Why Fast Trains Work: An Assessment of a Fast Regional Rail System in Perth, Australia

Perth’s new 72 km long Southern Rail System opened in 2007. With a maximum speed of 137 km/hr and an average speed of almost 90 km/hr this system acts more like a new high speed rail than a suburban rail system, which in Australia typically averages around 40 km/hr for an all-stops services. The Southern Rail Line was very controversial when being planned as the urban areas served are not at all typical of those normally provided with rail but instead were highly car dependent and scattered low density land uses. Nevertheless it has been remarkably successful, carrying over 70,000 people per day (five times the patronage on the express buses it replaced) and has reached the patronage levels predicted for 2021 a decade ahead of time. The reasons for this success are analyzed and include well-designed interchanges, careful integration of bus services, the use of integrated ticketing and fares without transfer penalties and, crucially the high speed of the system when compared to competing car based trips. The Southern Rail Line in effect explodes the current paradigm of transfer penalties, exposing this as a myth. The lessons for transport planning in low density cities are significant, and are explored further in the paper.


Introduction
This paper will try to suggest why a fast rail line that runs within the urban region of Perth works so well and why this may help in the assessment of high speed rail projects. We will initially provide a background to the Southern Rail infrastructure and the catchment which it serves, followed by a description of the Southern Rail patronage model, the rail line's operational model, and a description of its station configurations. This background information will lead to a more detailed discussion of the time and financial costs, or "generalized cost" comparesons between private vehicles and public transport within the Southern Rail's catchment.
Through this lens the paper demonstrates that the Southern Rail line's success in generating a significant mode shift to public transport has occurred through physical, operational and regulatory integration. The paper concludes by examining the possible implications of high speed integrated urban rail systems for making travel in dispersed cities more sustainable.

Background to the Perth Southern Rail
The Perth Southern Suburbs rail line opened on the 23rd of December 2007, and it runs 72 km from the Perth CBD to Western Australia's second largest city at Mandurah at speeds up to 137 km/hr. The rail line runs for 70 km of its 72 km length at grade in the median and along the western edge of the Kwinana Freeway.
The line has eleven rail stations along the way at major centers and major road intersections including the underground Perth CBD and Esplanade Stations (see Figure 1), with 18 new bridges and 5500 park and ride bays [1].
The Perth Southern Suburbs rail line replaced an existing separated Rapid Busway-based service with an integrated multimodal bus-rail service. The introduction of the rail service cut the journey time from approximately 68 minutes to 48 minutes for the journey from Mandurah to Perth, and increased the existing busway's patronage from approximately 14,000 boarding's/day [2] to the rail line's 48,000 when opened in early 2008 and to 70,000 boarding's/day in 2013.
The Perth Southern Rail Line catchment density (presented in Figure 2) is mainly between 6 -15 dwellings per hectare, which is a very low dwelling density for rail-based public transport services to be considered viable. South of Murdoch station, the average density of development is even lower, with significant gaps of rural land between settlements.

Perth Southern Rail Patronage Model
The Perth Southern Rail Line was the second in Australia to be implemented in the median of an existing freeway (after the section to the north of Perth). As a result, there is only limited pedestrian catchment along the alignment. The business case patronage model for the Perth Southern rail line assumed [1]:  9% pedestrian (walk-able) catchment.
91% patronage to come from outside the pedestrian catchment, comprising.  28% of total patronage to come from the 5260 park and ride bays. 63% of the total patronage to come from kiss and ride and the feeder bus services. The low pedestrian catchment model for public transport infrastructure is a major shift for the roles of the different public transport modes, with the park and ride facilities and closely integrated bus interchanges designed to extend the rail line's catchments into the low density surrounding suburbs, using the regional bus network as a feeder service for areas that had previously been car dependent.
As discussed later, actual patronage has been significantly higher than forecast, particularly from the feeder bus system. As a result, 85% of the Southern Rail patrons access the train by a bus service, (or line to line transfer), around 8% from the park and ride and the remainder from pedestrian catchment and kiss and ride [4]. This patronage model appears to be a major part of the suc-cess of the rail system.
The other major difference to other public transport systems in Australia is that the public transport system was designed to be competitive with the private motor vehicle in both time and cost (Generalized cost). This low generalized cost model is a function of the station configurations and the rail line's patronage and operational model in this low residential and commercial density environment. The "low generalized interchange cost" station configuration will be described in detail later in the paper and show how all the time and financial cost components of the bus to rail interchange have been minimized.

Perth Southern Rail Operational Model
The transfer based operational model for the Southern Rail Line is based on a high level of convenience and a low generalized cost to passengers. The Western Australian Public Transport Authority (PTA) achieved this by providing an integrated multi-modal public transport service able to compete with the car for journeys from Perth's CBD to the key centers and southern suburbs.
The PTA maintains Strategic, Tactical and Operational (STO) control over Perth's public transport network, with bus operator contracts managed by the PTA, with the PTA retaining the service fare box revenue. The STO framework [5] defines the three levels, or tiers of activeties of the transport agency (see Figure 3).
The retention of Strategic and Tactical control over the network has enabled the PTA to deliver Perth's transfer oriented public transport system for the southern suburbs and through the introduction of specific operational attributes primarily:  Integrated network design and timetabling;  Integrated ticketing;  A zone based fare system that enables seamless bus interchange;  Additional elements such as the free Central Area Transfer buses.

Integrated Ticketing and Fares
The introduction of integrated multi-modal electronic ticketing prior to the opening of the Southern Rail Line enables convenient transferring between modes without the need to purchase separate tickets. In addition, the fare system used means there is no financial penalty for changing between modes, regardless of modal operator (all of whom are under tactical control of the PTA). The ticketing model for Perth is presented in Figure 5.

Zone Based Fare Structure Facilitating Bus Interchange
Perth's public transport broad zone based structure (shown in Figure 5) has enabled most, if not all, of the bus based feeder services to be "intra-zonal" trips. Whereby the bus based trip to one of the Southern (or Northern) Rail line stations for a multi-zonal journey does not incur an additional financial cost for the journey. The zone based fare structure therefore enables the feeder bus services to travel along the urban highways that cross the Southern Rail Line stations to quickly interchange people at no financial cost to the passenger thus removing the financial cost penalty to intermodal transfers for the line.
The feeder buses are integrated into stations by placing these at the points where highway overpasses occur thus leading to a direct pedestrian feed-in down an escalator from the bus stop to the train. The feeder buses are integrated into stations by placing these at the points where highway overpasses occur thus leading to a direct pedestrian feed-in down an escalator from the bus stop to the train.

Regional Bus Network Structured to Feed
The feeder bus patronage

Area Transfer Bus Services
r-based modes, and in particular public transport (it wholly

igurations the Southern Rail Line
Southern Rail Line's broad catchment is served by a comprehensive bus network that has been structured to enter the low density suburbs and quickly bring the bus patrons to the Southern Rail stations, as well as stations on other lines, to transfer to the rail network.
These transfers are not "forced" as the buses continue onto other key trip attractors and centres. However the system provides bus passengers the opportunity to transfer to a fast trip to the CBD (and other centres across the public transport network) at generally no additional financial cost as the majority of these bus trips are contained within a single travel zone (illustrated in Figures 4  and 5).

Central
The Perth "Central Area Transfer" bus (CAT bus) se vices operate free of charge from the rail stations at Fremantle and Joondalup and within the Perth City Council municipal boundaries as a "Free Transit Zone", thus minimizing the cost of travel at either end of the trip origin and destination at these major centers.
To fund the CAT services, the Perth Parking Management Act, 1999 operates within the Perth CBD, and hypothecates the revenue generated to alternative non car funds the capital and operating costs of the Perth CAT bus services). This not only acts as a public transport revenue source (and PT trip cost reduction due to the CAT services), but as an additional cost (and disincentive) to private vehicle trips to the CBD.

Perth Southern Rail Station Conf
As previously mentioned the Perth Southern Suburbs Rail Line corridor is located in the median of the Kwinana freeway for the majority of its distance, and as such the rail stations and their intermodal interchanges are located in the middle of the freeway as well. This constrained location reduces the opportunities for pedestrian catchments for the station. However it does have a number of advantages, in that the station does not require additional land, while the negative externalities of the rail operation are confined to an existing transport corridor.
The integration of a significant bus interchange at the entrance to each of the stations minimizes the transfer time cost between modes (as part of the low generalized cost trip model), and given its controlled nature makes a very safe environment to transfer between modes, and for accessing the park and ride facilities both during the day and at night. The challenges associated with its location are that it limits the amount of ancillary infrastructure that can be co-located within proximity of the station

Murdoch Station
Murdoch Station (shown in Figure 6) is no tion and operational attributes of one of the Perth South-Central). This is an increa Station had 6733 boardings per day, and is now up to 8383 passenger boardings in March, 2012 with 4950 (59%) of these being bus to train transfers, with approximately 700 park and ride spaces (8.3%) [9].
In the absence of any data for the remaining patronage that is unaccounted for (2733 daily boardings), it is expected that these are made up of Kiss and Ride, Cycle and Ride and pedestrian access modes. The station has effectively no residential pedestrian catchment given the walking distance from property to the station is over 800m. This is due to the configuration of the road network adjacent to the station, and the lack of any pedestrian walkways (this is illustrated on Figure 7). The station is however within walking distance to a new and existing hospital precinct (approximately 700 m -800 m) and is close to Murdoch University (approximately 1500 m), though designs for redevelopment of a TOD over the car park have been drawn up by the State Government.
The key attribute of the Southern Rail Line stations, and Murdoch station in particular, is the extensive feeder bus services that travel at relatively high speeds along the crossing urban highways (such as South Street at Murdoch Station), these attract patronage to the rail line due to its speed, as it is generally at no additional cost as the feeder services are "intra zonal" trips.

Perth Southern Rail Operational Modal Choice Model
The generalized cost (GC) to the user of eit private transport trip is characterized as a function of e component, a monet component, which can be expressed as [11,12];   ther road based or public transport congestion, that could be spent doing something more productive. The conce ic or private ip is integral development tr rat be pt of generalized cost of a publ in transport planning for the tr a ansport link's modal choice model through the prepaion of a modal utility (a quantitative measure of the nefit/dis-benefit of a particular transport mode). The Southern Rail modal choices can be viewed in a Nested Multinomial Logit model (Figure 8).
The modal utilities assigned to each mode are used to determine the probabilities of which mode would be used for a trip for the Southern Rail catchment to the CBD. Due to data limitations this paper has provided a quailtative presentation of the generalized cost input time cost, financial cost and opportunity costs components to understand the generalized cost for private vehicles and the Perth Southern rail line. The actual patronage figures presented later, validate the operational model proposed.
Whilst the financial cost of the car trip is increasing significantly due to the increases in cost of fuel and parking costs into the CBD (see Figure 9), it is the competitiveness of the Southern Rail Line with the private vehicle due to congestion on the competing Kwinana Freeway which gives the Southern Rail its real edge.
The effect of road based congestion on Perth's com-  trip to the CBD during the AM peak, and these are used to frame the mode choice discussion with regard to the Southern Rail catchment.
lized cost for the trips to th CBD, and although this is only qualitative, the impact rth Southern Suburbs rail in is model was independently re- muters has been verified by the Royal Automobile Club of W conucted their annual survey of 400 Perth businesses on the sp Table 1 clearly demonstrates that the public transport trip can provide a lower genera estern Australia (RACWA), which in 2012 e d on passenger's modal utility and subsequent mode choice is obvious. This is reflected in the growth in the patronage of the Southern Rail line.
effects of road based congestion. This survey showed clearly that congestion was having a significant impact on the operation of their businesses and staff. This is reflected in Figure 10, which disturbingly shows that 82% of business believed congestion was causing lost productivity and 74% believed it was reducing staff punctuality. This road based congestion on Perth's road network is at its peak on the northern Mitchell freeway and southern Kwinana freeway, both of which are served by the high

Perth Southern Suburbs Rail Operational Performance
The overall Perth rail network has increased patronage by 83% from 2007 to 2012, (sees Figure 11). The transport business case for the Pe eed rail network, and it is argued that it is this road based congestion and high speed nature of the competing rail journeys that is driving the modal shift over the public transport in the Northern and in particular Southern catchments. Table 1 qualitatively presents the generalized costs for a private vehicle and public transport "Journey to Work" line was developed for the WA State government, by the consulting firm Sinclair Knight Merz (SKM) initially 1999, and then th viewed by Planning and Transport Research Centre in WA (PATREC) in 2004 [2]. Neither SKM nor PATREC were able to model how successful the Southern Rail Line would be at attracting patronage from the low density southern suburbs of Perth (as shown in Figure 2), as SKM's 2031 patronage forecast for the project was exceeded in the Southern Rail's first year of operation. The significantly more optimistic pa for the perth southern rail mode choice model.

Discussion
tronage model prepared by PATREC [2] (shown as PATREC Rail in Figure 11) still underestimates the impact of the patronage from the Southern Rail line, where the estimated 2021 total annual boardings for the network were exceeded in 2012.
Indeed the 2031 patronage forecast for the line and the greater Perth rail network is likely to be exceeded in 2014/2015 assuming that rail line does not get affected by a capacity constraint incurred by insufficient rolling stock provision which subsequently limits the growth of the network (Figure 11).

Lessons of Fast Rail for Low Density Cities
One of the most contentious debates in public transport planning is the role of transfer costs in deterring patronage due to time losses. This has been used to stop many rail projects [18,19]. The Southern Rail was designed to minimize transfer penalties by well-integrated bus interchanges and bus services. As a result these are responsible for 85% of the patronage, a figure much higher than for most rail systems. This seems to be acceptable to the atrons as the speed of the train means that the overall p The current fare structure for public transport trips is shown in Figure 10 [7] The introduction of the electronic ticketing system enabled zero cost transfer penalty between the bus and rail modes, hence removing any transfer cost penalty associated with transferring bet us ween modes. The zone based fare structure also means that the majority of the feeder b s to the Southern Rail stations are intra zonal trips and are therefore nil cost to the overa trip cost. See Figure 11 for the Transperth zone map [7] Road Very low pact of Road Congestion has been significant for road users and is significant modal distribution for the Southern Railway catchment Public Transport nil med ortunity Cost Ability to use tr e leisure noticeable with th ange in this area w d conduct online lei major change in how public transport travel time is viewed when compared to private vehicle use where this time is dedicated to the purpose of driving on increasingly congested roads crowding Public transport crowding on the Southern rail and its feeder services is increasing, and this issue will continue until more rolling stock is added to the line during peak times.
(illustrated in Figure 9, [15]) The positive utility of rail based public transport travel time is becoming increasingly e introduction of electronic media and devices. A major ch Opp avel tim productively for work or low high as the introduction of WIFI onto the Southern Rail line, thus enabling passengers to work an sure activities on their journey to and from work. This is a journey saves time compared to the private vehicle. There would appear to be a lesson here for all rail planning but especially high speed rail: it is essential to minimize transfer penalties and create speeds that mean the generalized costs of choosing to travel on fast rail are lower than the alternatives (car, bus or plane).
The Perth Southern Rail Line has maintained a strong focus on the competitiveness of the public transport against the private motor vehicle in a time and financial generalized cost model for access from the southern suburbs to the Perth CBD and the rest of the rail network. This is illustrated by:  The competitiveness of the train to the car in time due to the high speed of the trains (maximum speed 137 km/hr) compared to the car (freeway speed limit 100 km/hr).
T the ind des would like to thank members of the We Australian Department of Transport and Western Australian Public Transport Authority for their support in providing data for this research. ilway_station ger Transport' Round T and speed of interchange between the bus and modes, minimizing the time penalty for transferringbetween modes  The introduction of the electronic ticketing system that enabled zero cost transfer between the bus and rail modes, hence removing any transfer cost penalty associated with transferring between modes.  The majority of the feeder bus services that transfer to the rail stations do very quickly and conveniently, and operate within the single transport ticketing zones, therefore not adding to the trip cost for the overall journey. This model for the minimization of public transport generalized cost when compared to the private vehicle has enabled Perth Southern Suburbs rail line to be very competitive with the private vehicle in a region where the Southern Rail Line is located. This area is where the private vehicle has been historically extremely dominant due to low dwelling and population density, and long travel distances for the journey to work. he low generalized cost network design minimizes time and financial cost for the multimodal trip and uces traditionally car based ridership onto the transfer igned feeder services to the stations, and this is the key success for the intermodal public transport model that operates for the Southern Rail Line.
The application of the Southern Rail model to the design of high speed intra-city and regional rail for large, low density cities as well as very fast intercity train systems would appear to be clear.