How will the Metro Tunnel benefit your line?

The Metro Tunnel, our most anticipated suburban rail project since the City Loop, is due to commence service next year. If amply scheduled its effect could be transformative; not just for travel on the lines linked by the tunnel, but for benefits cascaded across other lines due to freed up capacity and overdue frequency upgrades. 

Except for assurances of 'more trains across Melbourne', the public doesn't yet know the services they will be getting. While you can appreciate sensitivities over peak service levels and stopping patterns, even broad specifications like maximum waits at various times of day have yet to be released. Sydney, in contrast, tends to be more open about this sort of stuff.

The nearest we have (which may or may not reflect current thinking) is the Day 1 service plan in the 2016 Business Case, with the important table 2-2 reproduced below.

 

Off-peak was really good in some respects (6 trains per hour for Sandringham, Craigieburn and part of Upfield) but niggardly in others (eg 3 trains per hour service beyond West Footscray). The table doesn't list six lines (Mernda, Hurstbridge, Belgrave, Lilydale, Alamein and Glen Waverley), so presumably there was then no intent to revisit their timetables (30 or 40 min maximum waits notwithstanding). Or it assumed there'd have been faster progress on the 2012 Network Development Plan than there was.  

While Melbourne doesn't have a great record of backing infrastructure with service, short-changing Sunshine, a station with comparable patronage to Dandenong, with half a service (ie 20 min gaps) seems implausible. Even for us. Also pandemic-driven working patterns have lowered the commuter peaks, especially on Mondays and Fridays. And Airport rail (aka SRL Airport for a while) was off, on, then now apparently off. Anticipation with regards to the Western Rail Plan and Geelong Fast Rail has also been on and off. 

Thus the Business Case specification represents guidance from a different time. The timetable ultimately adopted may well be different to that envisaged 8 years ago. Although given that it is part of the pitch to justify public funding one would hope that any revised specification would have similar if not better distributed community benefits. I speculated on the Metro Tunnel timetable a few days ago, giving three options of what frequency might be like at various times of the day.

My emphasis then was on off-peak service and maximum waits. Such improvements could have been done on most of the existing network without the Metro Tunnel. Victoria's emphasis on infrastructure  before (or even without) service artificially extended waits for the latter, as opposed to delivering service when the capacity existed to do so (the NDP - Metropolitan Rail approach). 

My emphasis today is what the Metro Tunnel enables with regards to peak service. Key benefits include (a) new rail access at Arden, Parkville and Anzac, (b) rapid transit under the Swanston St corridor that enables cascading tram upgrades, (c) improved core network robustness and (d) support for a growing CBD by adding capacity (likely the project's main justification). 

Benefits by line

The Big Build website lists claimed benefits per line.  I will discuss peak service levels in more details here.

As with my notional off-peak service levels there will be a bit of speculation without there being published up to date service specifications to go on, though I've tried to stick to published material as much as possible.

I will use current (2024) timetables as references rather than older ones that might have been used as references in earlier material. I think that is fair since today's timetables incorporate reforms (eg the 2021 Caulfield and cross-city changes) that get more out of our existing pre-Metro Tunnel network.

Also, for maximum rigour (as opposed to attributing or inflating claimed benefits to bolster a case for a project) analysis of what benefits new infrastructure will provide should be based on optimally using existing infrastructure (ie a higher base) than an existing perhaps less efficient use of what we have (ie a lower more favourable base).  

Craigieburn and Upfield

Considered together due to their sharing of the City Loop. 

The reference to the project 'creating room for 54 000 passengers per week' for the Craigieburn line implies a focus on peak capacity. That's 10 800 per weekday which they put at 27% more capacity. This gain would come from Sunbury line trains being removed from the City Loop. 

Upfield would gain a 71% increase in capacity, or 9 000 passengers per weekday (45 000 per week). Like with Craigieburn this gain would come Sunbury line trains being removed from the City Loop. 

The current (2024) timetable records 18 Sunbury line trains arriving at Flinders St between 7:00am and 8:59am. Similar arrival numbers for Craigieburn and Upfield are 17 and 7 respectively in two hours. 

Adding 27% to Craigieburn's 17 arrivals gives 22 trains. Adding 71% to Upfield's 7 gives 12 trains. That's adding 10 arrivals in those two hours. That allows room to grow since fewer are added than the 18 Sunbury line trains removed. Totalling this gives 34 trains, or less than two minutes between trains. There are however capacity constraints elsewhere on the Upfield line as this video highlights so we may not get all of the promised increase.

Craigieburn should see an average of 5.4 minutes between trains in its busiest two hours, with the Upfield line having 10 minute headways. There will be some unevenness due to the Craigieburn line needing to slot in V/Line services and leave gaps for the Upfield services in the City Loop. But overall it should be a solid turn-up-and-go service during the peaks except (probably) for the last few stations of the Upfield line.  

Table 2-2 in the 2016 Business Case's service plan gives higher capacity numbers. In the am peak two hours it has 28 trains for Craigieburn and 12 trains for Upfield, making a total of 40 for the 'Northern Loop'. In both cases not all trains would necessarily start at their outer termini. 

As for the people count, does one passenger count twice if they take the train home? It would seem so. As the number of people who can fit on a train is nearer to 1000 than 2000, if we're adding 10 arrivals across both lines then 10 000 makes more sense than 20-odd thousand. 

Sunbury and Cranbourne/Pakenham

Considered together due to their connection through the Metro Tunnel. 

I mentioned before the current 18 Sunbury line trains. The Big Build website claims a rise of 60% in capacity, or 113 000 more passengers per week (22 600 per day). Such an increase could mean 29 trains, or a shade over 2 minutes between trains. But that doesn't factor in the higher capacity of the HCMTs which will increase people throughput even if the number of trains remains unchanged. However you calculate it it will be a good turn-up-and-go peak service for the whole line, even if some trips start at Watergardens.  

The Table 2-2 tally is 31 Metro services, including 7 for the West Footscray turnback. Presumably 6 of those 7 are reserved for extension to Melbourne Airport given Melbourne Airport Rail's projected 10 minute frequency. Unless priorities are shuffled and Melton or Wyndham Vale electrification happens first. 

Currently the peak am two hours have 25 arrivals from the Cranbourne/Pakenham direction. A 45% increase is stated for Cranbourne/Pakenham. This is 121 000 passengers per week, or 24 200 per weekday. A 45% increase on 25 arrivals could mean 11 trains added though one must note that the new HCMTs in service have already significantly increased peak capacity. 

Table 2-2 in the Business Case service plan has 35 Metro trips and 4 V/Line trips in the am 2 hours. 

Sandringham and Werribee/Williamstown

The Sandringham line currently has 15 arrivals at Flinders St in the peak morning two hours. The Big Build website's 48% increase means 22 arrivals, or an increase of 7 trips. Either of these represents a pretty good peak service (ie every 8 vs every 5.4 minutes). This would create room for 72 000 passengers per week or 14 400 per weekday. This video queries whether that 48% is correct given terminating capacity constraints at Sandringham. 

This is similar to the Business Case plan which shows 21 arrivals (excluding South Yarra turnbacks). This also shows the Sandringham line replacing Frankston as the eastern portion of the cross-city group. The scheduling and operations of this will be critical to deliver fast cross-city travel that those used to the Metro Tunnel will expect on this group too. Although the cross-city group was meant to offer this in theory, actual reliability for what should be simple fast trips like South Yarra to North Melbourne is currently poor with trains often terminating, dwelling too long, turning back or transposing at Flinders Street.  

There are 24 arrivals from the Werribee side in the 7 - 8:59am slot at Flinders Street. This comprises 12 of these are from Werribee, 6 from Laverton and 6 from Williamstown. In other words current frequencies are 10, 20 and 20 minutes respectively. 

The Business Case's 24% increase would bring that up to 30, while Table 2-2 has 36 trips in the two hours. In any event I would expect most if not all of the increase would be starters from Werribee since that contains the group's busiest stations and has the highest growth prospects. 

Frankston line

The Frankston line currently has 23 arrivals in the am peak two hours at Flinders St. A significant proportion originate at Carrum. A 15% rise on that means 26 or 27 arrivals. 

The Frankston line enjoys a lot of service but loadings are less than other lines, especially post pandemic. Reasons could include factors like: a. demographics (a mix of white collar CBD workers with WFH options and local workers who largely drive is less conducive to high patronage), b. endless line shutdowns and bus replacements eroding rail as an option, c. the Mordialloc Freeway increasing driving's attractiveness, d. Often poor or unreliable feeder buses, exacerbated by b above. Frankston also typically enjoys double the frequency of lines with more favourable catchment, especially Craigieburn but also Mernda due to serving historically marginal electoral seats.   

Still this is not a reason to write off Frankston line usage post Metro Tunnel. Removing Cranbourne and Pakenham trains from South Yarra, Richmond, Parliament, Flagstaff and Southern Cross will likely mean more changing at Caulfield station to the Frankston line.

It won't all be one way though as some Frankston passengers for Anzac, Parkville and possibly other stations may switch to ex-Dandenong trains to save about 15 minutes. Although those who value the seat they probably have will likely stay on until Flinders Street, especially given Caulfield's unfitness as an efficient high capacity interchange station. 

Sandringham passengers will likely switch to Frankston trains at South Yarra to access the City Loop. There may also be some changing to/from the Burnley group at Richmond, depending on travel directions.   

Overall these changes should mean that the Frankston line in from Caulfield has high passenger growth while its outer half grows by much less, at least while intermittent shutdowns continue. For the next few years optimising the timetable against usage might mean that more peak trains start (and finish) at locations like Mordialloc, Moorabbin or even Caulfield rather than Carrum or Frankston. The attractiveness of scheduling such a more intensive inner service could increase given that Frankston will become Melbourne's only line that gets a City Loop portal all to itself.  

Mernda/Hurstbridge and Belgrave/Lilydale/Alamein/Glen Waverley lines

These do not gain services from the Metro Tunnel or the resultant cascading of capacity. However the project benefits page cites 5 to 10 minute travel time savings for trips to new station precincts such as around Anzac and Parkville (with major health and education destinations).

Outside that though the Metro Tunnel project's benefits for the remaining 6 out of 15 train lines are slim. Though this needn't be the case if a network rather than a project-based approach to service is taken. 

As mentioned last week the Metro Tunnel risks becoming a project of division rather than (deserved) pride if it creates a two class 'haves and have nots' metropolitan rail network. In this case the haves could get all day turn-up-and-go 10 minute service while the have nots remain saddled with widespread 30 to 40 minute waits, even at busy stations like Box Hill. Especially for people changing from the Metro Tunnel to one of the less served lines the customer experience of potential unaddressed 30-40 minute waits risks lowering the entire network's standing. 

How do we overcome this without spending much?  

The answer is the selective pursuit of something almost unspeakable in Melbourne public transport policy, that is increased service frequency. Instead of something that governments only grudgingly add when overflowing train loads threaten to roast them at the ballot box, frequency needs to be reimagined as a necessary feature of a connected and more useful network. This has the potential to transform rail travel, especially if accompanied by a vastly simpler weekday timetable on the Burnley group. 

As mentioned in More Frequency More Go, the cheapest and best way to do it is to start by eliminating 40 minute waits (~0.2% more trips per week needed) and then 30 minute waits (~5% more trips per week needed with more manageable smaller steps for a staged program possible). Again, as with Frankston, it may be possible to reduce implementation costs with 'swings and roundabout' changes that have some small minuses for a few stations but generate overwhelming gains for the majority. 

Together such initiatives halve maximum waits across the network from 40 to 20 minutes 7 days per week from early morning to late at night. This is a major usability gain for not much money. A further roll-out to 10 minute maximum waits would have further benefits with weekday interpeaks for Ringwood and at least shoulder peaks on lines like Mernda and Craigieburn (if not otherwise upgraded) being priorities. 

The Metro Tunnel project's benefits could be extended if its timetables, especially in the peaks are carefully examined. The aim here could be to see if there is scope to spread service hours more widely across the network, potentially almost doubling the lines that would directly benefit with more services. There may be cases were cutting out one or two peak trips and shuffling other trips may result in an almost unnoticed frequency drop. Yet shifting those service hours to off-peaks to cut maximum waits to 30 or preferably 20 minutes during a key time period would have a large and very noticeable gain.  

Conclusion

The Metro Tunnel promises to be transformative for our train and tram networks. Key gains include new stations, reduced travel times, connectivity across a growing city and freed-up capacity with benefits extending to the metropolitan fringe.   

The main uncertainty at this stage is the service levels of the Metro Tunnel-associated timetables. Above anything else this indicates how serious Melbourne is about becoming a city with public transport useful for diverse trips. And they determine whether the Metro Tunnel delivers its full potential or not. 

For example, will the improvements be concentrated in the peak or be spread across the week as modern travel trends demand? Will it be only Metro-Tunnel related lines that gain, creating a two tier haves vs have nots network? Or will benefits be more widely distributed, with the network's 30 - 40 minute maximum waits, stubborn for decades, finally slashed? 

There's one thing for sure, we'll be in for an interesting ride as we find out. And I intend to be there for the journey.