A new train map is coming (and: network status boards)

PTV are trialling a new train network map. They’re seeking feedback on it, and you’ll see it at some stations now (Bentleigh, Malvern and Moorabbin, I think).

Note, just to remove all doubt: unlike the PTV network plan, it’s not a concept for new rail lines; it’s a prototype of a map of the existing network.

PTV Rail network map: concept design, April 2014

View the map larger, in a new window

My initial impression: I quite like this.

Colour-coding the lines helps make sense of the way the network actually runs (or will run in the near future). It allows them to add detail such as the stations usually skipped by expresses on particular lines, which lines run via the City Loop, and which sections run as shuttles. This helps people navigate — for instance if you’re coming from the Dandenong line going to Armadale, you’ll probably have to change trains at Caulfield.

The caveat here is that the train network is not currently operated consistently. Loop operations (even leaving direction aside) are very confusing. Express stopping patterns are all over the place on some lines. The Frankston and Newport lines are connected… but only on weekdays.

The operational variations on the various lines might need some work. See the difference between Williamstown and Alamein, for instance; potentially confusing.

A big difference is this map also adds V/Line services. With Myki now phased-in for short-distance (commuter-belt) V/Line services, one barrier to city people using them (the need to buy a separate ticket) is gone. This is an interesting move. It does take extra space, thus makes everything smaller — is the benefit worth it?

The part-time Flemington Racecourse line is shown prominently in black. I suppose that’s a good (for occasional users) and bad (implies it’s fulltime). I’m told it’s showing terminating at Southern Cross because that’s how it’s likely to be (at least on weekdays) in the near future, due to rail viaduct capacity issues, so they’d rather encourage people to change there instead of Flinders Street.

Somehow the order of lines shown at Flinders Street seems wrong, but I think that’s because I know Glen Waverley direct services don’t actually terminate next to Sandringham services.

The Skybus connection is shown, but the Broadmeadows to Airport Smartbus connection isn’t. Neither are the 401 and 601 university weekday high-frequency shuttles, specifically designed to connect to the rail network.

In the first version of the map that got out in the wild over the weekend, there were at least two errors: Violet Town and Euroa had been transposed, as had Ballan and Bacchus Marsh, and the colours indicating Myki validity had crept beyond where they should have. The stations have now been corrected (though Myki still creeps beyond Wendouree, Eaglehawk, Marshall and Traralgon) and PTV expect to do quite a few more tweaks over coming months as a result of feedback.

They don’t expect a more general rollout of the map until Regional Rail Link opens next year. It costs a small fortune apparently.

But what’s wrong with the current train map?

PTV Metro train map, 2013Everyone will have their own views, but the current train map (below) has a few problems. For instance:

It doesn’t show where the lines go. Someone unfamiliar with Melbourne might assume there’s a line from Sunbury to Upfield, for instance. And it doesn’t show any operational detail; the map implies all trains run via the Loop, for instance. It gives little hint as to where the best places to change trains are.

Meanwhile, we’re losing two-zone trips next year, so there won’t be a huge need to show zones as at present. The new map started being designed well before this, but it’s good to be able to take advantage of it to show other useful detail.

What about multi-modal?

I think the new map is a good step in the right direction.

But if they’re starting to mix things up on a map (Metro and V/Line), I think another thing they should be looking at is showing the network frequent trams and buses that back up the train network… though of course, that would be a much more complicated and difficult visualisation to get right.

But other cities are moving into this, and you can see the benefits from it, as described by Vancouver’s Translink:

People traveling along FTN (Frequent Transit Network) corridors can expect convenient, reliable, easy-to-use services that are frequent enough that they do not need to refer to a schedule. For municipalities and the development community, the FTN provides a strong organizing framework around which to focus growth and development.

(My emphasis. That’s the most important point. For public transport to be competitive with cars, this is essential. It’s not like, as Jarrett Walker describes, you can only drive out of your driveway every half-an-hour — but that’s what most PT users face.)

The train-only network map is still useful — good for showing the mass transit, backbone of the public transport network. But a frequent network map would be great for showing all the places you can easily get to in Melbourne on public transport — which is a lot more than just the rail network.

Also: the status board, and the bigger picture

Are maps even in important?

Sure they are. Good maps mean people can navigate their way around more easily, so they’re more likely to use the system. More passengers means more impetus to keep upgrading services.

Bentleigh station: "Rainbow" network status board

But this is about more than just a map. Related is the trial rollout of “rainbow” network status boards, installed this week at Moorabbin, Bentleigh, Malvern, and in the PTV Hub at Southern Cross. The colours on the board match those on the new map… including Alamein, which has a distinctive colour on the map to draw attention to the fact that you usually have to change at Camberwell.

It’s a little early to judge these, though I note that they don’t show next train departures — this is present on other displays at Malvern, but not at Moorabbin and Bentleigh and most other stations.

I’m told they can modify the design based on feedback, so it’ll be interesting to see how this evolves. One issue I think is that line-specific info is shown at the bottom — only a “traffic light” indicator is shown at the top, which means the information you need may not be easy to find.

I’d hope that once these boards are running well, they roll them out quickly to the bigger interchange stations, where they’re likely to be most useful.

Both the map and the status board are part of measures to standardise train operations: the slow move towards more predictable routes, consistent stopping patterns, consistent platforms at the larger stations, and “metro”-like frequent operation on dedicated tracks. And there are also moves to improve the flow of information from operators (on all modes) through to PTV so a better view of the overall network is available, including online.

Clearly they’ve got a long way to go, but this is a step forward.

Other maps:

See also:

Pictures from Violet Town

It was 44 years last week since the 1969 crash of the Southern Aurora into a goods train at Violet Town in northeast Victoria. (Upgrades to safety systems should ensure such a crash doesn’t happen again.)

The CFA has published a set of photos and a fascinating article about the disaster — well worth a read.

As it happens I went through Violet Town just after Christmas, with time to spare, and snapped a few pictures.

The station, about a kilometre north of the accident site, is a very peaceful spot.
Violet Town station

I assume as the result of remodelling works, this odd crossing within the station grounds has been half fenced-off, but is open from one side.
Violet Town station

On the citybound platform are recognisable Myki reader mounting points and one of the big Myki/Kamco equipment boxes — with the change to the Myki rollout plan in 2011, this and other stations beyond the V/Line commuter belt might never get Myki. (I wonder if anything is inside the box?) A newer Albury-bound platform didn’t seem to have the mounting points.
Violet Town station

The modern-day daytime equivalent of the train that crashed, a southbound Countrylink XPT zoomed through… notice how much it bounces around — perhaps a result of continuing problems with the track upgrade on the line.

Train myths: a hundred years ago the Ballarat line was quicker – No it wasn’t

The Committee for Ballarat has an excellent campaign going to improve rail services. Unfortunately as sometimes happens, some myths crept into the rhetoric around the launch of the campaign.

“In the early 1900s you could do the Ballarat to Melbourne trip by steam rail in an hour — these days it takes an hour and a half.”

Committee for Ballarat — quoted in the Ballarat Courier

It’s tempting to automatically assume things are crap now, and were better back in the “good old days“. But this is entirely wrong.

Southern Cross station, V/Line train

We know this because Mark Bau’s excellent web site of old timetables includes a timetable from 1905. This shows, for instance, a stopping train departing Spencer Street at 7:40am, arriving at Ballarat at 11:08am — a trip of some 3 hours and 28 minutes.

An express train (stopping only at Melton, Bacchus Marsh, Ballan, and Ballarat East (quite similar to many express services today) departed at 4:40 and arrived at Ballarat at 7:25, for a travel time of 2 hours and 45 minutes.

Nowadays a typical train stopping along the way might be the 9:07am, arriving Ballarat 10:33 (1 hour and 26 minutes), whereas one of the fastest appears to be the 4:36pm express train, arriving at 5:41pm, or 1 hour and 5 minutes.

In other words, it’s now about twice as fast to catch a train to Ballarat as it was a hundred years ago. (And the services are more frequent, too, though only about hourly at off-peak times.)

Of course, this doesn’t invalidate the aims of the Committee’s Fast Track campaign, which aims for more frequent trains, more reliable services, faster travel times and better mobile phone coverage along the route.

V/Line’s possible timetables circa 2021

Following up my post the other week about proposed possible Metro timetables in 2021

The other thing that’s become apparent from the Travel Demand document is how the Regional Rail Link (and the other V/Line routes) could work.

Frankly it’s a relief they have some idea of how RRL would run, as so far they’ve been unwilling or unable to articulate what type of services will use the new line, which is pretty poor considering it’s costing them billions of dollars to build.

V/Line train approaching Clayton station

Caveat: this information from the document is a 2021 plan based on current and confirmed infrastructure. Doesn’t mean the funding for service upgrades will eventuate, or that they’ll be in place as soon as the Regional Rail Link project is completed in 2014ish.

Also: it appears that longer distance trains are not included here, and would be in addition to those listed below. And note that the stops on the electrified network are annotated as being “set down only” (inbound).

And I must emphasise that the below descriptions are my interpretation of the document, and are not guaranteed accurate. (eg I might have messed something up.)

Geelong line

In peak hour:

  • 4 trains per hour: Marshall, South Geelong, Geelong, Nth Geelong, North Shore, Corio, Lara, then express to Footscray, then Southern Cross.
  • 4 trains per hour: Geelong, Nth Geelong, Lara, Little River, Wyndham Vale, Tarneit, Footscray, Southern Cross.

In the inter-peak hours you’d have:

  • 4 trains per hour: Marshall, South Geelong, Geelong, Nth Geelong, North Shore, Corio, Lara, Little River, Wyndham Vale, Tarneit, Footscray, Southern Cross.

So in off-peak hours, a big upgrade from today’s hourly service — close to a suburban level of service, in fact — just as well because it includes the new suburbs of Tarneit and Wyndham Vale. A train every 15 minutes is not too bad for those areas (but of course for it to be workable there needs to be frequent connecting buses for those not lucky enough to be close to the two new stations).

What’s less clear is how the longer route will impact on travel times. Potentially the expresses could move through the new line at up to 160 kmh, with probably closer to 80 kmh on the inner-suburban section. That could overall be as about quick as the existing route, but it’s not really clear if this will be possible given the operating patterns and other constraints.

No hint as to what will be provided for the current people travelling between Werribee and Geelong, though some kind of shuttle bus from Werribee to Wyndham Vale has been mooted in the past. For some (eg between Geelong and the Newport and the Altona area) it might be easier to double-back via Footscray.

And of course the other issue worth pondering is if Tarneit and Wyndham Vale boom, will we have the same problems there that we currently have at Sunbury — suburban loads trying to fit into regional/long distance services? RRL apparently includes provision for additional electrified tracks, but it’s unclear if these would ever be funded for construction.

Ballarat line

In peak hour:

  • 2 trains per hour: Ballarat, Ballan, Bacchus Marsh, Footscray, Southern Cross. (Some to also stop at Melton)
  • 3 trains per hour (implied 5 over 2 hours): Bacchus Marsh, Melton, (Caroline Springs), Deer Park, Ardeer, Sunshine, Footscray, North Melbourne, Southern Cross.


  • 1 train per hour: Ballarat, Ballan, Bacchus Marsh, Footscray, Southern Cross.
  • 2 trains per hour: Melton, Rockbank, (Caroline Springs), Deer Park, Ardeer, Sunshine, Footscray, North Melbourne, Southern Cross.

Not terribly ambitious, but a half-hourly service to the western suburbs such as Melton and Deer Park, Ardeer would be a big improvement over the service there now, which is roughly every two hours in some cases.

But what would you do if you want to travel between Ballarat and Melton?

It’s not clear how many trains would originate at Wendouree.

Bendigo line

Bearing in mind the Sunbury line will be electrified by this point, in peak hour:

  • 1 train per hour: Bendigo, Kangaroo Flat, Castlemaine, Malmsbury, Kyneton, Footscray, Southern Cross
  • 1 train per hour (implied 3 over 2 hours): Bendigo, Kangaroo Flat, Castlemaine, Malmsbury, Kyneton, Woodend, Macedon, Gisborne, Riddells Creek, Clarkefield, Sunbury, Footscray, Southern Cross
  • 1 train per hour: Kyneton, Woodend, Macedon, Gisborne, Riddells Creek, Clarkefield, Sunbury, Footscray, Southern Cross


  • 1 train per hour: Bendigo, Kangaroo Flat, Castlemaine, Malmsbury, Kyneton, Woodend, Macedon, Gisborne, Riddells Creek, Clarkefield, Sunbury, Footscray, Southern Cross

An hourly service outside peak is not great. At least the addition of Sunbury to the suburban network should see the end of trains that are crowded between the City and Sunbury, but relatively empty further out.

It’s not clear how many trains would originate at Eaglehawk.

Seymour line

In peak hour: 2 trains per hour: Seymour, Tallarook, Broadford, Kilmore East, Wandong, Heathcote Junction, Wallan, Donnybrook, Broadmeadows, Southern Cross

Inter-peak: 1 train per hour, same pattern.

Gippsland line

In peak hour: 2 trains per hour (implied 3 over 2 hours): Traralgon, Morwell, Moe, Trafalgar, Yarragon, Warragul, Drouin, Longwarry, Bunyip, Garfield, Tynong, Nar Nar Goon, Pakenham, Dandenong, Caulfield, Flinders Street

Inter-peak: 1 train per hour, same pattern.

Thoroughly unambitious; roughly the same as now. And why no stop at Clayton (or even better, given the 601 bus now in service) Huntingdale for Monash University?

Note it says these trains would terminate at Flinders Street, presumably to reduce congestion on the viaduct to Southern Cross, as well as avoid terminating trains there, using up limited platform capacity… at the cost of regional travellers needing to change trains to reach Southern Cross.

Interchange at Richmond and North Melbourne

We already know that trains using the Regional Rail Link lines (Geelong, Ballarat, Bendigo) will not have interchange at North Melbourne, because someone’s decided that having spent a bazillion dollars on the project, they can’t spend a little more and do it properly, with platforms to serve this major interchange.

It also appears (from the document at least) that the Gippsland line will not stop at Richmond.

It’s unclear what passengers requiring interchange from suburban lines or the City Loop will do. Depending on how the Loop operates (and the document provides some clues on this too), it could be okay, or it could make life very difficult for some.

Already it’s an issue for people from the northern/western suburbs on weekends, for instance, who have to catch a suburban train that goes all the way around the Loop before arriving at Southern Cross, because V/Line trains don’t stop at North Melbourne on weekends.

One can only hope they’re looking at this carefully, and will provide interchange opportunities where they’re needed.

How many catch V/Line in peak hour?

Last night on the TV news they seemed to be struggling for an accurate figure of how many were affected by the closure of the Geelong line. One said “hundreds”, another said “up to a thousand”.

Figures on V/Line’s web site, which summarise the number of people on each train so you can plan your trip to avoid the packed ones, indicate that about three thousand catch the Geelong line each peak hour.

The figures appear to show 100% when the services are over-capacity — eg when people end up standing or sitting in aisles on the trip.

POTD: Overcrowding on V/Line

Looking at all the lines, the figures (into Melbourne before 9am; out of Melbourne between 4pm and 6:30pm) are:

I knew they’d grown strongly since the Regional Fast Rail upgrades were completed mid last decade, and the 2007 price cut, but I’m almost surprised to see the Bendigo and Ballarat lines up within about 10-15% of the Geelong figures. This probably emphasises why V/Line and the Department of Transport have been so keen on the Regional Rail Link project, to get all the busiest lines on their own tracks within the suburban area.

The Bendigo figures are likely to drop when Sunbury and Diggers Rest stations join the electrified Metro network later this year. This will also free up some carriages to run on other lines.

Obviously off-peak passengers are also affected by line closures, and we don’t have figures for them. To a greater extent than Melbourne suburban services, V/Line services are very concentrated in the peak (trains every few minutes in some cases), but quieter outside it (mostly hourly). Something they could/should do to help spread the peak load is upgrade off-peak frequencies.

PS: I see some real figures have made it into an Age Online story this morning.

Regional Rail Link works continue

There’s a good view from North Melbourne station of the works on the new Regional Rail Link line that will come in from Sunshine and the western suburbs, bypassing North Melbourne (unfortunately, with no interchange platforms) into Southern Cross. The idea is that V/Line trains will be able to bypass the suburban tracks, allowing both more V/Line and Metro trains to run.

Regional Rail Link works at North Melbourne

Nearby at Southern Cross, the new platforms are looking increasingly close to complete, though the track is still missing.

Regional Rail Link works at Southern Cross

Note the glass wall. When Southern Cross was built/renovated last decade, they did include provision for the extra platforms 15+16, which is why these have taken shape so quickly. But the glass wall on the western side of the station will actually sit between these platforms.

So platform 15 will be inside, and platform 16… well, that could be a little chilly and wet on cold rainy days.

Regional Rail Link works at Southern Cross

There are still questions about the overall project. There’s still little or no public information on an operating plan of any kind — which should be a prerequisite before you start building.

That is, you should work out what train services you want to run, then build the infrastructure to allow it. We still don’t know if the V/Line trains originating in Geelong will stop at the new stations in Wyndham Vale and Tarneit. We don’t know if passengers at Deer Park and Ardeer will get any extra trains stopping. We don’t know if Geelong trains will take longer on their trip, coming into Melbourne the long way around, even if the tracks they use are faster.

It seems the project wasn’t planned that carefully — despite being one of the most expensive infrastructure projects ever undertaken in Australia.

Victorian transport department secretary Jim Betts said at a conference last week that the $5 billion Regional Rail Link, which has blown out by $1 billion, was budgeted for haphazardly. ”The budget for that project was basically haggled over between the state and the Commonwealth one weekend and we end up with a number written on the back of an envelope,” he said. It was reported in the Australian Financial Review.

The Age, Baillieu sorry for transport blunder

Hopefully that planning is going on behind the scenes. Alas, if it is, like much of the planning of our public transport network, how they’re intending to spend taxpayers’ money is being kept secret from taxpayers.

RRL: We must build more rail lines. This is a rail line, therefore we must build it.

Regional Rail Link is a $5ish billion project to separate out V/Line trains, by running the Geelong line via new stations at Wyndham Vale and Tarneit to Deer Park, then into the City (along with Ballarat and Bendigo line trains) on dedicated tracks. Yesterday it was confirmed that it would go ahead.

Regional Rail Link proposed route

The idea of separating fast, limited stops V/Line trains from slower stopping Metro trains is a good one, of course.

But there seem to be a lot of people thinking that this combined with the new stations in growing areas automatically means the whole project must be a good idea.

Problem is there are a number of big questions about it that remain unanswered:

What service will the new stations at Wyndham Vale and Tarneit get? Will it be the kind of appalling V/Line suburban service already seen at places like Deer Park, Ardeer and Rockbank, where outside peak hour there’s a train only about every two hours? (Indeed, will Deer Park and others see any better service from this?)

Will the peak hour service to those new stations be adequate for the expected commuter population?

Will Geelong trains double as suburban trains to those new stations, and if so will they cope? (Are we setting up to repeat the problems seen now on the Sunbury line?)

Will residents along the line only be able to use the new stations if they are lucky enough to live within walking/cycling distance or get up at the crack of dawn to get a car spot? Or will good quality connecting buses be provided?

Southern Cross StationHow much longer will Geelong trains take to get to the city once they detour via Tarneit? (It’s theoretically possible to build the line so the expresses move through at 160 kmh, but it’s not been stated that it will be the case.)

What will Geelong to Werribee passengers (apparently several hundred per day make the trip) do once their trains are diverted away from Werribee? Maybe (if they can) they’ll just get in their cars and join the traffic.

Will V/Line trains still stop at North Melbourne so passengers can continue to use the 401 University shuttle, or change to the City Loop? (They can change to the City Loop in the morning at Southern Cross, but with current loop patterns, the only afternoon choice without interchange at North Melbourne is going to Footscray, which includes catching a train on the most crowded line in Melbourne, the Sydenham/Sunbury line).

Will there be problems with V/Line trains and Metro trains on the Bendigo line on the 15 kilometres between Sunbury and Sunshine, where they will still share tracks? (Ditto between Wyndham Vale and Sunshine, if it’s a mix of express and stopping trains. And the project won’t touch Seymour/Shepparton trains between North Melbourne and Craigieburn.)

And finally, given the huge cost, were other projects considered, such as extra tracks on the Werribee and Sydenham lines, and Smartbuses or bus ways into the Tarneit area from nearby stations?

Maybe someone’s looked at all these issues, but nothing’s been announced. Despite the multi-billion bill being presented to taxpayers, there’s no transparency.

Good public transport planning should dictate that you first work out what services you want, then you work out the infrastructure needed to provide them. Given continual responses of “we don’t know” when basic questions have been asked about, say, what the line’s timetables will look like, it appears the planning for this is going backwards — and still hasn’t been completed.

While it’s clear there will be some good benefits from the project, it’s less clear that it’s the best option available, or that it’s good value for money.

Fare Free Friday

With public transport free today, it looks like some are taking advantage — there seemed to be more Seniors on the train this morning, enjoying a free trip into the city (even though it’s only $3.40 normally).

But overall the train wasn’t markedly more crowded than usual, and it’s not like everyone abandoned their cars for the day.

Traffic on Latrobe Street
Latrobe Street, 9:30am

Many simply don’t have PT that is anywhere near time-competitive with driving, even if it’s free.

So what’s the real cost of today’s Fare Free Friday (making up for, as one wag put it, No Trains Tuesday)?

The Minister said it would be “something upwards of $1 million”. I think it’s closer to two million.

How much is fare revenue?

Fare Free FridayFare revenue formulas have changed now, and I haven’t dug around to get more recent exact figures, but the old formula from last year is a quick easy way of checking it. The old formula was a 40/40/20 split; 40% to Connex, 40% to Yarra Trams, and 20% to the government to pay for buses.

Page 7 of Track Record 40 (July-Sep 2009, the last full quarter before the new contracts came in) shows a farebox payment of $60.95m to Connex for the quarter ending Sep-09, that being their 40% share.

That makes total metropolitan revenue for the quarter about $152.4 million, or about $609 million per year. (One could use the exact payments made to them during 2009, but there’s a steady upward trend as patronage increases, so it’s probably fairer to use the last quarter figure x 4.)

Page 8 also shows the V/Line fare revenue, which is $18.19m for the quarter (again, it’s trending upwards), or about $73 million for a year, making total metropolitan+V/Line revenue about $682 million.

So how much is free travel costing today?

Divide that by 364 days per year (because Christmas Day is always free) and you get $1.87 million — and that assumes that weekday revenue is the same as weekend, which it isn’t.

How about we assume that weekday revenue is 50% higher than weekend (though it’s probably much higher). Taking into account ten public holidays, that would make it $2.09 million for a weekday, and $1.39 million for a weekend day.

One could theorise about how many people will inadvertently touch-on or validate and end up paying, and how many weekly/monthly/yearly and V/Line ticketholders will be claiming their free day’s travel.

And one could also theorise that Authorised Officers (who check tickets) may not be rostered-on, or may take the day off, since they would have nothing to do — though in fact a large number of them help with general customer service at big events like tonight’s football (another reason more revenue will be lost today than the average day) and planned disruptions for improvement works.

Some people may have moved their trips/outings/errands from another day to today, adding to the lost revenue.

All in all I think it’s not unrealistic to say it’s probably around $2 million in cost to run PT free today.

Could the money be spent better? I have no doubt people will claim their compensation, and some will enjoy travelling free today. And politically, the government had to eat humble pie. But I wonder if that $2 million might have been better spent on maintenance or upgrades, or planning to make the network more resilient to be able to isolate faults better in the future.

(Heck, a few more free days would pay for a station at Southland.)

Another thing: the free PT debate

The above figure of $682 million per year is the shortfall you’d need to make up if you made public transport free all the time. Less costs of Authorised Officers (inspectors) and running the ticket system ($135 million per year for the next ten years, and theoretically about $50 million after that). Let’s say about $500 million per year in funding you’d have to find to make the system free.

And as illustrated by the picture above, plenty of people would still drive — even into the CBD, because they don’t have good enough services to use, at any price.