The Dandenong corridor wish list

In a series of media releases, Spring Street has announced it has dumped the previous Coalition government's Dandenong corridor project - which was originally pitched to them by a private consortium - and have announced their own.

Much noise was made about the PPP/Napthine scheme only removing four level crossings between Caulfield and Dandenong as well as a smaller number of trains that were to be ordered to run services on this major route. In one of the media releases, the following table was provided as a comparison between the Napthine/PPP scheme and the Andrews Government vision.

Comparison of project features. Source: State Government media release

According to one media release, expressions of interest are to be called for the grade separations within months and the Premier has preemptively apologised for the shut down of the entire corridor in the name of progress while individual project areas are worked on.

Fair enough Premier, however could we have more detail please?

Given the relative "out of the blue" nature of the project announcement with not much detail beyond the points listed in the image above, here's a wish list of changes Urban Melbourne would like to see happen either in conjunction (priority 1), in the medium term (priority 2) and in the long term (priority 3) with the Dandenong corridor upgrade.

Priority 1

Link the infrastructure investment with more intensive land uses adjacent to the rail corridor

This is a golden opportunity to provide more affordable housing in the middle ring suburbs and Spring Street should make inclusionary zoning as mooted by the Planning Minister (or any other policy targeted at increasing the diversity of urban housing stock) apply to all areas adjacent to newly re-built stations as well as throughout any existing or new structure plan in the corridor.

The Government should work with councils to update structure plans within station vicinities, or create new ones where none exist, in order to capture the benefit of the rail investment. The larger number of trains in the Andrews' Government proposal appears to be linked to increased frequencies of service and therefore the major nodes in the corridor scope: Caulfield, Oakleigh, Clayton, Springvale and Dandenong should see more intense and inclusive development made possible through the planning scheme.

Plan Melbourne has previously designated these areas as Activity Centres with Caulfield and Dandenong as National Employment Clusters, further reinforcing their importance within the wider network.

Dandenong should be a special case building on the existing revitalising central Dandenong principles. According to the Southern sub-region growth fact sheet as part of Plan Melbourne, it is predicted that the area will see another 400,000 to 480,000 people living, for the most part, within reach of a Dandenong corridor station by 2031.

Spring Street should take a leaf out of the New South Wales Government's book and further decentralise public sector organisations and further encourage more private sector investment and employment growth in Dandenong like Macquarie Street has with Parramatta.

Cranbourne-Pakenham Transformation video from the Premier's website.

Priority 2

Mass reconfiguration of the bus network to feed the upgraded rail line

A public transport network is a failure if users are still dependent on a private vehicle to access it. Public transport is a logical extension of walking and where distances are too great between the point of origin and the mass transit mode, other forms of public transport need to be either created or reconfigured to ensure there is no dependency on private vehicles.

Beyond Caulfield and Carnegie, the only other major mode of public transport other than the rail services is the bus network. While a few of the Smart Bus services operate and act as feeder services in the area already, every station - including along the Pakenham and Cranbourne segments - should see less focus on car parking and more on bus service expansion.

Rather than focus on the bells and whistles of a Smart Bus service, bus network/feeder policy should focus on these core principles: frequency, route directness and operating hour expansion. The end-game should be to correct the sheer waste of land at stations like Berwick which has hectares of parking that should be put to better use, especially in the scenario where VicTrack own the land.

Springvale Station only had two platforms built as part of the recent grade separation project. Source: Daniel Bowen's blog.

Priority 3

At the very least keep one eye on corridor expansion and tell the public about it

No mention was made of expanding the track within the existing corridor however the track and station design should take into account that the corridor will be expanded in future. The logical next step would be to quadruplicate the track between Caulfield and Dandenong thus enabling two tiers of service - stopping-all-stations and express - to operate independently over the longer term.

The future track layout around Oakleigh and Clayton stations should cater for four tracks and four platforms, all other stations should cater for four tracks with only two platforms.

The fastest services on the Sydney Trains network between Sydney's Central station and Parramatta is 27 minutes. This is achieved through extensive and frequent express services which for the most part run on track which is rated for speeds in the vicinity of 80kph.

A direct Flinders Street service from Dandenong according to the current timetable is 41 minutes.

The City-Dandenong corridor (29km as measured from Flinders Street to South Yarra along the projected alignment of the Melbourne Metro Tunnel project; South Yarra to Dandenong station along the existing corridor) is 6km longer than the Sydney-Parramatta corridor (23km measured from Central station along the existing Main Western Line to Parramatta).

Given the advantage of a straighter alignment thus enabling higher speeds with express services, should track be quadruplicated - perhaps as part of a second stage of the Melbourne Metro Tunnel project - it is within the realms of possibility that a similar CBD to CBD trip, time-wise, could be undertaken in Melbourne. This should have the effect of improving Dandenong's profile as a employment and growth centre for people living 'up' the line toward the city, not just for those who live beyond Dandenong and commute to the city at present.

At the very least this should be the ultimate plan: creating another fast, frequent transport spine that caters for short and longer trips right through South Eastern Melbourne in both directions, not just for the current peak-time commuters.

Lead image credit: recently grade separated Springvale Station, via Wikipedia.


Llib's picture

This is a missed opportunity to design for future quadding, if it is not factored in the level crossings and upgrades as you stated it will be a lot more expensive when it needs to get built.

Back to top
Riccardo's picture

Good announcement though more needed, and the article is mostly OK

Sandringham was actually the more significant announcement technically, now all that is needed is commitment to remove the remaining crossings (4?) from that line so that the line is completely segregated from ground transport, and remove the connection to the balance of the legacy system, so that the line can be converted to automatic operation and operate at a higher frequency.

With all the backward and forward over the Mandurah-ising of the Doncaster route, I wonder whether it might be better to build a new Dandenong-Pakenham-Gippsland line down the M1 corridor, with the GWY line cannabalised instead. While the legacy line is wide enough for four tracks (except between Oakleigh and Caulfield) I worry that between Caulfield and Melbourne is too constrained now.

If anything, a useful future for the Caulfield-Richmond corridor is to have two tracks for express Frankston line trains, and the other pair for all stations to Cheltenham mixed with all stations to Dandenong, and run the trains to beyond Dandenong on this other new line.

Holmesglen to Glen Waverley could then just be a simple shuttle, the way Alamein ought to be.

Back to top
Simon wxtre's picture

'While the legacy line is wide enough for four tracks (except between Oakleigh and Caulfield)'

That is untrue. The railway line is wide enough for four tracks between Oakleigh and Caulfield. The railway line easements where designed to allow for 4 tracks. Someone actually did the measurements on this section and the room is available.

Back to top
Llib's picture

The room is available although if they remove the level crossings without catering for quadding they will have to demolish recently built bridges and stations to cater for the new tracks.

Back to top
Alastair Taylor's picture

While the legacy line is wide enough for four tracks (except between Oakleigh and Caulfield)'

That is untrue. The railway line is wide enough for four tracks between Oakleigh and Caulfield. The railway line easements where designed to allow for 4 tracks. Someone actually did the measurements on this section and the room is available.

I would be surprised if there are rail corridors in Melbourne where there is no space for an extra track pair on an existing 2 track pair corridor.

It would just mean - in most cases - track and associated infrastructure would be built right up to property lines. And it's been done before: classic example is between South Yarra and Hawksburn just to the east of the Chapel Street bridge - the existing 4 tracks are right on the property lines.

Caulfield - Oakleigh no different.

Back to top
Riccardo's picture

"It would just mean - in most cases - track and associated infrastructure would be built right up to property lines."

Yes Alastair Tayser, MP for Glen Eira, look forward to seeing your election campaign on that one.

Stick this in your pipe and smoke it,145.046964,3a,75y,14.48h,90.81t/data=!3m4!1e1!3m2!1snZFYu3OMOyvGqhwjIotvdA!2e0,145.076786,3a,75y,158.83h,91.54t/data=!3m4!1e1!3m2!1sX-ZiY1oWs69DAuGdHIBjag!2e0

Back to top
Riccardo's picture

With Victorians having so little experience of four track railways, I've posted a great shot off a bridge on google maps from Padstow on the East Hills route. This is what four track railways actually look like these days.,151.031801,3a,75y,273.37h,89t/data=!3m4!1e1!3m2!1syGuYz0zsCgTEQfmrSSRQig!2e0

Back to top
Ian Woodcock's picture

All of which raises the question of the best way to do the level crossing removals. There are two main aspects to this. The first concerns the transport function and the cost-benefits of achieving grade separations. The second concerns the issue of development around stations and the amenity of the corridor in general.

Taking the question of the transport function first: Is it easier and less disruptive to add extra tracks in a trench or with a viaduct? These are the only two acceptable options for level crossing removal in an urban area if there is already an activity centre, or the desire is to have one in future around the station. VicRoads last year very clearly and sensibly vetoed the 'road under rail' proposals by the PPP consortium, and the legacy of road-over-rail grade separations around Melbourne must make it clear that no-one in their right minds will propose these. So, it's either trenches (rail-under-road) or viaducts (rail-over-road). For some reason, the state has been very shy to include rail-over for level crossing removals in recent times. Yet, from a transport functionality view, they are far superior to dropping the line. The most energy efficient way to run trains is to have the line rise into stations and drop on the way out - it can be 10 times more efficient than the other way around, which requires more wear and tear on brakes and far more power accelerating out of stations.

Then there is the cost and disruption associated with digging massive trenches, moving all of the services that run along rail corridors, lining the trenches with retaining walls, building bridges for the roads over the top, and then laying new tracks in them. And once the trench is complete, before trains can be run, the entire thing must have anti-suicide fencing along the top for the length of the trench.

The rail-over option for grade separations is far preferable on the basis of the two criteria above - transport functionality and cost. Building a viaduct to run the new tracks on can cost as little as a third of the cost of digging a trench.

But there are other benefits that go well-beyond these criteria (which could be regarded as mainly focused on the engineering and cost alone). What about the amenity of the corridor? What about the wider potentials for intensifying the precincts around the stations?

One of the significant advantages of rail-over is that it provides dramatically improved ground level permeability and connectivity without having to build expensive decks - decks that are not strictly part of the transport task. Many architects, planners and urban designers like to see the massive expense of such decking as an investment in the future, since at present, most suburban stations are unlikely to attract the level of development that would pay for them within the project itself. Just for a moment, reflect on the costs at Glen Waverley's Ikon development. It may be 10 storeys, mixed, and very popular, but VicTrack only made enough from it for a station upgrade - not to re-level the tracks and put the line beneath the building so it could be extended (which is the kind of situation that would be comparable with the idea that developers will pay to build over trenched rail stations throughout Melbourne's suburbs). So, except for extremely high value locations, such as the central city and putative suburban CBDs (and even then, the land values may not work for a very long time), sinking a rail line with public money so the private sector can build over the top isn't realistic. And asking the private sector to build-in affordable housing looks even less likely in such a scenario. It would be far better to recognise the significant value-uplift that will accrue to privately-owned land in the catchment of the improved rail line and stations and enforce inclusionary zoning there.

So, rail-over-road effectively provides 'free' real estate beneath the elevated tracks. This means that it's much easier to attract the kind of development that most transport planners and sensible urbanists would prefer to see co-located with stations: non-residential land uses that encourage economic and social activity - so retail, commercial and community uses, that can easily be built at ground level and help to integrate the station and its local precinct far better than if there is a 9m-deep trench lined with anti-suicide fencing either side of the one place to cross every few kilometres. And between Caulfield and Dandenong that is what we're talking about: 18km of line, 9 level crossing removals - that's an average of 2km between each crossing point.

So, thinking about this as an urban corridor rather than a set of 9 level crossings with redeveloped stations, what broader urban implications are there between these two ways of approaching grade separation? Currently, the surface level tracks create an 18km-long division between the people living and working on either side of the tracks, with only 9 places they can legitimately get across (with a few pedestrian-only crossings also, here and there). With one long trench, or a series of trenches, this dis-connection remain. The traffic would flow better at the 9 roads that cross the corridor, and that's a great benefit, but really, not much else is going to happen. There won't be any decks on top of the trenches for development, so no increased permeability or urban intensification. And the VicTrack land will have to be kept in reserve for the addition of extra tracks in future, so no development along the corridor either side of the tracks either.

But if the entire corridor, or all of those sections where it is feasible were to be done as viaducts, then the corridor could be re-connected at ground level, opening up immense potential to improve the amenity of the places either side of it, as well as of the corridor itself. It could become an 18km linear public spine, with parklands, recreation facilities, walking tracks, bike trails, active and passive sporting areas, markets, community gardens, retailing, community uses and other economically and culturally productive land uses developing over time. This increased connectivity would raise value of the private land all the way along the line. And this way, the flow on or agglomeration benefits of grade separations on this corridor, with much improved network effects between trains and buses, walkability and cycle access would be far superior to that possible from trenching. And if the land values do ever reach sufficient levels to warrant significant intensification at stations, then there are plenty of precedents from Sydney as well as overseas that show how this can occur around an elevated track (think: Chatswood, for example).

A few people may have concerns on reading this about the appearance of elevated tracks and stations. Elevated rail is very common throughout the world and there are many, many precedents for it that show how it can be an opportunity for great design far more than sinking stations in trenches has been to date in Melbourne.

Back to top
Riccardo's picture

Ian I agree, but you've got your work cut out trying to persuade Melbourne people on this. Shame we can't rip their brains out and send them to Singapore or Japan for 'cleaning' and when they reinstall them, they can see how right your ideas are.

Back to top
Riccardo's picture

I think too, Ian, elevate rail in Melbourne was cursed by the Yarra-fronting monstrosities built in 1914 and 1976. And some very poor road over rail overpasses around town.

The 1976, and the 1914 works may not need to have gone ahead, and definitely had a remedy when the 'City Loop' debacle was under design.

There is no reason why all country trains apart from Gippsland could not have terminated at Spencer St, as they do, and all Gippsland trains terminated at Flinders St, as is common in many overseas countries. The demand for travel by train FROM Flinders St itself TO Spencer St itself is low.

A better approach than the City Loop would have been a dedicated underground metro from Richmond to North Melbourne. This line, or embryonic network, could then have picked up other journey sets, including via Spencer St to Docklands, via FSS to St Kilda Road, via Parliament to wards Clifton Hill and on it goes.

As for elevated, you would then use it where it needed, but not along river frontages!

Back to top
Ian Woodcock's picture

Personally, I like the viaduct along the northbank, it's one of my favourite places in town. What's a shame is that the market buildings that were there were demolished!

* Elevated rail is *not* the same as a road overpass, and this is extremely important for those interested in the urban design implications. Compare, for example the following:

* Elevated rail stations: Glenferrie, Auburn, Canterbury, Balaclava, Kensington, North Richmond, Collingwood, Victoria Park, Gardenvale - note how at all of these stations, the urban streetscape is relatively uninterrupted by the rail viaduct, and in some, the rail-bus / rail-tram transfer is extremely good (i.e. very closely connected). In some case, particularly Glenferrie and Auburn, the station forms part of the streetscape. You don't hear calls from these communities to have their stations dropped into a trench because of its 'impact' on the local area, or concerns about property values. These are some of the most desirable places in Melbourne.

* Road overpasses over rail: Burnley, Sunshine, Newport, Oakleigh. Very unpleasant effects on the local area, hard to imagine how a decent urban streetscape could be created around these structures. Where do the pedestrians get to go? Either footbridges or underpasses. Road-under-rail underpasses are similarly negative urban environments, no matter how much they are dressed up. It's hard to see how they could be 'urbanised' with retailing or anything else.

This distinction is important, because apart from these largely historic examples (which are part of Melbourne's heritage) we don't really have any evidence of grade separations from recent experience in Melbourne other than rail-under-road (like Mitcham, Springvale, Nunawading) or stations-as-footbridges (Sunshine, Footscray, West Footscray, Williams Landing, Roxburgh Park, Coolaroo, to name a few) where intermodal transfer access may not have been improved as much as it could have been had a full grade separation been undertaken. And yes, it would have been prohibitively expensive at many of these latter locations, but not the former necessarily. In any event, a very large number of the planned 50 level crossing removals would be prime candidates for rail-over-road, and it is to be hoped that those charged with the responsibility for making them happen will have the courage to actually do it. And then, we will have more up-to-date evidence upon which to discuss the merits of the various approaches than we have from recent experience that has been hampered by prejudice about what can work.

Back to top
Simon wxtre's picture

If you look at Blackburn station as an example the railway line will need to be lowered as there are shops on either side of the road and a street adjacent to the railway line. It is similar with the Dandenong line the tracks will need to be lowered in places out of necessity. Viaducts are quite good as you point out Auburn and Glenferrie stations but they can also be ugly as evidenced by the Flinders St viaduct and viaduct's in the UK. Underground metro is the modern method of contructing railway lines and most cities choose this option for new lines. But the Dandenong line is an existing mainline, freight and regional trains share the tracks with metropolitan trains so it is impractical. Quadruple track on mainlines improves services as evidenced by London and Sydney and other European and North American cities. This should be the main priority in creating proper express services from the outer suburbs.

Back to top
Ian Woodcock's picture

I disagree that underground is the 'modern method' of doing things. Firstly, underground railways have been around since the 1860s, though the idea was proposed much earlier than that for central London. Underground railways are just one way of building a railway, and usually, they're only deemed necessary in high-density urban environments, or for crossing large bodies of water (like the Channel between France and England, for example). If you look at some of the current urban rail projects, they combine a mixture of above ground, surface and below ground alignments according to context and topography - for example, the North West Rail Link in Sydney. Around the world, in the last decade, urban rail has been built in all of these alignments - there is nothing more modern about any particular alignment, only better ways of doing things to make systems work within their urban contexts.

Regarding Blackburn, it is not a high density urban environment, and there is quite a reasonable distance between the shops on either side of the tracks - the rail reserve through that whole area is quite generous overall. The current proposal to lower the tracks is not going to put the line underground, it won't disappear out of sight into a tunnel. It will be put in a trench. To make the levels work, the incline into the station from the east will be very steep, putting extra stress on the brakes and the engines as trains move through it. It is not an ideal outcome from a transport operation view, and in my view, putting the station in a trench is not the best outcome for Blackburn either.

At the moment, there are just a couple of ways to get from one side of the Blackburn shopping strip to the other - one via the underpass beneath the station, the other via Blackburn Rd. Neither is ideal from a pedestrian amenity view.

If the station were to be elevated, not only would the congestion issue at Blackburn Rd be solved, but it would create enormous potential to connect the communities and shops on either side of the rail alignment. While there are examples of unattractive viaducts around the world, equally, there are examples of attractive ones that add to the character of the place they're in. For example, some of the boulevards of central Paris have metro lines on viaducts, with markets in the spaces beneath them. They are very pleasant places to be. With a bit of imagination, I'm sure that Blackburn could be enhanced by such an approach - we have some of the best architects and urban designers in the world in Melbourne, we should have more faith in their abilities.

Back to top
Riccardo's picture

Leaving aside that Simon Doe is just wrong, I think the issue Ian is that the design principles you have espoused with elevated rail are very good, but have not yet been followed to date.

The Yarra-side rail viaducts encroached on public riverbank (and as I mentioned, weren't even necessary). The Banana Alley section showed that you could reuse the undercroft but this was not followed elsewhere. The older viaduct is quite noisy, a consequence of all its steel.

Other parts of the system where you would expect to see good rail over road outcomes include Richmond (just awful, worst of 1960s); Patterson (not bad); Newmarket/Flemington; Hartwell/Burwood (really should be one station over Toorak Rd); Canterbury (terrible, again worst of 60s). Probably others.

Agree road over rail is a killer -Newport for example.

Back to top
Riccardo's picture

No. This is classic dribbler/foamer exceptionalism (Australia/Victoria/Melbourne is 'different/special' when in fact is isn't).

Nothing 'necessary' about it at all. Just neurological pain from tribal response to change.

Back to top
Thomas The Think Engine's picture

Hey to my old mate Riccardo! I remember having many good conversations with you over at Transport Textbook and on my blog. Made this account mainly just to say gday.

Back to top
Michael Bell's picture

Much of Vacouver's SkyTrain (existing and new lines currently under development) runs on elevated viaducts. It is visually appealing, it has well integrated transport interchanges, and is apparently much loved by the locals.

A 4km section of Sydney's new north west rail link will also be elevated.

The level crossing removal authority would have to be nuts not to consider overhead rail as a viable and sensible option for rail separation in Melbourne.

Back to top
Michael Bell's picture

The Office of the Victorian Government Architect (OVGA) has made some startling and I believe misguided statements about the suitability of elevated rail following a review of Melbourne grade separations projects:

VOGA advises avoiding elevated rail grade separations, on the premise that they "have a significant physical presence and impact on a place" and "can impact on visual amenity, permeability, viability of activity areas, the value of land and appetite for future private development". When I read this, it struck me that the same arguments apply equally to rail trenches, which can stretch up to 1000m either side of a road crossing with high safety fencing, creating an impermeable barrier, having a visual impact and virtually eliminating the possibility of beneficial land use due to the high cost of decking over rail.

The VOGA paper leaves two questions unresolved:

1. On which Melbourne grade separation projects were the lessons learned that suggest elevated rail is not preferable? Have there actually been any elevated rail grade separations in recent times? (Narre Warren-Cranbourne Rd, Taylors Rd and Andersons Rd, come to mind, but those were road trenches under, rather than elevated rail.)

2. Does the guidance actually refer to rail on earth embankment and buttressed bridging over road, (like Princes Hwy, Malvern) rather than continuously elevated viaduct? If so, this distinction should be clearly stated, as these forms of elevated rail grade separations clearly differ.

Back to top
johnproctor's picture

Ovga are crap.

Back to top
johnproctor's picture

I expect the inner 6 will be trenches but the outer ones (heatherton, chandler, corrigan) might be rail over.

As Riccardo posted modern railway standards mean 4 track railways are wider than anticipated when this rail reserve was made and if you were building caulfield-South yarra now that corridor would be much wider as well. There are sections of the dandy corridor where 4 tracks won't fit - generally short pinch points. As I said elsewhere though the capacity generated by the grade separations allowing more trains, use of higher capacity trAins and then in time longer trains means 4 tracks wont be required for quite some time. Hastings will probably drive quadding if it ever happens.

Back to top
Riccardo's picture

Hi Again Thomas the Think Engine

I'm still trying to fly the flag for rational transport policy - sadly you can see with the EWL schemozzle that the public are still addicted to irrational transport policy and the political class gladly oblige. And to make matters worse, blogs like this one then give voice to even more irrationalists.

Though if you want truly nutty read here

What is distressing about some of the posters here suggesting that Caulfield to Dandenong can be easily quadded is their implication that the 80km/h transit speed for long distance trains through that section is satisfactory.

Back to top