Growth Impacts from SWIRL in the Illawarra and Greater Sydney

Vision Illawarra is SMART’s multi-purpose regional simulation model used by clients (like local governments) to better understand and plan for their infrastructure requirements in the future. Stepping away from traditional trend analysis, this model integrates demographic, economic, land use and transport dynamics as endogenous processes, allowing for more realistic projections.

The model was upgraded to include the areas identified by the Greater Sydney Commission in their Greater Sydney Region Plan. Previously the Vision Illawarra study area only included the Wollongong, Shellharbour and Kiama LGA’s, it is now calibrated to include the following five districts:

1. Western Parkland City: Blue Mountains, Hawkesbury, Camden, Liverpool, Campbelltown, Penrith, Fairfield and Wollondilly.
2. Central River City: Blacktown, Cumberland, Parramatta and the Hills.  
3. Eastern Harbour City: Bayside, Burwood, Canada Bay, Sydney, Inner West, Randwick, Strathfield, Waverley and Woollahra.
4. North District: Hornsby, Hunter’s Hill, Ku-ring-gai, Lane Cove, Mosman, North Sydney, Northern Beaches, Ryde, and Willoughby.
5. South District: Canterbury–Bankstown, Georges River, and Sutherland.

The Illawarra region is geographically defined as the narrow strip of land between the Illawarra escarpment and the Pacific Ocean bordering the southern edge of Sydney in the north and extending to north of the Shoalhaven or South Coast region. The region comprises the three local government areas (LGAs) being Wollongong, Shellharbour and Kiama. The residential population of the Illawarra is around 306,576 which represents 3.9% of NSW population and 1.3% of Australia’s population (Australian Bureau of Statistics, 2018).

In 2017, Illawarra’s Gross Regional Product (GRP) was $22.324 billion, which represents nearly 4% of the NSW economy and 1.3% of the Australian economy (Remplan Economy, 2017). It’s estimated that 144,494 people work in the Illawarra which is 3.7% of people working in NSW and 1.2% of Australia’s employment (Australian Bureau of Statistics, 2018).

Source: Australian Bureau of Statistics (2018) and Remplan Economy Profile (2017).

Although it is geographically close to Australia’s largest labour market, the Illawarra region has historically suffered from relatively high unemployment, particularly youth unemployment. The Illawarra economy has been traditionally based on coal mining, steelmaking, as well as exporting coal from the port at Port Kembla.

However, in more recent years, the Illawarra economy has benefited significantly from substantial regional public investment in local health and education infrastructure, supporting tens of thousands of jobs in the region. For instance, based on the 2016 Census data, the three greatest employment industries in the Illawarra are Health Care and Social Assistance (with 16,920 jobs), Education and Training (11,962), and Retail Trade (11,393), which together support 40% of jobs in the region (Australia Bureau of Statistics, 2017).

The connectivity within and out of the Illawarra region is supported by two rail lines, the Princes Highway, local motorways, local roads and Port Kembla.

The road transport network affords around 21,614 commuters daily between the Illawarra region and Sydney for work, study (Australian Bureau of Statistics, 2016). Besides road transport, Sydney airport serves as the nearest major airport to this region. An Illawarra regional airport is located at Albion Park, with Fly Corporate offering flights to Brisbane and Melbourne. Port Kembla plays an important role in supporting motor vehicle, mining, agricultural, manufacturing and construction industries in NSW.

As an economic driver in the Illawarra region, the port of Port Kembla is home to the state’s largest motor vehicle import hub and grain export terminal, and is the second largest coal export port in NSW (NSW Ports, 2015). Currently, 60-65 percent of freight travelling to and from Port Kembla is transported by rail on either the Illawarra Line or the Moss Vale to Unanderra Line. Operations on both lines are limited by passenger rail services in the region, resulting in disruptions to freight scheduling. Queuing of up to 11 hours is common as passenger services are given priority.

As a part of NSW Ports’ 30 Year Master Plan, the further expansion plan for the port of Port Kembla predicts a bright future and tremendous opportunities for new business and job growth (680 extra jobs and $170 million would be generated from ship arrivals alone) in the Illawarra region (NSW Ports, 2015). However, there remain a number of key challenges to be addressed, among which, one of the most pressing issues is land transport connection to and from the port. Inadequate rail freight capacity may lead to a substantial increase in road freight, further constraining the Illawarra region’s road network.

It has been identified that Port Kembla would face capacity constraints without any additional rail network improvements. Port Kembla is a significant economic asset for NSW and Australia, which is why maintaining efficient movement of freight to and from the port is essential for the Illawarra region. 

The limited capacity of the existing Sydney airport is a significant constraint to aviation and economic growth in the Sydney basin. To meet the need of a growing and changing population, the Greater Sydney Region Plan seeks to transform Greater Sydney into a metropolis of three cities, including the Western Parkland City and a potential Outer Sydney Orbital connecting it to Greater Newcastle, Wollongong and Canberra (Greater Sydney Commission, 2018).

Western Sydney’s industrial growth, both in and around Western Sydney Airport and Badgerys Creek Aerotropolis, will stimulate large residential growth in the South West of Sydney (around 35,000 houses will be developed), and create additional demand for freight connectivity between Western Sydney and the port of Port Kembla (Regional Development Australia Illawarra, 2017). Moreover, an additional 65,000 people are expected to live in the Illawarra and Central Coast, and an additional one million people in Western Sydney by 2031. The broader Western Sydney employment area is expected to provide 212,000 new jobs in the future (Infrastructure Australia, 2016).

In the absence of long term planning and enhanced transport connectivity, the existing transport network will reach its capacity and result in congestion and long travel time. On the Infrastructure Priority List, a multi-modal transport corridor connecting Central Coast, Western Sydney and Illawarra region is expected to be identified, which comprises a motorway, a north-south freight rail line and proper passenger rail line.

As one of the major rail lines connecting Illawarra region, the South Coast Line (SCO) is facing growing demand of passenger and freight, while its infrastructure facility and service ability falls behind the demand. According to the prediction by Infrastructure Australia, the line would reach its capacity by 2020.

The SCO runs from Bondi Junction through Central Station to Port Kembla, further south to Bomaderry station in Nowra, Shoalhaven and has been recognised as a key strategic passenger rail corridor (Department of Infrastructure and Regional Development, 2017). The line is 153 kilometres in length with predominantly single-track and diesel trains serving the South of Wollongong between Kiama and Bomaderry for both passenger and freight transport. North of Wollongong, it is dual track apart from the Clifton Tunnel (1 kilometre) between Scarborough and Coalcliff. Due to the unique features of the Illawarra’s escarpment, landslides occur unpredictably, and this type of safety issue is always a concern for the SCO, usually resulting in line closure.

For freight service, the SCO also serves as a main line to connect Port Kembla to Sydney, and regional NSW by providing an average of 23 freight services (mostly overnight) every day. In the meanwhile, the SCO runs 20 hours daily and provides 28 scheduled services from the Illawarra to Sydney and 27 services in the opposite direction for passengers. It takes 87 minutes to travel from Wollongong to Central Station (82 kilometres) in Sydney during weekday morning peak. To reach Parramatta station at the geographic centre of Sydney, it takes 114 minutes at best. On average, the travel time between Wollongong and stations in southwest Sydney, such as Campbelltown (131 minutes), Leppington (130 minutes) and Liverpool (125 minutes) are all over two hours. Hence, many commuters consider taking “park-and-ride”; driving 40 km roughly along Princes Motorway and parking at stations like Waterfall, Heathcote or Sutherland before traffic congestion becomes serious in the city centre. This lengthy commute for workers, students and day-trippers reduces economic and social opportunities for both Illawarra and Sydney residents in terms of accessing a wider range of jobs, business, trade, education, leisure and housing choices.

Illawarra First and the SMART Infrastructure Facility, University of Wollongong, released a detailed report examining the cost and benefit of a new rail line, referred to as the SWIRL. The SWIRL would complete and electrify the Maldon-Dombarton Line (35 kms), together with electrification of a small segment of the Unanderra-Moss Vale Line (7kms). The SWIRL would provide a critical passenger and freight service, creating a vital transport connection between the Illawarra and Sydney.

The SWIRL will play an important role in connecting Australia’s largest economy (Sydney), the third largest economy (Western Sydney) and the Illawarra, NSW’s third largest economy.

In SMART’s view, SWIRL will meet the transport connectivity objectives set by Transport for NSW and provide many economic and social benefits for residents in the Illawarra and Western Sydney. For instance, the SWIRL and the SCO operating together would increase total passenger and freight rail network capacity and open up jobs, business, trade, education, leisure and housing opportunities for both regions. A new line could also limit the cost of congestion, short-term closures for upgrades or a geological failure on the South Coast Line.

Shutting the South Coast Line in the event of a significant rock fall/landslide on the Illawarra escarpment would severely disrupt the region’s economy, preventing coal exports and container imports as well as forcing rail commuters onto already congested roads in peak times. The availability of a second passenger and freight line between Sydney and the Illawarra would greatly reduce these costs.

The economic model acts as the centre that constrains the factors of production (land, labour and capital) such that the VI simulations of future population and employment growth are reasonable. To this end, VI forecasts population and employment growth in order to understand future land use patterns including, for example, land supply requirements for augmented transport networks, new housing and new public services such as hospitals and schools. In other words, more people and more jobs in the Illawarra creates additional demand for economic and social infrastructure to support those new residents and workers.

As a general observation, in modern market-based economies, the pool of available labour is usually close to being ‘fully employed’. Full employment is not defined as an unemployment rate that is equal to zero, rather, the point where the unemployment rate is at its ‘natural rate’. The natural rate of unemployment in Australia is 4.5 to 6.0%; however, this rate can exhibit wide variation across regions and time.

Therefore, when forecasting employment growth in the long-run it is best to first align those forecasts to long term trends in employment growth, labour force growth and population growth such that (by assumption) the long term average unemployment rate remains constant. If we assume no ongoing structural change to the labour market, then employment growth is simply a function of population growth.

At a regional level, structural changes impact the relationship between population growth and employment growth. However, in the long term and for this model the employment of the Illawarra and Greater Sydney regions will depend on the population growth of each region.  These employment figures will be used as the inputs to the transport and land use modules.

The population module provides main demographic indicators such as population, household and dwelling estimations and projections to support economic, transportation and land use simulation. In the simulation module, the total population data are inputs that act as constraints to developments and changes of economic, transport and land use.

The population module compares and integrates multi-sourced population projections from: Transport for NSW (TfNSW); Department of Planning and Environment (DPE) of NSW Government; the Wollongong City Council (.id); and SMART’s population estimation based on ABS LGA population data 2001-2017. It was observed that while WCC and SMART estimations are presumably based on natural population growth related to migration, births and deaths; the DPE and TfNSW projections take into account the upcoming infrastructure projects, in particular, the changes in Western Sydney. Therefore, in order to meet the objectives of this project and the upcoming commencement of planned infrastructure facilities, the decision made was to use DPE’s projections for population, household and dwelling as the main population data source. This projection, however, was limited to the year 2036 so the data was extended and forecasted for a further 20 years from 2036 to 2056.

With the simulated changes in population, it is noted that family structure and composition, for the Wollongong region maintains almost the same proportions (see below). However, couples and lone or single person households will increase more quickly than families with children. For example, in 2016, the couple only and single person households account for about 51.1% of the total households, it increases to 53.7% by 2036 and 55.2% by 2056; however, the family with children household type drops from 31.2% in 2016 to 29.6% in 2036 and 28.7% in 2056. Another noticeable change in household composition is that the proportion of group and multiple-family households drops. The total percentage of group and multiple-family households in 2016 is about 6%, while the figure drops to 5.44% in 2036 and 5.0% by 2056. These changes could impact local industry structures, transportation and social services.

The changes to the age of the population is another issue observed from the population module. According to ABS data, the 65 year old and over population in Wollongong is about 16% in total population in 2016. It is expected that the proportion of 65 year old and over population will exceed 20% around 2036 and approach 25% in 2056. The increasing aged population and the demands associated will need to be considered in government strategy planning and decision making.

The chart below shows approximate forecasts from 2011 to 2036  for those in the Illawarra who are of working age and those who are not.

Vision Illawarra is a regional interaction model is based on the theory of spatial interaction, it explains the flow of people, goods and services between two or more locations in space. It determines the movement of jobs and people between regions in the study area based on relative attractiveness of regions and the ease by which they can be reached. In addition, the regional model converts the number of jobs and number of people specified by the economic and demographic models into a land area requirement based on the productivity in each region before passing the same data down to the land use model for spatial allocation.

This type of model is known as a constrained Cellular Automata (CA) based land use model and has been used in many applications around the world, both in its original form and its variants. The main variable of the land use model is a land use map in raster format which is iteratively updated in yearly time periods over the course of the simulation. The demands for the land are specified by the economic and population models, hence the land use model is constrained. The land use model then allocates this demand spatially. Land use transitions are driven by an endogenously calculated transition potential for which neighbourhood effect, physical suitability, accessibility (mainly to transport) and zoning contribute as key factors.

The transport model is based on a classical four step approach, but has been made dynamic. The land use model provides input to determine origins and destinations of trips, while the transport model influences land use change by means of accessibility, which is fed back to both the land use model and the regional model. This two-way communication provides the dynamic basis to the transport model.

The regional model is primarily set up to cover three aggregate regions: the Illawarra, Western Parkland City and the rest of Sydney. The land use model is set up with 7 land use classes out of which two dynamic land use classes, residential and commercial/industrial land, are mapped to the number of people and number of jobs in the study area. The Transport model recognises three modes (car, bus, train); four purposes (work, shopping, recreation, home); and four times of day (morning peak, interpeak, afternoon peak, and evening).   

An important input into the transport model is a set of matrices of trip counts, average trip distance, and average trip time by mode (car, bus, train) by time period (morning peak, interpeak, afternoon peak, evening) by purpose (work, shopping, recreation, home). The morning peak period is defined as the hours between 7.00 and 9.00, the inter peak from 9.00 to 15.00, the afternoon peak from 15.00 to 18.00, and evening period from 18.00 to 7.00 the next day. These matrices were required for 2011 and 2016 to calibrate parameters in the transport model.

These matrices for Metronamica transport model were derived from a set of matrices originally provided by the Transport and Performance Analytics (TPA) at Transport for NSW. TPA matrices were produced from the Sydney Strategic Transport Model for 2011 and 2016.

The results of the economic module shows that for the base case scenario, employment in Wollongong increases from  96,707 in 2031 to 128,626 in 2050. The total Illawarra employment figures, increase from 129,216 in 2031, to 171,865 in 2050. This equates to an increase of  1.44% on average per annum for the Illawarra for the study period.

Simulation outputs suggest that the Illawarra region will benefit significantly from the SWIRL project. Enhanced connectivity brought about by SWIRL is likely to influence the activity (people and jobs) movement in the greater region to the advantage of the Illawarra region. According to the simulation outputs, SWIRL can potentially increase the Illawarra’s population and jobs growth rate by up to 0.2%.

SWIRL will influence a substantial shift in the modes of travel. Under the base case scenario, car trips made between the Wollongong CBD and Western Parkland City are likely to grow by about 0.9% per annum during the period 2016 to 2050. For the same period, however, under the project case scenario car trips between these two destinations grow at a significantly slower rate (0.1% per annum). In contrast, train trips between these two destinations grow at a rate of 1.2% per annum in the SWIRL scenario compared to a slower rate of 0.7% per annum in the baseline scenario. This shift in transport dynamics between the baseline and the SWIRL scenario can also be assessed at the end of simulation period (year 2050) as observed below.

This mode shift is even more apparent for trips between the rest of the Illawarra Region and the Western Parkland City (WPC). In the baseline scenario, the car trips increase by 0.4% annually while train trips drop by 1.9% annually.

Opposite trends are observed in the project case in which car trips drop by 0.1% and rail demand increase by a significant 6.9% annually. It should be noted that the 2016 rail demand between the rest of the Illawarra Region (which excludes Wollongong Station and North Wollongong Station) and WPC Station was small, (approximately 300 trips daily), hence the large annual growth of demand for rail in the project case in which SWIRL is operational. Nonetheless, these growth rates underline the significant impact of SWIRL to the shift toward rail for trips between the Illawarra Region and the WPC.

Annual growth rate of car trips and train trips between the Illawarra region and the Western Parkland City.

Note that the two main stations in the Illawarra region, Wollongong Station and the North Wollongong Station, were assumed part of the Wollongong CBD in the calculations of the rates in Table 7. This explains the opposite trend of rail demand in the baseline scenario between the WPC and the Wollongong CBD (growing at 0.7% annually) and between the WPC and the rest of the Illawarra region (dropping by 1.9%).

However, the simulation results do not reveal any significant difference in transport activity within Wollongong CBD for comparisons made between the baseline and SWIRL scenarios. Given this model covers a large geographic area (most of Sydney GMA) and a detailed calibration has not been performed, it is concluded that the model in its current form is not sensitive enough to simulate such micro level changes. This shortcoming will be addressed in the next version of the Vision Illawarra model to be released in 2019.

Overall, the results indicate the road network will be much more congested by the year 2050 compared to the present unless substantial infrastructure upgrades are carried out.

The images below illustrate the relative congestion (shown in red) in the road network between 2016 and 2050 for the base case scenario.

SWIRL-induced population and employment growth inevitably increases the trip generation in the Illawarra region causing the road network to congest further. On average, a trip made by car within the Illawarra region under the SWIRL scenario is going to take about 4% more travel time compared to a similar trip made under the baseline scenario in year 2050.

The land use model successfully allocates new growth in anticipated areas, for example in and around Badgerys Creek airport, earmarked Western Sydney growth areas, as well as Illawarra’s growth areas around Horsley and Calderwood

Simulated land use change in the Western Parkland City and the Illawarra between 2016 and 2050.

The study confirms that the SWIRL project will create a positive change in the Illawarra region both in terms of the population growth and employment growth. According to the simulation outputs, SWIRL can potentially increase the Illawarra’s population and jobs growth rate by up to 0.2%.

The population and employment growth will lead to an increased demand for travel within the Illawarra region putting more pressure on already congested parts of the road network. This has implications on the travel time within Illawarra during peak demand periods, and the model in its current form estimates this to be about a 4% increase in travel time for SWIRL scenario compared to the baseline in year 2050. Furthermore, SWIRL is shown to induce a shift in transport mode share towards train. Between Wollongong CBD and the Western Parkland City, car trips grow at a much slower annual rate of 0.1% in the SWIRL scenario compared to an annual rate of 0.9% in the baseline. Train trips, on the other hand between these two destinations grow at an annual rate of 1.2% in the SWIRL scenario compared to the annual rate of 0.7% in the baseline.

The extended Vision Illawarra model successfully allocates key land use types in correct locations, for example, growing areas around Badgerys Creek airport and Horsley/Calderwood. Transport model has also been able to identify network links that are likely to congest over time.

Although results described in this section provide confidence to use the extended Vision Illawarra model to understand broad implications of a project like SWIRL, it should be noted that the limited scope and timeframe of this pilot project meant that the model is only partially calibrated. It is recommended that the extended Vision Illawarra model be further calibrated and validated prior to using it for accurately estimating the magnitude of change caused by policy decisions and infrastructure investments.