thediagramofeverything.com

A truly Western Sydney rail line

Due to successive generations of postwar expansion, the western suburbs of Sydney now comprise approximately two million residents, almost equal to the combined population of the inner and coastal suburbs. At the heart of the western suburbs is Parramatta, a metropolitan centre in its own right, anchoring the activities of these two million people, though struggling to do so with far less infrastructure and far less national and state political attention and financial resources than central Sydney. Yet Parramatta is also a commercial centre with a central busines district of approximately 160 hectares, similar to the Sydney CBD's 170 hectares, though vastly underdeveloped by comparison.

One key investment missing from the western suburbs is a serious commuter rail network; one that is serious about catering to hundreds of thousands of western Sydney workers, students and schoolchildren per day, which a rapid bus network cannot do; and one that is serious about developing the commercial centres of western Sydney to their full potential, and not on simply funnelling workers on hour-long journeys into Sydney and the lower North Shore.

The following exploratory drawings, produced in 2004, propose a new commuter rail line that runs north-south through the western suburbs, centred on Parramatta, but also connecting it directly with the commercial centres of Norwest Business Park and Castle Hill, the industrial centres of Wetherill Park and Smithfield and the region's most rapidly expanding residential areas in the northwestern and southwestern suburbs. They also propose a second railway station within the Parramatta City Centre to the north of the river, to fully realise the development potential of those districts. They are published in the hope of provoking further reflection and debate on the future of one of Australia's largest growing regions and of its oldest city, Parramatta.

Kerwin Datu, May 2009

Contents

A truly Western Sydney rail line
Preface: Inhabiting a metropolis | Parramatta as an independent city
Metropolitan nodes v. urban precincts
Introduction
Comparison of City Centres
Public Transport Usage
The Christie 'Long Term Strategic Plan for Rail'
The new line drawn to scale
List of stations
Private Transport Usage
Public Transport Commuters

Paris
London
Sydney
Berlin
Barcelona
Munich
Parramatta
Brussels

Town Hall
Wynyard
Central
Martin Place
Circular Quay
Museum
St James
Conclusion

Preface: Inhabiting a metropolis | Parramatta as an independent city

We tend to think of the Cumberland plains (the Hawkesbury-Nepean basin) as Sydney and its suburbs. But in areas such as the Kansai region of Japan, the Veneto of Italy, the Randstad of the Netherlands, we see regions of similar geographic area to the Cumberland basin inhabited by a cluster of mature, interdependent cities, rather than a single focus, each city having its own cultural legacy, economic centre and international reputation. Witness Osaka, Kyoto, Nara and Kobe in the Kansai, Venice, Mestre, Treviso, Padua, Verona and Vicenza in the Veneto, Amsterdam, Rotterdam, The Hague and Utrecht in the Netherlands. If the cities of the Cumberland basin - Parramatta, Liverpool, Penrith, Richmond, etc. - do not develop these identities for themselves, there will be increasing social inequality between 'inner' and 'western' Sydney - evident already in areas such as private transport dependency.

Parramatta itself has much to build on already as an independent city: a dense commercial core, a rich public heritage, abundant parkland, stadium and theatres. Within the city fabric, car traffic is already successfully guided around its perimeter, creating a highly pedestrianised centre. There is a chance to extend the pedestrian core of Church Street into a great boulevard, linking Belmore Park, Prince Alfred Park and Civic Place into a rich chain of public spaces. The commercial and residential development required to spark this transformation would itself be underpinned by an expanded network of new rail lines and new stations throughout the city centre.

Metropolitan nodes v. urban precincts

The following research is concerned with the current condition of planning in Sydney, in particular, that of city centres. The planning of city centres often fails to reconcile their role as metropolitan nodes with their role as urban precincts. As a metropolitan node, a city centre is a point at which transport routes intersect and redistribute commuters. As an urban precinct, it is a concentrated develpment of residential and working populations which may cover a large area of land.

The success of a city centre in either of these roles relies upon success in the other. Transport routes will not intersect at a point where there is too little development, and concentrated development will not occur without intersecting lines of transport. Therefore, the two must be planned together.

Conceptually, this is obvious, but at the metropolitan scale this relationship is easily obscured. In Sydney, metropolitan plans are rarely more than abstract diagrams drawn at the regional scale, with city centres depicted simply as points in space. Because the plans do not broach the scale of an individual city centre, the centre's requirements as an area go unresolved.

If a plan does not address city centres' requirements as urban precincts, its success as a metropolitan network is unforeseeable. Therefore, while metropolitan plans remain as abstract diagrams, without strategies resolved at the local scale, there is little impetus for communities to invest in their proposals.

The following research investigates elements of the relationship of metropolitan planning to the scale of the city centre, using drawings to understand how metropolitan nodes materialise within urban precincts.

Introduction

The issue of passenger rail services is at the core of the conflicting roles of the city. It is also a crucial issue in the Sydney region. As Sydney swells from four to five million people by 2022, it is no longer feasible for transport infrastructure to centre on the City of Sydney. The planning community is greatly concerned with redistributing the residential population and the employment centres of the region throughout a network of satellite cities.

A number of strategies reconceive of Parramatta as Sydney's second city centre. There are strong arguments for this conception - it is close to the demographic centre of the region, at the intersection of major arterial roads, has an established commercial core and has a rich public heritage.

However the various rail proposals fail to address one or the other of the two roles that Parramatta would have to perform.

The following research attempts to reconcile the two types of strategies through drawings at the scale of the region, the city centre, the station precinct, and the individual station. Throuh these drawings, Parramatta becomes an exampl of a more resolved approach to integrated transport planning, whose lessons can be applied to other centres.

The first set of drawings, at 1:200 000, places Parramatta as a node in the existing rail network, and investigates the feasibility of a new rail line through the western suburbs.

The second set of drawings, at 1:10 000, considers Parramatta as a precinct of development, in comparison with cities throughout the world, and investigates the feasibility of a new railway station within the city centre area.

The final drawings, at 1:5000, describe the pedestrian network required to distribute passengers from Sydney's city stations.

Comparison of City Centres

The facing maps compare the core precincts of Parramatta and Sydney, as defined by their respective city councils.

While there is high rise development throughout the Sydney CBD, in Parramatta this is limited to a few blocks around the station. North of the Parramatta river, many blocks have floor space ratios below 0.3:1, used for little more than on-grade carparking.

The NSW Department of Infrastructure, Planning and Natural Resources (DIPNR) forecasts that by 2016 to 2025 there will be 80,000 jobs within the Parramatta City Centre. This is a 167% addition or a 4.0-2.8% annual job growth since the 1991 census recorded 30,000 working there. DIPNR's target is that 60%, or 48,000 workers, will commute by public transport.

If the Parramatta to Chatswood line was completed as planned, only 25,580 people would exit Parramatta station during the 3.5 hour morning peak. Assuming 72%, or 18,400 of those are workers, there are still at least 30,000 people failing to use public transport to meet DIPNR's target (not to mention the remaining 40% of workers, local residents, shoppers and schoolchildren who should be using the system as well!)

The Greater Western Sydney Economic Development Board (GWSEDB) makes the point that while there are seven rail stations for the Sydney CBD's 200,000 workers, there will still only be one for Parramatta's 80,000. (Two if Harris Park is included, but currently only one out of four trains stop there.)

It has soundly proposed a second Parramatta station north of the river. Unsoundly though, this would lie on a new rail line running from Westmead to Rydalmere, parallel to the Main Western line. There is no evidence given that this new station will actualy convert car users to public transport, or whether it will simply redistribute existing users of the Western Line.

The proposal correctly identifies how a new station will create a viable urban precinct, but fails to understand that the new station must also function as a viable metropolitan node.

Public Transport Usage

To understand how Parramatta should function as a metropolitan node, it is necessary to place it in the context of the regional layout of Sydney.

Sydney's fastest growing regions are the west and south west, where residential subdivisions are rapidly filling in the Cumberland plains. However, these areas (around Rouse Hill, Bossley Park and Austral on the facing map) have developed with very little public transport connections.

The most recent additions to the network were motivated by factors other than population growth. The Olympic Park loop and the Airport line were built to connect major facilities to the network. The Parramatta to Chatswood line was to integrate existing rail suburbs and alleviate congestion on the Main Western line. The postponement of the line between Parramatta and Carlingford shows that serving the growth regions was a lower priority than alleviating congestion.

Current initiatives for the west and south west regions are the North West rail line linking Epping to the Hills District, and the Parramatta to Liverpool rapid bus transitway (T-way).

However, the North West rail line will not serve the bulk of the growth regions, while the T-way may not have sufficient capacity to meet the regions' needs.

The facing map shows that, because of the extent of sprawl, public transport usage in Sydney is more closely tied to the rail network than any other mode. Only in the dense inner suburbs are buses an effective alternative.

More effective than the current initiatives would be to establish a new rail line (shown in yellow) straight through the western growth regions, intersecting the Main Western line (existing lines shown in blue) at or near Parramatta. This arrangement would strengthen Parramatta's position as the focal point for the entire metropolitan network.

The Christie 'Long Term Strategic Plan for Rail'

In 2001, Ron Christie, the former NSW Coordinator General of Rail, set out his 'Long Term Strategic Plan for Rail'. This proposed a network with numerous line extensions, some additional 80 stations, and three entirely new lines. One of these new lines would run north-south through the west and south west growth regions, just as discussed above.

The proposal resolved many of the rail services so abstractly suggested in the NSW Integrated Transport Strategy of 1993, notably the Macquarie-Castlereagh corridor.

The new lines would firmly establish Parramatta as a viable metropolitan node, which the GWSEDB proposal would not do. Drawing from the currently car-dependent growth regions, the new north-south line would convert thousands of commuters to public transport, ensuring that DIPNR would far exceed its target of 60%.

The new line drawn to scale

One criticism made was that it was still too intensively focused on the City of Sydney. There are a number of new stations and lines throughout the central business district and the inner suburbs where buses already complement the rail system.

Also, new suburban extensions to Bringelly, Box Hill and Kings Park are routed to flow directly into the same congested inner suburban network, rather than creating a peripheral system in their own right.

The proposal lays our a viable metropolitan network, but it is unclear whether individual urban precincts so created will feed the new infrastructure. It is also unclear that a solitary station at Parramatta will not become overcongested.

However, if Ron Christie's proposal for a metropolitan node at Parramatta is aligned with the GWSEDB's proposal for a developed urban precinct around a new railway station, it is clear that these two visions support each other.

To strengthen the role of Parramatta, I have adjusted Ron Christie's layout slightly. Whereas under his proposal, a passenger commuting from Rouse Hill to Parramatta would have to change trains at Castle Hill, I am proposing that these two stretches of line be continuous.

List of stations

Rouse Hill - a growing town centre, future extension to Marsden Park and the Richmond line
Mungerie Park - major residential land release areas
Burns Road
Norwest Business Park - an expanding major employment centre
Hills Civic Centre
Castle Hill - an expanding commercial centre
Baulkham Hills - an established town centre
Winston Hills
Northmead
North Parramatta - cnr. Church Street and Victoria Road
Parramatta - interchange with Main Western Line
Merrylands West
Wood Park
Smithfield - major industrial employment centre
Prairiewood
Greenfield Park
Bonnyrigg
Busby
Miller
Hoxton Park - future connection to Bringelly

Private Transport Usage

The revised Ron Christie proposal for a western suburbs line is shown here with the GWSEDB proposed station highlighted. When compared with the geographical distribution of private transport users, it is immediately evident that this proposal would redress a major imbalance in the commuting patterns of Sydney.

It also becomes clear that this line is more urgently required than many other elements of Ron Christie's Long Term Strategy for Rail.

Public Transport Commuters

In residential areas dominated by single dwelling lot housing, it would be easy to suggest that these suburbs might not have the sufficient population density to support a new rail line. The facing map shows that this is not the case.

Because of their small lot sizes, these new suburban development areas are in fact denser than many established parts of the network. The southwest suburbs have typical densities of 3000 to 5000 people per square kilometre, comparable to the Cronulla line, whereas the Main Northern, North Shore and Campbelltown lines typically have less than 2500 people per square kilometre. Over time, the new areas will only further outpace the established suburbs in density, strengthening the need for the new line.

Along the proposed Western Suburbs line shown dotted, there are three areas that have noticeably less density than the other new suburbs. One of these is Stanhope Gardens - a residential area developed later than this 2001 Census Data, which is expected to have densities similar to the southwest suburbs. The other two are less dense because they are in fact major employment centres, which the Western Suburbs line seeks to support - the Norwest Busines Park and the Smithfield/Wetherill Park industrial area.

Therefore, not only is the Western suburbs line more densely populate than many existing sections of line, it also better integrates a chain of existing employment centres with their workforce. This is something which many other lines do not achieve. The Main Western, Campbelltown and North Shore lines may run for up to twelve kilometres between major employment centres.

Paris (11,293,200 residents)

Just as the GWSEDB proposal clearly attempts to establish Parramatta as an urban precinct, but fails to activate it as a metropolitan node, the Ron Christie proposal does the reverse. His new lines definitively position Parramatta at the focus of the expanded network. But while he proposed five new stations for the City of Sydney, he proposed none for Parramatta. This would effectively overcongest the existing Parramatta station, defeating its purpose as an interchange.

For a node to function, a precinct must be developed through which many alternative routes of travel may weave together. The precinct requires sufficient commercial, residential and recreational activity so that the node is as much a destination as a point of interchange. The more people who live and work in the precinct, the greater the frequency, capacity, and choice of direction in transport routes the node can sustain.

When seen at the scale of the city centre, the metropolitan node is a complex network of infrastructure, whose form cannot be resolved at the regional scale. It is necessary to investigate the urban scale itself to determine appropriate distribution of rail services within an urban precinct.

The following plans investigate a number of such city centres from around the world, in comparison with Sydney and Parramatta.

London (11,230,500 residents)

Paris and London both exemplify mature rail networks serving large metropolitan populations, in which certain patterns may be determined.

Firstly, the city centre is served not by a single major station, but by twenty or more stations. These are scattered within a few hundred metres of each other, so that wherever one finds oneself, there are a number of directions of travel available within walking distance.

The city centre is criss-crossed by a number of lines which rarely fork (as Sydney's stations do at Strathfield, Town Hall, or Granville). As a result, individual directions of travel do not compete for shared sections of the same line (as Cronulla and Waterfall compete for the Eastern suburbs line).

Along these lines, stations are still scattered within a few hundred metres of each other. The line behaves much more like a bus route, serving a continuous corridor rather than a chain of nodes.

Sydney (4,289,900 residents)

By comparison, the City of Sydney contains very few stations within the same geographical area, and very few alternative directions of travel. The number of individual platforms may in fact be comparable to Paris or London, but these are aggregated into larger single stations.

Understandably, the reduced level of service available in Sydney may be due to the substantially smaller metropolitan population. However, building heights within the City of Sydney are typically higher than in the centre of Paris or London, therefore it is possible that a similar number of passengers arrive through Sydney's seven rail stations each day.

Berlin (3,942,600 residents)

However, Berlin, a city with slightly smaller metropolitan population than Sydney, still provides a greater level of rail service to its city centre.

It must be noted that the distribution of stations follows the same pattern as that for Paris and London. A number of alternative directions of travel intersect loosely within the precinct. Stations are scattered within walking distance of each other, and each line behaves like a bus route, with stops at every second city block.

Barcelona (3,871,400 residents)

Barcelona, slightly smaller than Berlin, maintains twice the number of stations. At this point in the investigation, we must rule out the notion that density of railway stations is in proportion to metropolitan population.

We must also rule out that Sydney's rail service is adequate in comparison to cities of similar populations.

Munich (2,341,700 residents)

Even a city with half the metropolitan population of Sydney comprises more stations. If the centre of Sydney were to contain a proportionate number to Munich, it should have twenty stations - as many as London has in the same area.

The rules established in the Paris Metro and the London Underground still apply to the Munich U-bahn.

- an urban precinct is sufficiently viable only if it is within walking distance to a train station; or,

- an individual station is of very little use, since it will activate only a few surrounding city blocks of development.

Parramatta (1,780,000 residents in Greater Western Sydney)

Both of these new formulations apply to Parramatta. If we accept GWSEDB's argument that Parramatta be considered the centre of a region with a population of about 1.8 million, then we must expect it to conform to the same rules for other city centres.

Indeed, the distribution of stations compares closely with development levels across the area. New office towers are being built within 500 metres of Parramatta station, but there is little development pressure north of Phillip St, or along Victoria Road, or even in Harris Park, which is just outside of walking distance to four stations.

Brussels (1,740,300 residents)

We can compare Parramatta to Brussels, which contains eighteen stations in the same area as Parramatta's four, or London's twenty. The Brussels system demonstrates a clear logic driving the layout of the entire urban precinct - two perpendicular axial routes intersecting a peripheral loop.

These drawings reveal important strategies for individual lines and individual stations. In many cities, stations are positioned as close as 300 metres apart along the one line. Rather than positioning the stations as dots on the map, and connecting the dots with the rail line, a different approach is taken. The line cuts a swathe across the city centre, providing rail service to a contiguous band of development. The rail line typically runs beneath a main avenue, the two supporting each other.

Town Hall (29,393 commuters between 7 and 9am)

It becomes apparent that the individual station has substantial limitations. Commuters will not generally walk more than about five minutes away from a station. Dense development is restricted to a few surrounding blocks. Because of these limitations, the surrounding streets must be reorganised to allow commuters speedy thoroughfare.

This reorganisation is intrinsic to the efficiency of the station, and is an evolving endeavour, but unfortunately the two planning issues are rarely integrated from the outset.

If this is not planned integrally with the station, it will often occur belatedly in an unsatisfactory fashion. Many stations in the City of Sydney have been retrofitted with sprawling pedestrian underpasses. It is now possible to walk for four hundred metres away from Town Hall station without sky passing overhead.

Wynyard (27,861 commuters between 7 and 9am)

The preferable alternative is to reserve key surrounding streets for pedestrian use only. This has occurred in pieces around Wynyard station, with Regimental Square and Sesquicentenary Square. However, these pieces have not been strategically integrated with the station itself, creating another sprawl of underpasses to its north.

Central (25,058 commuters between 7 and 9am)

At Central station, a state-significant piece of infrastructure severely restricts the effectiveness of a local scale commuter node. On the Surry Hills side the station aligns perfectly between Devonshire and Foveaux streets, creating the (unrealised) potential for finely detailed human-scale transitions between the station and the surrounding fabric. On the Railway Square side, again one may walk for up to four hundred metres before arriving on the natural ground plane.

Martin Place (10,137 commuters between 7 and 9am)

Martin Place station is well integrated with the public promenade, exploiting its changing levels to discharge passengers efficiently. However, the whole system is lopsided - nearly half of the station precinct is taken up by parkland. The station may have been more efficient, and more heavily patronised, if it were one or two blocks west, relieving Wynyard station, and also discharge passengers at its eastern end. It may have been placed further southwest, closer to the Pitt St Mall, relieving congestion at Town Hall. It may have been placed further north, allowing St James to be more heavily used.

Then again, the previous research shows that the city as a whole precinct could easily have sustained an additional station at each of these locations.

Circular Quay (6,062 commuters between 7 and 9am)

The precinct around Circular Quay suffers from the same lopsided condition as Martin Place, where nearly half is parkland or water. In this case, the station is also somewhat isolated from immediate development.

The central discharge point created intense pedestrian flows to the south, forcing the closure of Alfred Street between Loftus and Young Streets, and rerouting the bus network to suit. This is a happy compromise for pedestrians and for the quality of the public space, but it is an indication of what further circulation systems must be taken into account in the planning of a railway station from scratch.

Museum (3,897 commuters between 7 and 9am)

In the case of Museum station, the relationship between pedestrian amenity and station efficiency is quite distorted. Located beneath Hyde Park, commuters are forced to walk up to a hundred metres before reaching the surface, or any other form of activity, and without any of the benefits of being in Hyde Park.

And while the creation of a city loop is wise, more of the lines from Central must be directed anticlockwise to reach Museum before Town Hall, for Town Hall to be substantially alleviated.

St James (3,176 commuters between 7 and 9am)

Again, buried under Hyde Park, St James is the station most isolated from immediate development.



RailCorp, who manage Sydney's CityRail service, require single entry points to each station, to minimise staffing costs at additional ticket barriers. As these drawings show, this prioritisation runs counter to the philosophy of the whole station precinct.

There are approximately ninety hectares of development that may be activated within 500 metres of a station. If commuters are forced to double back up to 200 metres through one ticket barrier, the increased walking time may curtail up to thirty hectares.

Laying out a radiating network of ticket barriers and concourse areas around a station, at the project's inception, and streamlining the surrounding pedestrian environment may encourage an additional twenty hectares of development. The land revenue that may be raised as a result far outweighs the cost of additional staff.

Furthermore, with the introduction of integrated ticketing systems, Sydney's rail stations may adjust to layouts less dependent on staff surveillance.

Conclusion

In recent years Sydney has expanded largely in ignorance of its mounting land, energy and water resource shortages. Alternative ways of developing our built environment must be established on two fronts.

Firstly, we must develop metropolitan infrastructure that can accommodate rising urban densities. Our rail network is one of many systems that has become neglected. However, proposals for new rail lines are meaningless if they cannot account for the types of urban precincts that will be created around them.

So, secondly, for the appropriateness and viability of our infrastructure strategies, we must choose what kind of urban environments we wish to live and work in, and take responsibility for the resources needed to support our choices.

A key decision for Sydney will be the future role of its regional centres. If Sydney's sprawling suburbs are to integrate and develop infrastructure efficiencies, the regional centres - Parramatta, Liverpool, Hornsby, Penrith, etc. - must become viable city centres in their own right to complement the City of Sydney. As metropolitan nodes, they must develop as focal points for new infrastructure; as urban precincts they must provide attractive environments to become the social heart of their immediate suburban areas.

This research has attempted to demonstrate that city centres must be planned at the metropolitan, urban and station scales simultaneously. The different scales must not be conflated, such that one station is expected to focus an entire sub-region, as is often presumed can be done in Parramatta. Each of these regional centres must be provided with a number of new lines and new stations. Without an adequate distribution of rail services in each centre, it will not develop sufficient density and activity to harness future growth.

Kerwin Datu, December 2004