This is because, unlike bus reform in Perth or our LXRP, DTP leadership has yet to transform the department into a well-oiled delivery machine that, amongst other things, is able to reform bus services at the pace and scale needed. The result is lengthening bus reform backlogs across Melbourne and reduced value from its $1.5 billion annual bus operating budget, despite a public obligation to efficiently spend this for the greatest good.
2021's Victoria's Bus Plan correctly identified the history/needs mismatch, as below.
"There is also often a mismatch between demand and frequency
with low service frequency on some high-patronage routes
or high frequency on low-demand routes"
(Victoria's Bus Plan, Page 8)
with low service frequency on some high-patronage routes
or high frequency on low-demand routes"
(Victoria's Bus Plan, Page 8)
In practice though the 1145 day-old bus plan has been a fly-by-nighter; two great initiatives right at the start but less since, with nothing seen of the promised implementation plan.
So it's back to history to explain why things are as they are in Melbourne's south-east.
Oakleigh and Dandenong areas
The late Dr Paul Mees grew up in Melbourne's east just beyond the Burwood Hwy tram's original terminus (it's been extended several times since). In A Very Public Solution he mentioned that the Oakleigh-based Ventura typically ran its buses every half-hour (every 20 min on its flagship Route 700) whereas the Dandenong-based Grenda's ran them hourly (apart from its flagship Route 800 - again every 20 minutes on weekdays).
Dr Mees didn't mention that 40 or more years ago Grenda's had a skilled ex-London scheduler in Ian Wall who intricately offset times to give even 30 or 20 minute combined frequencies where two or three hourly routes overlapped. That maximised benefit from limited resources. Evidence of this remains on the network today with key examples being the 802/804/862, 811/812 and 814/848 corridors. Also, until the new Route 816 replaced the 815, there was a combined 20 minute corridor between Dandenong and Keysborough comprising the hourly but co-scheduled 812, 813 and 815.
Greater and lesser minds
The weakness is that while a great mind can produce an elegant intricately scheduled maximally frequent network, the lesser minds that follow can destroy it, or at least fail to communicate the combined frequent service.
So it's back to history to explain why things are as they are in Melbourne's south-east.
Oakleigh and Dandenong areas
The late Dr Paul Mees grew up in Melbourne's east just beyond the Burwood Hwy tram's original terminus (it's been extended several times since). In A Very Public Solution he mentioned that the Oakleigh-based Ventura typically ran its buses every half-hour (every 20 min on its flagship Route 700) whereas the Dandenong-based Grenda's ran them hourly (apart from its flagship Route 800 - again every 20 minutes on weekdays).
Dr Mees didn't mention that 40 or more years ago Grenda's had a skilled ex-London scheduler in Ian Wall who intricately offset times to give even 30 or 20 minute combined frequencies where two or three hourly routes overlapped. That maximised benefit from limited resources. Evidence of this remains on the network today with key examples being the 802/804/862, 811/812 and 814/848 corridors. Also, until the new Route 816 replaced the 815, there was a combined 20 minute corridor between Dandenong and Keysborough comprising the hourly but co-scheduled 812, 813 and 815.
Greater and lesser minds
The weakness is that while a great mind can produce an elegant intricately scheduled maximally frequent network, the lesser minds that follow can destroy it, or at least fail to communicate the combined frequent service.
The former can happen if other considerations trump an originally carefully offset timetable, such as happened with 426/456 on Ballarat Rd and 802/804/862 between Chadstone and Mulgrave in 2014. As for the latter, PTV's one-size fits all credo can present timetables for overlapping routes at stops separately, making it difficult for passengers to visualise the more frequent combined service. Poor and non-sequential route numbering also makes it harder to discern frequent corridors, such as most conspicuous with 600/922/923 and to a lesser extent 281/293 and 802/804/862.
Multi-route bus corridors are most prevalent in bus-heavy cities with CBD-centric networks. Melbourne uses them less as most of its bus routes are not radial. But local examples exist, including 200/207, 250/251, 302/304 and 411/412. Modern planning has shied away from creating many more, with the tendency to simplify them, eg reduce 3 or 4 route corridors to 2, or even, in 826/827/828's case, just one.
Constraints
Creativity in bus network design can be constrained due to:
(a) current bus contract funding being based on live service hours, with maximising bus vehicle utilisation or frequencies a lower priority;
(b) a lack of a patronage maximising culture in DTP top management with good networks and service specifications a consequence of individual planners' pride of work rather than an institutional standard;
(c) the conditional nature of GAIC bus funding (which can force duplicative, inefficient and infrequent networks in some growth areas), and
(d) the presence of multiple bus operators in an area that can make network reform harder.
In relation to (d) DTP assure us that its bus contract reform will facilitate area rather than operator-based planning. While the former 'network view' approach is better, the difference is minor if one operator exclusively covers a substantial contiguous area, such as Grendas around Dandenong (or even wider under the Ventura group) and CDC around Werribee/Tarneit.
The difference widens where there are two, three or even four bus operators in an area, as is common in our inner north and inner west. This is because the most efficient network design may require some routes to be extended, shortened, rerouted, merged or deleted. In some cases operators may need to share a new route, as with the 900 between Caulfield and Rowville introduced in 2006.
Multi-route bus corridors are most prevalent in bus-heavy cities with CBD-centric networks. Melbourne uses them less as most of its bus routes are not radial. But local examples exist, including 200/207, 250/251, 302/304 and 411/412. Modern planning has shied away from creating many more, with the tendency to simplify them, eg reduce 3 or 4 route corridors to 2, or even, in 826/827/828's case, just one.
Constraints
Creativity in bus network design can be constrained due to:
(a) current bus contract funding being based on live service hours, with maximising bus vehicle utilisation or frequencies a lower priority;
(b) a lack of a patronage maximising culture in DTP top management with good networks and service specifications a consequence of individual planners' pride of work rather than an institutional standard;
(c) the conditional nature of GAIC bus funding (which can force duplicative, inefficient and infrequent networks in some growth areas), and
(d) the presence of multiple bus operators in an area that can make network reform harder.
In relation to (d) DTP assure us that its bus contract reform will facilitate area rather than operator-based planning. While the former 'network view' approach is better, the difference is minor if one operator exclusively covers a substantial contiguous area, such as Grendas around Dandenong (or even wider under the Ventura group) and CDC around Werribee/Tarneit.
The difference widens where there are two, three or even four bus operators in an area, as is common in our inner north and inner west. This is because the most efficient network design may require some routes to be extended, shortened, rerouted, merged or deleted. In some cases operators may need to share a new route, as with the 900 between Caulfield and Rowville introduced in 2006.
Small reform on a per-operator basis can work if this does not cause gaps or overlaps that need other operators' services to be modified. The 2021 Transdev network timetable adjustments along these lines were successful. Scope for similar with other operators remains, with the relative extent of opportunities graphed here. The main thing to avoid is large single operator reforms in multi-operator areas that lose out, such as with Transdev's aborted 2015 greenfield network.
Encouragingly, medium sized area-based network reform can still happen in multi-operator areas if the Department is determined enough to broker (sometimes reluctant) acceptance from operators. Brimbank in 2014 is an example of this.
The clear lesson from the past are that neither a new operator nor new contracts will necessarily deliver bus service reform if will is lacking to do it. As an example, the first lot of bus franchising in the 1990s saw Met Bus split into operations run by Melbourne Bus Link and National Bus. Melbourne Bus Link hardly reformed its services during its entire period while National Bus did a lot (maybe too much?). Its successor Transdev succeeded with its 2014 network but weak public consultation, single operator thinking and failure to accommodate political change made its 2015 attempt fail.
While the new 'ZEV' bus contracts may in theory have some benefits including easier network reform, realisation is by no means guaranteed. Given that there have been examples of bus network reform under existing bus contracts it does not seem reasonable to use 'awaiting the new contracts' as a reason to delay even small reforms.
The clear lesson from the past are that neither a new operator nor new contracts will necessarily deliver bus service reform if will is lacking to do it. As an example, the first lot of bus franchising in the 1990s saw Met Bus split into operations run by Melbourne Bus Link and National Bus. Melbourne Bus Link hardly reformed its services during its entire period while National Bus did a lot (maybe too much?). Its successor Transdev succeeded with its 2014 network but weak public consultation, single operator thinking and failure to accommodate political change made its 2015 attempt fail.
While the new 'ZEV' bus contracts may in theory have some benefits including easier network reform, realisation is by no means guaranteed. Given that there have been examples of bus network reform under existing bus contracts it does not seem reasonable to use 'awaiting the new contracts' as a reason to delay even small reforms.
Brighton/Moorabbin area
Established areas a little nearer the bay (but still beyond the trams) had higher frequency routes. For example every 20 minutes weekdays off-peak (as you can still see today on routes like 630, 824 and 825). Higher frequencies often operated on Saturday mornings with reduced (or no) Saturday afternoon service.
If there was a Sunday service at all it was sparse and if privately-operated likely fell victim to the severe bus service cuts of 1991. The latter was not only because the state was broke - there was also mismanagement including an acrimonious bus contracts dispute that led to duplicative routes being added while established routes were cut.
MetBus routes in the St Kilda/Prahran/Sandringham area largely escaped the cuts, with these retaining their weekend and evening services. 35 years on the key determinant of whether a bus route runs on Sunday evening remains if it is an ex-government service or not rather than a more objective criteria such as patronage or social needs. Ex-government routes exist in this area due to (a) Met Bus taking over Melbourne-Brighton Bus Lines in 1985 and (b) the area having Victorian Railways trams that became government-run buses such as the old 901/current 600 before being franchised later in the 1990s.
Further east though was a patchwork of small private operators. The modern 3-digit bus route numbering system commenced in 1971, with network maps published soon after.
Very roughly Grendas around Dandenong got 800-series numbers while Ventura routes around Box Hill - Oakleigh got 700-series numbers. Each operator's 'main' highway route got a round number like 800 or 700.
The 600s were mostly short routes in the Brighton - Sandringham area (where there were a lot more than now). Most stayed west of Warrigal Rd though a few extended a few kilometres east to as far as Westall. Route 800-series were densest around Dandenong though some had continued west as far as Southland. But everything west (and many east) of the Frankston line were 600-series routes that were relatively frequent, with some examples from Daniel Bowen here.
Very roughly Grendas around Dandenong got 800-series numbers while Ventura routes around Box Hill - Oakleigh got 700-series numbers. Each operator's 'main' highway route got a round number like 800 or 700.
The 600s were mostly short routes in the Brighton - Sandringham area (where there were a lot more than now). Most stayed west of Warrigal Rd though a few extended a few kilometres east to as far as Westall. Route 800-series were densest around Dandenong though some had continued west as far as Southland. But everything west (and many east) of the Frankston line were 600-series routes that were relatively frequent, with some examples from Daniel Bowen here.
Encouraged by the government's 'go or grow' policy, smaller bus companies amalgamated or got bought out by larger bus operators. This led to depots, fleets and eventually routes being consolidated. This provided more one-seat rides. However where a lower frequency route was merged with a higher frequency route the latter would often lose trips, effectively redistributing service from established to growth areas. But not always, as I'll show later.
The most notable merger was the consolidation of various small bus operators under the Moorabbin Transit brand following the Grendas takeover in 1988. Most 600-series route numbers were retired, with more 800-series routes from the Dandenong area extending westward towards Moorabbin, Brighton and Hampton providing coverage instead.
The 1992 network (below) illustrates this, with comparatively little subsequent change to the network structure in the 30 plus years since. Example long east-west routes include 824 (Moorabbin - Keysborough), 811/812 (Brighton - Dandenong) and 826/827/828 (Hampton to at least Dandenong - but later simplified to just 828).
The most notable merger was the consolidation of various small bus operators under the Moorabbin Transit brand following the Grendas takeover in 1988. Most 600-series route numbers were retired, with more 800-series routes from the Dandenong area extending westward towards Moorabbin, Brighton and Hampton providing coverage instead.
The 1992 network (below) illustrates this, with comparatively little subsequent change to the network structure in the 30 plus years since. Example long east-west routes include 824 (Moorabbin - Keysborough), 811/812 (Brighton - Dandenong) and 826/827/828 (Hampton to at least Dandenong - but later simplified to just 828).
Service level differences
In broad terms (and there are exceptions) the western part of most of these long east-west routes is relatively posh, the eastern end relatively poor with some low patronage 'green wedge' sections in the middle.
However as foreshadowed earlier, service levels along some of these routes vary.
The first example is 811 and 812. On weekdays each route operates once an hour. Because these entirely overlap over their western half, this gives a combined 30 minute frequency. Where they split at Dingley Village, service is halved to hourly each. Again due to history where the operator merged higher frequency western area routes with lower frequency routes further east.
Route 812 has little unique coverage, overlapping routes to similar destinations for a large section between Springvale Rd and Dandenong. In contrast Route 811 has high usage and significant unique coverage between Springvale and Dandenong. On usage and demographic grounds this Heatherton Rd portion justifies at least equal service to that which the Brighton end gets but doesn't get it. A complication exists on weekends where the combined frequency on all portions is hourly as the hourly full length 811 is fed by a shortened 812 (eastern section only).
In broad terms (and there are exceptions) the western part of most of these long east-west routes is relatively posh, the eastern end relatively poor with some low patronage 'green wedge' sections in the middle.
However as foreshadowed earlier, service levels along some of these routes vary.
The first example is 811 and 812. On weekdays each route operates once an hour. Because these entirely overlap over their western half, this gives a combined 30 minute frequency. Where they split at Dingley Village, service is halved to hourly each. Again due to history where the operator merged higher frequency western area routes with lower frequency routes further east.
Route 812 has little unique coverage, overlapping routes to similar destinations for a large section between Springvale Rd and Dandenong. In contrast Route 811 has high usage and significant unique coverage between Springvale and Dandenong. On usage and demographic grounds this Heatherton Rd portion justifies at least equal service to that which the Brighton end gets but doesn't get it. A complication exists on weekends where the combined frequency on all portions is hourly as the hourly full length 811 is fed by a shortened 812 (eastern section only).
Route 824 from Moorabbin to Keysborough is another example of a route whose frequency has been kept relatively high in the west but halves in the east. This is done not by having an overlapping pair of routes that fans out but a single route that finishes about half its trips part-way along, giving the west more trips.
Again, like with 811/812, weekday and weekend patterns vary. On weekdays the higher (20 minute) frequency operates between Moorabbin and Westall. Which, although a railway station, is a weak terminus.
Again, like with 811/812, weekday and weekend patterns vary. On weekdays the higher (20 minute) frequency operates between Moorabbin and Westall. Which, although a railway station, is a weak terminus.
On Saturdays the higher frequency portion applies over a shorter section of Route 824 - between Moorabbin and Clayton only. However Clayton is at least a stronger terminus. The western portion of the timetable reflects the old Saturday morning rush hour where a 30 minute service is provided. Similar for the Route 825 timetable that the 824 routes through to at Moorabbin. However on Saturday afternoons, Sundays and evenings the short-workings vanish, with the full route getting an hourly service.
Some time ago Route 828 had a similar Saturday pattern (ie western portion every 30 minutes on Saturday morning, eastern section hourly) but this got adjusted so the entire route operates to a 40 minute frequency, thus improving service in Dandenong/Doveton, which has high social needs and is more remote from other routes compared to the route's western section.
Another instance of history trumping need is where routes that hug the frequent Sandringham train line (such as the Brighton - Sandringham portion of the Route 600) have longer operating hours and later finishes than buses in areas that provide unique coverage. Again this is a case where an ill-advised amalgamation decades ago joined disparate routes with different service level requirements but no one since has succeeded in fixing it despite the zero to low cost of doing so.
Summary & solutions
While routes like 811/812 and 824 gained minimum service standards (including later evening trips and Sunday service) these 2006-era upgrades generally did not address other network issues such as service frequency (unless worse than hourly) or a poor allocation of resources relative to patronage or social needs.
Given the absence of significant implemented bus network reviews since, the result is that timetables can be basically static for years if not decades, with some dubious choices (like the quiet Route 704 frequency boost) sometimes subsequently made where upgrades are funded.
While some 1980s/1990s route amalgamations were a pragmatic response to problems faced at the time, they are not necessarily the best today, especially if one wishes to economically simplify the network, deliver a legible service hierarchy (something the Bus Plan proposes, though vaguely) or selectively upgrade services to best effect.
As one example tacking a relatively frequent and direct route (eg the 824 between Moorabbin and Clayton) on to a meandering local route with little unique coverage between Clayton and Keysborough is not a good fit if you want to build a frequent and direct network without short-working complications.
You might instead opt to have the 824 as an entirely main road service between Moorabbin, Clayton and Waverley Gardens for a more legible east-west route (replacing the eastern portion of the 704 and taking advantage of a level crossing removal). Meanwhile the western portion of the 704 could be joined to the eastern portion of the 824 to form a coverage style Oakleigh - Clayton - Keysborough route with a matching service frequency.
Summary & solutions
While routes like 811/812 and 824 gained minimum service standards (including later evening trips and Sunday service) these 2006-era upgrades generally did not address other network issues such as service frequency (unless worse than hourly) or a poor allocation of resources relative to patronage or social needs.
Given the absence of significant implemented bus network reviews since, the result is that timetables can be basically static for years if not decades, with some dubious choices (like the quiet Route 704 frequency boost) sometimes subsequently made where upgrades are funded.
While some 1980s/1990s route amalgamations were a pragmatic response to problems faced at the time, they are not necessarily the best today, especially if one wishes to economically simplify the network, deliver a legible service hierarchy (something the Bus Plan proposes, though vaguely) or selectively upgrade services to best effect.
As one example tacking a relatively frequent and direct route (eg the 824 between Moorabbin and Clayton) on to a meandering local route with little unique coverage between Clayton and Keysborough is not a good fit if you want to build a frequent and direct network without short-working complications.
You might instead opt to have the 824 as an entirely main road service between Moorabbin, Clayton and Waverley Gardens for a more legible east-west route (replacing the eastern portion of the 704 and taking advantage of a level crossing removal). Meanwhile the western portion of the 704 could be joined to the eastern portion of the 824 to form a coverage style Oakleigh - Clayton - Keysborough route with a matching service frequency.
Similarly 811/812 reform might remove duplication in the Keysborough area and add service to Heatherton Road between Springvale and Dandenong via Dandenong Hospital. Even better if, thanks to other reform in Dingley Village, it could be simplified into a single route operating at double frequency. This would end the anomaly where, for historical reasons, the Brighton end gets double the weekday service of the more important Heatherton Rd segment.
Similar cases exist across Melbourne, with discussions on many available below.
No comments:
Post a Comment