Friday, August 23, 2019

Building Melbourne's Useful Network: Part 17 - Melbourne Airport (it's all about Sunshine)

The construction of a railway to Melbourne Airport is planned to start in 2022. It is expected to take up to 9 years. By that time Melbourne's population will have added nearly 2 million more people to hit 7 million. Much of that growth will be north and west of the CBD.  

Melbourne Airport is tipping even greater growth, with a doubling of passenger numbers expected in the 20 years from 2018

With this projected growth and Airport Rail not coming until the decade after next, there's a need to rethink the airport's public transport network. Not just as a stop-gap measure until Airport Rail starts, but also long term to better serve surrounding areas that Airport Rail won't benefit.

This is what I will attempt in today's Useful Network. 

Existing Useful Network

Ten years ago there was no useful public transport to Melbourne Airport. There were just occasional buses to a few places, with multi-hour gaps between trips. At the time I suggested some upgrades here

And there were improvements, starting the following year. Most significant is the SmartBus 901 orbital via Broadmeadows and Epping to Frankston. You can see this on the map below (7 day interactive maps here). 

Other regular bus routes operate from Melbourne Airport. These include 478 and 479 to Airport West and 479 to Sunbury. They're also better than they were 10 years ago with an upgrade in 2014. But they're not yet up to a Useful Network frequency and service level (explained in more detail here). 

Despite its size Melbourne Airport's public transport has just one frequent route from one direction - far less than major destinations like Chadstone and Monash University or smaller hubs like Coburg or Mentone. Even the much smaller Adelaide airport has better local bus connectivity with its surrounds. 

My assessment omits Skybus to the CBD. This express service operates frequently day and night. However it is not on the myki ticketing system. The premium fare makes it unsuitable for budget travellers or low-income airport workers. Neither is it fast for local trips where backtracking is involved like to Broadmeadows or Niddrie. Skybus also runs to other destinations like Werribee and Frankston but at lower frequencies. 

Public transport versus driving speed

Melbourne Airport has a metropolitan (and indeed statewide) catchment. Its service area is bigger than any university or shopping centre. A change to a less-than-frequent train at Broadmeadows (not frequency harmonised with the bus) helps for some destinations to the north and south. However, despite its importance, the airport is  not reasonably accessible from many other directions. And even if you factored in the CBD Skybus, the geometry makes it poor for access to surrounding areas.

How does public transport (as it currently exists) compare with driving to the airport from surrounding areas? I  used trip planners to compare travel speeds. The 'snail map' below shows public transport travel time as a ratio of driving time from selected locations.

Origins without snails have quite fast public transport access. Often because there's a direct route or a single change only. Places with big snails have the slowest access.  More detail is in the table below. I used the average for each mode to determine the ratio.

The above numbers don't include parts of the trip time for both driving and public transport. For example driving time excludes time spent parking and subsequent access to the airport. Public transport time excludes the wait for the first trip. Also the effects of low frequency have been omitted. This is most notable for Sunbury which has only two trips to the airport on Saturdays. In all cases starting locations are train stations or tram stops, except for Caroline Springs where it's the town centre. The overall effect is that, if anything, this comparison favours public transport. Include these factors and actual ratios will be higher than given.
Slow from Sunshine

Where are the Irish joke origins?

That is the places you wouldn't want to start from if going to where you wanted to go - in this case the airport.

Sunshine stands out.

People can run from there to the airport much faster than any regular public transport option. The latter is both indirect and requires a minimum of two changes.

Sunshine is not only an important place in its own right. It's also a major transport interchange. All the state's busiest regional rail lines go through there. As do major metropolitan train and bus routes.

What about the other 'snails'? The other two big ones, Tarneit and Caroline Springs, have direct train or bus routes to Sunshine. As does Melton. So if you sort out Sunshine you'll improve airport access from all of these locations plus others like Watergardens, Altona North and Footscray.

Greensborough is also quite slow, even though a one-seat bus ride from the airport is available. Because this route is indirect the fastest way involves a change of buses. But this doesn't have to be so and travel could be faster. Coburg also has potential for improvement, but that's longer term. More on both later.

Speeding things up with a better Useful Network

What could an upgraded Useful Network for Melbourne Airport look like?

Here's a list, in order of priority.

#1. New bus route 500 Sunshine - Melbourne Airport. Limited stop. Every 20 minutes Monday - Sunday with wide operating hours, timed to connect with trains from Geelong at Sunshine. This service would support growth of the Sunshine National Employment and Innovation Cluster, bringing forward some of the benefits expected to accrue from airport rail. Equally important is improved  airport access from a large catchment (including many regional rail lines) where it is currently difficult.

A 10 to 15 minute train-bus interchange time allowed at Sunshine would assist the less mobile, those with luggage and provide a buffer for minor train lateness (reliability being more important than speed for to-airport trips).

This route would operate until 2031 or whenever airport rail starts. It would likely need 4 to 6 buses to operate. Much of Melbourne's west (and beyond) would benefit, including Tarneit, Wyndham Vale,  Geelong, Melton, Caroline Springs,  Deer Park, Watergardens, Altona North, Footscray and more. This is top priority with the wider travel time savings summarised below.

I haven't mapped the route. I'd go with the fastest between the termini. There would be few (if any) intermediate stops. This is because good speed is desirable and no intermediate stop is likely to attract more than a fraction of the patronage of the two end stops. But if they're on the way you might have a couple of stops in the industrial area before the airport and one or two nearer Sunshine. The 40 to 50 minute bus travel time estimate from Sunshine is conservative and is higher than the longest car travel time indicated above. It could be more during peak times but is likely to be less at most other times.

#2. Swap routes 902 and 901 at Broadmeadows so that Route 902 goes to Melbourne Airport and 901 to Airport West.  This would speed access to the airport from Greensborough and thus the Hurstbridge line. Easy connections will be possible to the 86 tram and thus the La Trobe National Employment and Innovation Cluster (including the university and an international student market). Swapping would also pave the way for future improved Upfield line connectivity - see later.

Apart from transition expenses, this change is cost-free. It would not increase the amount of service kilometres nor deny any stop a SmartBus service. A change will continue to be available at Broadmeadows for those who need it. However adding after 9pm Sunday evening service, even if a single bus shuttling between Melbourne Airport and Broadmeadows for a couple of hours, would be desirable to avoid people being left with no service to anywhere.

#3. Upgrade and simplify bus route 478/479. Operate all trips (including short-workings from Airport West to Melbourne Airport) as Route 479 so people have only one route number to remember and frequency is better communicated. Increase service to provide 7 day 20 minute frequency between Airport West and Melbourne Airport. Areas like Essendon and Moonee Ponds would be the biggest beneficiaries, with an easy connection to Tram 59 at Airport West. This would be time-competitive with Skybus, which requires an indirect trip via the CBD.

While a lower priority, the Sunbury portion could increase from every 60 to every 40 min, providing better connectivity with local buses (which, like the train operate on a basic 40 min pulse). A weekend frequency upgrade would also be desirable. This upgrade would require 1 or 2 extra buses to operate, depending on what is done at the Sunbury end.

#4. New bus route 416 Caroline Springs - Melbourne Airport. Route could overlap Route 418 alignment until Arthur St then continue east via Green Gully Rd, Keilor and Old Calder Hwy to Melbourne Airport. If operated every 40 minutes the times could be staggered with Route 418 to provide a combined 20 minute Caroline Springs - Keilor Plains - Keilor Shopping Centre service.

This is the lowest priority as Caroline Springs already gains from the new Sunshine route and this route may not be well used. Still there are local network benefits. For example it would connect Keilor to its nearest station and shopping centre. And Caroline Springs  Town Centre would gain a more frequent connection to the Sunbury line at its closest point. You might commence this service when Airport Rail starts, using resources no longer needed for Route 500.

Associated rail service upgrades, rail infrastructure upgrades, information and marketing

The above are all bus network changes. The following train service and infrastructure upgrades would complement these and improve connectivity to the airport.

#1. Stop Bendigo trains at Sunshine. This will allow improved connections to other trains and buses including the 500 bus to the airport.

#2. Upgrade weekend Geelong line trains from every 40 to every 20 minutes. As well as being highly desirable on its own (due to crowding) this would enable a high quality airport connection at Sunshine to the 500 bus (above).

#3. Construct a new station at Campbellfield on the Upfield line with convenient access to the 902 bus.  In conjunction with the airport swap (see previous) this would aid local network connectivity including to Melbourne Airport. The timing of this is not critical.

#4. Adequate information and promotion. PTV isn't strong at explaining and promoting bus services, especially as part of a multimodal network. Recent marketing has been creative but single modal, poorly targeted and sporadic. Even though a service every 15 minutes is qualitatively different to and more useful than one every 60 minutes for some reason they don't highlight the former and sell its benefits to drive patronage growth, which recently has been feeble.  Information at major interchanges is also limited compared to cities with more advanced network thinking, like Perth.

Because most people only travel to the airport occasionally promotion and information are important if this network (particularly Route 500) is to be successful. On-system information needs to be good and enduring, not only at the termini but also over a broad section of the western suburbs and Geelong within an easy connection to it. Online marketing, including through website, app, social media and journey planning channels, needs to reach other potential passenger segments including airport area workers who live locally and budget travellers, leaving time-poor and price-insensitive  passengers to the existing faster Skybus services.

Relationship with Airport Rail and the Suburban Rail Loop

All the above network improvements are designed to build patronage for, and eventually complement, Airport Rail. When that starts the Sunshine - Melbourne Airport bus route would no longer be needed, at least at its suggested service level. It could either be deleted or kept as a lower frequency all-stops service to connect areas away from stations if intermediate stops prove well used. Any buses saved could be put towards new or improved routes from surrounding areas that airport rail does not directly serve such as Caroline Springs (416 above), Keilor East/Essendon (Route 465 extension) or Greenvale/Craigieburn (new route).

Not only could the Sunshine - Melbourne Airport bus be considered to be the first step towards Airport Rail, but also a first step towards Sunshine - Broadmeadows connectivity culminating in the Suburban Rail Loop. This trip starts off by being bus to bus, then train to bus, then, finally, train only. Again the early presence of bus access shapes peoples mental maps of places they can get to and thus life decisions like where to live and work.

When can we have it?

Processes for introducing new bus routes are currently slow. These include not only the need to appoint a bus operator, procure buses and recruit drivers, but also a lot of 'back-end' data work with PTV. All that might normally take a couple of years. But if funding was available and there was a determination to fast-track the service, the flagship Sunshine route could be up and running sooner.

In any event a high-profile and well-promoted airport bus network that's operating by the 2022 election would show that the government is serious about airport transport, access to jobs, inbound tourism and connecting the state. It would also boost its credibility when it comes to bigger plans like the on-again/off-again airport rail, which has been the butt of public cynicism for years.


The network presented would greatly improve airport access from a large part of Melbourne's northern and western suburbs. Instead of being up to 4 times slower public transport may only be twice as slow as driving. While still slower, public transport has other benefits that may still make this a reasonable trade-off for people such as students, budget travellers and local airport workers. Even though most of the latter will likely continue to drive, the size of the airport precinct workforce could still mean patronage even if only a small percentage opt for the bus.

The Sunshine - Airport bus, in particular, could induce a culture of using public transport to the airport, especially from places like Melton, Tarneit and Sunshine. Its implementation would effectively build up the travel patterns of Melbourne Airport Rail so that when it starts there is already an established passenger base that just needs to switch from bus to train.

The above upgrades would require 6 to 8 new buses operating 7 days per week. Is this a far-sighted visionary upgrade that sets us up for the future? Does this network  represent good value for money? Or is it distributionally regressive, compared to if we used these buses elsewhere? Maybe there's routes and connections I've missed.

Please use the box below to leave your comments. Better still, if you want something like this  email
Public Transport Minister Melissa Horne MP or your local member in St Albans, Tarneit, Kororoit, Werribee, Niddrie or Essendon to let them know the difference improved airport buses would make.

PS: Want to see expanded Useful Networks in other areas? An index to them all is here.


Michael Angelico said...

If you're fiddling with the 477/8/9, would you also reroute it slightly so its catchment doesn't overlap with the 59 tram?

Peter Parker said...

478 & 479 don't duplicate the tram at all. They used to but reforms in 2014 removed all that, starting these routes at Airport West rather than Moonee Ponds. They got increased frequency as a result.

As for 477, that's a matter for another time as it doesn't serve the airport. Some have suggested routing it into Essendon Fields. It could also go via Hoffmans Rd but I had other ideas for that (a frequent corridor formed by removing 476's overlap with the tram) discussed here

Andy said...

As an occasional user of the route via Broadmeadows from Footscray I'd wondered why Sunshine didn't have a bus to the airport. With the planned rail link it would make sense to get people used to the idea of Sunshine as a major hub now too!

Kevin Murray said...

Thank you for giving this serious consideration. I sometimes take the PT option from Brunswick if I have a good book to read. The problem is usually the connection between the 59 tram and the Airport West bus. I'm surprised at how empty that bus is. The cost of a myki ticket is so much less than a taxi or Skypbus (and you don't have to watch obligatory video ads).

Malcolm M said...

Any thoughts on why the 901 and buses connecting with the 59 tram are so poorly patronised? If the target market for these improved services is the lower-paid airport workers, some stats on them would be worthwile, such as current suburbs of residence, turnover time in their current jobs, start times, finish times, etc.. It costs empolyers $82/month for a pass to the staff parking area, for which they would also be liable for fringe benefit tax. I suspect it would be perceived lack of reliability, poor service span, lack of service to their dormitroy suburbs. If occasional shifts start at 4 am, and others fininsh after midnight, a worker will need a car even if most of their shifts are within the service span of the bus service.

A new 500 bus would need to be promoted for staff rather than passengers, because we don't want a subsidised service to appear to compete with unsubsidised premium services run by Gull to Geelong, Skybus to Weribee and Tarneit, or the Ballarat Airport bus.

Re the tram 59 connection, there don't seem to be good waiting facilities at the tram terminus. Ptv could promote it much better, such as on the timetable of the 59 tram at stops, eg indicating which services connect with a bus to the airport.

Peter Parker said...

Malcolm - all good points. My own recent trips on 901 have seen about 7 to 10 passengers on it during the day on weekdays. If we imagine that as 4 x 15 min trips (2 there/2 back very tightly scheduled) that's >28 passengers per bus operating hour. Which is actually OK.

On airport workers it would be interesting to look at the pay structure. As well as the age structure. But there'd be lots on not particularly high incomes - especially all the retail workers etc. If there was ever to be 24 hour transport on Sunday - Thursdays then a Broadmeadows - Airport West - Sunshine route could connect the airport with the wider world (even if airport rail itself didn't run).

As for non-competition, the commercial bus services can continue to promote door to door service, probably faster travel overall, possibly hotel drop-offs. For a lot of tourists who hate interchanging at unknown places, and for whom the bus fare is a small percentage of what they paid for the flight these commercial trips will still be the best option for them. And the 901 via Broadmeadows doesn't seem to have hurt Skybus to the CBD in the slightest.

Agree on the promotion of airport services on other parts of the network. Something that PTV is terrible at - whether it's airport routes or access to shopping centres, universities etc. They simply don't have a strong patronage-maximising business mindset.