Monday, 13 March 2017

Suburban Activity Centers in Melbourne - Getting it Right with Rail

Melburnians, it's a GREAT time to be a suburban orbital rail enthusiast.

A number of quite momentous things have come to a head this week that all point in the right direction for that long time hobby-wombat of this commentator.

Firstly, we saw the election of a Labor government in WA who ran a good old fashioned Labor values campaign that spoke a lot about new rail infrastructure, particularly... you guessed it ... a suburban orbital rial line.

Also just published was this new article in Crikey calling for the same for Melbourne

This week also saw the release of the latest Plan Melbourne Refresh, which provided some very interesting new visual data about actual current job centers in Melbourne. 

Plan Melbourne: How Fresh is my Refresh?

There is much to praise in the Refresh from a pure planning/zoning perspective, but it retains I fear some of its predecessor's problems in terms of poor poor linkages between fine words about activity/employment centers and the necessary transport infrastructure and other obvious enabling factors in making these centers work to achieve the policy goals we are asking them to.

Transportation repeatedly appears in the document as an adjunct, rather than giving the impression of being core to the document. The headline transport policy for CADs is Policy 1.2.1 - "Support the development of a network of activity centres linked by transport" and it contains no actual transportation policy whatsoever.

We should already be able to answer the question of exactly how 80,000 people are going to get to work in the Monash NEIC,  and we should already be laying the necessary tracks.

The Refresh suffers for once more designating too few "Metropolitan Activity Centers" and too many "Major Activity Centers". This is a problem because

  1. A "Twenty minute city" concept would entail all residents having access to a Metropolitan Activity Center, but Plan Melbourne designates only 11 (9 current, 2 future) such centers for a city heading for a population of 8 million. This doesn't give adequate geographic coverage to enable the goal.
  2. This creates two enormous "activity center blackspots" across both South-East Melbourne and the Inner North, whereby basically by definition we're never going to be able to offer a "Twenty minute city" to people living in either geography
  3. It fails strategically in that it doesn't draw on or make any policy virtue of existing successes, instead seeking to somewhat reinvent the wheel

Plan Melburne's designated around 120 Major Activity Centers are listed as having variable development potentials, and being subject to local planning provisions. This is too much a "one size fits all" approach, providing little policy support to  facilitate development at a number of sites that actually have significant potential.

The policy and state-level planning are all directed towards the Metropolitan Activity Centers, therefore failing to designate even existing large employment centers as such seems again to be planning to fail.

Furthermore, the designation of the airport as a "transport gateway" seems to subsume its arguably more important function as an employment center.

And once again, the document fails in emphasising the necessity of developing RADIAL and maximally networked transport solutions in maximising those centers' success.

Same Data, Different Plan

Let's see what I mean by that. I find it somewhat bemusing through this entire exercise I'm using wholly Plan Melburne data and imagery to draw dots the authors of the Plan itself appeared unable or unwilling to.

What I've done here is take the employment "heat map" on page 36 of the document and plonk it into Google Earth, to create this ...  I've had to chop Frankston off these maps for display's sake, but it is included in the plan.

Employment density for Melbourne metro region - Plan Melbourne

Overlaying that with the Plan Melbourne designated activity centers in yellow and relevant designated transport gateways - ie the airport and the designated NEICs in red gives us this ...

Employment density with Plan Melbourne-designated Activity Centers

And so the  gaps in Plan Melbuorne appear staggeringly and immediately obvious even using its own data. We know how difficult behaviour change is in planning, so why are we planning to reinvent the wheel with suburban employment centers while ignoring the following already successful regions?

Existing employment centers not designated MACs or NEISs under  Plan Melbourne
Roxburgh Park

So, adding those 9 centers to the plan would look like this, with those in white below. Readers will note the far more even and judicious coverage across the geography here.

We then need a comprehensive transport plan to better enable radial travel in to those destinations. Enter Plan Wombat Refresh.

Employment density with Plan Wombat-designated Activity Centers

Transportation choices shape cities - not the reverse

I refer readers at this point quickly back to the earlier Wombat Plan.

The aim and the point here is all about the primal importance of getting your transport infrastructure right first, and planing around that in order to facilitate modern networked economies. And making that network as network as possible is what exponentially facilitates usage. We can see, for instance, and no, pedants this isn't causality, but it's interesting ... the huge employment mass radiating across inner Melbourne from the CBD correlates pretty neatly with the more "networked" components of the tram system. The "jobs effect" doesn't accrue to the same extent where single lines are extended into suburban setting. See below.

Melbourne Employment Densities and Centres with tram network

One thing I believe we can learn from all human history is that networking complex systems reaps exponential benefits. I believe Paul Mees was right to make networking an important locus of his transport planning work for this reason.

Add the heavy rail network to the picture, and we can begin to use it to identify the biggest gaps between the service the network provides and the actual transport needs Melburnians are expressing. One would like to have seen such a "needs/gap analysis" take far more primacy in all this grand planning from IV and Planning.

Melbourne Employment Densities and Centres with tram/rail networks

Some of my own wombat sniffings:

  • The network terminates just south of the major Roxburgh Park employment hub, with the Cragieburn line essentially missing the actual employment center 
  • The network provides no connectivity to the Knoxfield/Rowville employment hub
  • The network provides no connectivity to the Airport/Keilor employment hub
  • The network provides no connectivity to the Chadstone employment/retail hub
  • The network provides no connectivity to the Doncaster employment/retail hub
  • The network provides no connectivity to the Laverton employment hub
  • The network is poorly geographically optimised to cater for most of the large Monash/Clayton, Moorabbin and to a lesser extent Braeside hubs

It's interesting too that where we think of 'Bradmeadows' as being  the employment hub, the jobs are all actually just north of Broady proper in more like Roxburgh Park.

And of course, the network is as per my broken record, poorly designed to provide radial travel in to most of these employment hubs other than the CBD.

We can do better than this.

Plan Wombat Refresh - Heavy Rail Projects

So to recap from the previous Plan W. we are trying to achieve the following:

MAXIMISE the connectivity of suburban CADs
MAXIMISE the potential of rail catchments to operate radially into those centers
Provide new heavy rail coverage to existing major rail blackspots

And here's how I think that's best done.

1. Upfield-Craigieburn Extension

This is such a no-brainer. Because the rail reservation exists, and it's only 2kms of single track - though this needs duplicating already. Again, this is a massive existing employment center and currently unserviced by rail. So extending the Upfield Line to re-join Craigieburn should actually be the highest cost-benefit network expansion that PTV have on the table right now, and with this week's announcement of further land releases out North along Craigieburn, this element of PTV's existing Network Development Plan (NDP) surely warrants fast-tracking. No pun intended.

But of course this is in the NDP not because we've left a massive employment center unserviced by rail for thirty years, but rather because there will soon be CBD-bound commuters on the other side of it. Melbourne's radial mindset really does seem to have hampered our planning in so many myriad ways. In Sydney, heavy rail exists to get you to work. In Melbourne, it exists to get you to the CBD.

2. Urban Orbital Rail Stage One

This is something like what it should look like. The only dramatic change from what regular readers will have seen before is the doglegs via Roxburgh Park and West Heidelberg.

The other obvious option here is to run Donaster - Box Hill instead. To a large extent that choice would be informed by what choices were made for Stage Two below.

I favour Ringwood as a higher future potential location than Box Hill because of the tremendous rail catchment we'd be creating by doing this. It's  the only hub where we can create the "four spokes" effect by adding just one new spoke. There's very little office or industrial stock around the station already, but oodles of potential. This, I think is where we build our "Parramatta" rather than Box Hill, albeit that BH has similar natural factors favouring it IF we built an option of orbital rail that took us there.

So, the services would run

Airport - Ringwood
Metro style service, ideally smaller capacity, higher frequency, possibly driverless.

Southern Cross - Airport
Probably luggage-capacity modified three car sets of existing rolling stock types.

There would also be the option of creating the triangular junction shown at the airport to allow another less frequent SXS-Ringwood service, but probably less so if you'd be mixing rolling stock, and it's unclear there'd really be a demand or need for it.

The Southern Cross route MUST stop at commuter stations along the electrified Albion freight corridor because that runs through one of the largest existing heavy rail blackspots in Melbourne.

I don't intend to get bogged down here in construction detail. I've given some thought to how much of this could be done with combined tunnel/skyrail, where your skyrail could be plonked say down the middle of an existing traffic sewer like Springvale Rd, but I'd rather save all the issues around that for a future post.

The section Doncaster-Airport is 38.4 kms of what would need to be mostly tunnel. Doncaster-Ringwood is 9.5 kms, Doncaster-Box Hill 3 kms.

Plugging that into our very back of the envelope calculation from the previous plan, this would cost approximately
$14.37bn (Ringwood)/$12.42bn (Box Hill) at Swiss rates $24.45/20.7bn at Sydney Metro rates, and abut three times that at Melbourne Metro rates

The electrification of the Albion corridor was costed by the Liberals at around $2bn.

None of these calculations include expenses for new stations or rolling stock.

3. Urban Orbital Rail Optional Stage Two

So, we have a few options here. One would be not doing a stage two at all. In my next post I will look at alternatives to this entire stage via light rather than heavy rail, so let's set that aside for now.

Option one would run Donaster-Moorabin like this ...
Eastern Outer Orbital - Option A

Or, you could instead take the more suburban via Ringwood to Braeside, or you could do BOTH ...

Eastern Outer Orbital - Option B

Or maybe we think the expense for Dandenong and Braeside is limited, in which case this is another option, although by the  same principle as above, we're creating Melbourne's only "five spoke" employment hub in Dandy by including it ...

Eastern Outer Orbital - Option C

Or if you really were about the maximal network effect, this is building ALL the options ...

Eastern Outer Orbital - Option D

Expense-wise, Doncaster-Moorabbin is 28.2 kms, Ringwood-Braeside 30.5 kms, and Ringwood-Moorabbin 28.5 kms. So the cost would likely be north of $10bn for any of these options.

My inclination would be to build option B above in two stages, with Clayton coming first because of identified need. But all this would be dependent on how much of this you thought could be achieved by a tram/light rail solution. Which just fortunately I've also turned my noggin to recently. But more on that in the next post.

The plan at this stage leaves open the question of how best to deal with the Laverton blackspot. The obvious solution is an extension of the orbital rail to the west, but you then start to have issues with mixed rolling stock as the metro stock is only intended to run to the airport under the plan as I conceive of it.

An important footnote too, to be fair to Plan Melbourne's authors, it is possible that many of the omitted employment hubs were deemed problematic because these heavy industrial areas are quite dispersed in terms of employment destinations, and thus difficult to provide public transport solutions into and to create actual activity centers within. But some data on that, or some indication of the thinking/planning here would have been a bit more reassuring to see.

So to recap our goals again ...

MAXIMISE the connectivity of suburban CADs
MAXIMISE the potential of rail catchments to operate radially into those centers
Provide new heavy rail coverage to existing major rail blackspots

How do you think I did? Could you do it better? I'd love to hear some commentary/discussion here.