PPT - Bureau of Infrastructure, Transport and Regional Economics
Download
Report
Transcript PPT - Bureau of Infrastructure, Transport and Regional Economics
An inland rail route between
Melbourne and Brisbane?
The North-South rail corridor study
BTRE Transport Colloquium
14 June 2007
Structure of presentation
Introduction
Total freight market
Mode choice modelling
Access prices
Passenger market modelling
Project outcomes
Areas of future research
Slide 2
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)
Slide 3
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
Slide 4
Path of far western route
Slide 5
Slide 6
Freight movements (excl. coal)
5.4m
5.1m
0.8m
19.3m
9.2m
7.0m
4.5m
1.2m
9.0m
2.9m
5.3m
To/from WA & SA
into NS Corridor
4m
10.3m
Slide 7
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
Slide 8
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
Slide 9
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
Slide 10
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
Slide 11
Future demand modelling results
30,000
25,000
Ktonnes
20,000
15,000
10,000
5,000
2004
2009
Melbourne-Brisbane (case A)
Melbourne-Sydney (case A)
Sydney-Brisbane (case A)
2014
2019
Melbourne-Brisbane (case B)
Melbourne-Sydney (case B)
Sydney-Brisbane (case B)
2024
Melbourne-Brisbane (case C)
Melbourne-Sydney (case C)
Sydney-Brisbane (case C)
2029
Slide 12
Road, rail, sea or air ?
Current market shares by mode
Drivers of mode choice
Sydney
Convenient departures
Logit model
Results
Slide 13
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
Slide 14
Mode performance 2004 – M to B
Road
Relative
price
Sea
0-20%
20-40%
above d-d below d-d
rail
rail
Rail
linehaul
Rail door
to door
30-45%
below
road
0-20%
below
road
Reliability
95%
90%
35-45%
35-45%
Availability
99%
10%
40-45%
40-45%
3-3.5 days
36 hrs
42 hrs
Transit time 21-27 hrs
Slide 15
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?
Slide 16
Reliability - problems in Sydney
Brisbane
ARTC
Network
(exception of
north of the
Queensland
boarder)
Trains traverse
the ARTC &
Railcorp
Network from
MelbourneBrisbane
(arrow does
not follow
actual train
path)
40% on-time reliability
( average for both
directions)
80% on-time
reliability
Loss of 50% reliability
in Northern Sydney
30% on-time
reliability
40% on-time
reliability
Sydney
RailCorp Network
60% on-time
reliability
Loss of 20% reliability in
South Western Sydney
ARTC
Network
Melbourne
Slide 17
Transit time & availability
% of Total departures per hour
8%
7%
6%
Current
departure
time from
customer
5%
4%
2009
departure
time from
customer
3%
2%
1%
0%
1
2
3
4
5
6
7
8
9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Time of day
24 hour distribution of truck departures
Slide 18
Survey results – all customers
6%
7%
10%
41%
15%
21%
Price
Reliability
Availability
Transit time
Flexibility
Loss/damage
Slide 19
Survey results - manufactured
23%
51%
26%
Price
Reliability
Availability
Slide 20
The manufactured market
MelbourneBrisbane
MelbourneSydney
SydneyBrisbane
Express freight
5%
5%
5%
Freight sensitive
to reliability and
availability
60%
70%
70%
Price-sensitive
freight
35%
25%
25%
Slide 21
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
Slide 22
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 + variable1 x coefficient1 + etc
Slide 23
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
Slide 24
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
Slide 25
Scenarios
Case A (reference), Case B (high rail), Case C
(low rail)
Fuel price assumptions
Road and rail labour cost assumptions
Slide 26
Elasticities and coefficients
•
•
Choice elasticities were estimated
Price
Reliability
Availability
Melbourne- Brisbane
-0.5-1.2
0.4-0.7
0.5-0.8
Melbourne-Sydney
-0.7-1.15
0.3-0.7
0.08-0.36
Sydney-Brisbane
-0.3-0.9
0.3-0.7
0.3-0.8
All routes
-0.3-1.2
0.3-0.7
0.08-0.8
From these – coefficients were calculated
Slide 27
Testing of the logit model
• Observed performance (Melb-Perth 2000)
Price (ave
charge $/tonne)
Reliability
Service
availability
Observed
mode share
Road
186
95%
99%
30%
Rail
113
66%
80%
70%
• Logit modelling of same input variables
Price (ave
charge $/tonne)
Reliability
Service
availability
Observed
mode share
Road
186
95%
99%
27%
Rail
113
66%
80%
73%
Slide 28
Results – rail market share
Slide 29
Revenue maximising access charges
Revenue ($ million)
60
50
40
30
20
10
0
$0.00
$1.26
$2.53
$3.79
$5.05
$6.31
$7.58
$8.84
$10.10
$11.36
$12.63
Access charge ('000gtk)
0% Efficiency gains
15% Efficiency gains
5% Efficiency gains
20% Efficiency gains
10% Efficiency gains
Slide 30
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
Slide 31