Are you being served? Commentary on the service aspects of public transport in Melbourne, Australia. Covers networks, routes, timetables, planning, co-ordination, information, marketing and more.
This state Auditor-General report is about the much-anticipated (and now under trial) move towards credit card acceptance on myki readers and an eventual transition to account based ticketing. It was tabled on Wednesday.
- The project is on track to its revised time-line after previous delays (due to contracting squabbles) and a cost increases ($137m more).
- 26% of revenue collected by the ticketing system goes to run it. VAGO puts that number at $2.96 out of a $11.40 full daily fare which seems high. If you looked at it differently (eg attributed a flat amount from each passenger to run the ticketing system) there'd be some ticket types (eg the $3 concession daily for a short regional trip) where the fare hardly offsets myki running costs. And note that this report was done before under 18s got free travel so ticketing system cost will be spread over even fewer paying passengers.
- Verifying concession entitlements is complex and high-risk. It will be done in later stages after the system is up and running for full fare passengers travelling in the existing Myki area. In other words full fare passengers will be able to pay with credit or myki cards but concession passengers will just have myki for a while yet.
- VAGO found that DTP couldn't really quantify benefits or demonstrate value for money. Thus continuing a tradition of haphazard administration going back at least 35 years with four ticketing systems under eight premiers.
There is a much better summary than the above and some discussion on the Reddit thread below:
History almost always repeats with public transport ticketing systems in Victoria. No other Australian state has made heavier weather of rolling out new ticketing systems than Victoria. Symptoms include much-publicised project time and cost blow-outs. And megalitres of newspaper headline ink.
There have also been performance issues but efforts to fix them for both Metcard and Myki made reliability at least acceptable in the last few years before the next system took over. In contrast scratch tickets were never successful (except for fare evaders) so the system got replaced.
The above did not mean that there were not some functionality limitations that the subsequent ticketing system tried to fix. And one poor decision can set a bad path that ripples for years.
Myki was intended to be a dual ticket system. Regular passengers would have durable cards that they could top up while occasional passengers would have a disposable short term ticket option.
They could have gone two ways with the short term ticket. Either it could be an expensive to produce cardboard ticket with the electronic smarts to open station barriers or a cheap paper ticket that needed to be shown to an attendant.
The project opted for the electronic cardboard ticket. Which came into use when myki started in Geelong. There was a time that you could see discarded short-term Mykis on the ground there. Their spiral antennas were visible when you held them up to the light. Short-term tickets worked but were expensive to produce, especially relative to concession fares.
The incoming Baillieu Coalition government of 2010 reviewed the Myki project. The project too far advanced to scrap, it proceeded in a reduced scope form. The cut meant that some parts of the state would not get Myki. There would also be no Myki ticket vending machines on trams and no short-term ticket option.
Hence even casual travellers would have to find a Myki outlet, pay for a piece of plastic they might never use more than once and then top it up with enough to pay the fare. And it would be awkward if they wanted to get a refund of unused credit or return their card.
The descoped Myki was fine for regular train commuters but a pain for tourists and other occasional users. Myki's clunkiness won it no friends amongst various civic and opinion leaders (the same people who also gripe about our lack of airport rail).
A bidding war during the 2014 election campaign led to the (counterproductive) CBD free tram zone being created, possibly exacerbated by Myki's issues. And the state government contracted with Conduent to update myki to allow credit card and mobile payments - something else that would help visitors and occasional users.
Things might have worked out differently had paper been chosen for the short-term ticket medium. Per-ticket costs (a few cents) would have been lower. Thus it might have escaped the descoping under the Coalition. There would have been fewer hassles for occasional travellers, less fare evasion on buses and likely less of a perceived need for updates to support credit card payments.
Claimed advantages of cardboard short-term tickets opening barriers may have been over-stated as the vast majority of passengers entering stations would be using durable Mykis. Governments could instead have concentrated on matters more central to providing good service, such as not descoping infrastructure on projects such as Regional Rail Link and Metro Tunnel and improved buses, instead.
Back to today's reality. Below is the time-line from the VAGO report.
Right now we are in the start of Phase 2, with the full fare paying public invited to try credit card payment on selected V/Line and Metro lines. If results are good this will be extended to all train lines in the existing Myki area. Followed by tram and bus.
Phase 3 involves extensions to more passengers (concession holders) and more areas (those not in the myki areas). The latter is potentially beneficial for areas like South Gippsland that are relatively close to Melbourne yet are still on paper tickets because their train lines closed years ago all their transport is provided by coach.
The audit flagged verifying concession entitlements as a complication. That's both for DTP administratively (with 150 agreements needed with concession authorities) and for the customer with use of an online portal involved. The concession platform is expected to cost $34m over 10 years (on top of $1.96b over 15 years for the main contract).
Then there's the political angle. This is an election year. The current long-standing state government is struggling in the polls. Even though it's generally poor policy, cutting fares has been politically fashionable as a 'cost of living' measure in some states. As has extending 'free' travel to more groups such as this government has done with under 18s (under the "Youth Myki" - confusingly the concession fare "Child Myki" is now just for 18 year-old adults!).
Especially if Myki costs about as much as it recoups in fares for some concession passenger trips the temptation to descope aspects of modernised Myki Phase 3 while also extending free (or very cheap) travel to some groups might just be too tempting for some in or out of government.
The next several months will be very interesting for transport fares and ticketing in Victoria!
Past fuel supply squeezes have involved policies like speed limit reductions (such as in the US in the 1970s) and odds and evens registration plate rationing at petrol stations. High fuel prices in the 1970s (again due to Middle East instability) led to a revival of inner city living and public transport in some cities. In Melbourne's case it was confirmation that trams were here to stay with new trams being ordered and extensions on routes such as 59, 75 and 86.
Some 30 years later it was fuel price rises coupled with a strong economy and booming CBD employment led to rail crowding in the 2005-2010 period. Pressures on the network led to a revival of investment in it, with projects such as Regional Rail Link and the Metro Tunnel being the result. Not uncoincidentally researcher Jago Dodson released the VAMPIRE index gauging the vulnerability of Australian suburbs to oil price shocks in 2006.
Today more of our cars are electric and working from home is more common. There has been population densification in some areas. On the other hand our cities have continued to sprawl. Corner stores have shut in established suburbs while new suburbs never got them. Smaller families, the rise of private schooling, school consolidations, car ownership, both parents working and increased traffic has sharply reduced walking and cycling to schools.
Then there's the burgeoning delivery economy. Yesterday I received a small (non-essential) item that could have fitted in a padded post bag and put in the letterbox at the start of my driveway by a postie on an electric scooter. Instead it came to me in a huge packing-material filled box from a courier (who was likely not well paid) driven to my door in a big truck. Surely that doesn't make sense in an oil conscious world.
What are some responses to the pressures to maximise fuel availability for essential uses for which substitution is not easily or quickly possible?
Below is a personal transport pyramid that I've made. It looks a bit like a food pyramid. To conserve fuel you want to encourage people to do as much as possible near the base and as little as possible of what's near the top.
Behaviour change might involve sometimes jumping one or two steps to a less energy intensive option. Or it might involve doing the same thing but less often. Eg combining several tasks in a single car trip. If homes have multiple cars you want the ones with the highest fuel usage to hardly ever be driven.
Many can't or won't make big leaps. But enabling smaller movements might still be desirable. For example someone going from public to active transport during crowded peak periods creates room for someone else to go from driving to public transport.
Current hybrid commuting patterns may also offer opportunities. Currently Tuesdays, Wednesdays and Thursdays are busier than Mondays and Fridays. But are there opportunities to encourage some commuters to switch to the quieter days? That's not to everyone's taste but even if a few people do it there may be a cascading effect where space is freed for others to shift modes.
Incentives need to have some thought about them and not just be knee-jerk reactions. They need to me mainly about making desired alternatives good, not just cheap. We should emphasise what can be done on a large scale, quickly and cheaply with existing assets. It needs the direct opposite thinking to the "Big Build" approach that has dominated transport planning and depleted our budgets. For example tactical reallocation of road space to favour more energy efficient / higher people throughput transport.
Pop-up bike lanes, zebra crossings at all busy roundabouts to reduce severance they cause, walker-friendly traffic light signalling, opportunistic land acquisitions to gridify the active transport network, more shade and seating, delivery lockers walkable from all homes and pop-up convenience stores in growth areas without them could all be part of the story.
Yesterday the state government quietly released Victoria's Active Transport Plan. The release could not have been softer despite active transport being central to reducing non-essential uses for fuel and freeing up supply for harder to substitute for essential uses such as farming. There was no media release, no mention in parliament and (critically) no new funding.
The state might try gentle persuasion to encourage councils to do their share for active transport but something dramatic, like monitoring (or sacking) a recalcitrant council like Yarra for pulling up bike lanes, might send a short sharp message to all councils with regards to expected behaviours in a fuel and climate emergency.
Perverse incentives need to be avoided, even though they are superficially attractive. And it is essential to offer no less incentives for active transport as may be offered for public transport. If you don't do that you risk the reverse, ie people switching from active to public transport, potentially crowding out those who might be switching to the latter from driving.
"Free" public transport (like suggested by The Greens) is one example that is likely to attract as many (if not more) people from active to public transport than from driving. Plus its distributional benefits are dubious as it helps those nearest the best service the most while not assisting regional and suburban dwellers with no or sparse service. There may well be an argument to review the equity of fares (especially for shorter trips that we charge quite heavily for versus our insanely cheap entirely diesel long distance V/Line fares) but "free" public transport is not the way to go and could even be counter-productive.
On the other hand service upgrades on routes that are busy and/or serve catchments sensitive to cost of living pressures is likely a better policy response that has a higher potential to attract people from driving. Economy with regards to fuel consumption is likely maximised if bus networks are reviewed to reduce inefficient overlaps.
Cheaper to operate electric buses sitting idle in a fuel crisis? Yes it's a thing in Melbourne. For all the hype about electric buses, we are doing a terrible job at using the fleet efficiently to carry the most number of people all week because network reform failed to accompany electrification. For example just 1 in 6 of the bus routes at Ventura's electric Ivanhoe depot operate 7 days.
The one that does (the 527) is only every 50 minutes on Sundays and is inefficiently overlapped by other routes. Scope exists for simplified bus networks to make timetables less lumpy, even out intervals and likely equalise loadings to optimise bus occupancy without overcrowding in high activity areas like Coburg and Preston, with an example involving the 527 presented here. The government got cold feet on northern suburbs bus reform in 2023 but a revival is justified given changed circumstances.
Another electrified depot, Kinetic's at Preston, is the base of other northern suburbs bus routes like 503, 506 and 508. The first two have short hours and lack Sunday service. 503 is the only public transport directly serving a significant high-rise and social housing area in Brunswick West. 506 is Melbourne's busiest bus route without Sunday service. 508 does have Sunday service but only at 40 minute intervals. As the major east-west route across Melbourne's inner north it has even higher patronage potential than the 506.
Because the inner and middle north have significant north-south routes (trains and trams) but limited east-west connectivity (entirely mostly infrequent buses) bus route simplification could assist in modal shift as the network transitions to a more versatile grid with more consistently easier connections.
Opportunities also exist in areas hardest hit by cost of living increases. Noting that higher fuel prices are inflationary, with the effect likely cascading through to essentials such as food. 7 day upgrades to routes like 802 and 804 cut food deserts by enable easier access to fresh food destinations such as Dandenong Market. Similar boosts to limited service routes like 538 (Campbellfield), 559 (Thomastown) and 844 (Doveton) would also deliver gains.
Acceleration of bus roll-outs in unserved growth areas is another winner. Time-lines for implementation are typically long. However at least some of this is a matter of political choice from a government that has got good at removing level crossings in less time than it takes to do the simpler job of adding a new bus route.
Some potentially redundant or over serviced routes have been rationalised since this item was written in 2020. The soon to happen upgrade of Sandringham line trains to every 10 min weekday off-peak might make the rationalisation of poorly used north-south bus routes in the Brighton area parallel to the train worth doing, especially if it enabled service kilometres to be transferred to higher patronage potential bus corridors in busier, denser or higher needs areas.
In Melbourne we are not used to upgraded timetables taking much less than 2 or 3 years from planning to implementation. Active transport projects might also take a while to plan. But things can happen much faster when there is the will such as a fuel emergency might encourage. Here this government can draw inspiration from the Bracks-Brumby government that in one 18 month period added a massive 8000 extra bus services per week and what cities like Paris have done with regards to active transport.
Hopefully 'never waste a crisis' becomes a guiding principle for DTP in the next little while so that a legacy of a better active and public transport network endures even if/when current tensions subside.
Since January 1, 2026 you haven't been able to rock up at a Melbourne train station and instantly see how much a trip, or a day out, costs.
Even though successive governments have reduced the number of fare zones across the state such that it's close to a flat fare (with a few cheaper or free exceptions such as Zone 2 only or before 7:15am).
Even though, especially in this time of surging fuel prices, the fact that you can travel all day anywhere for under $12 full fare, should be seen as a bargain Transport Victoria would want to promote rather than an impost to apologise for.
And even though because the replacement poster (which requires scanning a QR code with a smartphone to find out fares) has an effective date it is no cheaper to have a poster that omits fares than display them.
The usability impact of this is to make finding out fares more complicated, as follows:
Introducing unnecessary hurdles in finding fares makes passengers less aware of the value of all-day travel and features that should be selling points. Such as mobility across modes and across the state for no extra cost, Zone 2, weekend and after 6pm discounts and free travel period such as Early Bird on weekdays. While doing nothing to arrest rampant fare evasion that parts of DTP choose to deny, thus sapping the system of revenue.
Timetables
A month later came the "Big Switch" Metro Tunnel timetable with changes for some other lines as well.
That saw West Footscray to Dandenong get trains every 10 minutes first to last while service to Sunbury, Cranbourne and East Pakenham improved to a maximum 20 minute wait. Most other lines remained with 30 to 40 minute maximum waits with improvements coming later in the year for Upfield and Craigieburn. Gaps can even be 60 minutes on Sunday mornings as late as 9am in outbound directions.
In other words outside peak times the vast majority of the Metro rail network remains one where passengers need to check train times to avoid significant waits, especially if making connections. This is a peak-heavy commuter/regional railway as opposed to a true 'metro' system. Yet passenger information decisions are based on us having the latter, which is not the case at ~90% of Melbourne suburban train stations due to slow NDP implementation.
Notwithstanding the continued importance of specific times (as opposed to frequency information) on most lines for most of the day due to aforementioned periods with low frequency, Metro stations that had timetable changes did not get suitably updated wall timetables.
In case you haven't paid much attention, wall timetables came in two formats. First of all there were the whole line timetables. These showed times for every train at every station on a line. They were handy for working out travel time between two stations on a line. However their print was small so they were probably not as used as they could have been. And the practice of ruling a red line under your station's times to make finding it easier was not always adhered to. These were typically on two sheets, one for weekdays and one for weekends, with these not always being correctly installed next to one another.
Much easier to read were the large print station specific timetables. These showed departures from each station, not the whole line. That was good for legibility but did not give travel times. However you got a good idea of frequency at the times you were travelling at.
Exact times are particularly important here, because unlike Sydney (where most stations have a maximum 15 minute wait over wide hours) our train frequencies can be 10, 15, 20, 30, 40 or 60 minutes, with many possible on the same line within a short time. Noting that the usability and connectivity of a 10 minute service is vastly different compared to one every 40-60 minutes.
Also, despite DTP thinking is that real time is always better than printed timetables, not all stations yet have visual PIDs. Displays that exist can be too small to display trains on all lines (particularly at stations served by multiple lines such as Caulfield and Malvern). Reliability can be an issue too, with some displays being broken or partly working for months.
Even the best (and fully working) displays cater for the 'here and now' traveller. That's a different function to wall timetables that convey time and frequency information for later in the day (or on another day).
Both are needed - the former to help people on their current trip and the latter to open possibilities for future travel. DTP's stodgy anti-growth/anti-business bureaucratic culture generally resists communicating value (like the prior example with fares), selling features of a service (eg frequency) or, more radically, creating demand for something that customers did not previously know they wanted or even knew existed.
Some individuals in it may be aware but collectively the department has a poor grasp of marketing or behavioural psychology such as promoted by the marketer Rory Sutherland (who cites many transport examples including the use of multimodal transport maps - video below).
This causes DTP to put all its eggs in information systems (such as its app and journey planner) that either undersell the service or, like an old-fashioned general store, relied on customers knowing what they wanted and how to ask for it. As opposed to a supermarket approach that actually expands demand for an item through merchandising. I first identified this as an issue with PTV in 2020.
That mentality, which tends towards decline, continues unchanged today under DTP/TV. It is particularly corrosive today given that, with pressured state budgets emphasis for the next few years needs to shift to maximising the usefulness of our existing transport infrastructure as a first rather than a last resort. Asset utilisation, service levels, network reform and communication, as repeatedly mentioned here, need to become priorities rather than afterthoughts, especially as governments urge a shift to public and active transport as the fuel crisis bites.
At the time of writing this is 'MR4' with the next agreement 'MR5' currently in franchisee selection stage.
11.6 of the Passenger Experience Module (p25) talks about information to be provided at stations. Requirements include:
(i) the Master Timetable as it applies to Passenger Services of the Franchisee which
stop at that Station or at the Southern Cross Station Access Areas (as the case
may be) and each master timetable of such other Train Operators as they apply
to passenger services which stop at that Station or at the Southern Cross Station
Access Areas (as the case may be); or
(ii) if PTV agrees, a frequency timetable as it applies to Passenger Services of the
Franchisee which stop at that Station or at the Southern Cross Station Access
Areas (as the case may be) and each frequency timetable of such other Train
Operators as they apply to passenger services which stop at that Station or at the
Southern Cross Station Access Areas (as the case may be).
Presumably the Master Timetable was the line timetable discussed before. Unlike some tram stops we don't have frequency timetables but that is because no line is yet individually frequent enough to justify them. The station-specific departure times mentioned before would have been a good substitute.
It then goes on to say:
(c) For the purposes of this clause 11.6, publish means:
(i) making the information available upon request in one or more booklets or in other
similar form at all Staffed Stations and at the Southern Cross Station Access
Areas; and
(ii) displaying it on walls or information displays at Stations.
(d) The Franchisee must ensure that each Station and the Southern Cross Station Access
Areas have on display:
(i) information on fares;
(ii) a map of the metropolitan fare zones;
(iii) information on other public transport services which operate in the vicinity of the
Station or the Southern Cross Station Access Areas (as the case may be); and
(iv) information about the roles, functions and services provided by the PTO, as
reasonably requested by PTV, and contact details for the PTO.
(j) The Franchisee must display a local area map at each Station and at the Southern Cross
Station Access Areas, provided that PTV may require the Franchisee to utilise the space
in which that map is ordinarily displayed to instead display any other notice specified or
provided by PTV. The multi-modal public transport map provided by PTV titled 'Getting
Around Melbourne', or an equivalent map as notified by PTV, must be displayed at all
Stations and at the Southern Cross Station Access Areas.
11.7 of the Passenger Experience Module (p26) is about printed system information, for example timetables. This must be done by the Franchisee according to the Master Style Guide.
However there is some flexibility. Eg 11.8 refers to a Practice Note issued by PTV where standards of published information are set down. Also 11.2(c) gives flexibility for PTV to designate any medium for providing Transport Information to passengers either in addition to or as a substitute for any of the existing Transport Information Systems.
So there may have been some chopping and changing, although the overall effect is that what's at stations is not as comprehensive as 11.6 specifies. Direct information on fares, for example, has been removed. Information on other public transport services are typically not provided right at stations (though they might be outside at bus and tram stops). Also multimodal maps at stations are rare (despite PTV/TV producing them and putting them on their website).
How station passenger information should be
To maximise both current day utility and future patronage station passenger information should both tell people what they need to know know while widening travel possibilities for later.
The latter requires what you might call 'push' promotion. This is bold messaging that's unavoidably in your face when you travel the network. This is not something you have to go looking for. Nor is it something only available in small print on your sun-faded phone screen because some overpaid DTP exec (with fewer sales skills than a student casual at JB) thought they were keeping up with technology.
Minimum standards for passenger information at Metro stations could be along these lines:
1. Fares displayed. No matter what you need fare information displayed at station in poster form. And we need to go one better than we did in 2025. The information up then failed to display fares for Zone 2 only (which are cheaper than Zone 1). While probably done with good intentions in the name of simplicity, it exacerbated a genuine issue with fares (short trips are seen as too expensive) and ignored that a fair number of passengers living in Zone 2 are under some financial pressure. Having information that overstates what people making Zone 2 only trips need pay is poor on both counts.
2. Rail network map (as current)
3. Melbourne-wide multimodal frequent network map A new item showing frequent routes only
4. Local area multimode PT map These are already produced by DTP (erroneously called 'bus maps') and are on their website in a not very prominent location. The cost of having them up on stations is tiny but they would greatly assist multimodal connectivity. An example of an underused asset as DTP goes to the work of producing them but does not install them much on the network. Ideally frequent routes would be shown in bold.
5. Local area walking and cycling catchment map that would also show bus/tram stops and major attractions
6. Station precinct map Already up at some stations but more detail needed including connectivity to nearby trams and buses
7. Large print style station specific train departure timetable Such as at stations prior to 2026
8. Departure lists for buses with times, route numbers and destinations at stations As is standard practice in Perth which has a much stronger multimodal planning, marketing and passenger information culture than Melbourne as its PTA is consistently institutionally stronger than our DTP
9. Disruption advice as currently done
10. Other posters that promote value and opens possibilities for travel
11. Real time displays at station entrances and on platforms sufficient in size to list at least the next two trains of each stopping combination
Conclusion
In 2026 the amount of passenger information displayed at Metro stations has been cut. Rail network maps persist but direct fare information has gone, with the latter replaced by a QR code link to the Transport Victoria website.
Timetable information has also been removed even though all week service upgrades have so far been insufficient to justify alternatives, eg metro style frequency guides, on the vast majority of the network. The existence of timetables are more important at Melbourne stations than those elsewhere due to the wide variations in service frequency across the network, including gaps of up to 60 minutes between trains.
DTP will no doubt argue that printed timetables are less relevant due to the widespread ownership of mobile phones and more information displays at stations. However the roll-out extent, adequacy and reliability of the latter can vary. Also, due to the lack of a will and culture to grow patronage, the department is not as cognizant as it should be in the need to provide information not just for the trip at hand but also in a manner that fully reflects the network's usefulness and value for money.
The Metro Tunnel has been in full service for nearly 6 weeks. Handily for data wonks, the 'Big Switch' happened on 1 February, giving us a full month's of performance data in the Track Record that has just come out.
That data describes the delivery and punctuality of the 18000 train services scheduled to run that month. It affects payments to the operator, and, if performance dips below a certain level, compensation to passengers.
Expectations
Contrary to expectations built up for the better part of a decade (most recently the premier's new timetable in place everywhere comment), the non Metro Tunnel lines didn't get much out of the February 1 'Big Switch' timetable.
The Craigieburn and Upfield line timetables were particularly underwhelming given that the Metro Tunnel's business case proposed an upgrade from every 20 to every 10 min off-peak, similar to what the Frankston line has had for over a decade. Craigieburn got a handful of extra trips while Upfield's was a 'rob Peter to pay Paul' job with trips shuffled and early evening frequency cut.
Without much in the new timetables to sell, the government switched to spruiking other potential benefits, such as less platform crowding and improved reliability. These would come about due to the Craigieburn and Upfield lines no longer having to share its City Loop portal and platforms with the Sunbury line, eliminating knock-on effects if one line gets delayed. In theory anyway.
Especially note this comments: "Craigieburn and Upfield lines will be getting a reliability boost with more space and less crowding in the city". We'll test that in a moment.
What about other lines?
Werribee, Williamstown and Frankston got new timetables with through-running removed and Frankston re-entering the City Loop. That may improve reliability numbers (as cross-city operation was always fairly fragile) though the bogey of loop bypasses is back for Frankston line passengers.
Sunbury, Pakenham and Cranbourne lines of course use the Metro Tunnel instead of the City Loop. The Sunbury line is no longer subject to delays in the City Loop from late Craigieburn and Upfield line trains. That's significant as the Craigieburn line is crowded while Upfield suffers from a single track on parts. On the other hand the through-running may make it susceptible to issues on the other side of town, with this equally so for Pakenham and Cranbourne.
Notwithstanding this, the higher frequency and upgraded signalling now in effect on the central part of the Metro Tunnel line should allow faster recovery from delays. The effectiveness of this should be discernible from performance data that we'll look at in a moment.
Those analysing punctuality should also consider dwell times at the new stations. Too short dwell times mean poor recorded punctuality while excessive dwell times is both annoying and inflates recorded performance. New systems often have a settling-in period as passengers get used to new travel patterns, with running and dwell time reviews highly desirable later.
The Mernda, Hurstbridge, Belgrave, Lilydale, Alamein, Glen Waverley and Sandringham lines got no significant timetable change in February. Thus no performance change attributable to the Metro Tunnel can be expected.
What happened? (the big picture)
Your choice of baseline is critical when making comparisons.
Especially as February is not a typical month. It's busier than January but, with universities not fully started, it's quieter than March. Lower loadings is typically good for train punctuality. But hot weather, which February has a lot of, is not.
To account for these variations I compared Metro train service delivery and punctuality for February 2026 against the 12 month average for 2025. I also looked at the last five years' worth of February performance numbers to better compare like months. Numbers are reported here for those interested (use the PowerBI app).
First of all the big picture.
Network wide service delivery in February 2026 was the same as the 2025 calendar year average at 98.9%. But on time performance was slightly better at 94.3% versus 94.1% network average.
Both service delivery and on time performance was at least as good as (and more often better) than that recorded for all Februarys going back to 2022 if not further.
Big new train timetables don't always go to plan but this one has done pretty well on a macro scale. Metro managers won't be unhappy with these numbers. That's important right now given that MTR is one of the three shortlisted franchise groups for MR5.
What happened? (performance by line)
Overall network results can mask big variations in performance trend by line. In this case actual results deviated from expectations, including some raised by government. I'll present more detail later, but for now below is a snapshot of where performance improved and where it didn't.
To summarise the Clifton Hill and Burnley groups (not listed as they didn't get a timetable change) maintained relatively good performance. And delivery on the old cross-city group (ie Frankston/Werribee/Williamstown) improved.
Both contributed to the strong network average cited before, offsetting poor delivery results on the Craigieburn and Upfield lines. Performance of lines now serving the Metro Tunnel was close to their past averages except for Sunbury's punctuality whose strong rebound contributed to a better than average network result while delivery was not so good.
The main movers are summarised below (bar lengths relative only).
Craigieburn and Upfield
These lines should have got a performance lift. All Sunbury line trains were removed from a shared loop portal and hardly any new services were added in their place.
The state government said reliability would improve with these lines specifically mentioned.
Only thing is that it didn't.
Instead service delivery fell relative to both the previous year's average and the four previous Februarys on both lines. Craigieburn's delivery was so poor that if the whole Metro network behaved like that then compensation would be paid due to it falling below the 98.5% threshold.
February's punctuality on both lines (see below) also dropped below the 2025 average, although it was still better than February of that year. Again Craigieburn's was substandard, with punctuality consistently below the 92% network compensation trigger. On the numbers Craigieburn line passengers should be getting compensation for late running but do not because the government drew up the franchise agreement to only pay when the whole network average is poor.
February 2026 was especially poor for Craigieburn relative to other lines. Its 88.7% made it Metro's most delayed line by a long shot. The next tardiest were Stony Point (91.3% on time) followed by Belgrave and Frankston (92%). Thus Craigieburn passengers are about twice as likely to experience delayed trains in February 2026 given the network average was 94.3% on time.
Unlike Craigieburn, the Upfield line's punctuality last month (93.9%) was only slightly below the network average. But there's a catch as you'll see later.
People naively expecting Upfield line trains to continue much past Coburg were especially short-changed last month, in several senses of the word. More than 3% of trains did not go the full distance, with the rate of trains terminating short between 3 and 6 times previous February averages. While short trips get reported there appears to be no separate passenger compensation trigger for them, unlike for delivery and on-time running.
To summarise, February 2026 was a terrible month for the Craigieburn and Upfield lines. Not only did they miss out on significantly upgraded timetables but promised reliability improvements failed to materialise with service delivery getting worse, not better.
If I was in government I'd be worriedly monitoring future months' performance closely given growth pressures in the outer north and marginal seats like Pascoe Vale closer in. Unless it can turn things around quickly, this timetable risks being a net negative for Craigieburn and Upfield passengers due to reduced delivery (both lines), chronic delays (Craigieburn) and early terminations (Upfield).
Sunbury/Pakenham/Cranbourne
It was hard to forecast how these lines would go, with these now running through the new Metro Tunnel.
Sunbury line's numbers were the most interesting. February's service delivery deteriorated to 98.5% versus a 99% average for 2025. Expressed in another way that's a 50% increase in the proportion of trains cancelled, or an avoidance of compensation by the narrowest margin if replicated network wide. On the plus side the improved Sunbury line frequency means that except for major disruptions the waiting time to the next train is likely to be halved if services do get cancelled.
Out of all the numbers surveyed, the largest statistical gain was undoubtedly Sunbury's on-time performance. This was much higher under the new timetable, exceeding anything in recent previous years and the 92% compensation trigger (if replicated network wide). While the February 2026 result appears creditable the real test is what happens after usage patterns have stabilised and (allegedly excessive) dwell times have been reviewed.
Pakenham and Cranbourne line statistics were less interesting than Sunbury's. Both delivery and punctuality in February 2026 were slightly below the 2025 annual average. Delivery on both is similar to previous Februarys while punctuality is better (Pakenham's numbers below, Cranbourne's not a lot different).
Frankston/Werribee/Williamstown
I've grouped these due to their past association in the cross-city group up to January 31, 2026. All lines have had works related altered operating patterns and/or significant periods with replacement buses.
All three lines had a good story to tell regarding service delivery with this higher in February 2026 than either the yearly average or recent Februarys. Maybe splitting operations at Flinders Street helped, although it is also true that frequent and reliable through-city running lines make trains useful to a wider variety of trips as Metro Tunnel users have now experienced.
As for punctuality, Werribee's February 2026 performance was similar to the 2025 average while Frankston and Williamtown's is about 0.5 to 1% worse. On a percentage basis Williamstown was similar to the network average (94.3% punctual) while Frankston scored near the bottom of the pack with 92%. Werribee's numbers are below.
Other Metro lines
Lines that did not get a timetable change performed relatively well in February 2026. This helped offset problems with the northern and Metro Tunnel groups, pushing the average up.
V/Line
1 February included changes V/Line's timetables, including extra trips on the Seymour line. 96.7% of scheduled V/Line services were delivered in February. This is inferior to the 2025 annual average (97.2%) but better than the very poor January 2026 result (91.1%). Punctuality had a similar pattern.
Seymour, which had extra services scheduled, fared a bit worse last month with 95.5% of trains running in February 2026 versus 97.8% for 2025 and 93.2% in January 2026. Seymour's punctuality was 85.5% last month, the lowest number since April 2024.
The publicly-operated V/Line's operational performance has been sub-par for years, with the franchised electrified Metro network typically cancelling or delaying a far lower proportion of services. This government's record of avoiding electrification on lines that could benefit, including Wyndham Vale, Melton and Wallan, has increased the number of people using less reliable V/Line frequently as suburbs continue to expand beyond the electrified network. With the last major V/Line to Metro electrification being way back in 2012 (when Melbourne had one million fewer people), expect outer suburban rail electrification to feature prominently in this year's state election campaign.
Conclusions
February 2026's stable delivery and punctuality is mostly attributable to good performance on the lines that did not get a timetable upgrade in that month, ie the Burnley and Clifton Hill group. The Werribee, Williamstown and Frankston lines also played their part though the cross-city group being broken.
Did the Metro Tunnel and its timetable contribute to trains running reliably and on time in February 2026? The short answer is no. Except possibly for Sunbury's improved punctuality, the Metro Tunnel and the accompanying timetables had less benefit than hoped. This can be demonstrated by the sharply reduced performance of Craigieburn and Upfield and the generally middling performance of the Metro Tunnel lines.
March's numbers, available around April 10, will also be of interest to see whether February's trends are sustained or reversed. The weather is cooler but March sees more passengers with universities resuming, a packed major events calendar and concerns over access to the MCG from the south-east.
Key matters arising from these numbers for rail operations and service include (i) restoring service delivery and reliability on the Craigieburn and Upfield lines, (ii) examining run and dwell times on the Metro Tunnel lines and (iii) creating a reliable cross-city group when Werribee, Williamstown and Sandringham are linked later in the year.
After an early release, deletion and then re-release, details for the new Lara - Avalon Airport bus are now available. The new Route 18 will replace the existing very short Route 11 (which didn't serve the airport). It is intended to provide access to jobs and a basic connection for airport passengers from the nearest station at Lara on a standard Myki fare. Previously Avalon Airport had no public transport except for private shuttles too expensive for regular commuter use.
The new service will start this Sunday, March 15 with the premier's media release here. Its introduction follows a local campaign for worker transport to airport precinct jobs.
Demands for better transport to Avalon go back a while. Most famously then opposition leader and later premier Ted Baillieu promised an Avalon rail link in 2010. This was a rash promise that, along with lines to Doncaster and Rowville, cemented that government's reputation for promising big but building little in transport.
A response proportionate to Avalon's usage would have involved a simple bus route from the nearest station. After waiting 15 years we're finally getting that with Route 18 in a few days.
Trust central to airport transport success
Travel to an airport is very different to a casual outing to the shops a few kilometres away. Whether it's bus or train, just one thing matters when it comes to airport transport for passengers. Trust.
That is trust that transport will show up so you don't miss your flight. And, when landing, you can trust that transport is available even if your flight gets delayed.
If people don't trust one form of airport transport then they will pay a premium for something that they do since: (a) the consequences of unreliable transport are very high, (b) most people spending their own money only fly occasionally so a bit extra can be shrugged off and (c) most frequent flyers either have their expenses paid for by others or are big spenders. And second only to reliability the ease of buying a ticket is more important than its absolute cost.
Where the bus dumps you at nowhere significant (sorry Lara residents!) and relies on a train connection for major destinations like Geelong, Tarneit, Footscray and Melbourne, another layer of trust is required. That means (i) reliable and enforced timed connections with trains, (ii) the trains themselves to be reliable and (iii) acceptable connectivity even when replacement buses are operating.
If Route 18's timetable does not have trust at the centre of its design then the new service will fail, ie it will be more symbolic than useful, especially for airport users. I'll analyse this later.
Route 18's timetable
Transport Victoria's news item says that Route 18 will run 21 trips on weekdays and 20 trips on weekends. That counts both directions so it's about half that each way. They claim a service span of early morning to 'around midnight', though the latter is only true for trips from Avalon; trips from Lara finish around 10pm. However that's still a wider span than most Geelong area bus routes, especially on weekends. Travel from Lara Station to Avalon will take approximately 12 to 14 minutes.
If we're talking buses in Melbourne and the major regional cities, the general expectation is something that runs every 30 to 60 minutes and hopefully 7 days. Some routes are more frequent while others are less.
Route 18's timetable cannot be easily summarised. As mentioned before its operating hours are longer than average but its frequency is highly variable. So much so that intervals between services can be 4 hours or more as shown below.
Something you might have noticed from the timetable above is the variability in origins. All trips from Lara operate to the Greater Avalon Employment Precinct, with most continuing to Avalon Airport. In contrast about 1 in 2 trips to Lara skip the Greater Avalon Employment Precinct according to the timetable. The route map (below) does not show this variation at all.
This isn't so much of a problem for those wishing to leave the Greater Avalon Employment precinct around 2:20pm on a weekday. This is because they can board a trip to the airport, wait a few minutes (presumably on the same bus) and shortly be off to Lara.
As seen below only 5 out of 12 weekday trips that depart Avalon (either the airport or the employment precinct) are formed by inbound trips arriving from Lara.
Where do these trips come from? Possibly from dead runs from some other routes that terminate at Lara, eg 10 and/or 12.
My understanding is that many bus operator contracts are structured so that operators are paid by in-service hours and kilometres. This incentivises operators to devise schedules that minimise dead running as this increases their margin. However this may also reduce the incentive for DTP to maximise passenger benefits as converting an empty trip to a live trip (desirable to improve frequency) may involve additional payment for kilometres that would be operated anyway.
* Route 18 will connect with trains at Lara Station, mainly for services to and from Geelong, making it easier for people working at GAEP to travel by public transport.
* Route 18 will connect with most GAEP shift times. Buses will usually arrive and depart about 20 minutes before and after shifts.
* Route 18 will connect with most flights at Avalon Airport. Buses will arrive 75 - 120 minutes before departing flights and 30 minutes after flights arrive.
* There will be stops at GAEP on Canberra Drive and at Avalon Airport near the terminal entrance making it easy to access these places.
Train connections
The above says that access to and from Geelong is a priority. There's big populations at Wyndham Vale and Tarneit on the Geelong line but train connections for them are a secondary priority. However with trains every 20 minutes most of the day hopefully waits for connections won't be too long even in the Melbourne direction.
What about train to bus connections at Lara from Geelong that Transport Victoria says have been optimised? A 10 minute connection time is generally allowed. On weekday mornings where it can be less people have a practical option of taking an earlier train which those getting flights would do.
However connections to the first bus on both Saturday and Sunday are a tight 5 minutes. The late start of weekend V/Line trains (unlike Metro which runs all night) means that there is no alternative like catching an earlier train. In my view this is too tight, especially given V/Line's fairly ordinary punctuality and cancellation record (which, despite some public perceptions, is inferior to Metro's).
Are connections enforced?
Transport Victoria do not mention whether Route 18 trips are held either for late arriving planes or V/Line trains.
When service is frequent, like Skybus from Tullamarine, you would depart buses regardless (with the possible exception of the last trip for the day). That's because it's never long until the next bus and holding buses hinders more people than it helps.
But when bus service is both infrequent and intended to connect then you might have a holding policy to guarantee connections for all but the longest delays. An example is where hourly Night Network buses wait up to 20 minutes for late arriving trains (also hourly).
In Route 18's case you might consider holding services (i) on the first trips that are meant to meet the first trains and (ii) where planes are delayed and there is a long gap to the next bus. However such decisions may be dependent on effects on other routes that may be interlined with Route 18.
Will people use and trust the 18 bus?
Workers might. But I'm not sure about airport users for whom trust is everything. Trust requires a timetable that does not have such huge gaps or at least some form of holding policy to ensure connections.
Route 18 is a cut price attempt to provide an industrial style route for Avalon precinct workers with the airport passenger transit role a happy by-product. Its resourcing works out at maybe half a bus per day with the rest being either idle time, dead running or operating another route.
In contrast, had the 18 been allocated a dedicated bus then it might have been possible to operate a 30 min peak / 40 min off-peak service meeting every second train. Still basic but it has resilience for late planes that the current timetable lacks. While some trips may be very quiet, the trust such a regular service engenders may make its trips operated at other times much busier. Such a service also provides a 'safety net' for workers with part-time shifts or who need to go home in an emergency.
A consistent frequency on Route 18 would provide a level of trust (and therefore patronage) that the proposed skeleton service would not. Until the former is provided my tip is that the latter will struggle to attract (especially) airport passenger patronage, especially if no means are provided for passengers to purchase myki tickets (which they can at Tullamarine where there is a machine, albeit not at the expected location in the bus interchange).
