Some thoughts on PTV’s rail wishlist, umm I mean Network Development Plan

I had been writing a blog post about proposed rail lines, and even went to the trouble of drawing a map of what was known about the various proposals floating around. Yesterday a very detailed PTV Network Development Plan for the rail network was released with lots of much prettier and more comprehensive maps.

The portents of its release have been there. In the past few weeks, several studies into proposed rail lines have been released: Doncaster, Rowville, and the Airport. The PTV document incorporates these, and lays out how they would build them, and a few more besides.

PTV rail network: Stage 4

What some people hadn’t registered until now is that the wish list now includes not one rail tunnel, but two.

The “metro” (north-south) rail tunnel has been proposed for a few years now — that’s the one leading from the Sunbury line at South Kensington via Parkville, the city to Domain, then to South Yarra to connect with the Dandenong line. The Rowville and Airport studies both conclude that their new lines rely on this for capacity through the CBD.

The Doncaster study had flagged the theory that a second tunnel — re-routing the South Morang line from Clifton Hill via Fitzroy and Parkville to Flagstaff — is necessary to provide enough capacity for the Doncaster line to run into the city via Jolimont. The PTV document says this should then extend to Southern Cross and eventually to the new suburb at Fishermens Bend.

What’s good

Wishlist maps for the Melbourne rail network are a dime-a-dozen. Every gunzel has drawn one. But it’s rare to see something official, and PTV are to be congratulated for publicly putting out the Plan.

Forward planning is essential. Vicroads do it all the time, and put their proposed motorways into the Melway. It’s been lacking in public transport, leading to debacles like the Footscray pedestrian bridge being partly demolished only just after it had been built.

And there’s a lot to like in the document. It explains how each project will build on the overall capacity and reliability of the network. The overall strategy is a good one — to move towards self-contained lines, with a minimum of junctions and interactions with other lines that limit train throughput.

Even stage 1, the 2016 service plan (basically “how do we run the trains once Regional Rail Link is built”), includes some great outcomes for passengers, including seven day services every 10 minutes from the city to Newport, Dandenong, Ringwood, Glen Waverley, Sandringham, Sydenham and Craigieburn (joining Frankston, which already has it).

Governments might baulk at the cost of some of the later more expensive upgrades, but those initial service upgrades should be a priority for funding. They bring great benefits for little cost.

PTV rail network: Frequency upgrades

The Plan also includes things you don’t normally hear about, like modern in-cab signalling systems, which allow up to 50% more trains to operate on a line, for a relatively low outlay.

And while it only covers Rail (apparently the Bus and Tram plans will be out later), it does have a section talking about connections between modes, including timetable coordination.

What’s bad

There are heaps of good ideas in the Plan. If I have anything bad to say about it, it’s that the running theme with both it and the Doncaster/Rowville/Airport studies appears to be along the lines of: there won’t be enough central core capacity in future year X, therefore we can’t build any new rail lines until the metro and/or Fitzroy rail tunnel have been built first.

It’s a kind of innate conservatism in the planning: that new lines have to be built with all possible future growth catered for, and if we can’t do that, we shouldn’t build them at all.

This is quite unlike the road lobby, who will happily build virtually anything, knowing that if later growth blows the capacity, it will just provide the political impetus for the next big project (and in the mean time it will still provide some “benefit”).

For instance, the Eastern Freeway was built knowing full well it would terminate at Hoddle Street, creating a snarl there.

The risk of wanting the multi-billion dollar tunnel before anything else can happen is that if funding for that is delayed, everything else is delayed.

In reality, there is scope for building lines and extensions now. Extending South Morang to Mernda, for instance, should be a no-brainer. Dipping a toe in the water towards Doncaster, by building the first, easy bit to Bulleen, would get scores of buses and their passengers out of inner-city traffic. The line to the city would cope for a good few years yet, especially if high capacity signalling was included.

The Plan has the Altona Loop duplication waiting until stage 4. It could easily be done earlier. Even much of the signalling won’t get upgraded until after the metro rail tunnel is built — when the signalling would deliver similar benefits in terms of capacity, but years earlier and at a fraction of the cost.

The Road Lobby knows all about salami tactics. The Rail bureaucracy need to learn the same strategies.

The way forward

Criticisms aside, it’s great this plan is out. If the relatively cheap upgrades that are part of the 2016 stage haven’t yet been fully funded, the government should show it’s serious about the rail network, and fund them pronto. (Yes, Southland station is included in that.)

It’d be nice to see signalling upgrades across the most congested parts of the network in the short term. It’s cheaper than building tunnels, and although there’s some complexity in ensuring all trains on those sections have the right equipment, the capacity benefit of up to 50% is obviously beneficial.

Whatever the precise order of implementation, the government (and politicians on both sides) need to start pushing these projects. In particular, the current government would do well to remember that they were voted in on the back of public transport issues — not the proposed east-west road which was barely mentioned during the election but somehow has morphed into their top infrastructure priority.

As for plans for the rest of our multi-modal network, while the tram and bus Plans haven’t yet been released, let’s hope they’re not too far away… and that they’ve been prepared in tandem with the rail Plan!

More reaction:

By Daniel Bowen

Transport blogger / campaigner and spokesperson for the Public Transport Users Association / professional geek.
Bunurong land, Melbourne, Australia.
Opinions on this blog are all mine.

27 replies on “Some thoughts on PTV’s rail wishlist, umm I mean Network Development Plan”

Would the Fishermans Bend Station be overground? Looks like the whole area is currently empty and it should be easy to reserve a 22m wide corridor for a new railway to be built there.

An underground FB station would obviously be hugely expensive.

Lots of other interesting titbits – Upfield line extended to Wallan through Roxborough Park (and Seymour services rerouted via Upfield). The Geelong line services extended through to South Yarra (which would require terminating facilities there), and considered part of the Metro (electrificiation?). Frankston line through routed to Craigieburn via the city loop – very interesting as it would require the Caulfield and Northern Loops to be used as conventional unidirectional tunnels (note there are only three services through the loop) with significant changes to the current access ramps.

And what about the troubled Gippsland line and the possibility of rail return to Leongatha, maybe beyond? As usual, this part of Vic is not mentioned – so more and more, larger and larger trucks, on less than adequate roads, will continue to be the norm in this area

It’s excellent to see that service provision, especially interpeak and evening frequencies, is a big priority and mentioned throughout the report. Hopefully this means that high quality of service, not just during peak hour, is now a prominent goal for PTV.

Certainly cancelling the East West Link and putting money towards this would have a much better impact on traffic.

I agree, they need to just start saying “this is the plan” rather than putting out endless wishlists. At least it is a vision, something which has been missing for years.

As for the Eastern Freeway/Hoddle St thing — according to my 1984 street directory, there was supposed to be a freeway from there to the north, linking up with the Hume Freeway (presumably heading south along Hoddle/Punt as well).

Personally I think the plan is horrendous. I can’t see why we new to spend billions on creating another tunnel when we already have ample options to get across the city. Just like I can’t see why we need to come virtually anywhere in Melbourne and have to change to another train to get to flinders street. It’s melbourne biggest station, ad this the busiest. The beauty of the way things are now is that every train goes there. Things really aren’t that bad at the moment, I travel daily and I’m not 100% sure what everyone’s complaining about?? In cab signaling, now that another matter. I don’t understand it fully, but as i believe, the train is effectively driven by computer. So if that’s the case, aren’t we putting people (drivers) out of work? I think governments were supposed to be making job? We’ve seen nothing but people losing ther jobs lately, I’m sure this is a great way to to PTV, really moving forward. In regards to the Dandenong line, how do you suppose we an double metro services? Even if there was a train running from Dandenong to the city every 5 minutes (which is unnecessary) how do you slot freight and v/line trains in there? Sometimes it can take a freight train 5 minutes to pass a point! And Vline timetables have already blown out to ridiculous lengths. Seriously hoping they price themselves out of this “plan” ever happening. So what if it benefits someone going from sunshine to south yarra? They can do that now!

Why is Doncaster and Rowville so pressing anyway?

if these are the only two major gaps in rail coverage where buses are being used instead then Melbourne is still doing prettt well as a whole.

Its not like Vancouver or Perth has the density of rail lines Melbourne does.

Why not just provide proper BRT right now? Sorta like how Brisbane has the BUZ network filling in gaps between the heavy rail lines.

My view is that while the plan does have positive attributes, it also has two
major flaws-

1. It ignores the tier concept and assumes that a commuter network will work
instead of splitting into Metro and Interurban.

2. It’s all about ‘this is what we’ll do’ rather than ‘this is how we came to
the conclusion that…’. They haven’t considered any other options.

@Jacob, good question. If the reservation can be grabbed before the area is built-up, then it’d make sense to not build it underground, given how expensive that is.

@Ros, the Plan certainly concentrates on metropolitan Melbourne. I wonder if a separate regional rail plan is being prepared.

@Campbell, I agree, it’s good to see an aim for frequent services 7-days until late.

@Simon, see

@Andrew (#7), I think one needs to put the plan in the context of anticipated growth that they reckon will double Melbourne’s population.

I’m not quite sure what you’re getting at Re: Flinders Street. Even in the Stage 4 plan, it says all but 3 lines will still go through there.

@enno, what are you talking about “your loop”?

But if you’d read the comments above, or the Plan itself, or looked closely at the diagram, you’d see why.

@Ros, this plan is only for metropolitain rail (which apparently now includes Geelong for some reason). Future plans will look at trams, buses and the regional network.

@Simon, I’d be interested to see a scan/photo of that map.

@Andrew, the system is running out of capacity, and it will cost billions to fix. That doesn’t necessarily mean tunnels (at least in the short term), but it is true nonetheless.

The problems with capacity are evident when you have people standing for an hour or more – the generally accepted threshold seems to be around the 20-25min mark, even in peaks.

In-cab signalling is different from driverless trains, but much of the technology is shared. Eventually driverless trains would be the ideal – we should be creating a system that benefits millions of people, so if a few thousand lose their jobs over that I think it’s a fair price to pay. Besides, it’s so far off that the solution may well be to simply stop hiring (in a few decades) and let nature handle the rest.

Regarding freight and V/Line on the Dandenong route – this is something that none of the PTV plans, past or current, have addressed adequetely. The best solution is four tracks to Dandenong (and possibly further), allowing one pair of tracks to take the stopping (slow) trains and the second pair to take all the faster trains (freight, v/line, express, empty cars, whatever).

In the short term, this could be achieved by spending $2.65bn to grade separate Caulfield Junction (closing the Queens Avenue underpass and using its vertical alignment), giving a four-into-four-plus-four split. You then extend four tracks to both South Road (Moorabbin) and Warrigal Road (Oakleigh), giving a place for the expresses to overtake the stoppers. You even get nine grade separations in that, improved bus/train interchange, urban renewal and you can delay construction of the Swanston Street tunnel by up to ten years!*

*Pending the success of the Domain Interchange

I wonder how they plan to get a train on the Baxter – Craigieburn line that enters the City Loop at Richmond to exit at Nth Melbourne?

The City Loop is not organised that way – they are separate tunnels.

Or have they talked about reconfiguring the loop by digging new connections?


As Daniel said. The plans are from the Network Devt Plan – their not his own.

Answer is, second plan shows staging of ten minute headway provision across the whole (current) network).

First plan shows ultimate (20yrs+ – stage 4) network proposed – Still four lines through the ‘loop, it’s just that only two continue to be used as loop services (green and dark blue), whilst two existing lines are used for the Craigieburn to Baxter cross town route (orange line represents two existing lines to be reconfigured).

So on the first map, there is three lines between Melbourne Central and Parliament. That’s three. Blue , green and orange. That’s the three I was counting.

@enno the blue and green are single track, the orange is double track – so four tracks all up.

They are proposing to dig new connections from Richmond to the Northern loop and from Nth Melbourne to the Caulfield loop to make this possible.

So the orange line is a regular bi-directional track, using 2 tunnels ? And the blue and green lines are the one-way loop tunnels.
2+1+1 = 4. Ok, got it. Thanks for clearing that up !

David Stosser #12

Why would a grade separation at Caulfield cost $2.65 billion? The exactness of this figure, and the reference to the Queens Ave underpass, suggests that you are referring to a particular document. What is this please and is it available?

Enno #13

The Baxter-Craigieburn route requires a new chord from Nth Mel to join the Caulfield loop at Flagstaff, and a new chord from Richmond to join the Northern loop at Parliament. The freed up Caulfield and Northern loop sections between Flinders St and Cross are used to make a new track pair from the west becoming the Burnley relief line.

About 2 track kilometres of new dive and tunnel and preferably 2 track km for related work at Nth Melbourne** gives 20 extra peak direction peak hour train paths from the west (from bypassing the chokepoint where the Craigieburn and Sunbury lines merge to the Northern loop), and about 10 from the east (from removing the constraint of reversing Burnley relief line trains at Flinders St).

Compare the Metro Rail Tunnel: 18 track km of tunnel and 5 new stations gives 20 extra peak hour peak direction paths from the west and NONE from the east (it only replaces an existing track pair between city and Sth Yarra).

To make the comparison even starker: for each 4 new train paths, in round figures –
– the Northern Caulfield loop connection needs about ½ track km;
– the tunnel needs about 4 track km plus 1 new underground station;

New chords for the loop connection would no doubt be dearer per km than the tunnel, because of the set up costs associated with shorter distances, and the difficulties of connecting new to existing tunnels underground. Even allowing for that, I hazard a guess that the loop connection must be HUGELY more cost effective than the tunnel per new train path.

Accordingly I don’t understand why the loop connection is not being taken as the higher priority project. The report, though it claims to have examined alternatives before reaching its recommendations (p22), does not address this issue. I wonder whether this issue has been deliberately avoided so as not to upset the regrettable, premature political commitment to the tunnel.

** I would propose track slewing and a flyover west of Nth Melbourne to allow up/up/down/down with cross platform interchange at 2/3 & 5/4 (Craigieburn/Sunbury/Craigieburn/Sunbury – Werribee uses 6 and a new 7)

PS My comment above on the capacity of the metro rail tunnel, to compare like with like, assumes existing signalling on the Sunbury and Craigieburn lines which would feed it.

The tunnel without more does not increase capacity to Dandenong. Works that would increase capacity to Dandenong – resignalling or using longer trains – can be done on the existing tracks in any case and so should be considered as a separate project.

Wondering if they mention doing anything about Carnegie/Murrumbeena/Hughesdale in all of that. There were public meetings some time ago and I’ve heard nothing since then. I can’t help thinking it would make sense for the roads type people to get on board with Metro with areas like that so speed up traffic considerably.

The report does not discuss the priority of grade separations very clearly. It does not address the probability that putting a peak load of 24 per hour each way on the Dandenong line, before complete grade separation, may be impractical or politically unacceptable because of the effect on level crossing delay.

This is a good reason why longer trains should be a higher priority than the metro rail tunnel – they increase capacity with very little extra level crossing delay (only the few seconds that it takes the extra carriages to cross; but that is insignificant in context of total closed time).

Forgive me if I’ve misinterpreted the maps, but it seems they plan to squeeze 26 services an hour during peak onto the ‘new’ viaduct between Flinders Street and Southern Cross, creating a pinch point where the Grovedale-SY, Wallan-Sandy and Gippsland routes all combine. This seems contrary to the goal of untangling the network and leaves no margin for error on the two longest metro routes. Am I right?

Polly #25: There’s a clearer map on p124 of the report. Sandringham/ Sth Yarra continues as the southern track pair on the viaduct then splits at N Melbourne to go to Grovedale via the Regional Rail link tracks and to Upfield via a new flyover.

The suggested numbers add up in terms of capacity, but it does break down the sectorisation somewhat.

Comments are closed.