done about ten years ago but only minor things got implemented. We may have been closer to some larger reforms but it got knocked on the head ('deferred' might have been the official term) at the last minute.

In the case of new bus networks a deferral is as good as a cancellation, government assurances notwithstanding. If there's a chance of getting a small network improvement through you should grab the opportunity as it may be a decade or more before the chance comes again.

Bus network reform quickened during Terry Mulder's four years as minister. Good things were done but more remained undone. It basically stalled under the infrastructure-focused Jacinta Allan. And the pace has yet to increase under current "the jury's out" minister Melissa Horne. The result is that ten years on the 770 and 771 continue how they've always run. Why is this bad? Keep reading!

Both 770 and 771 provide the main service from Frankston to the Karingal area. That’s important because while a frequent service (789/90/91) operates along Cranbourne Rd, there is poor walking access north of it due to hilly topography and pedestrian-hostile impermeable street layouts.

Consequently an internal bus network for Karingal, as provided by routes 770 and 771, is essential. The 770 and 771 are closely related routes. In fact some trips, such as to Karingal Shopping Centre, require you to get the 770 in one direction and 771 on the return trip. The map below shows them both along with the 777 (covered previously).

Both routes have large loops. Route 770 has a loop at its end with Karingal Shopping Centre about midway along the loop. Route 771 has a loop part-way along it, then a section of two-way running from Karingal Shopping Centre then a small loop at its end near its Langwarrin terminus. Also shown on the map is a 770 variant. This is an occasional service covering the Orwil St area very different from the regular Route 770.

You can see how it relates to the broader Frankston network from the map below. Interestingly it does not show the deviation crossing Frank St.


Route 770 gets above average usage for a Melbourne bus route with 29 boardings per bus service hour. Its usage is relatively constant throughout the week, with similar Saturday to weekday usage and Sunday dropping slightly to 25 boardings per service hour. Strong weekend patronage is quite common in coastal areas such as Frankston. Older, often working class, middle northern and outer eastern suburbs often have much less weekend trading and bus patronage, particularly on Sundays.

Route 771 is a bit quieter with 22 boardings per service hour on weekdays, dropping to 16 on weekends. This is probably because the route is longer, extending into low density parts of Langwarrin where car ownership is higher and bus usage is less.  


Both routes operate at close to minimum standards, that is 7 days per week until approximately 9pm. The main exception where service falls short is 770’s early last departure from Frankston on weekends (8:23pm).  

Headways are uneven. On weekdays Route 770 operates every 30 to 40 min during peak periods and around 50 to 60 minutes off-peak. Weekends are a flat hourly service. Two off-peak trips operate between Frankston and Orwil St – there is very little in common between that and the other Route 770 services except for the route number. 

Route 771 has a more consistent timetable. This includes a 40 minute frequency on weekdays and hourly on evenings and weekends. Even though it is the quieter route its last trips are later at night than 770’s last trips, with the difference greatest on weekends. 

Despite having roughly similar service levels, no attempt appears to have been made to evenly stagger departures from Frankston to provide a doubled combined frequency for those near both routes, even during peak times. The story is similar on weekends where, instead of offset by an even 30 minutes, the departures are offset by 20 and 40 minutes. This may have made sense when Frankston trains were every 20 minutes on weekends but does not now that trains run every 10 minutes. 


What would you do with 770 and 771? Would you swap them over in the Karingal area so that both routes are simple bidirectional services? Is there scope for better connectivity with the frequent Cranbourne Rd services? And should their timetables be tidied up to provide more even frequencies and offsetting to maximise effective frequencies for those near both routes? Please share your thoughts in the comment below.