Finding a way to Western Sydney Airport - by rail?
After more than 50 years of studies, the Commonwealth of Australia is funding and building another new airport in the fast-growing western part of the Sydney metropolis. This airport is about as remote as it is possible to be from population centres and CBDs in the region before running into the mountainous terrain that surrounds the Sydney Basin.
In Australia, responsibilities for such a major piece of infrastructure are shared – the Commonwealth/federal government is responsible for the airport itself and the State of NSW for the land transport access infrastructure that is needed to service it – noting that, however, the Commonwealth provided substantial funding to the State for such projects.
The State is also committed to building an “aerotropolis” or employment precinct adjoining the airport.
This report looks at the current airport access planning - and at what improvements could and should be made.
The lightly edited report is provided by Peter Thornton, Director, Transportation Associates and does not necessarily represent the views of CAPA - Centre for Aviation.
Sydney is finally building a second airport; useful rail links are needed
Fixed land transport links especially rail, which are considered the key to the commercial success of both an airport and an aerotropolis, are very expensive and can take years to develop patronage levels which warrant the investment in them. Of road and rail, road transport is far more flexible and, in the context of this airport, a network of existing and new or improved roads will permit easy access by road modes in today's generally unconstrained conditions.
The airport site itself provides massive land area for cheap at grade carparking as only one runway will be initially developed; but significant provision had been made in the Master plan for rail to service both the airport and the Aerotropolis.
But....While the upgraded road system is diffuse and will allow many different journeys in many different directions to and from the airport, a rail link is highly determinate and efficient for moving large volumes of people in a single direction in a well-defined corridor. What has surprised many commentators and transport professionals has been the decision by the Governments to build a 23 km metro spur line – albeit proposed to be a part of a future larger project – which links to the existing metropolitan rail system. This is instead of first building a connection to the Parramatta and Sydney CBDs.
The precise rationale for this decision is difficult to find though it is also relevant to note that both federal and state Governments need to retain the favour of electors in the region in order to stay in office. This railway looks like an investment for the region in the region.
The concern is that neither the new airport or the Aerotropolis will be served to maximum advantage in terms of accessibility; meanwhile this piece of railway may become an underutilised stranded asset.
In the case of the airport this may well reduce its value when the Commonwealth eventually decides to offer it for sale to the private sector in line with all other major Australian airports.
Then of course, one thing has been suffocated under the host of political goals which emerged once the long delayed (and long-obvious - work actually began on the site 30 years ago, but was cancelled for political reasons) decision was finally made. That is that the airport was intended to be a second Sydney airport, to relieve pressure on the existing, very established - and here to stay - capacity constrained Kingsford Smith Airport, Sydney Airport (SYD), 9 kms south of the major Sydney CBD.
As it stands, the new airport will do little to resolve that problem and the new owners of SYD, having paid over AUD20billion to buy it, will be anxious to pressure the government to relax the current antediluvian operating restrictions.
Western Sydney Airport (WSA) is being built 41 kms away, and, while surrounded by several regional centres, even the closest of these is 11 kms away, as the figures below show.
A key issue has been if and how this airport should be serviced by rail.
Railways are no longer viable for wandering around the countryside and providing public transport to all and sundry. They need to be used to link, in as straight a line as possible, places of high employment and / or high residential density – like CBDs. With employment generation of the order of 750 jobs per million passengers transiting, an airport certainly is a mini-CBD, and with both the onsite business park and the adjoining NSW Government developed “Aerotropolis” industrial development zone said ultimately to be generating more than 200,000 jobs.
Rail links, of course, are both very expensive to build and can take decades to build up to patronage levels which justify their operations. But this itself is a chicken-and-egg conundrum: if there's no egg, the chicken will be slow to arrive.
Sydney’s Airports and Land Transport System
The need for fixed infrastructure transport links
In Nov-2017 the then Commonwealth Urban Infrastructure Minister Paul Fletcher said, “a rail link will not be vital to the new airport at Western Sydney from day one” and that “a new rail line would not be justified in the airport’s early operations.” Given the extent of road upgrading already underway in the form of the Northern Road and the M12, on opening day in 2025, the airport is going to be very accessible by more flexible forms of transport such as high-capacity fast buses and, if Liverpool Council has its way, a trackless tram system, not to mention various forms of personal transport.
And in the start-up phase of WSA, such means of access for residents and local business travel within an arc about 20 kms will be cheap and effective. My own assessment has been that, insofar as serving the airport itself is concerned, and given the relatively uncongested and upgraded road system around it, a rail link is probably not needed and not likely to be economically justifiable before the airport is transacting something like 25 million passengers per annum, which, incidentally, was about where SYD was in 2000 when its rail link opened. And, while twenty years on it is operating modestly successfully, even that project went into receivership within a year or so of opening with patronage being only one quarter of that forecast.
However, that is another, and longer, story. It was a less than optimal compromise operation, integrated into the suburban rail system and offered prohibitive prices if more than one person was travelling.
Regardless of hard economics, railways have a kind of motherhood appeal in the world of public transport and the then-Shadow transport and infrastructure minister (and now Opposition leader) Anthony Albanese said in 2017 that “the project should be about more than the airport itself” and committed that a Labor Government would build such a rail link, though precisely what link he did not elaborate. There is a federal election looming, so we may find out soon.
And while I agree that an airport rail system should clearly not provide for airport passengers only, the problem is that no one can predict where the workforce at the new airport and at the adjoining “Aerotropolis” employment zone is going to come from.
So investment in earlier phase very high-capacity and very expensive fixed links which can realistically only point along direction or route is very risky.
It takes time for people to switch, given they are locked in by mortgages, schools, family commitments and the like. Road based transport will enable workers from residential locations from all points of the compass around WSA to efficiently access the employment opportunities. Given that WSA is utilising only about half of the site in the early stages of development, there is ample space to develop cheap at-grade car parking for workers and passengers alike. And airport operators love car parking revenue.
The WSA transport access plans
By March 2018, however, the Australian and NSW governments had undertaken a joint Scoping Study to “identify a long-term Preferred Network that sets out a vision for passenger rail to service both Western Sydney and Western Sydney Airport”with the outcome as indicated in the following figure.
Proposed Rail Links
The Scoping Study identified two preferred direct rail links to Western Sydney Airport – firstly a North-South Link marked (1) at an estimated overall cost of AUD15-20 billion (2017 dollars) and an East-West Link from Parramatta marked (2) AUD12-15 billion (2017 dollars) as well as identifying an extension of the existing line from Leppington to the airport at an estimated cost of AUD6 billion (2017 dollars).
The latter was not included in the Government’s Preferred Network as a direct airport rail link as this was said to “provide fewer long-term benefits”. It did not mention that this link, which would provide both an operationally direct connection to SYD and the Sydney CBD, would need to be operated by the existing Sydney Trains double-deck rolling stock rather than the single-deck driverless stock in operation on the first part of new Sydney Metro and strongly favoured by the NSW Government for new urban railways.
The current - questionable - state of play
Of Options I and 2, a part of Option 1 from a junction with the Main West at St Mary running south through the Airport to the site of the Aerotropolis has been chosen and is to be a 2021 AUD11 billion, 23 kilometre entirely new railway which, so it is proclaimed, will link residential areas with job hubs including the new Aerotropolis, and connect travellers from the new airport to the rest of Sydney’s public transport system.
Albeit a disconnected piece of railway to start with, it is intended to be a part of bigger network extending to Campbelltown in the south and to Rouse Hill in the North. As such it will be operated by driverless single decker metro style rolling as now operating on Metro Northwest.
Proposed Sydney West Airport Metro
More or less simultaneously with the airport and airport rail, the Greater Sydney Commission pronounced that Greater Sydney was, in fact, to become three 30 minute cities, “where most residents live within 30 minutes of their jobs, education and health facilities, services and great places” the Eastern Harbour City (read Sydney), the Central River City (read Parramatta) and the Western Parkland City (read an airport and co-located business precincts surrounded at various distances from 11kms to 24kms by regional conurbations - not by any standpoint point a traditional city structure).
City of Three Cities
This seems a nice, neat concept in principle - until, of course, you add in people and all their complexity and multiplicity of choices about jobs, culture, and lifestyle into the equation, as mentioned previously.
Cities and the people in them simply function in a more complex manner than lines on maps of land use zones and transport links imply.
Nevertheless, if that aspiration is ever to be fulfilled there is no doubt that those three Cities need to be linked by high-capacity public transport systems, in addition to the motorway systems in place already, and the GSC’s diagram does suggest this. Already there is advanced planning for a metro link between the Eastern Harbour City and the Central River City and corridor reservation for the extension of this to Western Sydney Airport is under way.
This is all good - but it then begs the question:
Given the massive cost of railways, why is Government not first building the railway which directly links these three major foci of employment and economic activity, as well as being the most likely sources of patronage for WSA beyond a 20 km arc, within which people will mostly use road-based transport? Why is government building a railway which could well be an under-performing stranded asset? It is very difficult to find the real answer to that question, at least in practical terms.
It rather appears that the Commonwealth, instead of insisting on its funding be used for a rail link that would make the airport more valuable, decided to accede to the State’s wishes for a new suburban railway - which happens to link to the airport.
Some relevant history from my personal experience
25 years ago, I led the study for a rail link to the site of WSA – the report is on my desk as I write! At that time the site for airport had been fully acquired at Badgerys Creek and there was renewed interest in planning the airport, but there was scant detail as to what form such an airport that might take. A rail link to Leppington, 10kms southwest of Badgerys Creek was already in the planning stage and the task was to find an alignment to extend that link.
It would be an extension of the existing metropolitan rail network and able to operate double deck trains. With the building of the Airport Line to Sydney Airport (SYD), had that extension been built it would then be possible for passengers transiting WSA to take a train to/from the Sydney CBD via Sydney Airport to/from WSA.
At that time, the job was one of route selection and not thinking about the calibre of the service – rolling stock, travel times, sources of patrons and the like - that might be needed to make WSA attractive to patrons, but one important criterion was to minimise the impact on private property and hence the need for property acquisitions.
Given the grief now being experienced by some landholders over compulsory acquisition and the calls for “a significant review into the state’s compulsory acquisition laws”  it was a prescient requirement. A route was duly found which reasonably achieved the latter (or so we thought, though of course, it has never been tested) and had a terminal point inside the airport site boundary. But there was little consideration of how this might interact with the airport terminal facilities and achieve an efficient interchange of passengers. But, as mentioned earlier, this route has been discounted by Government, although again at least the step of corridor protection is proposed.
Roll the clock on to 2010 and, with a lot of work on Sydney Airport’s rail link and high-speed rail in between, I was fortunate to lead the airport planning and engineering team which proved that Badgerys Creek was – while maybe not the world’s best new airport site – certainly the best one that Sydney had! During that work I had discussions with a representative of the Davies Commission which was looking into additional airport capacity for London.
Their view was that the single most important site selection criteria was the calibre of ground transport and accessibility. In the case of WSA, of course, the site was already determined so, by reverse, that left providing ground transport of the appropriate calibre which would contribute to the airport’s success. That means, in my opinion, linking it as directly as possible to the River City and Eastern Harbour City.
And so as, by now I was thinking of airport transport access far less as civil engineering but far more as a service to users, I did wonder why anyone from the Sydney CBD would bump along for well over an hour on a suburban train through one airport to access another – and it was this that started me to study and rethink what an airport railway to WSA ought to be and ought to do.
After 2014, I was a member of the team preparing the airport master plan and business case and, if I made any useful contribution at all, it was to promote that the master plan should provide for 4 rail tracks running through the site and for pedestrian access within the terminal complex to/from rail to be very low friction with no pedestrian flows crossing roadways, and causing traffic delays, such as happens at most Australian airports. This would enable a mix of services to be provided to and through the airport similar for example to the new Berlin Airport and as at Copenhagen. The provision also extends to two stations – one to serve the terminal precinct and the other to serve an onsite business park. That provision of space for two pairs of tracks through the airport site, in my view, is one of the most valuable transport assets in metropolitan Sydney and that provision needs to be used for its highest and best use.
The burning question was, and still is, how best to use them – to where should they connect and what type of service should be provided?
Well, as noted earlier, NSW Government has already decided on this matter by proposing to use one pair to connect to the existing heavy rail system at point 22kms west of the River City CBD and 41 kms west of the Harbour city CBD, necessitating that passengers interchange between rail services. Strangely, a reading of the Western Sydney Airport Rail Needs Scoping Study - seems to offer little conclusive evidence why this was selected while making much of the need for an east–west connection.
The best that it does is to agree that it’s not needed on opening day for the airport but is needed to effect the city shaping potential of a north south rail link between Campbelltown and St Marys and to offer what seems to be a “political” planning rationale of the need to build a link in Western Sydney, which does not point to the Eastern Harbour City i.e. Sydney.
As such, what is being built is about land development, not about servicing Western Sydney Airport and as such like the Sydney Airport Rail link before it, once again a hybrid compromise, with more than a sprinkling of politics.
What can benchmarking other airport rail links teach us?
To understand the issue of airport rail links better – especially given that Sydney’s first attempt at an airport railway ultimately led to it being not a dedicated service but a part of the suburban network that incidentally passes through and serves SYD and accordingly sub-optimal as an airport railway – I undertook a number of global benchmarking research studies looking at the characteristics of over 50 airport around the world to find how they vary and what were the characteristics of those which attract have a high proportion of airport travellers.
The figure below shows the variability of patronage of such links.
Mode Share to Rail for 51 Airports on 5 Continents
This showed that, in the absence of expensive patronage forecasts (which so often prove wrong anyway), it is possible to make a prediction of likely patronage of an airport railway based on airport passenger forecasts and linking likely sources of high patronage generation such as CBDs, based on factors such as:
- Road distance to common CBD location;
- Best road time to common CBD location;
- Worst road time to common CBD location;
- Rail travel time to a common CBD location;
- Rail service headway;
- Taxi fare;
- Airport parking cost;
- Rail fare;
In simple terms though, such global benchmarking made it evident that, inter alia:
- In cultures such as Australia, airport railways need to take airport passengers as close to where they need to go as possible – sort of obvious really! Otherwise, they will just use road-based modes.
- Rail’s global average market share of ground transportation to /from airports is about 20%;
- Market share for airport rail links around the world is seemingly most influenced by cultural attitude to use of rail transport – no airport link in North America, Australia or Africa exceeds 20% mode share; some do in Asia, and many do in Europe - where, it should be said, integrated and fit-for-purpose planning makes the offering much more attractive, in both pricing and timing.
Some Key Points for Successful Airport Railways
Following on from this global assessment I looked more closely at the situation in Sydney, with one very established airport SYD close to the major CBD and WSA to be built over 41 kms away and even more than 11 kms away from the closest major regional centre, as these figures show.
The optimal rail link
In summary, my conclusions were that to be successful, as an airport railway, a railway to Western Sydney Airport would need:
- To link places of high passenger generation capability and growth potential locations - Parramatta and Sydney CBDs are the obvious targets;
- Have a Sydney CBD to WSA time of 30-35 minutes;
- Offer fares competitive with alternative modes – average global rail fares being about AUD15 one way (the road tolls alone to WSA will far exceed this);
- Provide fully seated, not commuter standing and crowded, airport style rolling stock with luggage storage provisions;
- Less than 15-minute headways;
- Have a clearly identifiable CBD point of “low friction” access;
- Similarly, a “low friction” airport station that is more findable than a taxi rank;
- To focus on a few key interchanging locations – to enable connections to other transport systems and regional locations.
vs The reality
In contrast, my assessment of what is being built is that it would:
- Deliver a travel time of about 83 minutes from Sydney CBD which simply is uncompetitive;
- Lack a clear CBD point of access to the service, although arguably because of using the existing metropolitan system many points of access are available in the CBD - albeit many are medium or high “friction”;
- Not offer airport style rolling stock as, for example, on links like Arlanda Express in Stockholm and Gardermoen in Oslo;
- Requiring rail system interchange at St Marys with at least a degree of interchange “friction”;
- At least on the positive side, those who are prepared for a lengthy journey have many ways to get to St Marys to access the service – one just wonders if there will be very many;
- Not facilitate running a regional fast rail service to enable convenient access for Southern Highlands passengers.
It is not just the success of the airport at stake
And a further and potentially much more serious problem for what is now being built, given quite evidently it is not being built to focus on supporting the success of the airport, is that, and accordingly Dr John Kasarda, the person widely credited with the concept of aerotropolis employment precincts, it will be “dead in the water unless there is a fast rail link to Sydney”.
As such, it may suffer a worse fate than the “multifunction polis” – a futuristic city of high technology promoted by the Japanese Government and proposed by Australia to be built on saltmarshes north of Adelaide – at least in that case nothing was built, although by 1998 the cost of the failed project to the Australian taxpayer was reputedly already AUD150 million.
Infrastructure Australia released an evaluation report in Feb-2021 on the project now proceeding wherein they assessed that this project should not be included on the Infrastructure Priority List “at this time”. Specifically, this was because they found that the P90 capital cost of the project would outweigh benefits by AUD1.8 billion in NPV terms, because less than 40% of capacity would be utilised in peak periods and, of the AUD5.545 billion in benefits, 64% would go to “urban development (land value uplift, sustainability avoided infrastructure option value” , with only 18% to public transport users. There seems to be no mention of how much of that 64% would accrue back to the taxpayer through value capture mechanisms of which there has been much debate in the press.
Accordingly, I was recently moved to write this letter to the Sydney Morning Herald (SMH) of which they published an edited but unfortunately mangled version – somewhat missing the key point about linking the three designated cities under the Greater Sydney Commission’s plan.
“It comes as no surprise to me, as a consultant in transportation and having led the planning and engineering team in the joint Commonwealth–State study which proved Badgerys Creek as the best site metropolitan Sydney had available for another airport, as well as participating in the following business case studies, that the proposed airport metro running south from St Mary’s does not stack up in IA’s economic evaluation terms.
It was always considered that rail access to the airport would not be needed till well into the 2030s, based on access to the airport alone. However, the proposed metro is, like the railway providing access to Sydney airport, a railway which (only) incidentally provides airport access while attempting to serve other urban and employment access needs as well. Therein lies the problem – hybridising purpose can lead to dual inefficiencies.
My own research indicates almost no one within a 10-20 km radius will use rail to access the airport - roads and parking will be too convenient - and I doubt they will do so for proposed employment centres either.
The Greater Sydney Commission has sold the Government on the concept of the three cities and yet, instead of investing scarce capital into a fast rapid transit link between the three foci of that strategy – the CBD, Parramatta and the Airport/Aerotropolis - and capable of extension to the Southern Highland and beyond, Government has chosen to spend our money on a disconnected piece of railway which cannot deliver competitive travel times to the places which people who might want to use rail will really want to go.”
Useful and successful airport rail links
So, what might a useful and successful Western Sydney Airport rail link look like? In my view, there are now several good models to consider, with the Scandinavian cities having some good examples.
Comparison of Airport Rail Links to Major CBDs
What do these do well? Among other things:
- Specific well identified places to journey to and from in major CBDs;
- In the case of Copenhagen and Oslo, the ability to access long distance services – analogous to a regional service in NSW;
- Very good low “friction” on foot accessibility at both the airport and the CBD ends;
- Rolling stock that is – if not always exclusively for airport passengers - very well-designed with plenty of seating, on board facilities and luggage space. Sydney Metro Trains have only seating for 30% of capacity – though if my assessment is correct getting a seat will not be a problem for a long while, at least to/from St Mary’s – to/from there to anywhere will be a different story;
- Frequent and speedy services, at what seemed reasonable fares at least compared to the alternatives.
But possibly still the best model we could follow is one of the older ones - Hong Kong’s airport railway to the new airport on Lantau Island at Chep Lap Kok which opened in 1998.
In stark contrast to the situation at WSA with the Commonwealth planning the airport and the State the land transport access systems, that airport and its land transport access were planned and designed by one Authority which was able then effectively balance the competing needs of fast airport and all stopping commuter services. This was achieved by designing the infrastructure so that each service type “perches” on mostly the same physical infrastructure and with the services being highly effectively scheduled so that a traveller on the airport service is unaware of the other commuter service and vice versa.
Corridor Reservation for Direct Connection to Parramatta
This could have been achieved if the Sydney West Metro now under way from the CBD to Parramatta were to have been configured like the Hong Kong Airport link and continued to WSA as is in fact planned. But construction started on Sydney Metro West in 2020. Based on the performance of the Sydney Northwest metro now in operation, I estimate a WSA to CBD travel time of about 55 minutes.
This is woefully more than the 30 – 35 minutes that my analysis said was needed and nothing like the Hong Kong airport railway's performance. Furthermore, if rolling stock is configured like Sydney Metro stock (and there is nothing being said to the contrary), then the service to airport passengers will be poor.
Finally, while there is to be a Sydney station for this railway in Hunter Street there is no information emerging of how that station should function and be set up to handle airport travellers.
Neither is there any planning it seems that integrates this railway with the City Circle Railway which could be achieved if this railway continued onto St James Station and made use of the spare pair of platforms that JJC Bradfield thoughtfully provided. A fast service on this alignment could not only serve Western Sydney Airport from the east but be extended to serve Goulburn and the Southern Highlands as a part of TfNSW’s Fast Rail project.
In conclusion: not a transport plan for Western Sydney Airport
In conclusion, NSW planning for servicing the new Western Sydney Airport by rail seems to be almost absent with the focus on providing commuter rail services to its pet urban development project.
This is disappointing in the extreme not the least because even there NSW looks like building the wrong rail link first - all of which risks the success of the Aerotropolis. Moreover, instead of enhancing the AUD5 billion investment the Commonwealth is making in Western Sydney Airport and the another AUD5 billion in the metro link to St Marys, the rail link may end up looking like a stranded asset and fail to use those 4 tracks most effectively through the airport precinct.
The latter is all the more surprising as maximising passenger throughput at WSA would allow the Commonwealth in due course to sell a long-term lease of the airport.
This is a major public investment with the potential not just for all the direct benefits of a new airport (and they are manifold), but for the many billions of dollars' return on investment once the airport is sold, probably in the 2030s. The recent private sale of Sydney KSA valued that airport at some AUD24 billion. It is clearly in the medium to long term taxpayers' interests to maximise the value of WSA, something that will depend greatly on surface accessibility. Inadequate airport and Aerotropolis access will both push back the sale date and reduce the asset value for the taxpayer.
And a further comment from CAPA: Remember the environment?
The prospect of tens of millions of car journeys annually to the airport may be mouthwatering for an airport that has carparking facilities. But it will be many years before all cars approach anything like carbon neutrality in this country.
Any infrastructure planning that involves ultra-long term activity like an airport in this time of accelerated need for reducing greenhouse gas emissions should today begin with designing appropriate access modes. Fast rail links to key centres are an essential part of that equation.
For some earlier CAPA reports on the issue, please see:
Peter Thornton is Director, Transportation Associates
The comments, views, assertions and the like herein are those of the author personally and not of any clients of any of the work the author was involved in as mentioned herein and are made only in regard to information that is in the public domain.
 Led by EY/L&B
 Which failed commercially a private sector build-own-operate venture within a few years when patronage through the station turnstiles did come up to the forecasts.
 “Low friction” is my jargon for easy of movement for passengers – flat and level access suitable for rolling luggage is low friction – think Hong Kong Airport – up and down lifts and escalators multiple levels and long tunnels is “high friction” – Think London Piccadilly line or Sydney International Station.
 Ultimately a technology park development based on a pre-existing campus of the University of South Australia was built with the adjacent residential development Mawson Lakes and a centrally located bus and train interchange. It is relevant to note that only 3.6% of people use rail service in or out https://en.wikipedia.org/wiki/Multifunction_Polis