Extending the Sydney Strategic Model to represent toll road and park-and-ride choices

Document Type

Conference Paper

Publication Date


Subject Area

mode - park and ride, mode - car, place - australasia, mode - rail


Sydney, toll road, park and ride, updated models, train


RAND Europe were commissioned by the Bureau of Transport Statistics, Transport NSW, to re-estimate the travel demand model components of the Sydney Strategic Transport Model (STM). As part of this work, they have extended the scope of the model to represent toll road and park-and-ride choices. The STM model is a tour-based model system, comprising: - a cohort model for predicting aggregate changes in licence holding by age and gender; - models of couple and individual licence holding; - models of household car ownership, with special attention to company cars; - a population synthesiser to forecast the future population by zone and segment; and - models of travel frequency, mode and destination choice at the individual level. A key component of the work undertaken was to update these models to use more recent (2004-2008) household survey data and to make a significant increase in the geographical detail (884 to 2690 zones). The paper will focus on the two key extensions to the model scope, both of which involved adding additional choice decisions to the mode-destination structure. The choice to travel by tolled or untolled routes was modelled as a nest beneath car driver, with separate alternatives for tolled and untolled routes. Separate assignments were made to provide the level of service for these two alternatives, and these will be discussed in depth in the paper. A key finding from the runs was that long-distance travellers were much more likely to choose to use toll roads, even in those purposes like commute where willingness to pay was segmented by income. Park-and-ride was modelled by explicitly representing different access mode options to train (park-and-ride, kiss-and-ride, other), and for car choice of station. The approach used was based on the methodology developed for the PRISM model for the West Midlands region of the UK, presented at the 2005 ETC. The paper will report on specific adjustments that were made to the model in order that the observed variation in observed access mode shares with distance to stations was replicated in base year model predictions. Finally, the paper will summarise the structural parameters in the models, given that the following choices are simultaneously represented in the nesting structure: - main mode choice - public transport mode choice - train transport access mode choice - train station choice - destination choice - toll road choice The tests, including elasticity tests, that were done to obtain the best hierarchy representation will be explained.


Permission to publish the abstract has been given by AET, copyright remains with them.