Melbourne Rail Link: has it been properly planned?

As I’ve written already, both the Metro Rail Tunnel and the Melbourne Rail Link provide similar benefits in terms of rail capacity in the central part of Melbourne’s rail network. In those terms, they are roughly equivalent.

But MRL does have problems. For example, I think it connects the wrong lines.

Connecting lines

Both MRL and the Metro Rail Tunnel (I’m going to abbreviate it as MRT, in lieu of another convenient acronym) create an extra track pair through the city, connecting two existing lines together, freeing up capacity elsewhere. The whole idea is to isolate rail lines, to let them run independently.

MRT does it by connecting the Sunbury and Dandenong lines through a tunnel via Domain and Parkville, creating a cross-city connection.

MRL does it by connecting the Frankston and Ringwood lines through a tunnel via Montague and the City Loop, creating a connection from the south to the east, via north and west of the CBD.

Such a connection has obvious impacts on passenger movements. Let’s look at one example: Richmond.

Interchange at Richmond

Does anybody want to take the long way around?

Richmond is a major interchange, but isn’t significant as a destination at peak hour. It is during big events in the sporting precinct. As such the handling of big crowds is a huge issue, to the point where special measures are often in place to deal with the large numbers of people, particularly just after events conclude.

The current trip from South Yarra to Richmond is 2 minutes. Via MRL it could easily be 15 minutes or more. And remember, it doesn’t just affect the Frankston line — it also affects passengers at the MATH stations (Malvern, Armadale, Toorak, Hawksburn), who in the future are likely to not have Dandenong line trains stopping at their stations.

A trip from Malvern to Richmond is currently 8-11 minutes, depending on stopping patterns. Via MRL it’ll be around 21-24 minutes or more.

So then, what is the consequence of this?

Is this extra travel time enough to prompt large numbers of people to try and change to another train to avoid going all the way around?

At times of big sporting events, when the train system is trying to shift the bulk of a 100,000-strong crowd out of the MCG, will the Dandenong and Sandringham lines be completely swamped by Frankston line people trying to get home as quickly as possible?

Will people who are actually trying to make a connection from the south to the east be happy to take the 15 minute detour (say, students heading to Swinburne in Hawthorn), or will they also want to use the other lines to cut their travel time? What effect on dwell times (and thus, track capacity) would there be from large numbers of people changing trains at Richmond and South Yarra?

I don’t know what the answer to these is, but you’d hope they’ve been looked at.

Flagstaff station, peak hour

Lots of other issues – have they been studied?

This is only the tip of the iceberg. There are questions about how they’d build extra platforms at/under South Yarra (something MRT never resolved, and so ruled out).

How will it get under the river? Where will the extra platforms at Southern Cross be? I expect there are solutions to these (including the sewer), but have they been worked out, and costed?

Is Montague, with scads of potential users, really more important than Parkville, which has scads of users now? If the rail tunnel can’t run north-south to relieve St Kilda Road trams, what will be done to help them cope? I can think of lots of potential ideas on that one, but it’s unclear if this has been properly thought about and measured against prospective demand.

All trains not serving all CBD stations is inevitable, but what is the likely passenger demand from the Frankston and Ringwood lines for Flinders Street, and where are those passengers likely to change? Are the Frankston and Ringwood lines likely to be well-matched as far as future passenger/train demand goes? Ditto the other pairs: Sandringham and Newport, Sunbury and Dandenong.

And this is the real problem: the Metro Rail Tunnel project has its faults, but has had years of study (much of it published) done into it, part of a broader network development plan that studied not just where the tunnels would go, not just the number of trains flowing through, but also the effects of different upgrades on where and how passengers travel.

MRL in comparison has come from nowhere. There’s a complete lack of evidence that it’s gone through the kind of thorough study and planning that a multi-billion dollar project should have to get to a point where the State Government is funding it*.

Maybe in the few months since the project materialised, all these questions have been resolved. Maybe. But it doesn’t look like it.

That doesn’t bode well for avoiding cost blow-outs, nor for Melbourne getting the best solution for the billions that will go into it.

  • *The State government might claim MRL is fully funded, but the budget allocations so far are minimal — $40m this year, $50m in 2015-16, $140m in 2016-17, $600m in 2017-18. So around 90% of the cost of it is as-yet unfunded. It could easily slip off the funding radar after a couple of years, in the same way the Metro rail tunnel has.
  • In comparison the East West Link, you know, the road they said before the last election they weren’t even thinking about building, is being pushed along. For instance, the western section has $3.2 billion against it by 2017-18.

By Daniel Bowen

Transport blogger / campaigner and spokesperson for the Public Transport Users Association / professional geek.
Bunurong land, Melbourne, Australia.
Opinions on this blog are all mine.

16 replies on “Melbourne Rail Link: has it been properly planned?”

I’ve got two contrasting theories about this, neither of which I’ve been able to verify unfortunately. The first is that this alignment was on the cards a while ago, but was put in the bottom drawer because it seemed the less attractive proposition during the Eddington study, and like almost all infrastructure documents published by Victorian governments, “options papers” are not what is done, and instead single projects are put forth and then investigated further. The project was then dug out later because the current mob wanted to look like they’d been doing lots of work on public transport and wanted their own “stamp” on the project.

The second is that there was simply a ministerial brief sent to DTPLI stating they needed to quickly come up with an alternative to the MRT for the same reason.

Either way as you say Daniel, there is very little evidence this project has had the due diligence. Also, its neigh time Victorian started looking at options papers. “This is how much projects A, B, C & D will cost, this is the their benefits, this is their drawbacks.” etc.

I thought the original proposal would solve a lot of transport issues. Problems about digging up Swanston Street and the disruption is disingenuous. That happens the world over and is managed. After tonight seeing five packed trams pass by waiting St Kilda Road passengers, and knowing how a direct underground line from Domain Interchange to Flinders Street Station would solve this matter, I despair at the alternative. The original proposal ticked so many boxes. I don’t see the more recent one ticking many at all.

As someone with previous experience on the network, I can’t understand how separating lines actually helps in any way. Surely having the flexibility to have a train from Dandenong weave its way across the points at Richmond and form a service out to Belgrave due to a train fault or late running is using the network to its full benefit. Surely the controllers have the knowledge and skill to be able to utilise trains the best way in order to keep the timetable. Something I think has been done quite well in previous years. By separating lines you are forcing two lines in their entirety to run late if one is late (as we see with the current separation in the system, something that’s sure to get worse in future years if these plans go ahead). If a train from Frankston is running 10 minutes late, it means the Werribee service it will form probably departs around 5 late. The controllers would be better off taking a train from elsewhere to keep the Werribee on time and use the service from Frankston to form something else. This used to happen a lot, not so much now. Of course in a perfect world everything would run right on time and fanciful line separations may work, but any realist knows that’s never going to happen in Melbourne. There’s too many disrespectful people who cause delays (ie rock throwers, people forcing doors, trespassers, threats to self harm, plus signal and equipment faults etc)

South Yarra presents several issues. The biggest is the capacity of the station entrance and concourse. It already has crowding problems and needs a northern concourse in the near future, with an entrance to the Forrest Hill precinct (including Melbourne High) and possible Darling St.

The second is that it would be a very hard station to integrate MRL/MRT platforms into and this potentially would add to the Toorak Rd entrance capacity issues.

The non-yunnel solution of St Kilda Rd tram capacity is simple. Just quadruplicate the tram track in St Kilda Rd between the junction and Flinders St and Flinders St between Swanston St and Elizabeth St and then through route the Elizabeth St trams with some St Kilda Rd routes.

There’s been a lot of question over how issues like you’ve raised at Richmond for peak events will be managed, and the simple answer is they still have the ability to run special trains – much like they do to Flemington.
Saying the lines are completely disconnected isn’t quite genuine – they’ll still of course have routes which they can use to switch trains between tracks, and that includes around these major areas. This will allow either a train to run directly through Richmond as a special service, either as a complete service or to a certain point on the adjoining line where people can then swap trains down the line on an adjacent platform.

What truly bugs me is the amount of time it takes to complete construction projects in Australia. It’s for this reason that St Kilda Road was ruled out – if only we could complete projects quickly it would cause less disruption and create a significant number of jobs. Both St Kilda Rd could be pulled up as a cut and cover project, as could Burwood Highway and Doncaster Road. The alternative is to elevate trams down St Kilda Road, freeing up that space as traffic, with the outer lane segments for cycle traffic – or vice versa.

@Ben, separation of lines reduces flow-on effects of delays. Every time a train crosses tracks, that can delay another train. Reducing this helps punctuality, especially in a situation where you’re aiming to run trains every few minutes to cope with demand. If there are no regular conflicting moves, then you can throw lots of trains (up to about 30 per hour) at the system and it runs fine.

Yes, there are obviously implications if two lines are through-routed, but what we have now is the situation where (on weekdays), Werribee, Williamstown, Frankston, Pakenham and Cranbourne all interact (as well as the Geelong/Warrnambool line), and these have single-track sections. Delays can quickly spread. Separating them out to even two lines in a group limits that.

@Tim, special event services make sense in the Flemington/Showgrounds context where there aren’t normal services. But they make less sense in the context of existing frequent services, particularly at peak times such as around the time crowds head into the MCG for Friday night football.

There’s also the question of how people know there are specific trains running, and whether they should have to wait for them. eg there are trains every 10 minutes on weekends, but if you are going to the footy you’ll have to wait 30 mins for your special direct train. Would people do that? It’s better if they can know all trains go to X, and they can just hop on the first one that arrives – no matter if they’re arriving early for the Reserves game, or staying late for a drink with mates afterwards. It also keeps it simple, and also lets people see the potential for them making other trips on PT.

Similar: my experiences going to PAX at the Showgrounds last year. The special train services were infrequent, and only suited people who wanted to stay all day. Plenty of people didn’t, and had to find the tram stop (and crowded trams) instead.

It’s interesting that if the MRT were to go ahead, there would presumably no longer be a need for Platforms 3 and 4 at Richmond. Therefore the remaining platforms could be made wider as part of a possible redevelopment of the station and associated trackwork.

That’s not to say that MRT doesn’t have other issues, of course; I’m just putting it out there.

Hi Dan,

There is definitely a lot to be desired about the current state government’s approach to transport policy (‘policy on the run’, trying to rush through projects with insufficient planning – see exhibit A, East West Link), and there are indeed concerns about the MRL proposal, but I think you’re looking at the wrong areas in the above.

Firstly, the Richmond issue:
No matter which party’s proposal you look at, one southeast line (i.e. Sandringham, Frankston or Dandenong corridor) will lose out it’s direct connection to Richmond. Yes, that will mean that those passengers will either need to backtrack around the CBD to get back out again, or get on an alternative train out of Richmond then change further down the line, but this is the case for either proposal. In fact, the Libs’ MRL proposal means that FKN line passengers can get on any Sandringham train OR Dandenong train and change at Sth Yarra or Caulfield to a FKN line train; the Labor proposal does not include platforms at Sth Yarra, meaning those passengers (Dandenong line passengers, if Dandenong is routed via the new tunnel in Labor’s plan) only have the option of a FKN line train to get out, because a Sandringham train will no longer connect with Dandenong at Sth Yarra. In other words, they have LESS options than the MRL proposed.
On top of this, as the Frankston line services will no longer use the track between South Yarra & Richmond, and as the Dandenong corridor will have had its signalling upgrades, we can assume that there would be a greater number of services in peak times (peak hours, or after MCG events) from Richmond out to Sandringham and Dandenong for Frankston passengers to hop on then change further down the line, due to the freed up capacity on these lines.

On the ‘lack of north-south corridor’ in the proposal (i.e. no train up St Kilda Rd from Domain to CBD south/north):
The MRL does include a north-south corridor, it just follows a different route from the Domain through to Southern Cross. Anyone who’s gone near St Kilda Rd on a weekday morning knows that there are serious capacity issues that trams simply cannot address, but to suggest that the MRL does not address this is false. It will just be that those passengers coming from suburban train services who currently change at FSS for a St Kilda Rd tram will instead change at Southern Cross for a train to Domain. This will free up capacity along St Kilda Rd for those who are either coming from Swanston St itself, or connecting from other east-west tram lines along Collins, Bourke, La Trobe Sts. This also means we’re moving the connecting traffic from the overcrowded FSS to the much bigger Southern Cross, which can and should take on additional connecting capacity. This won’t increase commuters’ travel times, rather just change the routine of where they change services.

I do certainly agree that the hospital/university precinct is the loser of the MRL, as is the Arden precinct, but it is not impossible to address this with increased tram services along Flemington Rd and Royal Pde, along with the increased train services along the Craigieburn & Upfield line that should come with the RRL coming on line next year.

We also have to remember that the tunnel’s aim was in any case to increase capacity through to CBD, not to service Richmond after football matches or address tram crowding around Melbourne Uni. Those were just periphery benefits, and I must say I’d much prefer the periphery benefit of a supposedly cheaper proposal that allows some $$ to build an airport rail link than one that services the university precinct that I can already reach with frequent tram & bus services along with some of the best bike lanes in the inner city.

Finally, when we think about future passenger demand, the MRL is arguably designed to better meet that. The heart of the CBD is slowly gravitating towards the areas that are experiencing huge residential and commercial development right now – along Spencer St, Docklands, Southbank, and, eventually, Fisherman’s Bend. The commercial spine along Swanston St will lose its relevance as populations increase around these other areas, and with it, the current gravitation towards the small and increasingly expensive and crowded pocket between Melbourne Central and Flinders St Station will move away.

The best thing about the MRL and MRT is that if the MRL goes ahead what is to stop the MRT going ahead afterwards and having both systems?

One of the other issues around MRL is that it effectively rules out one of the future options for capacity increases to the network. The Network Development Plan anticipated a Stage 4 through routing of the northern and Caulfield loops. MRL does this now (albeit Caulfield and Burnley).

There may be other cost effective ways of delivering that long term capacity benefit, but they certainly haven’t been spelt out, and I’d guess not really considered at all as part of the MRL-on-the-run decisions.

Makes a mockery of the libs coming to power promising to establish an independent PT agency (PTV), only to completely disregard what was a logical, sensible, and well recieved strategy for progressively adding capacity to the network. With all the hoopla about MRL, there’s not even been any discussion about reviewing the NDP – so how do we know if we’re losing the most effective longer term capacity benefit by throwing all our eggs into MRL

I only wish we could have both rail projects done.

That way everyone wins. Sadly the major issue is the cost of digging a tunnel, and the limited funds we have for that.

Arh well we can always dream hey.

I wish they had metro out to gisborne because most people who live in gisborne use sunbury I do both depends on if I am with pop (i am a asberges 10 year old who wants to become part of the public transport not traffic campaign) my pop uses sunbury mummy and daddy use sunbury but if they had metro out to gisborne they wouldn’t have to pollute nature by driving 15 minutes and they could catch bus to station and the timetable would be
weekend 9-10
sunbury-city 20 min
wartergardens-city (runs direct to flinders street are services only going to wartergardens) 10 min gisborne to city 30 min.

Comments are closed.