1. An inland rail route between Melbourne and Brisbane? The North-South rail corridor study 14 June 2007 BTRE Transport Colloquium

2. Slide 2 Structure of presentation Introduction Total freight market Mode choice modelling Access prices Passenger market modelling Project outcomes Areas of future research

3. Slide 3 The task Background: freight growth, Auslink upgrades, lobbying for inland rail route To consider options for the Melbourne - Brisbane rail corridor over the next 25 years Consortium: Ernst and Young (project leader, financial anaylsis) Hyder (route options, other infrastructure, environmental) ACIL Tasman (demand, access prices)

4. Slide 4 Analytical approach Identify current total freight market Forecast growth in total freight market for 25 years Determine current mode shares Estimate mode shares over 25 years and their sensitivity to changing service quality Estimate rail freight over 25 years Integrate with other models in consortium

5. Slide 5 Path of far western route

6. Slide 6

7. Slide 7 Freight movements (excl. coal)

8. Slide 8 Analysis of base (2004) Origin - destination tonnages by commodity Data from rail operators, BTRE, ABS, FDF FreightSim for forecasting model. Structure: production, imports, consumption ACIL Tasman model of freight inducement effects

9. Slide 9 Drivers of future demand GDP growth (production, imports, consumption) Transport freight to GDP growth ratio Growth of industrial concentration Growth of agricultural production Growth of imports, service sector Scenarios – High/medium/low GDP – Growth transport/GDP ratio

10. Slide 10 New East Coast demand with an inland route Coal in southern QLD, northern NSW But questions about which route, which port Little else Mainly a through route with a stop at Parkes

11. Slide 11 Freight diversion – northern NSW 27% of grain from Northern Plains to Brisbane from Newcastle 50% of cotton from Northern Plains to Brisbane from Port Botany Brisbane-Perth freight via Parkes No other material freight diversion

12. Slide 12 Future demand modelling results

13. Slide 13 Road, rail, sea or air ? Current market shares by mode Drivers of mode choice Sydney Convenient departures Logit model Results

14. Slide 14 2004 snapshot Current Rail mode shares Melbourne-Brisbane ~30% Melbourne-Sydney ~7% Sydney-Brisbane ~11% Rail more price competitive on longer routes, less on shorter routes because of PUD time and costs Rail outperformed by road in service quality

15. Slide 15 Mode performance 2004 – M to B RoadSea Rail linehaul Rail door to door Relative price 0-20% above d-d rail 20-40% below d-d rail 30-45% below road 0-20% below road Reliability95%90%35-45% Availability99%10%40-45% Transit time21-27 hrs3-3.5 days36 hrs42 hrs

16. Slide 16 Explanatory variables Price $/tonne (but complexities) Reliability within 15 minutes of scheduled arrival time Availability % of times the freight carrier is available within an hour of customers’ preferred time Survey: how does demand for rail (at the expense of road) vary with changes in price, reliability, availability?

17. Slide 17 Reliability - problems in Sydney

18. Slide 18 Transit time & availability

19. Slide 19 Survey results – all customers

20. Slide 20 Survey results - manufactured

21. Slide 21 The manufactured market Melbourne- Brisbane Melbourne- Sydney Sydney- Brisbane Express freight5% Freight sensitive to reliability and availability 60%70% Price-sensitive freight 35%25%

22. Slide 22 Mode choice Logit model to predict modal shares Calibrated to explain current shares Forecast changes based on expected route characteristics – Price (incl fuel price, driver shortage) – Reliability following AusLink upgrade – Availability following AusLink upgrade Key parameters estimated from Surveys Econometric analysis of past data

23. Slide 23 Logit model Gives probability (%) of freight forwarder choosing mode n Simple logit (two modes) formula Where U is utility of using rail or road linear utility functions: constant + variable 1 x coefficient1 + etc

24. Slide 24 Logit model – hierarchical structure Logit model can be used at each level of the freight decision making process Nested logit useful for inland rail analysis

25. Slide 25 Values/inputs to use in the model Current estimates of road and rail performance and mode share were derived from Rail operators, BTRE, ARTC, surveys Future estimates of road and rail performance were derived from ARTC, BTRE, ACIL Tasman, Hyder, freight operators Standalone road and rail freight pricing model developed Access prices, fuel costs, labour

26. Slide 26 Scenarios Case A (reference), Case B (high rail), Case C (low rail) Fuel price assumptions Road and rail labour cost assumptions

27. Slide 27 Elasticities and coefficients Choice elasticities were estimated From these – coefficients were calculated PriceReliabilityAvailability Melbourne- Brisbane-0.5-1.20.4-0.70.5-0.8 Melbourne-Sydney-0.7-1.150.3-0.70.08-0.36 Sydney-Brisbane-0.3-0.90.3-0.70.3-0.8 All routes-0.3-1.20.3-0.70.08-0.8

28. Slide 28 Testing of the logit model Price (ave charge $/tonne) Reliability Service availability Observed mode share Road18695%99%30% Rail11366%80%70% Observed performance (Melb-Perth 2000) Logit modelling of same input variables Price (ave charge $/tonne) Reliability Service availability Observed mode share Road18695%99%27% Rail11366%80%73%

29. Slide 29 Results – rail market share

30. Slide 30 Revenue maximising access charges

31. Slide 31 Financial and economic results Summary of results Interpretation of results Coastal route: problems and solutions Inland route: problems and solutions Full report including Ernst & Young and Hyder chapters not covered here, on www.aciltasman.com.au www.aciltasman.com.au