Sharing tracks in Sydney
AUSTRALIA: High-profile passenger operators battle for train paths with other urban, regional and national passenger and freight carriers. Sydney provides a case study of issues that confront rail operators across the globe.
Ian Imrie,Technical & Commercial Analyst, Plateway Pty Ltd
The volume of container traffic passing through the port of Sydney is expected to double over the next 17 years, posing a significant capacity challenge for both freight and passenger operators in the city. With the cost of separate infrastructure for freight traffic likely to be prohibitive for the foreseeable future, greater use of planning software, traction and braking technology and some further grade-separation of junctions at key bottlenecks would permit more efficient use of limited paths.
Australia is an example of a medium-ranked economy undergoing rapid growth fuelled by the Chinese and Indian resource boom. Australian manufacturing industry now exists only in niche pockets, and 'old' industries have been replaced by containers of goods arriving at the ports from sources predominately in Asia.
The dominant Australian exports are bulk mineral and agricultural products, but increasingly both raw and processed agricultural products and the residual manufactured goods are shipped in containers. Mineral exports are generally handled by purpose-built ports and railways, whereas container traffic is handled at ports associated with the state capitals, particularly Brisbane, Sydney and Melbourne. Container traffic at Australian ports is predicted to jump from less than 5 million TEUs in 2001 to 11 million TEUs in 2025.
The state capitals grew up in proximity to navigable water, but today the cities sprawl far from the original port areas. Sydney has a current urban population of 4·2 million, but it is forecast to reach 5·3 million by 2031. The Sydney Statistical District forms the major part of a conurbation that stretches as far as Newcastle 160 km to the north, Kiama 100 km to the south, Katoomba 100 km to the west and Macarthur 60 km to the southwest.
Container traffic through Sydney is forecast to grow from 1·6 million TEUs in the year to June 2007 to 3·2 million TEUs by 2025. Traffic in the eight months to February 2008 was up 8% from the same period the previous year.
Such growth projections have serious implications for network capacity around the city. In fiscal year 2007, 300 000 TEUs were moved by rail, and the state government has set a target of 40% of incoming container traffic to be moved from the port by rail by 2011. This would require a quad-rupling of the number of boxes moved by rail. However, 90% of the containers arriving at the Port of Sydney are for local delivery, with over 85% for destinations within a 40 km radius.
A greater market share is possible if rail can rise to the occasion and infrastructure constraints are not a limiting factor; the already congested road system would be stretched to provide this capacity, and plans under discussion will result in significant increases in traffic, pollution and the urban 'footprint' devoted to road. The cost of a projected road tunnel as a partial solution will far outweigh any expenditure on rail projects associated with container traffic to and from the ports.
The consequences for rail
Rail can accommodate the task mainly on the existing local network, with encroachment only needed in a few locations where it is necessary to create space for additional tracks and grade-separated junctions.
This predominantly short-distance traffic from port to land terminal requires a different type of operation to the long haul container trains, in some cases double-stacked, which are used on interstate and regional services.
The biggest difference will be train length. Sydney basin short-haul container trains will be 600 m, whereas the long-distance container trains can vary from 1 500 m to 1 800 m depending on the sector. The shorter lengths are dictated by the portside loading and unloading processes, the desire to limit the amount of disturbance to the train consist during the transfer processes to ship and road transport and the optimum train path impact on suburban passenger operations.
Container trains will travel to some destinations accessed by freight-only lines, but several of the new and some of the existing terminals can only be reached by using passenger tracks.
Sydney's 1·5 kV DC suburban network is used by more than 2 300 daily services carrying 900 000 passengers between 302 stations. These services are politically sensitive, with commuter belt voters traditionally deciding which party controls the state legislature.
Under current arrangements, a curfew on freight access to the Sydney suburban network operates in peak hours, and only limited paths are available for container movements during daylight. However, the mandated 40% growth cannot be satisfied if operations are limited to night hours.
Plans to accommodate anticipated growth in passenger traffic call for a 4% increase in capacity using existing train paths by lengthening all trains to eight cars. A further 8% capacity increase is envisaged by the provision of additional services when the 'Clearways' project to recast the CityRail timetable is completed in 2012 (RG 5.04 p253).
Freight routes to the hinterland interface with the electrified system at the extremities of the network, and then share to a large extent the same tracks into the city. All the major freight terminals are accessed via flat junctions with low-speed turnouts, and a freight train entering or leaving a yard often blocks both tracks of the suburban line. Low-speed junctions are also found at the connections between the dedicated freight lines and the mixed traffic suburban lines at Lidcombe, Sefton Park and Meeks Road (Fig 1).
Transit times and the lengths of the freight trains mean that junctions can be blocked for a considerable time.
A team comprising port and landside container terminal operators, train operators and consultants from Plateway Pty Ltd has studied the ability of the network to handle the anti-cipated increases in container transport, allowing for enhancements, including the Southern Sydney Freight Line and the Clearways projects.
The studies considered the total processes involved, from ship unloading to the dispatch of containers by road from terminals distributed throughout Sydney. They showed that some off-peak daytime movements will be essential if these targets are to be met. The studies also show that with further infrastructure enhancements the target can easily be exceeded and a greater share of container transport taken by rail.
As a result, operations have been revamped with altered terminal processes and reductions in terminal shunting resulting in tighter turnarounds and increased reliability. The team considered not only the growth in container traffic through to 2017, but also the growth in passenger and other freight traffic.
The studies used the OpenTrack network simulator and Viriato timetable modelling programme (RG 3.08 p183). Once the market demand was established, Viriato was used to develop timetables delivering desired traffic volumes to the targeted destinations. These concept timetables were then proven using OpenTrack. OpenTrack is well suited to mixed traffic environments, in particular modelling train and network behaviour at complex junctions and the freight/passenger interfaces which call for more than a 'headway' solution. The tool has also been used for modelling the rail aspects of terminal operations, including all shunting and light engine moves.
The network simulator enables the operators to test the effectiveness of proposed infrastructure changes and ensure that investment yields the anticipated gains. Comparative performance encompassing timetable robustness, reliability and the behaviour of the network when random incidents are introduced, can all be observed.
Fig 2 shows the current movement of commodities in the Sydney area. Two key issues are the inability to refuge long freight trains on the outskirts of the city, and the cross-city nature of many of the freight moves, requiring multiple interfaces with passenger traffic.
Southern Sydney boost
For freight train operations, the most important infrastructure improvements will result from the opening of the Southern Sydney Freight Line which will provide curfew-free access to the south via a dedicated freight line extending from the southern extremity of the Sydney electrified area at Macarthur to the existing boundary of the Metropolitan freight network at Sefton Park. Fig 1 shows the extent of the SSFL and the freight terminals at Leightonfield and a new terminal to be constructed at Moorebank.
The SSFL, which is scheduled for completion by 2010, will be grade-separated from the electrified lines, providing connections from Macarthur on the Main South to the Old Main South Line as well as the Metropolitan Freight Lines. The Clearways projects at Lidcombe, Homebush and Hornsby, together with the new Epping - Chatswood passenger line which is due to open by the end of this year, have been designed to facilitate improvements in passenger traffic. They will also have a positive impact on freight traffic performance by reducing the number of flat junction conflicts. Service restructuring could also provide paths for additional traffic. Robust discussion will determine how many of these paths will be available for freight.
Current CityRail passenger services include a circular route from Central via the City Circle to Bankstown. From Bankstown, the service continues to Lidcombe, entering Lidcombe station via the Main Southern Line (Fig 3, DMS and UMS), joining the Suburban group of lines to Homebush (DS and US). At Homebush the service diverts to the Local lines at a flat junction (DL and UL) before continuing to complete the circuit at Central. This service also operates in the reverse direction. Rearranged terminal platforms at Lidcombe and Homebush would enable terminating trains to turn back without blocking the running lines (Fig 3).
These changes will release four paths/hour for trains on the Western Suburban lines thus providing capacity on the Western Main lines (DM and UM) for additional passenger traffic in the peaks and for freight traffic entering from the freight lines in the off peak period.
The Epping - Chatswood line will provide grade-separated junctions with the Main North and North Shore lines, which will remove the current necessity for services using the North Shore line and proceeding beyond Hornsby to use flat crossings.
Bi-directional signalling on the Carlingford branch and reconfiguration of the entry points to Clyde yard will allow for a doubling of frequency of passenger trains and clear access for freight trains to move from the Sandown container and fuel terminal into the yard without conflicting with suburban services.
Despite these improvements, which are in the process of being implemented, numerous other interfaces between freight and passenger trains will remain low-speed, flat junctions.
Preliminary plans have been announced to create a grade-separated junction for freight trains in the vicinity of North Strathfield, and a grade-separated junction at Flemington West Junction in Lidcombe is high on the wish list. Other critical issues that must be addressed include additional refuge loops for long trains on the approaches to the city and on the suburban network.
Long and short trains
The Freight Infrastructure Advisory Board study released in July 2005 recommended that the length of a container train for the port/land terminal link should be 600 m. This was chosen as it matches the length of the siding at the port used during the train stripping and loading process and it minimises the impact of the freight train operations on suburban network. However, these local container trains are significantly shorter than the interstate and long-distance intrastate services.
What is required therefore is the ability to amalgamate the shorter rakes into longer trains for interstate operation. Currently this means two rakes for 1 200 m or three rakes for 1 800 m trains. Few terminals can accommodate such long trains, but the design of new terminals will provide the ability to make them up and break them down.
With a purpose-designed land terminal, containers can be optimally loaded. Currently, trains arriving at the port have containers destined for different ship-loading terminals and different ships. This necessitates shunting at Botany yard, which is already very congested. Future arrangements mean that the short haul rakes should arrive at the port already sorted for their destination, and Botany will become a staging-yard acting as a buffer for the ship-loading sidings.
In addition to the work done as part of the port freight study team, Plateway has also considered the impact of further innovation with short distance freight trains.
The short haul container trains from the ports should ideally be capable of matching the braking and acceleration performance of the suburban EMUs that run on the same track. If this was the case, then signalling block distances and operational headways, as distinct from theoretical headways, could be reduced, releasing additional capacity.
In general, Australian freight trains still use the automatic air brake as standard. Work by the US Federal Railroad Administration has shown that electro-pneumatic braking can provide substantive benefits. The general application of ECP braking has until recently been limited by cost, but as more US railroads adopt the technology it can be expected that costs will fall. A semi-permanently coupled 600 m container rake with higher running speeds and better braking characteristics could help to improve path utilisation.
A standard train formation would also permit easier automation of the loading and unloading process. This would result in shorter turnround times and thus greater utilisation of the train in revenue service. Automatic and/or high intensity stripping at the terminals will result in pressure to restrict the movement of unprotected personnel on the ground in the loading area. This will call for innovative procedures for coupling and uncoupling locomotives and rolling stock, or alternatively the development of fixed rakes with distributed power.
A focus in the USA on the development of modern, medium-powered locomotives using multiple generator sets could result in a change away from using cascaded older locomotives to purpose-specific units that are cost-effective for short-haul trains.
Connections to and from the mixed traffic network remain a problem. Conflicts at the critical flat junctions impact on the robustness of the current freight and passenger timetables and limit the ability of freight traffic to utilise spare main line capacity in the contra-peaks. Grade-separation would provide benefits for both passenger and freight services and enables increased utilisation of the existing suburban corridors.
Surprisingly, given the New South Wales state government's intention to shift a minimum of 40% of the containers from the port by rail, some recent enhancement projects seem only to have considered the impact on freight as a by-product and not as part of the prime focus.
A happy co-existence
Freight and suburban traffic can and must coexist in urban environments. The social and economic cost of separate infrastructure is prohibitive, except in cases of intense corridor utilisation, such as in city centres.
The use of simulation tools with timetable and planning software enables existing services, future services and planned infrastructure to be modelled. Adjustments can be made to ensure stable and robust service delivery for both passenger and freight traffic.
Urban railways designed for peak services normally have surplus track capacity available for freight during contra-peak and off peak periods. Our studies have shown that it is possible for freight and passenger to co-exist, and the first step towards a harmonious relationship comes in recognising that freight rail is only a component of a logistics chain.
Market studies and the associated service planning must consider the holistic picture. Access providers must consider the impact of their infrastructure configuration at key interface points. For maximum utilisation of the network, access charges should encourage the use of modern freight rolling stock with braking performance characteristics that match that of passenger applications.
- CAPTION: A Patrick Portlink service from Sandown to Port Botany joins the Main Western Line at Clyde from the Carlingford branch on a 40 km/h speed limit.
- CAPTION: Flemington West Goods junction at Lidombe is high on the wish-list for grade-separation.
- Fig 1. The rail network in the Sydney basin showing projects under construction and planned, including the Southern Sydney Freight Line.
- Fig 2. Freight traffic flow by commodity group through the Sydney basin.
- Fig 3. Current track arrangement between Homebush and Lidcombe showing the interface between metropolitan freight lines and suburban lines.