Melbourne Rail Network
As most of you have probably guessed my interests in transport lie primarily within the optimisation and efficiency side rather than the more photography side of transport.
This post is mostly a very quick look at optimisation of the Melbourne Rail Network.
As we know, the current setup has approximately 16 lines converging onto four single track loops into the CBD, which creates a signficant bottleneck, reduces operational efficiency and can lead to cascade delays very rapidly. (a cascade delay is where one train runs late, causing every train behind it to run progressively later, and can really only be stopped by cancellation and transpositioning trains).
This proposal breaks the four Melbourne sectors into 7 sectors, (8 including interurban/intercity) and increasing frequency, using predominately the existing rollingstock and only minor additional infrastructure. This is shown in the image below
This proposal assumes the following: (figures from VicSig)
To do this i’ve had to make a sample/hypothetical timetable for the groups which probably have some mistakes within them, and i freely admit this) – to show that the proposal is vaugely workable. Each will be discussed very briefly below.
Upfield Line – 8tph – 10 trains required.
two new crossovers at Gowrie to enable trains to Upfield to bypass Stopper
Cost $2 million.
Sydenham Line – 12tph – 14trains required
Upgrade to Sydenham Station turnback facilities – 5million (Enable fast shunts into yard to minimise blocking of main line for Bendigo/Kyenton trains)
Seymour Line – 7 trains required. 1tph 2tph peaks – No upgrades
Werribee-Sandringham Line – 16 trains required – 6tph
Cragieburn – Frankston – 22 trains required – 6tph
Upgrade to Broadmeadows Line Flyover – Indicative cost 10million. (to enable Cragieburn trains to enter track 5-6 without interfering with Werribee trains.
Ringwood Group – 32 trains required – 6tph (3 stopping patterns)
Increase dual track for an additional 1000m south of Lilydale Station to enable 10minute service – 10 million.
Pakenham Line – 34 trains required. (6tph (3 stopping patterns)
4th Platform at Dandenong – 10million
3rd Platform at Pakenham 10 million
Central Westall Turnback (Under Construction?)
Glen Waverley Line – No upgrades – 6tph
Geelong and Newport Shuttles 6tph (Geelong/Williamstown) 3tph (Altona)
3rd Platform at Newport (Preferably 4 however) 10 million
Clifton Hill Group – 12tph Greensborough, 6tph Epping, 1tph Eltham, 1tph Hurstbridge
Dock Platform at Greensborough – 10million
Bendigo Line – No Upgrades 2tph Kyenton/Bendigo
Belgrave Line – 6tph
Additional Points at Ringwood to enable platform 3 to be separated from running lines 2million
So for 70 million in additional infrastructure you can separate out the lines and increase frequency on all lines (when additional rollingstock is ordered), increase reliability with the exisitng rollingstock (assuming Vicsig’s numbers are correct). Though that said some of the timings used are very tight, and probably wouldnt be feasible in the real world, but as a discussion point, is usable.
This shows that there can be an increase in services operating without the Eddington Tunnel, with predominately existing infrastructure. The Sandringham/Frankston groups could recieve a doubling of service without any additional infrastructure, and the Clifton Hill Group requires duplication to Epping for increased services.
Thoughts?
Dunno if Lilydale requires a 10″ service?
Not sure what you mean by “South of Lilydale station”.
High ‘Notch
Lilydale just makes a convienient turnback location nothing more. If the single track was any longer you’d only run the intensive service to Mooroolbark and peak fleeting through to Lilydale.
I’ve updated it slightly so the wording makes more sense for the dual track extension to Lilydale.
I assume you plan to do away with the flip flopping of directions on the city loop and let people transfer instead?
Which direction would you run each loop in? Personally I think running the Caulfield and Burnley loops in opposite directions would be wise, so people can simply transfer at Richmond for the fastest trip to suit their needs.
It really shits me some times when I’m at Parliament and trying to get to Richmond direct. Something like 30 trains an hour and every single one of them goes the long way around.
I actually had all loops (bar Clifton Hill) Running via City Loop, as they all had contra lines that ran direct – The Northern Group has the Werribee/Cragieburn, the Caulfield Group has the Frankston Line and The Burnley Loop has the Glen Waverley Line.
The Loop makes it a real pain to do the short distance hops like those. I was only thinking about the medium to longer distance journeys when i made them, and as an ‘outsider’ to Melbourne (From Sydney), i find the Loops operations now difficult enough to understand, especially that blasted reversal at midday – i always end up going to the wrong station and go the long way around.
As a melburnian it still catches me out at times, especially making connections between lines and to trams.
The contra line thing only works for the two ‘vaiduct’ stations, it’s not much use in the three tunnel stations. However the answer is relatively simple I think, run two tubes clockwise, two anti, and send the caulfield and burnley loops in opposite directions.
So like this:
Northern clockwise
Clifton hill anticlockwise
Caulfield clockwise
Burnley anticlockwise (these last two could swap, whatever is best with the trackwork)
This would require the long lost flyover at the Clifton hill portal to avoid the conflicting movement.
The benefit here is that anyone using the south eastern half of the network can transfer simply at Richmond to go in the most direct loop direction, and (pehaps more importantly) they can do the same heading back from any loop station. Running those two the same way basically guarantees you have to go the long way round one per return trip.
Running the northern and Clifton in opposite directions achieves much the same, but using southern cross or flinders as the transfer point.
Really think you’re overservicing past Ringwood. As much as I don’t understand some of the odd frequencies out that way (30″ off peak and morning counter-peak is ridiculous IMHO), 10″ is too much based on passenger numbers and actual population.
James, I think this is good, as a starting point for where inefficiencies might lie. A quibble though, unless I am misreading you, you have 2 lines too many between Flinders and Spencer St. Or rather, to put it another way: the real viaducts have two lines too few.
Excellent stuff.
Agree with the shuttling -though my position is that the “Lilydale” line is really the Mooroolbark line and as others have said, a bit of peak fleeting might be plenty, and a shuttle at other times.
James, you might as well add South Morass and the double track to your map, apparently the contract has been let.
I’ll look over the analysis and timings when I have more time.
He only has those two lines running at a total of 12tphpd, so the two viaduct-only tracks should be able to handle that just fine.
If you think about it, with four tracks on the city loop and two extras on the viaduct, the viaduct-only lines should carry one third of all services (excluding any that terminate at Flinders St directly).
Nick, fair point I was thinking graphically and conceptually, more-so than the specific proposal. The Sandringham and Frankston groups are actually the same group, by virtue of having to share the two viaduct-only tracks. Which, needless to say, puts constraints on both groups that don’t apply to the others.
Great analysis. Most of these ideas have great potential. Don’t you think you’ve screwed the outer end of the Hurstbridge Line a bit much though? I’m not going to talk it up and suggest the area is flooding with potential patronage but cutting back services from what they are currently seems a bit bizarre and would only drive people away from the service making it even less viable. Eltham deserves a train every 20 minutes in the peak. Take a look at the loadings on the express services each morning and evening. There is most certainly demand. I personally catch the train from Greensborough and know from daily experience that getting a seat on the morning express services is rare. Almost all are occupied before Greensborough. I realize standing isn’t unusual but cutting back services to every half an hour seems reckless when Eltham / Montmorency is generating such loadings.
Hi All,
I apologise about hte delay in replying.
Some points: In part i wanted to make this realistic, in using the amount of trains that exists now (1200 cars), which is why there is some odd frequencies, such as only 6tph Sandringham/Frankston, and the Hurstbridge shuttle.
Originally the Sandringham line was seperate as it is now, and the Werribee line trains acted as the stopper service to Cheltenham, and the Cragieburns were express services Caulfield – Cheltenham. Removing this saved 11 trains (66 cars). The same occured with the Hurstbridge Line where the Shuttle was cut back to half hourly, instead of every 15min to Eltham, which saved 2 trains (12 cars). That added up to 84 cars, or 14 new 6 car trainsets (10% maintenence requirements.
Adding in South Morang requires a minimum of an additional 4 trains, bringing up the reuqirement of 114 cars or 19 trains
At the current costing of new trains for Melbourne being 5.5 million per car or 32.5 million a train, this reduced the overall cost by 618million.
It is why there was those strange compromises.
If the Comeng trains were retrofitted to have longditudal only seating (See Riccardos current post) they would have an average of 144m2 floorspace per trainset (once seats are taken out) (24mx2m) per three car set, which would provide for approximately 576 standees (at 4ppl/m2 – which is standard capacity) and 200 seats (rough estimate 50cm per seat, and 6m per three door sets), would provide the same amount of (standard) capacity as the existing 6 car sets (776 people).
This would enable certain lines (Upfield, Epping, Sandringringham) to be run as three car sets, enabling the full possible timetable to be run.
In the end when the need for services to Werribee and Cragieburn exceed 12tph (each), the next viaduct track could be built, enabling the Werribee – Sandringham line to have exclusive tracks. (That said i much prefer BleakCity’s Docklands tunnel)
Chris good point although you’ve rebutted your own to some extent.
If you are getting on at Greensborough and the seats are gone, that tells me that Greensborough needs a high frequency, reliable service, and taking trains away from further out to boost that capacity would be the go.
I would cut the Melbourne rail system at Greensborough, and offer Nillumbik Sh the opportunity to use a small railcar on the line beyond as a shuttle, subsidised if they wish from local rates. They might find deelectrifying would save them money.
At the end of the day, running suburban rail services at considerable expense to the taxpayer is because of congestion, and I’m not convinced there is such heavy congestion that rail is required beyond Greensborough.
Using a dock platform at Greensborough and a shuttle would mean that any delays from the shuttle are not transmitted to the main system, and the need for heavy, high capacity stock between Greensborough and the city is not wasted beyond Greensborough.
I would differ from James proposal in that at the moment, there are spots for 9 sets IIRC at Eltham, I would ‘fleet’ those onto the system, especially if an extra couple of runs were inserted onto the main system anyway. Fleeting means no waiting for crosses, and any passengers in the reverse direction would be put on a bus during the fleeting.
So for example if Greensborough was running 6ph from 5am to 6:30, but then another 2ph were added, you’d bring them from Eltham sidings, give Eltham and Montmorency pax a couple of direct services, but with no waiting because there would be no contra-peak movement, any actual pax put on a bus. Then at say 9:30 you’d need that bus again, because unfortunately those extras need to be put away in the sidings, and they will clash with a shoulder peak directional movement. But any dangers of delays ricocheting at this point are low, as you are coming OUT of peak, not into.
Down the track, I could imagine double track to Eltham, but not to Hurstbridge. This latter did not receive electrification or suburban trains because it deserved it, but to save the use of a steam loco for such a small length of track in the 1920s. The suburban area has crept out there because of this fortuitously generous extension of wire, misleading the public that a suburban service was available when it wasn’t. If it had the DERM like Healesville or Mornington it would be gone by now.
So I think offering them a railcar and the chance to keep the rail line at all is generous. And much has been concrete sleepered now, so I don’t see much benefit in ripping it up. But simplifying the signalling, removing the wire and disconnecting it from the main system might be a cost saving that would enable the locals to support it.
As I was saying, at the end of the day a heavy rail system costs hundreds of millions a year in subsidies, money best spent on those that actually need the advantage of that technology, the highly congested inner and middle suburbs
James, I would go for Lilydale under the same arrangement as outlined for Greensborough. A shuttle from Lilydale to Mooroolbark, plus fleeting out of the sidings. Australians are understandably delicate about shuttles because of long standing unreliability of connections; but it is chicken and egg. Until we get the main system reliable, by breaking all the constraints around passing on single track, swinging points and reseting junction signals, we will not be able to improve reliability.
Sydney has seen the benefit of this although they too have a very long way to go before traffics are physically isolated from each other.
But at the moment a failed cross at Mooroolbark can impact on services 1 hour later at FSS, and potentially 2 hours later at Upwey or Ashburton. And once that pattern sets in it can ruin the system all day.
Whereas if we accept that the Lilydale shuttle missed a connection at Mooroolbark so they caught the next one, 10 minutes later, but no repercussions at Upwey or Ashburton as a result, a much better arrangement.
Thanks for that Riccardo -
Though with Lilydale when you think about it, its what 4 kilometres from Mooroolbark, through relatively benign conditions especially in comparison to the Belgrave Line – It’s probably just easier to duplicate the section and be done with it.
Hurstbridge is much harder to justify upgrades to due to the heritage wooden viaduct, and then the very low population base past Eltham. I’ve personally wondered whether the SmartBus if it operated fast to Greensborough rather than how it operates now, would actually be faster than the train, given the circuitous nature of the line.
Belgrave is similar past Upper Ferntree Gully, and would most likely be impossible to duplicate through Upwey.
The main reason for why i havent shown fleeting is that it requires a full timetable to be developed, and all i’ve done there is a quick snippet to find out how many trains would be required.
That said i would be interested in doing an integrated transport solution for Melbourne, but lack the data – especially GIS data like roads. If i had that sort of data i could probably do a similar proposal to what i did earlier this year for Sydney. Given that the bus network is so poor in Melbourne, and with targetted street-car and modification of existing street-car lines into Light Rail, Melbourne has massive potential for upgraded services, and would easily be the transit capital of Oceania, with relatively little work.
James,
I have some experience in regards to timetabling of trains, and I have to say I am very impressed with the quality of the work you have produced.
So impressed in fact, that I am running it through my train graph simulators – a very close representation of the tools that Metro uses.
I have started with the Burnley timetable that you have created, and there are only a few issues that need adjustment in order to make the timetable work.
First, it appears you do not have data on train headways, or the minimum time between services due to signalling and safeworking. In the Ringwood-Lilydale section the minimum is 4 minutes; because of this I have adjusted the timetable to suit. Because of this adjustment I found clashes at the city end as well, and to fix this I had to alter the timetable so that both the Mooroolbark and Lilydale tiers stop all stations to Camberwell, rather than the latter running express.
Also, it appears you were not made aware of minimum layover times at Flinders Street station – all trains require a stop of at least 2 minutes here, to allow for driver changeover and the larger volume of passengers.
Next, in regards to the Alamein line services; the adjustments I made translated to a 2-minute layover at Alamein station, which means that you would need compulsory drop-on-drivers at that location; the only other option with the current infrastructure is to run only half the trains to Alamein, and terminate the remainder at Ashburton with a Shunt-and-Redock procedure, i.e. the Ashburton train would shunt to the siding to let the Alamein through, then return to the platform to form the next up train with the Alamein service following it. In the medium term another option would be to save time by rebuilding the Alamein line, with stations only at Riversdale Road, Toorak Road and High Street Road, as [PLUG] seen in the recently released Smart Passengers Inc. publication.[/PLUG]
The final adjustment necessary was in regards to the single track from Mooroolbark to Lilydale; assuming you have not truncated those services at Mooroolbark, the section cannot under current conditions maintain a service of 6tph; 4 is the maximum under normal conditions and 5 with a bit of luck. However, adding Cave Hill station (as a crossing loop with simultaneous arrivals as per Upwey) fixes this problem. Full duplication is not necessary.
I have also assumed that all Camberwell-and-beyond trains are using Platform 2 exclusively at Flinders Street, and Glen Waverley trains are using Platform 3 exclusively. And, it is necessary for up Lilydale services to use the centre track between Camberwell and Burnley (Box Hill to Camberwell is indiscriminate for stopping trains, although up expresses run faster on the centre track).
The end result:
· Up Lilydale (63min) forms down Alamein* (25min) – See below
· Up Alamein* (35min) forms down Lilydale (57min) – Requires 20 trains
· Up Mooroolbark (60min) forms down Mooroolbark (52min) – Requires 12 trains
· Up Belgrave^ (25min) forms down Belgrave^ (27min) – Requires 7 trains
· Up Glen Waverley (35min) forms down Glen Waverley (36min) – Requires 8 trains
*2min less if cut back to Ashburton, no change to number of trains
^1min layover added at Upwey and Upper Ferntree Gully in both directions to make crosses more smooth.
Dwell times at termini are as below:
· Flinders Street (Camberwell services): 2min#
· Flinders Street (Glen Waverley services): 4min#
· Ringwood (Belgrave services): 8min
· Belgrave: 10min (offset)
· Alamein: 2min# (alternatively 12min ALM & 14min ASH)
· Lilydale: 14min (offset)
· Mooroolbark: 6min
#Requires driver changeover
Total trains required are 47, against your original estimate of 48. If we assume that only 85% of the fleet will be in use at any time, that means 55 trains will be consumed by the Burnley Group. (For reference the current value is around 94% in peak hour, the ideal is 90% and 85% is a luxury.)
Edit: Of those 47 trains, 8.5 stable at Lilydale, 3 at Belgrave, 7* at Upper Ferntree Gully, 2 at Bayswater, 12.5 at Ringwood, 4 at Camberwell, 5 at Glen Waverley and 5 at Burnley Sidings. That leaves 5 sidings at Bayswater and 1 at Burnley Sidings for maintenance, and 1 siding each at Mitcham, Ashburton and Burnley for other purposes.
However, that means that any extra trains required for peak hour services would need to be sourced from Epping or similar locations.
*Upper Ferntree Gully currently only stables 6 trains, but C Siding is being upgraded with fencing and lighting to make it suitable for the 7th train.
EDIT 2: On further simulation, it turns out I was wrong about the viability of a Shunt and Redock procedure at Ashburton; if the siding were facing in the down direction it would be possible, but under current circumstances it cannot be done because of the extra time required for a driver to change ends three times.
It may be possible to work the system using Competent Employees, who basically work as a lookout from the back cab of a train that is reversing. However, they communicate to the driver using Guard’s Bells, a feature not provided on Siemens or either series of Xtrapolis trains. But even allowing for all that, the timetable would be so tight that even half a minute of lateness at Alamein or Ashburton could throw out the entire sequence of trains at Flinders Street, and take hours to recover – at the same time becoming a perpetual latenss.
Because of that, I will amend the above; all trains will need to run to Alamein, and drop-on-drivers are the only option. Any train running late at Ashburton would need to be truncated there and run back to the city, using a spare driver stationed there (the driver of the late running train would replace the new driver as the Ashburton Spare).
Further to my last post(s), I have now analysed your Clifton Hill timetable.
I have started by adding the South Morang extension, so all figures/data below include that automatically.
Aside from dwell time at Flinders Street, the only real issue I can find is the single track between Heidelberg and Rosanna. This section of track requires at least 4 minutes between the departure of a down from Heidelberg, and the arrival of the next up train.
Because of this, it is practically impossible to support a 12tph service beyond Heidelberg. But I am wondering, why do you provide such an intensive service? The patronage is not there at all except during peak hours, and in that case you’d fleet trains from Hurstbridge to the city, then run empty to Epping Yard where they could remain until the evening peak where the reverse procedure would apply.
In regards to off-peak services beyond Greensborough, I think this would be the perfect place to use the Sprinters when/if they are displaced by extra Vlocities – just like your other shuttles (aside from Belgrave) and the Stony Point line. It will save sparks, and Sprinters are generally nicer to ride in.
Because a sprinter only has a capacity of around 90 seated passengers, there would be a minimum service of 3tph to Eltham, 1 continuing to Hurstbridge (although Hurstbridge-Eltham may soon outstrip that capacity). A better option may be to run 3tph to Diamond Creek with 1 of those extended to Hurstbridge, this method ensuring that all capacity is distributed with the least wastage. However, this would only work if around 5% of passengers from Hurstbridge got off at Eltham; otherwise those services would become too crowded.
Of course you’d leave the wires up to allow for peak hour fleeting, and during that time the three Sprinters would be coupled to run Hurstbridge shuttles.
So assuming 6tph to both South Morang and Greensborough, you’d need 10 trains to run each service. Unlike the Burnley timetable, South Morang trains would only form South Morang, and Greensborough trains would only form Greensborough. Beyond there, you would need three sprinters – one running nothing but Greensborough-Diamond Creek services, and the other two running alternating Diamond Creek and Hurstbridge services.
The 20 sparks could be sourced from Epping and Eltham (once both receive their upgrades; Epping is supposed to gain 3 sidings as part of the South Morang Project, and Eltham is being increased to 6). That leaves 3 trains at Macleod and 5 at Hurstbridge for fleeting in the morning and evening peaks, and/or as standby trains.
The use of 3 Sprinters beyond Greensborough would leave 14 for the remaining services, such as Cranbourne, Altona, Williamstown, Stony Point, a few in the workshop at any time and the remainder as spares for regular V/line services.
The end result:
· Up South Morang (43min) forms down South Morang (46min) – Requires 10 trains
· Up Greensborough (41min) forms down Greensborough (45min) – Requires 10 trains
· Hurstbridge^ (24min both ways) & Diamond Creek^ (16min both ways) shuttles – Requires 3 Sprinters
^I have added 2 minutes for all trains passing through Eltham, to allow for trains to cross.
Dwell times at termini are as below:
· Flinders Street: 2min#
· South Morang: 8min
· Greensborough (Citybound): 12min
· Greensborough (Shuttles): 4min
· Diamond Creek: Alternating 4min/24min
· Hurstbridge: 36min
#Requires driver changeover
Incidentally, the dock platform that you suggested for Greensborough station isn’t strictly necessary. If you want cross-platform interchange with three minutes between services, then you do need it. But if you are willing to increase this to 8 minutes between services, then the Sprinter shuttles can be offset to avoid the times when both platforms at Greensborough are occupied by sparks.
Dave, it’s not just about timings, but about reliability. You need a way of making certain your delays don’t spread to other parts of the system. Services on one line or service should be physically separate from another; save an emergency crossover or transfer lines used after regular services are closed.
Ricc, I assume you are referring to my through-routing of Lilydale and Alamein. I have only done this because there is no other option, short of using more platforms at Flinders Street or converting Alamein to a shuttle.
Keep in mind that I have dropped the idea of shunting at Ashburton. (I’d really like to be able to edit the above posts…)
With the current infrastructure, conflicts at single track locations and at junctions are unavoidable.
Why can’t you just send it back to where it came or run a shuttle? I do think your plan is slightly too detailed.
Somebody, if it wasn’t detailed I wouldn’t be able to tell you why Lilydale trains can’t form Lilydale trains
In regards to Alamein, I was following James’s proposal; my personal preference was to make it a shuttle. Basically, Lilydale had to go somewhere and since Alamein goes through the loop instead of Glen Waverley, Alamein gets the through route. Basically a “right place, right time” sort of thing.
The short answer is that there were a handful of errors in James’s timetable, and fixing these caused conflicts that required a bit of mix-n-matching of trains at Flinders Street to make it work.
The most prominent were:
1. Lack of dwell at Flinders Street
2. Headway issues – 2min to Blackburn, 2.5min to Ringwood, 4min to Lilydale is ok, but it’s nice to have a bit of leeway between trains.
3. Single track Mooroolbark – Lilydale. In its current form you need at least 13 minutes between a down departure and up arrival at Mooroolbark (14 if switched out). Adding an Upwey-style crossing loop at Cave Hill helps to fix that (full duplication is unneccesary).
4. On the three-track sections to Box Hill, express trains take different amounts of time depending on which track they use because the centre track has a higher speed limit, but sometimes junctions force you to use outer tracks for expresses and you can’t tell which is which by looking at the public timetable, which might also have changes for extended dwell times.
5. Some section running times were wrong.
Who wants the long answer?
Regarding some of the points about Eltham to Greensborough:
* There is probably unlikely to be much of a time saving running the Smartbus express from Greensborough to Eltham. I don’t get the impression that they stop a lot along that stretch anyway (although that may change as people change their travel patterns due to the vastly improved service now on offer compared to every 40 mins and the long way round to Greensborough on offer before) and the real holdups are traffic related, especially along Bridge St. It only takes 4 mins longer than the train anyway most of the time. Driving the same distance by car takes between 6-8 minutes most of the time, so I doubt there’s much possibility of it being faster than the trains, nor by implication of being sufficiently faster to be equivalent to the train and therefore able to replace the train in part or entirely.
* If you’re prepared to spend money duplicating the short distance from Mooroolbark to Lilydale, I don’t see why you wouldn’t do the same for Greensborough to the trestle bridge (just under 4km) – apart from the trestle bridge the rest appears relatively straightforward. That way you can run a Hurstbridge-Eltham off peak shuttle with whatever you like without risk of overcrowding a single Sprinter.
Krustyklo – duplication of a railway has nothing to do with the class of train used, only the frequency. If a Sprinter gets overcrowded, you can couple a second one on.
Sprinter service-
Pros:
Proper allocation of resources relative to patronage offering
Allows electric trains to be used where there is more patronage
The Nillumbik Shire can pay excess costs in trade for a nameplate on the Sprinters, “Nillumbik Numbat” as per Ricc’s suggestion
Greater travel comfort on shuttle
Toilets provided? If so, don’t need them on stations?
Optional on-train conductor? Saves AO’s and improves passenger experience?
Cons:
Need for empty runs to/from fueling point in city (although this run could be with passengers)
Requires up to 20 drivers at Hurstbridge/Eltham with Sprinter training
Need to hire the Sprinter fleet from V/Line (i.e. paper problem only)
Slight reduction in service to Wattleglen and Hurstbridge stations
Did I miss any?
Do the pros outweigh the cons?
Krustyklo – duplication of a railway has nothing to do with the class of train used, only the frequency. If a Sprinter gets overcrowded, you can couple a second one on.
I don’t disagree, but the basic idea of the original post was that the Melbourne network should be a core double track railway (based on where it currently exists in both an infrastructure and patronage sense) with the single track parts seperated out into seperate connecting services to remove unreliability. Given there was some scope introduced in the comments to duplicate small sections of single track at the ends of double track lines to avoid the need for a shuttle running for a matter of minutes to connect to the core service, I suggested that Eltham to Greensborough should be equally considered. Not because I care about what rollingstock is used on this section, but because I doubt the traffic on offer even off peak would suit the operation of a Sprinter or equivalent. If the traffic is greater than or potentially greater than that of a single Sprinter or 2, then it is probably worth considering including it in the core network, along with a short duplication to make it feasible in the context of the original suggestion. Without having specific figures in front of me, Eltham is a reasonably well used station and railhead for bus services from further east/north, many of which were recently improved in frequency and hours of operation quite significantly, presumably with improvements in patronage and potential for further improvement as people change their habits. In addition it generates quite good peak hour loadings and I suspect that a fleeted peak hour service of 6-10 trains with Sprinter shuttles would probably be seen as unacceptable by users and even motorists on roads in the area as people drive to another railhead served by better quality services. As an aside, the roads in the area are quite congested in peak hour now and many of the pinch points are unlikely to be upgraded in the near future eg Bridge Rd, Eltham, Main Rd between Eltham and Fitzsimons Lane, etc. I doubt that either of the roads just mentioned would be easily upgraded cheaply either in construction or politically given the nature of the area.
On the other hand, whilst the population of Diamond Creek is growing, the Eltham to Hurstbridge section could well be operated as a shuttle quite easily with a few fleeted trains in peak hour (as almost used to happen in the days of single carriage Tait motors), and I think the Sprinter suggestion is worth considering for this section.
To express it in terms of your pros and cons, I suspect that proper allocation of resources to Eltham – Greensborough would be an electric service at the very least in peak hour. If it is to be provided as part of the core network for peak hour, it may as well be included full time. In which case, do the long promised duplication and provide a proper service as part of the core network. I suspect that that in turn would improve the patronage as many people from that area drive to places like Greensborough, Watsonia or Rosanna / Heidelberg to use the better service provided from those places due to the presence of the double track and ability to run a better frequency, or alternatively just drive to their destination. Duplicating the track to provide a frequent reliable service in turn will free up car parking spots at those stations just mentioned for those who genuinely could make a case for needing them (ie not living near the train line or decent connecting bus service.
However, in one sense I think we are asking and answering the wrong questions in this post. The capacity and reliability issues we are trying to address are in peak hour, not off peak. In which case, don’t bother with fleeted trains at all, or at best a 3 carriage shuttle to a railhead. If there is a need to move trains from outer suburban stabling yards at places like Hurstbridge, run them as the first few trains for the day to get them onto the core, then run a 3 carriage electric shuttle for the rest of the day. I find it hard to believe that a 3 carriage electric train would save much if any money compared to running 2 Sprinters back and forth (and possibly even just the one) including the hidden costs of needing a specialised pool of drivers and having a specialised fleet outstationed an hour from home with extra requirements for maintenance, servicing, etc, as well as extra costs moving it to and from the city and swap overs. Before someone mentions that it already happens with Stony Point, I would pre-empt the discussion by pointing out that Stony Point isn’t currently electrified and part of the electric network, nor would most people here run Stony Point to Frankston the way it currently is either.
I think in Melbourne that rather than the alternatives being big bang projects in the far off future, or cutting off / reducing viable branches to keep the trunk alive and avoid small projects with significant benefits, we need to ask some hard questions about which parts of the network are worth running as a metro now, which parts are worth including in a metro on the basis of the promise they show to be feasible, which parts are worth running as an electric shuttle on existing single track (or even double track singled) as suggested here, and which parts are worth running as buses or light rail. Faffing about with half way houses such as Sprinters is false economy and just avoiding answering the real issues. In the case of Eltham to Greensborough I believe it justifies a good electric service, in the case of Hurstbridge to Diamond Creek it could be converted to a bus tomorrow for the number of people it would affect (and the bus service could be more frequent than at present for lower cost – although I’d probably run it directly from Diamond Creek to Greensborough for best effect, and consider running it as a wider network to say Panton Hill or St Andrews to make it more politically palatable to the area). Diamond Creek to Eltham I’m not sure what you’d do with. At present it could be a bus, but it has potential to be viable but you’d never be able to duplicate the track from Diamond Creek to Eltham. Maybe just improve the roads, provide a decent bus and otherwise abandon the area to the motorist, which it more or less is now anyway?
Hi all,
I have just come across these comments, and as a resident of Montmorency, wish to make a few informed comments.
There seems to be a view on this site that Melbourne doesn’t exist beyond Greensborough, well the truth is that suburbia has well and truly pushed out through Montmorency into Eltham, Research and now into Diamond Creek. The suggestion that the line beyond Greensborough does not require full duplication and upgraded services is a fallacy.
In Monty, the streets around the station are full of commuter cars, something that was not the case 10 years ago, the same at Eltham. Even off peak, the number of rail travellers from this area is visibly increasing yearly.The line needs duplication to at least Eltham, and longer term to Diamond Creek. It has been estimated this duplication could be done for $200 million. There has also been talk of a new station at Eltham North, and I think this will be needed in time, so full duplication to Diamond Creek should be the target.
Traffic congestion in this area in rapidly increasing due to population growth and vehicles taking short cuts from the Ring Rd to the Eastern Freeway. The Smart Bus is also getting caught up in the congestion.
As an average person, I believe any rail line in a large urban area like Melbourne should have full duplication of all tracks. That should be a given fact. The number of single tracks in Melbourne is a disgrace. Provide the tracks and a high frequency service and people will use it, people are fed up with traffic jams, that is why people are using trains!
Graeme – I think it isn’t so much that we think suburbia stops beyond Greensborough, rather that we’d like “commuters towards Melbourne” to stop there, with Greensborough becoming a city in its own right so that it can support the population from there to Diamond Creek, or Hurstbridge.
In the medium and long term I agree with you, when you say that duplication to Diamond Creek will be neccesary; but in the short term a service of 3 trains per hour (aside from peak) should be sufficient as far as Diamond Creek, and no more than half that to Hurstbridge.
Krustyklo – I completely missed your post above, I’ll respond when I have time, probably the weekend.
Graeme – I will take up the challenge and respond to you.
My bias is that I don’t see a generic 2 track ‘commuter’ standard rail service, such as operates now, as a cure all. As I keep saying, what works for Watsonia ain’t gonna work for West Richmond. The latter needs a metro standard service all day, every day, 10/hr 18/7/365. This is the only way the system is going to be attractive enough to Joe Public to actually use. And it needs to be combined with high frequency buses, so you don’t use the car.
The problem is, the service YOU advocate has to be paid for by someone, and its true cost is much higher than you could imagine. The costs rise exponentially the further from the city you go.
Congestion, as you correctly mention, is the only justification for such a large subsidy. Remember we are talking about congestion beyond Greensborough, which I contend is not as serious as before Greensborough.
Even if we did accept that it was serious at Mont then where does it stand against a thousand other priority areas across Melbourne, from Wyndham Vale to Doncaster or Rowville or Casey or Mornington or Clayton or the Airport…the list is endless.
Rail will not work unless it starts getting really serious ridership levels peak and off peak and which start to make a bite into the congestion.
I do not see the line to Hurstbridge as able to do its share of the heavy lifting for the congestion of the city as a whole. We are not talking 5000 people per hour, we are talking 4 times that.
Graeme, just to add to what David and Riccardo said. The discussion here is at one remove from the traditional demand management approach, in that the focus is on improving long-term efficiency given the many physical constraints, before any network expansion takes place (and potentially makes matters worse).
And there is a good reason for this: the current trains are pathetically slow. Take Diamond Creek, my partner has been travelling there from the inner-west for the past month. Despite being a 32km trip, it is actually faster (and more convenient, and more reliable) for her to cycle there (a 1h40 trip) than take the train (cycling to and from relevant stations). That’s a pretty serious failure of a system, to be beaten by a bike (riding uphill mostly) over that distance.
You won’t get the serious ridership levels Riccardo says are needed to reduce congestion with rail – to the extent that is even possible at all and the law of induced demand says it is unlikely – with that level of inefficiency. The rail system just doesn’t serve many trips as presently constituted.
While I understand that local residents like yourself want railway lines, stations and frequent services to their local area, a fully duplicated railway is a sledge-hammer approach to the problem, with the cost to match.
How do you solve driver/union issues regarding train operators not being able to do the same line within each cycle of their roster?
The rostering of drivers is complicated enough as it is, and you need to appease this group in order for it to happen.
Cal_t,
The easiest method is to simply get more drivers. We need more anyway, as last I heard the retirements were outweighing the new recruits.
If you have more drivers available, you are more likely to find one who can fill a given gap in your timetable.
It’s pretty dangerous to suggest people need to be replaced. Especially not in the industry we are all in.
If people choose to leave, we need to hire more people to replace them. There’s nothing wrong with that. Nor is there anything wrong with getting more people to run more services.
And we cannot reject ideas only because they disadvantage us personally. That is corruption.
Cal, how is suggesting people be replace dangerous? It happens all the time out in ‘the world’. Sport, politics, civil society, personal relationships. No-one is truly irreplaceable.
You seem to have brought that attitude you’ve shown on Railpage with you to TT. Leave it at the door please.
Other options for dealing with driver shortages include automated trains (the person occupying the cab opens/closes the doors when safe to do so) as on at least one of the Tokyo Metro lines or driverless trains (like Vancouver’s Skytrain). Sure, they’ll cost a bit and the unions will squeal, but in the end, there’s a shortage of train drivers around Australia and those displaced will easily find work somewhere else.
LS
I’m not worried about change. In fact, I embrace it. I did not know that TT was actual a place that welcomes constructive ideas to be placed.
Been told to get back in my box, both in RP and at work.
Look. I actually encourage staff replacement through positive attrition rates. In term of attitude, I actually have pretty positive outlook on those things. I just didn’t know on TT I am free to display concepts without fear of retribution.
Hi guys, sorry about all the spam, now removed.
Agree with LS, no-one is irreplaceable, and some industries are populated with people with the wrong mindset. A generational thing but there you go.
The first thing I would look at is why drivers want to be interchangeable between spark and loco driving. Sure one is a boring city job, doesn’t need to be well paid, the other is high pay, potential long hours away from home. The drivers and their union seem to want to preserve one career path. The reality is the railways should be able to recruit a spark driving workforce who never aspire to drive locos, and thereby cut out a great deal of the complexity of the job – safeworking, equipment and line knowledge for a start.
You could build a driverless metro according to the Riccardo plan, from Sunshine to Domain. There are no level crossings, most of it underground and little danger of people crossing tracks on the above ground bit.
Cal_t, you are free to say what you like. Please don’t be surprised though if you come up against some hardheads like us who show no mercy to the sorts of commentary seen on Railpage – fetishing the old culture and equipment, or making special pleadings. Rail transport needs to rise and fall on its merits.
Okay. Thanks for the pep talk, Riccardo. So as long as its justifiable (within cost benefit ratio limits, demand exists, etc), no idea is too absurd to discuss here on TT?
I don’t think you folks are too hard at all. There is a *no* can-do attitude that exists and where at work, severely affects the way rail moves forward in the state I am in, or indeed the country.
As for safeworking, it is seen as a privilege. Only those who can please the people who can assign you to that course can be promoted. You see why I am scratching my head half the time ;p?
The reality is the railways should be able to recruit a spark driving workforce who never aspire to drive locos, and thereby cut out a great deal of the complexity of the job – safeworking, equipment and line knowledge for a start.
Riccardo, this is pretty much what TransPerth do already, teaching trainee drivers enough to handle their EMUs and drive the 5 suburban lines. This is mainly a response to the Pilbara operators long term poaching of their fully trained drivers. Of course, the unions don’t like de-skilling of the workforce, which leads to the epidemic of ‘blue ‘flu’ that hit Perth’s trains last week as part of a wider industrial campaign.
LS
Thanks LS.
drivers should be able to aspire to what they like – but the suburbans service is not there just to act as a conveyor belt into loco driving.
Perth’s system is modern and fairly standard so there should be little complexity compared with basket cases in Sydney and Melbourne.
Consider how there was a split between city and country signalling and safeworking, but both cities have sprawled along their intercity routes (and are about to do so again, with Sunbury electrification).
No surprise that 3 significant prangs in recent times – Glenbrook, Waterfall and Cowan Bank all occurred on the boundary between suburbs and country, and involved quasi suburban services.
NSW considered it “OK” that a de facto suburban service was running ‘dark’ beyond Penrith with no automatic train stopper and using country signalling. On a risk mgt basis, this might be OK for 1 per day XPT but not for what is really a suburban service.
This is what I rail against – this political tendency to say “Lets run subbies to X” on a country main line but not spend the money on upgrading the signalling and other stuff to meet this higher level of safety requirement.
And they do it based on artificial criterion like electrification, not the actual use of the service. So we will end up with Velos every 5 minutes out of Geelong, but no suburban operating practice because they are ‘country’ trains. Whereas if they electrified to Geelong, they’d say “Geez, we need to change over to suburban rules”.
Risk mgt should be based on practice, not just a standard boxes of tricks related to using suburban stock.
I do believe there is a weird thought process when it comes to motive power, as ‘Riccardo’ pointed out, about suburban vs country services. Cityrail up North has kind of blended the two together, with V Sets and DMUs used to service inter-urban services, for which we would consider Geelong services.
And as aforementioned, we have two competing systems of signalling. RFR (TPWS) vs train stops. Ideally, we should stick to one, and they are trialling out RFR induction loops at Dandenong. That’s not to say the two can’t co-exist. They already do at Werribee.
To increase headways, moving block signalling should be the de-factor standard, for areas that are not bound by level crossings. The drivers/unions/organised labor/lower management are always scared of job reforms. The best way to appease these groups is to say: there will be no loss of jobs, and if there are, setup a subsidiary company, let them have their consultation, then take their considerations and let it disappear into ether; leaving them with remuneration and enough super to live on.
On another topic, why is it that drivers see Spark driving as a gateway to being a Loco driver? The Spark drivers get enough money compared to other operations/customer service staff in the company. Are Loco work conditions better? (read Traditional and not open to change), that’s why they go there? If so, send them there
I suspect it is just money. If you go to the Pilbara with some train driving experience, it is better than none. Although some companies up there aiming for driverless trains, so who knows?
Sunbury will be the first example of extension of the wire onto track and signal upgraded for RFR. We will lose 160km/h over that section if they force suburban operating practices onto it.
Agree Cityrail has looked more holistically at the way service is delivered than Victoria – the ‘interface’ is at places like Scone or Dungog where the inconsistency has little actual effect on practices or on passengers. And the Endeavour local stock and Xploder long distance stock are pretty close to identical.
Where the side was let down was signalling. If full track circuiting, radio and stoppers were put in on all the routes where interurbans run suburban services, would be more honest.
Here is another topic which I want to throw open.
Which level are resistant to change? Operators and DoT must take some of the blame, but at the same time, you have lower management, workers demanding respect, yet, they do not respect the wishes of their executive, ie the wishes of the CEO and the visions of the operator?
Where must the line be drawn? What steps have been taken or will be taken?
For some strange reason, it’s like the workers don’t want to welcome the new bosses. What do you think?
I am starting to believe that Connex would have had a lot more things in the pipeline, only to have changes met with strong resistance, with certain working groups sighting ‘cultural differences’ as the reason why it cannot be done?
As expensive as it is, maybe they should have exchange programmes? Drivers from both companies swap roles for a bit, train each other up, give improvements on each other, and same goes for staff from all departments?
Riccardo: I suspect it is just money.
Spark drivers get enough money. W/c allowances, meals, overtime, loop allowance.
My! If customer service staff did it their way, I’d have myki trained allowance, BOM usage allowance, platform attending allowance…