The Five Group Railway – good or bad for passengers?

The “five group railway” is something that’s been on the cards for a while. The Age highlights it today in this article: Metro plan to split Melbourne rail network into five lines hangs on union fight

It has its origins last decade — the 2008 Victorian Transport Plan talked about “creating a Metro system”, in the context of strong patronage growth and the evolution of Melbourne’s commuter railway into one that can cope with far more passengers than it can currently handle, reflecting Melbourne’s growth, and in particular the CBD and inner-city.

Melbourne’s railway system was designed as an old style commuter railway, carrying people to and from the city with branches, junctions and single track lines to maximise reach. The City Loop was a major boost to the rail system at the time but the four tunnels have to cope with trains from 10 separate inbound lines

The operation of this sort of railway is complex, the capacity of lines is not maximised and reliability of the service overall falls. One delayed train can affect the whole system.

Melbourne will soon be at the point where we cannot run more train services on key lines. Metro rail systems are designed to run higher capacity trains from end to end of lines using dedicated tracks – the trains can run at higher frequency without interfering with other routes. The focus is on simple timetables, frequent services and consistent stopping patterns.

Metro systems like those in London and New York have key interchange stations to allow people to change trains easily or switch to trams and buses to get to where they want to go.

PTV’s plans for this have also developed over the years. It’s reflected in the 2013 Network Development Plan — this map shows the rail network as it is now (apart from delayed Southland station), with the opening of Regional Rail Link. As you can see: five groups. (Yes, they left Stony Point off the map… I assume because it doesn’t go into the City.)

PTV rail Network Development Plan - stage 1

Metro (to be precise, the current operator, Metro Trains Melbourne, MTM for short) has gone in hard supporting the idea, and bit by bit, is splitting the network into the five groups. The first four are named after the four Loop tunnels, though not all trains would use the Loop:

Northern — the Sunbury, Craigieburn and Upfield lines, all using the Northern Loop.

Clifton Hill — the Hurstbridge and South Morang lines, all using the Clifton Hill Loop.

Burnley — the Belgrave and Lilydale lines, using the Burnley Loop, and also the Glen Waverley and Alamein lines, running direct to Flinders Street.

Dandenong — the Pakenham and Cranbourne (eg Dandenong) lines, using the Caulfield Loop.

…and the fifth…

Cross-city — the Frankston line running direct to Flinders Street then out to the Werribee and Williamstown lines. And also the Sandringham line, direct to Flinders Street and likely to be fairly independent. (In practice some Craigieburn trains share the Cross-City tracks into Flinders Street.)

Independence is the key. At present trains and drivers move from line to line across the day, meaning a delay on one line can quickly flow to another. In fact MTM claim in peak the reduction in interdependencies could mean a punctuality improvement of up to 10%.

From that point of view, separation makes sense for passengers.

Taking it further

Behind the scenes, PTV and MTM seem to be moving towards partial segregation of the train fleet. The High Capacity Metro Trains project will see Dandenong line get its own dedicated fleet of trains, with a maintenance centre at Pakenham, though the same model of train is likely to roll out to other lines later on. More on this in a minute.

But MTM want to take it further, and this is where it starts to get controversial.

Complete operational independence of the groups, including separate stabling and maintenance facilities

Removal of some points that provide connections between groups of lines, which Metro says will speed up services (though in some cases only by ten seconds or so), and reduce track faults
This could help, but of course it comes at the expense of flexibility, such as being able to route trains around obstructions, and limiting the places trains can be terminated during disruptions — leading to longer-than-otherwise sections of line replaced by buses. They wouldn’t want to go too far.

Fleets dedicated to specific groups: in 2013 the plan was for Cross-City to be Siemens, Comeng for Dandenong and Northern, X’trapolis for Burnley and Clifton Hill. Since then the plan has changed to get X’trapolis trains onto the Cross-City lines as part of the Bayside Rail project, which means more Siemens will presumably move onto the Dandenong and/or Northern lines until the new fleet arrives
Provided they don’t actually take steps to make parts of the fleet incompatible with some lines, or with each other. Over time, trains will need to be cascaded through the system as new ones come into service, this makes some sense, as it means more consistent performance (eg acceleration) on each line, and should simplify maintenance.

Restriction of most drivers to specific groups… including not training them to drive on the other groups (thus cutting training costs and time). Drivers don’t like this, though it’s not unusual on networks overseas. MTM would keep a central pool of drivers qualified for all routes for disruptions and other operations.
While getting drivers trained would be quicker, perhaps meaning they can ramp up additional services more quickly, for passengers this could be a problem if the central pool of fully-qualified drivers isn’t sufficient, leading to worsened delays during disruptions. Perhaps the answer is to simply not swap drivers between groups on a particular day? Not sure.

Driver de-centralisation, where most drivers start duty at outer termini rather than changing over at Flinders Street, including complete segregation of staff within line groups.
Again, controversial with drivers, but it’s not hard to see the benefits of cutting delays at Flinders Street, though theoretically it should be possible to swap drivers there pretty quickly. It largely happens today on the Clifton Hill group.

Some claim that MTM wants to run each group as a separate company — in fact Labor claimed that the original 2014 Dandenong line upgrade proposal was basically geared to be a PPP whereby MTM gained control of the line long after their contract for the rest of the network might have ended.
This of course makes no sense to passengers — we’ve seen the problems before of separate companies running separate lines, leading to competition between operators, fleets made incompatible with some lines, and unnecessarily inflexible operations.

Belgrave train arriving Southern Cross


Many drivers are campaigning against some of these changes, as seen on the MTM Memes blog, for instance this post: Bungled timetable or saving us from Liberal deceit? — I don’t necessarily agree with everything in it, but it raises some interesting points.

I suspect there’s a balance to be found here. Day to day operation, including high frequencies, predictable patterns and reducing the cascade effect of delays across lines would absolutely benefit from operational separation.

Less useful to passengers might be the kind of organisational separation that MTM seems to be pushing for.

It’s going to be interesting to see how this plays out.

PS. Thursday morning: This blog post a couple of years ago talked about limiting drivers to specific lines. Note the extensive comments from a train driver. Also worth noting, there are currently limitations on how many times a driver can drive a train on the same section of track each day, for safety. But two things make me question the safety reasoning: firstly it doesn’t apply to shuttle runs, and secondly I’ve been told this was only introduced in the 1980s, with the opening of the City Loop. It’s not some long-held standard. It’s not hard to see how it restricts operations, and to me it would make sense for it to be reviewed.

Productivity while travelling

Peter Martin in The Age yesterday, on driverless cars:

Imagine studying, reading books, watching TV, sleeping or (legally) playing with your mobile phone on the way to work. Whichever way you look at it, the freed-up time will boost productivity.

Sounds great!

Wait a minute, thousands of us do that stuff every day by catching public transport to work.

Being productive on the train 1Being productive on the train 2

Even when it’s crowded and you’re standing, it’s generally possible to steady yourself with one hand and play with a mobile phone with the other.

Oh, and pardon me for being a little cynical, but I suspect the technology has a way to go yet given reports that it still relies largely on highly detailed mapping data rather than actually being able to “see” what’s in front of it:

But the maps have problems, starting with the fact that the car can’t travel a single inch without one. Since maps are one of the engineering foundations of the Google car, before the company’s vision for ubiquitous self-driving cars can be realized, all 4 million miles of U.S. public roads will be need to be mapped, plus driveways, off-road trails, and everywhere else you’d ever want to take the car. So far, only a few thousand miles of road have gotten the treatment, most of them around the company’s headquarters in Mountain View, California. The company frequently says that its car has driven more than 700,000 miles safely, but those are the same few thousand mapped miles, driven over and over again.

The article also notes the risk of maps being out of date, of not being able to “see” a temporary traffic light because it’s not mapped.

Of course, computing power is improving by leaps and bounds all the time, but I wouldn’t be holding my breath on this.

Nor should we assume that driverless cars will solve traffic problems (especially if having delivered you, they drive home again to a parking space) or pollution problems (which are down to the nature of the engine and its fuel).

And anyway, has anybody actually tried to work in a moving driverless car? Are they really smooth, or — even if you’re able to take your eyes off the road — is travel sickness a risk?

I have to admit, I can’t read for long on a bus — and I know I can’t read for long in a car when I’m a passenger — I’d imagine it’d be an issue in a driverless car.

The next timetable – delayed but hopefully still on the way

Today’s Age highlights the Metro timetable changes that were planned to be implemented with the RRL opening last month.

As many would know, the introduction of Regional Rail Link was meant to be accompanied by a big timetable change on Metro to take advantage of the extra capacity unlocked, particularly on the Sunbury, Williamstown and Werribee lines.

Despite the moves towards “Metronisation” (a term the government and/or Metro appear to have coined to describe the gradual change to dedicated lines, simplified stopping patterns and more frequent services), the rail network still has numerous inter-dependencies.

This means many of the proposed changes were tied together, and it appears when the government got cagey about one of the changes, the whole upgrade unravelled, which is why the only Metro change when RRL opened was one lousy service each way on the Werribee line.

At this stage, it appears the rest of the changes are postponed until the end of the year, and I hope they all still happen, because it turns out there are some great upgrades in the package.

Notably this is not the only delay to Metro timetable changes that has occurred recently — there are improvements that have been proposed, but held back under the Coalition, as well as under Labor.

Trains near Footscray station

What was planned?

Here’s a summary of the timetable changes planned for 2015 — as well as changes originally proposed to have happened during 2013 and 2014 which were postponed.

This is based on my reading of information I’ve obtained. Any errors or omissions are mine.

Sandringham — Peak: upgrade to every 6-7 mins (currently 7-8)

Frankston — Peak: expresses and stopping trains each every 7-8 mins (currently 9ish). Weekday evenings: every 10 mins until 9pm.

The biggest change on this politically sensitive line would be no more Frankston Loop trains at all. This would mean trains will run via Flinders Street and Southern Cross and then through to Newport, as currently happens most of the time on weekdays.

It would affect weekend and peak (currently every second peak train runs via the Loop). This is what The Age says the government got nervous about. This will have obvious impacts as many Loop station passengers will change onto other trains at Richmond. Will that station, and those other trains cope?

In peak I usually catch these trains to and from Flagstaff Station. But I’m usually headed for Bourke Street — it’s only slightly further to walk from Flinders Street Station, so that’s what I’ll do, and I hope everybody else considers their options carefully rather than automatically still use the same station even when it involves a change of trains that makes it a longer journey than just walking an extra block or catching a tram. Werribee passengers got used to it; we could too. Ditto Sandringham, Glen Waverley AM peak, Alamein PM peak, Frankston peak express users…

The pay-off is more trains overall. The only nagging question I have is… will they use that space freed up in the Caulfield Loop? It appears not yet, not until a later timetable change to boost the Dandenong line. So why do it now? It seems to be tied to the Werribee line upgrade, as Frankston trains will mostly through-route to that line.

Glen Waverley — Peak: every 6-7 mins. Weekday interpeak: every 10 mins (currently 15). Weekday evenings: every 10 mins until 9pm. No more Loop trains weekday PM — see above.

Alamein/Belgrave/Lilydale — Peak: every 10 mins each to Belgrave, Lilydale, and Ringwood/Box Hill. Weekday interpeak and weekday evenings: until 9pm every 10 mins to Ringwood (currently 15-20), alternating to Belgrave and Lilydale, every 20 mins to Alamein.

Note, this is a reduction in service for Alamein during interpeak periods, but an upgrade in the evening.

Upfield — Peak: the source I have says every 11/22 mins, which I assume means 11 between the city and Coburg, as beyond that the single track makes more than 3 trains per hour difficult. (This is similar to operations during the Commonwealth Games, when extra trains ran to Coburg.)

Craigieburn — Peak: every 11 minutes via the Loop, and every 11 minutes direct via Southern Cross and Flinders Street, so every 5-6 minutes combined — (currently uneven, about every 6-12 minutes, with a small number of direct trains).

Sunbury — Peak: every 11 mins to Sunbury (currently 12), plus extra services to Watergardens.

Off-peak on the Sunbury line there was to have been a doubling of frequencies in 2016, every 10 minutes to Watergardens, 20 to Sunbury, but it’s unclear if this change was meant to tie in with RRL, which was originally expected in 2016.

Note the peak pattern here: Upfield/Craigieburn/Sunbury all share the Northern Loop, and would be on an 11 minute peak cycle… mind you, the direct Craigieburn trains would also share tracks with the Werribee/Williamstown lines in peak.

Werribee — Peak: every 7-8 mins (currently about 10-12) — a big boost taking advantage of the fact that Geelong trains are no longer on the line.

Altona Loop — Weekday interpeak: every 20 minutes all the way to Flinders St (currently shuttles to Newport)

Williamstown — Weekend daytime: every 20 mins (currently shuttles), making a combined Werribee/Williamstown service every 10 mins as far as Newport.

Spreading the peak: Although the grisly detail (such as actual timetables) isn’t available, there is also talk of spreading the peak over a longer period, so in some cases the highest frequency remains about the same, but it’s provided over a longer time span, encouraging more people where possible to travel outside the “peak of the peak” busiest times.

Train exits City Loop approaching Richmond

Good? Bad? Indifferent?

Overall good, though a couple of notes of caution.

The Loop changes would be painful for some, but the pay-off (as usual) is more frequent services, roughly a 20% peak boost on some lines, up to 50% in the evenings (from two trains per hour to three) and further moves towards ten minute services all-day everyday.

And this is the key to running more trains: you can’t have all the lines converging on the four track City Loopthose tracks are pretty much full. To make better use of the substantial track capacity in the CBD, some lines have to go direct into Flinders Street and Southern Cross.

Better separation of services should also help reliability, though there may be longer dwell times at interchange stations such as Richmond and Caulfield.

They’ve already started upgrading Richmond to provide more weather cover for interchanging passengers. Granted, the shelters above the ramps make it look like a jail, but given the centre and eastern end subways are often congested, at least people will be encouraged to use the subway at the city end, even if it’s raining — since widening the centre subway is a huge project that isn’t going to happen any time soon.

No doubt the timetable is tied into a myriad of other less obvious changes which move the rail system towards the “5 group railway” as Metro calls it, designed to improve reliability and frequency. For instance the reduction in seating on many trains would also be a factor here, providing more standing space for those short hops in and out and around of the Loop — though even the new layout lacks handholds to encourage people to move down the aisles.

Better separation and more consistent operation also means the lines can more clearly shown on the new rail map, improving the overall legibility of the network.

And I don’t think you’ll find any Altona Loop people who would be disappointed to see the end of interpeak shuttles.

We know the train fleet has been increasing in size. I wonder just how many trains sit idle at peak hour while people are packed into the services that are running?

I’ll be affected by the removal of peak Frankston trains out of the Loop. And perhaps this should have been held over until a big boost on the Dandenong line. But that could be years away, and I can see the big picture. I can work with it.

Overall this package of changes is good for the rail network, and helps move us ahead towards more frequent services overall. We need the extra services this provides. So yes, there’ll be some inconvenience, but hopefully the government will get the upgrades to Richmond in place, and publicly explain the changes, how they benefit people, how those impacted can work with it — then push ahead and do it.

Visually awful, make a lot of noise

Abbott and Hockey dislike the aesthetics of windfarms, describing them as “visually awful” and “making a lot of noise”. Funny, I think the same about motorways.

Beauty is in the eye of the beholder, of course.

But motorways certainly have problems with noise, with pollution causing health problems, and with the space they take up for what they do, and cause dispersed travel patterns rather than assisting the centralised dense city cores needed to grow economies based primarily on information workers and the “knowledge economy”. They also often cost more than the benefits they produce, and almost any mobility benefits for individuals are limited to those who can drive — and peak-hour congestion/time benefits are short-lived.

The PM wouldn’t see it this way of course, but then he has a blinkered view, an irrational dislike for alternatives such as public transport. (He does cycle, but only for sport, not for transport.)

Joe Hockey is calling for a mature debate about the budget. Fair enough.

Perhaps he and the PM could participate in a mature debate in other important areas, including power generation and transport — assessing the options on their economic and environmental merit, rather than subjective attributes such as aesthetics.

The CBD bus ride that #Myki thought was in Brighton

So apparently the installation of GPS equipment to track buses stop-by-stop in realtime hasn’t helped Myki zone detection at all.

On Tuesday at lunchtime I caught a bus from Queensbridge Street (aka Casino East, the brand new tram/bus platform stop) to Queen Street.

Tram/bus stop, Queensbridge Street, Casino East

It’s all within 1 kilometre of the city centre — about as far from zone 2 as you can get. And it’s on a route with realtime information, so at least some of the equipment in the bus knows almost precisely where it is.

So, what happened? Myki charged me for a zone 2 fare.

It thought I was in Brighton, in the zone overlap area.

Myki charging: I was in the City, but it thought I was at Brighton

It seems to have got the route number right. “out” indicates it thought it was an outbound trip, though given it’s a crosstown route, I have no idea that’s correct or not. Perhaps they should have different indicators for crosstown routes, such as “se”/”nw”?

The silver lining is that the zone 1 fare cap meant I was charged the correct amount for the day’s travel: a total of $7.52.

(I normally use a Yearly Pass, but it’s run out, so I’m using Myki Money for a while.)

Zone detection on buses (and trams) has been a problem for years, and it’s only the zone changes in January that have hidden the issue for Melbourne users, but it remains a problem on regional town buses — there are regular reports of overcharging.

Clearly it’s is something they still need to work on.

Oh, and the new platform stop? Nice, though some of the bus drivers seem a little uncertain about how close to the platform edge they should stop. The bus/tram lanes seem quite effective at helping them get past the traffic.

And I wonder if, when commissioned, the realtime screens there will show bus as well as tram?