UN 131: Greensborough FlexiRide - Will they bust the silos this time?

Victoria's transport portfolio operates as a series of silos within which things get done. 

Hence a new train station or 'accessible' tram stop may be moved further from buses or the cross road it serves. New highway pedestrian signals might not align with the (also new) footbridge they were installed for.  A level crossing removal or new station might present connectivity opportunities but bus routes were not always reformed to reap all benefits. And, biggest of the lot, there remain concerns about whether the Suburban Rail Loop will connect well with intersecting trains, trams and buses.

Not surprisingly the Auditor-General found that we had no overall transport plan, despite a full agenda of 'Big Build' road and rail projects. 

The Department of Transport (and predecessors) even struggles with network integration within the one mode. Sometimes lazy planning plonks new bus routes over existing unreformed routes, with the 903 SmartBus orbital between Altona and Northland being merely our most expensive example. This presents big opportunity costs (eg not having buses every 10 minutes), especially when it's been so hard to get funding for even small service uplifts (as opposed to big infrastructure).   

More recently we've had Rowville FlexiRide introduced. This replaced one form of flexible route bus with another. Again this was layered over an existing unchanged fixed route network. Both network have such limited operating days and hours that it is fair to say that the 40-odd thousand people in Rowville/Lysterfield have two half-networks rather than one full network. Rowville FlexiRide even won a UITP design award for this, presumably because opportunity cost and maximised passenger benefit were outside its terms of reference.

FlexiRide in the Lilydale area was next. There were some minor changes to bus routes but otherwise it was still the basic 'grafted on' approach. Having said that, the redundant 673 bus later did get removed. Like Rowville this is an area with sparse bus services including many routes that do not run seven days.  

Then Melton South got FlexiRide. This is the best FlexiRide of the lot with 7 day service. However the public holiday start was made to align with Sunday and not the Saturday service pattern, making it different from regular bus routes in the area. While relatively minor it indicates the non-connected way in which the DoT specifies services and the lack of a 'network integration' oversight. 

Early 2022 saw FlexiRide start on the Mornington Peninsula. This was better integrated with local network changes (even if PTV information wasn't). Its introduction allowed the shortening and simplification of the extremely long and twisty 787 bus. However operating hours and days remain very limited with last buses departing before 3 or 4pm.  

The Department of Transport is slowly getting better. It is now less likely to introduce a FlexiRide bus without reforms to fixed routes than it did when Rowville's started. It still has a way to go in that frequent and local buses should be reviewed and reformed first. A flexible route (with its high cost per passenger) should only be introduced after other options are proved unsuitable. That is if you want to maximise benefit per dollar spent (which should be an important DoT and public policy aim).

Future FlexiRide services

What's coming in the FlexiRide world? Northern Tarneit's getting theirs in late 2022. Next is likely to be Greensborough/St Helena. This was funded in the 2022 State budget.  More about it here. 

The wording in the budget paper is encouraging. "... network changes to deliver Flexiride services in Greensborough and St Helena" implies that, unlike Rowville, planning will consider the existing bus network. 

What part of the existing bus network will be considered? The July 2022 PTV network map below gives some clues.  Routes 517 and 518, between Greensborough and St Helena, are most notable. Both routes have indirect unidirectional loops near their end at St Helena shops. Loops provide coverage for those travelling to Greensborough but mean that residents along them cannot go to and from St Helena shops.

Route 517 is (mostly) a main road route starting at Northland. It becomes more local in character east of Greensborough. Weekday buses operate at a hideously unmemorable 24 minute frequency that doesn't consistently connect with trains (every 20 min). Weekend service on this, and Route 518 service on all days is every 40 minutes. 

Route 517's boardings per kilometre is about average for buses in Melbourne. It has large scope to do better with a minor timetable boost to every 20 minutes. Route 518's usage is much less, at around half. This is likely because the 517 serves larger destinations (eg Northland) and has better catchment demographics (including Heideblerg West) than the heavily car-owing St Helena neighbourhood that 518 exclusively serves. Notwithstanding 518's quietness, its 14 boardings per hour on weekdays is still better productivity than almost any flexible route bus can provide. 

What choices exist if you were were going to reform regular fixed routes in the area as part of introducing FlexiRide? If they're wedded to FlexiRide, it's unlikely that 518 would survive as a fixed route in its current form. However Route 517 between Northland and at least Greensborough would definitely have to continue with an upgrade to every 20 minutes on weekdays (and preferably weekends) highly desirable. There may also be a case to continue Route 517 to at least St Helena shops and maybe Diamond Creek via a straighter bidirectional alignment. This would make it the area's main trunk route, providing connections to the train at two or three stations as well as Northland Shopping Centre.   

Something like the above would be close to a 'minimum change' network. FlexiRide would have to run 7 days per week given that 517 and 518 both do. There is a risk in that if you just had the 517 on a single linear route there would be a large area (including some main roads) without their own fixed route bus. Retaining a simplified Route 518 could help here. While this would only be a short route scope exists to combine it with another to increase the number of people with a one-seat ride to the area's biggest destination of Greensborough.  

If you are going to get significant gains (noting the poor productivity record of flexible route buses) then a wider consideration of the areas's buses will likely be necessary. 

Need for a wider network review

The outer north-east around Greensborough / Eltham / Diamond Creek is pretty much Melbourne's capital when it comes to having lots of buses carrying fresh air. Just 3 in 10 of its local bus routes perform near the metropolitan average. 2 of those 3 extend significantly outside the area (where they would get many of their boardings). The remaining 7 routes hover at around 60% of Melbourne's average and attract loadings below what Infrastructure Victoria considers a viable bus route. 

Local demographics play a part but so must bus network planning. Some areas are over-serviced with overlapping routes while others suffer poor connectivity with trains or no coverage at all. Also key destinations like La Trobe University (education), Heidelberg (health) and Northland (shopping) are rarely easily reached by bus from the area. If there was any part of Melbourne that could be given a cost-effective bus network makeover it would be this, as discussed here.  

Such prospects present wider opportunities for those willing to look beyond Greensborough - St Helena. Routes potentially in the mix include 293, 343, 381, 385, 513, 514, 580, 582 and even the 901 and 902 SmartBuses. Most do not harmonise with trains and/or attract low patronage. Some reforms would be in Diamond Creek/Eltham rather than Greensborough/St Helena but have benefits spreading wider.  

Possibilities include: 

1. Straightening 517 and extension to Diamond Creek and boosting to every 20 min on weekdays (and preferably weekends too). This will be the area's main trunk route with a connection to a major out of area destination (Northland) as per above. 

2. Further straightening and shortening of 517 by skipping Viewbank. Viewbank could instead get a new Heidelberg - Macleod route (potential number 516) to replace the 517. Scope exists to extend this to La Trobe University via Springthorpe Estate to replace the indirect 548. A simplified 548 enables a simpler connection from La Trobe University to the inner-eastern suburbs. If extended to Caulfield via the 624 alignment this would pass close to three universities (La Trobe, Swinburne and Monash) and serve half of Melbourne's rail lines including Metro Tunnel trains at Caulfield.

3. Consolidating Route 513 and 514 into a single more frequent route 513 between Glenroy, Bell St and Greensborough operating at least every 20 min 7 days per week. This would shorten an overly long route, lessen duplication and simplify services with a single route on much of Bell St. A slightly rerouted 902 would replace the 514 on Mountain View Rd while much of the 513 Lower Plenty portion retains overlapping routes. 

4. Deletion of the poorly used and duplicative Route 385 in the area. Unique stops could retain service by being serviced by Route 580 operating from Eltham to Greensborough via Aqueduct and Plenty River Dr. This route would serve the shopping centre at St Helena with potential for this to be an interchange point with the upgraded and more direct Route 517 as well as FlexiRide. Route 580's current 30 minute frequency is possibly excessive for its weak usage and doesn't harmonise with trains. Nevertheless a stronger terminus at Greensborough and connectivity to St Helena shops may assist usage.

5. Potential extension of the (admittedly poorly used) Route 381 to Greensborough via a replacement of Route 518. Diamond Creek is a weak terminus for a long distance bus route like the 381. Unless there's particular demand (eg school traffic) it should either not run there or extend to a major centre like Greensborough. Thought could also be given to providing local coverage around Diamond Creek if Route 580 is moved from the area (4 above). This could be done in conjunction with a reappraisal of Route 343. 

6. Reduction in frequency of the poorly used 578 and 579 routes (currently hourly each) to every 80 or 120 minutes each to reflect their semi rural nature and to harmonise with trains. Resources freed could be used to extend network coverage to the Bolton St area of Eltham which currently has no service. 

7. Conversion of Route 582 from a unidirectional Eltham circular route to a bidirectional Greensborough - Eltham route to improve directness and add connectivity to a larger centre. Routing could be via the 293 alignment to Greensborough and also compensate for the deleted 513 portion to Eltham. It could even replace 293 in the Montmorency area if desired as per 8 below. More here.

8. Swapping the 901 and the 902 SmartBus in the Greensborough area. Currently the two biggest centres in the north-east (Doncaster Shoppingtown and Greensborough) have no direct SmartBus between them. This is despite Greensborough having two SmartBuses to the Mannigham area. Route 902 runs indirectly via Eltham while the 901 veers off to The Pines. Route 293 provides the most direct access but has short operating hours and a limited 30 - 60 minute frequency. Nevertheless Route 293 has good patronage performance, indicating a travel need exists. 

SmartBus routes 901 and 902 can be swapped with no stops being missed. Operating 902 via the 901 alignment, approaching Greensborough from the south, provides the direct connection from Shoppingtown. Meanwhile 901 can be run via Eltham on the less direct 902 path. The direct 902 makes the highly duplicative 293 redundant. The resources from that could be folded into the 281, thus providing a simple 15 minute frequency service between Templestowe, Doncaster Shoppingtown and Box Hill.   

Most of the above is represented on the 'Expanded Useful Network' map here. What's shown on it as 'New 514' was described above as the 516 as the 514 is now a currently used number.      

Conclusion

Planners looking at network changes to complement the introduction of the Greensborough - St Helena FlexiRide have a choice. They can either change only one or two existing routes. That gives some benefits but they are narrowly confined. 

Alternatively they can think broader and look at the huge opportunities presented by the huge fleet of poorly used buses running right next door at Eltham and Diamond Creek. With care it should be possible to transform that into an enhanced and better connected bus network serving a huge arc from La Trobe University around to Box Hill and many areas in between. 

Which option will they choose? Only time will tell. If you have any thoughts then please leave them below.  

See other Useful Network items here

TT #170: 828 Berwick to the Bay

Running a broad sweep across Melbourne’s south and south-east, Bus 828 is a long route that few would ride end to end. It runs from the swank City of Bayside in the west via bustling Southland, jobs and shopping near Moorabbin Airport, very white Dingley Village (emphasis on the latter), multicultural Keysborough, diverse Dandenong, down at heel Doveton, car-choked Fountain Gate and rustic main street Berwick to terminate at its station just to the south. 

Route 828 intersects radial rail at Hampton, Cheltenham, Dandenong and Berwick. It sees heavy use for people making local trips, sometimes involving connections at one of these stations. It serves Southland Shopping Centre bus interchange but not the new railway station there.

The route is mostly direct except for some indirectness between Hampton and Southland where it runs more like a neighbourhood coverage style route. 

Service levels

Route 828 operates to ‘minimum standards’ for local bus routes in Melbourne. That is service between 6am-9pm weekdays, 8am-9pm Saturdays and 9am-9pm Sundays. Weekday service is mostly every 20 minutes with some 40 minute gaps in the early morning and at night. Saturday service is every 40 minutes day, 60 minutes night. A 60 minute service previously operated during the day but services were upgraded a few years ago. Sunday service remains at hourly intervals. Overall weekend service (in particular) is low for a route of 828’s network importance. 

The Southland - Dandenong portion of Route 828 appear on the Principal Public Transport Network. These mostly main roads are set out in planning documents to inform denser, and presumably more transit oriented, land uses. Some PPTN corridors have frequent SmartBuses operating until midnight while others only get a lower neighbourhood style frequency and span of hours. 

Funding to address this was limited and for years we had no bus plan. Last year we did get a bus plan and some following service upgrades, including some on PPTN routes. The plan itself however doesn't mention the PPTN indicating a continued disconnect between land use planning and transport service provision. Hence it's common in Melbourne to get the development but not much transit, including in areas served by the 828.     

History

Route 828 is the consolidated result of several less frequent largely overlapping routes including 826, 827 and 828. The 826 was the longest of this group, operating from Hampton to Pakenham. The 827 went via Hallam Gardens Estate. 

Over the last 30 years the history of the 826/827/828 group of routes is the direct opposite of what happened to the route 600/922/923 group. Whereas the latter started off simple (just the route 600) and got complex as it was split into three infrequent routes, what is now just the 828 corridor was three routes (826/827/828), became two routes (827/828) and finally the simple and mostly direct 828.

If you go back further the corridor was simpler before it got more complex. Out of the three routes mentioned the 826 can trace its history back the longest. This goes back to the country route 238 that started running from Cheltenham to Dandenong in 1933. That got extended to Southland when that opened in 1968, becoming 826 in the 1971 renumbering. By that time there was also a 827 from Dandenong to Cheltenham Station only. The Highett/Hampton area at that time had many short routes numbered in the high 630s and 640s while southern Doveton had the short 844 from Dandenong plus a group of longer routes on Princes Hwy. 

The 827 had disappeared by 1978 with that year's network map only showing the 826 between Southland and Dandenong. The 641 ran from Southland to Hampton in a manner similar to today's 828 alignment. By that time southern Doveton had gained another route in the 842, extending east along Frawley Rd before heading south to the Hallam Hotel. 

The 1980s saw many Melbourne bus operators merge and combine routes. Passengers could travel more places on the one bus but frequencies in older suburbs were often reduced. This was most notable in the south where many Brighton and Hampton area bus routes ran to the growing Southland centre and then on to Dandenong and sometimes beyond. 

This through-running across the Frankston line was far less common in the eastern suburbs because bus routes north of the Ringwood line were government (MMTB) while those south were private (mostly Ventura). Except for the three orbital SmartBuses these divisions largely remain today as does the split between operators.  

The full extent of the 826, 827 and 828 can be seen on the 1992 map. They took identical routes from Hampton to east of Fountain Gate. That provided an improved service through Doveton, including a Fountain Gate connection, replacing the 842 which was abolished. 827 and 828 terminated at Berwick with 826 continuing on to Pakenham. 828 turned off the highway into William Rd to provide north Berwick residential coverage.

2006 saw further change. The weekday-only 826 to Pakenham was replaced with a dedicated route 926 from Fountain Gate. Pakenham also gained some short local routes providing new coverage. An ATRF research paper by Currie and Bell discusses the Pakenham upgrades, finding benefits in these and other bus upgrades. 

827 and 828 remained as Hampton to Berwick routes with Berwick station sometimes also served. Later that year evening and weekend service was improved with longer hours and Sunday service. Each route operated every two hours on weekends to provide a combined hourly frequency. 

827 and 828 services remained complex with not all trips operating to Berwick Station. Route 828 ran at uneven intervals, spaced at 20 and 40 minutes with an hourly Route 827 filling the gap. Also there remained short 828 trips between Hampton and Southland to provide a 30 minute frequency on Saturday morning. This  heightened service is a remnant of a bygone Saturday morning shopping rush that extended shopping hours and the big centres killed off. Nevertheless it remains a feature of many bus timetables in pre-1980s suburbs even in 2022.  

Route 827 and 828 had an odd pattern with more 828 than 827 trips especially on weekdays. But with the 826 gone the routes were 80% the same. A subsequent simplification in 2010 saw every trip made an 828 with consistent serving of Berwick station. The most recent change a few years ago saw the 30 minute Saturday morning frequency between Hampton and Southland replaced with a 40 minute morning and afternoon service over the entire route. Sunday service however remains hourly. 

Patronage

Route 828 enjoys somewhat above average patronage productivity for buses in Melbourne. In late 2018 this was 28 or 29 boardings per kilometre on all days of the week. This flat pattern is likely due to the low weekend service levels mentioned above.

Opportunities and conclusion

Route 828 is mostly a simple and direct alignment. The biggest opportunities for it to do better is through improved service frequency, notably on weekends. An increase to every 20 minutes would greatly improve connectivity to numerous shopping areas including at Hampton, Highett, Cheltenham, Southland, DFO Moorabbin, Keysborough, Dandenong, Doveton, Fountain Gate and Berwick. Earlier starts and later finishes would also assist some work trips.

Parts of Route 828 have been talked about as a SmartBus corridor. It would certainly fill a network need as a direct connection between some of the south-east's strongest destinations. In this archived 2010 advocacy paper BusVic advocated a Sandringham - Southland - Dandenong - Princes Hwy (Fountain Gate?) SmartBus encompassing most of the 828.

The direct connection from Sandringham to Southland is, if anything, more important today with a new station at Southland and the proposed Suburban Rail Loop starting there. It could well be included in a reformed Route 828 provided that the Hampton - Southland portion was merged into another route. 

What are your thoughts on the 828 and its future? Please share them in the comments below. 

Index to other Timetable Tuesday items

UN 130: Ignore the ideologues!

Recently I saw an interesting tweet from the Association of British Commuters about how application of UK competition law could threaten that country's ability to reform bus services and create integrated local networks such as operate in many European and Australian cities. 

As background, Britain de-regulated buses outside London in 1986. The planning role for local councils was greatly reduced while private operators competed against one another with their own routes, timetables and fares. This led to wasteful duplication or 'cherry picking' along busy routes while other routes were cut or abandoned. There were some local successes but overall it failed, with patronage falling greatly. In contrast usage held up much better in London, which kept its integrated service. 

Hence there's been pressure to re-regulate buses with councils wishing to take back at least planning and network integration functions through 'enhanced partnerships'. As you can read here the Competition and Market Authority (the UK's enforcer of competition law) hates removing single operator tickets as it will have a potentially significant impact on the incentives of operators to compete against each other. This is based on a doctrine that public transport operators compete with each other (and not, say, against the car).

Later on the CMA says it wants bus operator branding to be prominent on vehicles. We tried that in Melbourne 20 years ago with our trains and had an unstable period of continual rebrandings and a weak network identity. Not to mention the monetary waste involved. Common liveries is way for cities to effectively communicate they have an integrated transport network, so recommending anything other than that would be a step backwards. 

There's other things as well, for instance the potential of bus operators to veto progress towards integration.  It is clear that CMA is driven by an ideology that market competition in public transport and while planning and co-operation in areas like fares policy and network design is bad.

And in Australia?

The same armchair economist doctrine that disintegrated Britain's public transport in the '80s and is resisting their moves towards integration today is active in Australia as well. 

It resides in taxpayer-funded organisations such as NSW's IPART and our Infrastructure Victoria. They profess expertise in matters including public transport network planning. And they claim an independence with the public interest at heart. Unlike those rent-seeking vested-interest leeches in the transport unions, private operators or even a hidebound Department of Transport! This distance however puts them remote from the 'coal face' of public transport network planning and operations which is not always a good thing. 

It is possibly no accident that Sydney lacks fully integrated public transport fares while IV, so far wisely ignored by the Victorian State Government, wants to dis-integrate ours with its crazy modal fares scheme. 

In a self-serving move to put people who think like them in control of our network, IV has also recommended removing fares policy from the Department of Transport (who plans services) to a separate IPART-like body, presumably with less political control. Splitting service planning from fares policy reduces the effectiveness of the former since there is an interactive relationship. For instance you can't have a full integrated network without multimode fares. 

IV's conceptual model of public transport seems to be one of duplicative overlapping modes that compete against one another, thus providing the all important (in their minds) consumer choice informed by modal pricing (no trains for poor people, thanks). In this they have learned nothing from the de-regulation that was so disastrous for buses in the UK and which gave birth to their model of rail franchising (now being wound back).  The passenger, who must navigate an often confusing system, has a different view, seeing no reason why they should pay more for trips involving a transfer or for buses being unevenly spaced or not connecting with trains.  

IV's thinking seems to come from how their types see everything as a competitive market. If it isn't then it probably should be, including for areas previously considered public services or utilities. Even if their thinking originates in industries that lack an equivalent of public transport's network effect, coverage and frequency (which all make the service more useful and thus build patronage). Yet these bodies make pronouncements about transport planning (like assuming public transport modes compete rather than coordinate) that are too infrequently challenged. 

The thing that most strikes me is that, despite claims to the contrary, efficiency and maximising service seem to play second fiddle to ideology about competition between modes. If you have a given service kilometres you can run in a year (basically fixed by government funding) then you need a network that, amongst other things, maximises the number of people near frequent (ie most useful) service while also maximising network coverage. 

Jarrett Walker talks about a frequency versus coverage trade-off. As resources are not unlimited we in Australian cities typically settle somewhere between the two, ie some people get frequent service but nearly everyone gets some coverage. Planners juggle this with their limited resources and concerns about what amount of network reform their political overlords will wear (which regrettably up to very recently has been very little!). 

If you want to deliver the best service to the most number of people, especially for connectivity across a region as opposed to just the CBD, then you definitely do not duplicate radial train and tram routes with radial bus routes. Like what you might have in an unplanned network where each player seeks to maximise their share but in doing so makes the overall network offering worse (*). 

Doing that wastes service kilometres that are better spent improving coverage and frequency in suburban areas that need it. That means accepting that people will need to change for some trips; one cannot run a frequent network by giving everyone a single bus from their door to all places. The best you can do is to make the network as frequent and transfer-friendly as possible to win acceptance. 

A major part of this is integrated fares that charge according to something like distance, zone and/or time rather than mode. Something that Infrastructure Victoria, apparently under the spell of competition ideology more than maximising efficiency or service, wishes to dismantle. 

This is despite modal fares providing a disincentive to network reform that would improve efficiency. It's hard enough to sell changes to the public without also introducing a fare hike. Which could happen with modal fares where a network reform involves some previously all-bus trips including a train or tram component. For example if you wanted to split or reform the orbitals. 

It's very strange then that those who have championed economic reform with words like 'efficiency' and 'public benefit' are seeking to stymie same when it comes to reforming bus networks by making the job harder at best and politically impossible at worst. IV has also not been on efficiency's side when they overestimate the potential role of flexible route buses as opposed to just reforming fixed routes (which, even if quiet usually carry more passengers per hour). 

The lessons in all this? Study history. Do what works. And just ignore the ideologues!

(for more on this see my Twitter thread here in response to the UK developments) 


See other Building Melbourne's Useful Network items here

(*) Note that such an unplanned network needn't involve private operators nor privatisation. Conversely you can have private operators participating in an integrated network. Brisbane is an example of the former where the state government runs trains and its large city council runs buses, with negligible connectivity between them and both networks being almost entirely radial. Basically two rival bureaucratic fiefdoms unable to deliver the best results for passengers. In contrast Perth can contract out its buses to private operators but still have planning that delivers them as part of an integrated network feeding trains. 

Overall access to frequent service and the ability to make cross-suburban trips is higher in Perth than Brisbane due to the former's strong network planning culture. This culture also means that Perth's trains also carry more passengers despite having half the number of lines, stations and track length, aided by having double Brisbane's frequency and better bus connectivity. Productivity measures such as annual passenger trips per rail employee is also likely to be lower in Brisbane due to the Perth system's leaner staffing.