Friday, February 26, 2021

Building Melbourne's Useful Network Part 83: How fast can you add a bus service?


Unlike trains and trams, buses are touted as being flexible services that can be readily extended to new growth areas as people move in. Or even ramped up temporarily for a summer shopping or holiday season. Provided the scale required isn't too much, capacity is easy to increase with buses cheaper than trains and drivers quick to recruit and train. 

That's the theory anyway. Let's see if we can establish whether this is really the case. How much lead time does a bus service need? That's the question I'll tackle today. 

The short answer is that it depends. It can range anything from an hour to several years. 

Here are some examples that I hope can explain that wide time-scale. 

The unplanned rail replacement bus 

Despite their physical heft, railways are incredibly fragile. So many things have to work to get a train from A to B. When just one element fails, such as a signal, points, overhead lines or a train then service is suspended. The train operator phones bus companies and (normally) within an hour or two bus replacements start flowing. 

This responsiveness is aided by train operators having standing arrangements with bus companies to provide substitute services. Response can't be instant, particularly in peak times when the fleet is almost all out and spare drivers may be hard to source, but it's still impressively quick. The main issue is peak period capacity; a packed train of 1000 people needs to be replaced by 20-odd buses to carry its load.    

The council-funded summer bus 

Here's an example from last year. On December 10 the Surf Coast Shire Council was considering a shuttle bus between the popular resorts of Torquay and Jan Juc. In little more than two weeks, on December 26, the service was on the road. It ran for 32 days over summer. 

It didn't get a lot of use although that doesn't concern us today. What was impressive was that the council was able to introduce a service with so little lead time. Consider this when we draw some later comparisons.  

The planned rail replacement bus 

Sometimes railways have to be shut down for maintenance or construction works. This also means replacement buses for passengers. However the prior warning allows sufficient information, staffing, buses and drivers to be arranged so the scenes at stations are more organised than the apparent chaos during an unplanned  rail disruption. Work might be planned several months or even longer in advance, based on the maintenance or building program.  

The regular Department of Transport route bus service 

Here's the main thing I want to talk about. Permanent bus services. These take more time to set up. A family can get a house built in much less time than it takes to make even small changes to a bus timetable. And a new suburban route can take longer to establish than many infrastructure projects. 

Don't believe me? Bentleigh MP Nick Staikos has documented what he did to get the new route 627 bus up and running. I took a dive through his Facebook page to find the key dates involved. 

To put this in a wider context, at the time Bentleigh was a marginal seat key for whichever party or coalition formed government. Nick Staikos was one of many first-time MPs whose victories flipped the state to Labor in 2014. 

His Facebook page gives a useful account of when and how the 627, as we know it, came to be. 

As far as I can see, his first foray into buses (as an MP) was in May 2016 where he held a local forum on buses. This sought views on how buses could be improved. 

Then in October 2016 he asked the then transport minister Jacinta Allan about running the 703 all the way to Brighton on Sundays. Doing this would correct a timetable anomaly where the 703 only ran its full route 6 days per week with the Bentleigh - Brighton portion unserved on Sunday. 

The 703 Sunday upgrade happened soon after - in December. Daniel Bowen's analysis on it here. That was quick! Maybe October's question was a Dorothy Dixer? Anyway the 703 upgrade was much needed.

The May 2017 state budget included funding for bus improvements in Bentleigh. Formal consultation was promised later in the year

In September 2017 Staikos, possibly hoping for a pre-election implementation, was getting restless and did this Benny Hill skit to emphasise bus coverage issues in parts of Bentleigh. 'Later in the year' had arrived but no sign of consultation yet. 

However there was consultation in April 2018 with two options presented for the as-yet un-numbered route. 

By September 2018 there were more details including a route number - 627. It would commence in early 2019. Hence he had something funded and concrete to sell for his November 2018 re-election campaign. 

In June 2019 minister Melissa Horne visited to announce that the 627 was starting on Sunday 16 June. Hence the schedule had slipped slightly from 'early 2019' mentioned before.  

I was on the first bus (as was he) and wrote about it here

So how long does it take?

The time from the first meeting to implementation was 37 months (3 years, 1 month). The period from when the money was committed to in the budget was 25 months (2 years, 1 month). There's a whole process of pre-budget work involving various departments that would be several months before the budget so even if it was something the Department of Transport was working on the lead time would be around 30 months (excluding its earlier design and planning work).  

I should mention that the 627 was a new route. With a 30 minute run time and a 30 minute frequency it would have required two new buses to have been bought. Parts of the route overlap other routes (like the 822 - inefficiently in my view) but parts did not. That would have meant that new stops, unique to the 627, would have had to be created. That involves survey, resident communication, data system and signage work. You'd also need to factor in time for public consultation early on if there were significant network reforms, though some recent changes, such as the Rowville FlexiRide, went in with zero consultation. Plus there's driver recruitment and training, though you can take that off the critical path if done concurrently with other work (possible if requirements are known early). 

In case you think this is atypical, consider the 469 bus through Valley Lake (in the current transport minister's electorate of Niddrie). This was funded in the 2018 budget (1 May) and commenced service in July 2020. Hence a 26 month time from budget to delivery, which is similar to the 627 experience. 

The next state election is just 21 months away. Hence if we want any new bus routes by then it is too late with current business processes and pace of work. Even added frequency on existing routes may be a stretch. 

Fixing these is something that must be looked at. Especially if the Department of Transport is sincere in its stated aim of adding 78 million more bus passengers per year (which would require substantial network reform and service addition). 

However a faster roll-out of improved bus services should be doable.  

There was skepticism early on that level crossings would be able to removed at the promised rate. 

Yet they have been, with some ahead of schedule. 

These have been much bigger and more complex projects than adding train and bus services. Doubly so for upgrades that can be done with the existing fleet and with minimal consultation, such as more off-peak/weekend trips or minor route straightenings. 

The level crossing experience has taught us that when there is the will and the budget then the means will follow. 

If we want better buses then the Department of Transport will need to ditch its quaint 'cottage industry' one-by-one practices. It will need to think about how it can efficiently deliver industrial-scale mass bus and train service upgrades and network reform, involving thousands of extra services, such as what we've seen Sydney do in recent years (under their Liberal government). 

With such ramped up capability then we might just be able to get things happening before (or just after) the next election. But only if there is the budget, will and direction from the government. 

See other Building Melbourne's Useful Network items here

Tuesday, February 23, 2021

Timetable Tuesday #106: Where are our truly full time bus routes?

Something that defines our suburban trains is their long operating hours. They will almost certainly be running whenever you’re awake. Even including the wee hours on weekends since Night Network started. Trams are more defined by their frequency. One will turn up within 20 minutes at almost any time. Buses are most defined by their coverage, with over 350 routes serving over 90% of Melbourne’s population. 

To the latter must be added the condition ‘when it runs’. It’s an important rider. Melbourne buses generally have short operating hours. The average person is still awake when the average bus route shuts down for the night. Those who work shifts, in retail or in hospitality can rarely rely on buses to be there when they need to travel.  

Big strides were made ten or fifteen years ago when something like 100 to 150 local bus routes were upgraded to a 7 day service with a 9pm finish under the ‘Meeting our Transport Challenges’ plan. That program though was never finished, with maybe 50 or 60 residential area short hour bus routes missing out. No successive state government has seriously picked up what the Brumby government did so much on but left unfinished. 

Also from that era were SmartBuses. Some provide main road orbital transport to complement the radial train and tram networks. Others enabled access to areas without trains such as Doncaster in the north-east and several major shopping centres and universities. Though their Monday to Saturday hours are long these top tier ‘premium’ routes still have a 9pm Sunday finish, just like many local back street routes. Thus not even SmartBuses qualify as full-time routes like trams and trains do (save the mickey mouse Stony Point line). 

Despite adding more than Adelaide’s population in the last thirty years, Melbourne has introduced not one full-time bus route for its own sake. All of the dozen-odd that operate are either (a) inherited from when the Victorian Railways or Tramways Board ran buses, sometimes as replacements for closed lines, or (b) compensation for more recent decisions (including some broken promises) not to extend a tram or to divert a rail service. 

Don't believe me? Here's the map showing all full time bus routes in Melbourne. The dozen or so shown accounts for less than 5% the total number of bus route in Melbourne. Thus over 95% of bus services in Melbourne are part-time. As opposed to that proportion of train and tram services that run full time (which I define as roughly 6am - midnight, 7 days). With few exceptions our full-time buses rarely extend beyond pre-1960s Melbourne. More in these after 10pm network service maps

Full time bus routes and background

Below is a run through the entire list of full service bus routes in Melbourne and their reason for existence. If anything I've been lenient as some listed still start later than trains and trams, especially on Sundays.  

190 Wyndham Vale - Werribee: Introduced in 2015 when Geelong trains were rerouted to operate via Sunshine's Regional Rail Link instead of Werribee and the electrified Werribee line did not get extended to Wyndham Vale. Geelong passengers could catch any train, alight at Wyndham Vale and ride the connecting bus to Werribee. While this undoubtedly happens Route 190 is likely of greatest use as a direct, long hours and moderately frequent feeder bus to Werribee. 

200/207 City – Kew – Bulleen/Doncaster: Old ex-Met routes inherited from the Tramways Board. Service was greatly simplified in the 2014 Transdev network restructure. Serves former Collingwood cable tram alignment on Johnston St closed in 1937.  

216 Sunshine – Footscray - City: Ex-Met/Tramways route serving a busy bus corridor through Footscray and Braybrook. Simplified and upgraded when the overlapping 219 was amalgamated into the service. Some trips extended as far west as Melton in the 1990s but service successively shortened to Burnside (sometimes wrongly described as Caroline Springs) and, more recently Sunshine with Route 426 commenced in the outer portion's place. Previously ran to Brighton Beach but this was pruned back when 603 was created. Spencer St portion replaces West Melbourne cable tram route that closed in 1935. 

220: Sunshine – Footscray - City: Ex-Met/Tramways route serving a busy bus corridor through Footscray and Braybrook. For several decades it extended south to Gardenvale but this was pruned back when 603 and 604 were created. Barkly St portion is a remnant of the closed Footscray tram network. 

223: Yarraville – Footscray - Highpoint: Ex-Met/Tramways route with little unique coverage but full service retained. Poorly used Yarraville portion a remnant of Footscray tram network but today services nowhere useful or is partly overlapped by other routes including the 472 bus. Trains at nearby Yarraville have been upgraded over the years to now run a 7 day 10 minute day/20 minute night service. Scope also exists to involve the 223 in a network simplification that could deliver a 10 minute frequency between Footscray and Highpoint.  

234 Garden City – Port Melbourne - City: Ex Met/Tramways route serving a long established government housing area. Became a single number in the 2014 Transdev restructure that simplified local services (previously part of the 250/251/253 group that ran through to Fitzroy North/Northcote). Bay St/City Rd portion replaced Port Melbourne cable tram which closed in 1937. 

246 Clifton Hill – Richmond – Elsternwick: Popular ex Met route serving the major Punt Rd/Hoddle St north-south corridor. One of just two individual regular bus routes in Melbourne that operate every 10 minutes or better interpeak weekdays. 

250 La Trobe University – Northcote – City: Ex-Met route providing a CBD service on a radial corridor between tram and train lines. Previously ran to Garden CIty before 2014 restructure.  Parts overlap Route 251 (a daytime route) to provide a combined higher frequency. Was numbered 256 many years ago. Replaced former North Carlton cable tram route closed in 1936. 

426 Caroline Springs – Sunshine: Ballarat Rd route that got its long hours because its Met-operated predecessor (the longer 216) also had long hours. Its introduction a few years ago brought full span service to the Caroline Springs Town Centre, something it lacked for decades when the 216 ran only to Burnside. For a while it enjoyed even spacing including train connectivity at Albion with the overlapping 456 along Ballarat Rd but a timetable change in 2019 broke this evenness. 

429 Sunshine South – Sunshine: A very short relatively new route that got its long hours because its Met-operated predecessor (the longer 219) also ran until late. Two thirds of the 429 overlaps the busier 428. Usage is understood to be very poor and scope exists to amalgamate the route into an upgraded Route 428 operating at a higher frequency and potentially better hours. 

600 St Kilda – Sandringham – Southland: Part of the complex and indirect 600/922/923 group since a botched restructure in 2002. Prior to that the 600 was the only route in the area connecting to trains at Sandringham. The Sandringham – Beaumaris portion of the 600 is an ex Victorian Railways route (901) that replaces a former tramway (hence one of the local street names). Ditto for the St Kilda - Brighton section. Operation then passed to The Met during the early 1980s restructure, then Melbourne Bus Link (1990s franchising) and finally Transdev (2010s refranchising). 600 is almost entirely duplicated by parts of other routes such as 603, 606, 922, 923 and the Sandringham train line, however the Department of Transport has lacked the will to implement substantial network reform along the lines of this

603 Alfred Hospital – Brighton Beach: A relatively new remnant of the previous longer ex-Met Route 216/219 pair that started in the Sunshine area and routed through the CBD and then along St Kilda Rd. This portion of the route is poorly used but largely kept (and in some cases) increased its long operating hours and frequency. 

604 Alfred Hospital – Gardenvale: A relatively new remnant of the previous longer ex-Met Route 220 that started at Sunshine and routed through the CBD and then along St Kilda Rd. This portion of the route is relatively poorly used but kept its long operating hours.

732 Vermont South – Knox City: Labor promised a Route 75 tram extension from Burwood to Knox City in its successful 1999 election campaign. That promise got half-honoured with a tram extension to Vermont South and a ‘Knox Transit Link’ bus, connecting with each tram, for the last few kilometres to Knox. That Transit Link came in the form of trips added to the Route 732 bus that on full length trips runs from Box Hill to Upper Ferntree Gully. Because the tram operated over long hours, so did the bus. However in 2016 when the 75 tram was chosen as one of the six routes that would operate Night Network service the 732 bus did not get extra trips to match. Usage is relatively low, partly because much of the 75 tram is a slow 'bring a packed lunch' route and connections to the not very frequent Alamein train line (that could have sped travel) it intersects with are weak. 

Mention should be made of two former but abolished long-span routes. The 571 was a 'Trainlink' bus that ran from Epping to South Morang. It was born after the government reneged (temporarily as it turned out) on a 1999 rail extension promise. The 571 was deleted when the train extension eventually opened in 2012. 

Then there was the 896. A Cranbourne East station was another 1999 Bracks promise that was even less honoured than the Knox tram. Later a bus meeting every train commenced, operating on a confusing circular route with several deviations. It was the nearest to a 24 hour service (on any public transport mode) Melbourne had with service from roughly 4am to 1am to feed and meet every train at Cranbourne. However the 896 was deleted in favour of simpler and straighter routes serving more of Cranbourne in the 2016 bus network restructure. 


Many of these routes are skewed towards the inner suburban transport 'haves' rather than the 'have nots'. Examples of 'protected species' routes that likely don't justify full time service include 223, 429, 603 and 604 due to their low usage, low catchment population or closeness to full-time trains and trams. 

In contrast whole local government areas such as the cities of Hobsons Bay, Melton, Hume, Whittlesea, Manningham, Maroondah, Monash, Greater Dandenong, Cardinia, Casey, Frankston and Mornington Peninsula contain not a single full time bus route despite wider spacing between, and in some cases no, train or tram lines. Our population growth has overwhelmingly been in these areas but the number of full time bus routes has not increased to suit. And jobs tend to be only rarely near stations in these areas. 

It would thus appear there is little relationship between whether a bus route is full time and local needs. History is a far better guide to whether a bus runs full time than any objective data like populations, demographics, demand or any sort of conscious planning. Hence, despite its formal duties in bus network planning, the positive influence of the Department of Transport (and predecessors) in the area of full-time bus routes is best described as minor for at least thirty years.    

Conclusion: Ask for a train if you want a good bus

My advice remains that if you want a full time bus route you should always advocate for a train or tram, despite sneers from professional transport planners who might dismiss your scheme as inappropriate, expensive or not cost-effective. 

They might suggest a bus instead and they may well be right. The problem is that advocacy directly for the full-service bus of the type discussed here has had a 100% record of failure. I remind you again that although Melbourne has grown by about the population of Adelaide in the last 30 years, we have not added a single full service bus route for its own sake.  

Planners' barbs about 'crazy schemes' or dark mutterings about 'poor cost-effectiveness' are not fatal to a project getting political support and then funding. All it needs is for some numpty in the department or government to think it a good idea for some non cost-effective proposals to get up. 

For example the government is dragging the chain on reviewing dysfunctional bus networks in areas like Knox. But introducing service styles known to either collapse under pressure, overlap other routes and/or underperform on patronage such as FlexiRide Rowville are apparently top priority. Such styles of service are apparently spruiked by (not necessarily expert) executives in places like Infrastructure Victoria and the Department of Transport who, in the latter, wield more power than its rarely heard and seen expert planners.     

If asked privately before it was announced as a political promise, many such planners would not have endorsed the biggest project of the lot, ie the Suburban Rail Loop. Yet it was instantly publicly popular and is now government policy.

Urban and transport planners are less influential than some think, especially on the big infrastructure decisions. However if transport executives want some intellectual firepower to help explain why certain things that aren't policy should not be done then their department's on-tap cadre of planners is a handy resource to have as a sort of " 'no' army" when responding to correspondence etc. 

Notwithstanding that, what something that the Suburban Rail Loop (along with low interest rates) has done is to lift the threshold on what you can suggest without being laughed out of the room. After all, even if your pet project is large it will almost certainly still be cheaper and quicker than the SRL. That makes it easier for advocates who want something badly enough to hold firm, even in the face of a lukewarm or negative official response. 

And if you campaign for rail well and a hungry political opposition or desperate government takes the bait then you may get something that marks an improvement. Even if they break their rail promise in favour of an enhanced (ie full-service) bus like what South Morang, Cranbourne East and Knox City got. 

This might seem a roundabout, and some might say underhanded, way of getting stuff. But there appears no history of anything else working, with even SmartBuses (which Doncaster and Rowville, who also asked for rail, got) not qualifying as full services that genuinely match trains and trams with regards operating hours and frequency. 

Friday, February 19, 2021

Building Melbourne's Useful Network Part 82: Transport to Sunshine National Employment and Innovation Cluster

Various plans over the last 60 or more years have tried to define suburban centres of more intensive development. Some have succeeded, some have not. If you go back to the 1954 plan, for example, you’ll see that both Box Hill and Moorabbin were nominated as district centres. However Box Hill succeeded while Moorabbin failed (although its light industrial part succeeded). There has also been intensive development at sites not foreseen in the 1954 plan, including around drive-in shopping centres like Chadstone and Northland and inner suburban precincts like Docklands, Southbank and South Yarra.

Motivations for developing suburban subcentres varies. It could be to relieve pressure on the crowded CBD, allow more jobs nearer where people live or to make optimal use of transport infrastructure with more bidirectional trips. Or, more cynically, it could be to salvage some benefits from previous bad land use planning decisions (eg locating Monash University’s Clayton campus in an unserviced paddock) or benefit local landholders and developers.

Clusters and their transport network needs

About 12 years ago Melbourne@5 million came out. It defined a number of ‘national employment and innovation clusters’. These included precincts around major universities and established suburban centres, though there were a few greenfields ones, like Werribee East, as well.

The term ‘cluster’ is important. They are not centres. With a centre you expect a small area that is reasonably walkable from end to end. Whereas ‘clusters’ are not so dense. Parts can be five or more kilometres from one another. This greatly affects how we should plan and run transport to ensure good access within as well as to it.

Box Hill and Dandenong are established suburban centres. They typically have a train station with a bus interchange close by. Bus routes terminate at the interchange. In all but one case if you wish to travel from slightly south to slightly north of the centre you will need to take two buses. Buses may twist and turn to enter the interchange, be infrequent and be difficult to understand. Even a short 2km trip within a centre precinct could take 30 minutes or more, with walking often faster due to indirect route alignments and low frequency.

Trams in the Melbourne CBD follow a different model. For a start most through-run. Hence trams from the south will pass through the CBD, often terminating a couple of kilometres north at Melbourne University. Similarly trams from the north finish at the south end of the CBD (or even go well beyond). The same applies for trams from the east that finish in western areas like Docklands or Port Melbourne. Although trams aren’t the fastest mode their routes are mostly direct with no deviations into off-line interchanges. And if people want to change then they alight at a corner and change to another tram, that, because of the overlap of routes, is never far away.

NEICs such as Sunshine are several times bigger than the Melbourne CBD in land area. Therefore they need a CBD style of service. You can’t just provide good service at one point, such as just around a station. Instead it needs to be well connected within itself with routes that travel from one end to the other without deviating in the middle. Plus they should extend at least slightly into hinterland areas where more intensive residential development can support the cluster’s employment and other functions.

This means a network of direct and turn-up-and-go routes with easy interchange between them. An acceptable service standard might be every 10 minutes for connections to surrounding centres and every 3 to 5 minutes for short trips within a centre. While a 5 minute service seems high, it is appropriate for a dense area with short trips where waiting would otherwise be an excessive component of total travel time.

The movement and storage of cars is highly energy and space inefficient. At least some journeys need to be on more space efficient active and public transport for NEICs to develop to the density desirable and for them to provide opportunities to all. Improved train and bus networks are key to this. 

Enough with generalities. Today I’ll zero in on Sunshine’s NEIC and its public transport needs. Again I need to emphasise the area’s size. It’s not just the area around Sunshine Station. Major parts of it are not really walkable from it, including the Victoria University campus and (especially) the so-called Sunshine Hospital which is really in St Albans near Ginifer Station. That’s a long and unattractive hike from anywhere in Sunshine proper.

(map from VPA)

Existing Sunshine 10 and 20 minute networks

Sunshine is quite well served with train and bus services operating every 20 minutes. However nothing runs every 10 minutes or better outside peak. The best it gets are a few routes such as the 216, 220, 410 and 903 every 15 minutes on weekdays and as infrequent as hourly on weekends. The 903 SmartBus is the highest profile route but its 30 minute weekend frequency and less useful destinations make its benefits less than they should be. However it is the only route that travels both sides of central Sunshine in a pattern that could be replicated more widely.  

10 things Sunshine’s future network needs

Here are my ten top things Sunshine’s public transport needs if it is to fulfill its potential as a NEIC and make itself accessible to the wider surrounding region. Most are relatively modest service upgrades possible within three years with the political will and funding.

1. Trains every 10 minutes to Watergardens. Current off-peak trains are only every 20 minutes (day), 30 minutes (night) and 40 minutes (Sunday morning). The template for service should be what the Frankston line has had since 31 January, ie a 10 minute daytime Monday – Sunday service and a 20 minute maximum wait early morning and at night. Key beneficiaries of the improved access include Sunshine Hospital and Sunshine CBD. The importance of this upgrade cannot be overestimated; without it waiting time can exceed travel time for many trips. 

2.  Ballarat Rd MegaBus/future light rail between Sunshine and the CBD. Ballarat Rd is home to two VU campuses and will be near the proposed new Footscray Hospital. It has several bus routes along it but none have the high frequency and profile needed. Also the corridor’s main routes like the 220 and 410 don’t consistently run along it. The MegaBus would deliver a single simple route with an all day 5 – 10 minute frequency between Sunshine, Ballarat Rd, Footscray, Docklands and the CBD. It would have high quality stops and bus priority. Quiet electric buses could run along it within two years with potential for light rail within ten years. More here

3.  SmartBus to Highpoint. Rerouting the 903 orbital to run via Highpoint would provide a high quality link from Sunshine to Highpoint and Essendon via Braybrook, better connecting local activity centres. The cost of doing this is negligible but there would need to be some local bus reforms. More here.

4.   Fast airport bus.  Airport rail is proposed but won’t start until 2029. We could start a high quality limited stops fast bus much sooner. This establishes travel patterns early and increases the attractiveness of Sunshine as a destination for business. Plus it would slash airport travel times for a large slice of western Melbourne including the fast growing Wyndham, Melton and Geelong areas with connections from upgraded train services. More here.

5.   Geelong and Wyndham Vale rail service upgrade. Current waiting times can blow out to as much as 40 minutes at night and on weekends. The first upgrade should add trips so that waits never exceed 20 minutes at least as far as Wyndham Vale. The next stage could further boost Wyndham Vale’s frequency to every 10 minutes.   

6.   Melton rail service upgrade. The 31 January 2021 timetable added many trips. However there remain some long gaps and weekend service is not as good as it should be. An upgrade could boost weekend and evening service so that maximum waits at Melton are capped at 20 minutes. This would be a precursor to Melton electrification, with Melton eventually getting a full 10 minute service.   

7. Local bus network reform and simplification with 10 or 20 minute maximum waits. Initiatives such as the Highpoint SmartBus and the Ballarat Rd MegaBus could provide the catalyst for reform to other local bus routes, although some could be done earlier. Some would have very low costs. Potential upgrades include:

·         Route 216 upgrade to every 10 min weekday, 20 min night and weekends. 

·         Route 410 reroute off Ballarat Rd to operate consistently operate via Churchill Av (currently 410 follows an inconsistent route with an hourly deviation). Operate a consistent 15 min peak/20 min off-peak 7 day service with longer hours.

·         Route 428/429 consolidate to operate as a single simplified route 428 every 20 minutes 7 days per week. Service could commence north of station to improve main street access. More here

·         Route 408 upgrade with longer hours and Sunday upgrade to every 20 minutes. Would improve access to Sunshine Hospital. Potential western extension along Furlong Rd to Brimbank SC for improved hospital access in conjunction with other network changes. 

·         Route 418, 423 and 424 upgrade with longer hours and boosted frequency to every 20 minutes. Would improve access to St Albans and Sunshine Hospital area.  Also longer hours and higher weekend frequency for Route 425. 

·         Route 426 upgrade to every 20 minutes to Caroline Springs Town Centre. This would provide a simpler Ballarat Rd service connecting with trains at Albion. Potential  to merge with Route 410 to provide east-west route through Sunshine.

·         New Watergardens – Sunshine route via Sunshine Av. Formed by extending Route 419 to Sunshine with Route 406 extended west to St Albans to serve Main St East. Would operate every 20 min or better 7 days and better connect Sunshine with close in areas to the north.

·         Extended hours on Route 420 to midnight. Peak frequency increased to 10-15 minutes.

·         Longer operating hours on Route 400 between Derrimut and Sunshine.

·         A new Sunshine – St Albans bus route via St Albans Rd for improved local access as the precinct develops. Potential to operate near Sunshine Hospital and via Victoria University St Albans. This could be one of the St Albans local routes (eg 425) extended south. 

·         Improved industrial area coverage to Laverton North via a new route, potentially to Tarneit via Dohertys Rd. This could help create a job ready network for improved access to local employment. 

8.   Sunshine Station rebuilding and revamp.  Although the station is relatively new, local walking and biking connectivity in the area remains poor, especially across the rail corridor near Sunshine Station. The south side has huge potential if connectivity was better. An ambitious concept suggested by some could be to move the station about 150m to the north-west to bring it closer to the major activity centre at Hampshire Rd and have the buses between where the tracks diverge.  Multiple entrances and connecting walkways, at both ends of the platform, are essential to maximise walking catchment (something Melbourne is notoriously bad at doing even for major stations like Frankston and South Yarra). 

9. Albion Station upgrade. Currently it’s a dump. A southern entrance and connection onto Foundary Rd could make it a second station for Sunshine CBD. Bus stops would also be improved for better waiting conditions for those using connecting Ballarat Rd buses.  

10.  Various walking and cycling access improvements. To improve access along and across roads. 

11. (added later) Bendigo trains to stop at Sunshine. Again a major boost to NEIC access.


This has been my wish list for better transport in Sunshine NEIC. Most are relatively short term, or at least can be if there was the same will to introduce bus upgrades as implement larger more infrastructure-based projects. Have I left any off or are there some that you think are unnecessary? If so please leave your comments below.  

Sunshine NEIC is in the seat of Footscray (Katie Hall MP) and St Albans (Natalie Suleyman MP). Both are regarded as safe Labor seats. 

See other Building Melbourne's Useful Network items here

Tuesday, February 16, 2021

Timetable Tuesday #105: How are our trains performing with hardly any passengers?

Go back ten or fifteen years and transport execs and politicians would nervously await the release of the 'Track Record' train performance figures each month. For it was nearly always bad news. As fast as patronage was surging operational performance was dropping. 

On the hard-hit and electorally sensitive Frankston line, lateness surged from 1 in 20 to almost 1 in 2 trains at the peak of the troubles. The Brumby government had lost control of the system and there were widespread calls to 'sack Connex'.

Which eventually they did in 2009 by not renewing its operating franchise and giving the job to Hong Kongers. But changing a few bosses at the top and leaving everything else the same gave no relief, with performance continuing to deteriorate in Metro's early years.

You can see the trend in the Track Record graph below. Good punctuality up to the end of 2003, a long-term fall up to the 2011 nadir, a fast recovery for a few years after then, as I pointed out here, a slow decline up to 2020. There were also some industrial issues, making it not a nice time to be a novice minister. Click graph below for clarity.

You'll see that the 2003 - 2011 slump coincides perfectly with the 2000s train patronage surge as graphed below. As trains got more crowded their dwell times increased as people struggled to get on. 

And it was important for them to get on; it was quite possible that the next one would be cancelled or even fuller. And timetables had numerous inconsistent stopping patterns (like Ringwood trains still do) so even if everything ran on time waits could still be longer than desirable to the next. 

The patronage graph stops at 2011. But growth has levelled off so there's not much to see since. For example annual boardings would not have exceeded 250 million trips per year in 2019. That provided breathing space for the big punctuality turnaround that happened around 2012. Contributing factors included reformed simpler timetables, particularly on the unreliable Dandenong, Frankston and Werribee lines and some improved peak frequencies.

There were at least two other factors as well, one enduring, one (fortunately) not so. 

Passengers noted some timetable padding with the 40-odd kilometre trip to Frankston (for example) now taking well over an hour. Mernda is another slow line. The longer run times made it easier to meet performance standards. Depending on how you saw it, it was either a recognition of reality or a 'free kick' to the train operator. 

Train cleanliness also slipped. Metro around 2014 was running a filthy and heavily graffitied fleet. Laxness here would have helped operational statistics if the threshold for taking trains out of service was higher. It's worth nothing that 2014 was also a mid-franchise year, a time when PTV's franchise contract management was weak and operators could cut corners and get away with stuff (and not just with trains either).

However as  contract renewal approached in 2017, Metro smartened its act and greatly improved train presentation. Apart from neglect of Saturday and Sunday morning carriage cleaning (an unresolved issue from Night Network), train presentation is now as good as it has ever been (for at least 30 if not 50 years).  

Returning to punctuality, you'll see an uptick at the end. This coincided with city commuters working from home rather than catching the train due to COVID-19 restrictions. Like Yes Minister's hospital with no patients operational performance is best when there are few customers. 

The critical difference between modes is what happens when patronage rises beyond a few people. If more try to use flexible route buses then performance slips greatly with every added passenger (as the bus has to deviate past their house, slowing the trips of all other passengers). Whereas an extra passenger on mass transit modes has close to zero impact provided the system is not over capacity. The graph below tells the story, with performance diverging as patronage rises. 

Most benefits of public transport (think space efficiency, economic, social and environmental benefits) accrue when it is well used. It follows then that we only get these benefits if the mode chosen can support high usage while remaining reliable and punctual. 

Trains and fixed bus routes can achieve that whereas flexible route buses do not. Where flexible bus routes are carrying more than a handful of people at a time then they should be upgraded to a regular fixed route bus for better punctuality. If a regular bus regularly reaches capacity then it needs higher frequency, larger vehicles, its own road space, wider doors, off-vehicle validation and all the other things that makes it more like a rail service. And if regular trains get overloaded, like ours did in the 2000s, then we need a better system and service, comprising improved maintenance, upgraded infrastructure, more trains, revamped signalling and refreshed timetables. They happened later than they should but we eventually got improvements here and performance rebounded. 

The past 15 years has thus taught us what happens to performance when patronage booms. What about if it slumps? 2020 was an interesting case history of that. Below is a close up of Metro train performance. When the COVID restrictions were more severe patronage fell and punctuality improved. Then when they were loosened some passengers returned and delays increased. However both remain well down on pre-COVID times. 

Hence the inverse relationship between patronage and punctuality appears symmetrical, whether usage falls or rises. Trams and V/Line appear to exhibit similar behaviour. Data for buses, being  Melbourne's 'invisible mode', is not reported. However buses may turn out to be the 'silent sufferers' as car traffic rises to exceed pre-pandemic levels. A little more on them later. 

Less clear is the performance on service delivery (ie trains that are not cancelled). The graphs below show the 20 year trend. There was excellent performance in the early franchising period that was never quite met since. In the most recent pre-COVID period there was a steady deterioration. It might not look much, but a drop from 99.5% to 98.5% means triple the chance of your train being cancelled. 

Below shows more recent times. Cancellation rates approached 2% in January 2020. This was close to rates experienced in the Brumby era (a bad time for train travellers). This happened due to a gradual drop in performance rather than due to high-profile network meltdowns that were more common 10 years ago. 

Performance recovered in September/October 2020, ie around the height of the lockdown. In those months 99.4% of trains ran. This was very close to the excellent 2001-2003 performance. However train patronage was depressed and the good numbers were not to last, even though patronage is still relatively low. In the most recent available month (January 2021) service delivery had dipped below 99%, ie similar to some earlier pre-COVID months. There must have been something that was done prior to 2003 that we haven't been doing lately since cancellation rates have never been as consistently low as back then.  

How has V/Line gone? It's been a bit different from Metro with their paths diverging. Metro improved about 9-10 years ago while V/Line showed no such gain. In fact the latter's service delivery continued to deteriorate. That's a worry given that whereas the wait after a cancelled Metro train is rarely over 30 minutes, it can be several times that for a V/Line train due to lower country area frequencies. On the bright side punctuality improved. Mostly to about 2004 levels but during the height of the COVID lockdown (when Melburnians couldn't travel more than 5km) to the best numbers for over two decades. That's been followed by a post-COVID slump (graph below). 

It is clear that V/Line has systemic issues with its performance, like metropolitan trains had about 10 years ago. The Ballarat Line Upgrade and new timetable that came on stream January 31 might help. Or it might not. If we look at history, the major V/Line infrastructure works and timetable upgrades in 2006-2007 and 2015 (ie the Regional Fast Rail and Regional Rail Link projects) appear to have had no discernible impact on performance. However they did boost frequency (reducing waiting time until the next service). Also patronage (which probably harmed punctuality) increased.   

Trams have a pattern more like V/Line but with less volatility. Cancellations are up on five years ago. Punctuality had a period of improvement but this had levelled out by 2015. However the reduction of road traffic during the height of the lockdown delivered big punctuality gains during our most confined months. We were freer in January 2021 than we were in October and accordingly traffic volumes have risen as have tram delays. 

Buses? Monthly performance information is not publicly reported as it is for trains and trams. However there are clearly issues with road congestion as advised by bus operators around Melbourne. 

To summarise, all modes had improvements in performance when we had little road traffic and few passengers. All have given up some of those gains afterwards. However there remain systemic issues, such as V/Line's unreliability and rising traffic for trams and buses that will have a large bearing on performance longer term.  

See all Timetable Tuesday items here

Friday, February 12, 2021

Building Melbourne's Useful Network Part 81: VTAG's 2021/22 State Budget Submission

 The 2020 state budget seemed only a moment ago, yet we are already talking about the next one. It's not that time is whizzing by, it's because last year's was delayed due to COVID-19. Preparation for 2021's would be well underway right now. 

This year's budget is not the last before next year's election.

But, because even very modest service upgrades have long lead times the 2021 budget will be the last time to fund improvements that have any chance at all of being implemented by the November 2022 election. Even then it will be very very tight. 

There are several reasons why a family can get a house built in less time than the Department of Transport can organise even a minor change to a bus route or timetable (let alone a local network review). However a large part of it is due to paralysed internal processes. 

Because policy and resources are so heavily skewed towards building infrastructure, service upgrades are considered exceptional rather than regular business. Hence efficient internal processes to upgrade networks and services are less developed than with (say) the level crossing removals which has a large, organised and well-funded program. 

Hence 2021's budget is really important and advocacy groups need to have got their budget proposals in. One of those groups that has is the Victorian Transport Action Group

VTAG dubs the 2010s as the 'decade of infrastructure' and welcomes the state government's investment on new railway lines, tracks, stations and level crossing removals. However it notes that service has been starved. Accordingly it advocates a swing to service, starting with the 2021 budget and beyond. This would transform public transport from a narrowly-based CBD peak hour service to a more broadly versatile network suitable for more diverse trips for more purposes at more times of the day.

As well VTAG is advocating smaller cost-effective infrastructure works to support some of the larger projects including the Metro Tunnel and Airport Rail. 

Four priority measures are proposed. These include: 

1. SRL SmartBuses. High quality SmartBus services that would connect Suburban Rail Loop stations to bring forward SRL benefits by decades and encourage early development and orbital travel patterns. 

2. Thirty major bus and train service improvements to simplify timetables, spread turn-up-and-go service and provide a 20 minute maximum wait on the rail network.  

3. A new concourse at Caulfield Station to relieve a pinch-point on the rail network that will only get more crowded with the Metro Tunnel and Airport Rail.  
4. Accelerated zero emission bus roll-out. An announcement was made in 2020's budget for a modest trial but other states have been progressing faster and more decisively than we have.  

You can read the submission and other papers here:

VTAG's 2021 budget submission

Caulfield Station - The missing Metro connections

- Electric buses, Now

I will discuss the first two measures in more detail below. 

SRL SmartBus

This initiative adds service to deliver high quality orbital transport at most SRL station locations. All but one routes proposed are upgraded or reconfigured existing routes. This keeps implementation cost low. Some of the concepts may be familiar to long-time readers here. 

a. Upgrade 733, 737 and 767 to SmartBus

These upgrade benefit already very popular routes that in some cases haven't had substantial service upgrades for decades. They serve major destinations including Box Hill, Chadstone, Southland, Monash University, Deakin University, Clayton, Oakleigh and more. 

The most important upgrade is the 733. This most closely replicates the SRL Stage 1. It serves proposed SRL stations at Box Hill, Monash University and Clayton. An extension to Southland would allow it to roughly parallel the entire Stage 1 of SRL. The Clayton to Oakleigh portion would be served by another route where coverage requires this. Despite exceptionally high patronage 733's Sunday service is currently just hourly. Weekday service is half-hourly off-peak. An upgrade to SmartBus would double (or better) service frequencies to every 10 min peak, 15 min off-peak and 20 min weekend. Some new buses would be needed for the peak upgrade and Southland extension but a high proportion of extra service kilometres would come from working the existing fleet harder for more of the day. 

Route 767 is included for a SmartBus upgrade as it would provide long hours and more frequent service between Box Hill and Deakin University (another SRL station). It would also provide better access to Deakin from the south (which is currently lacking). Chadstone would also benefit. The dotted line on teh map shows an optional extension of the SmartBus service to Southland, likely via East Boundary and Chesterville roads. However this would require a reconfiguration of local bus routes 627, 701 and 822 to maximise speed and efficiency. 

Route 737 is included as the third SmartBus upgrade as it is a popular route and links major centres. It would serve Monash Univesity and Glen Waverley (both SRL station locations). The SmartBus service would extend to at least Knox City. The dotted line shows a potential extension to Westall to serve a rapidly densifying and developing area. 

The above routes serve marginal seats in Melbourne's east such as Box Hill and Burwood that would be endangered if the government receives even a small swing against it.  

b. Sunshine - Melbourne Airport rail precursor

This is a precursor to the Melbourne Airport Rail Link (and a later part of the SRL). Services would operate at a good frequency with fast transit due to limited stops. It would strengthen Sunshine as a transport hub from a wide part of Western Melbourne and regional Victoria. 

As it is a completely new route, new buses would need to be bought. But it would become a high profile part of Melbourne's public transport network and help establish travel patterns ahead of airport rail. More detail in my blog post here

c. SRL SmartBus across the north

This is the cheapest SRL SmartBus since it involves a simple swapping of two orbital SmartBus routes at two places (Broadmeadows and Greensborough). This provides large benefits, including more direct airport access from many northern suburbs and a direct Greensborough - Doncaster SmartBus. The 902 orbital is effectively reconfigured to become the SRL SmartBus across the north. 901 then gets to serve stops that the 902 is removed from, leaving no stop without a SmartBus. 

Thirty major bus and train service improvements

This is a package of service upgrades that delivers new turn-up-and-go routes between major destinations. Many are relatively low cost as they consolidate existing multiple bus routes or make use of vehicles that would otherwise be idle off-peak.

Some measures might be the sort of things that a government might wish to do before an election because they benefit marginal seats. 

I won't go through all of them but key initiatives include: 

a. New 10 minute service on the following upgraded and simplified routes and lines: 

- Deakin Uni - Box Hill express shuttle (upgraded Route 201 consolidated with 768)

- Highpoint - VU - Footscray (consolidated 406/223 on revised route) 

- Coburg - Preston - Northland - Heidelberg orbital (new 904 formed from western 903 portion merged with part of overlapping 527)

- City - Ringwood train interpeak upgrade (Belgrave & Lilydale upgraded to every 20 min)

- City - Glen Waverley train interpeak upgrade

b. Extra/revised SmartBus links

- Heidelberg - La Trobe Uni formed by splitting and extending eastern part of 903 to campus

- Essendon - Highpoint - Sunshine formed by rerouting 903 via Highpoint

- La Trobe University - Caulfield with merged 548 and 624

c. Evening train service upgrades across Melbourne

- Upgrade from every 30 to 20 min 7 nights. Watergardens, Upfield, Craigieburn, Mernda, Eltham, Lilydale, Belgrave, Alamein, Glen Waverley, Pakenham, Cranbourne. Follows on from January 2021 upgrades to Frankston, Werribee and Williamstown. An upgrade to every 20 min would bring us closer to service standards in Syndey (every 15 min until midnight) and reverse large cuts made in 1978 (when Melbourne was little more than half today's population and night time activity was much less).    

d. Service upgrades on high patronage/high needs bus routes

- Mostly extended operating hours and boosted frequencies (particularly weekends) on over 20 bus routes across Melbourne (list in submission). Relatively low cost working the existing bus fleet harder as most are off-peak upgrades.


That's one group's ideas as to service priorities for public transport in 2021. What's your view? Please leave your comments below. 

See other Building Melbourne's Useful Network items here