Union Rd, Surrey Hills (via Philip Mallis) |
The government recently announced they'd be grade separating ten more level crossings, bringing it to a total of 85 pledged and/or delivered under the Andrews government. They also announced they'd be closing four crossings, so it's sort of 89, but their messaging consistently says 85 so let's stick with that. As Daniel Bowen pointed out on his blog, this means about half the level crossings in suburban Melbourne on the 2008 ALCAM list will have been eliminated - very impressive.
This also means that we're reaching the point where some lines will be completely level crossing-free by the time all the pledged works are completed. This is something observers had noticed quite early on - crossings on the MM1 corridor often seemed to be prioritised - and it's a feature of LXRA's prioritisation framework developed before the 2018 election. But for the first time that I'm aware of, their promotional material explicitly mentions this - noting that the Lilydale and Cranbourne/Pakenham lines will be crossing-free, and highlighting the progress made on the Frankston, Belgrave and Sunbury lines in their graphics.
Level crossing removals, with crossing-free corridors in yellow (via LXRA) |
The phrasing used in the prioritisation framework is "corridor completion". Doing a whole line is obviously the end goal, but doing a meaningful section of a line may also be beneficial. For example, the Cranbourne and Pakenham lines join at Dandenong, and there's also short-starter trains commencing at Westall in the peaks - so being completely crossing-free to Dandenong was a meaningful achievement even before they announced the final crossings on the Pakenham line. In fact, just getting to the turnback at Westall was.
The level crossing removal program has clearly been a winner for the Andrews government, so it's odds-on they'll announce more crossings in the leadup to the 2022 election. They seem to like big round numbers, so I'm guessing they'll announce 15 more to bring it to a round 100. Given these assumptions, and what we know about the prioritisation framework and the remaining crossings, which 15 are they likely to be? And what does that mean for completing crossing-free corridors?
Taking the 2008 ALCAM list of crossings ranked by danger, and taking out the ones that have already been removed (or are pledged or under construction) gives us the below - the most dangerous level crossings in the state. (1)
2008 rank | Location | Road | Suburb | Risk score | Line | 2021 rank |
---|---|---|---|---|---|---|
9 | Metro | Macaulay Rd | Kensington | 3,686.73 | Craigieburn | 1 |
21 | Metro | Riversdale Rd | Camberwell | 2,882.48 | Alamein | 2 |
24 | Metro | Station St | Fairfield | 2,680.65 | Hurstbridge | 3 |
39 | Metro | Tooronga Rd | Malvern | 2,042.40 | Glen Waverley | 4 |
44 | Metro | Gaffney St | Coburg North | 1,869.75 | Upfield | 5 |
47 | Metro | Maidstone St | Altona | 1,839.90 | Werribee | 6 |
51 | Metro | South Rd | Brighton | 1,721.93 | Sandringham | 7 |
53 | Metro | High St | Glen Iris | 1,674.00 | Glen Waverley | 8 |
58 | Metro | Puckle St | Moonee Ponds | 1,559.52 | Craigieburn | 9 |
61 | Metro | Prospect Hill Rd | Riversdale | 1,495.72 | Alamein | 10 |
62 | Metro | Normanby Av | Thornbury | 1,486.49 | Mernda | 11 |
63 | Metro | Box Forest Rd | Glenroy | 1,481.30 | Upfield | 12 |
65 | Metro | Bunswick Rd | Brunswick | 1,414.85 | Upfield | 13 |
66 | Metro | Hampton St | Hampton | 1,398.54 | Sandringham | 14 |
70 | Light rail | Bridport St | South Melbourne | 1,355.85 | 96 tram | 15 |
71 | Metro | Glenferrie Rd | Kooyong | 1,317.60 | Glen Waverley | 16 |
73 | Metro | Hudsons Rd | Spotswood | 1,308.99 | Werribee | 17 |
76 | Metro | Madden Gv | Burnley | 1,240.17 | Glen Waverley | 18 |
77 | Metro | Macaulay Rd | North Melbourne | 1,235.52 | Upfield | 19 |
78 | Metro | Glen Eira Rd | Ripponlea | 1,212.07 | Sandringham | 20 |
For whatever reason, this government hasn't made any attempt at crossing removals on the Sandringham line, and the "Light Rail" crossings seem to be out of scope too, so I've removed these from consideration.
Grade separating the top fifteen crossings listed above would complete the whole Alamein line AND the whole Glen Waverley line, as well as eliminating the only remaining tram squares in Melbourne (a "very high" priority elsewhere in the framework). It would also eliminate the innermost two crossings on the Craigieburn line, but unfortunately this doesn't get you to a turnback so it wouldn't complete a corridor. Lastly, it would eliminate one crossing each on the Werribee, Mernda and Hurstbridge lines, and three on the Upfield line - none of which help with corridor completion.
What about approaching it from the other angle - picking the 15 most-useful crossings for corridor completion, and seeing what their danger rankings are like?
The black dots are unpledged crossings (via Google Maps (2)) |
As Daniel mentioned in his post, the two left on the Sunbury line are very quiet, so they're hard to justify in the short term. The Hurstbridge, Mernda and Upfield lines all have tons of crossings left, and there's not much low-hanging fruit from a corridor completion perspective - except maybe Ramsden St in Clifton Hill, on the shared Hurstbridge/Mernda section, which is #61.
The remaining crossings on the Alamein and Glen Waverley lines are all in the top 20, so they're clearly justifiable from either angle.
The two remaining crossings on the Belgrave line, Hilltop Rd and Alpine St in Ferntree Gully, are ranked #83 and #101 respectively, so perhaps a little hard to justify on safety grounds alone. However, they bookend the innermost section of single track on the Belgrave line, so packaging crossing removals with duplication would be a worthwhile project. And when combined with the Alamein/Glen Waverley works, this would make the whole Burnley Group crossing-free, which could have implications for future purchases of driverless trains etc.
The tram square at Glenferrie Rd, Kooyong on the Glen Waverley line (via Liamdavies) |
The innermost two crossings on the Frankston line, Highett Rd and Wickham Rd in Highett, would get you to Cheltenham, where a lot of extra peak trains originate - they're #22 and #63 respectively, and are sufficiently close together to be packaged in the same project. So it's probably justifiable to do one as low-ranking as Wickham Rd alongside Highett Rd.
The direct Werribee line only has five crossings left in total. On current operations, the innermost two (Hudsons Rd, Spotswood and Anderson St, Yarraville) are #17 and #36, so fairly justifiable - this would get you to Newport, where the Williamstown line splits off, crossing-free. But if the plan is ultimately to build MM2 and send Werribee trains via Fishermans Bend, then the other three crossings are the relevant ones - Maidstone St, Altona (#6), and Champion Rd (#30) and Maddox Rd (#52) in Newport, the latter two of which are close enough together to be packaged.
The Craigieburn line also only has five crossings left in total. Almost all of them rank pretty high, and Park St, Moonee Ponds (the lowest one at #54) is about 600m from Puckle St (#9) so might need to be packaged with it. You'd need to do all five to, in any sense, complete a corridor - three would get you to Essendon station, which has a turnback, but it isn't in regular use for short-starters, it's just used in disruptions.
Trains only stop at Platform 1 at Essendon when there's disruptions |
Another big consideration is cost. The budget has been stretched by COVID-19, and crossing removals only seem to be getting more expensive, so I'm sure they'll prefer cheap and easy crossings to expensive ones. I won't go through the full list, but to give some examples - Maidstone St, on the direct Werribee line, is in an industrial area with few obstacles around it, so it'll probably be cheap to do as a road bridge, whereas the two crossings in Spotswood and Yarraville are right next to their respective stations, as well as a bunch of shops and houses - probably very expensive. Likewise, Macaulay Rd, Kensington is by far the most dangerous crossing on the list and has been for ages - it was #9 even in 2008 - but it's a very difficult site, so I wouldn't be surprised if it gets passed over again.
With all this in mind, which projects should be selected? Here's my top fifteen, with ten backups I wouldn't be surprised or disappointed if they chose.
2008 rank | Location | Road | Suburb | Risk score | Line | 2021 rank |
---|---|---|---|---|---|---|
21 | Metro | Riversdale Rd | Camberwell | 2,882.48 | Alamein | 2 |
61 | Metro | Prospect Hill Rd | Riversdale | 1,495.72 | Alamein | 10 |
39 | Metro | Tooronga Rd | Malvern | 2,042.40 | Glen Waverley | 4 |
53 | Metro | High St | Glen Iris | 1,674.00 | Glen Waverley | 8 |
71 | Metro | Glenferrie Rd | Kooyong | 1,317.60 | Glen Waverley | 16 |
76 | Metro | Madden Gv | Burnley | 1,240.17 | Glen Waverley | 18 |
86 | Metro | Highett Rd | Highett | 1,006.54 | Frankston | 22 |
86 | Metro | Wickham Rd | Highett | 252.25 | Frankston | 63 |
176 | Metro | Hilltop Rd | Upper FTG | 159.09 | Belgrave | 83 |
197 | Metro | Alpine St | Ferntree Gully | 101.02 | Belgrave | 101 |
9 | Metro | Macaulay Rd | Kensington | 3,686.73 | Craigieburn | 1 |
58 | Metro | Puckle St | Moonee Ponds | 1,559.52 | Craigieburn | 9 |
89 | Metro | Devon Rd | Kensington | 976.18 | Craigieburn | 23 |
107 | Metro | Gaffney St | Kensington | 670.32 | Craigieburn | 37 |
134 | Metro | Park St | Moonee Ponds | 381.67 | Craigieburn | 54 |
47 | Metro | Maidstone St | Altona | 1,839.90 | Werribee | 6 |
162 | Non-Metro | Hopkins Rd | Truganina | 210.24 | Melton | 70 |
62 | Metro | Normanby Av | Thornbury | 1,486.49 | Mernda | 11 |
173 | Metro | Woolton Av | Thornbury | 168.22 | Mernda | 80 |
24 | Metro | Station St | Fairfield | 2,680.65 | Hurstbridge | 3 |
65 | Metro | Brunswick Rd | Brunswick | 1,414.85 | Upfield | 13 |
103 | Metro | Park St | Parkville | 732.25 | Upfield | 34 |
157 | Metro | Union St | Brunswick | 228.85 | Upfield | 67 |
95 | Metro | Station St | Seaford | 825.25 | Frankston | 27 |
100 | Metro | Armstrongs Rd | Seaford | 750.36 | Frankston | 31 |
Which ones would be your priorities?
1. Using the 2008 ALCAM list and just getting rid of the removed crossings doesn't factor in changes to train or road traffic levels since 2008, or changes to crossing design (since approaching the tracks at a better angle can help visibility etc). So this mightn't be 100% as accurate as the 2021 data the government will be using, but should hopefully be fairly representative. The main exception to this is the Melton line, which has seen huge increases in both rail and road traffic, so these crossings probably rank higher than they did in 2008, relative to other unpledged crossings.
2. Several years ago, I came across a map someone else made that showed the status of all level crossings in the state. It was saved to my own Google profile and I've subsequently modified it to add missing single-track sections, add 'crossing free' sections, and update it as new projects were announced and old projects progressed - but the base map and the level crossing icon thingies are from the original map. I have no idea who made it, so please let me know if it was you and I'll happily credit you.
You mention the MM2 for Werribee trains and three crossings to remove. Could/would Ramsden St Clifton Hill be removed as part of the MM2 project?
ReplyDeletePrecisely how MM2 will interface with the existing Mernda line is a bit of an unanswered question at this stage. It's likely they'd send it underground before Clifton Hill though, so it'd probably make Ramsden St less urgent.
DeleteIs there a plan to extend Upfield to Craigieburn to allow VLine trains to use the Upfield line to bypass Essendon? I thought that was the idea, to provide for more short-starters? That would mean Park St (and of course, Macaulay Rd and Puckle St) would probably have a higher priority than Devon Rd and Gaffney St.
ReplyDeleteI think it's more likely that with the MARL upgrades to the Albion-Jacana link that Seymour trains go via there, would allow at least extra paths from Broadmeadows into the city if done properly.
DeleteSending V/Line via Upfield has certainly been talked about, but I understood it was for more capacity along the whole Broadmeadows-City section rather than short starters. In any case, the plans to do this have never progressed much so I don't think it's likely in the next few years.
DeleteNot only would Highett+Wickham get you to turnback at Cheltenham but it would also get you to Mordialloc.
ReplyDeleteRivesdale LXR could also involve combining Riversdale with Willison stations (then extend to Chadstone but that’s another story!)
they can't merge Willison and Riversdale together between the to though, as the rules in the book don't allow curved station platforms to be installed anymore
DeleteWould probably make more sense to close willison and move riversdale station further south towards riversdale road/under. That way its still straight platforms and a path could be included from the former willison to riversdale.
DeleteBy "merge", I assume Louis meant close Wilison and move Riversdale to go up to or over Riversdale Rd to be as central as possible to the two original stations.
DeleteAccording to my data, there would still be three crossings impacting Mordialloc short-starters - Latrobe St Mentone, McDonald St Mordialloc, and Bear St Mordialloc (which is between the station and its stabling yard).
Delete@john.s : yep ;)
Delete@simbara : Latrobe! True!
The assessment of the Williamstown line has forgotten about the Giffard St crossing near the end of the line.
ReplyDeleteRamsden St at Clifton Hill could be a contender for just being closed. There is enough nearby alternatives.
To clarify, I was saying that removing Hudsons Rd and Anderson St would remove all the crossings on the tracks between the city and Newport, which the Werribee and Williamstown lines share with each other (not that the whole Williamstown line would be complete).
Delete