Parramatta city circle: MacPark/Global Economic Corridor as "middle ring"

Malcolm Turnbull has announced Federal funding for an extension of the Gold Coast light rail.

In NSW the top two candidates for the next light rail line (amongst an original short list of four routes) are Parramatta to Olympic Park and Parramatta to Macquarie Park.  At the moment, there is an intense competition between these two routes for which one gets funded and proceeds.    Could further Federal funding for light rail make it possible for both routes to proceed?

This blog has argued the Olympic Park route should have priority, on the basis of the CBD-scale urban regeneration and land-value capture possible there.  On the other hand, Macquarie Park should first implement parking levies to bring it into competitive neutrality with other employment centres like Parramatta and North Sydney before it gets any more government funded infrastructure.  TfNSW's 2031 travel modelling shows very little justification for the MacPark to Parramatta connectivity.  In contrast, TfNSW projections for travel between Parramatta to Olympic Park are an order of magnitude greater than Parramatta to MacPark:

Parramatta's "city circle": world class ring of CBD train stations

Parramatta CBD is so much overshadowed by the Sydney CBD that it's characteristics as a CBD in it's own right are under-appreciated.  I believe the biggest reason of the "under-marketing" of Parramatta CBD is that it is seen only through the lens of the high rise building zone around Parramatta railway station - whereas both Melbourne and Sydney CBD have the concepts of a "city circle" or "city loop" of 8+ train stations.  Also, the boundaries of the Parramatta local government area are nonsensical, incorporating suburban areas far to the north of Parramatta River (bordering onto Hornsby LGA) whilst missing out on immediately adjacent employment zones like Silverwater and Olympic Park that are south of Parramatta River.

So it's important to redefine the conventional notion of Parramatta CBD.  It needs to be more than a highrise zone around a single train station.  It needs to incorporate diverse and vibrant precincts in the same way Sydney CBD has districts like Ultimo, Pyrmont, Darlinghurst that add character to the high rise business core.

The catch is therefore broadening out the definition of Parramatta CBD to the corridor contained within the "city circle" of train stations between Westmead, Strathfield and Rhodes:

This definition is important, as Parramatta station precinct on it's own is not in the "world class" league.  However, Parramatta "city circle" is clearly world-class as highlighted in the poster below:

Tax reform: investment neutrality

The tax system distorts investment decisions.  On the other hand, some forms of investment have a "positive externality" spill over into the welfare of society and rightly should attract concessions.  The aim of tax reform should be to have the greatest possible degree of tax neutrality once positive externalities have been taking into account.

Most current tax subsidies (superannuation, negative gearing, franking) favour mature, and generally income generating investments over early stage growth investments.  The cost of superannuation tax concessions alone costs $25 billion per year and is growing rapidly.

To incentivise growth, the tax subsidies need to be recast.  Close off loop holes for mature stage investments that can stand on their own two feet.  Then provide tax incentives to grow early stage investments like IT & biotech that at $2.5b per year are a drop of water compared to the ocean of superannuation and property investments.  These tax incentives need to be targeted yet not be so complex or specific it amounts to "picking winners".

Sydney Metro: 2036 projection calls for Liverpool extension

The last post looked at current 2014 and projected 2024 rail system patronage and capacity numbers.

Below are projections out to 2036.  A simple across the board 3% growth rate has been used on the basis of the "KISS" (Keep it Simple, Stupid) strategy.  (Explanations for each of the columns as per previous post: Sydney Metro: 2024 patronage projections & it's hidden "West Metro" branch).

This is because local suburb-by-suburb rezonings and developments are just rounding errors in overall rail system growth rates.  For a land use change to have dramatic effects on growth rate on a line, it needs to be "CBD scale" or a cluster of strategic centres, or a very large scale land release over many years.  (To put the latter into context in a simplistic manner, the 20+ years of large scale land release across the Northwest from Cherrybrook through to Rouse Hill translates into patronage of the order of an extra 10,000 peak hour pax, or 12.5% of total North Shore/NWRL rail patronage).

Nomenclature of sectors above:

Sector  F0 = Northern line's freight/intercity tracks (also services upper northern stations north of Epping).  Terminates at Strathfield platform 1.

Sector T0 = Epping to North Shore on dedicated northern line tracks (post quad Strathfield to Epping), then via Harbour bridge. (Upper northern line stations remain in F0)

Sector T1 = Western express tracks to central

Sector T2 = Merrylands to Inner west, Airport and Revesby

Sector T3 = Sydney Metro, extended to Liverpool via Bankstown Airport, then taking over Cumberland line to Guildford, & continuing to Blacktown via Parramatta

Sector T4 = Illawarra main tracks to ESR

Sydney Metro: 2024 patronage projections & it's hidden "West Metro" branch

I've been building a patronage and capacity model.  This model covers the years 2014, 2024, 2036 and 2049.  First let's look at the 2014 rail system numbers, and my projections for 2024.  I've taken a pretty simplistic approach of just taking the 2006-2014 compound annual patronage growth rates and projecting these out to 2024, but this more or less aligns with TfNSW's line-by-line growth projections to 2019 (they use a slightly higher growth rate for the Western line).

Some more specific notes to explain the tables below:

1. Baseline 2014 data: taken from TfNSW Bureau of Travel Statistics train station barrier counts.

2. Forecast 2024 data: projected from 2014 data using historical growth rates 2006-2014.
* NWRL data taken from TfNSW projections and added to North Shore & NWRL 2014-2024 growth rate.

3. AM peak entries: apply to the period 6am to 9:30am.

4. Suburban transfers/exits: see explanatory note after tables.

5. Peak Pax to Capacity Ratio: interpretation is as follows
* ratio ~ 0 to 1: well below capacity even during peak hour (8am-9am)
* ratio ~ 1 to 2: at capacity during peak hour, spare capacity in shoulder periods of AM peak
* ratio ~ 2 to 3: at capacity during peak hour & during shoulder periods of AM peak
* ratio ~ 3+: exceeds capacity even in shoulder periods

Ok, here are my spreadsheet tables for 2014 & 2024.  (Future post will cover 2036 & 2049).

Some interesting details discussed below are the following:

1.  Sydney Metro already has a hidden "West Metro" branch

2.  The critical role of suburban interchange at Strathfield in balancing overall system patronage.

Sydney Metro extension to Liverpool: system-wide consequences

Liverpool council has called for Sydney Metro to continue to Liverpool, instead of current plans to terminate at Bankstown.

This idea seems simple on the surface, but it will have system-wide consequences on deeper examination.  Are these system-wide consequences positive or negative?  After extensive thought, my conclusion is that an extension to Liverpool (using a direct route via Bankstown Airport) will be a game-changer for Sydney Metro, and make it arguably the most important line by far in all of Sydney's rail network.

This is all without any further branching of Sydney Metro, an idea I've over time shifted to become less in favour of.  In fact, branching would be highly undesirable and a single unbranched line will provide supreme benefits of simplicity, high frequency and reliability whilst being a good match for the increased capacity requirements of serving Liverpool.

The key to why the Liverpool extension has such system-wide implications is what happens to the Cumberland line.  Given it's continuity with the Liverpool extension of Sydney Metro, it makes sense to incorporate at least part of the Cumberland line into Sydney Metro as well.  This would bring Sydney Metro connections to Parramatta (a key centre that currently does not have Sydney Metro access).  The result is shown in the rail diagram below (which also contains other rail plan ideas, but of main relevance to Sydney Metro is the Sector T3 orange curve).

As seen above, incorporating the Cumberland line into Sydney Metro will finally allow rationalisation of the complex interactions of the Cumberland line with Sector T2's Innerwest & South line.  What are these interactions?  There is a tension between whether the stations from Liverpool to Merrylands should have single seat (ie: interchange free) access to Parramatta CBD or to Sydney CBD.  As a result, rail planners have tried to do both by having Cumberland line and Sector T2 share tracks up to Merrylands, where the former branches west (via Harris Park) and the latter to the east (via Granville).  This has resulted in compromised frequency and increased complexity for both lines.  The lack of frequency is a major problem, as with wait times of 15-30 minutes, the Cumberland line is an unattractive option compared to just using interchange at Granville.  Likewise, the inner west line only offers frequency of 4-6 trains per hour at Newtown, which is unsatisfactory.

Parramatta Centric Revisualisation: Geographical Accuracy

Key features of the Parramatta Centric revisualisation of Sydney's rail network are the middle and outer orbitals centred on Parramatta.  These orbitals were diagrammatically shown as ellipses.  However, as seen in the map below, this is surprisingly accurate geographically, as key railway stations and centres are located very close to a "best fit" ellipses drawn through them.  The long axis of these ellipses is restrained to perfect east-west by my drawing program, and it may be that the fit would be even better if there was a slight rotation to the ellipses.  Either way the fit is very good: