Compared to comparably-sized cities in North America, Sydney does very well on Public Transport (Transit), with a pre-Covid 26% transit commute share. Compared to cities in Europe or Asia, it does poorly, indicating significant room for improvement. 

Much of that difference has to do with wealth and space. Despite the complaints,  Sydney is rich (money doesn’t grow on trees, but it does grow in rocks), so most families have cars. Sydney is also far less concentrated than cities in Europe or Asia, so distances are more amenable to the automobile and less to public transport, and the accessibility indicators show that.

Still, it’s clear more can be done.

There have been a forest of expired plans for public transport in Sydney. There are more plans still in the works. They almost entirely focus either on Trains (and especially Metros), or on specific lines that a particular party is pushing. But a detailed comprehensive look at the layer below the trains is missing.

I believe that removing rather than recapitalising most of Sydney’s Tram network was a mistake, and the evidence of Melbourne’s popular Tram service is as close to case-control comparison you can find in this field. Sydney has higher transit mode share than Melbourne, but that is because the Trains are so much better (higher frequency), not because people like the buses better than the Trams.

Still, that does not mean the trams should all be put back. First, Sydney cannot build everything on the historic maps, at least not all at once, even Chinese Metros have been built over phases. Even more significantly, you do not want to:

1. It may have been overbuilt. Many historic tram lines were abandoned fairly early, indicating the proponents missed their mark.
2. We have to prioritise. Some things are more important than others, or have a higher benefit/cost ratio. 
3. Conditions have changed. The places where historic trams were built may not warrant tram service today because the land use has changed, and of course because people’s travel preferences have changed with the widespread introduction of the automobile, the broader train network, and telecommunications and other technologies. 

I have been thinking about a “blank slate” redo of Sydney surface public transport (buses and trams). This is something the Government could do in a decade if they applied themselves.

You can see a detailed version in Google Maps here. You will need to zoom out, as it centers on the Liverpool CBD (the geographic center of the region). As always feedback is welcome.

Extending the LRT System

The Figure below shows the Existing, Under Construction, and the Sydney FAST 2030 Proposal LRT Lines. 

Design Principles

1. History: The routes on the historic maps were reasonable starting points, they knew what they were doing and often reshaped the landscape to fit trams, and the human landscape reshaped to adapt to tram corridors. Trams added access.
2. Comport: The new lines should comport with their environment and take almost no existing buildings, but instead use existing street space as much as possible, especially former tram lines, as well as former rights-of-way where appropriate. 
3. Significance: We want to connect places that were significant 100 years ago because they are most likely to be significant 100 years from now, lines should follow historic RoW.
4. Directness: Routes should proceed in a fairly direct (non-circuitous) routes between the origin and destination.
5. Parsimony: We should have a few core lines with a maximum of one split (two branches) at either end. The branches can be numbered differently (L2 vs. L3), but they share a core. Spurs with high frequency can be used if branching becomes a problem.
6. Complementarity: All lines should complement existing higher capacity public transport (Trains and Metro), not substitute for them. (Net riders on Trains and Metro should increase after the LRT opens)
7. Terminals: Lines should start and end at key transfer points (e.g. a Train Station) or destinations (e.g. the Beach)
8. Gap-filling: Lines should serve areas that are today underserved, even if it violates the above (e.g. Lane Cove)
9. Rebalance Movement and Place: Motorway construction allows us to repurpose roads that presently have a conflict between Movement and Place function (Parramatta Road, King Street/Princes Highway, Military Road)
10. Reconcentration: The collection of new lines (these plus that already committed) should serve all areas of the built up parts of Greater Sydney, and support infill (and brownfield) development rather than speculative greenfield development. 
11. Exclusivity: The designs assume LRT to get high ride quality on exclusive tracks with lower operating costs. These are not classic trams that shared space with roads.
12. Frequent: Most lines are served by single (articulated) car (L1), at a high frequency rather than fewer but longer trains (L2) at low frequency. 
13. Electrical: Electric powered, electricity delivered by wire (more efficient than battery storage)

Proposed LRT Lines

The Proposed Lines are discussed below:

1. L1sx: (Red) L1 Southern Extension: Central to Green Square and Mascot via Elizabeth and O’Riordan, Rationale: Serves existing high density areas and potential redevelopment sites. Elizabeth and O’Riordan are most feasible for Tram services due to Street widths among parallel routes and centrality. Provides capacity relief for T8 service, as well as serving areas in between the far-spaced stations. FAST Buses would serve parallel routes.  Extends L1 to maintain balance of flows (not split CBD frequencies too much on L2/L3, single car trams appropriate for this street running service. (~5.8 km)
2. L2/L3liz: (Dark Blue) Elizabeth Street: Relocate the L2 and L3 on the eastern side of the Sydney CBD from George Street to Elizabeth Street (Phillip Street), and then circle around to George St. Rationale: Provides service to Eastern CBD via Tram (currently missing), allows George Street to serve L2 and L3 western extensions.
3. L2ex: (Light Blue) Coogee Extension: Extends L2 from Randwick through The Spot to Coogee Beach. Rationale: Connect to a popular beach from the CBD without a transfer
4. L2wx: (Light Blue) Broadway – Wolli:   This line takes over the George Street LRT (which meets (and through runs with) the Elizabeth Street LRT. At Central it proceeds west along Broadway, South along City Road, down King Street in Newtown, down Princes Highway, to Wolli Creek. Rationale: Provides high capacity services to part of University of Sydney (Camperdown) currently without rail service, Newtown. It has the potential to pedestrianise King Street in Newtown (like George Street in the CBD)  by terminating City Road at Carillon Avenue and terminating Prince’s Highway at Sydney Park Road, which should be now feasible in a post-WestConnex world.
5. L3ex: (Dark Blue) Little Bay Extension: Extends L3 along Anzac Parade from Kingsford through Maroubra to Little Bay
6. L3wx (Dark Blue) Broadway – Five Dock: The line splits with the L2wx line and runs along Parramatta Road to Norton Street in Leichhardt, and turns West at Marion, to proceed through Haberfield to Five Dock, where it terminates at a Metro Station.
7. L4: (Green) Oxford Street/Victoria Road:  From West to East: West Ryde, via Top of the Ryde, Gladesville, Huntley’s Point, Drummoyne (assumes A40 tunnel under Drummoyne), Rozelle, sharing the existing L1 line (Alt: take two lanes from the Anzac Bridge), Museum, Darlinghurst, Paddington, Woollahra, Bondi Junction  to Bondi Beach
8. L5: (Purple) North-South: Wynyard to Northbridge via Harbour Bridge, North Sydney, Cammeray to Northbridge. This project restores Tram service to Wynyard Station and the Harbour Bridge, providing local services to North Sydney, and enabling interchanges with regional Trains and Metro services.
9. L6: (Purple) Wynyard to Lane Cove via Pacific Highway. Sharing track with the L5, it branches to provide local services on the dense Pacific Highway corridor and connecting with the historic regional center of Lane Cove, which is in an area underserved by rail.
10. P1nx: (Orange) Castle Hill Extension: (Female Factory through Baulkham Hills to Castle Hill)
11. P1ex: (Orange) Camellia Service (Rosehill – Camellia – Silverwater – Newington – Olympic Park)
12. P2nx: (Orange) Epping Extension (Carlingford to Epping). Extends Parramatta LRT Phase 2

Creating a Rapid Bus Network

Buses have not received the attention they deserve. We could do much better with buses than we actually do, but elite projection (elites can imagine themselves riding trains but not buses) is hard to overcome, so buses are regarded as inferior to trains for reasons that mostly have to do with how we use buses in the system, rather than the technology itself.  [Recognising that the ride quality of buses on streets is not as high as trains on exclusive rights-of-way]. This vision for Rapid Buses is not T-Ways (on exclusive rights-of-way) everywhere, but more akin to the Arterial Bus Rapid Transit services that Metro Transit in Minnesota provides. (You can see a nice video about the service). In short, buses are the workhorse of the public transport system, and need more attention to make them as excellent as possible.

Principles

• Gridded: We should design a Grid-like network, so that people can get to their destination with at most one-transfer. I don’t think this actually holds because of an inconvenient river. All of the proposed V-Lines stop south of the Parramatta River. Rail service north of the River is good, the B-Line already exists (which can be thought of as V-01), and the proposed LRT extension above are fairly comprehensive, so the areas north of Sydney Harbour and the Parramatta River get more H-Lines and no V-Lines. The existing Railroad rights-of-way present another barrier, as there are few surface street crossings of the lines, those are taken advantage of where possible, but wind up distorting the network from an idealised grid. (See Bambul’s post for a similar idea with a slightly different implementation or my earlier version focusing on Inner Sydney.)

• Directness: We should minimise bus circuity for these routes (there can be other minor routes after these are fixed, I won’t bother with that here), so that travelers are not riding all over creation as the bus operator seeks a few extra passengers or to meet some arbitrary standard about distance to the nearest bus stop. 

• Complementarity: Routes should complement not compete with existing Train, Metro, LRT, BRT routes. So even when there are bridges over the Parramatta River, they are already served by existing rail lines, so the principle of complementarity reduces the ability to provided continuous V-Line services from the south to the north, relying instead on transfers. This proposal assumes everything under construction and budgeted will be built (major Motorways, Metros, etc.)

• Feasibility: The cost should relatively minimal (achieving High Return on Investment), so essentially no new roads, bridges, and so on, are built for the bus routes, and a minimum number for the proposed  LRT links.

The services are gridded, so they are divided into H-Lines and V-Lines. Specific lines are itemised below.

H-Lines

Shown in pale blue, East-West or Horizontal (“H”) Lines (always even numbers, major lines end in 0, lowest numbers in the North, highest in the South)

• H10: Parramatta – Eastwood – Macquarie Park
• H20: Chatswood – Cremorne – Mosman – Manley
• H30: Crows Nest – Cremorne – Mosman – Taronga
• H40: Guildford – Lidcombe – Olympic Park – North Strathfield – Concord – Canada Bay – Drummoyne – Birkenhead
• H50: Bonnyrigg – Cabramatta- Yagoona – Chullora – South Strathfield – Enwood – Burwood Heights – Croydon – Haberfield
• H60: Bankstown – Belfield – Ashfield – Leichhardt – Annandale – Glebe – Usyd – Redfern – Surry Hills – Moore Park – Waverly – Bronte
• H64: Stanmore – University of Sydney – Redfern – North Waterloo
• H68: St Peters – Randwick – Clovelly
• H70: Liverpool – Canterbury – Dulwich Hill – Petersham – Enmore – Newtown – Alexandria – Green Square – Kensington (via Old Canterbury Road)
• H80: Bardwell Park – Earlwood – Marrickville – Enmore – Newtown – Erskineville – Alexandria – Green Square – Kensington – Coogee
• H90: Sydenham – Mascot – Rosebery – Eastlakes – Kingsford – South Coogee – Maroubra
• H100: Revesby – Padstow Heights – Beverly Hills – Bexley – Arnclife – Kyeemagh – Botany – Pagewood – Eastgardens – Maroubra – Maroubra Beach

V-Lines

Shown in light purple, North-South or Vertical (“V”) Lines (always odd numbers, major lines end in 5, lowest numbers in the East, highest in the West)

• V03: Bronte – Vaucluse – Watson’s Bay
• V05: Rose Bay – Double Bay – Eastcliffe – Randwick – Maroubra
• V11: Botany – Eastlakes – Rosebery – Zetland
• V13: Potts Point – Zetland – Green Square – Mascot – Botany – Malabar
• V15: Botany Road – Circular Quay – The Rocks – Barangaroo – Redfern – Green Square – Botany – Pagewood – Eastgardens – Maroubra
• V21: St Peters – Waterloo Metro – Redfern
• V23: White Bay – Annandale – Stanmore
• V25: Balmain – Leichhardt – Petersham – Marrickville – Sydenham – Wolli Creek – Miranda
• V29: Dulwich Hill LRT – Earlwood – Bardwell Park
• V31: Summer Hill – Hurlstone Park
• V35: Abbotsford – Five Dock – Croydon – Canterbury – Bardwell Park – Banksia
• V41: Bexley – Rockdale – Brighton-Le-Sands – Kogorah
• V45: Sans Souci – Carlton – Bexley North – Cabarita
• V47: Mortlake Spur 
• V55: Ramsgate – Allawah – Bexley – Kingsgrove – Belmore – Strathfield
• V65: Carrs Park – Hurstville – Beverly Hills – Wiley Park – Flemington
• V67: Penshurst – Lakemba – Greenacre – Chollura – Lidcombe – Olympic Park
• V71: Mortdale – Riverwood – Punchbowl
• V73: Padstow – Bankstown – Yagoona – Regent’s Park – Auburn
• V75: Rose Hill – Sefton – Revesby
• V85: Parramatta – Merrylands – Chester Hill to Panania
• V95: East Hills – Villawood – Fairfield – Westmead

Physical geography is always a factor. Because of the width of Sydney compared to the height, there are more V-Lines than H-Lines. Also, based on the principle of non-redundancy, more vertical bus routes are provided, as there are more existing and under construction horizontal train lines.

Note this service stops in Liverpool, as the areas west are not developed yet. We have ideas about that, and I am currently doing work in the area, so will abstain.

If any of these H- or V-Lines are successful, they can of course be upgraded, and as the physical rail network changes, one expects these lines will evolve as well, taking advantage of the flexibility offered by buses. I have not conducted a formal accessibility analysis of this FAST network proposal, but if you are interested in funding something, find me.

This vision is essential if public and active transport are to be the preferred choice for most Sydneysiders, which is critical for achieving the environmental goals of eliminating CO2 emissions.