Sydney’s Constrained Rail Future

1     Introduction

The document Sydney’s Rail Future - Modernising Sydney’s Trains - proposes a three-tier set of train operating principles and a five stage plan for the rail network over the next 20 years.

Stages 1 to 3 appear to present a workable and understandable way forward, given the history of rail planning confusion and a decision to prioritise construction of the North-West Rail Link (“NWRL”), despite the inconvenience of additional interchanges that many users will experience. However, as well as these inconveniences, features of the full five stage network appear to be inconsistent with the three-tier principles and/or the intent of the metropolitan strategy such that the longer term benefits for the whole of Sydney could be compromised.

This document expands on the above and concludes that some aspects of Stages 1 to 3 should be seen as a stop-gap measure only, to be later amendable, and outlines appropriate consequent modifications to the NWRL project that would enable a more considered long term plan to be developed.

2     Three-tier principles

The introduction of metro-style single-deck trains, described as Tier 1 rapid under the three-tier principles, offers a number of advantages, including:

Such services are generally operated on a point-to-point basis, or at the most two branches at the ends of a central core, in order for “turn up and go” frequencies to be maintained for all users. This is in contrast to timetabled Tier 2 suburban services, where a higher number of through seats over greater distances, and often serving multiple branches, is expected by passengers. The use of Tier 1 rapid services from inner areas would also be consistent with employment location aspects of the metropolitan strategy (see later).

The application of these principles to the rail network, before and after a second harbour crossing, is addressed next.

3     Before a second crossing

The decision to prioritise the NWRL ahead of a second crossing has flow-on implications for operation of the North Shore and Western Line, where present train operations already appear to be close to the limit for frequency and passenger loading before reliability becomes unacceptable. Patronage growth plus an NWRL contribution is likely to add to current difficulties.

A key issue in the AM peak is the combined impact of regular commuter exit flows from the west and large contra entrance flows of north bound passengers joining from other lines or modes at the key Sydney CBD stations. The difficulties are compounded by double-deck trains and narrow station platforms. It is similar in reverse for the PM peak, but with more passengers waiting on Sydney CBD stations for a particular train. To ensure a reliable 20 trains per hour through the CBD with higher average loadings while retaining the contra flows mentioned above, Sydney's Rail Future proposes to reduce the number of routes and stopping patterns to have most users board the first available train and thereby more evenly distribute the commuter flows. Effectively, the proposal suspends part of the traditional convenience of a Tier 2 suburban service by requiring more passengers to interchange at an intermediate station.

The above proposal leads to the NWRL needing to be operated as a shuttle service before a second crossing can be completed. Observations on choosing to use single-deck metro-style trains for this shuttle include:

4     After a second crossing

In the longer term, the NWRL will extend to the Sydney CBD with a limited-stop link, and probably to the west through Marsden Park and onto the Western Line. As such it will more resemble a Tier 2 suburban, rather than an inner area Tier 1 rapid, service. Further, the average station spacing is not close enough for there to be a major benefit from better traction and dwell time performance, and the NWRL alignment is already based on double-deck train requirements.

Transferring most of the contra flows previously mentioned to the new Sydney CBD link would reduce the conflict with commuter flows at the existing busy Sydney CBD stations. Accordingly there would also then be scope to restore a traditional Tier 2 suburban service to these stations with branches on both the North Shore and the Western Line. Introducing a Parramatta-Chatswood shuttle service to provide an alternative route from Parramatta to Macquarie and the lower North Shore would help support such a restoration, as would an expected transfer of some patronage from the Western Line to the NWRL after a through service to the Sydney CBD is established on it. Making closer headway management on the approach to busy stations a feature of ATO could also be beneficial.

Elsewhere in the network, what can be achieved has been constrained. For example, the Inner West to Strathfield, Eastern Suburbs and Revesby via Airport services, particularly the latter with its growing number of passengers having luggage, cannot be independently converted to more suitable single-deck trains, and many double-deck commuter services from the West and South West will be slower than today. This is because all trains from the South West will run via the Airport and eight of the additional 14 trains from the West will slow down Liverpool services by sharing the local (all stations) tracks east of Strathfield, and themselves be slower than the other six additional trains from the West. In additional to the doubtful utility of the slower trains from the West, there is also doubt as to the need for so many additional services from the West and South West considering the employment location issues raised next.

5     Metropolitan strategy

There seems to be some tension between Planning and Transport for NSW over employment location, particularly for that served by public transport. The Planning ethos of jobs closer to home is supported by an aim for western employment to keep pace with western workforce growth. The middle diagram below (reproduced from the Metropolitan Plan for Sydney 2036, dated December 2010) illustrates how Planning envisage good quality cross-regional public transport links, such as a Parramatta-Chatswood shuttle, being introduced to help shape and grow the regional cities as attractive points of agglomeration. In contrast, Sydney's Rail Future, which looks more like the left hand diagram, will entrench the public transport focus on jobs in Global Sydney


The Planning position seems to have merit. The 2009/10 Household Travel Survey shows the commute share of trips at 16%, the distance at 28% and the time at 24% (see Annex section 1 for more detail from the data set). The excess percentage share for distance and time over trips has grown slowly over the last decade, and indicates a continuing disproportionate allocation of transport resources to the commute. The lower percentage for time than distance also shows that the commute is already faster than the average for all other trip purposes. In theory, transport networks would be used more efficiently if trips made at peak times were much shorter than average, rather than much longer. Although real-world employment considerations, such as the necessary scale for agglomeration to be effective, multi work destination households and site specific industries, tend to put a limit on what can be achieved, commute distance and time savings would help offset the negative impact of sprawl and release more resources for discretionary purposes.

It's generally true that less capital would be required to beef up capacity on the existing radial system, rather than establish new cross-regional rail links, and this, effectively a linear projection of current trends, probably lies behind the Transport for NSW position. However there would be operational cost and user time savings with the new links that would reflect into more liveable and sustainable outcomes. A shorter average commute could also lead to higher revenue per km, given the tapering of fare increases with distance, reducing the amount of community support required.

6     Being more consistent

It doesn't look like putting the customer first to establish sound principles and strategies, and then to impose technical constraints on their full and consistent application, as seems to be the case for the various stages of Sydney's Rail Future. The following looks first at north-side issues, starting with the new Sydney CBD and harbour crossing rail link, for more consistency.

As noted previously, the problem of simultaneous large egress and ingress flows at key Sydney CBD stations using double-deck trains and narrow platforms needs to be addressed, and transferring most of this problem to a better suited new Tier 1 rapid link so that the present route can handle its suburban commute role more effectively, looks logical. The new link would need to provide interchange with Martin Place station as part of such a transfer.

If the NWRL is to be operated as a Tier 2 suburban railway in the longer term, then other north-side services would need to be converted to Tier 1 rapid to match the new Sydney CBD and harbour crossing link. To this writer, the most suitable would be the present all stations to Gordon service plus a potential future link from North Sydney to Manly Vale or Brookvale to interface with an improved public transport system using mainly dedicated road space in the Peninsula area. Fortuitously, several factors work in favour of this possibility.

Assuming the 4.5% grade performance adopted for previous metro plans, the harbour crossing link could connect to a Victoria Cross station located closer to the surface and then to the existing line at Waverton for Tier 1 services. At the same time, new full profile tunnels would be constructed from just west of North Sydney to St Leonards to carry Tier 2 suburban services to the upper North Shore, NWRL and optionally the outer Northern Line. The less frequent ECRL double-deck services needed for a similar number of seats would, in turn, allow space for the previously mentioned Parramatta-Chatswood shuttles to be introduced. Additional fast tracks would then also be needed between St Leonards and Gordon, and the Tier 1 rapid service would need to be diverted to new underground platforms at Chatswood to make room for the turn-back of Parramatta shuttles at the surface platforms.

On the south-side, Sydney's Rail Future provides an illustration of the extent and limitations on what improvements can be achieved without significant additional expenditure on infrastructure. At the other end of the spectrum, this writer's plan for a more complete and consistent application of the principles and strategies through an imagined negotiation between what the planners want and what the rail system could provide, has already been submitted and an outline is presented in Annex section 2 to this document. Further information on the additional infrastructure required is available in the document CBD Rail Relief in Presentation Format that has also been submitted previously.

7     NWRL project changes

While operating the NWRL with single-deck trains initially has merit as a stopgap measure, the previous discussion has also shown that the situation would be different once a second harbour crossing is in operation. Accordingly, the credibility of the initial configuration depends on there being a clear migration path to a fuller and more consistent longer-term application of the previously discussed principles and strategies that also includes a timescale of commitment. For the NWRL project this means:

8     Annex

The following Annex comprises two sections as follows:

1: 2009/10 Household Travel Survey extracts for Commute and All Other Purposes

2: Description and diagram of this writer's rail network plan based on a fuller and more consistent application of the three tier principles along with the intent of the Towards a Network City diagram from Planning.

Peter Mills
July 2012

Annex 1: 2009/10 Household Travel Survey extracts for Commute and All Other Purposes

ITEM  Ref  Formula  1999/00    2001/02    2003/04    2005/06    2007/08    2009/10    HTS Ref.  
Population ('000)a3,9584,0674,1284,1814,2794.4374.1.1
Total Trips ('000)b15,11215,20715,28615,75716,26316,1714.1.2
Commute Trips ('000)c2,2582,2622,3812,3892,5412,5684.2.1
Commute Trips Sharedc/b0.1490.1490.1500.1520.1560.159
All Other Trips Sharee1-d0.8510.8510.8500.8480.8440.841
Commute Trips/Capitafc/a0.5700.5560.5770.5710.5940.579
All Other Trips/Capitagb/a-f3.2483.1833.2573.1973.2073.066
Total Distance ('000 km)h124,325127,560129,219129,401133,802133,5574.4.1
Commute Distance ('000 km)i32,19432,39834,06133,74736,60837,5814.4.4
Commute Distance Shareji/h0.2590.2540.2640.2610.2740.281
All Other Distance Sharek1-j0.7410.7460.7360.7390.7260.719
Commute Distance/Capita (km)li/a8.137.978.258.078.568.47
All Other Distance/Capita (km)mh/a-l23.2823.4023.0522.8822.7121.63
Total Time/Capita (min)n78.678.879.078.781.378.24.5.1
Commute Time/Trip (min)o32.533.
Commute Time/Capita (min)po*f18.518.519.018.920.419.9
All Other Time/Capitaqn-p60.160.360.059.860.958.3
Commute Time Sharerp/n0.2360.2340.2410.2400.2510.254
All Other Time Shares1-r0.7640.7660.7590.7600.7490.746
Commute Speed (km/h)l*60/p26.325.926.025.625.225.6
All Other Speedm*60/q23.323.323.123.022.422.2
Commute Trips % Shared*10014.914.915.015.215.615.9
Commute Distance % Sharej*10025.925.426.426.127.428.1
Commute Time % Sharer*10023.623.424.

Annex 2: Rail Network Plan

This plan applies the principles of a three-tier railway (rapid, suburban and intercity) while minimising the impact of flat junctions, and builds on spare capacity within the rail network to support key cross regional services. The plan also shows some extensions to radial rail routes. The number of cross regional rail corridors is necessarily limited by economic considerations but should still be effective because the rail network can support somewhat more circuitous routing before running out of shaping power. The cross regional services include: The plan implies a shift to more emphasis on serving the area around each centre, and less on a growth of longer distance commuting to the Sydney CBD for which the Sydney double-deck train has been optimised. The first three cross regional shuttle services need to be reasonably fast and frequent to offset the advantage of staying on the express double-deck service to the Sydney CBD. Accordingly, a mix of double-deck and single- deck trains is envisaged for the overall network, separated where possible but sometimes sharing track where radial and cross regional routes coincide or for specific operational reasons. As more lines converge on Parramatta, its importance as a rail interchange point would also increase.

In addition to the heavy rail shown, there could be extensive opportunities to introduce light rail services to serve regional centres on the off-rail axes and at the same time provide good connections to heavy rail for onward travel.

or sydconstrainedrailfutureBt.png
July 2012 web counter