Routes with service levels that don't match their catchment or patronage explain part of their low productivity. On weekdays too many buses take indirect routes while on weekends they sit idle in depots rather than being in revenue service to the big shopping centres where the people are. Large middle-suburban trip generators often have busy routes with basic frequencies unchanged for decades.
It is also common for buses in low car ownership or low income areas (eg Brunswick, Reservoir, Springvale, Dandenong, Campbellfield) not to run much on weekends unlike the fresh air carriers through affluent overserviced suburbs like Brighton and Eltham. Even some semi-rural areas are over-served. We'd do better if our buses served people, not paddocks.
The Department of Transport does occasional network reviews but there are too few to make serious inroads. Or they just fizzle out with little to show. At the current pace today's youth will be old by the time most suburbs will have received reformed bus networks. Yet overcrowding is acute on some routes while others underperform; Infrastructure Victoria found that 40% of bus routes failed to attract enough patronage for what it considers viable service (20 passenger boardings per bus service hour).
Contrary to some currently fashionable thinking, 'demand responsive' flexible route buses are no cure. Those we already have are doing well to make 10 boardings per hour. I can't see how they could free resources for improvements elsewhere unless changes involve lower standards, large service cuts and/or shifting to a casualised and less paid driver workforce. In other words, the low road.
What you're reading right now might be a bigger help. It's the high road. The Useful Network series seeks to get productivity from buses by presenting alternative networks that would make them simpler, more direct, more frequent and therefore more useful. In some cases you wouldn't need many more buses than currently run. They just need to be put on better routes. And, where service upgrades are needed, these are mostly based on working the existing fleet harder. The result builds connections and creates jobs.
Nowhere is this more apparent than around Greensborough, Eltham and Diamond Creek. There's a lot of poorly used buses on duplicative routes around here. Many are confusing and/or don't harmonise with trains. Any resources saved could be used to boost connections to major destination like Northland, Heidelberg and La Trobe University (such as discussed some time ago). There's also scope to simplify routes and better serve popular local destinations as already proposed for southern Montmorency and Eltham.
Main electoral districts this network covers include Ivanhoe (Anthony Carbines MP), Bundoora (Colin Brooks MP), Mill Park (Lily D'Ambrosio MP), Bulleen (Matthew Guy MP) and Eltham (Vicki Ward MP).
Main electoral districts this network covers include Ivanhoe (Anthony Carbines MP), Bundoora (Colin Brooks MP), Mill Park (Lily D'Ambrosio MP), Bulleen (Matthew Guy MP) and Eltham (Vicki Ward MP).
Existing Useful Network
I explain the Useful Network concept here. It's those routes that are frequent enough and run over long enough hours to be useful for many trips. I've specified a 20 minute frequency on weekdays and 7 day service until 9pm. In other words the coloured lines on the Melbourne Public Transport Frequent Network map with the 20 minute frequency selected.
The map shows the existing Useful Network around Greensborough (click for better clarity).
Heidelberg
(bottom of map) has three Useful Network routes - the Hurstbridge train (to
Greensborough only), the 903 orbital SmartBus and bus route 513 to Glenroy. 513
also goes significantly east to Eltham, but its confusing split drops its
frequency to a no-longer-useful 40 minutes.
While the weekday service levels of the orbitals are good they don't go to the most useful destinations for people in the Greensborough area. For example none directly serve key shopping, educational and health facilities such as Northland, La Trobe University, Bundoora RMIT and hospitals at Heidelberg (a major jobs area). These have buses but their frequencies are less than useful and they don't connect with trains or other buses.
As an example, the 901 SmartBus connects to TAFE at Greensborough and hospitals at Epping but via a sparse, semi-rural route that attracts few passengers. 901 also offers airport access but is indirect from Greensborough. And going the other way it veers left into nowhere a few kilometres short of the area's biggest attractor - Doncaster Shoppingtown.
For this trip
from Greensborough you need the 902, but that's indirect via Eltham. Going west
the 902 in the western direction doesn't
have very strong trip generators, although there are some jobs in the
industrial area it passes through and a connection to the 86 tram towards RMIT
or La Trobe.
To sum up, the
SmartBuses offer high service but timetables don't harmonise with trains and
their destinations are not the most useful for local people. This compares to
the Smartbuses from Manningham southwards where they are more useful, better
connected, have denser catchments and are thus better used.
Finally there's bus route 582. This provides a useful though one-way service to Eltham Station. Holes in the Eltham train timetable mean the bus provides a more consistent frequency than the train at certain times. Its loop nature can make some trips long and there is no direct access to Greensborough, the area's largest suburban centre. I covered Route 582 reform before and will use the same approach here.
Finally there's bus route 582. This provides a useful though one-way service to Eltham Station. Holes in the Eltham train timetable mean the bus provides a more consistent frequency than the train at certain times. Its loop nature can make some trips long and there is no direct access to Greensborough, the area's largest suburban centre. I covered Route 582 reform before and will use the same approach here.
Existing not-quite-useful
network
Key routes serving popular destinations include:
513 Glenroy - Coburg - Preston - hospitals - Heidelberg - Rosanna every 20 minutes then split to Eltham via Greensborough or Lower Plenty (more here)
517 Northland - Rosanna - Greensborough - St Helena (every 24 minutes)
566 Lalor - Bundoora RMIT - Greensborough - La Trobe University - Northland (every 24 minutes - covered here)
These are backbone routes to the area's biggest trip generators. They often go along main roads with significant residential catchments.
513 is a good patronage performer, with weekday patronage exceeding 30 passenger boardings per service hour (statistics courtesy of Department of Transport). It serves a large number of medium sized attractions on a main road, especially on the Coburg - Heidelberg section. While the figures don't go into this detail it is likely from catchment demographics that the western and middle portions are busiest.
The same is likely true for 517, with Northland and Heidelberg West helping its western part. It and 566 are overall quieter, with only around 23 boardings per service hour. That's productive but only just according to Infrastructure Victoria's benchmark. There is no reason why route productivity can't be boosted, with poor coordination (around 24 minutes, missing trains every 20 min), poor directness, overlaps with other routes and in 517's case poor catchment currently holding them back.
566 goes to some useful places but is hideously complex with it crossing over itself. Its twenty-something minute frequency is also a handicap. I suggested extending it to Epping and splitting it at Greensborough here. This leaves 513 and 517 remaining to deal with. We'll return to those later.
Waste and opportunity
If you wanted to blow your money running empty buses on confusing routes that miss trains, this area looks a textbook example. You just need to look at the above network maps. Many routes overlap or terminate at dead-ends rather than continuing to a train station or shopping centre. Service levels are excessive where there's no people and too low where there are people.
Don't just take my word for it; data from the Department of Transport says the same thing. Most routes in the Greensborough area have less than 20 passenger boardings per service hour - the minimum that Infrastructure Victoria regards as being viable for a bus route. This contrasts with areas like the City of Wyndham where all but a few minor routes are considered viable.
In contrast this established and denser part of southern Greensborough only gets weekend buses every 80 minutes because it is beyond where the confusing 513 splits off beyond Rosanna.
Then there's the 343, a fairly new route that duplicates a train. As has become normal (but not good) practice this has been layered over existing routes. It has not been a success with less than 15 boardings per service hour. The same can be said for the recently added 381; again it's one of the area's quietest bus routes.
A lack of passengers isn't the only problem with the local bus network. Others are summarised on the maps below (click for better view).
Very notable is the number of routes
between the Plenty Rd area and Greensborough/Diamond Creek. We discussed the
901 before. However there is also the 385 (a connection under the old 520
before the South Morang and Mernda rail extensions) and the newer 381. These
have mixed catchment, with some semi-rural areas receiving a suburban-type
service 7 days per week. Both are unproductive and one might query their need,
especially at non-school times.
The up side is that it should
be possible to reform this area's bus network to provide more people with
useful bus service with similar or not much more expenditure. And if done
carefully there should be multiple spill-over benefits with improved
connections to adjacent areas such as La Trobe University, RMIT, Northland,
Heidelberg and Doncaster.
Existing resources
Let’s meet
some of the existing bus routes in the area.
Most recent is
the 343. Recently introduced for political rather than patronage reasons, it
duplicates the Hurstbridge train line. It runs on weekdays. Buses
operate every 20 minutes in the peak and 40 minutes interpeak on weekdays only.
It would use 3 buses peak and 2 interpeak.
Then there’s
the 580. This is efficient in terms of its own timetable, with the 45 minute
return trip requiring two buses to operate every 30 minutes. However it does
not mesh with local trains (which operate every 20 or 40 minutes). And one
could argue that the service frequency is excessive for the low density
catchment (there exist very busy routes that run the same interpeak frequency
in areas like Box Hill, Clayton and Glen Waverley).
517 was also
raised before. It’s a route that almost but not quite makes the 20 minute
frequency needed for a route to be considered useful. Its demographics change
greatly from west to east. The west (around Heidelberg West) is low income/high
social needs while the opposite is true beyond Greensborough. Plus its terminus
is weak. The 24 minute frequency, which doesn’t meet trains and is hard to
remember, is too low for the route’s west but possibly excessive for its
eastern portion. On the other hand it provides a handy connection to Northland,
the area's biggest shopping centre, and its boarding numbers are healthy. Maybe
you'd boost its frequency north of Greensborough to offset reductions to poorly
used services such as the 518 and 580 in the St Helena area.
The 518 from
Greensborough to St Helena efficiently runs as a return trip with a 40 minute
frequency (1 bus). That meshes with every second train. However, like the 517
its northern terminus is weak with a one-way loop. Like other routes in the
area (eg 582) it has early weekend starts but finishes are an hour earlier than
the MOTC minimum standards.
381 is a new
route arising from network reforms in the Plenty Valley. It runs every 20
minutes peak, 40 minutes interpeak and weekend. Diamond Creek is a weak
terminus and patronage is understood to be low.
Similar
comments apply for the 385 which has a similar service pattern. However its
southern terminus is stronger, being Greensborough Station. Its predecessor
(Route 520) dates from before Mernda and even South Morang had trains so the
rationale was stronger then.
South and west
of Greensborough are routes 561 and 566. It's important to be
aware of them as they provide connections to important attractions such as La
Trobe University, Bundoora RMIT and Northland. 566, like 517, has a non-clockface
twenty-something minute frequency that fails to consistently meet trains.
Swapping the
SmartBuses
Route 901 goes
directly from a strong trip generator (Greensborough) to a distant weak trip
generator (The Pines Shopping Centre). Route 902 goes directly from
a weak trip generator (Eltham) to a very strong trip generator (Doncaster
Shoppingtown). Travel between the two biggest trip generators
(Doncaster – Greensborough) involves either an indirect trip (902 via Eltham),
a 901/902 or a connection between the 902 and an unharmonized and sometimes
infrequent train.
A way to
overcome this is simply to swap the 901 and 902 between the Yarra River and
Greensborough/Eltham. This means that the 902 would run directly from
Greensborough to Shoppingtown. Whereas the 901 would run between Greensborough,
Eltham and The Pines.
902 would
become a stronger corridor. A frequency upgrade to 10 minutes would allow it to
harmonise with trains at Greensborough, Springvale, and, if Ringwood gets ten
minute trains, Nunawading. And if routes 901 and 902 were swapped at
Broadmeadows this change would give both Greensborough and Doncaster faster and
more direct access to Melbourne Airport.
The 901
SmartBus is already weak between South Morang and Greensborough as its
catchment is sparsely populated. Similar comments apply for it west of The
Pines Shopping Centre along Foote St/Reynolds Rd, Templestowe due to overlap
with other routes, low population density and poor demographics. This
change would weaken the 901 further. The existing weak justification
for the north-east portion of 901 to be a SmartBus would become even
weaker.
This lends weight to the view that Route
901 could be split to free resources for local reinvestment in more useful
service. The western portion could terminate at South Morang and the eastern
portion at The Pines Shopping Centre (or even Blackburn if 703 is extended
north). A local route could operate from South Morang to Greensborough or
Eltham to retain a connection. And there could be reforms on Reynolds Rd to
offset for 901's removal from there as well. Ideas
on the latter here.
Useful Network concept
A concept of an expanded Useful
Network for the Greensborough area is below.
You'll notice
how many more Useful Network lines there are compared to the existing Useful
Network. This is because routes like 517 and 566 (which are every
twenty-something minutes) would get upgraded to evenly meet trains every 20
minutes.
Also some
routes that have weak termini (eg 517 and 561) are extended to their nearest
major station. At least some resources for these initiatives may come from
poorly performing routes like 343, 381, 385, 578, 579 and 580 that could be
could be deleted or modified provided reasonable alternatives were made available.
Greensborough becomes a major hub with
this network. Although the number of routes is similar to now, the number on
the Useful Network approximately doubles. 561 and 566 together provides 6 buses
per hour to LaTrobe University (or fairly close), while 517 and 566 provide 6
buses per hour to Northland. Other destinations including Bundoora RMIT and
along Bell St get 3 buses per hour. The latter is made possible by simplifying
Route 513 so all trips operate to Greensborough. Finally, the 901/902 swap
gives a faster trip to Doncaster Shoppingtown.
There are a
few gaps and loose ends, not least because street patterns make some areas
difficult to efficiently serve. The thinner lines are potential local
routes that exist to provide coverage and enable the main routes to operate
more directly. I've tried to avoid loop routes so that every location gets at
least two useful destinations. An example is the extended 518, which replaces
parts of the existing indirect routes 566, 580 and 517 across the north as well
as providing new coverage near its western end. Similarly to the south a new
Route 514 between Heidelberg and La Trobe University allows 517 and 548 (discussed here) to be
made more direct. Ditto for 582 which
replaces the 293 to Greensborough (which becomes redundant with 902's
straightening).
So far I've
only discussed weekday off-peak service. Because it serves Northland the 517
probably deserves a 20 minute service on weekends as well, at least as far east
as Greensborough. 561 and 513 are also quite strong weekend performers so may
merit a similar upgrade.
Certain
coverage gaps remain. Most notable include Bolton St, Eltham and large parts of
Watsonia North. I also haven't said anything about Diamond Creek which
currently has the 580 providing a close-in 'town' service that also runs to
Eltham. If these areas do get a service it would be a neighbourhood type route
better connected with trains but less often than the main routes discussed here.
How does this
network match demographics? This Charting Transport SIEFA animation shows
that higher income households are around St Helena while there are more low and
middle income earners in the area south of Greensborough. The former are likely
to own more cars and use buses less than the latter. You'll notice from the map
above that the revised Useful Network throws a lot of service in to areas with
higher social needs and likely a greater tendency to use buses.
This network
is more complex than what I normally present on Fridays. However implementation
could be broken into simple stages each involving a few routes. This has been
done with some recent network reforms such as in the Caroline Springs
area.
For example,
Route 566 could initially be split into two routes (565 and 566,
through-running at Greensborough) with the same times as now. A second
stage could reform the ends of both routes (to directly serve bigger
destinations) and boost frequency to 20 minutes.
561's
extension to Greensborough could be done independently if funds are available
and there are no dependencies on something else to be cut.
The 281/293 and 582 package also
looks fairly self-contained. Although it would be nice for it to tie in with
the 901/902 flip to retain the fast Greensborough - Doncaster connection
(which would gain frequency). That in turn may have implications for part of
the 513 between Greensborough and Eltham.
517's
extension to Diamond Creek has the most interactions with other routes. Eg the
343, 381, 385, 513, 518, 548, 580 and the new 514. Even this could be divided
into two stages - Viewbank area and St Helena.
Conclusion
What do you
think about this network? Would it make buses better and more useful in areas
around Greensborough? Are the new connections to Northland and La Trobe
University a good idea or is it more important that frequency in less densely
populated areas is retained? If you have thoughts please leave them in the
comments below.
PS: More Building Melbourne's Useful Network posts are here. Do these and you've transformed buses and public transport generally in Melbourne for not much money.
1 comment:
It seems I'm the only person interested in the Greensborough area...
The 513 split has been a personal bugbear of mine over a number of years. Having the Greensborough leg as my nearest route the lack of frequency at all times is frustrating if I want to use it. I've often wondered how the split would work, and I like you're logical solution. The 561 terminating in the middle of nowhere at a wildlife reserve (a rather nice one at that) so the extension through to Greensborough makes sense and as you said adds the missing link to Latrobe University.
I also beleive that there is a large untapped commuter catchment that like myself use Watsonia station. The combination of the 513 and 561 would add a much needed boost to the frequency and hopefully would make the Greensborough to Watsonia corridor a turn up and go frequency and transfer to trains at Watsonia. I know I would use the bus a lot more especially on weekends when the 513 is basically non existent.
The 566 double-back in Grimshaw Street is surely one of the most confusing stretches of any bus route in Melbourne so thank-you for taking the logical step and splitting it. Removing the Watsonia North deviation is great as well.
I remember the days when the 517 used to do an alternating split loop service in the St Helena area. The 518 replaced the west branch of the St Helena loop, and nothing more was done. I've always thought that the opportunity was missed at the time to extend the 518 or 517 to Diamond Creek, but your 518 proposal goes further and seems like a great way to serve the less major suburban services, even if it is a bit windy.
Basically I can't fault your plan. I wish it was actually going to happen.
Post a Comment