Back in March 2020 I posted an article about tram patronage by route, across Australia.
Via data in this Age article in February which provided recent Melbourne data, here’s an update:
(I’m experimenting with interactive charts. If this is not viewable on your device, try this image.)
Some notes:
- Wrestling with the chart captions was a bit of a trial. For clarity, green is boardings (millions, mostly 2024-25, see below) and blue is boardings (millions) per route kilometre
- The data published by The Age did not include route 30 or 35 (City Circle) presumably because it was at least partly reliant on ticketing data
- Canberra’s data is a bit shaky – I could only find a 2023 number, and the troubles they’ve had with their ticketing system seems to have delayed more recent accurate patronage data
So, some observations
- Route length may play into boarding numbers, but other factors also come into play, particularly whether the line serves a big Central Business District, and the density of suburban areas it passed through
- Gold Coast Light Rail has surpassed Melbourne’s 109, and is catching up with the 96. (The GCLR data is approximate; I couldn’t find all the FY24-25 data, but it should be pretty close to accurate.)
- Newcastle still has the least busy line
- Sydney’s L2 and L3 lines have overtaken Melbourne’s 96 and 109 as the most patronised tram routes in the country
I am not a big one for Melbourne/Sydney rivalry, but should Melbourne, with the world’s biggest tram network, an icon of our city, be a bit embarrassed that the busiest lines are in Sydney? Perhaps.
Fact is, Melbourne’s trams rarely reach their true potential. Many Melbourne tram routes have weak outer termini, limiting patronage.
Service frequency is also lacking. In the evenings you can be waiting 20-30 minutes for a Melbourne tram. In the other cities it’s max 15 minutes (Adelaide 20).
Melbourne’s on-road tram priority (especially traffic light priority) is primitive compared to other cities. This results in slower trams which are less time-competitive with driving.
And although it has less of a direct strong effect on patronage, tram and stop accessibility is years behind where it should be. There are some more low-floor trams and platform stops coming into service this year, but hundreds more of each to go.
The Victorian Government’s plans for development around tram lines will likely boost patronage, but more investment to extend and fix accessibility and priority is needed to help their patronage get to what it should be.
Sources: Melbourne data from The Age. Sydney/Newcastle data from NSW Government. Adelaide from SA Government via Wikipedia. Gold Coast from Translink dashboard. Canberra from 2023 from CMET via Wikipedia. Route lengths: Wikipedia.
Spot any errors? Please leave a comment!
Update 26/5/2026: Thanks Tony P for the corrected Sydney L2 and L3 line lengths. I’ve found some slightly different numbers, but as you say, it increases the millions of passengers per km significantly. You can’t always trust Wikipedia!
50 replies on “Australian tram patronage by route”
That will be a surprise to many – thanks for synthesising and sharing this, Daniel.
What is the relative capacity of the Sydney light rail sets vs say an E Class tram ?
Thanks for the post. I like the interactive graph as well.
It would be great to see an analysis of how we could speed up our trams/improve the network as a whole.
L2 and L3 in Sydney have 5-section Alstom Citadis trams of a later design to a C2-class Bumblebee tram, and similar capacity to an E-class – but they run in pairs.
@Matt McDonald – A combined/coupled X05 carries 450 passengers, compared to 210 for an E Class tram.
The peak hour stats for Sydney are even more interesting. The L2/L3 combined is carrying up to 12,000 people per hour one way in peak hour, significantly higher than any other tram route or corridor in Australia.
But how? It is a maximum of 4 minutes, and 15 trams per hour times 450 = 6,000, about half of that! On top of that, conventional wisdom has placed “on street light rail” at around a maximum of 10,000 pax per hour one way.
It seems that there is significant turnover on the L2/L3 both north, and south of Central, with roughly equal peak hour patronage on both sides. So we are pushing beyond the “maximum capacity” with enough seat (or space) turnover.
What else is interesting is that we are also seeing large queues forming at Central outbound in both the AM peaks and PM peaks. AM due to people travelling to UNSW and the Randwick health precinct, and PM for people returning from city jobs. It might be one of the few transport routes that are balanced in both directions in peak! Most people now know that there is no chance to be able to board the first tram available. So we actually have much higher demand than capacity to serve.
Why is it only 4 minutes rather than 2 minutes? Well, apparently in modelling, they found that such a high frequency with traffic light priority would have significant impact on all east-west routes, causing delays all the way to Parramatta. Even if you had an attitude of “trams above cars”, this however had significant impact on all the buses – routes such as the (current 500x, former M52 and future 52) and 333 would have significant degradation in service and patronage. Melbourne had similar issues when testing absolute light priority on the route 96 about 10-15 years ago – it resulted in significant problems on all Doncaster-bound buses. There may need to be further testing to see whether shorter starting runs from Central outbound may be possible and avoid such delays, but alongside Anzac Pde, light rail vehicles and buses share the same corridor.
Ultimately, we will need Metro West to be extended to the East, before we can consider extending services to Coogee or La Perouse.
A feature of Melbourne’s network are the multi-route corridors (St Kilda Rd/Swanston St, Victoria Pde, etc, as well as the city streets). Most of the routes with lower figures play a part in these.
Great data and analysis. It would be interesting to know how data is collected? I assume more modern vehicles including the E Class would have some form of weight system to help identify loading/weight, similar to modern trains (Waratahs). I would assume data collection is more accurate/mature on the modern systems too.
Surely all this shows is that Melbourne tram patronage is split across multiple different routes rather than focused on a sparse few lines like it is in Sydney. The overall numbers are still magnitudes higher in Melbourne, just split up.
Great to see this list thanks Daniel. How come The Age gets to access PTV figures that we, the public, struggle to find?
I wouldn’t see it as Melbourne-Sydney rivalry because the circumstances are very different in each city. Melbourne developed a strong suburban railway system in the 19th century and led Sydney in rail patronage well into the 20th century. Sydney didn’t and the tram system was its de facto suburban rail system, the patronage of which was as great as those of Melbourne and Brisbane tramways combined.
There are still corridors in Sydney that should have a rail line but don’t, one of these being SE Sydney, so it’s hardly surprising that trams reintroduced on those corridors would be heavily used. Melbourne’s tram lines in general perform a more local function rather than as de facto rail lines. The SE light rail in Sydney also serves huge patronage generators – Sydney CBD spine, ferry and rail termini, sports and entertainment centres, a major university and hospital, as well as a cluster of medium density suburbs.
In terms of pax per km, Melbourne’s calculation is disadvantaged by its long route length, overall carrying about 0.8 million pax per route km, compared with the old 250+ km Sydney system at 1.5 million per route km and the new Sydney lines (about 38 km) at 1.7 million per route km. These figures are still very low compared to Europe, reflecting Australia’s poor level of public transport use. The busiest systems in Budapest and Prague (both over 150 route km) carry about 2.5 million passengers per route km.
It’s fair to say that, unlike the systems in other states, the Melbourne system suffers terrible neglect by government, being starved of traffic separation, priority, accessibility and decently funded service levels. One can’t be too critical of an otherwise fine system that operates under an enormous handicap imposed by government. When I read about your 30-40 minute headways, compared to Sydney’s typical 4 to 15 minute headways, I think you’re very hard done by and can’t blame people for driving instead.
I find it more surprising that the L1 is so low
@Real Metro is Woolies Metro. CSELR already has the design capacity for 2 minute headways on the core route and 4 minutes on the branches. It already uses this capacity between Central and Moore Park/ Randwick Racecourse for events.
There is no “rule” for tram capacity. The former Sydney system was moving 60,000 per hour per direction between the venue and the city.at sporting events. They were also running up to 180 trams an hour each way along George St in peaks. The Prague system has a design capacity, used in peaks, of 25,000 passengers per hour per direction. Such capacities are enabled by the use of turning loops rather than stub termini.
I understand the point about CBD cross streets, but there is the Cross City Tunnel. Road agencies will play bluff wherever they can.
@Michael. L1 reached 10 million, but was badly affected by the CAF tram failure, then covid, and has not fully recovered. It will in the longer term.
Interesting that the lowest patronised line on a per km basis in Australia is about to get the newest trams!
@ Tony P Thanks I had not heard about the problems the L1 was having with it’s Trams.
Daniel… what comes to mind is the amount /number of non-paying travellers … my observations are around Sydney / Canberra / Gold Coast where the Tap On Tap Off process is at the point of departure / arrival
anecdotally a lot of passengers just walk-on / walk off with no tapping and there is very little Authorised Officers checking, again anecdotally Melbourne has a higher % of actual paid travellers, due to the ease of on-board devices
Tap on / Tap off on these other locations being on platforms makes it easier to ignore
@Tony P, I’m guessing the Age took the trouble to lodge an FOI request.
@Steve, the current small trams on 57 and 82 may be limiting patronage. But there’s certainly been a lot of development in the inner NW (and will continue to be) as well as more frequent train connections at Footscray now and the new hospital, so I’d expect patronage to rise in coming years.
@David, it’s important to remember that not all patronage data is based purely on ticketing. We don’t have details for how these numbers are calculated, but we do know that many trams have equipment that can estimate loadings independently of Myki taps.
People not wanting to walk a single block in the Melbourne CBD contributes to ridership in the city. Metro line can get me to Flinders St on time if the loop is delayed through to Flinders St the normal way but I have to hustle.
@Tony P
Rules regarding tram capacity do in fact exist, and are extremely relevant in this day and age. With modern practices there is no way we can consistently run CSELR-length trams every 20 seconds each way at reasonable speeds. There are barely any heavy railway lines around the world that achieve 60,000 ppdph let alone tramways, and that’s contributed to the fact that we overestimate crush loads. People, especially Sydneysiders in 2026, are not going to cram in at 6ppsm-level crowded rolling stock on a regular basis. People were more likely to in the past but even then systems around the world have overestimated this.
The amount of cross traffic also will affect the frequency, reliability, and therefore the capacity of the corridor. Swanston St runs to a crawl and needs double/triple length platforms because of how many other road tram corridors cross through. Swanston Street is already the world’s busiest corridor, and it will effectively never get any higher than what it currently (and technically, frequencies will likely go down with the introduction of the Gs).
We need to be realistic here. Tramway capacity exists.
@Brian. Yet good European systems are capable of running much higher frequencies than theory dictates (Prague for example), but sure not 60,000 pphpd like the former Sydney Tramways with their eight-door cars and average dwells of 8 seconds. But calling on “conventional wisdom” that trams can only carry 10,000 pphpd is an unfounded generalisation. It depends on system design (traffic priority, loops rather than stubs, number of stops etc) and tram design.
Incidentally, that 450 capacity figure cited for the Sydney coupled Citadis trams is actually based on 6 ppsm. At a more standard 4 ppsm, the capacity figure is more likely closer to 400. The higher crowding is enabled by the door numbers and spacing. The Sydney CAFs, with less doors due to TfNSW deleting doors in the specification, are not capable of carrying more than 4 ppsm given the need to maintain reasonable stop dwells. There are a lot of variables to consider that often render standardised assumptions invalid.
Actually, I’ve just found a significant error in a couple of those figures. Whoever complied this fell for the TfNSW publicity mileage figure for CSELR of 12 km. This is actually the combined mileage of the system, with the common core line measured as one route. The individual routes are actually 8.5 km each. That would make the pax per route km about 1.82 million for L2 (not 1.29) and 1.97 million for L3 (not 1.4). This elevates those two routes into serious European territory in terms of intensity of use.
If I stood a place called QEII library for an hour and counted everyone going North on a tram I would count the number of passengers past per hour on the tram line at the point. If I got all the boarding stats for the hour I would also have the number who got on South of me tat got off South of me and also the numbers who got on North of me and got off North of me. I now have two figures the number passing QEIi going North, which only a very high performing tram can go over 10,000 per hour and a whole line heading North which can go higher.
Can someone supply me please with the raw data so I can put it on excel spreadsheet to work average weekday patronage.
@Tony P, thanks very much for that, I have updated the figures using what I’ve found digging around a bit more: 8.5km for L2/Randwick, and 8.6km for L3/Kingsford.
@David, I got your email about this. Will get back to you. Unfortunately I don’t have a weekday breakdown, just monthly.
@Tony P
Yes it is indeed possible to improve capacity through a range of measures, but I don’t think it’s that unrealistic to say that the max capacity of many tram corridors out there is around the 10,000 ppdph mark give or take, even if you account for possible improvements. That is, if all improvements are possible at once given it isn’t on many corridors. Most European systems around the world don’t do much more, only a select few do a bit higher.
Swanston Street runs at 50tph at best during regular peak times, assuming all routes used E-classes (which they don’t, and chances are won’t be possible due to their length and the amount of cross traffic – two limitations), that’s just 10,500 ppdph best case scenario.
Having higher frequencies than this along many tramways around the world would come at a cost of speed and reliability, especially if you factor in limitations (that of course vary depending on the corridor). For example if you tried to run super-high frequencies on the CSELR, even after some improvements which should’ve probably happened anyway (such as balloon loops, platforms extensions, right turn tracks at Moore Park, priority etc) and factoring in limitations (tram length, George Street, IWLR crossing), the system would likely slow to a crawl. Which isn’t helpful given the system has a fairly low seat recycling rate for a good tramway as it was an inappropriate mode for UNSW and Eastern Suburban commuters to Central and the CBD (it’s a heavy rail job), and people expect decent travel times. Sure you could maybe have a conga line situation for special events (to a certain extent) where capacity trumps speed, but is it really worth building a system to do that for special events only? You would also need a great deal of rolling stock that would otherwise be doing nothing during regular service. Realistically what they do now is not bad.
The only other difference they could consider other than the aforementioned is to use 45 metre long rolling stock with the double-length platforms rather than the single 67 metre consists. Which may or may not improve things.
@Daniel Bowen. Yes, L3 is 8.6 km. I was writing without consulting my notes from long ago. That should make it about 1.95 million per km.
@Brian. Naturally there is a point past which they should be building metro rather than light rail and indeed there is a metro line proposed for SE Sydney, but, due to the cost of it, there is a priority queue for that and the SE line is not at the head of that queue. Nevertheless, metro won’t provide the closer stop spacings that light rail provides in the dense inner region of the SE. CSELR does have a legitimate purpose that is intended to be supported in the longer term by metro.
Suggest you look at the Prague system for an intensive operation that doesn’t compromise speed or reliability one iota, but not every operator has such skills. The slowness of the Sydney operation is down to poor system design, poor operator skills and the wrong type of trams, not congestion on lines that typically have absolute traffic priority. None of the lines (except Dulwich Hill with its single track stub terminus at that location) have a problem achieving 2 minute headways if necessary. Like Sydney’s metro system, that capacity is in the design and presently in reserve for when it’s needed.
Not surprised when it’s all but impossible to board a tram in Collins or Bourke Street between dawn and dusk while the entire state of Victoria is designed to prioritise cars over public transport at all costs. Collins St is the worst of the two by far, solely due to all the useless A class trams on the 109.
@Heihachi_73. An A has the capacity of an articulated bus. Count your blessings that you don’t have standard buses like Sydney that have little more than half the capacity of an A. Then you’ll be really feeling it! Having said that though, it’s well past time that Melbourne’s tram fleet should be minimum 24 to 30 metre – and tram priority, absolutely yes.
@Tony P
Yes it’s true there are merits to a tramway especially through the CBD, where local travel is quite high. However, especially given how expensive CSELR was anyways, a tramway was not really the right thing to prioritise over suburban rail for the Eastern Suburbs. Travelling from the Eastern Suburbs is not a local trip, but a suburban/commuter one – a greater distance with lower seat recycling rates. Density has barely increased in the Randwick LGA especially within the walkable catchment of the Anzac Parade corridor, yet the tram is already stuffed with both general and university commuters. And cars still remain a fairly popular mode. The buses that were once there weren’t crowded because demand for local travel was high and a tramway was needed, it was because they were serving as the only public transport alternative to suburban/commuter travel as there was no suburban rail to fulfil that purpose. If one was built, the crowded buses would’ve been solved instantly. The CSELR already has longer stop spacing and large trams to attempt to act as a de facto suburban rail option (which has downsides) – sure capacity could be even higher but you can only go so far. It’s also not helpful during a housing crisis where one of the highest demand suburbs of Sydney now has an excuse to build less transport-oriented development.
Regarding Prague, yes it does employ a range of measures to increase capacity. And like many tramways, it also has limitations (all hell breaks loose when a single one of those unidirectional trams breaks down). But to say that the average tramway out there could boost it’s capacity that much higher than 10,000 ppdph, to the point where it could sit very close under a railway, seems misleading in many situations. As you said, there is no issue running the CSELR at 30tph, but with a realistic 400 vehicle capacity that’s just 12,000 ppdph. If we could somehow bump that to 40tph with some of those measures discussed earlier (balloon loop at Circular Quay, right turn track at Moore Park, priority etc) that’s still just 16,000 ppdph. Which isn’t nothing, but realistically one of the highest design capacities you could get during regular service. Anything more than 40tph with the current 67 metre long trams isn’t really happening even with those new measures.
Besides, the Prague network isn’t particularly fast either (and sometimes, not even as reliable as it could be) Which it doesn’t really need to be, given it’s supposed to be for local travel. But trams can’t replace trains, they don’t have the speed or the capacity. Most tramways out there will have a maximum design capacity of 10,000 ppdph. But critically, it’s also not realistic to say that the average tramway should have or need to implement all those aforementioned capacity improvements and upgrades we discussed, since in a lot of cases it is unnecessary. And that’s the same logic we apply to all modes of transport.
Trams won’t be trains. And honestly, that’s fine.
@ Brian. Definitely the SE suburbs of Sydney need a metro line, especially to enable the much higher levels of housing density that are planned – that’s not a point of contention. Extending CSELR, as some advocate, will not enable that. Its capacity is too low and its journey time too slow. However, the metro stations as planned south of the CBD would be at Zetland, Randwick, Maroubra Junction, then some more south of that. That leaves significant parts of Randwick, Kensington and Kingsford not closely covered by rail, in addition to Moore Park, Randwick Racecourse and much of UNSW campus, were it not for the existence of CSELR to complement the coverage of the future rail. CSELR is still needed (within 7 km of Sydney CBD is not seriously considered suburban any longer), but, no, it’s not a substitute for rail in the region.
Have you ever lived in or been to Prague? You would be in for a surprise. The average speed of the tram system is 19 km/h, which is high for on-street. Check the online timetables for journey times – you will be surprised how fast they are compared to Australian and many other European systems. As for punctuality, the popular view is that you can set your watch by them – the last DPP Annual Report discloses 99.8% timetable punctuality for the whole Prague public transport system. And you surely know that unidirectional trams can be reversed or pushed or towed out of the way? Prague also has a metro system that carries the highest patronage of all the modes. It’s not argued that a rail system does not have primacy, or that a tram system can do its job. I say that living in a state where some politicians think that buses can do the job of trams or trains!
I don’t see tram route 35 Melbourne City circle. I know this is a free tram but so is the CBD. Nobody taps on anyway. I would be very interested in the method of calculation
@Tony P
It really does depend what alignment they take. But at the end of the day buses do a fairly good job at feeder services and it does take a lot until a tram is needed. And that’s before cycling infrastructure. Yes CSELR will eventually be needed, but was more discussing the priority point. Suburban/commuter travel involves passengers travelling between suburbs as opposed to local travel involving travel within a suburb (or two) and given the presence of UNSW it is not unreasonable to expect significant patronage if suburban rail was built.
To be honest, I don’t believe the Prague network is significantly faster than the CSELR or the 86/75/109 in comparable sections. We are more car-brained, so of course many corridors experience traffic that brings the system average down. Yes, unidirectional trams can go backwards but Melbourne really does have the advantage here which is demonstrated by how quickly they adapt to delays by rerouting and diverting anywhere, relocating trams in sidings or centre tracks very easily and sending them in whatever direction they like, short-turning etc. Oftentimes if a delay happens on the Prague network you do end up with a big jam of trams for a fairly long time…
@Brian. The problem with buses in Sydney is that they’re a sub-standard operation. Low capacity (reduced even further on battery buses because of weight), caught in traffic, long dwells because of inefficient passenger exchange practice. For all their slowness, Sydney’s trams still do a better job at moving crowds. Buses are suitable only for feeders and cross-suburban work.
When metro is eventually built to the SE it will only come as close to UNSW as Randwick – about 500 metres walk to the top of the upper campus (not bad) but over a km to the lower campus. CSELR will clearly have an ongoing role servicing UNSW, with stops at both the upper and lower campus. The task of CSELR will eventually close off at the outer ends with interchange to the metro at Randwick and, with a short extension, at Maroubra Junction. As for mode construction priority, it was simply a case of budgetary constraints/priorities and an urgent need to relieve the inadequate UNSW bus service that led to CSELR being built first. Rail costs a hell of a lot more and they know they will need it eventually to service the density they are proposing for the SE. But reintroducing trams to do the job that buses were struggling to do was the first priority. They were not intended to do the job of rail but to sit between the two modes.
I’m staggered at your remarks about the Prague system and wonder if you have actually been there or studied it? It is significantly faster than the Sydney and Melbourne systems in equivalent sections, as you’ll see if you sit down and analyse journey times and distances over equivalent sections (or even ride a tram!). Like Melbourne, they are also very skilled at dealing with disruptions and divert trams just the same. They use turning loops, as Sydney used to in the city and busy suburbs (and still does at Central Station), to reduce terminus downtime and blockages resulting from shunting. The Melbourne tram operators are aware of this too and a few decades back considered turning loops and unidirectional trams for some busy routes, but space constraints led them to adopt alternative methods of improving the functionality of stub termini instead.
Apologies to @Daniel if we’re overloading your blog with this discussion!
@Tony P, all good, this discussion is interesting.
The idea of Sydney having a tram which is one day supplemented by heavy rail on a very busy corridor should not be too foreign to us… Melbourne just did it!
I recall some tram insiders years ago noting that there was a lot more they could do to boost St Kilda Rd tram capacity if they were allowed to.
@Tony P
Sure, buses in Sydney are quite poor for international standards. But it really doesn’t take much to improve them, and feeder buses really don’t have to be high capacity either. People only need to spend a very short amount of time on a feeder bus to reach a heavy rail station (which unfortunately in our context, currently does not exist). There are plenty of places in Sydney that do feeder buses fine and would be much worse off if they only had a tramway (such as CSELR) to rely on. Feeder buses + suburban rail as a combination would address the demand along Anzac Parade far better than a tramway compromise in this particular instance, as it could provide much higher capacity and speed to address the high number of commuters going in both directions to and from the CBD and UNSW/Randwick health precinct etc. And also, would probably better moving crowds to and from Moore Park than the tram.
Again, it depends what alignment they take; they have not finalised this and they’ve already changed it multiple times. True, CSELR would eventually have a role as demand for local travel increases (proportional to density increase) but feeder buses to stations would provide fairly significant capacity for a while. Trams will replace this role; however, it does take some time – but unfortunately the fact stands that at this present moment, people are using the tram as suburban rail due to its absence, and the system is filling up significantly. Rail has a much higher upfront cost, but with such high demand in the Eastern Suburbs the benefit is far greater. It would’ve solved the bus issues whilst also unlocking a significant amount of housing (both demand for residential and student is extremely high), so it’s more about the cost benefit. Besides, increased density would mean the tram would’ve easily been justified in the future. So, I 100% agree trams sit between the two modes and act as a very efficient local and feeder mode, but that isn’t the priority along Anzac Parade as it doesn’t reflect the demand. But honestly, it’s too late now to debate this. The only way forward is to build suburban rail ASAP because they definitely need it real soon…
Just to be clear, I’m not here to denounce the overall performance of the entire Prague network as a whole, because it does sustain a good level of service in almost all sections. I’m simply suggesting that what Prague has isn’t necessarily something we don’t have at all in Melbourne. The better sections in Melbourne are equivalent to many sections in Prague; they don’t really have anything special that we will never beat. Some comparisons of 6-kilometre sections at 0800 (tried to select contextually equivalent ones):
Lines 26/20: Sídliště Na Dědině – Dejvická vs Route 75: Vermont South – Deakin University
• Average stop spacing (m): 500 vs 500
• Mixed traffic (%): 0 vs 0
• Time (min): 16 vs 15
• Average speed (km/h): 22.5 vs 24
• Headways (tph): 15 vs 6
Lines 16/10: Motol – Moráň vs Route 86: La Trobe University – Collins St/High St
• Average stop spacing (m): 500 vs 333
• Mixed traffic (%): 3.3 vs 63.3
• Time (min): 19 vs 18
• Average speed (km/h): 19 vs 20
• Headways (tph): 15 vs 8
Lines 24/3: Libeňský zámek – Karlovo náměstí vs Route 59: Moonee Ponds Junction – Melbourne Central
• Average stop spacing (m): 429 vs 353
• Mixed traffic (%): 6.8 vs 40.0
• Time (min): 22 vs 25
• Average speed (km/h): 16.4 vs 14.4
• Headways (tph): 10 vs 8
Lines 3: Přístaviště – Jindřišská vs Route 70: Power St/Riversdale Road – Swanston St/Flinders Street (which I realised was not really a great comparison, see below)
• Average stop spacing (m): 500 vs 353
• Mixed traffic (%): 20 vs 53.3
• Time (min): 19 vs 25
• Average speed (km/h): 19 vs 14.4
• Headways (tph): 7.5 vs 8
Time/average speed rank:
1. Yarra Trams Route 75
2. DPP Lines 26/20
3. Yarra Trams Route 86
4. DPP Lines 16/10, Line 3
5. DPP Lines 24/3
6. Yarra Trams Route 59, Route 70
These are just some examples, and not all of the Melbourne ones are particularly good ones relative to the rest of the network either. Looking back, I think Line 3 should’ve been more comparable with the Route 96 to St Kilda, but I’m simply not bothered to compare anymore haha (it’s late). A quick check says there is only a couple minute difference. Headways are generally higher in Prague, and that reflects the density and demand for public transport over cars in sprawly, stroady Melbourne.
But given how much more mixed traffic there is in Melbourne and tighter stop spacings, it actually does really well for speed. Hence why Melbourne’s system average of 16 km/h isn’t (slightly) less than Prague 19 km/h because their trams are fundamentally superior, we simply have a (much) larger network here with a high proportion that just hasn’t been brought up to the standard of the better sections, thereby bringing down the average. In other words, Melbourne absolutely does know how to run trams quickly, we just need to expand what we know to the rest of the network that isn’t so good. I think because of the negative perception that Melbourne has a lot of mixed traffic operations in many poor sections makes us forget how good it actually is overall. Possibly the low-density nature moving everything far apart is also a factor, you simply just need to travel longer in general. And besides, trams aren’t all about speed, Melbourne simply relies on them too much for suburban travel when they work best for local travel and feeders to trains (which Melbourne sucks at). A reconfiguration of the network would be ideal, it’s very radial now.
In regard to dealing with disruptions, it’s the simple fact that unidirectional systems don’t have balloon loops everywhere and they are inherently slower to recover than bidirectional ones. It can get quite crowded when that happens in Prague. Melbourne has done a very good job without balloon loops (though they could consider them in various locations) so I think we have the best of both worlds. It’s not like we can’t have balloon loops in bidirectional systems anyways…
@Daniel Bowen Thanks for letting us discuss and tuning in :)
@Brian. I’ve figured out who you are now. You need to get out and about to experience the reality of other systems rather than working from paper alone. For a start, live in Sydney for a while and experience the daily reality of the inadequate bus and suburban train systems and understand why metro and light rail are now the preferred way forward in that city.
I’m not disparaging Melbourne’s system. It’s competence is high, but it’s thwarted by government neglect of its operating environment, notably tram priority. As for your route comparisons, it’s not valid to compare a system with compulsory stopping with one with optional stopping, especially as the one with compulsory stopping has to handle twice the patronage of the one with optional stopping. Despite these factors, the fact remains that that system’s average speed is at least 3 km/h faster than Melbourne’s – the latter largely down to Victoria’s government failing to give trams priority and rationalising stop spacings.
@Tony P
I don’t think you know who I am. I’ve lived in Sydney for years (Sydney-born too) so I am quite familiar with the city’s public transport. So much so, it’s the system I know the best. The metro is effectively a suburban rail line, it is geared for suburban travel/commuters and is definitely unlike a traditional metro system built for local travel in the inner city. The NWRL was going to be part of the CityRail network anyways with little changes to the nature of the line since those proposals. The Sydney Trains network with some upgrades has the ability to run as well as the metro, and honestly it already does in many aspects. I can assure you UNSW students would love to have a suburban rail station nearby or at least a few stations along the CSELR so there’s a little more wiggle room on those trams.
I’m 100% with you with about the Victorian Government not caring much about the trams in Melbourne. My point is that the shortfalls of the network are caused due to underinvestment in many sections, not because systems in other nations are inherently better at us at operating tramways to a high degree. Yes Melbourne uses request stopping, but this is less relevant in the peak. And even less relevant in the comparison as stop spacing is tighter so it roughly evens out. I can tell you (both on paper and from experience) on many of the high patronage routes in Melbourne like the aforementioned, there is always at least one person who wants to either get on or off – which is all it takes to trigger a tram to stop. Given how mixed traffic was present in significant sections of the examples I provided (other than the 75) and that tram priority lights don’t exist (well they technically do, except they currently operate as car priority lights that make the trams stop for right turning traffic) Melbourne is doing a really good job for what it has and could easily be just as good as Prague with just some minor changes. 25% or so of the network is already running in ROW sections so it’s not like they don’t know how to implement that, and all they have to do is make those priority lights work for them. I tried my best to compare equivalent corridors in that comparison, and the difference honestly wasn’t that significant.
As much as I appreciate the network in Prague (it certain has merits and I would love if Melbourne could take a thing or two out of its books) it’s not exactly so ground-braking that it’s necessarily going to be that difficult to beat. In fact I would argue the only thing that makes it hard is the fact that the Melbourne network is so big it’s going to take a while to catch up. A money, time, and priority issue rather than a skill issue. And if we compare Prague to other European cities, then it’s performance is pretty normal.
I really appreciate your efforts in engaging in this discussion in good faith. Let’s not jump into assumptions or target other opinions with them, and refrain from undermining the credibility of particular points with ad hominems. And at the end of the day, experience can be influenced by psychological factors, presumptions, and biases which everyone has. And I definitely have my own, honest with you there. I think our views are closer than it may seem.
@Brian. That the Melbourne tram system is unquestionably competently operated is exactly the point that I made when I wrote that its shortcomings are due, *not to the operator*, but to lack of (or tardy) government commitment to invest in an operating environment and service level that it can do its best in. The comparison with Prague is not so much about comparing operator competence as pointing out that the operating environment is much more sympathetic to trams in Prague. They have absolute right of way everywhere, are very frequent and reliable and the drivers are trained to drive fast. Its patronage is twice that of Melbourne’s trams, despite the city having only a quarter of Melbourne’s population.
Sydney’s metro is a rapid transit system and distance is irrelevant to its function. Sydney Trains will never be that, not least because of line entanglement. As an example, it can be noted that Rouse Hill (metro) and Macarthur (suburban express) are the same distance from Central (about 45 km), with the Macarthur service having one less stop, yet the metro journey time is about 15 minutes faster, and it is more frequent, more reliable and cheaper to operate. Big handicaps for the suburban system to overcome if it is to “catch up” in any significant way. Melbourne is lucky to have a system that sits between suburban and metro.
Apologies for confusing you with somebody else.
@Tony P
Noted and agreed, though I believe that apart from sections that have extensive mixed traffic (the Route 78 is entirely mixed traffic) the trams are actually acceptable speed wise despite possible improvements. As evident in the comparisons the Melbourne equivalents were either slightly faster or slower, and many routes should really be used to connect to suburban rail (which is not great in Melbourne) so it shouldn’t need to be so quick anyways. The patronage in Prague is attributed to its higher density; the overall population of the city is less relevant. As a result of the low-density sprawl in Melbourne, it has not only reduced the amount of people living within walking distance of public transport, but also incentivised the development of many large motorways, car parks, garages in all dwellings etc that makes driving more competitive. Not even Sydney gets to enjoy something like Melbourne’s extensive toll-free motorway network, if you know you know…
All metros are rapid transit, but not all rapid transit systems are metros. Interlining is quite the issue on the Sydney Trains network, but similar to our tram conversation that’s also something to do with investment rather than an inherent issue with legacy suburban rail. In fact, we know that even metros (well, almost all forms of heavy rail in general) can run well with branching and interlining, so there is also no reason why the entire suburban rail network in Sydney needs to have no interlining whatsoever.
Campbelltown to Central and Tallawong to Central are more accurate comparisons distance wise, with the former making 15 stops (via Airport stations) and the latter 18 stops. The difference is around 7 mins in favour of the metro. Yes, indeed the metro is a bit faster and efficient in this situation. However, you do need to account for the following:
• ‘Temporary’ line speeds that have progressively been implemented especially since Waterfall, with huge drops in line speeds on the East Hills line.
• Acceleration on Waratah trains being artificially limited, they have the capacity to accelerate similar to what the Metropolis stock does in practice (it never actually hits 1.2 ms^-2).
• Current double deckers having poor dwells, though this is influenced by the 3×2 seating (which a silly choice in 2026) and narrow stairs that could be widened (they really just need a little bit more width). 4 doors per car also isn’t helpful but that’s hard to change.
• Tons of slack in the timetable, especially recently.
• Signalling, but this will change as ETCS L2 is rolled out.
• The decision to run the Macarthur express (all stations to Holsworthy then Padstow, Riverwood, Wolli Creek then Airport stations) on the local tracks instead of the express.
• Somewhat following on from the above point, having local trains run via Airport and the express trains via Sydenham, which is why the Macarthur express doesn’t stop at Revesby and switches onto the locals. This will eventually change for the better though, with all trains via Airport stations in the upcoming timetable (as it should be). But apparently, they are keeping the Macarthur express that way, likely just so Padstow and Riverwood get 8 tph. Which is a choice. Eventually the plan is to increase frequencies throughout Sector 2 but that’s a while away at this rate with the government taking no interest in improving public transport.
• In terms of frequency, Tallawong’s 15 tph is only slightly higher than Campbelltown’s 12 tph (which isn’t very high for Sydney standards anyways, at least not historically) but the metro is set up for 30 tph which is higher than the 20 tph Sydney Trains (used to) run. 24tph is possible with ETCS L2 though and note the double decker rolling stock has a greater capacity (which could also be expanded with improved seating layouts). If they ever manage to get 24tph on Sector 2 happening (up from the current 16), then Campbelltown would likely keep its current 12 tph but the local service would be boosted from 4 to 12 tph. Airport stations would end up with 24 tph combined.
• In terms of reliability, that’s a signalling/investment issue + conflict between unions and the government.
The metro is new and engineered to meet high standards of what suburban rail could be, but its speed isn’t really something we can’t replicate throughout the rest of Sydney’s network. Many lines in Sydney were faster not that many years ago until slowing down the network became a consistent trend, which is disappointing to see. Like how the new Bankstown line conversion means trips are soon going to be as quick…as they were in the past before they began lowering line speed and padded timetables towards the end of its twilight years. I guess the metro means we’ve come full circle; it’s become the only reasonable excuse to upgrade the performance of suburban, when in reality major changes such as full conversions don’t exactly have to happen to achieve that (though in this particular instance the Bankstown line running into the City Circle was problematic frequency wise).
I’m not sure what you mean by ‘Melbourne is lucky to have a system that sits between suburban and metro’. Not sure if you’re referring to the trams or the trains, but the Melbourne’s suburban rail network is the same in nature as Sydney’s, just more radial and a little bit worse…I find the line speeds are way more consistent than Sydney though, unfortunately the issue lies more in the dwells and timetabling.
Unless we are to discuss more about Melbourne, I fear we may be getting off topic though.
@Brian. I mean that Melbourne dodged a bullet when it decided not to use double deck trains and stayed with single deck, three doors per side per car that are far more efficient and also have more *practical* capacity, which is about 1,400 in a 160 metre HCMT train, compared with about 1,200 in a Sydney 160 metre double deck. The double deck system was coming up against its capacity limits, with little scope for expanding capacity for future population growth, which is why the metro was introduced.
I know there’s a school of thought, particularly backed, unsurprisingly, by the union, that the double deck system can be vastly improved, but I’ve observed all the attempts at this since the 1970s and they’ve all come to nothing. More significant politically is that the public mood has now radically changed since metro opened. Everybody is weary of the constant unreliability of the suburban system. When councils, the public etc call for new train lines they only call for metro. Nobody is interested in expanding the suburban system any longer. Politically and practically, that show is over. The task ahead is to relieve pressure on it by expanding metro and incrementally improving suburban operation by the means you mention – if they actually work.
Don’t forget, in NSW we’re also saddled with a militant union that’s resolutely opposed to modernisation of both train and bus systems and inclined to take industrial action at the drop of a hat. While light rail and metro are not completely immune from the threat of that, they’ve at least been able to improve operational efficiency by design and practice – e.g. multiple extra doors on both modes, with all-door boarding, to expedite passenger exchange and flow. Also, unlike the double deck trains, both modes have fully unassisted accessibility to current DDA standards.
I have to call into dispute your calculations about Macarthur line trains vs NW metro. I checked the typical peak hour runs between Campbelltown and Central vs Tallawong and Central. The fastest Campbelltown expresses via Sydenham have 11 intermediate stops and take 55 minutes, compared to 52 minutes to Tallawong with *17* intermediate stops. For sure a train is going to perform well if it doesn’t stop, but that’s not the purpose of public transport is it?!
The trains via Airport line take 63 minutes with 13 intermediate stops, 11 minutes slower than metro despite 4 less stops. I did in the past find a couple of services with close to 17 stops to match the metro and the difference was 15 minutes, but the timetable must have changed and those services no longer run. The overall outcome is that the metro journey time beats any type of suburban service despite the “handicap” of more stops – which isn’t actually a handicap because that’s what a public transport service is supposed to do and people love that.
Stopping at all stations also serves the planning objectives of the multi-centred modern Sydney where activity and employment centres are scattered across the whole metropolis and not just the CBD. The old two-tier train service (express vs local) focussed on the old CBD doesn’t serve a decentralised city well.
Talking about what people like – your comments on trams remind me of a couple of Melbourne tram enthusiasts on the TDU tram forum visiting Prague and coming back saying they wished Melbourne trams could move along as fast as that! Public perception is important. Passengers need to feel that they’re really moving along, not crawling along slowly watching other traffic speed past. Those photos I see of Melbourne trams stuck in a conga line of cars actually driving on the tram tracks give me the creeps. That really has to change – coming back to topic!
@Tony P
The 1200 figure is heavily influenced by the poor seating layout though. HCMTs have longitudinal and 2×2 seating, as opposed to Sydney’s 3×2 with ridiculously narrow aisles and passengers shuffling across 3 seats. And the HCMTs use a 6ppsm metric for their maximum capacity, which we know is unrealistic especially for Melbournians in 2026. The floor space of the Sydney double deckers is still greater, and in fact Melbourne’s 24-metre non-HCMT carriages means they could have double deck trains with 6 doors per car, without sacrificing too much floor space. The Dandenong line has the capacity to take 225+ metre long single deck trains which is obviously better than Sydney’s situation of double deckers. However, realistically Sydney will never be able to run anything longer than 160 metres, unless they really do take the approach of terminating suburban trains at Sydney Terminal (which does have significant downsides).
I would argue they haven’t really made any attempt to improve passenger flow since the 1970s, aside from Comeng’s attempt of widening the stairs and doors when they designed the Tangara – which ended up going to Goninan, so all changes made were scrapped. I don’t think it is true that people are only interested in the Sydney Metro, now that the Sydney Trains network has vastly improved since the late CityRail years. Proposals such as the New Cumberland Line are likely to be part of the Sydney Trains network, so Sydney Metro isn’t necessarily always going to be the way forward despite its emphasis. Not to mention my point was more about improving the existing network rather than discussing what new project will be built as part of Sydney Metro or Sydney Trains etc. And besides, public and political opinion have also proven to not necessarily be something that has to be addressed. Especially in Melbourne, most recently the Skyrail debacle, or perhaps the public petition backed by politicians who are trying to bring back Dandenong line services to the City Loop as we speak…
Regarding the best possible design for Sydney 4 door double decker carriages, there are really only a few relationships we can work with. The number of seats is proportional to the dwell of the train and inversely proportional to overall train capacity, until train capacity reaches a certain point where it becomes excessive. Then, train capacity becomes proportional to dwells. But train capacity is also inversely proportional to the headways, when line capacity must be maintained. Trying to find this balance isn’t necessarily easy, but it also hasn’t really been considered either. Likely because there was no real need to in the (near) past, as our focus on public transport performance is recent excluding the days before car centric living.
Campbelltown to Central via Sydenham is 12 stops during the peak, however you aren’t accounting for the fact that the peak direction express that skips East Hills and Panania saves barely anything (the section is not quad track). I believe that don’t even do this to save time, but due to timetabling limitations alongside the East Hill/Panania stoppers and Macarthur express. The 63 mins via Airport during the peak you mentioned is the aforementioned flawed Macarthur express, which as I addressed earlier before, uses the local tracks. There are no standard expresses via Airport during the peak, as all expresses (save for the Macarthur express) go via Sydenham – only off peak does it revert back to the simple Revesby/Campbelltown split via Airport. And that 15-stop express takes 59 mins at most, just 7 mins slower than the newer 18-stop metro. And once again, as much as I understand your points about the metro this still doesn’t address the issues I raised with the East Hills lines today. If upgrades to signalling happened, greater acceleration on As and Bs were permitted, and line speeds were increased more consistently throughout to increase average speed (which as demonstrated in the past, it could be), service would further improve. The gap between Sydney Metro and Sydney Trains can be narrowed; at the end of the day, they are both very similar in nature, and arguably both are suburban rail except with one being newer and the other slightly neglected. It’s like how legacy tramways have the capacity to be just as fast and efficient as modern European tramways; they are fundamentally the same mode of transport.
Stopping at more stations obviously has significant benefits, however three other things are also true. Firstly, the East Hills line only has one local/express split, meaning to travel between local stations east and west of Revesby you only need to change once. This is why I’m supportive of 4 platforms at Kingsgrove first to expand express service, then possibly extending quad track to Holsworthy whilst adding an infill station at Voyager Point. Then, we could possibly consider single deck for the local service. Secondly, the fact that Sydney has sprawled so far and that demand for both employment and housing is still concentrated closer to the CBDs despite that, means we will be in a situation where a significant proportion of the city will be travelling longer distances for work, for quite some time yet. In fact, the mere reality that a city of less than 5 million has such a large urban footprint, means this will inherently be an issue unless the population and density ramps up significantly – bringing me to my third point.
The CBDs we have will still (and likely forever be) the most significant areas of employment, because that’s where the higher demand naturally is and will likely be the highest forever. Gadigal CBD will always have an edge when it comes to demand, that even forced decentralisation can’t necessarily undo. Not to mention the additional costs that are involved in this process; for example, Bradfield CBD will be the most inefficient waste of our money, building more infrastructure and sprawl further out where demand for housing and employment isn’t nearly as high, rather than building in existing areas where the demand and infrastructure already is. Most cities are not naturally polycentric without intervention, and even those that are require a significant amount of density to ensure it occurs. Which is a lot easier when you have a smaller urban footprint and high growth, Sydney only has the latter. Canberra has this issue of having dispersed employment centres despite being low density, and this was heavily influenced by poor planning practices. It now means residents spend large amounts of their time commuting, as you’re more likely to work in an employment centre away from you as employers cannot justify opening multiple offices/workplaces for a small, low-density city. This is exacerbated by poor public transport. Sure, a significant proportion of the workforce in Sydney do not work in the CBDs, but that’s where travel demand will be. And given how large the city is, express services do play an important role. Parramatta CBD is our first attempt of a new infill CBD because we have sprawled so far from demand due to poor planning practices and other factors, and it will take a fair while to develop. It will take even longer for other infill CBDs to exist, and a greenfield one will never occur unless the government initiates it (which it unfortunately has…). Point is, there will still be significant demand for Campbelltown residents to travel to the City Circle effectively forever, this demand isn’t going to shrink any time soon (likely never). I think express services are justified in this instance. And of course, there is also no reason to abandon the quad track infrastructure that is already there.
Back to Melbourne. I’m not so much saying there isn’t anything the network should improve on; however, to say most routes in 2026 are clogged up with trams isn’t really so true anymore. The conga line situation is so much rarer now, and just something that people love to see in photos and constantly complain about to claim that Melbourne’s network is worse than it actually is. This is a general comment, but many studies of psychological factors associated with the speed of public transport are mostly older and are less relevant in today’s world where people simply stare at Google Maps for the objective results. The gap between the speed of trams vs cars is narrowing, in addition to changing preferences of driving especially in the inner city where trams are. And also, let’s not forget the rule – when a tram stops, you stop! Motorists are often stuck behind trams as well; they don’t end up being that much faster in a lot of cases. And by the time they reach the inner city, trams are given their own separated ROWs whilst they get stuck in traffic…
Again, speed would also be less of an issue if more passengers simply took the tram to a station and zip to their destination on suburban rail. Unfortunately, Melbourne’s trams are so commuter oriented by sending almost everything to the CBD. That needs to change as heavy rail frequencies increase.
(I also do apologise for my increasingly long replies…)
I think at this point the discussion has moved too much into theory and speculation and a lot of it is not about the planning and transport directions that Sydney is moving, which are heavily influenced by the fact that the geographical and population centre of Sydney is near Parramatta. There are 2.5. million people living in the western half of Sydney, which is like having another Brisbane or Perth there, a situation that doesn’t exist in other Australian cities where the old CBD is more centred. The public transport effort is now largely in Western Sydney and metro is the rail mode chosen for the future, non CBD oriented network there. And it’s interesting that Melbourne has also chosen automated metro for its philosophically similar orbital rail. I don’t think we’re going to convince governments (least of all State Treasuries who have to deal with the higher operating costs of traditional rail) to do an about turn at this stage.
@Tony P
(Apologies for the slow reply, I wrote a more comprehensive one including some sources and thorough examples which I lost, so here is a simplified version below)
It isn’t really just about speculation. Parramatta may be the geographical centre of Sydney; however, it is not where the demand for employment or housing is the highest – and it has never been. Other Australian cities have attempted to sprawl and decentralise, this has always been a theme of planning in Australia), it’s not just Sydney that has this issue of a significant proportion of its population living in the outer suburbs. The only difference is that the demand is much higher yet historically we have continued to develop outwards, so the issue is more prominent.
I would argue that governments are actually shifting away from this idea of transport and housing decentralisation when it comes to planning; it wasn’t something new at all, it’s something that has historically been attempted countless times all across Australia, with almost none of them being genuinely more successful than the counterfactual. Politicians, planners, and other public servants are finally realising (years too late), that building against the demand is a waste of public resources. You can see this especially with housing policy for example, as we are trying to switch to building greater supply where the demand already is through upzoning programmes. But you can see this with transport as well; Australian cities have overall been far more focused with transport development in existing areas, rather than expanding outwards.
The financial feasibility of greenfield development (especially in Sydney) has plummeted in the private sector. More recently, especially with the introduction of various upzoning programmes in NSW such as the TOD SEPP, LMRH, and various LEP revisions, the feasibility of high-density development west of Parramatta has also decreased. Other Australian cities are following similar trends, including Melbourne. This is unsurprising; after all, the demand has always been concentrated towards the inner city and middle ring suburbs. Importantly this, amongst other data, indicates that not only are we moving away from greenfield development to infill development, but we are also focusing more on infill housing and transport development closer to high demand locations rather than the outer suburbs. Policy has both reflected and influenced this trend, but it is the demand that drives these changes. In Sydney, there’s a reason why there is little motive for projects such as extending Tallawong and St Marys to Schofields, Leppington to Bradfield, electrification south of Campbelltown, even extending Bankstown to Liverpool, and Metro WSI will likely be the last greenfield line for a while. But improvements on the existing network (or new infrastructure in high demand areas east of Parramatta) such as RSIP (replacing MTMS), MTIRP proposals, extending CSELR, Metro West, transport to the Northern Beaches, the Parramatta Road corridor as well as relevant upzoning are becoming more of a focus as opposed to our past approach of playing infrastructure catch up in rapidly growing outer suburbs. Other Australian cities are slowly moving towards this approach as well. Melbourne is improving their existing network and even opting to build new lines in existing areas as opposed to extending further out, complemented with the new Townhouse & Low-Rise Code and Activity Centres programme. Perth is realising that not even rail-induced sprawl is going to remain sustainable for their 150km-long city. Brisbane is still catching up with Springfield over 30 years later whilst dealing with some of the worst congestion in the world, all next to a neglected suburban rail network.
The latest IPART report also shows that the cost of operating the Sydney Trains network is not significantly higher, especially compared to other modes. Note that it has infrastructure such as the Richmond line which will inherently have a higher cost of operation per passenger until density within walkable catchments and services are increased. As public transport becomes a more significant mode in Sydney, this gap will also narrow and become less relevant.
At the end of the day, my previous comments were not really about discussing new infrastructure, rather the relevance of current infrastructure and the importance of improving it. However, even new infrastructure will likely involve servicing not only existing areas of cities, but particularly places with higher demand. And yes, housing and transport are particularly relevant in unison nowadays, with new transport infrastructure often justified on the basis of providing housing supply. But critically, as more housing supply is delivered along the existing transport network, the importance of improving existing infrastructure to support it will increase accordingly. And we are seeing our planning practices finally shifting to reflect that rather than the other way around, albeit very slowly. So, plenty of people will continue to travel from places like Campbelltown to the CBDs, decentralisation is not going to offset existing demand in a growing city to the point where this radial travel demand falls off a cliff. Speaking of Campbelltown, let’s not forget that the suburb itself was (re)developed in the 50-70s as planning practices at the time suggested it could become a new CBD of Sydney. Half a century later, it’s not even a Liverpool…
Bradfield will likely follow the same path, especially in the context of today where sprawl is not sustainable nor nearly as feasible as it was in the past.
@ Tony P & Brian
1. I think it was mistake for Sydney to go double deck.
Single deck trains have about the same capacity as double deck trains, once you allow for the extra trains-per-hour capacity enabled by designing to minimise dwell times (that means three doors per side). But with a lower proportion of seats, of course.
Sydney’s double deck trains are expensive, heavy and power hungry. They carry a full load for one or trips a day, then spend the rest of the day with very low load factors. The stairs are unfriendly to less mobile people (more of an issue as the population ages), and the lack of passive surveillance along the train reduces the sense of security at low use times.
Assuming 24 per hour in the inner area, single deck trains could carry the peak load, if supplemented by some peak-only services from the west to Sydney terminal.
But that’s water under the bridge, sadly.
2. Marketing Sydney’s metro lines as a high level different mode of transport (‘train/bus/light rail/metro …’) so that route and timetable information is not integrated with that of the legacy train lines, is really really stupid. Passengers don’t care (and may not even know) what the service is called; they just want information to be available in the most logical place.
3. I think the next priority for a four-platform station on the East Hills line would be Beverley Hills, because of the important bus interchanges towards Strathfield and Hurstville. That would be easy to do around the present alignment with side platforms. It would be possible but tricky to rebuild as double island platforms; but that’s hardly necessary as same direction cross platform interchange is available at Revesby.
Melbourne needs a concerted project to speed up trams running in mixed traffic on four-lane roads.
The key problems are queues at traffic lights and cars turning right off the main road blocking the tram.
I suggest the key actions would be:
1. Ban parking for a suitable distance on the approach to traffic lights; mark the road in this area so the traffic moves to the kerbside lane and the median lane becomes a short tram-only section.
2. Minimise right turns off the main road by marking out ’round the block’ turns – left/left/left or left/right/right to cross the main road.
I would not support extended clearways, since in shopping areas the kerbside parking has a traffic calming effect and improves the amenity of pedestrians on the footpath.
Away from traffic lights, letting cars drive on the tram tracks is usually not much of a problem (apart from the right-turning ones) because the whole traffic stream is going as fast as the tram would want to anyway.
50 trams per hour on Swanston St is really too many.
Here’s how you would improve things:
– Build new chords from Queens Bridge St to Flinders St and from Flinders to Elizabeth
– Run tram 1 from Sth Melbourne Beach via Kingsway, Queens Bridge, Flinders and Elizabeth to pick up one of the Elizabeth St routes.
This also gives better connections from Sth Melbourne to points north of Melbourne University, and it reduces congestion (both for tram movements and for people) at the very cramped and unsatisfactory Elizabeth St terminus.
You would need to build a turnback siding and third platform between Flinders St and Flinders lane to accommodate the trams that continue to terminate there (since you can’t really have trams terminating on a running line that’s also used by the through route.) There’s plenty of room for that as traffic is now prohibited in that area anyway.
You could so a similar thing without any new infrastructure by simply running route 1 via Kingsway/William St to terminate somewhere around Melbourne Hospital. But that might be less desirable from the point of view of passenger destinations. And it doesn’t help Elizabeth St.
and of course route 1 north of Melbourne Uni would be picked up by one of the routes that now terminate there.
@Brian. Parramatta is not only the geographical centre of Sydney, it’s also the population centre. What you’re saying is that the 2.5 million (and growing) people in western Sydney should commute all the way across the metropolis to the eastern half (which has roughly the same population) to go to work, study and play, but those in the eastern half are not expected to do the same in the other direction.
That’s an economically and socially unsustainable viewpoint. You may be too late for your pitch anyway. The NSW government itself and a number of corporates are already based in Parramatta and much industry, as well as tertiary institutes are established in western Sydney – not to mention Sydney’s second (and 24 hour) airport. The old Central Industrial Area of South Sydney is virtually gone. You’re promoting a transport model for an urban scenario that’s already died.
@Tony P
(Once again wrote a reply and it didn’t send, or perhaps it did so you may get this twice, albeit in a shorter form as again it did not save)
Well, I’m not actually saying we shouldn’t have people travelling to Parramatta, so I’m not sure why that was your response. Rather, the point was that Central station will continue to be a relevant hub. Parramatta may be the geographical centre, the ‘population centre’ (depending on your definition, that could be incorrect), but it most definitely isn’t where the demand is the highest. Where are the house prices higher? Where is the land value higher? Where is the employment density higher? Where are the vacancy rates lower? Where are most people commuting to even in 2026? These are just some of the questions we should be asking ourselves. Despite Parramatta being a special place to me personally, even I understand the demand to live and work near there isn’t higher than the east.
We cannot just equate growth to demand in our current situation. Again, the only reason why there are so many people in the west now is because of outdated planning practices limited housing supply in the inner city, creating artificial growth in the outer suburbs where supply is delivered at a greater rate. This form of decentralisation has historical been a practice done effectively in all major Australian cities. If this was revised (which it increasingly is) this growth will slow. We are already seeing this will sprawl becoming less feasible, but this also applies to infill development. Higher density is the future for Sydney; as LEPs are upzoned and updated, where are the developments being approved? Where are they getting built? It certainly isn’t the west…
This conversation began from the idea of moving away from express services to Central. Sure, eventually people from Campbelltown will be commuting to Parramatta. However, plenty will be travelling to Central for quite some time yet, especially given the population wouldn’t be static – this isn’t merely just a shift. It would be ideal if the east and the west could just commute to their respective CBDs, this has been a pipe dream for many famous planners in the past. However, this cannot occur until there is an increase in population growth and density throughout all of Sydney, which will take some time yet. Remember, it’s always much easier and efficient to bring the housing closer to where people need and want to be, than the other way around. And assuming supply was met equitably (which is currently not the case at all) what would likely happen is that the east will end up being the most populous with more growth and density. The only plausible reasons why that would not happen are a) Sydneysiders simply decided the east was no longer desirable for an extended period of time, or b) the government backtracks on its increasing efforts to permitted supply in the east, forcing growth away.
We can’t undo our past mistakes. The only thing left to do now is to stop further sprawl (which I hope we can agree is socially, economically, and environmentally unsustainable) and begin to allow more density within the west but especially the east.
Linking back to Melbourne…well…this applies to all cities really.
@Brian, your first comment went into the moderation queue because it didn’t have an email address on it. You reposted before I could approve it!
Thanks for the civilised discussion on this! It makes interesting reading.
@Daniel Bowen
Oh apologies, that’s interesting. Swear I did put it in (and usually you get a pop up saying it goes into moderation), maybe the cursor was still flashing and I submitted too quickly before the computer registered that I had typed in the email.
Thanks for letting the conversation continue despite it going off topic long ago haha