Tuesday, September 28, 2021

Transport lessons from the Bracks and Brumby governments (via Catch & Kill)


I've just finished reading Catch and Kill: The politics of power by Joel Deane. It's an account of the 1999 - 2010 Bracks and Brumby state Labor governments mainly based on interviews with participants. 

Against most expectations Labor won enough seats to form a minority government after the 1999 election. The previous Kennett Coalition government was praised for having reduced state debt but its asset sales and spending cuts in health, education and (to a lesser extent) transport had hurt. When the economy revived the benefits were seen as flowing to Melbourne rather than being spread across the state. Labor tapped into this sentiment with the famous 'two taps' advertisement.   

The result was a large swing in seats around large regional centres like Ballarat, Bendigo and Geelong. This was enough to deliver government with the help of regional independents who were upset with the Kennett government. 

Transport was intertwined with this. The Kennett government had famously closed several regional rail lines. In 1999 Labor promising a revival program, involving reopenings and 'Regional Fast Rail' to the major centres. It was politically popular and helped win regional seats.

The Bracks government was re-elected with a large swing to it in 2002. This allowed it to hold office in its own right. It retained power in 2006 with a reduced but still comfortable margin. 

However the regional rail program's cost blowouts, a view on what was considered fiscally responsible, surging population growth and poor advice from the Department of Transport sowed the seeds for a metropolitan public transport crisis that the (now) Brumby government was ill prepared for. 

Warning signs were in Track Record statistics as early as 2004. Metropolitan trains were getting less reliable and more crowded each year with no end in sight. However train performance issues were ignored year after year in budget after budget. 'Blame the private operators' was the line.   

This was partly due to a 'hands-off' franchise mentality that started under Kennett and solidified as conventional practice in the early Bracks years. Despite the frequent rebranding of trains and trams, a key aim of rail franchising, ie to allow the government to shirk responsibility of ensuring a reliable service, was never fully accepted by increasingly grumpy passengers. 

When the Brumby government did decide to act it was too little and too late. Promises of grand infrastructure did not address the reality of lateness for work, missed appointments and childcare penalty fees caused by increasingly frequent rail network meltdowns. 

Taxpayer-funded advertising campaigns only added insult to injury. Hence in 2010 marginal seats (particularly along the Frankston line) swung towards Ted Baillieu's Liberal-National Coalition to end eleven years of Labor rule. 


That's the broad sweep. For the details of who thought what when, here's a summary of the transport aspects of Catch and Kill. 

P157: Linking regional Victoria to rail went against the ‘natural inclinations of the bureaucracy’

P158: Bracks staffer who commuted from Ballarat (Dan O’Brien) gave Bracks idea for Regional Fast Rail. Rationale was spreading benefits of growth outside Melbourne. 

P158: RFR costs kept rising. 

P176: Treasury opposed putting significant transport goals into Melbourne 2030

P176: Assumption that people would blame Kennett’s privatisation for train problems when it was seen as a public service. 

P176: Peter Batchelor said that govt’s fundamental error was underestimating patronage growth.  

P177: John Brumby said that cost blowouts on RFR slowed metropolitan rail upgrades. 

P177: Bracks said that there weren’t many parts of govt where they got inaccurate advice but PT was one. Treasury consistently underestimated population & jobs growth.

P178: Bureaucracy and political staffers dismissed PT’s political importance as they thought only 10% of people used it. A pervading view that roads more important than rail. 

P181: Overarching theme of financial responsibility, balancing the budget and not being captive to Canberra (strong finances allowed Victoria to fund some things alone eg the Synchrotron). 

This limited spending on things that needed it – eg metropolitan train services – until too late. 

P181: DoT very road focused. Did not believe PT patronage would go up. 

P181: Opinion polling (GAMS) led Labor to a false sense of security. PT not seen as top priority and Labor were thoughts better managers of this than Coalition. PT not seen as vote changer. 

P192: Discusses National Express walkout. Had to find $400m. That took the entire spare amount Treasurer Brumby had set aside for other areas. Hence breaking a promise and building Eastlink as a toll road. 

P205: 2003 budget did not buy new trains. But it allocated $11m to bus services. Brumby said they put a lot of money into buses. But saw no political dividend – “not sure if people associate buses with the state govt” Note: at the time buses had private operator liveries, not Metlink as it could have been then nor PTV as it is now. 

 P283: Lynne Kosky was good minister but had bad fortune due to rail meltdowns and myki (which was high risk being an IT project).

P283: John Brumby said that core of myki problem was DoT gave wrong advice on awarding tender (KAMCO who had no experience). Regretted not challenging this. 

P283: Sharon McCrohan was critical of Rob Hulls for not seeking a more challenging/tougher portfolio like public transport. 

P308: Premier’s speech in late 2008 announced major rail infrastructure projects (but too late)

A continuing theme was that the government was (or thought it was) poorly served by advice from  the transport bureaucracy. Notable examples include its tepidness regarding regional rail, the underestimation of train patronage growth and of course myki (where, in retrospect, it is hard to argue that the contract wasn't wrongly awarded). 

This patchy record makes premier Daniel Andrews’ use of external advisers to develop the Suburban Rail Loop instead of the Department of Transport more understandable, despite the internal toes it trod on. The same might apply to any minister wishing to pursue radical projects where existing processes are wanting or incumbents may hold unhelpful or obstructive views. 

As an example Jeff Kennett's rail splitting and franchising was implemented by a 'Transport Reform Unit' in the Department of Treasury and Finance. As opposed to the department which could carry attitudes and people from the previous Cain and Kirner years including limitation on what might be possible due to union power and labour movement factionalism (which brought down the premierships of both John Cains). 

One could argue (as the state Auditor-general did) that the Transport Reform Unit achieved its objectives of cutting costs and even improving some aspects of performance. However The TRU's  failure to see the network as a whole led to a wilful fragmentation of the network that subsequent governments had to waste time bringing back together.  

If Minister Carroll wants more than glacial progress on bus reform, he will need to oversee massive increases in the capability to deliver. Currently bus reform is run as a slow moving small scale 'cottage industry', with reforms to only a few of Melbourne's 350 bus routes per year. Unreformed bus networks operate in about two-thirds of Melbourne's suburbs and some neighbourhoods have been without even hourly 'minimum standard' buses for 30 or more years. 

Transport internal processes are so borked that it takes about as long to add a new bus route as to do a major capital works project like grade separate a level crossing. Even timetable upgrades that merely work the existing bus fleet harder on an unchanged existing route take much longer than they should. 

Accelerated delivery could either be through a dedicated LXRP-style unit (good for focus but carries a risk of fragmentation) or a massive increase in the  DoT's ability to deliver. There is currently a process of recruiting more people to the part of the Department dealing with bus reform. Internal process improvements and possibly redirecting resources from other areas could further augment capacity.  

Anyway that's straying off today's topic. What doesn't the book mention? I'd have liked the book to cover various promises, schemes and expectations that were raised in 1999 but did not happen. For example we still don't have passenger trains to Leongatha or Mildura. We didn't get a tram extension to Knox (only half-way). Expectations of rail to Rowville were raised but not met.  Ditto for a third track to Dandenong for express trains.

I assume that these projects did not happen due to a wish to be seen as fiscally responsible in order to exorcise the ghosts of the Cain and Kirner era. Public transport would likely have been considered a money sink, particularly due to project cost blowouts and the abovementioned National Express walk-out

In contrast the policy context had changed by 2014, where expansive (and expensive) 'big build' infrastructure programs were back in fashion with the 2010 coalition government seen as doing too little. The Coalition still used the 'financially incompetent' line to bash Labor but it was tied to specific projects (eg myki and the desalination plant) rather than more generally. Also, by 2014 hostile voter memories of Cain / Kirner had dissipated compared to more recent memories of Bracks (who was either liked or generated no serious revulsion even amongst opposition voters).  

If you have any interest in Victorian politics then you should definitely buy Catch and Kill. If your interest is more narrowly transport policy then it's still worth reading as it gives you the reasons for some key decisions being made (or not being made).


You can get Catch and Kill via various websites including Amazon. Using this link supports Melbourne on Transit (at no extra cost to you).  


Friday, September 24, 2021

Building Melbourne's Useful Network Part 106 - Point Cook and Sanctuary Lakes network refresh

Melbourne has many local bus network that are just layers and layers of confusing routes and deviations. They exist due to the slow pace of bus reform. The silver lining though is I rarely run out of things to talk about on Fridays with revamped networks often proposed. 

I'm taking a different tack today. Today's topic is Point Cook in Melbourne's south-west. Unlike most established suburbs in the north, east and south it has had a lot of recent network reform. Point Cook's bus routes and timetables thus lack many of the legacy problems like indirectness, confusing deviations, wasteful overlaps, lack of Sunday service, non-standard public holiday timetables, etc that continue to dog unreformed bus networks in much of Melbourne today. 

That's not to say that Point Cook's bus network is perfect. It isn't. Continued suburban growth, high car traffic volumes and the paucity of routes in and out of the suburb has exacerbated issues. So much so that the community has on occasions banded together to run services the state apparatus has not seen fit to do. More on that later. 

The pre-Williams Landing Station network

Before Williams Landing Station opened the nearest stations to Point Cook were Aircraft and Hoppers Crossing. Both the area's two bus routes ran on roughly east-west indirect routes between them. Services were infrequent, delays were rife and huge new subdivisions to the south were being built with no nearby bus service. 

The first Williams Landing Station network

The two-route network had a major redesign in 2013 when Point Cook's new nearest station at Williams Landing opened.  413 and 416 were discarded. In their place were five new routes (493, 494, 495, 496, 497). Four (494, 495, 496, 497) were largely north-south routes. All five but the 496 fed into Williams Landing Station. Peak frequency on most increased to 22 minutes, chosen to meet Werribee line trains (typically every 11 minutes). A map from the Star Weekly article is below. 

This was almost exclusively a commuter network intended to get people to the new station. It did this very well, subject to delays caused by car traffic. Williams Landing has one of the highest proportion of commuters arriving by bus with the number being about that who drove to the station. This is an amazing result for Melbourne, a city that has traditionally been poor at planning efficient coordinated rail feeder bus services. 


The new network however wasn't so good for local trips. These are largely east-west involving destinations such as Hoppers Crossing (including Mercy Hospital and Suzanne Cory High School), Werribee and Werribee Plaza. The shops near Aircraft Station were also difficult to reach with buses removed and a treacherous roundabout preventing easy walking access. 

Williams Landing, where most buses went, was then a station without a town centre so it was a place where you'd get the train rather than being a useful destination in its own right. 493 (orange in the map above) provided some east-west travel but was distant from Point Cook South. It also had a dead-end terminus and forced people going to Suzanne Corey to 'go the long way around' with an inconvenient change. Also the design of the 496 cut Seabrook residents off from everywhere but Central Square Altona Meadows. It didn't even run to their nearest large shops at Point Cook Town Centre. 

Even though the new routes carried more people than the old ones ever did there were local calls to bring back the old network. Local MP Jill Hennessy (then in opposition) put out this release and the video below. 


There was enough momentum for change to amend the Point Cook bus network as part of wider Wyndham area network reform in 2015 (again linked to a rail upgrade - this time the Regional Rail Link via Tarneit). 

These changes kept the integrity of the commuter-friendly 2013 network while adding elements that improved local travel. The two key changes included (i) Extending 496 to cover Sanctuary Lakes (a low density golf course based development and (ii) replacing the 493 with the new 498. Map below. 

The 498 provided a simple east-west route that the 2013 network didn't do well. Benefits included: (i) restoring two routes to Seabrook, a direct new connection to the Point Cook Town Centre and a restored Hoppers Crossing link, (ii) more direct access to Suzanne Cory High School from more areas, and (iii) new coverage for the new and dense western Point Cook area. Even though it initially ran only every 40 minutes in the peak the 498 was well used to the point of being unable to pick up more passengers.   

Continuing issues

Point Cook was still growing. Train passengers were catered for with improved peak frequencies on routes like the abovementioned 498 and the popular 495 to Williams Landing. However there were many areas in the south that the 498 to Hoppers Crossing didn't help very much. 

Parents in the area were particularly concerned about continued poor access to Suzanne Cory while others would have wanted bus access to Hoppers Crossing. Not just to connect with the train but also to reach local jobs and shopping at places like Werribee Plaza. 

Just before the 2018 election the government announced a 'Pick My Project' scheme where people could nominate and vote on community-enhancing initiatives. One of the winners was Wynbus. This was a Point Cook community-led initiative to run mini-bus services in areas that the existing network was lacking. 

Two of its first three routes connected the area of Point Cook around Sneydes Rd to Hoppers Lane (though not Hoppers Crossing Station). Read my write-up on Wynbus here. Wynbus has since moved to other projects. But the demonstration indicated that others saw the gap in east-west travel from southern Point Cook. Especially when a road that could have supported a direct route existed. The high productivity of fixed routes in areas like Point Cook indicate that improving these rather than flexible routes is the path to patronage success, particularly on main roads. 

A refreshed network

What might an upgraded Point Cook and Sanctuary Lakes bus network look like? Here is one that continues the evolution of it from being largely a commuter-oriented train feeder network to a more comprehensive and more broadly useful grid. It involves changes to one existing route (496) and the creation of one new route; no route changes for others like 494, 495 and 497 are proposed.


Idea 1: Sneydes Rd bus. The big change is a second east-west link to serve central and southern Point Cook. This is achieved by creating a new Laverton - Hoppers Crossing route via Sneydes Rd. I've labelled it as the 499. It would run as per the existing 496 alignment from Laverton to Sanctuary Lakes Shopping Centre then westward via Sneydes Rd to Hoppers Crossing. It could even potentially continue to Werribee Plaza. 

This alignment has multiple purposes including better connections to destinations in the Hoppers Crossing / Werribee area and as an alternative train feeder to Williams Landing. Access to Sanctuary Lakes Shopping Centre is improved and there is some extended coverage in the south-west Point Cook area. A discussion on an earlier version of this idea is here

The 499 is close to the same length as the existing 498, serves the same termini (Hoppers Crossing and Laverton) and would have the same service level. This gives scope to stagger its times to provide an even 20 minute off-peak and weekend frequency along the 498/499 common sections. These include the Hoppers Lane, Seabrook and Altona Meadows portions.  

Idea 2: Sanctuary Lakes direct bus to Williams Landing. :Having the 499 follow the 498 through Altona Meadows means there is no need for the 496 to Sanctuary Lakes to do that as well. Hence it can be shortened to serve Williams Landing rather than Laverton Station. Using Williams Landing station provides a quicker trip to the train at a nicer interchange station. This should make it more car-competitive for commuters. That's important as Sanctuary Lakes people do own a lot of cars. The 496 could retain its current hourly off-peak frequency. But if if was boosted to every 40 minutes an opportunity exists to offset its timetable with the 497 to provide an even 20 minute frequency between Williams Landing and Sanctuary Lakes Shopping Centre.  

Idea 3: Service upgrade for 494/495. Point Cook's two most productive bus routes are the 494 and 495. Not only that but they are also very productive on a Melbourne scale. Currently they run every 40 minutes each off-peak. Extending operating hours and boosting both to every 20 minutes would provide a much more attractive service, especially if it is combined with 10 minute off-peak train frequencies at Williams Landing. In conjunction with the coordination mentioned in Ideas 1 and 2 large parts of Point Cook would get off-peak buses every 20 minutes as opposed to the current 40 minute gaps. 

One objection to the above might be that the new 499 doesn't add a lot of extra bus coverage. Instead its main effect is to provide an alternative that may be more convenient for some trips (especially local). However it should be noted that Point Cook's north-south road grid (especially between 494 and 495) means that some midblock pockets are more than 400 metres from a bus. And the existing network performs well on patronage grounds with good responsiveness when services are increased.

The area also has significant traffic issues due to high car usage and a limited number of roads going in. Hence  better buses, including improved access to local destinations, should be considered a tool to tackle this. Lastly, even after you add the 499, the network is relatively simple with overlaps mainly in areas where they could usefully provide a more frequent service. 

Conclusion

Described has been an upgraded Point Cook / Sanctuary Lakes bus network. It adds one new route and amends one other for greatly improved local connectivity. Comments on it are appreciated and can be left below. 

Index to all Building Melbourne's Useful Network items here

Tuesday, September 21, 2021

Timetable Tuesday #135: Night Bus 970 (a video retrospective)

A little different today. Instead of a whole lot of words it will be a video featuring the Route 970 Night Bus. Its last trip ran on Sunday departing from a terminus near home so it was too historical an occasion not to miss and document.   

Route 970, which ran from Carrum to Rosebud, with some trips finishing at Mornington, will be replaced by extended hours on regular routes as part of Night Network bus reforms. These changes bring Melbourne more into line with other cities that have regular bus routes running 24 hours on weekends, providing a simpler service overall. 

The route can trace its origins to the Frankston NightRider. This was one of the original NightRider buses of the late 1990s. It roughly followed the Nepean Hwy from the CBD with a deviation via Frankston North, Fraunkston Station then east to Karingal. Later the latter was dropped with it extending beyond Frankston to Mornington and then Rosebud. Instead of just being known as the 'Frankston NightRider', it, and other routes, got formalised numbers in the 900-series (which are rarely used for daytime buses). 

A rebranding around 2008 saw NightRider services increased from every 60 to every 30 minutes. And the higher cash fares were discontinued in favour of them being brought into the integrated Metcard (later myki) system. At this time trains and trams didn't run so buses typically all departed from the CBD (or connected with those that did). 2016 saw the launch of Night Network, incorporating hourly services on all electrified Metro train lines and half-hourly service on six tram routes. The bus network was also reformed with more routes connecting with trains rather than starting in the CBD. However usage was often low, possibly due to public confusion with different routes operating late at night versus during the day. The network that starts this weekend is an effort to overcome this with some of the 900-series Night Bus routes (including the 970) being replaced by added services on regular routes (in this case the 833 between Carrum and Frankston and a shorter version of the 788 from Frankston). 

Anyway enjoy the video. Please let me know if you'd like other 'Timetable Tuesdays' to be done in this format as it could become an occasional feature. 
 


Read more Timetable Tuesday items here

Friday, September 17, 2021

Flashback Friday: 70 years since 1951's bus crisis

70 years ago newspapers reported on a crisis that was the start of three or four decades of bus patronage falls and service cuts. 

This was the era when rights to run bus routes were allocated by government but there was no state subsidy. Bus routes had to earn their keep from their fare box revenue. 

1951 was very different from the more protected Melbourne bus scene of 2021, where, provided you've obtained or inherited a long term contract you can't really go wrong even if no one rides your buses. Today duplicative and poorly used bus routes that do not even stack up on social equity grounds routinely attract public subsidy for years without serious review or reform.  

Bus companies (which were numerous and small in those days) charged and collected their own fares. They could apply for fare increases but they were not always granted by the government that regulated fares. And if they were there was the risk that they could lose patronage to a cheaper operator running a nearby route. And, especially in older suburbs, routes were often both short and close together. 

The problems of 1951 were caused by two matters outside the bus industry's control.

Firstly there was the end of wartime petrol rationing in 1950. When that ended people could drive their cars further without restriction. Which they did. This contributed to the fall in public transport use off from its peak in about 1945. Public transport usage narrowed to become more housewives, schoolchildren and peak CBD commuters, although buses had some reprieve due to growth of suburbs beyond the tram tracks and away from stations. The non-conversion or closure of some trams also put more people onto buses, though the replacement routes were government rather than privately run. 

Secondly there was the Korean war, wool boom and the resultant high inflation (history of which is here). This increased costs for bus operators since wages were regulated and unionisation was high. With static or falling patronage they had to hike fares to break even, which risked further losses. Or, to stem losses, abandon their routes. Survivors generally did so by buying out weaker operators and running longer combined routes. This lessened the need to change (which was good) but frequencies and operating hours were often reduced (which was bad for passengers but reduced operating costs). 

That's the 'big picture' background. Here are some Trove articles, ordered by date, that discuss route by route details.

* 13 July 1951 https://trove.nla.gov.au/newspaper/article/244238441 & https://trove.nla.gov.au/newspaper/article/205340109 & https://trove.nla.gov.au/newspaper/article/244237896

Loss of East Malvern bus service and calls for extension.

* 27 July 1951 https://trove.nla.gov.au/newspaper/article/205340381
Self-help scheme to revive bus. (Note there have been other community-led schemes, most recently 
Wynbus in the City of Wyndham).

* 12 June 1954 https://trove.nla.gov.au/newspaper/article/205386496
Chamber of Commerce to run Hughesdale bus.

* 10 September 1951 https://trove.nla.gov.au/newspaper/article/205336844
Bus companies threaten to abandon routes (some listed) if government does not grant fare rises

* 17 September 1951 https://trove.nla.gov.au/newspaper/article/244240774
Elsternwick - Point Ormond bus route to close.

* 23 October 1951 https://trove.nla.gov.au/newspaper/article/205328065
Account of a route being abandoned despite operator being granted a fare rise (the route concerned is a predecessor to today's 623).

* 23 October 1951 https://trove.nla.gov.au/newspaper/article/205328097
Up to 13 routes to cease. 

* 24 October 1951 https://trove.nla.gov.au/newspaper/article/247847349
Higher fares blamed for lower bus use and a shift to cycling. 
Major bus routes in Northcote, Heidelberg & Watsonia would cease on 31 December

MP advocates for Heidelberg buses.

Entire bus network to be reorganised. Need to shift buses from stagnant established areas to growing outer areas.

Reprieve for some buses including Kew - Mont Albert and Heidelberg area. 

* 6 February 1952 https://trove.nla.gov.au/newspaper/article/246044577
Mont Albert - Kew converted to school bus only (today the area has few regular routes but many school routes especially around Barkers Rd)

* 1 March 1952 https://trove.nla.gov.au/newspaper/article/246052811
Heidelberg - Ivanhoe route to stop (except for school children)

Western suburbs Tramways Board buses

Criticism of Deer Park bus service due to its hourly frequency and lack of Sunday morning service. Train considered inadequate due to distance of station from settled area. 

Tramways board start Sunday morning Deer Park service. Mention that weekday service will be every 30 minutes after a driver shortage is resolved. 

Frequent Tramways board buses to west. Peak as frequent as 3.5 minutes. 

Tramways Board start running Deer Park - City buses. Article says that buses will provide competition with rail service which was found wanting. A very high frequency was provided, including service every 5 min peaks, 7.5 min off-peak and 10 min night. This became the 216 bus until Brimbank area reforms. 

Summary of buses in 1955

Things seemed to have settled down for a while after these cuts. Although many more were to happen with a general redistribution of service from established to newer suburbs. However newer suburbs never got the operating hours and frequencies that older areas lost.  

This 1955 Government Gazette summaries buses as they were in 1955. 
https://gazette.slv.vic.gov.au/images/1955/V/general/1.pdf

Earlier and later crises and successes

Zooming out a bit, 1951 wasn't the only year buses had problems maintaining existing service levels. The early 1930s saw reduced commuter usage as people lost their jobs during the Great Depression.  Others might have switched to walking or cycling to save having to pay a fare (our cities being more compact then with more small-scale industry nearer peoples homes).

A decade later we were in war. It's well-known that fuel rationing limited private travel. However public transport use was high and people were only encouraged to use it for necessary trips. Bus companies were required to trim their routes to lessen overlap, cut competition with the railways and reduce wasteful duplication.  

We've already covered the early 1950s. After these problems some bus companies did well as suburbanisation exploded and most new estates were beyond walking distance of trains. However rising car ownership led to buses role being reduced to specific markets including housewives, older people and those commuting to stations. Schoolchildren predominantly walked or cycled but they too became significant bus users later. 

Bus companies responded by reducing routes, merging routes (and with each other) and cutting service frequencies and operating hours, particularly at night and on weekends. This was only a short-term relief, with patronage falling so far that they could no longer make their services pay. This combined with wage rises led to a recurrence of 1951-type problems with governments this time stepping in to subsidise buses from the early 1970s. With subsidy came a degree of stability but also increased state control. This meant that bus companies, while still privately owned, ceased being genuinely 'free enterprise' businesses with  limited ability to change their product or pricing.

Dependence on state funding tied buses' fate to public finances. There were long periods in the 1970s and 1990s where buses simply did not expand to serve suburban growth areas. We remain with these 30 or 40 year service backlogs in areas like Knox, Chirnside Park and the Mornington Peninsula. There were also large state-imposed cuts, eg in 1990 and almost exactly 30 years ago in 1991 following a bus contracting bungle a couple of years prior. Again in 2021 timetables remain with the axe-marks of those early '90s cuts as visible now as the day they were made.  

The last 15 years saw a major revival in government interest in buses, with significant service upgrades under Meeting Our Transport Challenges launched by minister Batchelor. This was followed by accelerated network reform under minister Mulder, a stagnation under minister Allan (as infrastructure became all-important) followed by recent signs of a revival with minister Carroll's Victoria's Bus Plan. More post-1950s history in the Seven ages of Melbourne's Buses

Tuesday, September 14, 2021

Timetable Tuesday #134: 10 bus route upgrades that would make our train network more robust

The announcement came out on Sunday that all V/Line train services that day would be cancelled. This is due to COVID-related driver shortages. Instead substitute buses would operate at roughly hourly intervals. This not only affects regional locations but also heavily urbanised parts of outer Melbourne including Wyndham Vale, Tarneit, Melton, Caroline Springs and Deer Park who rely on V/Line for their nearest trains. 

Hourly buses arriving at undetermined times are better than nothing but still make travel inconvenient. Provided they know about it people may prefer to catch a regular bus (operating at a known time) over waiting for a substitute bus whose arrival time is not known (and possibly being unable to fit on the first that arrives). 

However only rarely do regular route buses operate at the same wide operating hours as trains. This includes all buses in areas like Tarneit, which has lower than average car ownership and a higher than average portion of essential workers and those whose jobs does not enable working from home. Limited operating hours lessens the usefulness of buses to function as a true substitute service.


The V/Line disruptions didn't just happen Sunday. There were some yesterday and they continue today, as I write this. It is not known long how long they will continue for. 

COVID isn't the only thing that can disrupt trains. Planned occupations is another. Our network still has too many points and track faults, signal failures, overhead power losses and level crossing incidents that disrupt services. And when it gets hot trains may have 'go slow' orders that lessen their reliability. Climate change will raise average temperatures and cause more extreme weather events that could disrupt train services more. So we really do need to think about how to both ruggedise our train lines so they can run reliably during more extreme conditions network and provide alternative when they don't.

A casual look at PTV network maps indicate that there are a lot of regular bus options to move between train lines at many stations. The diagram below shows how that might work if a portion of the Frankston line couldn't run. One could take a bus to another line and get the bus across. If you're in between two lines the extra time taken might only be 15 minutes more - often better than waiting for a substitute service. At their current frequencies regular routes might not handle peak loadings well but would provide some handy mitigation especially if there are multiple options that disperse people onto multiple routes, such as between the Sandringham line and parts of the Frankston line (below). 


A key issue though is frequency and operating hours. Buses must match train operating hours if they are to be a reliable substitute. Frequencies on all the orange lines shown can drop to 40 - 60 minutes at times. And 95% of Melbourne's bus routes do not run 'full time', which I'll define as operating hours similar to trains. Most routes cease at 9pm with no service before about 8 or 9 am on weekends. The map below shows Melbourne's entire 'full-time' bus network as it currently stands (though some improvements are coming next week in the Doncaster area). 


Also, where they do run frequencies are low, with 30 to 60 minute frequencies most usual. This can affect connectivity especially for trips that involve multiple changes (which is often the case if the rail line you'd normally take is knocked out). 

Bias against reliability & redundancy (until it gets really bad)

If you're deciding how to spend money on transport projects, reliability and resilience projects often miss out. This is because a piece of infrastructure that people use every day (like a new road or rail line) is both politically more visible and stacks up better than a reliability-enhancing project (eg extra sets of points that allow operation around a damaged section) that might only be useful on the 1% of days when the system fails. However when there's lots of failures then they make the front page and political pressure for fixes intensifies, with lax governments sometimes bundled out of office. 

Generally though the current fashion is to not design in redundancy and call in the buses when running the railway gets too hard. This is as opposed to the 'show must go on' rail heroism a century ago when rail played a more central part in community life and road substitutes were less developed. 

In a similar vein, you might not add full-time bus services if their only role was to provide back-up when trains failed and they would be poorly used at other times. 

Fortunately there is no need to. It so happens that the same bus routes that allow access to the nearest operating train line are also those which are useful and popular for feeder services and to serve destinations only accessible by bus. And usage may be above-average for buses. So these routes, as well as joining train lines during disruptions also more than justify themselves at other times. Hence they should be top priority for improvements with gains for network redundancy as well. 

Network resilience enhancing bus routes

Where are these routes? Here's my list. I'll give top priority to direct routes in areas with high existing usage in areas with high proportions of essential workers who might be using them during these times. It would also be desirable if the routes are locations where trains are regularly terminated as that would make changing to them easier. Because these improvements are all operating hours and off-peak frequencies they merely work the existing fleet harder and would require no new bus purchases. However some could justify subsequent upgrades that would require extra buses, such as 10 minute frequencies. I'll go through them roughly west to east. 

180: A popular and direct route between Tarneit and Werribee. It will gain 24 hour weekend service in the upcoming Night Network change. However it still has late am starts (particularly weekends) and before 10pm finishes on most evenings. Longer operating hours would provide Tarneit Station with a connection to the Werribee line if V/Line trains are knocked out while also supporting the Werribee line if Metro trains are suspended.

150: Similar comments apply to 180 above, though the route runs between Tarneit and Williams Landing. It serves a large and densely populated catchment near no full time services and relatively low car ownership. Hence the extra trips will be useful as feeders when trains on both lines are running. 

170: Like the 180 it runs from Tarneit to Werribee but via a different alignment that includes Werribee Plaza Shopping Centre. It needs trips added after 9pm and earlier in the mornings, especially weekends. 

All three of these routes have above average patronage. More on improving them here

420: Sunshine - Watergardens. A popular route that provides connectivity between three major stations (some on different lines) at Sunshine, Deer Park (though with poor interchange facilities) and Watergardens. Even if City - Sunshine trains were not running there are many (though slow) bus options for that portion of travel from either Footscray or the CBD. 

901: Roxburgh Park - Epping. A segment of a SmartBus orbital. The main things holding it back for reliable train connectivity are the operating hours (notably the 9pm finish on Sunday evenings) and the weekend frequency (only every 30 minutes). If you were to upgrade service on this route you would likely run the upgrade over a longer segment, eg Melbourne Airport or Broadmeadows to South Morang, even though only the central Roxburgh Park - Epping portion is strictly required for cross-line connectivity.  

902: Broadmeadows - Greensborough. Another SmartBus with similar service limitations to the 901.  A limitation is the long distance between stations, exacerbated by the lack of a station at Campbellfield which could have enabled a connection to the Upfield line. 

903: Coburg - Heidelberg. The last of the orbitals. I've not included Sunshine or Essendon in this segment as directness isn't great (though one could). It connects some major stations including Preston and destinations including Northland Shopping Centre. I have described an economical upgrade that delivers a 10 minute frequency in this item on a Route 904. Like the other 900-series it suffers from an early Sunday finish and limited weekend frequencies. 

Other contenders in this area (but further south) include routes like the 513 and 510. However these have weaker termini and trip generators. And in 513's case there's significant complexity and indirectness that make it not worth a large service upgrade without these being addressed. 

624: Caulfield - East Kew. A potentially strong north-south route that connects a lot of train and tram lines at right angles. Unfortunately it is a very complex route that performs well below potential due to its half-hourly weekday/hourly weekend service. Operating hours are also limited. 

903: Box Hill - Oakleigh. Another section of the 903 that was discussed just above. Arguably the upgraded section should extend to Doncaster or Heidelberg to complete the circle (if keeping as an orbital). Connects similar lines to the 624 and 733 but (often) at different stations. Also extends south to the Frankston line but geometry is less favourable as it approaches at an acute angle rather than 90 degrees.  

733: Box Hill - Clayton. This is the busiest section of a very popular route. It connects the Belgrave/Lilydale, Glen Waverley and Pakenham/Cranbourne lines. Existing usage is very high relative to its low service levels. It has similar hours and frequency issues to the 624 mentioned above.  The full route runs to Oakleigh but the Clayton - Oakleigh portion largely parallels other services and is less deserving of an upgrade.  

824: Clayton - Moorabbin.  Part of an existing popular route. A benefit of Moorabbin is it's a place that trains are often terminate in the event of disruptions (planned or otherwise). A future network reform might extend it west to the Brighton area providing better connectivity than the currently infrequent 811/812. Operating frequencies and hours are the normal offering from buses, eg a 9-10 pm finish, late weekend starts and limited weekend frequency. 

902: Nunawading - Springvale - Chelsea. Serves busy stations at Nunawading, Glen Waverley and Springvale. It's further out than the 903 so the distances are greater between lines. However it doesn't have the time-using deviation into Chadstone of that route. The geometry is less favourable towards the Frankston line but it is still the most viable substitute when that line is out.  Has SmartBus operating patterns including limited weekend frequency and an early finish on Sunday. 

901: Ringwood - Dandenong - Frankston. Similar comments to the 902 but more so due to the lines fanning out. However it does serve three major centres and rail junctions. 

Service upgrades for above to bring to close to rail standards could include: 

* Improved weekday frequency (to 20 min): 150, 624, 733
* Improved weekend frequency (to 20 min 7 day): 150, 624, 733, 824, 901, 902, 903
* Earlier morning starts: All non-SmartBus routes
* Evening frequency upgrades: All routes listed 
* 9pm - midnight Sunday span: All routes listed

The above and other routes would benefit from the Double Service Frequency on Everything plan. 

More?

There are more routes than those listed above that could also aid network resilience. And there are some cases where alternative connections, not provided by the current network may be merited.

For instance a direct Southland - Sandringham connection along Bay Rd is more direct than the current 822 bus via Cheltenham and better lends itself to becoming a high frequency / all day service. 

One might argue the case for a Caulfield - Camberwell Burke Rd connection, with both being major rail junctions and the route serving both Metro Tunnel and Airport Rail. It's the sort of corridor that probably needs a medium capacity option like light rail or busway on its own way. Then it could make a large contribution if disruptions happen during peak times as well as being a popular connector at other times. 

The SRL SmartBus concept, especially if associated with limited stops and faster speeds could also strengthen the network at all times and raise the profile of the SRL as something that's really coming.

The point with all these is that a resilient network in many areas provided by circumferential buses is not just an extra cost but something that has many wider network benefits useful in non-disrupted times as well.

Maybe I've missed some? Ideas for others are welcome and can be left below. 

Friday, September 10, 2021

Building Melbourne's Useful Network Part 105: Done in the wrong order? FlexiRide coming to Lilydale

Many think of flexible route buses as new up-and-comers that could transform transit. They're not. They go back to at least the 1970s as 'dial a ride' services tried in many places. Most failed, with low ridership and high costs per passenger.

Despite their chequered record, 'demand responsive' buses are a recurring recommendation in various transport plans over the years, old and new. These include 1988's MetPlan and 2021's Victoria's Bus Plan. More on why flexible route buses often fail here

 Telebus in Lilydale

One of the 'dial a rides' that did survive is 'Telebus'. It's operated near Croydon - Lilydale in Melbourne's outer east since the late 1970s. Telebus was established to service sprawl and demographics that could not be more hostile to bus usage as subdivisions turned their back on highway and station access.  

High car ownership, high incomes, low density and, most critically, street layouts did not enable direct and efficient bus routes. That is even if governments were willing to fund them, which they weren't in much of the 1970s and 1990s. For Chirnside Park this was initially due to a golf course being in the middle of the estate. When this later got filled in with houses the neighbourhood remained bus-hostile with narrow, winding and disconnected internal streets. Telebus was thus an 'after the horse has bolted' attempt to minimally serve a bad layout that good planning should have prevented.     


Like a fixed bus route, Telebus has a defined origin and destination and some fixed intermediate stops that are always served. However the intermediate stops are fewer than for a regular bus route. With Telebus you could either wait at a fixed stop or phone the dispatcher before the bus leave the origin and order the bus to call nearer your home. A non-myki add-on fare is charged for that service. This early '90s Invicta bus timetable book explains it as follows: 


Telebuses are a daytime-only shopper, school and commuter-style service. There is sometimes Saturday service but never Sunday or public holiday service. While many local bus routes got 'minimum standards' upgrades including Sunday service about 10 - 15 years ago, Telebus timetables were left untouched. 

The 1992 Met network map below shows four shaded areas. Each has its own Telebus route. Chirnside Park is Area One. There remain four areas up to 2021 though boundaries are slightly different (for now you can see them on the PTV Yarra Ranges map).


The subsequent addition or extension of some fixed routes like 664, 670, 675, 676, 677 and 680 near or through some Telebus areas is notable. These would have sapped some Telebus patronage (see Knox Maroondah Yarra Ranges Bus Service Review, p45). Also Telebus' boardings per hour was higher in 1978  than it was forty years later, partly due to there then being fewer 'competing' routes with overlapping catchments. Nevertheless Telebus shares some characteristics of fixed route buses (eg some fixed stops) which probably made it a better performer than purely flexible route buses. 

Telebus has always been Invicta Bus Company's 'baby'. It has continued under subsequent owners, including Grenda and Ventura. A similar concept was introduced across the Yarra in Gowanbrae as Route 490 from Airport West.

Unusual for buses in Melbourne, Telebuses don't have conventional route numbers, making information on them hard to find on the PTV website. The same can be said for 'how to use' details, which requires effort to find the special page here. PTV's app has even less information with recourse to the full website item necessary to find the number for booking. 



As mentioned before, Telebus started around Lilydale in the 1970s. Rowville gained it maybe a decade or so later with services to Stud Park and Ferntree Gully. At the time both Ventura and Invicta (then different companies) had buses that ran to Rowville but neither ran what you'd call a full service. This remains an issue today with effectively two incomplete networks operating - one fixed route and one flexible route - with 'minimum standards' service on neither.  

Last year Rowville's Telebuses were replaced with FlexiRide, This removed Telebus' fixed times and stop. It also stripped both the PTV website and app of timetable information, with everything being shunted off to a separate app (although phone bookings were still taken). The area's fixed route network was not reformed and remains indirect and infrequent. My write-up on Rowville FlexiRide is here


John Usher of Invicta Bus Company presented Australasian Transport Research Forum papers about planning and operating Telebus in 1978 and 1994. I recommend reading these for detailed background on the Lilydale and Rowville Telebuses. 

FlexiRide in Lilydale

Following on from Rowville, now it's Lilydale's turn. From October 4, 2021 all their Telebus services (1, 2, 3, 4) will be replaced with FlexiRide. Buses will operate in three zones. Each zone will have two anchor destinations, typically a railway station and/or major shopping centre. An annotated version of the existing PTV area map (Telebus areas in orange) with guessed FlexiRide travel zone boundaries is below.  The FlexiRide app will likely have a fuller map, with the difference being the separate zones as opposed to just one as with Rowville. 


FlexiRide operating hours are 6am to 8pm weekdays and 8am to 6pm Saturdays. There will be no Sunday or public holiday service. The absence of the latter puts it out of kilter with the standard (but only partly implemented) pattern for Melbourne buses where routes with Saturday do run on most public holidays. Bookings can either be made via the FlexiRide app or telephone. The telephone number is the same as that used for the Rowville FlexiRide (8710 6377). 

Boarding points can include physical or 'virtual' bus stops (which are unmarked so may be harder to find). The PTV site says that FlexiRide is a 12 month trial, though I can't see the name going back to the original Telebus after then. Overall these changes move the service further away from a fixed route (albeit one with a variable path between some fixed intermediate stops) to a fully variable path. This is generally in the direction of less rather than more passengers per bus operating hour productivity. 

What about the door to door service for which Telebus imposed a fare surcharge? The surcharge has gone. There is still the closer service available but only for those with accessibility requirements. The ending of the surcharge removes the last vestige of 'pay the driver' cash bus fares in Melbourne, although prepaid ticket books were available.

Other network changes

As well as Lilydale Telebus becoming FlexiRider, two other changes are happening to local buses on October 4.   

Regular bus route 676 will disappear. This will be covered by one of the FlexiRides. There are fixed route buses on surrounding main roads but part of the area is hilly. 

Also Route 672 between Croydon and Chirnside Park will be simplified to become a fixed route bus at all times. Currently a section north of Croydon has off-peak Telebus running with it able to deviate in the Croydon Hills area. Now it will remain on Yarra Rd, simplifying the service and providing more predictable travel times. Some passengers will need to walk further but the general experience with Telebuses is that the fixed stops see the most usage. This makes the 672 a regular bus route rather than being 'neither fish nor fowl'.  

Both these changes are relatively minor. They don't much overhaul the network, despite local buses being amongst the least used in Melbourne. Not even the redundant and duplicative 673 is removed. And main roads that don't have buses don't get them. 

Is this the best we can do?  

The short answer is no. 

Just replacing Telebus with FlexiRide is replacing one low productivity bus service model with one even lower. All while neglecting the fixed route network with greater potential. The tendency to take the path of least resistance of rebranding an existing similar service, adding an app and calling it bus reform is a habit the Department of Transport needs to quit.  

This is because flexible route buses have low passenger boardings per hour. Most commonly this is around 5 to 10 boardings per hour (or less for operating models less successful than Telebus which is probably 'best of type'). Even quiet fixed routes in sparse suburban areas do about double that. And it's double again in areas with straighter streets, higher density and better demographics. 

Rowville does not have great demographics for buses. It also has a pitiful mix of a poor quality unreviewed fixed route network and a limited hours FlexiRide service grafted over it. Similar applies in the Mooroolbark and Chirnside Park areas where routes like 664, 675, and 677 operate near or through the Telebus area. The presence of both types of services may mean that neither thrive for the thin patronage market offering. 

A concept network

The order we plan various parts of the network is critical. Service planning is basically about providing the best service to the most number of people at an affordable cost. 'Best' typically means things like operating hours, frequency, directness and speed. This usually means running frequent service along main roads that are reasonably walkable to most. 

Such a network may have coverage gaps which routes on in between streets can cover. Planners may have to trade-off frequency and coverage depending on local needs and demands. They may have to insert local 'in between' routes, especially if neighbourhood walkability is poor. Both frequent and local bus services are fixed routes, with flexible routes a last resort due to their high costs per passenger carried and sharply reduced reliability if more than a few people start using the service. 

Their inability to efficiently serve passengers means that before you start fiddling with flexible routes a higher priority should be to reform fixed routes, as per the framework below. If flexible routes have a role it is a minor residual one serving certain special needs that justify the high cost per passenger. 

Just like in Rowville, the recipe in Lilydale is 'FlexiRide first'. 

It is being introduced while the local fixed route bus network is hardly changing.

Buses would become more useful and more reliable if the order of reform was reversed. 

Two concepts where regular routes were deviated or extended to throw coverage into areas currently served by Telebus (and would be served by FlexiRide) are mapped below: 

NETWORK 1


NETWORK 2

The pockets where any flexible route may be considered would get smaller if not vanish. This will result in a smaller and preferably zero proportion of buses being used for inherently indirect and inefficient operations. This should make it possible to deliver the most effective and economical bus networks to an area where service has long been substandard.  

Other network concepts worth exploring appear in 2010's bus service review for Knox Maroondah & Yarra Ranges. Only a few recommendations from that were implemented. 

Conclusion

On one level it is good that buses in the Lilydale area are getting attention. On another it's not the attention that would necessarily deliver the most efficient and economical service. 

This is because changes are being made in the wrong order, with flexible route services being adjusted without there being a review of fixed routes first.  Because the latter generally have higher boardings per bus hour (or potential for same) reforming fixed routes should be done first. 

Two concepts for fixed route networks that would lessen or remove the need for flexible routes have been outlined. Comments are appreciated and can be left below. 

See other Building Melbourne's Useful Network items here


Tuesday, September 07, 2021

Timetable Tuesday #133: 907 the biggest winner in oily rag bus revamp


Back in June we were promised some bus upgrades as the first instalment of Victoria's Bus Plan. The plan itself is not that detailed when it comes to specific measures. But a few were announced. These included a revised Night Bus network that added 24 hour weekend service to some regular bus routes, some funded but not yet implemented initiatives like the Parkville - Victoria Park shuttle (mentioned on Friday) and timetable changes to 19 (mostly unspecified) bus routes.  


The resurgence of COVID and a further lockdown (including a curfew) deferred the Night Network reforms. Then Minister Ben Carroll issued a media release last Tuesday about the other changes. Details were still limited but we learned more about the Victoria Park shuttle, a proposed route in Keysborough South (funded in the 2019 Budget) and extended Sunday evening bus service to Doncaster and Templestowe. All-up nine bus routes would get off-peak service upgrades. 

There was the odd hint. For instance Routes 223 and 293 were mentioned in places. But not other details. Hence little could be written until timetables came out on Friday. News of their availability was advised on the PTV website

How significant are the changes? The PTV website item doesn't give much guidance. Their style is like "here are the timetables, go compare them yourself". Even though a comprehensive comparison takes the best part of half a day that few people will spare. When people bemoan the low profile of buses (and improvements to them) in Melbourne, the Department of Transport must take significant responsibility.  


Without much official guidance, the task falls to others to explain the 19 changes. What's in them? Do they make buses more useful and easier to use? And do they signal a real revival in bus service reform? Keep reading and find out! 

It's all about Transdev (this time anyway)

All 19 changes affect one just one operator - Transdev Melbourne. They are very low cost overall. Some if not all are paid for by moving service from quiet routes to busy routes. Or from quiet times to busier times on the same route. So if you were expecting big all-day service gains on 19 routes you are likely to be disappointed. Though PTV did give warning with phrases like "better aligned to passenger demand" (which allows cuts as well as gains). Hence it's a 'robbing Peter to pay Paul exercise' but if Paul's routes serve many more people than Peter's then it's still a worthwhile thing to do.

Why Transdev? Of all the bus operators in Melbourne, Transdev (who inherited old government Tramways and Railways routes) is the one whose services have the greatest scope for pruning and redistribution. Some sensible reforms have happened but there was scope for more, as I explained nearly two years ago

Brief description of changes

The release of Victoria's Bus Plan marked revived official interest in bus service reform. With it being so long since its last big run out of the garage, the engine of bus reform needs a while to reach full speed. So right now they're just doing some low-speed test laps without complex manoeuvres involving route reforms and multiple bus companies. 

These changes, to start on September 20, are timetable adjustments only. They don't change routes. Some routes will see fewer trips while others will get more. Overall, it's a boost for some of Transdev's (and Melbourne's) more popular bus routes with the City of Manningham being the biggest winner.  

Watch this video for a quick taste. 


What do these changes mean for the network? As it turns out quite a lot. Not least because Melbourne has so few full-time 7 day frequent service bus routes. Fewer in fact than smaller cities like Brisbane and Perth. 

Within two weeks the number of individual metropolitan bus routes operating every 15 minutes or better during the day / all week will double from two (246 and 732) to four (234, 246, 732 & 907). 

Four SmartBuses, all in the Doncaster area (905, 906, 907, 908), will gain 9pm to midnight Sunday service, increasing the number of seven day 'full-time' bus routes in Melbourne from 14 to 17 (as one very minor route loses its full time status). 

This is the biggest single increase we have seen for many years. Also significant is that three of these routes (905, 907 & 908) will gain 24 hour weekend service as part of the revamped Night Bus Network (discussed here). 

Doncaster Road's Route 907 features in all three upgrades so sees the biggest gains. 

Some less frequent Transdev routes also feature. Weekend upgrades are most common. These include wider operating hours and/or increased daytime frequency, especially on Sundays. The aforementioned Garden City 234 and Doncaster 907 go from every 20 to 15 minutes. Highpoint's 223 improves from every 30 to every 20 minutes on Sunday morning. This breaks the decades-long association with old operating patterns that had a lesser pre-midday Sunday service (you still see this on tram and train timetables, and even bus 605). Also the popular 270 and 279 north from Box Hill double from every 60 to every 30 minutes on Sundays. 

293 between Greensborough and Box Hill gains with better weekend hours and a halving of Sunday waits from 120 to 60 minutes. Despite its overlap of other routes this is overdue as the 293 both gets good usage and follows a particularly direct alignment that should have become a SmartBus (instead of weaker alignments such as parts of 901 and 902 in the area). 

In June I listed 19 Melbourne bus routes that most deserved an upgrade. Three (270, 271 and 279) are operated by Transdev. Of these two (270 and 279) get better hours and weekend frequency similar to the middle option suggested. Furthermore 279 and 907 rank in the top ten most productive weekend bus routes in Melbourne so their upgrades are solidly justified by usage evidence. 

Now to the oily rag reference. There was little if any new money for the abovementioned upgrades. Instead funding came from cuts in Transdev's less used routes. Usage of these was low relative to the service frequency provided. Service reduction techniques included: 

* AM peak services: Where buses were very frequent and usage was not high, some inbound trips on long routes (eg 305, 309, 905, 906, 907) were started part way along it. This saved significant service kilometres that could be redistributed while retaining a still good 10 minute frequency at the route's origin. This is easier to do on inbound morning trips as you avoid the 'multiple destination problem' that would arise if some trips (particularly afternoon peak) were finished short. 

* Frequency reductions: This was most common in the evenings on quieter but well served routes. Instead of every 20 minutes the frequency might be cut to every 30 or 40 minutes. Or every 30 down to every 40 or 60 minutes. There are exceptions but operating hours were generally preserved or even extended (particularly on Sunday mornings where it's known that some unmet travel demand exists). The later it was the more likely frequency would be cut, with some routes having the second last trip removed. Daytime frequency cuts were generally avoided, with the main exception being the 908 on weekends (cut from every 20 to 30 min) so that this retained coordination with the 907 (upgraded from every 20 to every 15 min). Routes 305, 603 and 604 are the most affected though others like 270, 279 and 426 also got small reductions late at night. 

* Shorter operating hours: Again appears to be a last resort only done if usage was very low. The very short Sunshine - Sunshine South 429 stands out in this regard, with everything much after about 9pm being deleted. 

The full list of routes that will get their timetables changed from September 20 is: 

Route 215 Caroline Springs – Highpoint Shopping Centre
Route 223 Yarraville – Footscray – Highpoint Shopping Centre
Route 234 Queen Victoria Market – City – Garden City (Port Melbourne)
Route 237 Fishermans Bend – City (Queen Victoria Market)
Route 270 Box Hill – Blackburn North – Mitcham
Route 271 Box Hill – Ringwood via Park Orchards
Route 279 Box Hill – Doncaster Shopping Centre via Middleborough Road
Route 293 Box Hill – Greensborough via Doncaster 
Route 305 City – Doncaster Shopping Centre – The Pines Shopping Centre
Route 309 Donvale – The Pines Shopping Centre – City
Route 370 Mitcham – Ringwood via Ringwood North
Route 426 Caroline Springs – Sunshine
Route 429 Sunshine – Sunshine South
Route 603 Alfred Hospital – Brighton Beach
Route 604 Alfred Hospital – Gardenvale
Route 905 City – Templestowe – The Pines Shopping Centre
Route 906 City – The Pines Shopping Centre – Warrandyte
Route 907 City – Doncaster – Mitcham
Route 908 Doncaster Park and Ride - The Pines Shopping Centre

Some get increases and some get cuts. Some get a bit of both. Changes to some are major while others get only minor tweaks. Even more detail on each below. 

Summary by route - large changes

Route 223 Yarraville – Footscray – Highpoint Shopping Centre
+ Sunday am service increased from 30 to 20 min
- Late weeknight and Saturday evening service reduced from every 20 to every 30 min

Route 234 Queen Victoria Market – City – Garden City (Port Melbourne)

+ Sunday day service upgraded from every 20 to every 15 min

Route 270 Box Hill – Blackburn North – Mitcham
+ Sunday day upgraded from every 60 to every 30 min with large longer span
+/- Monday - Saturday night frequency dropped from 30 to 60 min but better span
(upgrade to minimum standards)

Route 279 Box Hill – Doncaster Shopping Centre via Middleborough Road
+ More consistent inbound am peak timetable
+ Large improvement in weekend spans (earlier starts, later finishes)
+ Saturday am frequency harmonised with trains at Box Hill
+ Sunday day service improved from 60 to 30 min
- Late weeknight frequency reduced from 30 to 60 min

Route 293 Box Hill – Greensborough via Doncaster 
+ Saturday am service starts earlier with some short trips added
+ Sunday frequency doubled from 120 to 60 min with better span

Route 305 City – Doncaster Shopping Centre – The Pines Shopping Centre
- AM peak frequency reduced with shorter runs
- Weeknight span reduction on inbound (but weekend unchanged)
- Weeknight and Saturday evening frequency reduced from 30 to 40 min
- Sunday evening frequency reduced from 30 to 60 min

Route 370 Mitcham – Ringwood via Ringwood North
+ Weeknight span extended from 7 to 9pm (hourly service)

Route 429 Sunshine – Sunshine South
- All trips much after 9pm deleted

Route 603 Alfred Hospital – Brighton Beach
+ Earlier Sunday am start
- After 7pm service reduced from every 20 to 40 min
- Saturday am service reduced from every 20 to 40 min
- Sunday evening service reduced from every 30 to 40 min

Route 604 Alfred Hospital – Gardenvale
+ Earlier Sunday am start
- After 7pm service reduced from every 20 to 30 min
- Before 8am Saturday service reduced from every 20 to 30 min
- Sunday night service reduced from every 30 to 60 min

Route 905 City – Templestowe – The Pines Shopping Centre
+ New 9pm - midnight Sunday evening service (every 30 min)
+ 24 hour weekend service (new Night Network)
- Some AM peak trips shortened to start at Templestowe Village

Route 906 City – The Pines Shopping Centre – Warrandyte

+ New 9pm - midnight Sunday evening service (every 30 min)
- Some AM peak trips shortened to start at The Pines

Route 907 City – Doncaster – Mitcham
+ Weekend day service upgraded from every 20 to every 15 min
+ New 9pm - midnight Sunday evening service (every 30 min)
+ 24 hour weekend service (new Night Network)
- Some am peak inbound trips converted to short-starters

Route 908 Doncaster Park and Ride - The Pines Shopping Centre
+ New 9pm - midnight Sunday evening service (every 30 min)
24 hour weekend service (new Night Network)
- Weekend day frequency reduced from every 20 to every 30 min (to meet 907)

Summary by route - small changes

Route 215 Caroline Springs – Highpoint Shopping Centre
- Minor reduction in weekday span

Route 237 Fishermans Bend – City (Queen Victoria Market)
+ Slightly longer evening span from Port Melbourne

Route 271 Box Hill – Ringwood via Park Orchards

- Small reduction in am span
- Some weeknight reductions
- Some trips terminate at Donvale

Route 309 Donvale – The Pines Shopping Centre – City
- AM peak short trips starting at Donvale

Route 426 Caroline Springs – Sunshine
+/- Sunday span shifted to include earlier start and earlier finish
- Some late evening frequencies reduced to 60 min



Loose ends and next steps

These are worthwhile but 'small target' timetable reforms. They deliver substantial benefits (particularly in the Doncaster area which wins longer hours and a train-like frequency on its Route 907) but are unlikely to arouse complaints. The (almost entirely evening) service reductions will likely hardly be noticed, and could be defensible on the grounds of low patronage and higher needs elsewhere. COVID-19 makes it hard to compare patronage performance but the revised timetables should win several times more passengers than they lose. 

It is arguable that the changes don't go far enough. Further opportunities, even just within the Transdev network, remain. However they cut some services (including during the day) by more than was done here and (in some cases) may be more controversial. Examples include: 

223 - The 15 minute Saturday morning frequency starts very early. It may be excessive for the patronage. Reducing this to 20 minutes for the first couple of hours could fund minor trip insertions elsewhere. A larger network reform would consolidate the 223 with CDC-run 406 to provide a simpler Footscray - VU - Highpoint route every 10 minutes. 

271 - This timetable only got minor changes. Its catchment demographics aren't favourable for patronage yet it gets a (good by Melbourne standards) 30 minute Saturday frequency despite not running on Sundays. An opportunity exists to run a low-cost 7-day service by halving Saturday service to hourly and introducing a new hourly Sunday service.  

279 - Because this was a timetable change only, the opportunity was not taken to remove the occasional (and complex) Blackburn deviation. The 270 and 279 Sunday upgrades could have been a good quid pro quo for this as it serves the same catchment. Hence I see this as an opportunity missed because public resistance to new bus networks lowers if you can show 'swings and roundabouts' improvements while cutting out complexities and straightening routes. Looking further ahead, the 279's high usage makes it a potential low-cost SmartBus route. However it needs to be simplified first, with the infrequent Blackburn and Templestowe variations removed. Once these are done all it then needs are some modest extensions to operating hours given the upgrade of Sunday service to half-hourly. 

293 - This got a Sunday upgrade to every 60 minutes (instead of 120 minutes). However further frequency and span upgrades would be desirable. The main thing holding this back is the network - the 293 route is a good alignment but the SmartBuses that overlap much of it are not. Hence further improvement might have to wait until much more ambitious network reform that rethinks the SmartBus orbitals. More here

309 - Covers an overserviced but generally low density corridor. This change shortened some peak trips. A more radical change could have removed interpeak service to make it peak only. However the area really needs a complete network overhaul which is far more complex than the timetable tweaks discussed here. These would address the future of the poorly performing and duplicative 280, 282 and 901 services on Reynolds Rd, while leaving the door open for new connections such as between Heidelberg and The Pines Shopping Centre

370 - Gained weekday service until 9pm, ie a significant improvement on the previous 7pm finish. However weekend services still finish early. A low-cost upgrade would be to improve weekend hours too so that the 370 meets 'minimum service standards' for local routes. 

429 - This route could have been deleted entirely however to do so might have required a small deviation to the non-Transdev Route 428 to retain coverage. However the 429's reduced operating hours (eg a 9pm finish to match the 428) probably makes such an amalgamation easier in the future. Such an incremental approach, where opportunities are taken over several timetable changes to gradually increase, reduce or reform routes towards a desired network goal, is successfully used in Perth as explained here

603 & 604 - These changes reduced evening frequency but kept operating hours pretty much unchanged. These are not well used routes and the case could be made for finishing service earlier if more deserving use could be made for the resources saved (which is possible). 

A few Transdev routes were left out. Arguably routes like 251, 302 and 304 could have had operating span increases to deliver minimum standards service seven days. 220 is arguably 907's western twin so could also justify 15 minute Sunday service. The popular 216 could be considered also. However reforms are best done in conjunction with amendments to the partly overlapping non-Transdev 410 to create a simple frequent Ballarat Rd route operating every 10 minutes or better. 

600/922/923 between St Kilda and Southland also didn't get a look-in. However proper network reform rather than just timetable adjustments is needed for the best results here. Similarly 232 is quite poorly used but has resources that are best used to drive Altona North area bus network reform. The 350, which has little unique coverage but serves a developing area, needs to be rethought. The same can be said for Fishermans Bend routes that send too many buses to Queen Vic Markets. There is also the about to commence 202 Melbourne University Shuttle with some suggesting an extension to Kew Junction. 

The fate of the three SmartBus orbitals is however the big design issue facing the Transdev bus network. These gobble up a huge percentage of its bus service hours resources. Some orbital sections are over-serviced while others should have more frequency. Train connectivity is poor, parts inefficiently overlap other routes while the 30 minute weekend frequency, in particular, is not adequate for a top-tier premium bus route in a major city. Addressing this is a matter for substantial orbital SmartBus route reform as discussed here


Conclusion

These changes are a good start. They get us closer to the 7 day frequent bus service Melbourne needs on its key routes. Due to their very low cost they are the sort of loose-ends 'timetable optimising' reform that should be the daily business of the Department of Transport (but, for too long, hasn't been).

All of the September 20 reforms could have done at any time in the last five or six years. But they weren't.

This was just one bus company's routes. Imagine what you could do with others, especially in the rarely reformed middle-north and south-eastern suburbs. And, even better, more sophisticated area-based reforms with a network-wide multi-operator approach taken.

It is not to the credit of previous ministers and department heads that they let opportunities for even low-cost bus service reforms such as these slip by for so long. 

It is good that we now have a minister who has restarted the potentially powerful engine of bus reform.

There is a lot of catching up to do with both smaller timetable tweaking and larger network reforms to make buses more useful to more people.

All effort should now be made to streamline processes and increase the department's capacity by about five-fold to address multi-decade backlogs and deliver more bus service reform. 

The minister might do well to encourage the department to think bolder and bigger to sweep away years of inertia on buses that has historically retarded progress. 

At the same time it is also essential to develop a 'small improvements' culture (with budget) so that the more potential worthwhile and cost-effective reforms (like these) in other areas could happen sooner. 

An index to Timetable Tuesday posts appears here