Tuesday, June 28, 2022

TT #169: V/Line rail-rail connections: How good?

How easily can you travel between Victoria's second, third and fourth largest cities exclusively by rail? 

All have train stations in their CBD. And their lines intersect somewhere in metropolitan Melbourne. So the answer boils down to how good the connections between them are. That's what I'll try to answer today. 

From Melbourne all three lines head west to Sunshine and then fan out. Bendigo branches north while Ballarat and Geelong separate just west of Deer Park. Hence Deer Park serves as an interchange point for Geelong - Ballarat travellers. 

Footscray does the same for those travelling to Bendigo. This involves backtracking from Sunshine as, despite hopes about it being a transport 'superhub', Bendigo trains don't currently stop there. 

I'll look at Saturday times only. Sunday timetables are generally similar except for later starts and earlier finishes. Weekday timetables are more frequent, particularly for Geelong and Ballarat (20 and 40 minute interpeak services respectively). Scheduling to ensure connectivity is more important when services are less frequent so I think looking at Saturday times is fair to gauge its quality. 

What is a good connection? You don't want it to be so tight that even the slightest delay to the arriving train will cause it to be broken. Conversely you don't want to be waiting an hour around on a platform in the middle of nowhere especially late at night. 

Due to V/Line's difficulty in meeting punctuality targets, the prudent passenger may distrust scheduled connections of less than say ten minutes, especially if a timely arrival is critical. I'm going to make a stab in the dark and say 10 to 20 minutes is OK as a planning aim. Some might see these somewhat longer times as legitimising poor punctuality performance while others view this as prudent robustness based on 'what is' today. 

Geelong - Ballarat

The earliest you can reach Ballarat from Geelong is just before 9am. Because both lines operate hourly at this time connections are a reasonable 19 minutes at Deer Park for the first two trips. Total travel time is 2hr 11 min. 

From late morning the pattern changes as Geelong trains improve to every 40 minutes while Ballarat trains stay hourly. These unharmonised frequencies mean varying wait times. Typically either 19, 39 or 59 minutes. In other words a 1 in 3 chance of making a reasonable connection. 

Note: Morning and evening ex-Warrnambool trains do not stop at Deer Park 

What does a 59 minute connection mean if the train you’re changing to comes hourly? It means that if you are arriving you will likely see the previous train departing. It’s a bad experience for passengers and should be avoided. It happens six times on Saturdays at Deer Park under current timetables. This puts the onus on passengers to carefully choose trips that connect rather than on the network to provide it as a matter of course. This reduces effective frequency, travel flexibility and thus network usefulness. 

A special problem happens in the mid-evenings where Ballarat trains drop to every 90 minutes before recovering to hourly. Waits are never less than 39 minutes between 18:55 and 21:35. The longest wait is 91 minutes. The 18:52 train from Geelong arrives Deer Park at 19:35. Passengers on it would see the lights of the Ballarat train that departed a minute earlier at 19:34. Thus they must wait at Deer Park until the next train at 21:06. The two 90 min evening gaps in the Melton/Ballarat timetable is an issue in itself, independent of connectivity concerns discussed here. 

Ballarat - Geelong

There’s fewer trips here as I’ve only shown Geelong trains that have a connection from the less frequent Ballarat service. Interchange times at Deer Park are mostly :04, :24 and :44. None are ideal with 4 minutes too tight for good reliability and :24 and :44 adding to journey time. 

The main after 7pm issues are low frequency (with a 128 minute gap between second last and last train) and an early finish with the last train leaving Ballarat before 9:30pm. There isn’t the very long mid-evening wait we saw from Geelong but there is one tight (4 minute) connection.   

Ballarat - Bendigo

Unlike what we saw between Ballarat and Geelong, Ballarat and Bendigo trains operate at close to an hourly headway during the day. This means that connections are likely to be either consistently good or consistently bad. 

The earliest a train – train trip will get you to Bendigo is 9:55am if starting at Ballarat. This is the second Bendigo train from Melbourne as there was a first, departing Southern Cross at 7:01am, arriving at 8:58am. The V/Line timetable shows an early coach leaving Ballarat at 5:35am. However this arrives Southern Cross at 7:05am, missing the Bendigo train by 4 minutes. This illustrates the effect of a few minutes in determining how early and late you can and cannot reach a place. 

As for the rest of the day, waiting times at Footscray are typically 36 minutes out of a 3hr 37 minute trip. Why Footscray and not Sunshine, when both lines go through there? The answer is that Bendigo trains don’t stop at Sunshine. 

If they did and timetables were optimised then Ballarat to Bendigo travel time could be cut by 20% to about 3 hours. Removing such backtracking and speeding overall travel times are the sort of ‘quick wins’ that should be the 'bread and butter' of Department of Transport and V/Line planning work. Where constraints exist to not currently permit this then this should drive investment and reform in areas like infrastructure and operating practices.  

What happens at night? There are some quick trips (3hr 9 min) but these are based on tightish (7 min) connections. Maybe they’re not tight by Japanese and Swiss standards but we don’t have their reliability. It can sometimes be hard to make the business case for better reliability and even basic maintenance stack up on assessment criteria used. Flashy new station roofs are sometimes seen as having greater electoral appeal even while tracks under them rot. However I should point out that the 2022 Victorian state budget has a $248m allocation for 'V/Line efficiency and reliability'. 

There is again a >2 hour gap between the second last and last inbound train from Ballarat like we discussed before. Those in Ballarat would be wise to catch the second last train if at all possible and forego any food or entertainment they might have had. Why? It’s because the last train reaches Footscray at 22:46. When did the desired train to Bendigo leave there? 22:45 – 1 minute earlier. The result is an unsatisfactory 77 minute wait at Footscray for the next and last train at three past midnight.

To summarise, unless you are lucky taking a train from Bendigo to Ballarat on a Saturday (or likely Sunday) is a mugs’ game and you wouldn’t want to do it unless a stopover in Footscray was part of the plan.     

Bendigo - Ballarat 

Travel is happier for those going to Ballarat, provided there are no disruptions. Waiting time is 8 minutes for much of the day.  Good if everything is running to time, but if not ...

The earliest you can reach Ballarat by train is 9:43am – a little earlier than going the other way. Again there’s an earlier train but the first train from Bendigo reaches Footscray 13 minutes too late to meet it. The subsequent and one other connection involves 47 – 49 minutes at Footscray until the pattern (mostly) settles down. Nothing operates at a suitable time to meet the train that arrives at 11:43am.

Evenings present 32 to 58 minute connections. The latter is because the last train (Footscray at 23:26) is two minutes after the Ballarat train at 23:24. However night owls have a coach option where they can go in to Southern Cross with a 15 minute change there. Total time is however still slower than earlier train trips.    

Note: Last trip is coach via Southern Cross.

Geelong - Bendigo

Here we're back to unharmonised frequencies and thus varying connection times at Footscray. Mostly they're about 26, 46 and around 60 minutes. The last happens if you choose an unsuitable train at Geelong; if you delay your departure to get the next train in 40 min time then you cut your waiting by that amount and end up on the same train to Bendigo. There are some reasonable 17 minute connections in the evening and a tight 6 minute connection in the afternoon. 

Bendigo - Geelong

Like with Geelong - Bendigo you can get there by 9am on the first trip. Waits can be up to 39 minutes in the morning but settle down to an alternating 4 minute and 24 minute connection time in the afternoon. A problem arises at 17:58 where a train from Bendigo arrives at Footscray at exactly the same time as a Geelong train departs. This results in an impossible connection unless the Geelong train is serendipitously delayed. 

Some evening connections are more consistent, though the 7 minutes for the last trip would be tight for the conservative traveller. Again Bendigo's 90 minute evening headways limit journey time  flexibility. Like Ballarat and unlike Geelong (which has later trips) the last inbound service leaves at about 21:30. 

Coach options

Coach services operate from Geelong to Ballarat and Bendigo. The Geelong - Ballarat coach trip takes about 1 hr 35 min while Ballarat - Bendigo is about 2hr 15 min. Both these are faster than all-train trips. However trains, though less direct and involving a transfer, win out on frequency, particularly on weekends. With a coach travel speed 4hr 30 min (including a 40 min break at Ballarat), trains are always faster for Geelong - Bendigo travel. 

Findings and solutions

Trying to travel between our largest regional cities in Victoria by train requires significant planning on the part of the passenger to avoid long waits or exposure to fragile connections with a high chance of failure. While trains on these lines are typically every 40 or 60 minutes at most times, the period between instances of good connections can be longer due to incompatible frequencies. For Ballarat to Bendigo frequencies may be harmonised but poor connections can repeat over most of the day. 

Amongst the most cost-effective way to speed regional train travel and make the rail network less 'Melbourne-centric'  would be to review timetables to optimise connections for inter-regional trips. This is likely to assume increasing importance given the regional-based nature of the 2026 Commonwealth Games

As Melbourne outer areas grow, shorter trips like Tarneit to Melton are likely to grow in importance. These lack direct buses between them. For them V/Line is their local train service. Thus what's been written here about connectivity between regional cities applies in equal force to them, with travel time comparisons even less favourable compared to driving.

Need for higher frequency

Even ignoring connectivity concerns, higher frequency is needed to accommodate crowding. Adding carriages (where available) may be cheaper but, while it improves comfort, does not reduce waiting time or end-to-end travel speed like boosting frequency would. The most urgent need for improved frequency is at night where services frequencies drop to 90 - 120 minutes on the Ballarat and Bendigo lines. Inserting one or two extra trips to improve this to 40 - 60 minutes would make a substantial difference, especially if the day timetable pattern can be extended consistently into the night. 

Also desirable are weekend daytime frequency increases, notably on the Geelong and Ballarat lines that have a significant suburban transit function. A 20 minute frequency would bring these lines up to weekend frequencies that operate on most electrified Metro lines. Added frequency on a line doesn't guarantee perfect coordination with other lines but it does put a ceiling on maximum waits and provides capacity to avoid scenes like above. 

A VictoriaTakt?

A network planning approach that explicitly considers and promotes to the public connectivity between lines is essential. The idea is to harmonise frequencies and optimise connections at key interchange points, which in this context means Deer Park and Footscray (or Sunshine). 

Because rail planners have done the painstaking planning and scheduling the passenger often does not need to. Thus they can hop on any train and expect good, reliable and repeating connections to important destinations.  

The Swiss and German approach to achieving this is "First the timetable, then the infrastructure". 

This is based on a vision of state (or country) wide mobility based on good connectivity across the network. This requires significant work in scheduling. Where this isn't efficient (eg a train's run time is too long for it to reach an interchange point in time) then consideration of infrastructure that would allow improved timing is given.

We could potentially make a start by reviewing the extreme cases of services missing one another by a few minutes and considering the scope for rescheduling. This is particularly the case early and late in the day where a change of a few minutes can extend the period that travel is possible by an hour or so. More detailed work could look at the practicality of a more regular and consistent harmonised pulse pattern over the day and night. 

The nearest we get to this is the Regional Network Development Plan and the Western Rail Plan. The recent Ballarat Line Upgrade increased frequency from 60 to 40 minutes on weekdays interpeak but not weekends.  Neither were the long weekend evening gaps plugged. A 40 minute off-peak service remains as an aspiration for the Bendigo, Seymour and Gippsland lines on any day of the week. 

Direct transport options

Faster rail to Geelong, Ballarat and Bendigo would mainly improve connectivity to Melbourne. There may be some indirect effects for inter-regional travel provided the speed increases are in the outer portions of the lines and interchange points are preserved (and if possible pushed out to improve geometry and reduce backtracking). The latter is not assured and results can vary. For example if Geelong trains are routed via Laverton to go the 'old way' (as proposed in the Geelong Fast Rail Project) then there's more back-tracking for Geelong - Ballarat travel while that for Bendigo is may be lessened (assuming interchange at Footscray). 

Electrifying Geelong and running via a Melbourne Metro 2 tunnel from Newport would further improve geometry and speed for the CBD but make the interchange an even less direct Southern Cross Station for regional trips to adjacent regions eg Melton and Ballarat.

Options could include continuing to run some Geelong trains via Wyndham Vale, Sunshine and Footscray. That gives extra choice but there's the risk of splitting frequency (and lengthening some waits) unless service is vastly improved over now. Ideas for restoring direct rail service between Geelong, Ballarat and even Bendigo also bob up with local advocates including the Rail Revival Alliance. Hopes were raised by the Coalition parties before the 2010 state election (which they won) but their feasibility study found it would cost $935 million and not be viable

Improved coach services to be more than a few trips a day are eminently possible and could offer good travel times, especially between Geelong and Ballarat. However in terms of public and political profile this seems to be even lower than metropolitan buses. Little has been done despite their cost-effectiveness. There's certainly some opportunities here. Notwithstanding this pursuing rail connectivity remains essential since there are growing populations in places like Tarneit and Melton that could and should benefit from better rail access to regional cities.


We’ve identified three main problems with regional train timetables. Low frequencies at important travel times, unharmonised timetables and planning that doesn’t necessarily consistently optimise regional to regional trips. 

Money invested in service, work done planning timetables and infrastructure priorities driven by network connectivity have big payoffs in saved time for passengers and likely higher patronage as more trips are made consistently easier.  

See other Timetable Tuesday items here

Tuesday, June 21, 2022

TT #168: The freeway express Route 350 - Do we need it any more?

Most bus routes have at least a section of road where they're provide unique coverage. That is if the route was taken away there would be areas left without bus service. However there are a few special routes that completely overlap others. Instead of providing coverage, their rationale may be that they provide convenient one-bus connectivity to important destinations that would otherwise require backtracking and/or changing buses. In other cases they provide capacity relief for high usage along a popular corridor.  

Route 350, between Melbourne CBD and La Trobe University Bundoora via the Eastern Freeway is one of these special routes wtih basically no unique coverage. Its map is below:  

Its relationship with the rest of the network is better seen on the PTV network map below. 

They are a bit hard to seen but the key overlaps are with Route 250 La Trobe University - Fairfield and numerous Doncaster area express routes between the Eastern Freeway and the CBD. The 350 also overlaps the very occasional 609 on a 500 metre section of Chandler Hwy. This area does however include the dense Alphington paper mill development which is testament to Melbourne's ability to place high-rise development on sites where some of its worst bus services run (Port Melbourne and M-City Clayton are other examples). 

Pretty much the only place where the 350 overlaps with nothing is on about a 500 metre section of Grange Rd in Fairfield. All parts of this are near more frequent 7 days routes including 250, 508 and 510. Also the abovementioned occasional 609 terminates in a back street about 100-200 metres away. 

The Route 350 on Grange Rd crosses the Hurstbridge rail line between Fairfield and Alphington stations. However its distance from either hobbles its potential secondary role as an efficient rail feeder. Moving Alphington station a few hundred metres west is one of those local network enhancement things unlikely to happen, despite connectivity benefits of aligning with perpendicular roads and bus routes.  


It's unusual. Route 350's timetable is somewhere between a peak direction only freeway service and a regular all-day service. 

The route runs between about 7am and 6pm on Monday to Friday only. Morning and afternoon services run in both directions. But there's a 4 hour midday gap (about 10am - 2pm) where nothing operates. Frequency when it does run is about 20 minutes. This is fairly constant across the route's 3 hour morning span and its 4 hour afternoon span. The intent of the timetable is basically to provide a daytime commuter service both to La Trobe University and the CBD. Last buses leave both ends a little after 6pm. 

There is no weekend or public holiday service. This includes public holidays that are university teaching days. Other university bus routes typically run a service on these days. 

End-to-end travel time on the 350 varies from a bit over 50 minutes (shoulder peaks) to around 70 minutes (peak of peaks) with scheduling obviously reflecting traffic conditions. At least 7 buses would be needed to run the service. This makes the 350 an expensive route to run despite its lack of weekend or midday service. Is this justified? Keep reading!


At 23 passenger boardings per hour (late 2018 numbers) Route 350 is pretty close to average for a bus route in Melbourne. On non-school days it drops to 20 boardings per hour. 

Most other bus routes have interpeak, weekend and some evening service whereas the 350 does not.  You would expect these times to be quieter than the peaks for a bus route. Viewed in this context, the  largely peak 350 is underperforming, even though its numbers (just) exceed what Infrastructure Victoria regards as a productive bus route. 

Route 350's achievement of this threshold does not imply that it nor the area's wider network is performing at its potential. And where a route does not contribute to unique coverage there is room to be firmer in calculating its opportunity cost (ie whether something better can be done with what the 350 costs to run). This is particularly where (a) a route uses a large number of buses, (b) the buses used are idle for most of the week, and (c) usage is not particularly high. Route 350 stands out on all three counts and is thus ripe for review in a broader network reappraisal. More on that later.  


The 350 has had much chopping and changing over the years, including both to route numbers and timetables.  Much of this is due to having a varying purpose with potentially better alternatives emerging. 

For example is it a CBD commuter route? Maybe it's the fastest way to reach the transport-starved La Trobe University. Or maybe its high bus-using Heidelberg West catchment makes it useful for non-peak trips too. 

Before we discuss 350's history it's worth looking at transport to La Trobe University, especially from the CBD. What we now know as the 86 tram was only extended there in 1985. Before then it was just buses, with a change from the tram terminus at East Preston. Alternatively you could have got Tramways Board bus 256. Network maps show that this stopped a few hundred metres south of the campus in 1971 with an extension into the campus appearing on the 1972 map. We know this route today as the 250. 

Both the 250 and 86 are slow routes with many intermediate stops. Hence there would have been a desire for a faster way to the expanding La Trobe University campus. The 350 (and predecessors) could have been a response to this. 

The Eastern Freeway has long been the preferred corridor for buses to the Doncaster area (where there are no trains). It was less needed for buses to the north and north east as these had the Hurstbridge and (then) Epping railway lines. 

There were also alternative ways from the inner city, such as via Westgarth St or Heidelberg Rd, for those buses that did run. Nevertheless, despite these routes crossing a rail line, the MMTB did experiment with peak-only Eastern Freeway CBD buses from West Heidelberg (257), Northland (258) and La Trobe University (259) commencing in 1979. It is the last that can be considered the 350's ancestor. 

These services continued through the 1980s. The 1991/2 Melway lists the 258 and 259. The 257 short-working had by then disappeared though it was still shown on the 1992 network map

The government-run Met Buses were franchised shortly after, with routes transferred to National Bus or Melbourne Bus Link in the '90s. Melbourne Bus Link hardly reformed its network while National Bus rerouted and renumbered almost everything. These changes included scrapping the 258  from Northland while replacing the 259 with the 350. This followed a pattern, applied to other routes too, where the 300s numbers were used for freeway express services, often running peak periods only. 

At around this time La Trobe University's 256 became the 250 and Northland's 254 became the 251. The 250 and 251 followed a common route on their inner portions with the timetable providing an even spacing. Oriel Rd in Heidelberg West had not only the 250 but also the 350 and some Route 246 trips (extended from Clifton Hill to La Trobe University).  

For a while in the 2000s there was both a Route 340 and 350. The 2006 Melway lists a 340 service operating all stops weekday interpeak. Service was roughly half-hourly. Route 350 ran express trips during peak times. Service between about 10am and 2pm was removed in the 27 July 2014 Transdev network changes with all services operating as the 350. With an emphasis on fewer but simpler and more frequent routes, these same changes increased 250 and 251 from every 30 to every 20 minutes while removing the 246 La Trobe extension.

350's future?

Early 2016 saw the introduction of more new express university shuttle services typically operating from the nearest train station. These follow the great success of the very frequent Routes 401 and 601 to Melbourne and Monash universities respectively. 

One of those introduced was the 301, operating every 10 minutes between Reservoir Station and La Trobe. Once you're on the Mernda train (every 20 min off-peak) you have a turn-up-and-go service to the campus. This has likely eaten into the 350's patronage further and weakens the route's reason for existence. Boosting the Mernda line to every 10 minutes (desirable for other purposes) would make this train/bus combination unambiguously better and further weaken 350's role. 

Macleod on the Hurstbridge line has the 561 bus but both it and the train mostly run every 20 minutes. 350's network role might drop even further if both got all-day frequency upgrades. 

Having said that, some areas served by the 350 do need better service, such as the Alphington paper mill development. And a freeway bus at the door may be more enticing than a walk to the train. It's just that apart from 9-5 commuters, the 350's weekday peak only pattern is unappealing. The route also uses a lot of buses for very average patronage productivity. That should raise the hackles of efficiency-minded planners.

Meanwhile there exists a high need for better connections to La Trobe University from the eastern suburbs and across the Yarra more generally. The bus resources the 350 uses would likely be key to achieving such as revised network economically, especially given current interest in the Eastern Express Busway project as part of North-East Link.

Network reform possibilities could include extending Route 567 north to La Trobe University and south to the Kew/Glenferrie area (also replacing 609), seven day service on Route 546 and moving Route 250 from Victoria St to Grange Rd to replace the 350 and deliver a 7 day service. The resources saved from not running Route 350 may be enough to boost 250 and 251's frequency from 20 to 15 minutes, with reform helped by all three routes all being run by Kinetic.    

What are your thoughts on the 350 and its future? Does it have a continued role? Are changes needed? Or is it more a remnant from before La Trobe got its tram or university rail shuttle feeder with better uses possible for its buses?    

See other Timetable Tuesday items here

Friday, June 17, 2022

What happened to Melbourne's bus audit?

The Victorian Auditor-General appears to have quietly dumped a planned performance audit that would have told Melburnians whether their metropolitan bus services 'integrate with the wider transport network and meet expected service demand'. 

Melbourne on Transit readers already know the answers are too often 'no' and 'no'. For instance we have convoluted bus routes every 22 to 25 minutes in areas where trains are every 20 minutes, making connections haphazard. Even many of our premium SmartBus routes are every 30 minutes on weekends, failing to frequently meet trains mostly every 20 minutes. Maps also show unproductive overlaps in places as new routes get layered over old without sufficient network reform. 

On the demand side buses in areas like Brighton and Eltham carry fresh air while busy routes (often in high social needs areas like Dandenong's Route 814) still finish midday Saturdays and not run Sundays. And main road routes may run only hourly on Sundays despite serving major shopping centres  that are busiest then. A 2020 splurge upgraded the little-used 704 bus while neglecting the nearby far busier Route 800.

State budget papers reveal a $1.4 billion annual spend on bus services (BP3, p333). It is in the public interest that we get maximum value from this. Such value can only be assured by delivering services efficiently (through appropriate contracting arrangements) AND delivering the right services (which depends on planning and monitoring with regards to routes, operating hours and frequencies). 

An audit here would have answered many of these questions. Findings from the Auditor-General carry much more weight than blog posts here. Most notably it would have forced the Department of Transport to respond. And encourage accountability with regards to how we plan public transport services. That's important because it's often seen as a secret art with little oversight nor discernible rationale for bus routes and timetables being what they are. 

Past VAGO bus audits

What's been the Victorian Auditor General's history in this area? In 2015 they looked at the tendering for Melbourne's Bus Franchise. This was the package of routes previously run by Melbourne Bus Link and National Bus in the Sunshine, Brighton, Heidelberg and Doncaster areas along with the three orbital SmartBuses. They were put out to franchise by the Baillieu government. The package included about 30% of Melbourne's bus routes including most of our busiest.

VAGO found that $33m cost savings were achieved in 2013-2014 but that the government hadn't got full value due to unreliable data and, related to this, their inability to withhold payments when performance was poor. The report was also critical of the (by then) Labor government for not extending the franchise model to more routes. 

Maybe they were constrained by limited scope, but one cannot help thinking the auditors' insight would have broadened if they walked a few hundred metres from their office and boarded one of the Transdev buses passing by. 

They would have likely found filthy buses that would have challenged whether the cost savings were real, ie not at the expense of quality. And apparently they missed looking at the conditions that gave rise to the biggest bus fleet safety crisis in recent memory when numerous Transdev buses were put off the road. While Transdev Melbourne lifted its game later, the government apparently remembered and awarded the franchise to Kinetic instead. More on that and better contracting here.

VAGO annual audit plans 

Annual Plans tell us what investigations VAGO is planning in the next few years. You can browse 15 years' worth of annual plans here. They come out before 30 June each year and are tabled in state parliament. 

Plans can be flexible and responsive, meaning that work might be shuffled around a bit. About 3 or 4 performance audits are on the go in the transport portfolio at any one time. 

Bus audit first mooted in 2019-20

The 2019-20 Annual Plan is the first recent one that mentions planning bus services. It had an audit planned for 2020-21 called planning and management of metropolitan bus services.

The audit aimed to determine whether metropolitan bus services were reliable, regular and integrated with other forms of public transport. It noted an Infrastructure Victoria finding that 40 percent of this network was underperforming (based on routes having less than 20 passenger boardings per bus hour). Also that buses were important to serving growing outer suburbs, new bus contracts had started, and that past VAGO audits (like mentioned above) had found poor performance management mechanisms. DoT's role included 'leading the development of an integrated transport plan' (which, in a separate audit tabled in 2021, VAGO found didn't really happen).  

2020-21 plan defers bus service audit

A year later the 2020-21 Annual Plan came out. The audit for bus service was still there but was deferred to 2021-2022. The reason for the audit was similar to above except that the number of routes had risen from 342 to 350. The audit objective had been slightly reworded with the words 'reliable, regular' removed, however 'reliable' still appears as an issue. 

2021-22 deferred again

2021-22 Annual Plan is here. The bus service audit this time appears as part of work proposed for 2022-23. 

The wording is similar to previous plans. 

As late as today, clicking on VAGO's website Annual Work Plan tab still takes us to the 2021-22 plan, with a page for the bus audit here

2022-23 plan - no buses!

The 2022-23 annual plan, tabled in parliament earlier this month has the bus audit dropping off the 2022-23 program. That leaves a thin agenda with just two audits. It's not even deferred to 2023-24. And unlike some previous years we don't have indications of 2024-25 planned audits. Maybe they're waiting until the election in case there are major programs announced that could be worth auditing then. 

While the Auditor General has repeatedly deferred and now apparently abandoned looking at bus services, this inertia has not been matched elsewhere in government. We saw the release of Victoria's Bus Plan in 2021, a little more kindness for buses in 2022's state budget and announcements regarding the electrification of the bus fleet with several depots undergoing conversion. 

It would have been good to have had the audit done before the Bus Plan came out. That could have  given reason for it to be a bit sharper and more specific than it was. An audit report out at a somewhat later time could have contributed to the Bus Reform Implementation Plans currently being developed. A still later release date could have enabled a look at some network reforms being done with lessons for those that follow. 

If VAGO ever does get around to auditing bus service planning, I have just one message. 


Get On The Bus. 

As demonstrated by the salutary 2015 experience, your report will be 10 times better if you do!

Tuesday, June 14, 2022

TT #167: Frankston's loopy 774 bus with the 127 minute gap


Frankston is the land of loop bus routes. There's at least six, depending on how you count them. And none of those have been substantially reformed for decades. Frankston's local MP is Paul Edbrooke MP. Today we'll look at the 774, Frankston's shortest but most productive loop bus route. 

The 774 map from the PTV website is below. It tells you what streets the buses goes down but not in the order that they travel. And that's important for a unidirectional route like the 774. Old-format PTV maps, which were manually drawn and human checked, had direction arrows but no such luxury carried through to PTV's new style of maps. 

PTV local network maps are however still manually mapped. These do have arrows. You can see from the one below that the 774 is a short roughly clockwise route starting and finishing at Frankston Station via Frankston South. Very little of it is unique as almost all of it is overlapped by the 772 which serves less built up areas to the south. Closer in to Frankston, on inbound trips there are overlaps with the 782, 783 and bidirectional portions of the 775 and 776. 

While the network map above shows the direction of the 774 loop it does not show the Nepean Hwy variations in central Frankston on certain midday trips. These need to be understood in conjunction with the timetable. 

Timetable and service levels

Route 774's complete timetable are below (click for clearer view) or can be found on the PTV website here. The first thing to note is that they are split even though the bus goes the full loop. Thus for trips north to Frankston you may need to first go south and then refer to the second timetable to find your trip and arrival time. This is a consequence of both complex route design (loop routes should be avoided where possible) and poor arrangement of data by PTV (including ordering that require timetables to be read in opposite sequence to presented to understand the artificial break at Frankston High School). 

Like other bus routes in Frankston South, Route 774 missed out on the minimum service standards introduced from 2006. Thus it is Monday to Friday daytime only. There is no evening, weekend or public holiday service. Most of 774's passengers do have a partial alternative in the 772 which operates Saturdays and most public holidays but not Sundays. 

Route 774 has a very late start for an outer suburban bus. The first bus arrives Frankston at 7:36am. Allowing some transfer time this means that the earliest one can arrive in the CBD with a train connection is not much before 9am. After that there is an over 2 hour (127 minute) wait until the next bus arrival at 9:43. Oddly it's only 30 minutes until the next bus after that. After that frequency settles down to about hourly until the last departure. This is at 7:17pm from Frankston. The latest one can leave the CBD by train to get this trip is about 6pm. Hence the 774's operating span isn't long enough to be a reliable CBD feeder, though as noted before its catchment has other routes like 772 nearby. 

What about the Nepean Hwy deviation mentioned earlier? The timetable has footnotes for three trips around midday - 2pm. These trips miss one stop near Davey St in favour of two stops to the west, including at Nepean Hwy. Presumably these are more convenient for Frankston CBD shoppers. 


You wouldn't think that that a route with such limited service would be successful. However the 774 is the most productive bus route in Frankston when measured on a boardings per hour basis. At a high 37 boardings per hour on weekdays also narrowly misses the top 40 bus route list discussed here. Some of its usage is dependent on Frankston High School that the route serves. But by no means all of it. On non-school days the 774 attracts a still above average 30 passenger boardings per service hour. 


Basically nothing. The big story is stagnation for 50 years if not more for bus routes serving Frankston South. Maps going back to 1971 are here if you want to confirm how little routes have changed. 


Bus routes in Frankston South enjoy above average patronage productivity on a boardings per hour basis. However as noted here the points where patronage productivity is highest and where patronage is highest are different. 

If you want to unlock all the benefits of public transport you need to maximise patronage. This means adding service until after productivity has peaked but while it is still high. 

With limited operating hours and gaps of up to 2 hours between buses with a complex network there is clearly great scope and need for bus reform in the Frankston area, notably Karingal, Frankston East and Frankston South. A large part of the latter must include reforms to routes like the 772, 773 and 774 to provide simpler and bidirectional routes with less overlap. 

See other Timetable Tuesday items here

Friday, June 10, 2022

UN 129: Highway robbery: How our grandparents walked there quicker

A stealthy theft of time and space is making driving faster and walking slower in suburbs across Melbourne. Individual small changes spread across hundreds of sites is creating a tsunami of slowness driven by state and council authorities who in words claim to support 'active transport'.

Decisions, made by faceless technocrats in road authorities, happen without public accountability or oversight. Yet they're changing the way that we get around, prioritising driving over walking and public transport, even for very local trips. Whether intended or not, the effect is induced car traffic and stronger business cases for more and wider roads. That repeats the vicious cycle as walking becomes even less attractive across increasingly severed communities. 

This theft has been happening for 60 or 70 years. Newer suburbs have known nothing else. Older suburbs got it imposed on them. And it's still going on. Including under the cover of apparently connectivity improving projects such as level crossing removals. In some cases old direct walking paths got obliterated as crossing removals merely substituted one type of barrier for another. In other instances added turning lanes widened roads and thus reduced connectivity for walkers crossing. Certain works, such as less direct paths and replacing zebra crossings with lights, is a walking time grab that just induces more driving despite being done under the guise of accessibility and safety.

Space taken

As mentioned before, this theft can involve space, time or both. 

An example of a space theft is where a road is widened. Instead of having to cross four lanes a walker may have to cross six or eight lanes. This is harder and scarier, especially for the less mobile. Particularly at unsignalised points, which is most places. The recent Getting to the Bus Stop study by Victoria Walks found that 60% of bus stops were on high speed roads. Of those fully 95% lacked a pedestrian crossing within 20 metres.  

Widening projects that incorporate roundabouts or slip lanes are even worse, especially where there are no zebra crossings on the approaches. As well as taking more space the geometry works against bus-bus transferring passengers as stops are pushed back from the intersection. Because reaching a stop requires multi-road crossings, multiple long waits at lights and long walks, this greatly reduces connectivity to and between bus routes. The diagram below paints the picture. Or you can tour a real example involving intersecting SmartBus routes at Springvale Rd/Wellington Rd here

Minimum parking requirements is another hidden space invader in our cities. They also add walking distances. Supposing an old library, shop, entertainment venue, school or health centre is demolished. The new one, that councils will lobby for and politicians will pat themselves on the back for having helped fund, will likely be subject to parking requirements that add to metres squared of asphalt. Some of the housing styles in our most highly valued older suburbs would not be buildable today due to red tape parking requirements that inconvenience walkers and entrench driving dependence.  

Time taken

Space and time are related in intersection design. The wider the road the wider the intersection. And the longer it takes to pass through it. And there are more possible movements when it's complex. That leads to longer cycle times. Long cycle times means that if you arrive just after your light has turned red you have a long wait until you can next go. Waits are even longer if the intersection design and light timings require you to complete your crossing in multiple stages as is frequently the case, especially for older people.   

Speaking of whom, do we wait longer to cross signalised roads today than our grandparents would have? Let's go back to the 1930s when traffic lights were proposed on some main roads after earlier trials. Articles here and here discuss the issues. 

Signals were proposed due to a high number of crashes (even back then they used the misnomer 'accident'). They could be synchronised (timed) or vehicle actuated with the latter suggested for Point Nepean Rd near Martin St (amongst other locations). The 1940 article says that the real danger spots (for motorists) were at 'busy but not obviously busy' intersections. Somewhere like Camberwell Junction was 'so obviously busy' that it engenders care do was less dangerous. Even then they knew stuff that is starting to be 'rediscovered'.

Nepean Hwy near Martin St got their lights in 1939. A reason given for their existence was to allow pedestrians to cross despite heavy traffic, especially on weekends and race days. 

The lights ran on fixed cycles of 45 seconds total.  Martin St got 11 seconds (green) with 4 seconds amber while the busier Pt Nepean Rd got 26 seconds (green) with 4 seconds amber. In the worst case scenario if you just missed a green on Martin St you would wait 34 seconds before your time comes around. Remember those times, noting that the then Point Nepean Rd was one of Melbourne's busiest.

There were several different styles of traffic signal in these early years. One style was developed by Charles Marshall who was concerned that existing signals did not give warning that they were going to change. His signals did with two dials at 90 degrees facing oncoming traffic on each road. These showed how long you needed to wait and how long you had. Like today's lights Marshallites had red, amber and green aspects. Some background on the Marshalites is here

In 1947 Charles Marshall offered to install his lights much further down Point Nepean Rd near Main St and Tyabb Rd (Mornington). Councillors expressed varying views on their desirability. Cr Bradford was concerned about 'heavy traffic that would be stopped every 40 seconds'. Cr Nunn said that it 'was a heartache for a mother to send her children to school' (at a time when most walked independently) and that (traffic) 'having to stop for a short time' should not be a worry. Cr Fielding agreed, saying that 'the motorist, instead of learning manners, was now worse than ever before'.  

Of interest from that discussion was the 40 second wait time cited. Below is a video demonstrating Marshalite signals as used on Nepean Highway around areas like Aspendale and Chelsea through to the 1970s before being replaced with conventional all-electronic lights. 

Look at the seconds elapsed as the hands move to where they start. You'll notice that the total cycle time is 45 seconds. The red and green areas are roughly equal, meaning that each direction gets about the same amount of time.  

A disused Marshalite signals sits in Chelsea's Bicentennial Park. While not currently operational they were restored so that they were when they were first moved there. The walk phase and its associated amber is about one-third the total cycle time. 

The Martin St Brighton lights mentioned before were apparently not Marshalites. However as you saw from the article they had a 45 second cycle time. As did the Marshalites you read about and saw demonstrated. Thus it's fair to conclude that 45 seconds was not an unusual cycle time for traffic signals, even those on major roads. 

While not perfect for dense walkable areas, a 45 second cycle time still stacks up quite well (a B level of service) for walking connectivity when viewed from the perspective of the Department of Transport's Movement and Place Framework for active transport. Lights with short cycles have far fewer people crossing against them than where cycles are longer.  

How long does it take today? 

Basic walking needs have not changed much in decades. What was poor for a traffic signal 80 years ago will still be poor today. This likely extends to matters like willingness to wait and risk-taking behaviour. What has changed has been the roads and the traffic signals. Below is a video case study from the intersection of Nepean Hwy and Martin Street in Gardenvale. 

This is not an isolated case. Projects other than road widenings, such as level crossing removals, can also increase walking times. This is especially if they foist current traffic engineering practice onto century old communities designed and built around trains and walking. 

You only need to go a bit further down Nepean Highway to find an example. Chelsea recently got its level crossings removed. Local walking access conditions changed. Gone was the old station underpass whose exit aligned with pedestrian lights across to the shops on Nepean Hwy. This got replaced by a footbridge that, unlike the old underpass, has no pedestrian lights near its highway exit. The result is that walkers from the bus or nearby streets have a less direct, unsheltered and unshaded walk before they reach inconveniently placed lights and often have to backtrack to their destination. 

This poor outcome is a direct result of LXRP-associated engineers ignoring walking as an important local transport mode and its need for good permeability at multiple points. See the consequences for yourself in "Hot, Wet & Disconnected" below.  

But wait, there's more! Not mentioned in the video was that pedestrian light timings have lengthened, further extending walking times.  For example the (inconveniently located) southern pedestrian lights at Nepean Hwy has a maximum wait of 60 seconds versus 45 seconds on the removed crossing (conveniently near the old station's underpass). 

Another highway crossing further north near The Strand has an even longer 80 second maximum wait at lights on both (busy) Nepean Hwy and (quieter) Station St. This is despite  Kingston Council advocating it as a major pedestrian connection and even partly funding a footbridge. Notwithstanding their effect on walking access and safety, signal timings are considered obscure technical business rarely discussed in public. 


Walking in metropolitan areas today is slower than what our grandparents (if they lived in Melbourne) would have experienced. This is due to less connected street patterns and double or triple the waiting at signalised intersections. Also, due to lower car traffic volumes, there would have  been more times they could cross mid-block without waiting long. All are direct consequences of thousands of small decisions made over many years. 

The Department of Transport is currently reviewing traffic light timings . However there's no transparency or word on the values that guide either it or technology like Smarter Roads. Especially for people who walk for transport in the suburbs. Something like the Movement and Place framework may list movement as a priority for a major road corridor and give car traffic free flow. However this risks downplaying the road's role as a barrier for walkers needing to cross it.  

Our grandparents would have waited half the time for a tram as we do today. This is because tram frequencies have dropped (notably in the 1950s and 60s) with only a weak recovery since. The most recent news is of a cut, rather than an increase, for tram frequencies. Also in-tram travel was faster due to fewer cars on mostly shared roads being in the way. Even major train lines, like Frankston, have longer end-to-end travel time than they did in the 1990s with the trip to the city now exceeding an hour while there's new parallel freeways that weren't there then.

Maybe it's these sorts of issues that our billions allocated for 'Big Build' projects should be tacking across Melbourne. Even if some (particularly traffic inducing freeways) are cancelled with funds diverted to (say) 10 000 'Little Build' projects instead. More people in more neighbourhoods could well be better off. 

Index to more Useful Network items here