Dandenong Line upgrade
Arguably the most important commuter line in Melbourne is the Dandenong line, simply because it connects Melbourne CBD with the proposed second CBD of Dandenong, as well as being the single longest commuter line in Melbourne being approximately 58 kilometres long.
The Dandenong line is essentially a dual track line from Richmond to Pakenham, with a single track branch to Cranbourne. There is a small segment of quad track between Caulfield and Richmond, however in usual use it is two tracks for the Dandenong Line, and two tracks for the Frankston line.
The high level of patronage, coupled with only two tracks means that the absolute majority of services stop all stations, to maximise potential capacity.
The only way to increase speed on the Dandenong line is to introduce skip-stop services, close low use stations in peak periods, or expand to four tracks.
The Victorian Government originally proposed for a third track to be constructed from Caulfield to Dandenong costing approximately 1 billion dollars, which has now been scaled back to a third track to Springvale. Third tracks are highly undesirable, as they can only increase speed in the peak direction, and also work best if the train is only doing one trip (for example, Belgrave – City Loop – Flinders St – terminate, then run empty to Burnley) as there is little capacity in the opposite direction for these trains to run back to the end of the line.
This proposal looks at three ways to increase the speed on the Dandenong line;
1. Close little used stations and replace with a high frequency bus service (min every 10-15min)
2. Introduce a skip-stop pattern
3. quad track.
1. The most used stations in order on the Dandenong Line from Caulfield are :
a. Caulfield
b. Oakleigh
c. Dandenong
d. Springvale
e. Noble Park
f. Clayton
g. Huntingdale
h. Narre-Warren
i. Carnegie
j. Berwick
k. Hallam
l. Murrumbeena
m. Pakenham
n. Hughesdale
o. Cranbourne
p. Westall
q. Sandown Park
r. Yarraman
s. Merinda Park
t. Beaconsfield
u. Officer
This means you could potentially remove Carnegie, Sandown Park, Hughesdale, Murrumbeena, Yarraman, Westall stations between Caulfield and Dandenong. This would speed up services in this section by exactly 6 minutes.
This shows that a reasonably direct bus route can be implemented to replace these stations; however the disbenefits of closing existing stations needs to be looked at seriously, with such a proposal. The peak hour only or permanent closure of 7 stations; most within the inner or middle rings of Melbourne (therefore with the greatest redevelopment potential) is something that is unlikely to occur. Therefore this is an option of absolute last resort.
2. Introducing Skip-stop patterns
Skip stop is a simple way of increasing speeds on an existing line. This means that stations are identified as being either:
All Station
B Station
C Station
In this example all trains will stop at:
Caulfield; Clayton; Dandenong Stations
B trains will stop at:
Carnegie; Hughesdale; Huntingdale; Springvale; Noble Park Stations and the all stop stations;
C trains will stop at
Murrumbeena; Oakleigh; Westall; Sandown Park; Yarraman Stations and the all stop stations.
This would speed up all trains by the amount of skipped stations, which is traditionally 1 minute at speeds below 90km/hr and 2 minutes at speeds above 90km/hr. Given that the Dandenong Line in this area, you would receive a timesaving of approximately 5 minutes on the B and C trains; bringing down the entire trip time from Dandenong to 36minutes, compared to 41minutes now on an all stations train.
Clayton was chosen as it is the midpoint between Caulfield and Dandenong station wise, over any other reason.
Pros of Skip-stop
Enables maximum frequency on a dual track line
Enables maximum efficiency of the existing infrastructure
Enables a maximum frequency of 20tph (10tph on each pattern)
Minimal infrastructure cost
Cons of Skip-stop
Unnecessarily confusing; requires passengers to know which pattern train to be on
Minimal time savings in comparison to true express-local patterns
Requires a highly efficient signalling system to ensure bunching does not occur
If delays occur, the system will cascade badly.
3. Quad track to Dandenong
Quad track is the most costly option of increasing speed on the Dandenong line, as it requires the construction of two additional tracks, and the associated infrastructure.
For this process we are looking at the area between Caulfield and Dandenong for the implementation of quad track.
Given that there is approximately 9 level crossings between Caulfield and Noble Park, and an additional 1 level crossings to Dandenong, this requires physical separation of trains and cars to enable optimum use of the railway network. The Level crossing between Noble Park and Dandenong could potentially be closed given its proximity to the grade separated Heatherdale? Road.
This proposal looks at the construction of a trench between Noble Park and Caulfield, to bypass 9 level crossings, and enable the demolition of two overpasses that have significantly impacted on the urban form of those two centres.
This project would have to be split into two components to enable trains to run on the majority of the tracks:
Dandenong to Clayton,
Clayton to Caulfield
I have costed the trench at 30million per kilometre, a local station at 7million and a major station at 14million.
The trench is approximately 16m wide and 6 m deep for its length; without stations. With a local station only it is 22m wide; and a major station it is 28m wide. This provides for approximately 1.25million cubic metres of spoil for removal.
The total cost of the trench is approximately $400million.
One potential way that the cost of the trenching can be recouped is to deck over the newly created tunnel and sell off the land above. The Dandenong Railway line is in a corridor that is approximately 40m wide for its length from the first bridge from Caulfield to Dandenong. The area of trench between Caulfield and Noble Park is approximately 15 kilometres, which given an average land value of $250m2 (approx $25,000 per 100m2), which is the absolute lower end for average land values, could recoup approximately $152million, towards the redevelopment. This assumes that the bridging of the corridor would be done by the developer.
This brings the cost of the trenching down to 250million.
The cost of bus replacement services for the line has been calculated below:
It assumes that there is a requirement for 70 buses between 5am and 11pm (18hours of service) each costing $60/hour to operate. Of these buses, 40 would be express services, stopping only Clayton – Dandenong/Caulfield, and 30 would be all stops services). This would provide for approximately 5,000 people per hour capacity in each direction. This means that each day would cost $63,000 or $11.5million for the 6months shutdown.
This means that the total cost for this shutdown would be 299million, after land sales and provide for a significantly improved speed and capacity on the Dandenong Line.
The tme of closure could be shortened by spending additional money.
Thoughts?
Excellent analysis of the options James.
I agree with your findings on skip-stop – it looks fine in theory (and in fact Adelaide seems to be pretty happy with their line that operates that way) but this is Melbourne and things DO go wrong.
Ditto for closing stations. Personally I’d love to see a Perth-style semi-interurban system, but the DoT seems intent on making our system a Metro. Apart from that, closing stations is a sure vote loser.
Quadding is definitely the best option. Grade separation should also be included – in fact, I doubt that stations like Murrumbeena, Clayton and Springvale (and all the rest that have a level crossing) are even possible to be quadded at ground level.
In fact, if you don’t mind me plugging my own group, Smart Passengers came to the same conclusion and a book is coming out on the subject – in fact I just sent it off to the printers yesterday. We’re only asking for quadding as far as Oakleigh for now, to keep the costs down. We also want a flyover at Caulfield to eliminate conflicts at the junction, that adds a bit to the bill.
> One potential way that the cost of the trenching can be recouped is to
> deck over the newly created tunnel and sell off the land above.
I think that’s a big ask. Who is going to want to build what over the top of railway cutting?
One of the things I’ve been looking at for Sydney is BRT as an alternative to stoppers on some of the inner lines. That concept might have an application here, though I doubt it would be as cost effective as in Sydney.
I’m increasingly of the view that modern heavy rail should not have stations any less than 5000m apart. Closer than that and the advantages of heavy rail are lost, and all you are left with is the cost.
djf – that’s the Perth theory. Apart from CBD walkup stations, others are being built so that the train gets a decent bit of cruise happening before having to slow down again. And buses as the bit in the middle.
The bus/train combo obviously disadvantages inner users as the extra time is a large part of journey time, but as in Perth it might be better to leave those people on the bus eg Canning Bridge.
However, Melbourne is at a disadvantage here because of the trams – they crawl. A tram from say Carnegie or Murrumbeena to Caulfield isn’t going to set any speed records.
I think the potential for station closure time savings is actually further out – Sandown Park, Westall, Huntingdale for example.
Sadly, the ones that you would close to save time are the obvious ones for park and ride or bussing from further out eg Yarraman next to Eastlink, Huntingdale on North Rd, Westall on Westall/Blackburn Rd and so on. Whereas Carnegie and Murrumbeena could be readily densed up and metrofied if you could put in extra track and grade sep.
Which then gets me to the ‘why even bother’ side of this. Maybe cheaper and easier to just build a new railway, underground from St K Rd all the way to Chadstone and Monash – and then just leave the existing route as a medium frequency all stations route from Westall to the city. The new line would cost a fortune but so would fixing the existing one and the outcome would still be poor.
I’ll look into the other ideas from a foreigners point of view, but I want to bring this one up
“The total cost of the trench is approximately $400million”
I wonder if you are being a tad optimistic there? The New Lynn trench in Auckland, which is just shy of a kilometre, had a contract cost of NZ$160 million and as far as I am aware, there was no gold-platting of that project.
Riccardo re “why even bother”: If we built a new underground line basically following Dandenong Road (thereby catching Chadstone, Oakleigh and Monash Uni) we could make the existing line a standard gauge freight line. The level crossings would still be a problem though, as would the shopping strips that have grown up around the existing stations. But we know Eastlink was designed to have a single track freight line added between Dandenong and Frankston, which would give us an option for rail access to the Port of Hastings. Of course the question of how to get from Caulfield to Dynon is still a sticky one.
Didn’t you do a post a while back advocating the use of the Dandenong Road tram line as a high capacity transport corridor? If it could be extended past Malvern it would just about do the job.
However, I have to disagree with your last sentence. The existing Dandenong line can be transformed into a four track interurban and metro combo for a reasonable figure and that would give us a good outcome for both suburban and country passengers. The only problem is the freight that comes up from Gippsland, although that can be allowed for as long as it never needs to go onto standard gauge.
What about the simpler option of running a combination of short runners and expresses all day?
For example, services between the city loop and Oakleigh run all stops, while services to Dandenong run express to Oakleigh (stopping at Richmond, South Yarra and Caulfield only) then all stops Oakleigh onwards. These could be timetabled to provide seamless connections between the inner and outer sections of the line: i.e: the all-stops terminates at Oakleigh platform two a couple of minutes before the outbound express stops at platform three, then it leaves for the city again a minute or so after the inbound express stops at platform one.
This means that an express can depart Flinders St every ten minutes with an all-stops short runner departing about a minute later, by the time the express has reached Oakleigh it has saved ten minutes and caught up to the preceding short runner and can pick up it’s patrons that wish to travel further out.
All the other various stopping patterns on the Dandenong group could be done away with so there was only one express pattern and one all stops that connected to each other half way and at interchange stations (i.e. no more short runs to Caulfield, no short runs to Springvale, no express to Malvern, no express to Caulfield etc.)
Vlines could simply ‘platoon’ on to the express stopping pattern during peak hours, following a suburban express at the minimum safe distance and taking advantage of the same all-stops connections. Monash Uni shuttles could be rerouted to primarily service Oakleigh rather than Huntingdale.
This gives a significant time saving from stations on the outer part of the line while maintaining the full user functionality, all without the need to build significant new infrastructure. The problem with skip-stops and three different service patterns is that it reduces the effective frequency of service for most origin-destination combinations, in some cases by as much as two-thirds.
Perhaps the only drawback is that it doesn’t address the level crossing issue, which could be a big problem if the gates are coming down four times every ten minutes at Grange Rd, Carnegie, Murrumbeena and Hugesdale.
Good post and you’ve included costings….
But I’d like to see where you got the patronage data from. I’m not too sure you’ve got Carnegie, Beaconsfield and Sandown Park right for patronage.
Also, with cutting and covering the line, what do you do about Westall’s maintenance depot and Dandenong’s stabling sidings?
After Oakleigh I don’t think you would bother continuing the full trench. From there the surface corridor is wide enough for four tracks with some major cross streets already grade separated, it would just be a case of trenching Clayton and Springvale stations and undertaking more conventional grade separation at the remaining cross roads.
Furthermore the value of land that far out probably wouldn’t make it worthwhile to deck back over the line, as it is mostly sprawlling industrial with a bit of suburban housing.
Have you had a look at Metro’s new timetable for this line?
Essentially, it was designed to stop loading breaches as much as possible.
Here’s the link:
http://junetimetable.metrotrains.com.au/
Interesting reading.
Hi All,
Sorry about the delay in replying – i was in Melbourne for a very long weekend. I didnt feel up to replying last night heh.
Land uses above a cutting:
After Oakleigh the line does run through mostly industrial lands. This corridor is approximately 40m wide, which is sufficient for a new industrial building or bulky goods building to be installed on the land (say Harvey Norman for Bulky goods etc). There is even potential for the existing industrial lands to expand into this corridor.
John-ston – The cost itself is probably being optimistic, but there is scope for sale of the spoil etc. Given the relative geology of the areas surrounding it (saying this only as a layman), there is potential for the spoil to be used for road base/fill, and if there is bluestone, for that to be sold off as well, which could theoretically cover some of the costs. It was also guessing that the ground is relatively stable, and wouldnt need significant underpinning.
That said, one of the biggest costs in trenching or tunneling is the removal of spoil – if a suitable site could be found (disused open cut mines hint hint) and dumped their cheaply, that negates a major cost.
Also the sale of the land above, i went for the absolute lowest price. In reality At the Noble Park end you’d probably be looking at $750sqm, and at Oakleigh end closer to $1200sqm, which would increase the money avaliable. It could also be an impetus for major urban renewal along the corridor, especially if done as part of a planning scheme amendment to allow for higher density around the major centres – say Noble Park, Springvale?, Clayton, Huntingdale, Oakleigh and Carnegie. This might be enough to enable for urban density within areas outside the Sydney Road – St Kilda Road Axis, to gain traction.
‘Notch – can i email you this? – And as for Westall/Dandenong – i left those parts alone overall. That said – with Westall, it is debatable as to whether its in the optimum position for what it is – and whether a location down the end of the line would be more appropriate. The trench essentially ends at Noble Park.
Nick – From a purely transport perspective, i’d agree with you, but from an urban design point of view, Huntingdale and Oakleigh have been devestated by the overpasses, and if they could be removed, a ‘blight’ so to speak would be removed, which is part of the urban renewal that could potentially happen in these areas.
And yes, i have seen the timetables – its a start definitely.
I have been tempted to put up a Melbourne style proposal like i did for Sydney.
One would have to consider the heritage implications of putting stations in trenches, one reason why I’d advocate leaving Oakleigh as it is.
Hughesdale and Huntingdale are concrete block modernist-functionalist slabs. I imagine no one would object to demolishing these and replacing them with new stations in the trench.
Cargegie, Murrumbeena and Clayton are weatherboard structures with some historical and architectural merit. Given the timber framed construction it would be simple to relocate these, perhaps down on the trench platforms but more likely as a surface level station entrance building (or otherwise relocated nearby and put to new use).
Oakleigh presents a problem. It is a historically and architecturally meritous structure so would need to be maintained, but it is made of brick and stone so would be very difficult to lower or relocate…. yet removing the horrific road overpass would be pretty critical to the revitalisation of the station area.
Perhaps the solution at Oakleigh would be to replace the Warrigal Rd overpass with and underpass (one wonders why they didn’t do that in the first place). The rails would still be at surface level and present a barrier to movement, but the road overpass would be gone and some development could then occur either side of the line.
James, my email address is a Gmail address with my full username.
As for fill – there’s a company called Lantrak that make an absolute motzer providing clean fill; except for the ballast layer, it’d be pretty sandy down below. Remember, you’re skirting the golf-course sandbelt.
Personally, I’d drop the line from South Yarra, even out the grade a bit, leave Malvern and dive at the down end of Caulfield and come out on the down side of Hughesdale. The rest of the alignment’s wide enough for quadding without having to worry about finding buyers for land.
Hi James,
Nice article, and I am just responding to the spoil issue. I am not sure about Melbourne, but here in Perth, civil engineering companies like mine will buy clean fill (generally sand, but whatever is suitable as Main Roads specification material) for around $5 per tonne. Your 1.25 million cubic metres of spoil will come to around two and a little bit million tonnes of spoil, but it will not be clean fill.
By the time you have cleaned it out, or just skimmed the cirty spoil from the re-useable material, you may only get around 75% of the material as saleable. The income may then fetch $7-8 million, but your disposal costs for the remaining 500,000 tonnes or so will easily eat up a good $12-15 per tonne. This will come to around the same cost as your income from the good fill material, so in effect will probably give you a break even on this.
On a very interesting side note, I am working on a tender right now which is located right next to a railway yard, and we were chasing one million tonnes of fill. I tried to get the railway companies to quote on the haulage for me, and they all came back to me and told me they were unable to haul those kinds of tonnages, and I should go to trucks, as they were more efficient. Yet, a two million tonne haulage contract for a mining company is hailed as a saviour for the rail system. My mind boggled…..
Anyhoo, keep up the good site here!
Cheers,
Matt
James, a well thought out post. The answer, as always, depends on what it is you want to achieve? If it is only to save a handful of minutes, then certainly, choose the most cost effective option. But, if that is the case, I am inclined to agree with Riccardo, why even bother?
The answer to that lies in the long term. I suspect that the Dandenong corridor is worth more than the existing setup and that, it is, almost certainly, not carrying what it potentially could. I’d step back from your analysis then and ask what the optimum setup would be, not just in terms of the corridor, but the whole network, and then look to plan, in steps, how to get it from where it is now, to where you’d want it to be.
Over the long term, I think the Perth optimum is a good one (certainly superior to what we have here). We should aim to have a 5km station spacing for travel around the metropolitan region AND light-rail/bus connections stopping every 400-800m in between. Closing stations is only a vote-loser if: a) the people in question vote for you anyway, in which case closing the MATH stations isn’t a problem for Labor; b) you don’t get them something better, which for Oakleigh outwards, a light-rail with walk up access is; but more importantly c) the rationale for why the changes ultimately benefit everyone isn’t properly explained. The biggest failure of the most recent plan is that it speaks vaguely about a “Metro” without ever defining what that would entail.
This isn’t a short term goal though. It needs to be planned over several decades, with each intervention into the system getting you a little closer to the optimum. Hence the need for a well produced rationale. A single intervention that is ditched by the subsequent government is worthless. A single intervention that prevents you reaching the optimum without it being unwound is worse than worthless. I have a whole post waiting to be written about that, but not now.
A well-thought out technical analysis James, good work. Just interested to take a step back and, following Riccardo and Russ’ trains of thought, ask not only “why?” but indeed “who?”
Who are we going to have to move over this corridor between the outer south-east and the city? Who is it that actually needs much faster, vastly more efficient, much greater capacity heavy-rail links from Dandenong or Pakenham to the Melbourne CBD, or even to intervening destinations like Caulfield? Not questioning that modestly improving the standards of the service in all these areas (and others) is not worthwhile for those using the service, but again, who is it for?
I don’t have data to hand for all the LGAs in the corridor, but a few seconds on Google got me the 2006 Census data for the City of Casey (which is most of the Pakenham and Cranbourne lines past Dandenong, excluding Pakenham itself) here: http://www.transport.vic.gov.au/DOI/DOIElect.nsf/$UNIDS+for+Web+Display/E3A72D73189B9410CA2573E9001E3F2E/$FILE/TDIA-Introduction-Chapter1.pdf (pg8)
According to that data, about 26% of Casey residents work within Casey. Another 24% travel only as far as the adjacent LGA of Greater Dandenong, and 10% go beyond that to the next LGA up the line, the City of Monash (likely to include a large number of University students and some staff, but in that I’m only guessing). About 12.5% travel completely counter to the spokes of the rail network, to Knox and Kingston. Only 6.5% go to the City of Melbourne each day – so why build a line, at great expense, with large express services to cater to those people?
I would humbly suggest that any money spent in the area might be better invested in developing a radial heavy and/or light rail network centered on Dandenong (Melbourne’s potential 2nd CBD, as you rightly point out), taking advantage of the existing corridors and reserves toward Frankstoon, Ringwood and the Port of Hastings. Alongside, indeed even before this, should come expansions of and improvements to local buses, and investigations of the potential usefulness of light rail/tram options in the future, with reserves set aside and so on.
Casey’s story of minority travel to the Melbourne CBD is repeated right across the metropolitan area, and to continue to pour funds into CBD-ward travel is foolish when improving local PT, cycling and walking options is so much cheaper and serves a much larger group of people. I’m not sure why our political leaders ignore these options, but Riccardo, you might say that you can’t cut ribbons on new bus stops, no?
Update: Like an idiot I didn’t look deeper into the .pdf I found. Other Outer suburban LGAs and their CBD-ward commuter populations versus their ’stay-at-home’ (i.e. in their own and adjacent LGAs) populations (in %):
Cardinia: 57/4
Casey: 60/6
Frankston: 76/6
Dandenong: 74/8
Hume: 60/15
Kingston: 67/13
Knox: 73/9
Melton: 52/17
Mornington: 76/3
Nullumbik: 50/14
Whittlesea: 63/13
Whyndam: 57/19
Yarra Ranges: 70/6
The ones with higher CBD commuter numbers might also make for an interesting study of some kind for someone who’s not meant to be doing something else
That said, obviously these numbers are descriptive, and we cannot employ them SOLELY prescriptively, as if they are set in stone. If people don’t move long distances for work, or don’t move in certain direction, it may be for social reasons, or even because of inadequate existing transport infrastructure…but I think we ignore such things at our peril. I hate to think of the day when PT advocates end up building white elephants like the last two or three Tollway projects around the country…the demand CAN be supplier induced, but isn’t necessarily so.
James,
Great post on the Dandenong line. Although, as Notch says, you’re station usage data is a bit off from my reckoning. Are you using passenger counts or validation data to order your stations? The validation data I’ve seen puts a different slant on the ranking of stations (albeit with very little variation in the top 5 and some movement in the top 10).
There’s some good spatial analysis of VISTA and ABS data at Charting Transport. The owner of that site loves analysing this stuff.
LS
@James – I think there is another “pro” you might want to add to your list.
By reducing total running time you also reduce the cost of providing the service. If you run 6tph (10 min headway), and you shave 5 min off the running time in each direction, one less train set is required to operate the same level of service.
I don’t know what the numbers are for Melbourne, but based on Sydney I’ll have a stab at $2mil/yr/train. I think it’s reasonable to present value these ongoing savings and claim them as a cost saving. Very rubbery, but say $25mil saving for 1 train set, and $50mil for 2?
these future savingsI know this isn’t enough to make a business case for the proposal, but I think you could reasonably price
Again, i’m sorry about the late replies.
As for the spoil – Bulbous, that was prettymuch the costs i was guessing for the spoil sale – it was good to get an actual figure. What i was meaning there was that with the waste – the unsaleable spoil, the disposal costs are generally significant. If it was possible to put this spoil into an already degraded environment (the open cut coal mines for example) a significant cost would be removed. IIRC – the disposal of spoil is one of the highest costs in any excavation/tunnelling work. – If you have any figures that would help there, it would be appreciated.
I agree that continuing the domainance of Melbourne CBD for the rail network is not the best option – rather a polycentric city (somewhat like Sydney) would be a better outcome, as you have bi-directional flows of people, and more office jobs located closer to where people live. However Melbourne for jobs is actually more decentralised than Sydney – Its just that Melbourne doesn’t have the subregional cities that Sydney does. A significant proportion of the jobs occur in the Monash corridor (all those industrial, and semi-commercial uses along the Monash Fwy/Mulgrave areas). That is in part why Melbourne dispite having fewer cars than Sydney, has a much higher VKT’s.
A proper bus network (Smartbus frequencies at a minimum) along the key arterials, would improve PT use in Melbourne.
The figures i got where given to me by someone i know. I did that very quickly so i might have put them in wrongly – given that it was a text file – very different data capture compared to Sydney’s counts.
Thanks for the charting link Loose Shunter – looks interesting.
I’ll have a followup post on this shortly, looking at the entire system.
James,
Good post, especially from an occasional interstate user.
An interesting note is that $2m was allocated in this May state budget to move the rail planning for the corridor along. Maybe this post is timely.
With the just approved urban UGB expansion and rapid development along the corridor, the numbers of commuters will continue to increase two new stations will have sod-turning photo opps later this year to commence construction. The usage is growing quite rapidly out there.
Anecdotally I reckon Narre Warren and Hallam have nearly triple the loading they had 20 years ago as they developed, and the new growth areas promise similar sort of usage. Potentially there will be 2 Canberra’s east of Dandenong in the rail catchments by 2040.
The 10 level crossings from Grange Road to Webster St are a major challenge, particularly as there are 6 Major activity centres and the “2nd city of Dandenong” in close proximity to the line where even more intensive housing is to be developed. Check out the recent Planning Panel Victoria hearings on developments on Poath Road beside Hughesdale station.
TOD Intensification is good for passenger catchments, but it also creates more NIMBYs.
As usual, any engineering and planning for removing level crossings will be justified by “congestion removal” for the 6 Arterial Roads a la Springvale Road Nunawading and Middleborough Rd, not rail efficiency. VicRoads will probably end up delivering the project rather than a rail agency like VicTrack.
Trenching might be efficient for operational separation, but much of the line was created there in the 1870s because it followed the top of the great Gippsland swamp. Drainage would be a major consideration for rail and/or the roads to be grade-separated from the rail.
I was suprised to see Brumby emphasising the complexity of grade seperation (re: Hoddle Street) for impacts on other utilities: water, electricity, telecommunications, sewerage, drainage and there is also a (the?) gas main along the netire length of this Dandenong corridor
Another primary future use of this line is for freight to the Port of Hastings once Port of Melbourne “exceeds” capacity in 15-25 years to replace the Frankston Line connection. That has 30 level crossings, is built on old sand dunes for 15 km s along the beach and a number of other limitations. If the new Webb Dock Rail is dismissed later this year, the Hastings development could be accelerated, with flow-on consequences for the Dandenong Line.
Most suburban stations had freight “goods handling facilities” in the 1969 Melbourne Transport Plan (vol 1, p.52a), mostly car parks since the 1980s. A number of 1950s-1960s industrial zones also had private sidings, some of which are still connected. General Motors station anyone?
The commuter focus on urban rail is a recent concept.
An intermodal freight terminus is being developed in Sth Dandenong which could increase rail freight use of the line before the 2035 Hastings expansion. Interestingly there are passing lines for freight/country services at Caufield, Oakleigh and Dandenong stations between the suburban tracks.
I intend to respond more about this, but babies and blogging don’t mix at the moment. Sorry if it wanders a bit.